Professional Documents
Culture Documents
insftjuiveo? ! « e * í S « S ( á W ^ O R d a z o * IÄW j
te>î!lî$ètor Uï ii i s $ Ü i3 ¡'íL íS S ííS tf i
■■•; s . ^ O u ^ S B R G l i i ! » !
>3-3’ m r
DEVELOPMENTS IN HYDROBIOLOGY
edited by
Victor Jaccarini and Els Martens
izwo
Instituut voor Zeewetenschappelijk Onderzoek ( vzw)
Institute for Marine Scientific Research
VICTORIALAAN 3-B -8400 OOSTENDE BELGIUM
Tel.+32-(0) 59-321045— Fax:+32-(0) 59-321135
Developments in Hydrobiology 80
Series editor
H. J. Dumont
The Ecology of Mangrove and
Related Ecosystems
Proceedings o f the International Symposium held at M ombasa, Kenya,
2 4 —3 0 Septem ber 1 9 9 0
Edited b y
V icto r Jaccarini & Els Martens
W
Kluwer Academic Publishers
Dordrecht / Boston / London
Library of Congress Cataloging-in-Publication Data
The E c o l o g y o f ma ng ro v e and r e l a t e d e c o s y s t e m s : p r o c e e d i n g s o f t h e
i n t e r n a t i o n a l s y m p o s i u m h e l d a t Mo mbas a, K e n y a , S e p t e m b e r 2 4 - 3 0 ,
1990 / e d i t e d by V i c t o r J a c c a r i n i & E l s M a r t e n s .
p. cm. — ( D e v e l o p m e n t s i n h y d r o b i o l o g y ; 8 0 )
" R e p r i n t e d f r o m H y d r o b i o 1o g i a v o l . 2 4 7 ( 1 9 9 2 ) . "
Includes indexes.
I S BN 0 - 7 9 2 3 - 2 0 4 9 - 2 ( h b : a c i d f r e e p a p e r )
1. M a n g r o v e s wa mp e c o l o g y — C o n g r e s s e s . 2. Mangrove f a u n a -
-Congresses. 3. Mangrove s w a m p s - - C o n g r e s s e s . I. Ja c c a rin i,
V ictor. II. M artens, Els. III. Series.
Q H 5 4 1 . 5 . M2 7 E 2 5 1992
5 7 4 . 5 ‘2 6 3 2 5 — dc 20 92-38659
ISBN 0-7923-2049-2
The publication of this volume was funded by the Commission of the European Communities,
and sponsored by the Hydrobiological Society of E astern Africa (HYSEA).
Printed in Belgium
V
Contents
P re f a c e ......................................................................................................................................................................... vii
P i c t u r e ......................................................................................................................................................................... ix
S T A T U S , P R IM A R Y P R O D U C T IV IT Y A N D T A X O N O M Y O F M A N G R O V E S,
SEA G RA SSES A N D ALGA E
Developing m anagem ent plans for the m angrove forest reserves o f m ainland T anzania
by A. K.Semesi ............................................................................................................................................. 1
S tructure and species distribution in C oringa m angrove forest, G odavari D elta, A ndhra Pradesh,
India
by J. A zariah, H. A zariah, S. G unasekaran & V. Selvam ........................................................... 11
S tructural properties o f tw o types o f m angrove stands on the northw estern coast o f Sri Lanka
by M. D. A m arasinghe & S. B alasubram aniam ................................................................................ 17
The current status o f m angroves along the K enyan coast: a case study o f M ida Creek m angroves
based on rem ote sensing
by P. O. G ang & J. L. A g a ts iv a ............................................................................................................. 29
N et prim ary productivity o f tw o m angrove forest stands on the northw estern coast o f Sri L anka
by M. D. A m arasinghe & S. B alasubram aniam ................................................................................ 37
The effect o f coastline changes on m angrove com m unity structure, Portuguese Island, M ozam bique
by J. C. H atto n & A. L. C o u t o ................................................................................................................ 49
The seagrass and associated m acroalgal vegetation o f G azi Bay (K enya)
by E. C oppejans, H. Beeckm an & M. D e W i t .................................................................................. 59
M icrophytoplankton o f the P itchavaram m angals, southeast coast o f India: species com position
and population density
by L. K an n an & K. V asantha ................................................................................................................ 77
USERS O F TH E M ANGALS
The ecology o f Belizean m angrove-root fouling com m unities: patterns o f epibiont distribution and
abundance, and effects on root growth
by A. M. Ellison & E. J. F a r n s w o r th .................................................................................................... 87
The m eiobenthos o f five m angrove vegetation types in G azi Bay, K enya
by S. Vanhove, M. Vincx, D. Van G ansbeke, W. Gijselinck & D. S c h r a m ............................. 99
Ecom orpholy o f Crassostrea cucullata (Born, 1778) (O streidae) in a m angrove creek (G azi, Kenya)
by J. F. T ack, E. V anden Berghe & Ph. P o l k .................................................................................... 109
A note on the Zooplankton distribution and diversity in a tropical m angrove creek system, G azi,
K enya
by M . K. W. O sore .................................................................................................................................... 119
vi
C H E M IC A L A N D P H Y S IC A L P R O C E S S E S
H ydrodynam ics o f m angrove sw am ps and their coastal w aters
by E. W o la n s k i............................................................................................................................................. 141
Intratidal fluctuations in stratification within a m angrove estuary
by R. J. U ncles, W .-K. G ong & J.-E . O n g ......................................................................................... 163
Physical processes in the Rufiji delta and their possible im plications on the m angrove ecosystem
by J. F r a n c i s .................................................................................................................................................. 173
D iurnal variation in physical-chem ical properties and prim ary production in the interconnected
m arine, m angrove and freshw ater biotopes o f K akinada coast, A n d h ra P radesh, India
by V. Selvam, J. A zariah & H. A zariah .............................................................................................. 181
M ethanogenic bacteria in m angrove sedim ents
by R. M ohanraju & R. N atarajan ......................................................................................................... 187
F IS H E R Y B IO L O G Y A N D M A R IC U L T U R E
M angroves as a habitat for fish and praw ns
by A. Sasekum ar, V. C. Chong, M. U. Leh & D ’C ruz .................................................................... 195
M ariculture developm ent in Kenya: alternatives to siting ponds in the m angrove ecosystem
by J. R asow o ................................................................................................................................................ 209
T he effect o f oocytic atresia on fecundity estim ates o f the rabbit fish Siganus sutor (Pisces:
Siganidae) o f K enyan m arine inshore w aters
by M. J. N tiba & V. Jaccarini ................................................................................................................ 215
Differences in biom ass production and carrageenan yields am ong four strains o f farm ed carrageeno-
phytes in N orthern Bohol, Philippines
by G . C. T rono & A. O. L lu is m a ........................................................................................................... 223
E X P L O IT A T IO N , C O N S E R V A T IO N A N D M A N A G E M E N T A S P E C T S
Status and exploitation on the m angrove and associated fishery resources in Z an zib ar
by M. A. K. Ngoile & J. P. Shunula ..................................................................................................... 229
The m anagem ent o f Insular C aribbean m angroves in relation to site location and com m unity type
by P. R. B acon & G . P. Alleng .............................................................................................................. 235
Regeneration strategy o f m angroves along the Kenyan coast: a first approach
by D. Van S p e y b ro e c k ................................................................................................................................ 243
Im pact o f p ast m anagem ent practices on the present status o f the M uthupet m angrove ecosystem
by J. A zariah, V. Selvam & S. G unasekaran ..................................................................................... 253
Preface
T he international sym posium on The Ecology o f Mangrove and Related Ecosystems, M om basa, Kenya,
24-30 Septem ber, 1990, w as organized by the D epartm ent o f Zoology, University o f Nairobi, in col
laboration with the K enya M arine and Fisheries R esearch Institute, and the Regional D ispatch Centre
o f the Regional C ooperation in Scientific Inform ation Exchange in the W est Indian Ocean, M om basa,
Kenya. It was officially opened by Professor F. N. O nyango, Principal, College o f Biological and Physical
Sciences, U niversity o f N airobi. The closing address was given by Professor M. Hyder, retired Principal
and Professor o f Zoology, C B PS, University o f Nairobi.
Scientific investigations o f the m angrove and coral reef ecosystem s in K enya have been going on at
least since the eartly 70’s and have been gathering m om entum with the involvem ent o f the University of
N airobi, K M F R I, and British, Belgian, D utch and Italian universities and scientific institutions. So the
tim e w as ripe to take stock o f the situation and to try to bring together w orkers in the same field from
as m any tropical regions as possible. A cursory glance at the list o f participants is sufficient proof o f the
w idespread interest in the them e o f the sym posium . C ontributions cam e from all the main tropical m arine
areas o f the globe, representing tw enty-three countries, in W est and E astern Africa, South and Southeast
A sia, C entral and South A m erica, and A ustralia. The thirty papers published in this volume were selected
from seventy presented at the symposium.
It is recognized th at the m angroves have intim ate and complex linkages with other coastal systems,
notably seagrass beds and coral reefs. So, starting from the m angrove forest and its users at the interface
betw een the terrestrial, freshw ater and m arine environm ents, the papers presented proceed to the fully
subm erged seagrasses and further out to the corals. The nature o f the coupling between the different
systems w as touched on at the symposium, but stands out as a serious gap in our knowledge. Next,
chem ical and physical processes which constitute the m echanism s o f linkage received attention. H um an
uses and abuses o f the ecosystem s form ed the last m ajor topic o f the symposium.
W orkshop discussions were held during the sym posium in preparation for the adoption in plenary
session o f two declarations: A Statem ent o f Priorities fo r Scientific Research and A Declaration on
Management Strategies.
The following priorities for basic research were agreed : the interlinkages between the mangrove, coral
reef and seagrass ecosystem s; the need for m ore inform ation on nearshore hydrological processes
operating at the small and m eso-scale and elaboration of predictive m odels ; m ore w ork on nutrient cycling
and energy flow including soil processes; biological processes and use o f bioindicators.
Priorities were also set for applied research. This, it w as agreed, should be directed tow ards m anage
m ent o f the forest or aquatic product on a sustainable basis, taking into account the integrated nature
o f the ecosystem s. Em phasis should be placed on minimizing perturbations to the systems and the
m aintenance of ecological integrity. B road areas needing investigation were identified: m ethods for
efficient and ecologically sound land use in the w atershed adjacent to the m angal, seagrass beds and coral
reefs; effect o f m anagem ent o f the m angroves on off-shore fisheries; im pact of industralization on the
m angal and related ecosystem s; effect o f dom estic and industrial effluents and m anagem ent o f their
im pact; com parative studies on destructive and non-destructive aquacultural techniques with a view to
m inimizing negative effects; the effect of hum an activities on physical, chemical and biological soil
processes; research on m anagem ent practices appropriate to different m angrove forest types and
prom otion o f a biotechnological aproach in m angrove silviculture; socio-econom ic studies o f hum an
viii
coastal settlem ents; natural products o f the m angals; the im pact o f global clim ate change; and active
rehabilitation o f seagrass beds and coral reefs.
The Declaration on M anagement Strategies was positive, concluding th at m uch can be done to stop the
degradation o f dam aged m angals, seagrass beds and coral reefs and to rehabilitate them , but th at new
techniques m ust be w orked out to allow use on a sustainable basis for long term econom ic return and
for the wellbeing o f coastal hum an settlem ents and a healthy environm ent.
To achieve the above goals the Symposium recom m ended the following strategies.
T he establishm ent o f National Coastal Management Plans including the following com ponents : survey
o f both natural and hum an resources and their present uses ; identification o f the problem s affecting coastal
ecosystem s; adjustm ent o f legal fram ew orks with regard to the uses o f these ecosystem s and enforcem ent
o f existing m anagem ent regulations; provision o f technical assistance and training for the introduction
o f econom ically and ecologically viable activities in these ecosystem s; rehabilitation o f dam aged areas;
creation o f natural reserves both for tourism and for the recuperation o f natural plant and anim al
populations; education and creation o f public aw areness on the value and problem s o f these ecosystem s.
In form ulating and im plem enting each national P lan all concerned parties should be involved, i.e., the
inhabitants and the outside users o f these ecosystem s, the policy m akers, the scientists and technical
experts, and also the national and international funding agencies. U se should be m ade o f existing
institutions and netw orks. The scientists gathered in the sym posium urged the establishm ent or strength
ening o f regional and international associations o f scientists to provide the scientific basis to national and
regional m anagem ent plans.
It is our pleasant duty as C hairm an and Secretary o f the Symposium Organizing C om m ittee to thank
the m any sponsors who generously provided financial support for the Sym posium : the African
Biosciences N etw ork for providing the seed funds; the U S N ational R esearch C ouncil; the C om m ission
o f the E uropean C om m unities ; the A ustralian International D evelopm ent A ssistance B ureau; the N ether
lands M inistry o f Agriculture, N ature M anagem ent and Fisheries; U nited N ations Environm ent P ro
gramm e, Forestry D epartm ent; U N E S C O - R O S T A ; and the T hird W orld A cadem y o f Sciences.
Special thanks are due to the Com m ission o f the E uropean C om m unities for providing funds covering
the publication costs o f this volume. A debt o f gratitude is owed to Hydrobiologia (K luw er A cadem ic
Publishers) and the C hief Editor, Professor H. D um ont o f U niversity o f G h en t (Belgium), for agreeing
to publish the Proceedings as a special volume in their series D evelopm ents in Hydrobiology. W e would
also like to thank the m em bers o f the Executive Com m ittee for their help in the organization o f the
Symposium . Finally, we wish to thank the University o f N airobi in the persons o f its previous Vice-chan
cellor, Professor P. M. M bithi, Professor F. N. O nyango, Principal, College o f Biological and Physical
Sciences, and Professor R. W. M wangi, D ean, Faculty o f Science, for their support. T hanks should also
go to Professor G . M. O. M aloiy who negotiated the initial funding with the A frican Biosciences N etw ork.
V. J a c c a r in i & E. M a rte n s
1. D. Onyango 15. N. 0 . Isirimah 29. A. M. Ellison 43. W. M. Kudoja 57. M. A. Ngoile
2. J. Kasyi 16. M. Hemminga 30. A. Robertson 44. J. G. Mutwiri 58. M. M astaller
3. C. Amadou 17. P. O. Gang 31. A. H. Moallim 45. A. Yanez-Arancibia 59. N. Revisa
4. J. Francis 18. M. A. Amakiri 32. F. van Averm aat 46. M. D. Fortes 60. E. Okemwa
5. J. K athupotha 19. M. D. Amarasinghe 33. V. Selvam 47. L. K annan 61. E. M artens
6. G. C. Trono 20. R. W. Mwangi 34. A. G. Tejan 48. E. Coppejans 62. K. Delbeke
7. V. Jaccarini 21. M. Vannini 35. A. Sasekumar 49. P. Akiwumi 63. R. M ohanraju
8. C. A. M uhando 22. S. G unasekaran 36. J. K. Muraya 50. S. Abdulrahm an Issa 64. M. Vincx
9. A. Thorhaug 23. J. P. Shunula 37, U. Helbing' 51. G. van der Velde 65. T. Done
10. A.-F. Woitchik 24. M. Vuai 38. K. N askar 52. G. W amukoya 66. D. van Gansbeke
11. J. L. Agatsiva 25. K. Deges 39. L. Vergeer 53. E. Wolanski 67. J. Schrijvers
12. L. Solorzano 26. J. D. Woodley 40. D. Prasad 54. J. H atton 68. M. J. N tiba
13. J. Azariah 27. B. Musyoki 41. J. Rasowo 55. F. W. Fong
14. E. Slim 28. N. Visser 42. J. M utere 56. F. M. Chale
Hydrobiologia 247: x - x i i, 1992.
X V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems.
List of participants
A bdulrahm an i s s a , S., Commission for Lands & Environ D ’C r u z , R., D epartm ent of Zoology, University of M alaya,
m ent, P.O. Box 811, Z anzibar, Tanzania 59100 K uala Lumpur, M alaysia
A g a t s i v a , J. L., Remote Sensing, D epartm ent o f Resource D a r o , M. H., Ecology & Systematics Laboratory, Free
Surveys and Remote Sensing, P.O. Box 47146, Nairobi, U niversity of Brussels, Pleinlaan 251, B-1050 Brussels,
Kenya Belgium
A i k e n , K., D epartm ent of Biological Sciences, Florida Inter D e g e o r g e s , A., Regional Environm ent Advisor, U.S. -
national University, Tamiami Campus, Miami, A ID , P.O. Box 30261, Nairobi, Kenya
FL 33199, USA D e g e s , K., State University of G hent, Institute of Zoology
A k i w u m i , P. K., O ceans & C oastal Areas Programme, Ac - M arine Biology, K. L. Ledeganckstraat 35, B-9000
tivity Centre (OCA/PAC) U nited N ations Environment G hent, Belgium
Programme, P.O. Box 30552, Nairobi, Kenya D e l b e k e , K., Kenya-Belgium Project M arine Science
Kenya M arine & Fisheries Research Inst., P.O. Box
A l l e n g , G., D epartm ent of Zoology University of the W est
81651, M om basa, Kenya
Indies, M ona, Kingston 7, Jam aica
D e n H a r t o g , C ., L aboratory of A quatic Ecology, Catholic
A m a d o u , C., U N D P/FA O Fuelwood Project c/o U N D P,
University of Nijmegen, Toernooiveld, 6525 ED Nijme
P.O. Box 1011, Freetown, Sierra Leone
gen, The N etherlands
A m a k i r i , M. A., Rivers State University Science & Techn.
D o n e , T., A ustralian Institute of M arine Science A IM S,
Faculty of Agriculture, P.M.B. 5080, Port H arcourt, PMB 3 Townsville MC, Q ueensland 4810, A ustralia
Nigeria E l l i s o n , A. M., D epartm ent of Biological Sciences, M ount
A m a r a s i n g h e , M. D., National Aquatic Resources Agency, Holyoke College Clapp Laboratory, South Hadley,
Regional Research Centre, Kalpitiya, Sri Lanka MA 01075-1484, U SA
A n d e r s o n , M., D epartm ent o f Biological Sciences, Florida Farnsw orth, E. J., D epartm ent of Botany University of
International University, Mamiami Cam pus, Miami, Vermont, 120 M arsh Life Sciences Building, Burlington,
FL 33199, USA VT 05405, U SA
A n y i e n i , Z. M., Perm anent Presidential Commission on Soil F o n g , F. W., D epartm ent ofZoology, University o f M alaya,
Conservation & Afforestation, P.O. Box 46191, Nairobi, Field Studies Centre, 59100 K uala Lumpur, M alaysia
Kenya F o r t e s , M. D., M arine Science Institute, University of the
A z a r i a h , J., D epartm ent of Zoology University of M adras, Philippines, Diliman, Quezon City 1101, Philippines
Guindy Campus, M adras - 600 025, India F r a n c i s , J., Institute of M arine Sciences, University of D ar
A z a r i a h , H., CAS in Botany, University of M adras, Guindy Es Salaam , P.O. Box 668, Z anzibar, T anzania
Campus, M adras - 600 025, India G a m b r e l , R. P., D epartm ent o f Soli Science, Rivers State
B a c o n , P. R., D epartm ent of Zoology, University of the University of Science, P.M.B. 5080, Port H arcourt,
W est Indies, M ona, Kingstone 7, Jam aica Nigeria
B a l a s u b r a m a n ia m , S., D epartm ent of Botany, University G a n g , P. O., Technical Officer, D epartm ent of Resource
of Peradeniya, Peradeniya, Sri Lanka Surveys and Remote Sensing, P.O. Box 47145, Nairobi,
B o r g e r , R., Ecology & Systematics Laboratory, Free U ni Kenya
versity of Brussels, Pleinlaan 2, B-1050 Brussels, Belgium G e e t s , G., Dept. Zook, K enyatta Univ., P.O. Box 43844,
C a r b y , B., D epartm ent of Biological Sciences, Florida Inter N airobi, Kenya
national University, Tamiami Campus, Miami, G h e r a r d i , F., D epartm ent Animal Biology & Genetics,
FL 33199, USA U niversity o f Florence, Via R om ana 17, Firenze 50125,
C h a l e , F. M., D ept, o f Zoology & Marine Biology, U ni Italy
versity of D ar Es Salaam, P.O. Box 35064, D ar Es G ic h i r u , M., D epartm ent o f Soil Service, Rivers State
Salaam , Tanzania University o f Science, P.M.B. 5080, Port H arcourt,
C h o n g , V. C., D epartm ent o f Zoology, University of Nigeria
M alaya, 59100 K uala Lumpur, Malaysia G ie s e n , W., L aboratory o f Aquatic Ecology, University of
C o p p e j a n s , E., State University of G hent, L aboratory of Nijmegen, Toernooiveld, 6525 ED Nijmegen, The N ether
M orphology, Systematics & Ecology of the Plants, K. L. lands
L edeganckstraat 35, B-9000 G hent, Belgium G o r d o n , V., D epartm ent o f Biological Sciences, Florida
C o u t o , M., Faculty of Biology, University Eduardo Mon- International University, Tamiami Campus, Miami,
diane, C.P. 257, M aputo, M ocambique FL 33199, USA
xi
G o r i s s e n , M. W., Laboratory of A quatic Ecology, Catholic M e n k i r , V., N ational Academy of Sciences BOSTID ,
U niversity of Nijmegen, Toernooiveld, 6525 ED Nijme H A 476, 2101 Constitution Ave., N.W., W ashington,
gen, The N etherlands DC 20418, USA
G u n a s e k a r a n , S., D epartm ent o f Zoology, University of M i c h e l i , F., D epartm ent Animal Biology & G enetics, Uni
M adras, Guindly Campus, M adras - 600 025, India versity of Florence, Via R om ana 17, Firenze 50125, Italy
H a l l , K. C., U N D P/FA O Fuelwood Project c/o U N D P, M il l e r , B., D epartm ent o f Biological Sciences, Florida
P.O. Box 1011, Freetown, Sierra Leone International University, Tamiami Campus, Miami,
H a r r y , F. O. C., D epartm ent of Soli Science, Rivers State FL 33199, USA
University of Science, P. M. B. 5080, Port H arcourt, M o a l l im , A. H., D ata Collection & Storing Centre,
Nigeria Ministry o f Fisheries and M arines Resources, P.O. Box
H a t t o n , J. C., Faculty of Biology, University Eduardo 2589, M ogadishu, Somalia
M ondiane, C. P. 257, M aputo, Mozambique M o h a n r a j u , R., Centre Advanced Study in M arine Biology,
H e i m s a t h , A., G erm an Agency for Technical Cooperation, Annamalai University, Porto Novo - 608 502, Tamil
Kilifi W ater and Sanitation Project, P.O. Box 666, Kilifi, N adu, India
Kenya M o n d e g o , Faculty of Biology, University Eduardo M ond
H e l b i n g , U., IBAMA, Av. Rio Branco 186, 490000 Ara- iane, C.P. 257, M aputo, Mozambique
caju-SE, Brazil M u h a n d o , C. A., Institute of M arine Sciences, University of
H e m m i n g a , M. A., D elta Institute for Hydrobiological Re D ar Es Salaam, P.O. Box 668, Z anzibar, T anzania
search, V ierstraat 28, 4401 EA Yerseke, The N ether M u n g a , D., Kenya M arine & Fisheries Research Institute
lands (K M FR I), P.O. Box 81651, M om basa, Kenya
H y d e r , M., D epartm ent of Zoology, University of Nairobi, M u r a y a , J. K., D epartm ent of Botany, University of
P.O. Box 30197, Nairobi, Kenya Nairobi, P.O. Box 30197, Nairobi, Kenya
I s i r i m a h , N. O., D epartm ent of Soli Science, Rivers State M u s y o k i , B. M., Kenya Wildlife Service Provincial Wildlife
University Science & Techn., P.M.B. 5080, Port Office, P.O. Box 82144, M om basa, Kenya
H arcourt, Nigeria M u t e r e , J., Kenya M arine & Fisheries Research Institute
J a c c a r i n i , V., D epartm ent o f Zoology, University of (K M FR I), P.O. Box 81651, M om basa, Kenya
Nairobi, P.O. Box 30197, Nairobi, Kenya M u t w i r i , J. G., Kenya Marine & Fisheries Research Insti
K a n n a n , L., Annamalai University, Centre o f Adv. Studies tute (K M FR I), P.O. Box 81651, M om basa, Kenya
M arine Biology, Porto Novo - 508 502, Tamil N adu, M w a n g i , R. W., D epartm ent o f Zoology, University of
Indiua Nairobi, P.O. Box 30197, Nairobi, Kenya
K a s y i , J. N., D epartm ent o f Zoology, University o f Nairobi, N a s k a r , K., Central Inland Capture Fisheries Research
P. O. Box 30197, Nairobi. Kenya Institute, B arrackpore - 743 101, W est Bengal, India
K a t u p o t h a , J., D epartm ent of Geography, University of Sri N a t a r a j a n , R., Ocean D ata Centre for W ater Resources,
Jayew ardenepura, Gangodawilla, Nugegoda, Sri Lanka Annamalai University, M adras 600 025, India
K u d o j a , W. M., D epartm ent o f Zoology, University of N g o i l e , M. A. K., Institute of M arine Sciences, University
Nairobi, P.O. Box 30197, Nairobi, Kenya of D ar Es Salaam, P.O. Box 668, Z anzibar, T anzania
L a r a - D o m i n g u e s , A. L., EPOM EX Programme, A uto N g u l i , M. M., Kenya M arine & Fisheries Research Insti
nomous University of Campeche, Ap. Postal 520, tute (K M FR I), P.O. Box 81651, M om basa, Kenya
C lm peche 24030, Mexico N t i b a , M. J., D epartm ent of Zoology, University of Nairobi,
L a r e w , H., U SA ID - Office of the Science Advisor Agency P.O. Box 30197, Nairobi, Kenya
for International D evelopment, Room 320, SA-18, O j a n u g a , A. G., D epartm ent of Soil Science, Rivers State
W ashington D C 20523, USA University of Science, P.M.B. 5080, Port H arcourt,
L e h , M. U., D epartm ent of Zoology, University of Malaya, Nigeria
59100 K uala Lumpur, Malaysia O k e m w a , E., Kenya Marine & Fisheries Institute (K M FR I),
L e v y , I., D epartm ent of Botany, University of W est Indies, P.O. Box 81651, M om basa, Kenya
K ingstone 7, Jam aica O l u o c h , A., D epartm ent of Zoology, University o f Nairobi,
L i t t a u , J., D epartm ent of Botany, University of West P.O. Box 30197, Nairobi, Kenya
Indies, M ona, Kingstone 7, Jam aica O n y a n g o , F., College Biological & Physical Sciences, Uni
L l u i s m a , A. O., M arine Science Institute, University o f the versity of Nairobi, P. O. Box 30197, Nairobi, Kenya
Philippines, Diliman, Quezon City 1101, Philippines O n y a n g o , D., N airobi University, D epartm ent of Zoology,
M a r t e n s , E. E., D epartm ent of Zoology, University of P.O. Box 30197, Nairobi, Kenya
Nairobi, P.O. Box 30197, Nairobi, Kenya O s o r e , M. K. W., Kenya Marine and Fisheries Research
M a s t a l l e r , M., Tournelystr. 16, D-8901 Stadtbergen, Institute, P.O. Box 81651, M om basa, Kenya
Germany O y i e k e , H., Kenya M arine & Fisheries Research Institute
M e i j e r , G. J., Stinapa, Bonaire, N etherlands Antilles (K M FR I), P.O. Box 81651, M om basa, Kenya
xii
PiS S lE R S S E N S , P., RECOSCIX-W IO Project Kenya Marine T h o r h a u g , A., D epartm ent of Biological Sciences, Florida
& Fisheries Research Inst., P.O. Box 95832, M om basa, International University, 555 N E, 15th St. P H H , Miami,
Kenya FL 33132, USA
P o l k , Ph., Kenya-Belgium Project M arine Sc, Ecology & T r o n o , G. C., Jr, M arine Sciences Institute University of the
Systematics Laboratory, Free University o f Brussels, Philippines, Diliman, Quezon City 1101, Philippines
Pleinlaan 2, B-1050 Brussels, Belgium V a n A v e r m a e t , F., State University of G hent, Institute of
P r a s a d , D., D epartm ent o f Botany, University o f the W est Zoology-M arineB iology,K . L .Leganckstraat 35,B-9000
Indies, M ona, Kingston 7, Jam aica G hent, Belgium
R a s o l o f o , V. M., Centre N ational de Recherches O céano V a n D e K e r k h o f , K., L aboratory o f Aquatic Ecology,
graphiques (C N R O ), B.P. 68, Nosy Be 207, M adagascar University of Nijmegen, Toernooiveld, 6525 ED Nijme
R a s o w o , J., D epartm ent of Zoology, Moi University, P.O. gen, The N etherlands
Box 3900, Eldoret, Kenya V a n D e r V e l d e , G., Laboratory of A quatic Ecology,
R e e s e , R., D epartm ent of Biological Sciences, Florida Inter Catholic University of Nijmegen, Toernooiveld, 6525 ED
national University, Tamiami Campus, Miami, Nijmegen, The N etherlands
FL 33199, USA V a n G a n s b e k e , D., State University of G hent, Institute of
R e v i s , N., Ecology & Systematics Laboratory, Free Univer Zoology - M arine Biology, K. L. Ledeganckstraat 35,
sity o f Brussels, Pleinlaan 2, B-1050 Brussels, Belgium B-9000 G hent, Belgium
R o b e r t s o n , S. A., Coast Forest Survey, N ational Museums V a n H o v e , S., State University of G hent, Institute of
of Kenya, .P.O. Box 162, Malindi, Kenya Zoology - M arine Biology, K. L. Ledeganckstraat 35,
R o d r i g u e z , M., D epartm ent o f Biological Sciences, Florida B-9000 G hent, Belgium
International University, Tamiami Campus, Miami, V a n S p e y b r o e c k , D., D epartm ent of Botany, University of
FL 33199, USA Nairobi, P.O. Box 30197, N airobi, Kenya
R o h r s t e d , T., Delegation of the Commission of the V a n ’t H o f f , T., Stinapa, Bonaire, N etherlands Antilles
European Communities, P.O. Box 45119, Nairobi, Kenya V a n d e n B e r g h e , E., D epartm ent of Zoology, University of
R u w a , R. K., Kenya M arine and Fisheries Institute Nairobi, P.O. Box 30197, Nairobi, Kenya
(K M FR I), P.O. Box 81651, M om basa, Kenya V a n n i n i , M., University of Florence, Dept. Animal Biology
S a s e k u m a r , A., D epartm ent o f Zoology, University of & G enetics, Via R om ana 17, 50125 Firenze, Italy
M alaya, K uala Lumpur, Malaysia V a s a n t h a , K., Centre Adv. Studies M arine Biology, A n
S c h r i j v e r s , J., State Universtiy o f G hent, Institute of namalai University, Parangipettai - 608 502, Tamil
Zoology - M arine Biology, Ledeganckstraat 35, B-9000 N adu, India
G hent, Belgium V e r g e e r , L. H. T., L aboratory o f Aquatic Ecology, Univer
S e l v a m , V., D epartm ent of Zoology, University o f M adras, sity of Nijmegen, Toernooiveld, 6525 ED Nijmegen, The
Guindy Campus, M adras 600 025, India N etherlands
S e m e s i , A. K., D epartm ent of Botany, University o f D ar Es V i n c x , M., State University of G hent, Institute of Zoology,
Salaam , P.O. Box 35060, D ar Es Salaam, T anzania M arine Biology Section, K. L. Ledeganckstraat 35,
S h a y , E. G., N ational Academy of Sciences BOSTID , B-9000 G hent, Belgium
HA 476, 2101 C onstitution Ave., N.W., W ashington, V i s s e r , N., Agricultural A ttache, Royal N ertherlands Em
D C 20418, USA bassy, P.O. Box 41537, N airobi, Kenya
S h e t e , M. H. A ., E ast African Wildlife Society, P.O. Box V u a i , M., Forestry D epartm ent, Z anzibar, T anzania
20110, Nairobi, Kenya W a m u k o y a , G. M., KEN -U -TAN E X PO R TS, P.O. Box
S h u n u l a , J. P., Institute of M arine Sciences, University of 84384, M om basa, Kenya
D ar Es Salaam , P.O. Box 668, Z anzibar, T anzania W i j n a n t , J., Kenya-Belgium Project M arine Sciences
S i d r a k , G., D epartm ent o f Biological Sciences, Florida Kenya M arine & Fisheries Research Inst., P.O. Box
International University, Tamiami Campus, Miami, 81651, M om basa, Kenya
FL 33199, USA W o i t c h i k , A. F., EEC Project Mangrove Ecology, Free
S l i m , E., EEC - M angrove Project, D elta Institute for University of Brussels A N C H , Pleinlaan 2, B-1050
Hydrobiological Research, Vierstraat 28, 4401 EA Brussels, Belgium
Yerseke, The N etherlands W o l a n s k i , E., A ustralian Institute of M arine Science
S o l o r z a n o , L., Instituto N acional de Pesca Letamendi 102 (A IM S), PMB 3, Townsville M C, Queensland 4810,
y La Rita, P.O. Box 5918, Guayaquil, Ecuador A ustralia
T a c k , J. F., Ecology & Systematics Laboratory, Free Uni W o o d l e y , J. D., Discovery Bay M arine Laboratory, P.O.
versity of Brussels, Pleinlaan 2, B-1050 Brussels, Belgium Box 35, Discovery Bay, Jam aica
T e a s , H., D epartm ent o f Biological Sciences, Florida Inter Y a n e z -A r a n c ib ia , A., E PO M EX , Autonom ous University
national University, Tamiami Campus, Miami, of Campeche, Av. Agustín Melgar, Ap. Postal 520,
FL 33199, USA Campeche 24030, Mexico
T e j a n , A. G .,U N D P/FA O Fuelwood Project c/o U N D P,
P.O. Box 1011, Freetown, Sierra Leone
Hydrobiologia 2 4 7 : 1-10, 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 1
© 1992 Kluwer Academic Publishers. Printed in Belgium.
A delaida K. Semesi
Botany Department, University o f Dar es Salaam, P.O. B ox 35060, Dar es Salaam, Tanzania
Abstract
The m angroves o f T anzania are currently no t being m anaged although they are legally gazetted as for
est reserves. The m anagem ent o f this overexploited and under estim ated natural resource rests with the
F orest and Beekeeping Division in the M inistry o f N atural R esources and Tourism . It is the D ivision’s
intention to conserve the m angroves o f T anzania, therefore steps are being taken to reach th at goal. So
far the Division, through its established m angrove project, has done an inventory o f all the m angroves
of m ainland T an zan ia with the assistance o f N O R A D . Aerial photography and ground checks were used
to assess the state o f all the m angrove reserves in the country. Literature search and a socio-econom ic
study w as also undertaken.
The inventory revealed th at the area covered by m angrove vegetation is 115901 ha while the previ
ous records indicate about 80000 ha. If the bare saline areas, w ater bodies, and salt pans which are part
of the m angrove reserves are included, the area am ounts to 172879 h a for the T an zan ia m ainland. The
m angrove reserves have already been m apped in 30 sheets. Some sheets are m apped using a scale of
1:25000 while others are m apped in 1:50000 scale. The m aps show the vegetation types, the area o f each
vegetation type, and the stand density and height o f each com partm ent.
The p aper therefore presents the current status o f the m angroves o f T an zan ia in respect to area,
distribution and vegetation types. Included are the strategies being used to develop the m angrove
m anagem ent plans for T anzania.
ture holding capacity, extrem e alkalinity and hard of m angroves w as found in the Rufiji D elta. Fairly
subsoil which results in poor drainage. Along river large areas were also found at Tanga, Kilwa, at
basins alluvium deposits are found. the m ouths o f the Ruvu, W am i, Pangani and Ru-
vum a rivers.
The m angroves are distributed in the different
Population adm inistrative regions/D istricts as shown in T a
ble 1.
The total population of T an zan ia is 23174336
(1988 census). T an zania m ainland has 22533 758
inhabitants and the islands of Z anzibar and Mangrove vegetation
P em ba have 640578 people. The growth rate for
1978-1988 was 2.8% and the m ean population T here are 8 species o f m angroves trees th at occur
density is 25.3 km 2 for T anzania m ainland and in m ainland T an zan ia belonging to six families:
260 km 2 for Z anzibar. R hizophoraceae, Sonneratiaceae, Avicenniaceae,
C om bretaceae, Sterculiaceae and M eliaceae as
has been previously reported by Semesi (1986).
Materials and methods M ost com m on are Rhizophora mucronata and
Avicennia marina (Table 2). M ost im portant in
An inventory o f the m angroves o f T an zan ia was term s o f pole production is Rhizophora mucronata.
carried out using aerial photographs which were The outer zone is m ore often covered by
taken in 1988/89. They were supplem ented by Avicennia marina and som etim es by Ceriops tagal
ground reconnaissance and topographic m aps to or Lumnitzera racemosa. The seaw ard zone is o c
assess the m agnitude o f the forests and to clas cupied by Sonneratia alba or Avicennia marina.
sify them into different vegetation types. 30 m aps M oving landw ards the Sonneratia alba zone gives
were printed depicting 10 adm inistrative blocks. way to the Rhizophora mucronata zone which is
A socio-econom ic study on coastal com m unities alm ost im penetrable because o f the arched aerial
and how they use their m angrove environm ent roots.
was carried out (von Mitzlaff, 1990). Finally a D ue to the lim itation o f fresh w ater com ing into
w orkshop on m angrove m anagem ent brought to the upper intertidal sites some locations like parts
gether resource persons and decision-m akers in of the coast o f Tanga, Bagam oyo, Kilwa, Lindi,
order to arrive at a consensus on a future policy and M tw ara, the concentration o f salts in the soil
concerning the m anagem ent o f this resource. m ay exceed the physiological tolerance o f all the
These activities were carried out w ith the support plant species and thus a vegetation-free sandy
o f the F orestry Division and N O R A D . area is found before other non m angrove halo-
phytes like Sesuvium portulacastrum and Sueda
monoica are observed. This area is norm ally
Results flooded by seaw ater twice m onthly during the
highest spring tides.
Mangrove forest distribution
PEMBA
PANGANI
DIS TRIC'
Bl o c k i
ZANZIBAR
b a g a Mo t o
O IS T R I CT
( BLOCK 3 )
•
}-DA
SAL AAffy
REV"
^BL OCK 4 )
K I S AR AWE
( BLOCK 5 )
MAFIA'
MAFIA___
"b D IS T RI CT __
(BLOCK 6*T
I d ist r ic t
(BLOCK 7 )
k ilw a o istric
S C AL E I-.IOOO 0 0 0
1 0 0 KM
(BLOCK
■m t w a r a
/d istrict
.^ B L O C K I O )
Table 1. Mangrove forest area in the different districts in are im portant. B arnacles and the spiny oyster
Tanzania.
(Crassostrea cucullata) are very com m on on stems
District/region Area quoted in Area found o f m angroves especially on Sonneratia trees. On
forest records inventory (ha) in 1989 (ha) the leaves o f m angroves especially those o f Bru
guiera, Sonneratia and Ceriops are seen species of
1. M uheza and Tanga 9,996.1 9,403.3 (12,931.3)*
2. Pangani 600.0 1,755.6 (3,035.4)* Littorina.
3. Bagamoyo 1,499.0 5,635.8 (9,183.0)* Bees are visitors to the flowers o f the m an
4. D ar es Salaam - 2,168.2 (3,213.3)* groves, the honeydew o f which attracts the weaver
5. Kisarawe 3,498.0 3,858.3 (6,050.3)*
6. Rufiji 40,470.0 53,254.8 (67,612.2)* or tailor ants, Oecophyllia which weave the leaves
7. Mafia 404.7 3,472.9 (4,365.0)* together to form a n est and are m ost com m on in
8. Kilwa 13,994.5 22,438.7 (36,747.5)* the Ceriops zone in M uheza.
9. Lindi 1,416.0 4,546.5 (7,300.8)*
10. M twara 7,996.9 8,941.5 (13,249.1)* Several dem ersal fish species inhabit the creeks
and the m ud skipper lives in the m angrove mud.
Total 79,875.2 115,475.6(163,688.9)* M angroves are also the feeding and hunting
* The figure in the brackets include also the area of water in places for a num ber o f birds: herons, egrets, ibis,
the creeks, saline bare areas, and salt pans. sea eagles, and several species o f w aders and
Kingfishers. The insect and crab eating kingfisher,
and Scylla serrata. On sandy portions soldier Halcyon senegaloiides is restricted to the m an
crabs (Dotilla spp) are seen. Dotilla fenestrata is groves while the other birds are also found in
restricted to the E ast A frican coast. other habitats. T he area between the Rufiji delta
The m olluscs Potamides decollatus and Littorina and M afia island contains large num bers o f w ad
scabra are the m ost com m on species found in the ers and terns during the w inter in northern E u
m angrove form ation north o f Tanga. The m an rope (D anish-T anzania IC B P expedition to T an
grove snails like Terebralia, Cerithidea and Nerita zania from Ja n u a ry -M a rc h 1988). The m ost
Table 2. The relative composition of Mangrove vegetation types found along the coast of Tanzania.
Classification Area ha V
/o
pans for solar evaporation in Tanga, Bagam oyo inadequate and dispersed to different institu
and M tw ara D istricts. They are often placed tions (e.g. Forestry, Fisheries). N o institution
within m angrove areas which entails the total alized channels o f com m unication have been
clearing of the forest, whilst topography and soil established to overcom e this problem .
are equally suitable for their location behind m an
groves or onbare patches. Policy issues
Pole cutting, if done in an unplanned way and The m anagem ent policy which consist o f control
on a large scale, can reduce the genetic pool and and prohibition only, w ithout having the m eans
endangers reproduction. and capability to carry it out (lack o f funds, per
The dem and for dom estic fuelwood endangers sonnel, tran sp o rt and infrastructure), has affected
the m angroves only in densely populated areas the m anagem ent o f the m angroves in Tanzania.
around tow ns. If the inland resource is already O n top o f this it has no t considered traditional
exhausted, people turn to m angroves as a second interests o f people in their resource or their in
choice for fuelwood. volvem ent in its m anagem ent. The N ational ban
Agricultural activities interfering with the m an on cutting m angroves spelled out in 1987 has poi
grove environm ent take place in the Rufiji D elta soned the relationship betw een people and gov
only and this quite recent developm ent has a p ar ernm ent bodies even further.
ticular historical background. The practice of
shifting cultivation for rice farm ing is becom ing a Legal issues
serious th reat to the m angroves. D espite the ban, licenses for cutting poles and for
The m angroves are also endangered by oil pol the construction o f salt pans are issued. C onse
lution and industrial pollution, by excessive silt- quently the users o f the m angrove resource, be it
ation because o f m an’s degradation o f catchm ent villagers or business people, are confused about
areas, by herbicides com ing from inland via the the actual legal situation and their rights. Also
rivers, and by the construction o f dam s or m ajor they lack any pow er to lobby for their interests.
irrigation schemes. The traditional rights o f the coastal village
The aw areness o f the overall im portance o f this com m unities to use their resource for their small
resource is low am ongst the people in general. local needs have been disregarded although they
They value it rather for its direct m onetary return have never been a threat to the m angroves.
than its m any indirect benefits.
- other areas will be allocated to individual vil Zone I are all forests which will receive prime
lagers or groups for com m ercial use, i.e. they protection.
will have preferential access to licenses for the Zone I I are all ecologically stable areas with
exploitation o f the resource (pole cutting, salt sufficient regeneration potential. They are consid
production). ered productive forests suitable for controlled
- training on the m anagem ent by rationally ex harvesting.
ploiting the resource will be available for those Zone I I I are degraded areas which will be
who are engaged in it. This extension service closed for som e time to allow recovery and reha
has to be provided by the M angrove F orest bilitation.
Officers. Zone I V are areas m arked for ‘development’ of
- villagers will be given support in developing different kinds.
alternative econom ic opportunities. The Bee
keeping Division will encourage and assist vil
Conclusion
lage w om en and m en in the establishm ent o f a
honey and wax industry, in its processing and
Ecological changes, uncontrolled exploitation by
m arketing. T he Fisheries Division will assist in
m an, uncoordinated and inadequate institutional
im proving the situation o f the artisanal fisher
m easures, deficiencies in policies and an unclear
m en through the supply o f gear or through
overall policy have all contributed to the present
loans, through training and through im proving
situation, where the survival o f the m angrove for
technologies o f processing fish (the latter con
ests o f T an zan ia is in doubt. T here is urgent need
cerning w om en mainly).
for planned, coordinated action, there is need for
- com plem entary m easures, like establishing fu
a new policy and there is need also for an orga
elw ood plots for villages or individuals, will
nizational set-up which can m eet these tasks.
receive every possible support.
- infrastructure im provem ents will benefit the
villages as well as the m angrove m anagem ent. Acknowledgements
- villagers will be em ployed as M angrove F orest
G u ard s I w ould like to thank the N orw egian Agency for
- contracts will be m ade betw een the villagers or International D evelopm ent (N O R A D ) for finan
groups o f villagers involved and the F orest D i cial support and the U niversity o f D ar es Salaam
vision with regard to their rights and obliga for granting m e perm ission to w ork on the m an
tions o f exploiting and protecting the resource. grove project. M any thanks should also go to
U ltim ate authority lies with the F o rest D ivi m essrs A. M itoi, S. K ipangula, A. M tw eta for the
sion who can w ithdraw these rights if the ob airphoto interpretation, to M r. H ak an Sjöholm,
ligations are n o t met. C oordinator o f C atchm ent F o rest Project for his
- through the establishm ent o f U sers’ A dvisory valuable discussions. D r. Von M itzalaff, Dr. K.
B oards villagers and other com m ercial users Howell, D r. F. U rasa, D r. A. Kivaisi and M an
will be given an opportunity to voice their in grove W orkshop P articipants m ust be thanked
terests. for their critical com m ents.
FAO, 1982. Management and Utilization of Mangroves in Saenger, P., E. J. Hegerl & J. D. S. Davis (eds), 1983. G lo
Asia and the pacific. FAO Environment paper No. 3. Rome: bal status o f mangrove ecosystems. Commission on Ecol
FAO. ogy Papers Number 3. G land Switzerland: IU C N . The En
Flekstra, G. P., 1989. Global warming and Rising Sea levels: vironmentalist 3. Supplement No. 3.
The Policy Implications. The ecologist, 19 (1): 4-15. Semesi, A. K., 1986. Zonation and vegetation structure of
Linden, O. & A. Jernelov, 1980. The Mangrove Swamp - An mangrove communities in Tanzania. In: Mainoya J. R. &
ecosystem in Danger. Ambio, 9: 81-88. Siegel P. R. (eds), Proc. of a W orkshop on Save the M an
Mackinnon, J., K. Mackinnon, G. Child & J. Thorsell, 1986. grove Ecosystems in Tanzania. 21-22nd February, U ni
Managing Protected Areas in the Tropics. IU C N Switzer versity o f D ar es Salaam. 174-181.
land. Semesi, A. K., 1988. Status and utilization of mangroves along
Mercer, E. & L. S. Flamilton, 1984. Mangrove ecosystems: Tanga, Tanzania. In: Mainoya J. R. (ed.), Proc. of a Work
Some economic and natural benefits. Mar. Resour. shop on Ecology and Bioproductivity of the Marine Coastal
U N ESC O , 20: 14-19. W aters of Eastern Africa. 18-20th January, University of
Von Mitzalaff, U., 1990. Coastal communities in Tanzania D ar es Salaam. 174-181.
and their mangrove environment. A socio-economic study
prepared for N O R A D -D ar es Salaam, 60 pp.
Hydrobiologia 2 4 7 : 11-16, 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 11
© 1992 Kluwer Academic Publishers. Printed in Belgium.
K ey words: C oringa m angrove, structure, distribution, atm ospheric pollutants, industrial im pact
Abstract
Coringa m angrove forest is located in the G odavari delta, A ndhra P radesh, India. The m angrove
com m unity consisted o f m ore than 13 species o f m angrove and other plants in the present study area.
The following three dom inant m angrove plants, Avicennia marina, Excoecaria agallocha and Sonneratia
apetala were found to be present on the banks o f a m ajor channel o f the G odavari river running through
the forest. The structure and species distribution o f m angrove, in the C hannel N ag ath an a K alava has
been described. T he area behind the belt consisting o f Acanthus ilicifolius and Myriostachya wightiana is
generally colonized by E. agallocha and A . marina. The zone has been called the A vicennia and Excoecaria
zone. A djacent to this zone species like Aegiceras corniculatum and A . officinalis were the com m on
species. In the flat clayey soil, Suaeda maritima w as found to grow. In areas o f high elevation, devoid
of inundation o f tidal seaw ater during the high tidal period, species such as M . wightiana and Acanthus
were found to colonize both the banks o f the channels.
A n analysis o f species diversity, indicated a definite trend in the distribution o f m angrove from the
m outh o f the estuarine region to the inland waters.
The levels o f atm ospheric pollutants such as sulphur dioxide ( S 0 2), oxides o f nitrogen (N O x), am
m onia (N H 3) and suspended particle m atter (S P M ) were within the legal limits.
w ater berths, extension o f break w ater to 3.35 km, tributaries, G ad eru and V asavakalava are note
dredging o f access canal and provision o f ancil w orthy (see Fig. 1).
lary facilities such as road, approach bridges and
other navigational aids. T he project, costing R u
pees 1335.8 million, is expected to be com pleted M aterials and methods
by April 1995 (Indian Express, 1990).
A second m ajor developm ental activity is the T he present study w as carried out in the N agath-
building o f the fertilizer com plexes namely, G o ana K alva (canal) w hich extends for a distance of
davari Fertilizers and Chem icals L td (C apital cost 10 km from the C oringa river. T he area is swam py
o f R upees 1080 million, w ork com pleted) and the w ith knee deep m ud.
N ag arju n a Fertilizer and Chem icals Ltd (capital A series of quadrats o f 5 x 5 m (Sidhu, 1963)
cost o f R upees 6700 m illion-work nearing com were laid at an interval o f 1 km along the 10 km
pletion). These tw o fertilizer factories are ex long canal. The num ber o f trees in each quadrat
pected to m eet the fertilizer dem ands within a w as counted and the d ata were tabulated. U sing
200 km range from the factory. In view o f such a certain indices o f species structure in com m unity,
rap id pace o f industrialization in the coastal zone d ata on the index o f dom inance, species diversity,
o f K akinada, a detailed study on the environm en richness o f species were calculated. Besides, d ata
tal im pact assessm ent was m ade on a m ultidis on the relative density, relative frequency and
ciplinary basis. In an earlier paper the results of relative dom inance were calculated. D etails o f the
the analysis o f the m angrove vegetation within a form ulae used are given in an earlier paper (A z
15 km radius o f the fertilizer com plex have been ariah et al., 1986). A few m icrom etereological as
given (A zariah eta l., 1990). The present paper sessm ents were carried out to obtain baseline in
examines the structure o f the m angrove com m u form ation on the level o f sulphur dioxide, oxides
nity and its distribution in the C oringa forest, of nitrogen, am m onia and suspended particulate
w hich is located on one of the deltaic branches of m atter. The details o f the analysis are given in an
the northern zone o f the G odavari delta, so as to earlier paper (A zariah e ta l., 1990).
provide quantitative baseline inform ation on the
current status o f the ecosystem together with a
survey o f the current levels o f atm ospheric pol .1Mi le
lutants such as sulphur dioxide, oxides o f nitro Kakinada d Mangrove forest
Town
gen, am m onia and suspended particulate m atter. ■ St udy area
Kaki nada
Description of site Bay
-16 55N
M angrove forests are seen in places like Coringa,
Y anam , K andikuppa, K akinada and S acram ento
which are in the E ast G odavari delta. The C or
inga river originates at Y anam which is about
6 km on the upper reach o f G odavari river. This
tributory along with G adeur form s a p art o f the
G o d av ari estuarine system. B oth the tributaries
open into the southern side o f the C oringa bay.
The bay is nam ed after C oringa (Koriangi, a well
know n ancient port). A group o f islets are located
in the vicinity o f C oringa bay. Am ong these, H ope Fig. 1. M ap showing the location of the Coringa mangrove
Island, B hairavapalend series and C oringa and forest and the study site.
13
Table I. Concentration of atmospheric pollutants (/ig m 3) in Coringa forest and K akinada neighbourhood
Coringa:
Dec 88-F eb 89 1.3 ± 2.4 10.6 ±5.7 14.4 ±8.1 64.9 ±23.8
(0-3.0) (2-20.2) (0-28) (31-97)
M ar 89-M ay 89 3.6 ±2.9 36.0 ±27.0 13.0 + 4.7 80.5 ±31.6
(0-10) (0-98) (0-19) (38-137)
Kakinada neighbourhood:
Dec 88-Feb 89 2.76 ±2.83 46.9 ±30.5 31.0 ± 17.3 104.03 ±52.8
(0-20) (0-330) (0-180) (24-294)
M ar 89-M ay 89 1.9 ± 2.26 45.1 ±36.2 9.3 + 9.46 106.86 ±56.8
(0-12) (0-276) (0-81) (28-302)
Maximum admissible limit 80 /tm m 3 80 fig m ""3 30¿ig m 3 200 /tg m ~ 3
14
SPM
A B
The concentration o f suspended particulate m at
A B A B A B
ter exhibited a direct relationship with hum an ac
Fig. 2. Concentration of atmospheric pollutants in the Kaki
tivity. It’s level w as high both during the post
nada town and Coringa mangrove forest.
m onsoon and sum m er seasons, the values being
24 to 294 /ig m -3 and 28 to 302 /ig m ~ 3 respec
w as found. The concentration was higher in tively.
places where there were m ore hum an activities in
term s o f handling nitrogen based fertilizers (load
ing and unloading). D uring the sum m er m onths, Analysis o f mangrove ecosystem
the coastal zone registered a value ranging be
tween zero and 40 /ig m “ 3 and it varied from 2 The num ber o f plants recorded in 10 quadrats
to 330 f t g m “ 3 during the post m onsoon season along the distance o f 10 km are given in Table 2.
i.e. D ecem ber to February. The vegetation in the m angrove sw am p consisted
Total 93 62 48 46 14 36 79 93 71 72 614
15
of 13 species o f plants out o f which 11 species w ard the inland w aters. O n the other hand,
belonged to the m angroves. Secondly, their dis Suaeda maritima D um ort w as recorded in the first
tribution varied in a distinct pattern from the tw o quadrats w hich were laid very close to the sea
m outh o f river tow ards the inland water. coast.
The distribution o f Avicennia officinalis Lam . In general, the following three distinct zones of
w as restricted tow ards inland w aters whereas distribution can be m ade.
Excoecaria agallocha Lam . and Avicennia marina Z one I: Excoecaria agallocha - Suaeda maritima
(F orsk.) Vierh. were found along the entire stretch zone, which extends 1 to 2 km from the river
of the river i.e. from the coast to the inland w a m outh.
ters. Between E. agallocha and A . marina the Z one II: A m iddle zone in which a mixed com
dom inance o f the form er w as seen throughout the munity of Excoecaria, Avicennia and Aegiceras
study area. H owever, E. agallocha was m ore corniculatum L. m ay be found. This zone extends
thickly distributed tow ards the m outh o f the river. upto about 6 km from the river m outh.
On the other han d for the latter species, although Z one III: W as represented by Excoecaria, Avicen
distributed throughout the study area, the m ag nia and Cyperus which were found at a distance
nitude o f colonization was relatively less than o f 7 -8 km from the sea. In this zone a mixed
E. agallocha. com m unity o f Excoecaria, Avicennia and A can
The zone o f m angrove can be divided into two thus m ay be recognized.
areas. First, the Avicennia area which is located It is interesting to note th at Suaeda maritima
near the inland w ater and the second region dom was found only near the swam py regions o f the
inated by E. agallocha. This region w as found coastal zone along with Excoecaria and Avicen
m ore tow ards the ocean front. A feature o f inter nia. A nother interesting feature is the occurrence
est is that, when the num ber o f E. agallocha in of the Rhizophora apiculata Blume and Lumnitzera
creased the num ber o f Avicennia decreased. It is racemosa Willd. B ut these were no t present within
also interesting to note th at Avicennia officinalis the 10 quadrats. Rhizophora apiculata was repre
was n o t distributed within the area o f 4 km from sented by a single tree, which w as located in the
the m outh o f the river. It is likely th at this species m iddle stretch o f the study area. W hereas L. race
may prefer a inland habitat with lesser m agniture mosa w as found to be sparingly distributed.
of salinity variation. W hereas A. marina did not The species diversity varied from 0.92 to 3.13
exhibit such a relationship. A non m angrove spe with the highest level tow ards the inland waters.
cies Ipomaea tuba Linn, which w as closely asso Similarly, an ascending value in species richness
ciated with m angroves show ed a discontinuous w as recognized from the m outh o f the river to
distribution. N ine species (item N o. 5 to 13 in w ards the inland water. W here Excoecaria and
Table 2) showed a salinity related distribution in Avicennia dom inated, as in the case o f first 3 km,
th at they were represented in regions the river, the richness index was p oor ( 1.02 to 1.67) w hereas
where there m ay be wide fluctuations in salinity, it ranged between 4.85 and 5.4 in the region of
and were recorded only in quadrat num bers 6 to freshw ater regime. The species evenness index
12. T he q u adrats 1 to 3 were located near the w as relatively high in areas o f low salinity (T a
m outh which w as characterised by the presence ble 3).
of neritic w aters throughout the year due to its
proximity to the sea w hereas quadrat num bers 7
to 12 were located in the inland area where fresh Discussion
w ater condition existed during the m onsoon sea
son and brackish w ater regime found during the Figure 2 com pares the concentration of atm o
post m onsoon and sum m er seasons (S arm a & spheric pollution between K akinada tow n and its
G anapti, 1971). Myriostachya wightiana H ook was neighbourhood and the C oringa forest. The vari
found in habitats which were located m ore to ation in the atm ospheric pollution is a function of
16
Table 3. Species diversity, species richness and species eveness values in different quadrats
1 2 3 4 5 6 7 8 9 10
Coast line Inland water
Species diversity 1.15 1.81 0.92 1.26 1.42 1.69 2.52 2.69 3.13 2.95
Species richness 1.02 1.67 1.19 1.20 1.75 1.93 3.16 4.05 5.40 4.85
Species eveness 0.73 0.90 0.58 0.76 0.90 0.84 0.90 0.85 0.90 0.89
the influence o f m onsoon accom panied by the contradictory feature betw een the earlier and the
direction o f wind speed. The behaviour o f S 0 2 present study is the variation and the dom inance
and N O x w as similar in th at the concentration of o f species within the com m unity. In the above
these com ponents w ere higher in C oringa during areas o f study, Avicennia w as the dom inant spe
the sum m er season w hen the wind w as blowing cies, w hereas in the present study, Excoecaria was
in the SE and SW directions. It is likely th at w ind found to be the dom inant species and no pure
borne S 0 2 and N O x m ay be concentrated at the stand o f Avicennia w as recognized at any place.
Coringa reserve forest. O n the other hand, the A nother feature o f difference is the association of
concentration N H 3 and S P M w as higher at the Avicennia w ith Acanthus.
K ak in ad a tow n which is due to wide spread
hum an activity and the use o f nitrogenous fertil
izers. References
The species structure in the m angroves o f
C oringa appears to vary from site to site. R ao Azariah, J., H. Azariah, P. Pandion & S. Gunasekaran, 1990.
et al. (1963) and V enkatesan (1966) reported Coastal wetlands of Kakinada, A ndhra Pradesh. A case
the occurrence of 26 species in the neighbour study in environmental impact assessment. In W etland
ecology and management, Bharatpur (in press).
hood o f K akinada. O f which Suaeda maritima, Azariah, J., P. Banth, H. Azariah & V. Selvam, 1986. Impact
Heliotrophium curassavicum L., Fimbristylis of urbanization on the status of mangrove swamps of M a
cymosa R. Br. and Aeluropus lagopoides (L.) Trin. dras. In D. D. H ook etal., (eds), The Ecology and M an
occur in considerable num bers. In the C oringa agement of W etlands, Croom Helm, London: 225-233.
Blasco, F., 1975. Mangroves of India. Inst. Fr. Pondicherry.
forest, B lasco (1975) recorded the occurrence of
Tran. Sect. Sei. Tech., 15, 175 pp.
Excoecaria agallocha, Avicennia officinalis, Phoenix Indian Express, 1990. K akinada Port-Gateway for trade and
paludosa, Derris sp., Dalbergio sp. and Ceasalpinia industry. March, 2: 1-15.
sp. L ater Sidhu (1963) recognized three m ajor Rao, T. A., K. R. Agerwal & A. K. Mukherje, 1963. An eco
forest com m unities. They are tidal ever green, logical account of the vegetation of Rameswaram Island.
Bull. Bot. Surv. India, 5: 301-323.
semi-evergreen and tidal decidous. H e recognized
Saenger, P., E. J. Hegal & J. D. S. Davie, 1983. Global sta
pure Avicennia com m unity, mixed Avicennia com tus of Mangrove ecosystems. Commission on Ecology,
m unity and Aegiceras mixed com m unity. H e also Paper No. 3, IU C N Gland. 88 pp.
recognized a total num ber o f 20 species in his Sarma, D. V. R. & P. N. G anapti, 1971. Hydrography of the
studies carried out the at Coringa C oup N o. 2 Coringa river of the G autam i-G odavari estuarina system. J.
mar. biol. Ass. India. 13: 234-248.
and 4 (coup = area alloted by forest authorities
Sidhu, S. S., 1963. Studies on the mangrove of India. I. East
for clear felling) and various region within the G odavari region. Indian Forestar. 89: 337-351.
C oringa reserves such as G undhuhalla, Varava- Venkatesan, K. R., 1966. The mangroves of M adras state.
calea and P adavakpalam . A n interesting and Indian Forestar, 92: 27-34.
Hydrobiologia 2 4 7 : 17-27, 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 17
© 1992 Kluwer Academic Publishers. Printed in Belgium.
K ey words: m angrove ecosystem s, com m unity structure, floristic com position, complexity index,
estuarine m angals, fringing m angals
Abstract
M angrove stands in P uttalam lagoon and D utch bay, tw o interconnected lagoons situated on the
northw estern co ast o f Sri L anka have been classified broadly into tw o groups, i.e., estuarine and
island/m ainland-fringing stands. S tructural diversity o f six m angrove stands, representing the tw o types
was studied in term s o f floristic com position, density, basal area, m ean stand diam eter, tree height,
standing above-ground biom ass and leaf-area index. Rhizophora mucronata and Avicennia marina were
the dom inant species. Higher m ean stand diam eters for the m angrove stands in P uttalam lagoon indi
cated greater m aturity than the estuarine m angrove stands in D utch bay. N evertheless, estuarine stands
in D utch bay were structurally m ore com plex (complexity indices 8.11-22.7) than the island/m ainland-
fringing m angrove stands (complexity indices 1.38-6.78). H igher num ber o f species present in the es
tuarine m angrove stands is the m ajor elem ent th at contributes to the higher values for the complexity
indices for those stands. This appears to m ask the contribution o f stand-age to the com plexity o f a
m angrove stand. Therefore complexity indices alone m ay no t be used to explain adequately the struc
tural diversity am ong m angrove stands.
DUTCH BAY
Ka l a V
INDIAN 1 SRI
LANKA
OCEAN Erumathivu
IV
100 200
Serakkuli/
Karathivu
Eul l upi di
Pallivasalthurei
PUTTALAM
LAGOON
LEGEND
Estuarine
mangals
Fr i n g i n g
mangals
Salt
marshes
Puttalam
Fig. 1. Distribution of mangroves in the study area and the location of the study sites. I - Serakkuli; II - Pullupidi; III -
Pallivasalthurai; IV - Erumathivu; V and VI - Kala Oya.
19
P uttalam lagoon and D utch bay, together with U sing auxiliary inform ation on physiognom y such
Portugal bay (location - 8° 15' N, 79° 50' E; as species present, vigor of the stand, relative
Fig. 1) form the largest brackish w ater body in Sri location etc., m angals (the area covered with
L anka. T otal w ater surface is approxim ately m angrove plants) o f the study area have been
46000 ha. M angroves cover nearly 1200 h a and classified according to Lugo & Snedaker (1975)
2100 h a o f inter-tidal land around P uttalam la and the m ethod m odified by C intron et al. (1985).
goon and D utch bay respectively (K anakaratne R epresenting the tw o types o f m angals in the area
eta l., 1983). i.e., estuarine and island/m ainland fringing (Am-
Am ong the few river and creek systems asso arasinghe & Perera, 1984), six study sites were
ciated with these lagoons, K ala O ya is the long selected (Fig. 3).
est with the largest river basin. A sem i-diurnal
tidal p attern occurs in the lagoon and the m axi
m um tidal range is 79 cm (P erera & Siriw ardena, Floristic composition
1982). Surface salinities vary between 0.00 pp t (at
K ala O ya estuary) and 45.00 ppt. M angrove areas on the m ainland were visited
A dry clim ate prevails (Fig. 2) and the average over-land while the islands and K ala O ya estuary
annual rainfall (100-110 cm) is seasonal with two were accessed by boat. Specim ens o f the species
distinct dry seasons, i.e., Ja n u a ry -M a rc h and encountered were identified using descriptive lit
Ju ly -S ep tem b er. M axim um rainfall is in the p e erature and taxonom ic keys (W atson, 1928; D as-
riod from O ctober to D ecem ber by inter-m on- sanayake & Fosberg, 1980; Blasco, 1984) as well
soonal convectional rains. as by com paring them with the specim ens in the
N ational H erbarium at Peradeniya, Sri Lanka.
E num erated species were categorized into true
m angrove and m angrove associated species
(Saenger eta l., 1983).
Community structure
Pallivasalthurei
Serakkuli/ Kàrathivu
Am
8-
t Am
4-
Arrw
Erumathivu
Pul l upidi
8-
4-
Rm,
8-
, Am ■Ami
4-
0-
0 20 40 0 20 40 60
Fig. 3. Vegetation profiles of study sites. Am - Avicennia marina; Rm - Rhizophora mucronata; Lr - Lumnitzera racemosa; Ct -
Ceriops tagal.
21
Relationship between dbh and tree biomass for ships betw een dbh and total biom ass and the
Rhizophora m ucronata and A vicennia marina biom ass o f other com ponents o f the plant for
each species obtained from the tw o sites were
As R . mucronata and A . marina are the dom inant com pared using S tudent’s t-test and found to have
species in the study area these tw o species were no significant difference at 5% level between the
chosen to determ ine dbh and tree biom ass. Thus, two sites, thus all the d ata for a species in the both
the relationships were obtained using the m ethod sites were pooled and the relationship between
described by C intron & Schaefer-Novelli (1984) dbh and biom ass was com puted for each species.
and D ay et al. (1987). G irth at breast height (gbh) Allometric relationships betw een dbh and bio
of the trees in the four perm anent transects m ass were no t calculated for other species en
(10 X 50 m 2) in the m angals at Erum athivu island countered in the transects, as they often occur in
and K ala O ya estuary (2 transects per site) were low densities. A conservative estim ate o f biom ass
m easured in order to determ ine the range o f gbh for such species w as obtained by assigning the
for R. mucronata and A. marina in these tw o m an lowest biom ass to the corresponding dbh. The
gals. These two sites were particularly chosen for lowest biom ass values were selected out o f the
biom ass studies as litterfall and gbh increm ent biom ass values calculated for R . mucronata and
m easurem ents (to estim ate the net prim ary p ro A. marina for the corresponding dbh.
ductivity, which will be discussed in a subsequent
paper) were obtained only from these tw o m an
gals which were presum ed to represent estuarine Standing above-ground biomass
and fringing m angals. Ten trees o f R . mucronata
(representing the whole range o f dbh) from each Standing above-ground biom ass o f the stands
study site (total o f 20 trees), nineteen trees of were calculated by determ ining the frequency dis
A . marina from Erum athivu and ten trees o f the tribution o f the diam eters in each stand and es
sam e species from K ala O ya m angals were se tim ating the biom ass o f each diam eter class sep
lected for harvesting. These trees were cut as close arately, using the com puted regression equation
above-ground as possible. The canopy i.e., for each species. T he average biom ass o f a diam
branches, twigs and leaves (flowers and fruits eter class obtained thus, w as m ultiplied by the
were absent at the time o f harvesting), was num ber o f stem s in th at class and the results for
stripped and the weight o f each com ponent w as all the diam eter classes were sum m ed to obtain
obtained. Sam ples from each com ponent (about the total biom ass o f the trees in the transect.
500 g were collected and the fresh weight o f each
sam ple w as m easured to the nearest 0.1 g in the Table 1. List of mangrove species enumerated in Puttalam
field, using a triple beam balance. Tree trunks lagoon and D utch bay mangals.
were cut further into convenient lengths o f about
1. Acanthus ilicifolius L.
2 m, and the weight o f each length w as obtained
2. Aegiceras corniculatum (L.) Blanco
with a spring-balance. A few sections from a few 3. Avicennia marina (Forsk.) Vierh.
lengths were cut as sam ples and fresh weights 4. Bruguiera cylindrica (L.) Blume
were obtained. F resh weight o f the prop roots of 5. Bruguiera gymnorhiza (L.) Lam
harvested R . mucronata trees were obtained and 6. Ceriops tagal (Perrot.) C.B. Rob.
7. Cynometra ¡ripa Kostei.
weighed sam ples were brought to the laboratory
8. Exoecaria agallocha L.
to determ ine the dry weight to wet weight ration 9. Lumnitzera racemosa Wiild.
(conversion factor). W et weights m easured in the 10. Rhizophora apiculata Blume.
field were converted to dry weights using this con 11. Rhizophora mucronata Lam.
version factor. These estim ated dry weights for 12. Scyphiphora hydrophyllaceae Gaertn.
13. Sonneratia alba J. Smith
each com ponent were sum m ed up to obtain dry
14. Xylocarpus granatum Koenig.
weight o f trees. Slopes o f the regression relation
22
Table 2. Relative density (Rde), dominance (Rdo), frequency (Rfr) and importance value (Imv) for the major mangrove species
(Based on the trees with diam eter>2.5 cm). Roman numerals in the parentheses indicate the location in Fig. 3. A.m. - Avicennia
marina; R.m. - Rhizophora mucronata', L.r. - Lumnitzera racemosa', C.t. - Ceriops tagal.
Riverine mangals
(Dutch bay)
Kala Oya (V) Rde 65.70 25.90 1.40 2.10 4.90
Rdo 88.53 8.85 0.39 0.69 1.54
Rfr 33.33 25.00 8.33 16.67 16.67
Imv 187.56 59.75 10.12 19.46 23.11
Kala Oya (VI) Rde 13.50 52.30 28.80 4.50 0.90
Rdo 27.18 62.75 2.96 7.00 0.11
Rfr 36.36 27.27 9.09 18.19 9.09
Imv 77.04 142.32 40.85 29.69 10.10
Fringe mangals
(Puttalam lagoon)
Karathivu/Serakkuli (I) Rde 100.00 - - - -
Rdo 100.00 - - - -
Rfr 100.00 - - - -
Imv 300.00 - - - -
Pallivasalthurai (III) Rde 100.00 - - - -
Rdo 100.00 - - - -
Rfr 100.00 - - - -
Imv 300.00 - - - -
Pullupidi (II) Rde 89.20 6.50 4.30 - -
Rdo 99.42 0.24 0.34
Rfr 60.00 20.00 20.00 - -
Imv 248.62 26.74 24.64 - -
(Dutch bay)
Erumathivu island (IV) Rde 36.20 63.80 - - -
Rdo 36.20 63.80 - - -
Rfr 50.00 50.00 - - -
Imv 122.40 177.60 - - -
( A)
91.0 12 - 0 *
y= 6-18 + 2-11 x
r= 0-973 ( P < 0 0 5 )
co.
o
X
Ul
V, 47-0 10 - 0 -
CtJ
E
o
CQ
«3
O
(A) (B)
87-0' 11.5-
V= 5-55 -t- 2-15 x
°
r = 0-960 ( P < 0 05
oo~ U
U ll
o CD
X E
O) o
in
cn
C/) CD
£ 44-0- 9-5-
o
bo
O)
o
1-a 7-5
2-0 6-8 11-6 0-6 3-4
dbh (cm) lo g dbh
Fig. 5. (A) Curvilinear and (B) log-log relationships between total biomass and dbh for A. marina in D utch bay.
24
3 (D "tí
tí > S tí I
títí °) o3 ¡ S
X m r-~
oo un
o en ÍN ,—i
t í ON t-~ r-
<! bß x w íN
£
o
o<
X 3 ■a
X
T3 X)
j
r - OO oo
00 en r -
un
ON
u° .a
"2 íN oo
ÍN
ÍN 1-H NO
tí; un íN en en ON
tí tí r- Ö t í en
ON o o r- <N
t í r- tí o oo
oo NO as NO oo
tí íN ÍN r"H
tí "Ö _ O n r - íN OO
X g S tí. O n r- CN tí
+2 tí ü
C/3 X c3
O n en en ON t í
t í <N en
ex
tí
íN t í NO NO NO un
On r- t í NO en en
ON íN íN OO oo en
t í CN CN
f - t ' NO ÍN un
O en en O ON -—<
3
tí O o r-H r-H un oo
CQ
Q
-o
§
tío T3 ."tí
H
o § o o
o o
o un
O O tí
O un en
o
o
SP 55 "O w NO t-~
e n íN
OO t í un CN
(N
T3 un o O un r - O
tí S íN o o íN NO <N
CX un ÍN NO en r - ÍN
o O t— o r - O
un en NO o NO en
NO íN NO en OO ÍN
bß
tí un O o un r - O
r- un o CN NO OO
'5b o un íN r- ON
tí en ÍN
bß
b ÍN
o o
o NO
en OO IN
en en NO
O
c O o On oo t í
ÍN
tí
A t í tí ÍN tí
O
r - un *—
h en íN
tí
bß
tí
tí T3
¿a .s tí
■SI S •tí
C/5
tí 2h ■M
3 c
3O 1 > > ^ co ;ss
tí tí tí bOß rtí| tí>
t í tí bß
en ex . ?n tí
tí —1T3tí ^ X 5 3
tí S-o o 'e x 2
^ *«, <% tí tí tí
1 3 ”tí Db ß J5 33 JÜ 2 tí
^“-CD 13 tí
Ö oi-H
^ *¡
•s s a £ tí tí
S E fe ex Q
25
associated and salt m arsh species enum erated in in P uttalam lagoon are occupied by a single spe
Puttalam lagoon and D utch bay are listed in T a cies, frequently by A. marina. A ccording to C in
ble 1. Vegetation profiles o f the study sites are tro n et al. (1985), the m ost co n stan t conditions of
depicted in Fig. 3. In P uttalam lagoon m angals, salinity and nutrients as well as the exposure to
A . marina had a higher im portance value (C intron expectedly higher kinetic energy regimes which
& Schaefer-Novelli, 1984) than in the D utch bay are relatively unfavourable to support a large
m angals and the reverse w as true for R . mucro num ber o f species m ay restrict the num ber of
nata (Table 2). species occurring in fringing habitats. In the es
tuarine m angals, nutrients and salinity m ay oscil
late widely and the kinetic energy o f w ater flow
Community structure m ay anticipated to be lower than th at of fringing
m angals. M oreover, during periods o f flood, con
S tructural features are presented in Table 4. siderable tidal and w ater flow energies m ay be
Stand densities o f the m angals in D utch bay dissipated in the riverine (estuarine) sites (C in
are considerably higher than those in P uttalam tron e ta l., 1985). A lthough such periods are rel
lagoon and a reverse trend occurs with regard atively short, they m ay favour m angrove estab
to stand basal area and standing above-ground lishm ent m ore than the co n stan t tidal m otion in
biom ass. Com plexity indices for estuarine m an the fringing m angrove habitats, especially in a
gals are greater th an those for the fringing m an low-tidal environm ent. Thus as a result o f the
gals. oscillating conditions and m ore favourable nutri
R elationship between dbh and biom ass for ent balances, estuarine m angals m ay be able to
R . mucronata and A . marina are graphically p re support m ore species than fringing m angals.
sented in Figs 4 & 5. Regression coefficients used G eneral zonation o f species within a m angal
to estim ate above-ground biom ass are presented reflects the gradient in flooding and soil charac
in Table 3. teristics o f the m angrove species. Kenneally
(1982) listed frequency o f tidal flushing, soil type,
soil salinity, drainage plant and anim al interac
Discussion tions as the m ajor factors th at appear to deter
m ine the survival o f individual species and thus
Floristic composition and zonation zonation w ithin a m angal. In the study area where
a dry clim ate and a small tidal am plitude prevail,
As in m any other m angals in the dry coastal re the frequency o f tidal flushing in the upslope may
gions in Sri Lanka, R . mucronata and A. marina be significantly low and as the terrain is flat, soils
are the m ajor constituent species of the m angals are poorly drained thus creating a soil salinity
of P uttalam lagoon and D utch bay. A . marina is gradient inland. M ore salinity tolerant m angrove
the m ost dom inant species in the m angals in P u t species such as A. marina and Excoecaria
talam lagoon with relative dom inance values agallocha as well as the salt m arsh species occupy
ranging from 60 to 100 and im portance values such landw ard portions while know n less tolerant
ranging from 248.62 to 300. Ability o f A . marina R . mucronata occupies the w ater-front zones of
to tolerate a wide range o f environm ental condi the inter-tidal land. Seasonal precipitation and
tions (W ells, 1982) such as salinity and air tem high evapo-transpiration rates m ay have largely
perature, m ay favour its distribution, no t only in contributed to the developm ent o f salt m arshes
frequently inundated areas but also in infrequently and salt flats which characteristically occur
flooded upper inter-tidal areas. R. mucronata is around P uttalam lagoon where freshw ater influx
m ore dom inant th an A . marina in the estuarine is apparently low.
habitats o f K ala Oya.
M o st o f the fringing habitats, especially those
26
Lugo, A. E. & S. C. Snedaker, 1975. Properties of a mangrove tus of mangroves. IU C N Commission on Ecology, Pap.
forest in southern Florida. In G. E. Walsh, S. C. Snedaker No. 3.
& H. J. Teas (eds), Proc. International Symposium on Bi Watson, J. D., 1928. Mangrove forests of the Malay penin
ology and Management of Mangroves, IFAS, University of sula. Malaya Forest Records 6: 1-275.
Florida; USA: 170-212. Wells, A. G., 1982. Mangrove vegetation of northern A ustra
Perera, W. K. T. & P. P. G. S. N. Siriwardena, 1982. Topog lia. In B. F. Clough (ed.), Mangrove ecosystems in A ustra
raphy of Puttalam lagoon. J. Inland Fish. (Sri Lanka) 1: lia: Structure, Function and Management. Australian In
97-104. stitute of Marine Sciences: 57-78.
Saenger, P., E. J. Flegerl & J. D. S. Davie, 1983. Global sta
g ft
Hydrobiologia 2 4 7 : 29-36, 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 29
© 1992 Kluwer Academic Publishers. Printed in Belgium.
The current status o f mangroves along the Kenyan coast: a case study of
M ida Creek mangroves based on remote sensing
K ey words: m angroves, K enya, rem ote sensing, species com position and distribution
Abstract
M angroves form a unique ecosystem o f limited extent covering an area o f about 53000 h a along the
K enyan co ast which need protection from overexploitation. B ackground inform ation on the areal ex
ten t and status o f these m angroves is limited and m akes their protection and m anagem ent difficult.
A m odel study has been carried out on the M ida Creek m angroves based on a double sam pling ap
proach starting with S PO T m ultispectral satellite imagery followed by ground checks to provide infor
m ation on species com position, density and distribution o f m angroves. The utilization o f m angroves and
uses around and within the m angroves were determ ined.
D uring this study, seven m angrove species were identified in M ida Creek, namely: Avicennia marina,
Bruguiera gymnorrhiza, Ceriops tagal, Rhizophora mucronata, Sonneratia alba, Xylocarpus granatum and
Lumnitzera racemosa. This report briefly outlines the present utilisation o f m angrove species. It is rec
om m ended th at a m ultidisciplinary m anagem ent plan be developed in order to conserve and m anage the
m angroves o f K enya on a sustainable yield basis.
m angroves has occurred in the p ast unnoticed or and w ithout the m angrove area in M ida
unquantified, as has been previously reported by Creek.
D oute eta l. (1981) who based their study on the (iv) To give possible suggestions and recom m en
use of L an d sat M ultispectral Scanner (M S S ) im dations th at m ay be applied for sustainable
agery. The study is now under review (Agatsiva utilisation o f the m angroves.
et al., 1990) to up-date it at the sam e time giving
m ore details which were never achieved in the
1981 study. The study area
The lack o f up-to-date and accurate d ata poses
planning problem s on conservation and m anage Location o f the study area
m ent of such m angrove species. It has been noted M ida Creek is located in Kilifi district on the
th at m angroves and their related ecosystem s have north coast o f K enya and is approxim ately 25 km
in the p ast undergone drastic changes w ithout south o f M alindi tow n. The area lies within ap
m uch attention being paid to their ultim ate fate. proxim ate bounds o f longitude 39° 58' E and lat
W hen exploiting a renew able resource, the con itude 3° 20' S (Fig. 1).
cept o f sustainable developm ent should be borne
in m ind. Conflicting land-uses such as fishing, Climate
cutting o f m angroves for different uses, salt m in The rainfall pattern is bim odal. The long rains fall
ing and various m arine activities are likely to ex in April to June, with a peak in M ay. The short
haust the m angrove resources. rains fall from O ctober to D ecem ber. T he area
The M ida Creek m angroves (G ang, 1988) have receives an annual rainfall o f about 900-1100 mm
been considered in an effort to design an appro (D istrict D evelopm ent P lant 1989/83).
priate m ethodology for having the K enyan m an
groves rapidly and cheaply assessed. The recom Soil types
m endations m ade m ay enable them to be properly Soils o f the m angrove sw am ps are poorly drained,
conserved, m anaged and where exploited, to be deep and excessively saline, olive to greenish grey
appropriately harvested on a sustainable yield loam to clay and often w ith sulfidic m aterial (D is
basis. M ida Creek w as preferred because the area trict Environm ental A ssessm ent R eport, 1985).
has a variety o f m angrove types as well as differ
ent regeneration stages and som e dense old L and use
stands. There are also several land use practices Agriculture is one o f the land use types in the
in and around the creek w hich m ay also influence area. Tree crops occupy m ore than one third o f
the m angrove ecosystem along the coast. The the area under cultivation. The m ost im portant of
bro ad recom m endations m ade on this area are which are cashew nuts, m angoes and coconuts.
therefore likely to represent a w ider spectrum of C oconuts are very m uch valued due to the role
inform ation concerning all m angroves along the they play in providing daily requirem ents such as
K enyan coast. cooking fat, woodfuel, building m aterials, local
beer and hair oil. C ashew nuts are sold to the
The objectives o f the study are therefore processing plant based in Kilifi. M angoes are sold
(i) To docum ent the extent and status o f the to the village, hotel industry and traders in M o
m angroves of M ida Creek. m basa.
(ii) T o assess the suitability o f S PO T m ultispec Fishing is another im portant activity in the
tral satellite imagery in providing baseline study area. M ost o f the catch is harvested by
d ata for the m onitoring and m anagem ent of individual fisherm en using their own boats and
m angrove areas. gear following traditional m ethods. They are often
(iii) To prepare a checklist o f plants growing in seen in their small boats in the creek which is
close proximity to the m angroves both within surrounded by m angroves.
31
M I D A C R E E K
All w e a t h e r r o a d
L oo se s u rf a c e
M a in t r a c k s
N P /G a m e r e s e r v e b o u n d a r y
H uts
4 0 * 0^ E o f G re e n w ic h
image. This w as done by use o f a tran sp aren t and counting the dhow s transporting m angrove
overlay put over the image and placed on a light poles to a collection yard at U yom bo. Finally an
table where all the distinguishable features were attem pt w as m ade to prepare a vegetation check
delineated. The differentiation w as based on co list o f plants growing in close proxim ity to m an
lour intensity, chrom a and hue in addition to other groves. T he m ethods used were those used by
image characteristics and interpretation elements K enya R angeland Ecological M onitoring U nit
including pattern and texture am ong others as (K R E M U ) in its R angeland M onitoring attem pts
described by Lillesand & Keifer (1979). (K uchar, 1979) and those used by botanists else
A topographic m ap at the scale o f 1:50000 where in com piling lists o f plants by occurrence
(sheets N o 193/1 and 193/3 from survey of in different habitats.
K enya) w as used to associate various m anifesta
tions in the image with contents o f the m ap re Results and discussion
ferring to land use and also for orientation and
navigation during field surveys. T he above double sam pling approach yielded
good results on the m angrove species distribution
Ground measurements and uses. Seven m ajor m angrove species were
identified (Table 1).
In each stratum obtained, representative stands The S P O T image delineated h ad the following
w ere sam pled to obtain inform ation regarding the cover classes (Fig. 2).
m angroves on:
A vicennia m arina
(i) M angrove Species
This is found in the landw ard fringe o f the Creek.
(ii) Height class
The underground is mainly covered by pneum ato-
(iii) C row n cover
phores. This covers 160 h a (2.9% ).
(iv) D ensity o f species
(v) D iam eter at breast height (dbh).
A vicennia/C eriops/R hizophora
R epresentative stands were selected within each This is characterised by species which have dis-
stratum , using 50 m by 50 m quadrats. T he total continous canopy form ations and generally in
count m ethod w as used as described in M uchoki poor state o f health, covering 145 h a (2.6% ).
(1988). All m angrove trees 0.7 m high and above
were m easured and quantitative attributes such R hizophora/C eriops
as height, dbh and crow n cover were recorded. D om inated m ainly by Rhizophora mucronata and
Visual observation was also done to determ ine Sonneratia alba at the edge o f the creek and
the species being exploited. This included actual Ceriops tagal w ith few Bruguiera gymnorrhiza and
counting o f stum ps o f the already cut m angroves Xylocarpus spp. The undercanopy consists of
L E GE ND
Avicennia
E S R hizophora / C eriops
I c I Mixed tr e e A griculture
1 * ^ 1 O p e n S w a m p s - Muddy
Kitilepwc
Uyombo
Fig. 2. Classification of M ida Creek mangroves and surrounding cover types based on SPOT imagery.
creek. Tree crops found in this zone are m angoes, Table 3. Plants growing in close Proximity to mangroves in
M ida Creek.
coconuts and cashew nuts. There is also natural
vegetation growing together with tree crops. This Species Family
includes Thespesia danis Olive and Azadrachta
indica A. Juss. This class covers 3805 h a (68% ). Aeva lanata Amaranthaceae
Anisotes parviflora Acanthaceae
Aristogeitonia monophylla Euphobiaecae
Open swamps ¡muddy areas
Asparagus africanus Lam. Liliaceae
These are sw am py areas th at no longer receive Azima tetracantha Lam. Salvadoraceae
flooding except during high tides, im mediately Cremaspora trifloa Rubiceae
within the vicinity o f these areas, small halophytic Cyperus tuberosus Rottb. Cyperaceae
plants such as Salicornia sp. and Sueda sp. are Canaralia carthartica Thouars Papilionaceae
Dactyloctineum gem inatum Hack Gramineae
dom inant. The category covers 292.5 h a (5.2% ).
Derris spp. Papilionaceae
Satellite imagery (S P O T image) proved to be Digitaria milanjiana (Randle) Stapt Gramineae
very useful in identifying m angrove areas and Eleaodendron schweinfurthianum Loes Celastraceae
other land use types in close proximity. Different Enteropogon macrostachysus (A. Rich) Gramineae
m angrove zonations are easily identifiable using Benth
Euclea schimperi (A. D. C.) Dandy Ebenaceae
tonal differences in the image. It is no t possible to
Flacourtia indica (burn, f) Merr. Flacourtiaceae
identify m angroves to species level in the image Fimbristylis triflora (L.) K. Schum Cyperaceae
but tonal differences interpreted were confirmed Fuirena ciliaris (L.) Roxb. Cyperaceae
to represent different m angrove types during field Grewia vaughanii exell. Tilliaceae
surveys. It was n o t possible also to detect the Hyphaene compressa H. Wendl. Palmae
Ipomoea pes caprae (L.) R. Br. Convolvulaceae
cutting o f the m angroves from the satellite im ag
Mormodica trifoliata Hook F. Cucurbitaceae
ery because the cutting is selectively done and it Panicum infestum Anderss. Gramineae
is n o t concentrated at one spot. A ctual counting Panicum piniplium Gramineae
in the sam pling plots indicated a cutting intensity Phoenix reclinata Jacq. Palmae
o f thirty poles per 50 X 50 m plot. Table 2 shows Phragnites communis Gramineae
Premna spp. Verbanceae
som e o f the m angroves species identified in the
Psilotrichum sericeum (Roxb.) Dal 2 Amaranthaceae
M ida Creek using the total count approach. The Saliconia sp. Chenopodiaceae
rest o f the plants in the surrounding areas are Salvadora persica L Salvadoraceae
indicated in Table 3. The cutting intensity de Sesuvium portulacastrum (L.) L Aizoaceae
creases away from the landw ard fringes probably Sideroxylon inerme L Sapotaceae
due to accessibility and convenience. The forest
departm ent controls the cutting by issuing licenses
and ensures th at poles are cut according to the
quantity specified, and suspends further cutting in
Table 2. Total number of species obtained through the total som e areas until regeneration is restored. There
count method from five sampling plots randomly chosen in
are no afforestation activities in the creek but
M ida Creek.
rather the m angroves regenerate naturally, grow
Mangrove Plots ing into exploitable poles after 3 -6 years. N o
quantitative m easurem ents were carried out on
1 2 3 4 5 Total
regeneration but it w as observed during field sur
Ceriops tagal 46 107 7 47 2 209 veys th at species like Ceriops tagal and Rhizophora
Xylocarpus granatum 7 - 12 1 - 20 mucronata have several seedlings actively grow
Bruguiera gymnorrhiza 2 1 - 2 1 6 ing. This m ay suggest th at the forest departm ent
Avicennia marina 1 - - - 82 83 need to regulate the m ode o f harvesting in order
Rhizophora mucronata - 32 27 23 1 83
to ensure sustainable exploitation o f the m an
Sonneratia alba - - - - 5 5
groves in this creek.
35
Table 4. Summary of mangrove data in plot No. 2 near D abaso in M ida Creek.
gain in-depth understanding of the param eters D .R .S .R .S ., N airobi, for his support and encour
controlling the distribution o f each m angrove spe agement. I also extend my gratitude to African
cies within a given habitat. Effects o f such p a Wildlife foundation, regional office in N airobi,
ram eters as land use activities and erosion in the through the fund for conservation endeavour for
hinterland, salinity, tem perature and depth of providing p art o f the funding for the survey ac
substrate type are im portant for designing advi tivities in M ida Creek. I am also grateful to m em
sory and protection schemes for the proper inter bers o f the E ast A frican H erbarium for their help
pretation o f the results from m onitoring p ro with plant identification and finally to technical
gramm es. staff, D .R .S .R .S . for criticizing the text.
K ey words: m angrove ecosystem s, estuarine m angals, fringing m angals, litterfall, prim ary productivity
Abstract
Productivity studies were carried out from Septem ber, 1985 to August, 1987 in tw o m angrove stands,
i.e. estuarine and island fringing, in D utch bay, a lagoon situated on the northw estern coast o f Sri Lanka.
N et above-ground prim ary productivity w as m easured by m onitoring litterfall and above-ground biom ass
increm ent. The average annual rate o f litterfall in the estuarine and island-fringing m angrove stands are
588.14 g m -2 (approxim ately 6 t h a _1) and 407.33 g m -2 (approxim ately 4 t h a _1) respectively. The
average annual rates o f above ground w oody growth are 614.74 g m ' 2 (approxim ately 6 t h a - 1) in the
estuarine stands and 286.8 g m -2 (approxim ately 3 t h a _1) in the island-fringing m angrove stands.
H ence estuarine m angrove stands record a higher annual rate o f above-ground net prim ary production
(N P P ; 1207.88 g m -2 or approxim ately 1 2 t h a _ 1 ) than the fringing m angrove stands (694.22 g m - 2 ;
approxim ately 7 t h a - 2). The annual rate o f N P P in the w ater front zones o f the stands (1300.47 g m “ 2
in the estuarine stands and 874.56 g m _ 2 in the fringing stands) are greater than those in the back-
m angrove zones (115.28 g m “ 2 in the estuarine stands and 513.88 g m -2 in the island-fringing stands).
T hese variations m ay be attributed to the differences in tidal flushing and influence o f freshw ater in the
two localities.
e s tu a r y
mixed back-m angrove zone) and 18 at E ru
m athivu (8 in the w ater-front R . mucronata zone
and 10 in the A . marina back-m angrove zone)
were used for the purpose.
Traps o f 1 m 2 in area, m ade out of w ooden
fram es and nylon m esh were placed random ly in
each zone, along the 10 m wide belt transects
which have been laid perpendicular to the
shoreline/riverbank across the m ajor zones o f the
m angals upto 50 m inland, at the tw o study sites
86 87 (2 such transects per site). T raps were installed at
Fig. 1. Monthly variation in salinity from December 1985 - a level higher than the highest tide level, using
January 1987 at the study sites. stakes/poles and nylon ropes.
39
. t ot al litterfall
CD
"O
CN
E
O) eaves
misc. + tw lg s
85 86 87
Fig. 2. Variation in total litterfall and components in the Rhizophora mucronata zone of the riverine/estuarine mangals at Kala Oya
estuary.
Litter (trapped) w as collected fortnightly, for a biom ass of the trees in the plots were estim ated
period o f two years from Septem ber 1985 to A u using the following regression equations (A m a
gust 1987. Collected litter w as brought into the rasinghe & B alasubram anium , this volume).
laboratory in plastic bags, sorted into species and
litter o f each species into plant com ponents, i.e., F o r R . mucronata
leaves, stipules, floral parts, fruits/seedlings, log,, (total biom ass) = 6.247 + 2.64 log,, (dbh)
bran ch w ood and m iscellaneous item s and they
were dried at 70 ° C to co n stan t weight.
F o r A. marina
log,, (total biom ass) = 5.551 + 2.153 log,, (dbh)
Above-ground woody biomass production
Table 1. Averages litterfall rates ± standard deviation (g m 2 yr ') by component in the major zones of the mangals at Kala
Oya.
R. mucronata zone
1985-1986 535.63 57.57 14.30 22.49 14.24 644.23
± 18.24 ±2.24 ± 2.11 ±2.25 ±0.85 ± 18.49
1986-1987 505.81 52.11 13.25 18.42 13.75 603.34
± 17.14 ±2.36 ± 1.33 ± 2.02 ±0.87 ± 18.23
Whole period 19.22 54.84 13.78 20.46 14.00 623.79
+ 17.36 + 2.26 + 1.73 + 2.10 + 0.84 + 18.03
Table 2. Average litterfall rates + standard deviation (g m 2 yr ') by component in the major zones of the mangals at Erumathivu.
R. mucronata zone
1985-1986 408.54 46.95 11.16 29.43 16.80 513.38
± 19.06 ±2.11 ± 1.68 ±4.99 ±1.26 ± 22.58
1986-1987 273.68 44.61 20.59 13.30 16.35 368.53
±9.25 ± 1.78 ±1.93 ± 1.45 ± 1.08 + 8.45
Whole period 341.11 45.78 16.13 21.37 16.58 440.96
± 15.86 ± 1.91 ±1.81 ±3.67 ± 1.15 ± 17.77
A. marina zone
1985-1986 304.77 - 34.10 20.13 34.80 363.80
± 16.68 - ±4.20 ±2.38 ±1.43 ± 15.87
1986-1987 300.00 - 10.17 14.48 28.92 353.57
± 17.29 - + 1.17 ± 1.49 ± 1.29 ± 15.90
Whole period 302.39 - 22.14 17.31 31.86 373.69
± 16.61 _ ±3.19 ±1.95 ± 1.35 ±15.63
bles 1 and 2 and the percentage contribution of L eaf litter com prised the largest portion (7 7 -
com ponent litterfall to the total litterfall in the 84% ) o f the total annual litterfall. N on-leaf plants
m ajor zones o f the tw o study sites are show n in (m angrove) litter accounts only for 2 -1 0 % o f the
Table 3. The average annual litterfall rates in the total litterfall in these m angrove ecosystem s.
tw o zones at K ala Oya estuary are higher than Four-weekly totals for the litterfall indicate a
those in the tw o m ajor zones at Erum athivu m an m arked seasonality in leaf, reproductive struc
gals. In both the m angals, the average litterfall tures and total litterfalls in the tw o m angals. In all
rates in the w aterfront (R . mucronata) zones the zones, the period o f high litterfall was during
were greater th an those in the back-m angrove M arch -A u g u st (with m inor variations), which
zones. coincides with the south w est m onsoon, generally
41
Table 3. Percentage contribution o f component litterfall to the total litterfall in the major zones of the mangals at Kala Oya es
tuary and Erumathivu island (R.m. - R. mucronata', A.m. - A. marina)',
the dry season for this area. T he period o f low ences in the m onth o f m axim um leaf fall. The
litterfall coincides with the north east m onsoon m agnitude o f litterfall o f the species is higher at
and the inter-m onsoon m onths are the wet K ala Oya than in the m angals at Erum athivu.
m onths o f the year (Figs 2, 3, 4 and 5).
V ariations in total and com ponent litterfall o f Annual above-ground woody biomass production
R . mucronata and A . marina in the tw o sites are
presented in Figs 6 and 7. In both the sites, leaf Average basal area increm ent, average woody
fall follows a similar pattern with m inor differ grow th per stem per year and total annual growth
total l i t t erf al l
eaves
OJ 1
e o r o a u c tiv
Parts
misc. + twigs
S O N D J F M A M J J A S O N D J F M A M J J A
85 86 87
Fig. 3. Variation in total litterfall and components in the mixed mangrove (back-mangrove) zone of the riverine/estuarine mangals
at Kala Oya estuary.
42
t o t a l litterfall
2
■>
CO
■O
CM
'E
Ul
£
n
’0 1
eaves
>
Q
repro.
Iparts I
m is c .+ tw ig s
85 86 87
Fig. 4. Variation in total litterfall and components in the in the Rhizophora mucronata zone of the fringing mangals at Erumathivu
island.
o f m angrove trees m easured in the m angals at tw o m ajor m angrove species, i.e., R . mucronata
K ala O ya estuary and Erum athivu island are and A . marina were greater at K ala O ya than at
given in Table 4. Erum athivu. In both sites, average w oody growth
T he average rate o f basal area increase and net o f the trees in the w ater front zone was higher
w oody above-ground production per stem for the than th at in the back-m angrove zone. A t K ala
Table 4. Average basal area increment (A) in cm2 per stem per year, average woody growth (B) in kg per stem per year and total
annual growth (C )in g m -2 y ~ 1 of mangrove trees measured at Kala Oya estuary (I) from July 1985 to M arch 1988 and Erumathivu
island (II) from July 1985 to October 1987.
Species A B C
I II I II I II
Water-front zone
R. mucronata 1.68 0.89 1.33 0.82 676.68 433.60
Back-mangrove zone
R. mucronata 1.66 - 1.24 - 210.35 -
A. marina 1.56 1.10 0.86 0.29 253.42 140.19
Other species 3.32 - 1.52 - 99.02 -
43
to ta l litterfall
2-
■>
CD
cm"0
'E
O)
-C
O)
'0
>
L_
Û
r e p r o . p<
misc. + twigs _
1 i 1----- 1-------- 1----- 1 j 1------- 1-------- 1------1------1-------1-------1---------1----- 1----- 1— “ i------- 1-------- 1----- 1-------- 1— i-----r
S O N D J F M A M J J A S O N D I F M A M J J A
85 86 87
Fig. 5. Variation in total litterfall and components in Avicennia marina zone of the fringing mangals at Erumathivu island.
Oya, A . marina had a smaller basal area increase Table 5. The N P P values for the tw o zones at
th an R . mucronata. A t Erum athivu, the trend was K ala O ya were greater th an those values calcu
reversed. In both sites, increase in w oody biom lated for the tw o zones at Erum athivu. A t each
ass o f R . mucronata w as greater than th at for site, the N P P for the w ater-front zone w as higher
A . marina trees. Although R . mucronata in the than th at for the back m angrove zone. Except in
mixed m angrove zone at K ala O ya h ad a higher the back-m angrove zone at Erum athivu, where
basal area increase and average w oody growth w oody growth accounted only for about 27%
th an A . marina in the sam e zone, its contribution of the N P P , in all other zones litterfall com
to total w oody grow th (210.35 g y r “ *) w as less prised nearly 50% o f the total N P P o f those m an
than th at by A . marina. gals.
The total w oody growth o f the trees in the
w ater-front zones at K ala O ya and Erum athivu
were higher th an those in the back-m angrove Discussion and conclusions
zones o f the tw o m angals.
Litterfall
Total above-ground NPP Am ong the different types o f m angals, the high
est litterfall values have been recorded from
T otal above-ground N P P values calculated for riverine/estuarine m angals (average 1100—
the m ajor zones o f the estuarine (K ala Oya) and 1200 g m “ 2 y r “ % and fringing m angals are re
fringing m angals (Erum athivu) are presented in ported to have a value o f 906 + 89 g m “ 2 y r “ 1
44
Leaves
b Reproductive Structures
-g 0-05 b. Stipules Ü 0-051
CM
'E
S N J M M J S N J M M J
d. Branch Wood 85 86 87
0-05-
Fig. 7. Variations in litterfall by components of Avicennia
marina at Kala Oya (--------- ) and Erumathivu (............ ).
S N J M M J S N J M M J
85 86 87 notable similarity o f these tw o m angrove areas is
Fig. 6. Variations in litterfall by components of Rhizophora th at they are dom inated by R . mucronata. F u r
mucronata at Kala Oya (--------- ) and Erumathivu ( - -.........). therm ore, the annual litterfall rates for A . marina
(at K ala Oya, about 356 g m -2 y r “ 1 and at E ru
m athivu, nearly 342 g m ~ 2 yr - x) are m ore com
(Twilley, 1982 cf. D ay et al., 1987). The average parable with those for the d w a r f s , marina stands
litterfall value obtained (average for both the in N ew Z ealand (W oodroffe, 1982) and A . nitida
zones), for the estuarine m angals at K ala Oya stands in South F lorida (S nedaker & Brown,
(about 588 g n C 2 y r _1) is m uch lower than the 1982), w hich are less than 1 m tail, in contrast to
average range o f values reported for the estuarine 3 -4 m tail stands in this study area. This could
m angals and this value is even lower than th at partly be due to the differences in freshw ater input
reported for the fringe m angals. Litterfall rate in and tidal am plitude. Pool et al. (1975) have sug
the fringing m angals at Erum athivu, i.e., gested th at the rate o f litterfall in m angroves is
407 g m “ 2 yr “ 1 is m uch below the average range related to increasing turnover o f freshw ater. Be
reported for fringe m angals (C intron et al., 1985). sides bringing nutrients required for continuous
N evertheless the total litterfall rate estim ated for productivity o f m angroves, within the m angal it
K ala O ya estuarine m angals (annual rainfall 100- self, w ater is im portant in bringing nutrients and
110 cm ) is com parable with th at o f the m angals oxygen to the ro o t system and rem oving C 0 2,
at Siar river estuary (annual rainfall 426 cm ) in toxic w astes and organic debris as well as in con
Saraw ak, E ast M alaysia (O thm an, 1988). The tinual m aintenance o f the soil-salt balance
45
Table 5. Estimated above-ground net primary production of mangrove species in the major zones of the estuarine mangals at Kala
Oya and the fringing mangals at Erumathivu island (percentage contribution of each species is given in parentheses).
A - R. mucronata zone; B - Mixed mangrove zone; C - A. marina zone.
A B A C
Litterfall (g m 2 yr ’)
R. mucronata 609.79 (46.9) 127.72(11.5) 424.38 (48.5) -
A. marina - 355.96 (31.9) - 341.83 (66.5)
Other species - 30.99 (2.8) - -
Miscellaneous 14.00(1.1) 37.82 (3.3) 16.58(1.9) 31.86 (6.2)
Total 623.79(48.0) 552.49 (49.5) 440.96 (50.4) 373.69 (72.7)
Average for mangal 588.14 407.33
(Snedaker, 1978). K ala O ya river system is ex ensue infrequent and insufficient tidal flushing
tensively utilized in the upstream areas for irriga- m ay also contribute to the decreased productiv
tional purposes. Eight reservoirs have been built ity o f these m angals. This also m ay explain the
in the upstream by dam m ing the river and the lower productivity in the back-m angrove zones
w ater stored in K ala W ewa, the largest reservoir than the w ater-front zones which subject to daily
in the system, is diverted to a num ber o f smaller flushing.
tan k s in the N o rth C entral Province to irrigate Leaves form the largest portion o f the total
paddy land. H ence the volum e o f freshw ater dis litterfall and hence constitutes a significant input
charged by the river m ay inevitably be low. Al o f organic m atter in both the m angals (Table 3).
though the nutrients in this have no t been qu an A lthough woodfall in these m angals appears to be
tified, due to low freshw ater inputs and associated unim portant, it m ay no t be the reality as dead
p o o r sedim ent tran sp o rtatio n into the estuary, trunk fall has no t been taken into account in the
it is plausible th at the nutrient influx is also present study.
low. Fringing m angals at Erum athivu, situated A general trend for peak litterfall to occur d u r
farthest from the rivers m ay receive even lower ing the rainy season has been reported by a num
input o f freshw ater and nutrients. Although the ber o f m angrove studies (H eald, 1971; Pool et al.,
elevation is sufficiently low, low tidal am plitude 1975; Leach & Bürgin, 1985; D ay et al., 1987).
46
& M. D. Correa (eds), The botany and natural history of Odum, W. E. & E. J. Heald, 1975. The detritus based food
Panama: La botanica e historia natural de Panam a IV se webs of an estuarine mangrove community. Estuar. Res., 1:
ries, Monographs in systematic botany, V 10, Missouri Bo 265-286.
tanical Gardens, St. Louis, Mo; USA: 53-56. Othman, S., 1988. The structure of vegetation and rate of litter
Day, J. W. Jr., W. H. Conner, F. Ley-lou, R. H. Day & N a production in a mangrove forest at Siar Beach, Lundu,
varro, 1987. The productivity and composition of mangrove Sarawak, East Malaysia. A paper presented at the Sym
forests, Laguna de Terminos, Mexico. Aquat. Bot. 27: 267- posium on Mangrove Management; Its Ecological and
284. Economic Considerations, Bogor, Indonesia.
Duke, N. C., J. S. Bunt & W. T. Williams, 1984. Observa Pool, D. J., A. E. Lugo & S. C. Snedaker, 1975. Litter pro
tions on the floral and vegetative phenologies of North- duction in mangrove forests in Southern Florida and Pu
Eastern Australia mangroves. Aust. J. Bot. 32: 87-99. erto Rico. In G. E. Walsh, S. C. Snedaker & H. J. Teas
Heald, E. J., 1971. The production of detritus in a South (eds), Proc. Int. Symp. Biol, and Management of M an
Florida estuary. Uni. Miami Sea G rant Bulletin, No. 6. groves, IFAS, Univ. Florida: 213-237.
Kanakaratne, M. D., W. K. T. Perera & B. U. S. Fernando, Rodin, L., N. I. Bazilavich & N. N. Rozov, 1975. Productiv
1983. An attempt at determining mangrove coverage in ity of the worlds main ecosystems, National Academy of
Puttalam lagoon, D utch bay and Portugal bay, Sri Lanka, Science, Washington D.C.
through remote sensing techniques. Proc. 4th Asian Con Snedaker, S. C., 1978. Mangroves: their value and perpetu
ference on Remote Sensing, Colombo: 1-9. ation. Nat. Res. 14: 6-13.
Leach, G. J. & S. Bürgin, 1985. Litter production and sea Snedaker, S. C. & M. S. Brown, 1982. Primary productivity
sonality of mangroves in Papua New Guinea. Aquat. Bot. of mangroves. In C. C. Black & A. Mitsui (eds), CRC
23: 215-224. H andbook of Biosolar Resources, 1: 57-70.
Lugo, A. E. & S. C. Snedaker, 1974. The ecology of man Teas, H. J., 1979. Silviculture with saline water. In A. Hol-
groves. Annu. Rev. Ecol. Syst. 5: 39-64. laender, J. C. Aller, E. Esptein, A. E. Pietro & O. R. Za-
Lugo, A. E. & S. C. Snedaker, 1975. Properties of mangrove borsky (eds), The Biosaline Concept. Plenum Press, New
forests in Southern Florida. In G. E. Walsh, S. C. Snedaker York: 117-161.
& H. J. Teas (eds), Proc. Int. Symp. Biol. Management of Turner, R. E., 1977. Intertidal vegetation and commercial
Mangroves, IFAS, Univ. Florida: 170-212. yields of penaeid shrimp. Trans, am. Fish. Soc. 106: 411-
Martosubroto, P. & N. Naamin, 1977. Relationship between 416.
tidal forests (mangroves) and commercial shrimp produc Woodroffe, C. D., 1982. Litter production and decomposition
tion in Indonesia. Marine Research in Indonesia, No. 18: in the New Zealand mangrove Avicennia marina var. resin
81-86. ifera. New Zealand J. mar. Freshwat. Res. 16: 179-188.
r<4¡i 1»
* z
Hydrobiologia 2 4 7 : 49-57, 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 49
© 1992 Kluwer Academic Publishers. Printed in Belgium.
Abstract
C hanges in coastline configuration and sand m ovem ents on Portuguese Island, a small island lying close
to In h aca Island 35 km east o f M aputo, M ozam bique have resulted in large areas o f the Island being
eroded aw ay and consequent restricted tidal recharge o f the m angrove. M easurem ents m ade during 1989
show th at flooding o f the m angrove occurred on only three occasions. Percentage m ortality o f the
d om inant species, Ceriops tagal (Perr.) C. B. R obinson., is high. How ever, at the tim e o f the study, the
co-dom inant, Bruguiera gymnorrhiza (L.) Lam ., show ed no signs o f stress. H alophytic conditions are
indicated by an invasive spread o f salt-tolerant herbs. All species of crustaceans and m olluscs have
disappeared from the m angrove.
A lthough m ortality o f C. tagal is high, the regeneration status o f this species is also high, com para
ble to th at in a regularly recharged m angrove on In h aca Island. It is suggested th at C. tagal propagules
were shed under conditions when seedling establishm ent w as favorable i.e. flooding. These seedlings m ay
represent the last cohort which m ay also suffer stress and death unless tidal recharge is re-established.
investigations related to ecophysiology, m angrove crab Uca urvillei (M . Edw .) is com m on along the
dynam ics or utilization patterns o f the M o zam b anks o f the canal. The central core o f the m an
bican m angroves have been carried out. grove is covered by a dense, mixed stand of
T he present study represents one o f several R. mucronata and Bruguiera gymnorrhiza (L.)
research projects related to m angrove ecology on Lam . In this zone the soils have a relatively high
In h aca Island. A prelim inary survey carried out organic m atter content (5 -6 .5 % ) and sesarm id
by the authors in 1988 noted the presence o f m or crabs are ab u ndant (Sesarma meinerti de M an.
ibund and dead trees in the m angrove on nearby and S. guttata M. Edw.). T he landw ard edge of
Portuguese Island w hich appeared to be related the m angrove com prises a less dense stand of
to im peded tidal recharge due to closure o f the smaller growing Ceriops tagal and Avicennia m a
canal feeding into the m angrove. A com parison of rina (F orsk.) Vierh. trees w ith occasional scat
aerial photographs (1958 to 1985) confirmed th at tered individuals o f Lumnitzera racemosa Willd.
m arked changes in coastline configuration had on the outerm ost fringes. T he soils o f this zone
taken place probably leading to this sequence of are drier and sandy with relatively low levels of
events. T he present paper represents the results organic m atter (0 .1 -1 % ). F iddler crabs {Uca
o f a m ore detailed investigation into the changes inversa (H offm an), U. annulipes (M . Edw .) and
in coastline configuration around Portuguese Is U. gaimardi (M . Edw .) are com m on faunal asso
land and the effects o f these changes on the m an ciates in this zone (M acN ae & K alk, 1962).
grove com m unity. The m angrove on Portuguese Island w as also,
originally, a creek m angrove. This m angrove is
m uch smaller than the P o n ta R asa m angrove and
Study area zonation patterns are less distinct. A dense stand
of R . mucronata lines the original course o f the
In h aca Island and nearby Portuguese Island are canal. T he outerm ost landw ard zone o f the m an
located 35 km east o f the city o f M aputo (see grove com prises a dense stand o f C. tagal and
m ap, Fig. 1). Tides at In h aca are sem i-diurnal B. gymnorrhiza. Only a few scattered individuals
with an average am plitude o f 3.3 m for springs of A. marina are found in the latter zone whilst
and 1.5 m for neaps. Average annual tem perature L . racemosa is absent altogether. The diverse
is 22.9 °C with m axim um m ean m onthly tem per m angrove fauna o f crustaceans and m olluscs
ature in F ebruary (26.2 °C ) and m ean m onthly found in the m angrove o f P o n ta R asa are absent
m inim um in July (19.2 °C ). Average annual rain from the P ortuguese Island m angrove although
fall is 874 m m w ith a distinct wet season during rem ains o f these anim als are present providing
O ctober to April and a dry season during M ay to evidence o f their recent disappearance.
Septem ber. E vaporation exceeds rainfall for all
m onths except January and February.
F our broad vegetation types m ay be distin Methods
guished: in tact forest, agricultural landscape with
scattered trees, sw am ps and m angroves (Fig. 2). Changes in coastline configuration were deter
The m ain study w as carried out in the m angrove m ined using aerial photographs (1:15 000, D ep art
on Portuguese Island with a com parative study m ent o f G eographic Surveys, M aputo) for 1958,
(‘no n -stressed ’ m angrove) being m ade in the 1967, 1985 and 1989. Changes in area o f m an
m angrove at P o n ta R asa, In h aca Island. grove and associated vegetation types were cal
The P o n ta R asa m angrove is a typical creek culated from enlarged ( x 4) images o f these pho
m angrove approxim ately 1 km in length running tographs. Extensive ground truthing to support
in a north-south direction. The banks o f the canal the photo-interpretation w as carried out in 1989.
leading into the m angrove are lined w ith a dense T he frequency and extent o f the tidal recharge
stand o f Rhizophora mucronata Lam . The fiddler during 1989 for the Portuguese Island m angrove
51
20 "
500 IOCOKm
Indian Ocean
15Km
MAPUTO PORTUGUESE
ISLAND
INHACA r
ISLAND /
M aputo B ay
Fig. 1. The geographical setting of Inhaca Island and Portuguese Island. Inset shows the location of M aputo in southern Africa.
w as determ ined by daily m onitoring. W hen flood D ensity (num ber o f individuals per ha.), % fre
ing occurred, the extent w as m easured by m ap quency, % m ortality and regeneration (num ber of
ping on the ground with reference to perm anent seedlings per m 2) o f tree species were m easured
‘m arker’ trees in the m angrove. in nine 5 x 5 m replicate quadrats in flooded and
COMMUNITY AREA (ha)
Ponta Torres
ED Forest 945.8
Fig. 2. Location of study sites (arrows) on Portuguese Island and Inhaca Island (Ponta Rasa). Also shown are the distribution
of the four main vegetation types.
53
2 Km
1958 1967
1985 1989
Fig. 3. Changes in coastline configuration, Portuguese Island between 1958 and 1989 (continuous line). The original 1958 con
tour is shown for comparison for 1967, 1985 and 1989 (dotted line). Shaded areas indicate mangrove. Arrow (1958) indicates
original entrance to canal. Arrows (1989) indicate areas where new sand banks are forming.
Table 1. Surface area changes (m2) of Portuguese Island and four days in Septem ber. A t these times (corre
mangrove. sponding approxim ately to the equinoxes) the
A rea (m2) Area (m2)
tidal am plitudes were sufficient to breach the
Portuguese Island Mangrove newly-form ed beach ridges on the northern shore
separating the m angrove from the sea.
1958 3027000 110150 A lthough tidal recharge o f the m angrove took
1967 2619000 73860
place, the reduced strength o f the tidal w ash re
1985 1519000 28430
1989 2161000 26240 sulted in only partial im m ersion o f the m angrove.
F looded and non-flooded areas o f the m angrove
corresponded to 72% and 28% respectively.
also been eroded away, the canal leading into the
m angrove h as been closed and the rem aining Vegetation
m angrove has becom e isolated from direct access
to the sea by the form ation o f beach ridges. The density (num ber o f individuals per ha.), fre
In co n trast to these erosion patterns, a new quency (% ), m ortality (% ) and regeneration
peninsula h ad form ed in the north east o f the (num ber o f seedlings per m 2) for Ceriops tagal,
Islan d by 1985, the process continuing until 1989. Bruguiera gymnorrhiza, Avicennia marina and
The prograding nature o f this peninsula is high Rhizophora mucronata in the Portuguese Island
lighted by the form ation o f sand banks in an area m angrove is given in Table 2. A com parison is
around the northern shore, the erosion o f this m ade between flooded and non-flooded areas of
shore apparently having halted by 1989 (Fig. 3d). the m angrove.
Table 1 shows changes in area o f Portuguese C. tagal and B. gymnorrhiza are by far the m ost
Island and associated m angrove. im portant tree species o f the m angrove. A . marina
and R . mucronata occur at low densities and fre
Tidal exchange quencies in both flooded and non-flooded areas.
M ortality o f C. tagal is highest with a significantly
D uring 1989 the Portuguese Island m angrove ex higher m ortality o f individuals o f this species in
perienced only three periods o f tidal flooding: the non-flooded areas. M ortality o f the other spe
three days in February, five days in M arch and cies is low. H ow ever, regeneration o f C. tagal is
Table 2. Density, frequency, mortality and regeneration of Ceriops tagal, Brugieria gymnorrhiza, Avicennia marina and Rhizophora
mucronata in flooded and non-flooded areas in the mangrove o f Portuguese Island. Standard errors about the mean are given in
brackets. Differences of significance are between flooded and non-flooded areas.
Flooded Non-flooded Sign Flooded Non-flooded Sign Flooded Non-flooded Sign Flooded Non-flooded Sign
C. tagal 22444 19777 n.s. 76.2 72.5 n.s. 17.9 56.0 *** 42 28 n.s.
(± 3 2 3 0 ) (± 6387) (± 5 .4 ) (± 5 .8 ) (± 4 .9 ) (± 6 .2 ) (± 8 ) (± 6)
B. gymnorrhiza 5933 4022 n.s. 19.8 14.6 n.s. 0.7 0 n.s. 12 12 n.s.
( ± 1869) (± 1126) (± 5 .4 ) (± 4 .4 ) (± 2 .1 6 ) (± 8 ) (± 3 )
A. marina 133 555 n.s. 0.1 9.8 n.s. 0 5.8 n.s. 0 0 n.s.
(± 9 5 ) ( ± 385) (± 0 .7 ) ( ± 8 .1 ) (± 5 .8 )
R. mucronata 622 311 n.s. 3.04 0.55 n.s. 0 7.9 n.s. 0 0 n.s.
( ± 327) (± 3 1 1 ) (± 1 -5 ) (± 0 .5 5 ) (± 7 .9 )
Table 3. Percentage cover of herbaceous species in the Por Sporobolus virginicus (L.) K unth. C over d ata (% )
tuguese Island mangrove. Differences of significace are be for these three species in flooded and non-flooded
tween floded and non-flooded areas (*** = p < 0.001). Stan
areas is given in Table 3. F o r all three species
dard errors about the mean are given in brackets (n = 9).
percent cover w as significantly higher in non-
Cover (% ) Sign. flooded areas than in flooded areas.
A com parison o f m ortality and the regenera
Flooded Non-flooded
tion status o f the tw o dom inant species, C. tagal
Sesuvium portulacrastum 5.5 ( + 3.6) 34.4 (± 7 .6 ) Hi** and B. gymnorrhiza, in the ‘stressed’ (restricted
Sporobolus virginicus 5.5 (± 5 .5 ) 33.3 (± 5 .8 ) *** tidal recharge) m angrove o f Portuguese Island
Chenolea diffusa 0 19.4 (± 5 .5 ) *** and ‘non-stressed’ (diurnal tidal recharge) m an
grove on In h aca Island (P onta R asa) is given in
Figs. 4a and b. F o r the P o n ta R asa m angrove
highest followed by B. gymnorrhiza; no differences d ata is given for both a Ceriops zone and a Bru
between flooded and non-flooded areas could be guiera zone. Although, as noted, m ortality of
detected. C. tagal is higher in the non-flooded area as com
Three invasive herbaceous species were noted pared to the flooded areas in the m angrove o f
in the Portuguese Island mangrove: Sesuvium Portuguese Island (Table 2, Fig. 4a), m ortality in
portulacastrum L., Chenolea diffusa Thunb. and b oth areas is higher than in the Ceriops or Bru
guiera zones o f the m angrove at P o n ta R asa. The
regeneration status o f C. tagal on Portuguese Is
land is similar to th at found in the Ceriops zone
at P o n ta R asa.
jC eriops ta g a l
Discussion
B ruguiera g y m n o rrh iz a
£30
1
2 20 It is evident th at between 1958 and 1985 a strong
a ) M o rtality
current moving in an arc ground the north of
10- Portuguese Island led to the erosion o f m uch of
J± L this shoreline. In the process large areas o f m an
grove have been w ashed aw ay with a loss o f an
area m ore than 75% o f the 1958 cover (Fig. 3,
Table 1). Between 1985 and 1989 new tidal forces
b) Regeneration appear to be operating resulting in deposition o f
sand at various locations around Portuguese Is
land. O ne o f the consequences o f this deposition
S1^ of sand is th at the Portuguese Island m angrove is
20
isolated by beach ridges which are only rarely
breached by tidal wash. Tinley (1985) has pro
posed th at a continental shelf counter-current,
moving in a direction opposite to the deep w ater
Non flo o d e d C eriops z o n e B ru g u iera z o n e
P o rtu g u e s e Islan d P o n ta R a s a
M ozam bique C urrent, tran sp o rts sand n o rth
w ards along m uch o f the M ozam bican coast. It
Fig. 4. Percentage mortality (Fig. 4a) and regeneration also appears from a study o f aerial photographs
(Fig. 4b) of C. tagal and B. gymnorrhiza in flooded and non-
th at the northern-m ost point o f Inhaca Island
flooded areas of the Portuguese Island mangrove and in Ce
riops and Bruguiera zones of the Ponta Rasa mangrove (In acts as a fulcrum, sweeping sand around this point
haca Island). Standard errors about the mean are shown. w estw ards tow ards Portuguese Island. A sand
56
b ar now extends betw een the tw o Islands and tion o f tidal exchange and im poundm ent of fresh
around the northern shore o f Portuguese Island w ater. N o im poundm ent o f fresh w ater w as o b
(Fig. 3d). These processes have probably resulted served in the Portuguese Island m angrove al
in the prograding beaches presently being form ed though it is possible th at the m angrove suffered a
on Portuguese Island and consequent im pound short-term im poundm ent o f fresh w ater due to
m ent o f the m angrove. rainfall w hich m ay have led to the death o f C. ta
W hatever the m echanism s leading to im pound gal. A nother cause o f die-back could be due to
m ent, by 1989 the m angrove was only occasion increasing salinity in the m angrove. The hydro-
ally flooded by tidal w ater and, because o f re logical and clim atic conditions on Portuguese Is
duced tidal flow, only partial im m ersion o f the land (little fresh w ater runoff, low rainfall and
m angrove occurred. Similar im poundm ents and high evaporation) w ould probably prom ote salin
restricted tidal exchanges have been reported by ization o f the soils. A n increase in salinity is fur
Breen & Hill (1968) for the K osi Estuary in South ther indicated by the invasive spread o f halophytic
Africa, 120 km south o f In h aca Island, and by herbs in areas o f the m angrove no t flooded by the
G o rd o n (1987) for the arid D am pier-C ossack re occasional tidal recharge. G o rd o n (1988) reports
gion o f W estern A ustralia leading, in both cases, th at salinity levels o f 90%o were associated with
to m ass m ortality o f m angrove trees. death o f C. tagal and other species in W estern
In the case o f the Portuguese Island m angrove A ustralia. Sm ith (1988), in laboratory grow th ex
the dom inant species, Ceriops tagal, appears to be perim ents, dem onstrated th at salinity levels of
m ost affected by restricted tidal exchange. M o r 60%o resulted in death o f C. tagal. T he salinity
tality rates o f C. tagal are m uch higher than indi and ground w ater conditions in the Portuguese
viduals o f the sam e species receiving regular ex Island m angrove are presently being studied. The
changes o f tidal w ash at the P o n ta R asa m angrove disappearance o f the once diverse fauna m ay also
on In h aca Island (Fig. 4a). Individuals o f play a role; M acN ae & K alk (1969) point out the
Bruguiera gymnorrhiza, a codom inant, do not ap close association betw een sesarm id crabs and the
pear to suffer the sam e consequences. N o exact roots o f C. tagal which enhance aeration o f the
d ata as to the im poundm ent o f the m angrove is ro o t system.
available, but it probably occurred some time dur Som ew hat contradictory is the observation
ing 1988. C. tagal, therefore, appears to be m ost that, although C. tagal has the highest m ortality
susceptible to the new adverse conditions. This is rates, regeneration status is com parable with that
further supported by the fact th at extrem e and o f the P o n ta R asa m angrove on In h aca Island
less-extrem e conditions exist in the m angrove i.e. (Fig. 4.b). This m ay be explained by C. tagal
non-flooded and occasionally-flooded areas. shedding its propagules at a time w hen conditions
D eath o f C. tagal is significantly higher in the in the im pounded m angrove were suitable for
non-flooded areas (Table 2). The susceptibility of generation viz. sufficient w ater for rooting and
C. tagal m ay be due to several factors. Along the germ ination. Propagules o f C. tagal and B. gym
E ast African coast, C. tagal is alm ost at its norrhiza on In h aca Island begin germ ination in
southern-m ost distributional limit (C hapm an, January and are shed in late January and F ebru
1976) and any environm ental shift w ould have a ary at about the sam e tim e as one o f the last tidal
proportionally larger effect on the growth o f this floodings o f the Portuguese Island m angrove
species. H utchings & Saenger (1987) report that (F ebruary and M arch 1989). D evelopm ent of
the root system o f C. tagal is particularly suscep these seedlings is now being studied and the re
tible to sedim entation by sand, a process which sults will be presented in a later paper.
m ay be occurring in the m angrove. The w ork of The future o f the Portuguese Island m angrove
Breen & Hill (1968) at K osi Bay, where C. tagal will depend on the build-up o f beach ridges
is rare, showed th at all m angrove species, includ around the m angrove. The rapid death o f C. tagal
ing C. tagal suffered death after short term restric m ay be followed by increased m ortality o f the
57
co-dom inant, B. gymnorrhiza. The new ly-estab groves in the Kosi Estuary. Trans, r. Soc. S. Afr., 38:
285-303.
lished seedlings o f these tw o species will certainly
Cott, H. B., 1929. Observation on the natural history of the
require a regular exchange of sea w ater if they are land crabs Sesarma meinerti, with special reference to the
to flourish. W ard ei'al. (1986) have show n that theory of warning colours. Proc. Zool. Soc. Lond. 18: 679-
re-establishm ent o f all m angrove species is pos 692.
sible even after m ass m ortality due to im pound Chapman, V. J., 1976. Mangrove Vegetation. Cramer, Vaduz.
Hutchings, P. & P. Saenger, 1987. Mangrove Ecology. St.
m ent providing th at physical factors preventing
Lucia, Queensland: University of Queensland Press.
tidal recharge are rem oved (in this case a sand bar Jacques, J. M., 1975. Aperçu sur un biocerose cohere; la man
blocking the m outh o f the K osi Bay estuary south grove. N at. Belg. 56: 45-50.
o f In h aca Island). If no natural tidal recharge of Gordon, D. M., 1988. Disturbance to mangroves in tropical-
the Portuguese Island m angrove occurs in the arid Western Australia: hypersalinity and restricted tidal
exchanges as factors leading to mortality. J. Arid Envir. 15:
n ear future and deleterious trends continue, then
117-145.
m anagem ent options such as artificially creating Kalk, M., 1959. The fauna of the shores of Northern M ocam
a canal m ay have to be considered. bique. Rev. Biol. Lisbon, 2, 1.
M acNae, W. & M. Kalk, 1962. The ecology of the mangrove
swamps at Inhaca Island, Mocambique. J. Ecol. 50: 19-34.
MacNae, W. & M. Kalk, 1969. A Natural History of Inhaca
Acknowledgements Island, Mocambique. W itwatersrand University Press, Jo
hannesburg.
The authors gratefully acknowledge financial sup Mogg, A., 1963. A preliminary investigation of the signifi
p o rt by the N orw egian O rganization for R esearch cance of salinity in the zonation of species in salt-marsh and
mangrove swamp associations. S. Afr. J. Bot. 59: 81-86.
and D evelopm ent for this project. W e thank all
Sim, T. R., 1910. The flora of Portuguese East Africa. S. Afr.
m em bers o f staff o f the M arine Biological R e J. Sei. 6: 294-300.
search Station, In h aca Island for their logistic Smith, T. J., 1988. Differential distribution between subspe
support throughout the period o f this work. cies of the mangrove Ceriops tagal: competitive interactions
along a salinity gradient. Aquat. Bot. 32: 78-89.
Tinley, K., 1985. Coastal Dunes of South Africa. South Af
rican National Scientific Programmes Report No 133.
References CSIR, Pretoria, 193 pp.
W ard, C. J., T. D. Steinke & M. C. W ard, 1986. Mangroves
Bennessalah, D., 1988. Manual on Mapping and Inventory of of the Kosi System, South Africa: Their re-establishment
Mangroves. FAO, Rome. since mass mortality in 1965/66. S. Afr. J. Bot. 52: 501-
Breen, C. M. & B. J. Hill, 1968. A mass mortality of man 512.
.
Hydrobiologia 2 4 7 : 59-75, 1992.
V. Mccanni & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 59
© 1992 Kluwer Academic Publishers. Printed in Belgium.
Abstract
The seagrass and m acroalgal vegetation o f G azi Bay (at approxim ately 50 km south o f M om basa) have
been studied by m eans o f 88 relevés along 7 transects. C orrelation betw een the distribution o f the
seagrasses and som e abiotic factors (particle size fractions, chem ical com position o f the substrate) is
not well m arked. N evertheless a general zonation and succession o f seagrasses could be established:
1. a transition zone between the m angal and the seagrass beds is covered by Boodleopsis pusilla',
2. the pioneer association Halophila ovalis + Halodule wrightii form s low sandy bum ps at the upper limit
o f the seagrass beds, but also occurs in the whole m idlittoral w here sandlayers have recently been
accum ulated (e.g. on coral platform s);
3. the climax vegetation o f the intertidal zone seems to be Thalassia hemprichii which som etim es is
associated with Cymodocea rotundata and C. serrulata, certainly in deeper pools and close to low
w ater m ark; Halimeda opuntia, Gracilaria salicornia and G. corticata are also frequent in this vegeta
tion type;
4. from low w ater at neap tide dow nw ards patches o f m onospecific Enhalus acoroides vegetation can
also occur;
5. from m ean low w ater dow n to approxim ately - 1 m mixed m eadow s o f Thalassia, C. serrulata, C. ro
tundata and Halodule uninervis are well developed; the seaweeds Halimeda macroloba and Avrainvillea
obscura are also typical for this zone; locally patches o f Syringodium isoetifolium grow on small bum ps
and Halophila stipulacea grows as a pioneer on bare sand;
6. from - 1 m dow nw ards the whole lagoon is covered by hom ogeneous, m onospecific Thalassodendron
ciliatum m eadow s, locally replaced by E. acoroides.
w aters and surface sedim ents and are therefore veloped from the lower m argin o f the m angal
critical plants in the control o f w ater quality of dow n to the subtidal zone. They cover the whole
shallow w aters (Patriquin, 1972). lower basin (lagoon) o f G azi Bay, stretching up
In G azi Bay the large quantities o f drifting sea to the fringing reef. Only a few sandy patches
grasses (especially o f Thalassodendron ciliatum) (G azi Beach) or fossil coral outcrops do not have
and seaweeds (Sargassum sp.) brought into the seagrasses. T he com plete area is subhorizontal,
m angrove at each high tide (certainly in some slightly sloping tow ards the tide channel and the
periods of the year) might account for a m ajor shallow lagoon. A t neap low tide m ost o f the area
im port o f organic m aterial in th at ecosystem. rem ains inundated; at spring low tide vast areas
N ext to these ecological roles o f the seagrass ( = the infralittoral fringe) are em ergent except for
ecosystem , the econom ical uses o f these m arine the deepest p a rt o f the tide channel and the la
angiosperm s are em phasized by different authors: goon.
production o f paper because o f the high cellulose/ The substrate is very diverse because of sedi
low lignin co ntent o f the leaves; young rhizom es m entation p atterns (C oppejans & Gallin, 1989:
as well as fruits o f some species are eaten in dif 49; O nyango, 1989): locally it is rather coarse
ferent parts o f the w orld; dried seagrass leaves sand, on other places one sinks knee-deep in silty
mixed with terrestrial plant m aterial form an ex sand. Large areas are characterized by closely
cellent fodder; seeds o f Zostera are ground as packed cone-like bum ps, ± 0,5 m high, + 1 m in
flour in Sonora, M exico; seagrasses like Thalassia diam eter, enclosing small tide pools. The bum ps
are excellent for green m anuring. are built up by still unknow n anim als which reg
In G azi Bay several biological research p ro ularly extrude sedim ents from a central crater.
gram s are carried out: Floristics, Faunistics and The solid sandy bum ps are devoid o f seagrasses,
Ecology o f K enyan coastal biotopes (F K F O the pools contain a mixed seagrass/seaw eed veg
2.0043.88: 1988-1991), D ynam ics and assess etation. This diversity in environm ental condi
m ent o f K enyan m angrove ecosystem s (E E C tions results in a zonation and patchiness of the
project TS2-0240-C (G D F ): 1989-1992), Sys- seagrass vegetation.
tem atics, Ecology and Biogeography o f M arine
O rganism s in the Indian O cean (F K F O
2.0009.92: 1992-1995). Sampling
The purpose o f this paper is to give a descrip
tion o f the seagrass vegetation. It is hoped to be Field w ork was concentrated in July and August
a m ain fram ework for m ore detailed studies of 1987 (D e W it, 1988). A dditional observations
this ecosystem. have been m ade in August 1989. Seven transects
through the seagrass beds have been studied d u r
ing spring low tides (Fig. 1). Five o f them (3 -7 )
Material and methods
are at right angles to the coastline, starting at the
Description o f the site low erm ost pneum atophores o f Sonneratia
(C oppejans & Gallin, 1989; G allin eta l., 1989)
A m ap o f K enya with the localisation o f G azi is going dow n to the rather hom ogeneous subtidal
given in C oppejans & Beeckm an (1989: 382); a m eadow s. The vegetation o f the tide channel was
general description and m ap o f G azi Bay ( = M af- not studied here. A detailed study o f it has been
tah a Bay), approxim ately 50 km south o f M om carried out by D e P auw (1990). Each transect
basa, in C oppejans & G allin (1989: 4 8-49, w as subdivided in hom ogeneous zones defined by
Fig. 1); clim atologieal d ata and m ean salinity val the presence/absence of seagrass species, visible
ues are sum m arized in G allin et al. (1989: 198). changes o f substrate an d /o r relief. In total 88
Except for some restricted seagrass grow th in relevés (1 x 1 m 2) were m ade in the different
m angal tide pools the seagrass beds are best de zones.
61
fie ld la b
rV man made c a n a i 5 0 0
h a rb o u r'
m an g ro v e
^ w a te r l e v e l a t
••sand e x tr e m e low t i d e
’bankyv'V-.
/Ya muddy s a n d
d ead c o r a l p la tfo rm
sandy i n t e r t i d a l zone
la r g e ly covered by se a g ra sse s
W
-vCU iiv
<
f i s h e r m e n 1s ^ ~ V« •*> ------- -
hut r¿ VC ~ ^rC'r-
W - ~ ~ ~ ~ rJ ~ ^
^ ~ ~ t" ~ ~ ~ ~
V» ~ ~ ~ iV
„ , _. ^«j rv/
~^
<v « j «J cu cv ™
The presence and cover o f the seagrasses as T W IN S P A N (Hill, 1979). T he tw o-w ay table has
well as o f the m acroalgae were recorded using the been m anually rearranged in order to show the
Londo-scale (L ondo, 1984). Specim ens which m ain gradient in the vegetation. A p art o f the
were unidentifiable on the field (mainly sea m atrix, consisting o f 32 taxa and 45 relevés, is
w eeds) were collected and either pickled in 4% com pared to another m atrix with inform ation on
form alin/seaw ater or prepared as herbarium 15 environm ental variables for the sam e 45 sam
specim ens which are deposited in the H erbarium ple units by the algorithm C anonical C orrespon
U niversitatis G andavensis (G E N T ) under H E C - dence A nalysis (C C A ) (C A N O C O , ter B raak,
num bers (H erbarium Eric C oppejans); duplicates 1988).
are present in the K enya M arine and Fisheries The involved environm ental param eters are:
R esearch Institute (K .M .F .R .I., M om basa, the distance (m) o f the plots to the starting point
Kenya). o f each tran sect ( = low w ater level), the concen
Salinity and w ater tem perature o f the tidal trations o f N a, K, C a, M g and P (m g/kg sub
pools along the transects 1, 3 and 4 were m ea strate), the quantity o f organic m aterial (% ), the
sured on the field with a com bined conductivity/ p H o f the substrate, the proportion o f seven p ar
therm om eter. Sedim ent sam ples were also taken ticle size fractions (% ).
along som e o f the transects. C anonical C orrespondence Analysis can be
considered as a m ultidim ensional direct gradient
analysis into which regression and ordination
Analysis o f the samples have been integrated (ter B raak, 1986, 1987;
B eeckm an, 1989). C anonical or constrained or
The unidentified seaweed specim ens were an a dination is based on H utch in so n ’s concept o f the
lyzed in the laboratory for P lant M orphology, niche (H utchinson, 1944, 1957, 1978; Schoener,
System atics and Ecology. T he chem ical soil fea 1989), which w as first introduced in a footnote to
tures were determ ined in the L aboratory for A n a 1944 paper in Ecology on phytoplankton peri
alytical and Agrochem istry: the organic substance odicity: ‘the term niche is here defined as the sum
by the m ethod o f W alkley and Black (1934); the o f all the environm ental factors acting on the o r
concentration o f C a, K and N a by flame p ho to m ganism ; the niche thus defined is a region o f an
etry, the concentration o f M g by atom ic absorp n-dim ensional hyper-space, com parable to the
tion and the concentration o f P by the colori phase-space o f statistical m echanics’ (H utchin
metric m ethod o f Scheel (C ottenie e ta l., 1982: son, 1944: 20). U sing this concept, C anonical
43). The texture o f the soil w as analyzed by the C orrespondence A nalysis is an appropriate
pipette m ethod in the L aboratory for Agronom ic m ethod to study ecological complexity and inter
Soil Science (for m ore details: D e Wit, 1988: 2 4 - actions (B eeckm an & Lem eur, 1989). T he general
29). aim o f C CA , w hen used in an exploratory way, is
to turn out an ordination diagram o f sam ples,
species and environm ental variables w hich opti
Numerical analysis mally displays how com m unity com position v ar
ies as a function o f environm ent. Biplot scores of
The d ata on benthic plant species and on abiotic environm ental variables give the coordinates of
factors were subm itted to m ultidim ensional n u the heads o f arrow s in the ordination diagram .
merical analysis (Legendre P. & L., 1984a, b). The rules for constructing and interpreting
The d ata were checked for possible clusterings species-environm ent biplots are the sam e as those
betw een seagrasses an d /o r seaweeds and for co r given in ter B raak (1987: 127-129) for P C A bi
relations betw een biotic and abiotic com ponents. plots: the direction o f the arrow indicates the di
Therefore a m atrix o f 37 species and 88 relevés rection in w hich the am ount o f the corresponding
w as subm itted to the F O R T R A N 77 program environm ental variable increases m ost; the length
63
o f the arrow equals the rate o f change in th at lected in G azi Bay by Isaac (1968: 32) has not
direction; the angle betw een arrow s o f a pair of been observed again.
variables (environm ental, species or sam ples) Following the m orphological classification sys
provides an approxim ation o f their pair-wise cor tem o f the seagrasses o f den H artog (1967, 1977:
relation. The biplot is constructed m ost easily by 92) the seagrasses from G azi Bay can be classi
draw ing separate plots o f species, sam ples and fied in six categories o f grow th form s which can
environm ental variables on tran sp a ran t paper, be linked with differences in ecology.
each one with its own scaling. How ever, within The parvozosterids (e.g. Halodule) with fine,
each plot the scale units o f the axes m ust have linear leaves; they can be found in all kinds o f
equal physical length. The biplot is obtained by habitats, from coarse sand to alm ost liquid m ud,
superim posing the plots with the axes aligned and from the intertidal zone dow n to considerable
the origins o f the coordinate systems coinciding depth, from open sea to the estuaries and m an
(ter B raak, 1988). grove swam ps.
T he m agnozosterids (e.g. Cymodocea, Thalas
sia) with wide linear leaves; they also grow on
Results various substrata, bu t they are typically subtidal.
The syringodiids (e.g. Syringodium) with long sub
Species list o f the plants ulate leaves; they do tolerate some em ersion, but
are usually restricted to the upper sublittoral.
Species o f plants listed here are exclusively from The enhalids (e.g. Enhalus) with leathery strap
the transects through the intertidal and subtidal shaped leaves; they are restricted to the ‘upper’
seagrass beds and partly sand-covered coral p lat p art o f the sublittoral and inhabit sandy and rocky
form. Large intertidal pools with coral boulders substrate, being rarely found on m ud.
an d /o r dead tree trunks as well as the tide chan The halophilids (e.g. Halophila) with delicate,
nel have a m uch m ore diversified algal flora; these fast growing rhizom es and elliptic, ovate or lan
species are n o t included here. ceolate delicate leaves; as with the parvo
zosterids they have a very wide ecological range.
Seagrasses (Helotiales) T he am phibolids (e.g. Thalassodendron) with
D escriptions and illustrations are to be found in: sym podially branched vegetative system, the
B rouns & Heys (1986), den H artog (1970), Isaac leaves being distichously arranged along the up
F. (1968), Phillips & Mefiez (1988), M oorjani & right stem s; as the enhalids they are strictly sub
Sim pson (1988). They have been sum m arized in tidal, euhaline, inhabiting sandy and rocky sub
D e W it (1 9 8 8 :4 6 -55). Eleven o f the twelve know n strates (dow n to - 3 0 m for T. ciliatum.
seagrass species from K enya (Isaac F., 1968) have
been found in the study area: Cymodocea Seaweeds
rotundata A sch. & Schweig. (H E C 8477), C. ser D escriptions and illustrations o f m ost o f the spe
rulata (B raun) Asch. & M agnus (H E C 8478), cies can be found in Jaa su n d (1976) and in M o o r
Enhalus acoroides (L.f.) Royle (H E C 8506), jani & Sim pson (1988).
Halodule uninervis (F orssk.) A sch. (H E C 8476),
H. wrightii A sch. (H E C 8475), Halophila minor Chlorophyta
(Zoll.) den H artog (H E C 8471b), H. ovalis Avrainvillea obscura J. Ag. (H E C 6069),
(B raun) H ooker (H E C 8503), H. stipulacea Boodleopsis pusilla (Coli.) Taylor (H E C 6741),
(F o rssk .) Asch. (H E C 8472) Syringodium Bryopsis hypnoides (H E C 6903), Caulerpa
isoetifolium (A sch.) D andy (H E C 8502), Thalassia sertularioides (G m el.) H ow e (H E C 6076),
hemprichii (E hrenb.) Asch. (H E C 8479), Chaetomorpha crassa (Ag.) Kiltz., Cladophora
Thalassodendron ciliatum (F orssk.) den H artog patentiramea (M ont.) Ktitz. f. longiarticulata
(H E C 8480). Only Zostera capensis Setch., col Reinb. (H E C 6840), Enteromorpha clathrata
64
present in them should be com ing from the sea 200), 200-100 and 100-50 p m fractions, in
w ater. B ased on the K /M g relation it can be con transect 5 to the 500-200, 200-100 and 100—
cluded (D e Wit, 1988; Table 10 p. 34) th at in 50 p m fractions.
transects 1 and 2 the quantity o f M g is also the A ccording to the texture triangle o f soil taxon
sam e as in seawater. In transects 3, 4, 5 there is omy (FA O 1977) the sam ples can therefore be
a m arked enrichm ent, up to 5 times. described as sand. G rady (1981) found 82-9 9 %
The am ount of calcium is considerably higher sand in seagrass flat sedim ents in Florida.
than in the seaw ater (up to 275 tim es) in the
sam ples o f all transects. A ccording to Patriquin
(1975) the C aC o 3 fraction o f seagrass substrate T W IN S P A N classification
is from biotic origin, com ing from the calcified
The 88 relevés are split up in 8 groups:
Halimeda (Chlorophyta), the Corallinaceae
(R hodophyta), corals, foram inifera and gastro - T W IN -group 1 with Boodleopsis pusilla as in
pods. dicator species, characteristic for the high in
The clay and loam fractions o f the soil sam ples tertidal zone (between the highest seagrasses
along the transects 1 -5 are generally less than and the low erm ost m angroves);
0,5% . The am ount o f very coarse sand is also - TW IN -group 2 with Enteromorpha flexuosa
very limited. The proportion o f coarse, medium and E. ramulosa as indicator species, charac
coarse, m edium fine and fine sand varies from one teristic on m uddy substrates in the intertidal
sam pling spot to the other (D e W it, 1988, Table 8, zone;
p. 32), even within a transect. In transects 1-2 - T W IN -group 3 with Laurencia papillosa as in
they mainly belong to the 500-200 and 2 0 0 - dicator species, characteristic for thinly sand-
100 p.m fractions, in transects 3 -4 to the (5 0 0 - covered coral platform in the intertidal zone;
B oo d le op sis pus i 1 la 01 00
8
011 010
H alophila s t i p u l a c e a
0101 0100
Enteromorpha spp.
01011 01010
Thalassondendron c il i a t u m
Fig. 2. TW INSPAN-dendrogram of the relevés indicating the subsequent splitting levels of the TW IN-groups and the indicator
species.
66
Table 2. Two way ordered TW INSPAN-table after manual rearrangement. The significance of the name codes is given after ta
ble 7. [1 = < 1%; 2 = l-2 % ;3 = 2 -5 % ; 4 = 5 -10% ; 6 = 30-50% ; 7 = 50-70% ; 8 = 70-90% ; 9 = > 9 0 % ].
12 7 5 6 6 7 7 7 7 1 1 1 1 1 2 2 2 2 3 4 4 4 4 4 4 4 5 5 6 6 6 6 7 7 7 7 8
9 0 12 3 0 9 8 9 1 2 3 5 0 3 4 6 7 1 2 5 9 3 0 1 2 3 4 5 8 7 8 0 3 6 7 4 6 7 8 0 1 2 3 4 5 6 7
BOO PUS 2 3
ENT FLE . 1 . 673
ENT RAM 9 6 7 .
ENT CLA . . . 3
EN T SPE .11.
ULV LAC . . . 3
LAU PAP 1 . 4 5 5
PAD BOR 5 5 . 2 .
ACA SPI 1 2 . . .
CHA CRA 1. 1
GRA SAL 1
GRA ED U
HYP N ID
CLA PAT
HAL OVA 2. 3. 2 2 2 2 3 2 2 3 . 2 3 5 2 . 2. 2 2 2 . 2 3 5 5 . . 2 2 5 . 255
HAL WRI 2 2 2 2 2 2 2 3 3 2 3 5 5 2 3 2 3 3 2 3 2 3 5 3 3 3 4 3 3 5 6 3 7 6 7 5 6 6 6 5 5 5 5 6
THA HEM
CYM ROT
HAL OPU
ROS ORI
HAL UNI
CER TRU
AMP FRA
HAL M IN
CYM SER
CAU SER
SYR ISO
ENH ACO
HAL MAC
HAL STI
THA CIL
EN T KYL
HYP CO R 1. 1
ULV RET 11
NER IM I
AM P R IG
HYP VAL
m m 0
T W IN -group 4 with Halophila ovalis and ered as the climax com m unity betw een low
Halodule wrightii as indicator species, which w ater m ark at neap tide and at spring tide;
can be considered as a pioneer com m unity including the relevés w ith Enhalus acoroides',
in the intertidal zone; - T W IN -group 7 with Thalassodendron cilia
T W IN -group 5 with Thalassia hemprichii and tum as indicator species w hich can be con
Cymodocea rotundata as indicator species, sidered as the climax com m unity o f the sub
which can be considered as the climax com tidal zone;
m unity in shallow intertidal pools and flats; - T W IN -group 8 w ith Halophila stipulacea as
T W IN -group 6 with Syringodium isoetifolium indicator species, w hich can be considered
as indicator species, which can be consid as an infralittoral pioneer com m unity.
67
1 1 1 2 2 2 3 3 3 3 3 4 4 4 5 6 6 6 6 7 3 3 3 3 5 5 5 5 2 2 5 5
4 5 6 2 5 8 6 7 8 0 1 2 4 9 6 7 9 0 1 2 4 5 9 5 6 7 8 1 2 5 6 3 4 3 4
BOO PUS
EN T FLE
ENT RAM 2 3 2 5
E N T CLA
ENT SPE
ULV LAC
LAU PAP
PAD BOR
ACA SPI
CHA CRA
GRA SAL
GR A E D U
HYP N ID
CLA PAT
HAL OVA 2 . . . . . 3 ............................ 3 . . .
HAL WRI 2 2 . 2 . . . . 2 . 2 2 3 . 5
. . . . 2 2 2
T H A H EM . 5 . 5 3 3 5 5 4 3 6 3 3 5 5. . 5 2 5 4 5 3 2 5 5 6 6 5 6 3
CYM ROT 2 5 2 5 3 4 5 5 4 5 5 5 5 3 3 5 3 4 4 4 2 5 . . . 5 3 . 2 5
HAL O PU . 3 . 2 1 2. 5 2 4 4 . . . 2
ROS ORI 1 . 2 . 1 1 . . . . 1 1 . .
HAL U N I 2 2 3
CER TRU 3 5 5
AM P FRA
HAL M IN
CYM SER
CA U SER 2 1
SYR ISO 5 2 4 3 5 5 5 3
E N H ' ACO . 5
HAL MAC 1
HAL STI 3 3
TH A CIL 8 9
ENT KYL
HYP COR 1 1
ULV RET
NER IM I
AM P RIG
HYP VAL
0 0
Correlations between the vegetation and abiotic fa c cal significance for explaining the variability are:
tors (CCA) distance to low w ater, P, organic m aterial, pH ,
fractions 2000-1000, 5 0 -2 , 2 -0 /im . W ithin the
Intercorrelations between environmental factors latter, the C anonical C orrespondence Analysis
Some variables are strongly correlated with oth was applied again.
ers (inflation factor > 2 0 ) and therefore do not The correlation m atrix specifies the intercorre
have an univocal contribution to the explanation lations between some o f the variables o f the ex
of the variation in vegetation. They are: N a, K, cluded group: N a: strong correlation with K and
Ca, Mg, fractions 1000-500, 500-200, 200-100, small particle fractions; K: strong correlation with
100-50 /um. P aram eters which do have an univo organic m aterial and small fractions; Ca: strong
68
o co VO «O oo t" Os VO OO O' O' O' oo VO related. Axis 1 is mainly defined by the am ount of
o © o 04 04 p 04 '3- 04 so 04 ©
r-î Ö © Ö Ö Ö Ö Ö Ö Ö Ö Ö o © © organic m aterial and the particle fraction 5 0 -2 ¿un
in positive direction and p H in negative direction
(higher pH values corresponding with smaller co
3. GC C
'S ^ ^ 'S 0 =3. =3. 5S. ordinates).
go 2*¡ bß O
o
O oO oO o 3. Sa. a.
o
E
6Û\ M 5? ^ cd - H i n M r H (N) O
Axis 2: the defining variables are the distance
S J g Ë »G I I I I I I I
O O O O O O 04 to low w ater m ark in positive direction and P in
'S A" 00
^ ff E O
ô öO O O O UO
> Q Z, *4 u S a* o o.S 2 IO (N -H negative direction. C oarse sedim ents are also
69
Table 5. Percentage variance accounted for by first 4 axes. the particle size fraction 2000-1000 p m , organic
m aterial and particle size fraction 5 0 -2 p m , P
Axis Percentage Eigenvalue
and particle size fractions 2 -0 )im; negative cor
1 37.6 0.67 relation between pH and organic m aterial or 5 0 -
2 22.5 0.40 2 fim size fraction.
3 14.2 0.25 Well m arked correlations betw een environm en
4 11.4 0.20
tal factors and species (diagram 2) are no t n u
m erous: Boodleopsis pusilla (5), Enteromorpha
Table 6. Inter set correlations of environmental variables with flexuosa (17) and E. ramulosa (14) are positively
axes. (Multiplier 100).
correlated with high values o f organic m aterial
Variables Axis 1 Axis 2 Axis 3 Axis 4 and particle size fraction 5 0 -2 p m and with low
pH values. In the field these seaweed species are
D ist low 157 518 - 128 - 15 typically found (in and) on the m ud flats devoid
P (mg/kg) 166 -3 7 7 -5 0 7 - 137 o f seagrasses. Boodleopsis pusilla is a tiny filamen
Org mat (% ) 535 - 133 154 270
tous algae, growing in dense m ats, the vertically
pH -3 9 1 102 -2 2 1 -2 4 7
2000-1000 115 316 478 -2 6 6 growing filam ents fixing enorm ous am ounts o f ex
50-2 414 -7 4 72 141 tremely small particles.
2 -0 104 -2 7 6 -3 5 189 C orrelations between the T W IN S P A N relevé
clusters and environm ental variables are ex
pressed in diagram 3: the T W IN -groups 1, 2, 3,
positively correlated, small sedim ents rather neg 4 + 5 are distributed along the axis distance to low
atively. w ater m ark [from subtidal (1, 2) over infralittoral
Axis 3: mainly defined by P in negative direc fringe (3) to m idlittoral (4,5)]; the T W IN -
tion and the fraction very coarse sand in positive groups 4 + 5, 6, 7, 8 (all from the m idlittoral zone)
direction. are distributed along the axes organic m aterial
Axis 4: m ost positive correlation with the and particle size fraction 5 0 -2 p m suggesting that
am ount of organic m aterial and m ost negative the distribution o f the seagrasses within the m id
correlation with the fraction very coarse sand. littoral zone is due to these environm ental varia
bles.
Ordination diagrams based on canonical correspon
dence analysis Conclusions and discussion
The m ost evident correlations appear from the
diagram o f the first and second canonical ordi A zonation pattern o f seagrasses and associated
nation axes. C orrelations am ongst environm ental m acroalgae can generally be recognized in G azi
factors are illustrated in diagram 1: positive cor Bay. This has been corroborated by the T W IN
relation between distance to low w ater m ark and S PA N and C CA -analysis.
70
AXIS 1
-540 0 540
CD
i_n
dist low
2000-1000
AXIS 2
CD
5 0 -2 ;jm
org mat
2-0 y CD
CD
CD
LTD
I
CO
CD
X
CD
CD
CD
CD
-SO
AXIS
AXIS 2
CD A OA
CD
CD
I
-GO GO
AXIS 1
71
The numbering of species used for the canonical ordination diagrams are added between [ ].
ACA SPI = A canthophora spicifera [1], A M P FRA = Amphiroa fragilis [2], AM P R IG = A. rigida [3], AM P SPE = A. species
[4], BOO PU S = Boodleopsis pusilla [5], CAU SER = Caulerpa serrulata [6], CER cTR = Ceramium cf. truncatum [7],
CYM ROT = Cymodocea rotundata [8], CYM SER = C. serrulata [9], E N H ACO = Enhalus acoroides [10],
ENT cCL = Enteromorpha cf. clathrata [11], ENT KYL = E. kylinii [13], ENT RAM = E. ramulosa [ 14], ENT SP2, EN T SP3,
ENT SP5 = E. species 2 [15], 3 [16], 5 [17], GRA SAL = Gracilaria salicornia [18], HAL MAC = Halimeda macroloba [20],
H AL M IN = Halophila minima [20], HAL O PU = Halimeda opuntia [19], HAL OVA = Halophila ovata [24], Halophila min
ima [20], HAL STI = H. stipulacea [23], H AL U N I = Halodule univervis [21], HAL W RI = H. wrightii [22],
HYP COR = Hypnea cornuta [26], PAD SP1 = Padina species [27], ROS O RI = Rosenvingea orientalis [28],
SPH SPE = Sphacelaria species [29], SYR ISO = Syringodium isoetifolium [30], TH A H EM = Thalassia hemprichii [31],
THA CIL = Thalassodendron ciliatum [32].
Fig. 3. Ordination diagrams based on canonical correspondance analysis: 1. Environmental factors (dist low = distance to low
water mark, org mat = organic material, P = phosphorus, 2000-1000 /im, 50-2 /im, 2 -0 /im = particle size fractions); 2. Species
(numbers corresponding with those given after Table 7); 3. Relevé clusters belonging to the TW IN-groups 1: 2: ♦ , 3: A, 4: 0 ,
5: A , 6: □ , 7: ■ 8: O .
72
A ccording to our field observations this zonation Coppejans, E. & E. Gallin, 1989. Macroalgae associated with
scheme can be generalized for the K enyan coast, the mangrove vegetation of G azi Bay (Kenya). Bull. Soc.
r. Bot. Belg. 122: 46-60.
except for Enhalus w hich seems to be a rather rare Cottenie, A., M. Verloo, L. Kiekens, G. Velghe & R. Cam-
species in this region. It also corresponds well erlynck, 1982. Chemical analysis of Plants and Soils.
with the one described by B rouns & Heijs (1985) I.W .O.N.L. Brussel. 63 pp.
and Heijs (1987) for S. P ap u a N ew G uinea (P ort Dawes, C. J., 1981. Marine Botany. John Wiley & Sons, New
York.
M oresby area) except for the absence o f
den Hartog, C., 1967. The structural aspect in the ecology of
Thalassodendron ciliatum in the subtidal zone sea-grass communities. Helgol. Wiss. Meeresunters. 15:
around P o rt M oresby; it is there replaced by 648-659.
Thalassia hemprichii which either form s subtidal den Hartog, C., 1970. The sea-grasses of the world. Verh.
m onospecific stands or is growing together with Kon. Ned. Ak. Wetensch., Natuurk. Reeks 2, 59: 1-275.
den Hartog, C., 1977. Structure, function, and classification
other species, forming extensive mixed seagrass
in sea-grass communities. In: McRoy & Helfferich (eds),
m eadow s. Seagrass Ecosystems: a scientific perspective. M. Dekker
Inc., New York: 89-121.
De Pauw, K., 1990. De vegetatie van de getijdengeul van Gazi
Acknowledgements Bay (Kenia). M. Sc. thesis State University G hent (Bel
gium): 142 + tab.
This study w as sponsored by N .F .W .O . (N atio De Wit, M., 1988. Vegetatie-ecologische studie van de zee
naal F o n d s voor W etenschappelijk O nderzoek), grasgemeenschap (incl. wiercomponent) in G azi Bay,
F .K .F .O . (F o n d s voor K ollektief en F undam en Kenia. M. Sc. thesis State University G hent (Belgium):
189 +81 pp.
teel O nderzoek) and E E C (E uropean Econom ie FAO, 1977. FAO-guidelines for soil profile description. Rome.
Com m unity). W e are also grateful to Ph. P olk for 66 pp.
giving us the opportunity to do botanical research Gallin, E., E. Coppejans & H. Beeckman, 1989. The man
in K enya within the fram ew ork o f the K enyan grove vegetation of Gazi Bay (Kenya). Bull. Soc. r. Bot.
Belg. 122: 197-207.
Belgian R esearch Project, to M r. Alíela for p u t
Grady, J. R., 1981. Properties o f sea grass and sand flat sedi
ting the infrastructure o f the K enya M arine and ment from the intertidal zone of St. Andrew Bay, Florida.
Fisheries R esearch Institute at our disposal. O ur Estuaries 4: 335-344.
gratitude also goes to I. W olfs, E. M artens and N. Harlin, M. M., 1980. Seagrass epiphytes. In Phillips R. C. &
Revis for the local organization o f and T. Beeck- McRoy C. P. (eds), H andbook of seagrass biology. An eco
m an for the help w ith the field w ork and labora system perspective. G arland STPM Press, New York:
147-152.
tory research. Heijs, F., 1987. Community structure and seasonality of macro
algae in some mixed seagrass meadows from Papua New
Guinea. Aquat. Bot. 27: 139-158.
References Hill, M. O., 1979. Twinspan - a FO RTRA N program - for
arranging multivariate data in an ordered two-way table by
classification of the individuals and attribute: 90 pp. Cornell
Beeckman, H., 1989. Mathematische verwerking van phy-
Univ. Ithaca, New York.
tosociologische gegevens. Groene Band 76: 1-36.
Beeckman, H. & R. Lemeur, 1989. Ecology, science of com Hutchinson, G. E., 1944. Limnological studies in Connecticut
VII. A critical examination of the supposed relationship
plexity and interactions. FT I 1989 Project Sheet Interface
R U G 9: 3 pp. between phytoplankton periodicity and chemical changes
in lake waters. Ecology 25: 3-26.
Brouns, J. & F. Heijs, 1985. Tropical seagrass ecosystems in
Hutchinson, G. E., 1957. Concluding remarks. Cold Spring
Papua New Guinea. A general account of the environment,
marine flora and fauna. Mar. Biol. 88: 13-50. Harbor Symp. Quant. Biol. 22: 415-427.
Hutchinson, G. E., 1978. An introduction to population eco
Brouns, J. & F. Heijs, 1986. Structural and functional aspects
of seagrass communities and associated algae from the logy. Yale Univ. Press, New Haven, Connecticut.
tropical West-Pacific. Ph.D. thesis Nijmegen (Nether Isaac, F., 1968. Marine Botany of the Kenya coast. 4. An-
lands). 431 pp. giosperms. J. E. Afr. N at. Hist. Soc. 27: 29-47.
Coppejans, E. & T. Beeckman, 1989. Caulerpa section Se- Jaasund, E., 1976. Intertidal seaweeds in Tanzania. Univ.
doideae (Chlorophyta, Caulerpales) from the Kenyan coast. Tromso. 160 pp.
Legendre, P. & L. Legendre, 1984a. Ecologie numérique. 1.
N ova Hedwigia 49: 381-393.
75
Le traitement multiple de données écologiques. Masson. beds at Barbados and Cariacon, W est Indies and its
Paris. 260 pp. ecological and geological implications. Aquat. Bot. 1: 163—
Legendre, P. & L. Legendre, 1984b. Ecologie numérique. 2. 189.
La structure des données écologiques. M asson. Paris. Phillips, R. C. & E. G. Meflez, 1988. Seagrasses. Smiths.
335 pp. Contrib. Mar. Sei. 34: 104 pp.
Londo, G., 1984. The decimal scale for relevés of permanent Schoener, T. W., 1989. The ecological niche. In: J. M. Cher-
quadrats. In R. Knapp (ed.), Handbook of vegetation rett (ed.), Ecological Concepts. Blackwell Scientific Publi
science 4: 45-50. cations, Oxford, 385 pp.
McRoy, C. P. & C. McMillan, 1977. Production ecology and ter Braak, C. J. F., 1986. Canonical correspondance analysis:
physiology of seagrasses. In: McRoy C. P. & Helfferich C. a new eigenvector technique for multivariate direct gradient
(eds), Seagrass ecosystems: a scientific perspective. M. De analysis. Ecology 67: 1167-1179.
kker Inc., New York: 53-88. ter Braak, C. J. F., 1987. Ordination. In: Jongman R. H.,
Molinier, R. & J. Picard, 1951. Biologie des herbiers de C. J. F. ter Braak & O. F. R. Van Tongeren (eds), D ata
Zostéracées des côtes françaises de la Méditerranée. C. r. Analysis in community and landscape ecology. Pudoc
Acad. Sei., Paris 233: 1212-1214. Wageningen. 301 pp.
Molinier, R. & J. Picard, 1952. Recherches sur les herbiers de ter Braak, C. J. F., 1988. CANOCO - A FO RTRA N pro
phanérogames marines du littoral Méditerranéen français. gram for canonical community ordination by partial de
Ann. Inst. Océanogr. 27: 157-234. trended canonical correspondance analysis, principal com
Moorjani, S. & B. Simpson, 1988. Seaweeds of the Kenya ponents analysis and redundancy analysis (version 2.1.).
coast. Oxford Univ. Press. Nairobi. 134 pp. Agrie. Math. Group. Wageningen. 95 pp.
Onyango, H. B. A., 1989. Numerical modelling of Gazi Creek Walkley, A. & E. A. Black, 1934. An examination of the
(South coast, Kenya): 1. Hydrodynamics model; 2. Degtjareff method for determining soil organic matter and
Descriptive ecological model. M. Sc. thesis FAM E. VU proposed modification of the chromic acid titration method.
Brussel. Soil Sei. 37: 29.
Patriquin, D. G., 1972. The origin of nitrogen and phospho Zieman, J. C. & R. G. Wetzel, 1980. Productivity in
rus for growth of the marine angiosperm Thalassia seagrasses: methods and rates. In: Phillips R. C. & McRoy
testudinum. Mar. Biol. 15: 35-46. C. P. (eds), Handbook of seagrass biology. An ecosystem
Patriquin, D. G., 1975. ‘Migration’ of blowouts in seagrass perspective. Garland STPM Press, New York: 87-116.
'jp w iy p
"jiUpj»)!)' Mjj
Hydrobiologia 2 4 7 : 77-86, 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 77
© 1992 Kluwer Academie Publishers. Printed in Belgium.
L. K an n an & K. V asantha
C A S in Marine Biology, Annam alai University, Parangipettai - 608 502, India
Abstract
Species richness o f phytoplankton o f the P itchavaram m angals was high. T here were 82 species con
stituted by 67 species o f diatom s, 12 species of dinoflagellates and 3 species o f bluegreen algae. The
diatom s form ed the bulk with 72% followed by the dinoflagellates with 15% , in which there were au
tochthonous (tem porary and perm anent) and allochthonous forms. P hytoplankton population density
exhibited a wide seasonal fluctuation with the m inim um during m onsoon and the m axim um during
sum m er, suggesting the possible differential influence o f various environm ental factors.
MONTHS 1985
800-
Fig. 2. Monthly mean of rainfall recorded at Porto Novo (common to stations 1 & 2).
the stations during the study period is given in toplankton was recorded which included 83.75%
T able 1. o f diatom s, 13.75% o f dinoflagellates and 2.50%
o f bluegreens. H ence diatom s were the dom inant
Population density phytoplankton group in the study areas. Khalil
(1990) also reported th at diatom s were the dom
The fluctuations in phytoplankton population inant group com prising 52 to 90% in lake M an-
density were wide in different m onths o f the zala (Egypt) owing to their relative tolerance to
present study period (Fig. 9). In general, the m ax salinity.
imum density w as recorded during sum m er (M ay) Generally, the diatom s ranked first in species
and the m inim um , during m onsoon and p ostm on com position followed by dinoflagellates and blue
soon seasons. A t station 1, the phytoplankton greens. This tendency w as observed throughout
density ranged from 360 (D ecem ber) to 685 260 the study period at both the stations except d u r
cells 1“ 1 (M ay) and at station 2, the density var ing the freshw ater flow in m onsoon season when
ied betw een 420 (January) and 784320 cells 1“ 1 the dom inance o f freshw ater phytoplankton con
(M ay). stituting green algae such as Pediastrum simplex
The phytoplankton com prised 67 species o f di w as observed.
atom s (Bacillariophyceae), 12 species o f di The perm anent autochthonous diatom species
noflagellates (D inophyceae) and 3 species of described from here by Sivakum ar (1982) viz.
bluegreens (C yanophyceae). A total num ber o f 76 Asterionella glacialis, Bacillaria paradoxa,
species o f phytoplankton w as recorded from sta Chaetoceros affinis, Coscinodiscus marginatus,
tion 1 which included 84.21% o f diatom s, 13.15% Odontella mobiliensis, Thalassiothrix frauenfeldii
o f dinoflagellates and 2.64% o f bluegreens. A t and the dinoflagellates, Ceratium furca and Pro
station 2, a total num ber o f 80 species o f phy toperidinium sp. were recorded from station 1 dur-
80
Station I St at i on 2
Sur f ac e w a t e r
35
33
31
29
27
25
35 Atmospheric
33
31
29
J J F F M M A A M M J J J y J y A A S S O O N N D D
MONT HS 1985
Fig. 3. Monthly variations in atmospheric and surface water temperatures.
16
o—o S t a t i o n 1 Station 2
12
36
32
28
20
16
J J F F M M A A M M J J J y J y A A S S O O N N D D
MONTHS 1985
Fig. 4. Monthly variations in light extinction coefficient and salinity.
Station I Station 2
o
o
>
J J F F M M A A M M J J J y J y A A S S O O N N D D
M O N T H S 1985
Fig. 5. Monthly variations in dissolved oxygen and hydrogen-ion concentrations.
83
S tation 2.
o f phytoplankton (K etchum , 1964; K rishnam ur-
o -o S tat ion I
thy et al., 1974; R am an et al., 1990). In the present
study, a low density o f phytoplankton (360
cells 1“ 1 at station 1 and 384 cells 1" 1 at station 2)
08
w as noticed during late m onsoon and early p o st
m onsoon seasons respectively w hen the salinity
w as low (20.23%o at station 1 and 21.21%0 at sta
tion 2) due to the inflow o f freshw ater from the
20 upper reaches causing greater turbidity. R ecord
ing o f a low er density o f phytoplankton during
m onsoon m onths was due to the heavy freshw a
ter inflow which flushed m ost o f the native or
ce. 0 8
ganism s including phytoplankton from study
areas into the sea. C adee (1984) also reported
th at excessive run-off might lead to dilution o f the
300
phytoplankton and produce ‘w ash o u t’.
•r 2 5 0
Subsequently the population density o f phy
3200
toplankton gradually increased during postm on
150 soon season (8 490 cells 1“ 1 at station 1 and
□ 100 19450 cells D 1 at station 2) and reached the peak
50 (685260 cells I -1 at station 1 and 784320
cells D 1 at station 2) during sum m er and contin
MONTHS 1985
ued till early prem onsoon season (102970
cells 1” 1 at station 1 and 127450 cells D 1 at sta
Fig. 6. Monthly variations in phosphate, nitrate and silicate
tion 2). The sum m er peak w as aided by the inflow
concentrations.
o f clear neritic w aters coupled with penetration
The im portance o f tem perature as a factor in o f m ore light into the w ater colum n and flourish
controlling the grow th o f phytoplankton w as em ing grow th o f m ostly the net phytoplankton
phasized by m any w orkers (Williams & M urdoch, {Bacillaria paradoxa, Chaetoceros affinis, Odontella
1976; G oldm an, 1977). In the present study, m ax mobiliensis, Rhizosolenia alata and Thalassiothrix
imum biom ass o f phytoplankton w as observed frauenfeldii), tolerating a higher salinity (32.32%0
when the tem perature o f surface w ater was high. at station 1 and 34.85%0 at station 2). Such a bi-
The peaks o f phytoplankton density (685260 m odal oscillation o f phytoplankton recording high
cells 1“ 1 at station 1 and 784320 cells 1“ 1 at sta density during sum m er and prem onsoon w as ob
tion 2) were observed when the surface w ater served by R am adhas (1977) in the Vellar estuary.
tem perature (31.0 °C at station 1 and 33.1 °C at In general, low concentrations o f nitrate and
station 2) w as high. H owever, a low phytoplank silicate recorded at stations 1 and 2 during p o st
to n density (360 cells 1 ' w as also observed in m onsoon to prem onsoon seasons suggest the
D ecem ber when the surface w ater tem perature possible utilization o f these nutrients in large
w as high (30.2 °C ) at station 1. Sivakum ar (1982) am ount by the abundantly occurring phytoplank
also noticed the coincidence o f low tem perature ton. H owever, these tw o nutrients were not lim
with low phytoplankton density in the backw a iting as revealed by the good population density
ters o f P o rto N ovo. H ence influence o f surface o f phytoplankton especially diatom s though sili
w ater tem perature on phytoplankton density is cate could becom e a limiting nutrient for diatom s
rather obscure in tropical m arine environs. (W etsteyn eta l., 1990).
Salinity acts as a m ajor ecological factor co n High standing crop of phytoplankton (685260
trolling the distribution and seasonal succession cells 1“ 1 at station 1 and 784320 cells D 1 at sta-
84
0-15
□ Station 1
0.13 ■ St a t i o n z
0.11
0-09
ct 0-07
E
0-05
o
CL
0-03
0-01
J J F F M M A A M M J J JyJyA A S S O O N N D D
MONTHS 1985
Fig. 7. Monthly variations in particulate organic carbon.
MONTHS 1985
6 -
- 5
) i F FM M A A M M J J J y J y A A S S O O N N D D
1985
Fig. 9. Monthly variations in phytoplankton population density.
tion 2) dom inated by large diatom s were recorded 784320 cells I -1 recording low density during
during sum m er in which season high contents m onsoon and high density during summer. Sum
(0.115 mg C l -1 at station 1 and 0.148 mg C 1 " 1 m er crop o f m icrophytoplankton contributes to
at station 2) of P O C and low bacterial counts the production o f P O M and retardation o f het-
(1.213 X IO4 C F U m l -1 at station 1 and erotrophic bacterial growth.
1.643 X IO4 C F U m l -1 at station 2) were re
corded. This revealed the im portance of phy
toplankton for the production o f P O M (S arm a & Acknowledgements
N agesw ara R ao, 1989) and retardation of het-
erotrophic bacterial grow th owing to their extra W e th an k the D irector o f the CAS in M arine
cellular release o f antibacterial substances Biology and the authorities o f the A nnam alai
(K ogure eta l., 1979; Sieburth, 1979). U niversity for facilities and encouragem ent. F i
nancial help by the U G C , N ew D elhi, is gratefully
acknowledged.
Summary
Abstract
The aerial prop roots o f the neotropical red m angrove, Rhizophora mangle L., begin growing well above
highest high w ater (HE1W) and often extend well below lowest low w ater (LLW ) before rooting in the
benthic substratum . In Belize, C entral Am erica, prop roots growing below LLW are colonized by di
verse assem blages o f organism s, including m acroalgae, hydrozoans, ascidians, sponges, anem ones, hard
corals, and isopod crustaceans. M angroves, root-fouling epibionts, root herbivores, and benthic pred
ators engage in com plex interactions th at are m ajor determ inants o f m angrove growth and production.
Species richness o f root epibionts increases with distance from the m ainland and with proxim ity to the
barrier reef. Species richness decreases with variability in w ater tem perature and salinity. A scidians and
sponges transplanted from L ark Cay into the coastal Placencia Lagoon failed to survive, but anem o
nes from L ark Cay survived in Placencia Lagoon. R eciprocal transplants survived off-shore. The gas
tro p o d predator, Melongena melongena L., present only in m ainland estuaries, reduced local barnacle
abundance and epibiont species richness in Placencia Lagoon. Isopod species richness also increases
with distance from shore, but the num ber o f roots bored by these species decreases. These isopods can
reduce root relative grow th rate (R G R root) by 55% . On off-shore cays, sponges and ascidians amelio
rate negative effects o f isopods. In m ainland estuaries where epibionts are less com m on, isopod dam
age to roots is m ore severe. Experim ental studies in m angrove sw am ps throughout the w orld would
clarify the im portance o f plant-anim al interactions in these w idespread tropical ecosystems.
Reef
BuC
Study sites
Fig. 1. M ap of southern Belize, showing the six study sites.
P L : Placencia Lagoon; BgO. Big Creek; LO. Lark Cay; BuO.
Belize, located in northeastern C entral Am erica,
Bugle Cay; NO. N ortheast Cay; TO. Twin Cays. Names and
includes the largest continuous barrier reef in the locations follow British Survey Sheets of Belize, nos. 32
w estern hem isphere (R ützler & M acintyre, 1982). (1966), 36 (1976), and 40 (1982).
89
Table 1. Taxa encountered on mangrove roots in various locations o f coastal Belize (Site abbreviations as in Fig. 1).
Cyanobacteria1
Scytonem a polycystum Barnet & Flahault y y
Chlorophyta2
Acetabularia crenulata Lamour. V y y y
Caulerpa mexicana (Sond.) J. Ag. y y
Caulerpa racemosa (Forssk.) J. Ag. y y y
Caulerpa sertulariodes (S. G. Gmelin) Howe y
Caulerpa verticillata J. Ag. v/ v/ y y
Derbesia osterhoutii (L. & A. H. Blinks) Page y y
Halimeda spp. y y y y
Valonia macrophysa Ktitz. y y
Ventricaria ventricosa J. Ag. y y
Phaeophyta2
Dictyota spp. y y y y y
Lobophora variegata (Lamour.) Worm. y y y y
Padina sanctae-crucis Borg. y y y y y
Rhodophyta2
Amphiroa rigida var. antillana Borg. y
Bostrychia spp. y y y y y y
Corallinaceae y y y y
Wrangelia argus (Montagne) Montagne y y y
Porifera3
Amphimedon viridis (D. & M.) y y y y y
Chondrilla nucula Schmidt y
Halichondria melanadocia de Laubenfels y
Haliclona implexiformis (Hechtel) y y y y
Haliclona curacaoensis (van Soest) y y y
Haliclona pseudom olitba de Weerdt, Riitzler, and Smith y
Lissodendoryx isodictyalis (Carter) y y y
Mycale microsigmatosa Arndt y
Mycale sp. 2 fide Riitzler y y y
Pellina carbonaria (Lamarck) y y
Tedania ignis (D. & M.) y y y y y
Ulosa ruetzleri Wiedenmayer y y y y
C nidaria1
Aiptasia pallida (Verrili) y y y y
Bartholomea annulata (Lesueur) y y
Condylactis gigantea (Weinland) y y y
Halocordyle disticha (Goldfuss) y y y y
Halecium spp. y y y
Millepora alicornis L. y y
Palythoa caribaea D. & M. y y y
Sertularia turbinata (Lamour.) y y y
Annelida
Sabellidae y y y y
Serpulidae y y y y
91
Table 1. (Continued)
Crustacea (Cirripedia)1
Balanus eburneus Gould V y V y y y
Chthamalus angustitergum thompsoni (Henry) y y
Crustacea (Isopoda)4
Dynamenella cf. angulata (Richardson) y
Paracerceis caudata Say y
Phycolimnoria clarkae Kensley & Schotte J y
Sphaeroma terebrans Bate y V
Mollusca (Bivalvia)5
Crassostrea rhizophorae Guilding J y y
Isognomon alatus Gmelin y d d y y
Chordata (Tunicata)6
Ascidia nigra (Savigny) s/ y
Botrylloides nigrum Herdman J y y
Botrylloides cf. perspicuum Herdman J y y
Clavelina picta (Verrili) y
Didemnum conchyliatum (Sluiter) d y y
Didemnum psammathodes Sluiter y
Diplosoma glandulosum (Milne Edwards) J y y
Distaplia corolla Monniot s/ y y
Ecteinascidia minuta (Berrill) y
Ecteinascidia turbinata Herdman y y y
Eudistoma olivaceum (van Name) d s/ y y
Eudistoma obscuratum (van Name) y
Perophora formosana (Oka) y
Perophora regina Goodbody & Cole y
Perophora viridis (Verrili) ■J y y
Styela partita (Stimpson) J y
1 Nomenclature follows Sterrer (1986).
2 Nomenclature follows Littler et al. (1989); algae identified by D.S. Littler, Smithsonian Institution.
3 Sponges identified by K. Riitzler, Smithsonian Institution. Haliclona nomenclature follows de W eerdtetal. (1991).
4 Nomenclature follows Kensley & Schotte (1989); isopods identified by B. Kensley, Smithsonian Insitution.
5 Nomenclature follows McLean (1951).
6 Tunicates identified by I. Goodbody, University of the W est Indies.
only on aerial roots, while other cnidarians (Pa found m ost com m only in the ro o t tip ( F < 0 .0 1 ,
lythoa, Millepora, and Condylactis) occurred only G -test), and barnacles only occurred on the in
on ground roots. A scidian species richness also tertidal portions o f the roots (P < 0.001, G -test).
w as m uch higher on aerial roots at these tw o sites Bostrychia w as m ost com m on at the low w ater
(Lark: 5 species only on aerial roots; Twin: 11 line f P < 0.05, G -test). All o f the other species
species only on aerial roots). At Lark and Twin, occurred only below LLW , and were found with
isopods only attacked aerial roots. equal likelihood at all locations along the subtidal
Few clear patterns in vertical zonation o f epi portion o f the root (F > 0 .5 0 , G -test).
bionts on roots were observed. Isopods were
92
too
Effects of epibionts on mangroves
75
Interactions with M elongena
50
T o determ ine the effects o f these faunal interac
25
100
ft=- A tions on the trees them selves, we m easured tree
height and stem diam eter (‘d b h ’: above the high
est prop root) in the Melongena experim ent
(above). T here w as a trend tow ards increased
height and ‘d b h ’ in control and cage-control trees
75 where snails foraged on barnacles, relative to trees
50
Fig. 6. Results of the snail exclosure experiment on epibiont
25 % cover (x + 1 SE) at Placencia. A: Algae (Bostrychia spp.
and Caulerpa verticillata) on aerial roots; B: Algae on ground
roots; C: Barnacles (Balanus and Chthamalus) on aerial roots;
D : Barnacles on ground roots; E : Sponges (Pellina) on ground
0 50 100 150 200 250 300
roots. D ay 0 is 1 February 1988.
DAYS
95
Discussion
prevents isopod colonization o f roots (Ellison & trophic com plexity and determ inants o f com m u
F arnsw orth, 1990). In C osta Rica, isopods nity structure no t predicted by observation alone
(Sphaeromaperuvianum R ichardson) also reduced (e.g. Connell, 1961; Connell, 1975; Paine, 1966;
ro o t grow th rate by 50% and net root production Paine, 1988; M enge eta l., 1986; M enge & S uth
by 62% (Perry, 1988). Perry’s experim ental re erland, 1987). Similar studies in tem perature zone
sults also indicated some reduction in the effects salt m arshes have docum ented m ajor effects of
of isopods in roots exposed to w ater colum n benthic invertebrates on plant production (Bert-
predators, but she w as unable to identify the iso ness, 1984; B ertness, 1985). In m angrove
pod predator. sw am ps, recent studies have uncovered unex
In Belize, predation m ay also am eliorate p o pected interactions betw een herbivorous craps
tentially negative effects o f epibionts on tree and plant zonation (Sm ith, 1987; Smith eta l.,
growth. In the m ainland estuary at Placencia, 1989), and betw een epibionts and root p roduc
Melongena predation reduces barnacle cover on tion (Perry, 1988; Ellison & F arnsw orth, 1990;
b o th aerial and ground roots (Fig. 6). T hose trees R odriguez & Stoner, 1990). Long-term experi
on which Melongena reduced barnacle cover grew m ental studies on plant-anim al interactions in
m ore rapidly than trees with heavy barnacle cover. m angroves throughout the w orld w ould likely
In the only other study o f the effects o f barnacles yield new insights into com m unity structure,
on m angrove roots, Perry (1988) found th at Bal functioning, and complexity.
anus spp. reduced R . mangle root grow th rate
(c m m o n th “ 1) by 30% and net root production
( g r o o t- 1 m o n th “ % by 52% along the Pacific Acknowledgements
coast o f C o sta Rica. Perry (1988) found th at b ar
nacles were preyed upon by snails ( Thais E A R T H W A T C H (C enter for Field R esearch),
kiosquiformis D uclus and Morula lugubris A dam s) the N ational G eographic Society, and the Sm ith
and herm it crabs (Clibanarius panamensis Stimp- sonian Institution N M N H (with additional sup
son) and proposed th at predators indirectly fa port from E X X O N ) supported this research. W e
cilitate tree growth. th an k six team s o f dedicated E A R T H W A T C H
Epibionts are found w orld-wide on m angrove volunteers for field assistance, and Chris, Skip,
roots (M attox, 1949; M aenae & Kalk, 1962; Jessica, and M eredith W hite for friendship and
M aenae, 1968; Riitzler, 1969; H utchings & Re- logistical support at L ark and Placencia.
cher, 1974; Sasekum ar, 1974; Pinto & W ignara- M eredith W hite collected the salinity d a ta at P la
jah , 1980; Sutherland, 1980; A lvarez I., 1989), cencia. This is contribution no. 303, C aribbean
b u t studies o f their roles in the m angal have only C oral R eef Ecosystem s (C C R E ) Program , Sm ith
ju st begun (Perry, 1988; W ada & W ow or, 1989; sonian Institution, and paper 2 in the series ‘The
Ellison & F arnsw orth, 1990; R odriguez & Stoner, ecology o f Belizean m angrove-root fouling com
1990). Historically, ecological studies o f m an m unities’.
groves have em phasized floristic inventories (re
viewed in C hapm an, 1976; Tom linson, 1986) and
References
descriptions o f energy flow and nutrient cycling
(Golley e ta l., 1962; O dum & H eald, 1972; Lugo
Alvarez, I. A., 1989. Establecimiento, desarrollo y manten
& Snedaker, 1974). Although an oft-stated goal imiento de una comunidad epibentonica tropical. Ph.D.
o f faunistic studies in m angal is the determ ination Thesis, Universidad Central de Venezuela, Facultad de
o f the trophic structure and energy budget o f a Ciencias. 150 pp.
given m angrove swam p (Sasekum ar, 1984), this Bacon, P. R., 1971. The maintenance of a resident population
of Balanus eburneus (Gould) in relation to salinity fluctua
goal can n o t be achieved w ithout detailed exper
tions in a Trinidad mangrove swamp. J. exp. mar. Biol.
im ental studies. Intensive m anipulative experi Ecol. 6: 187-198.
m ents in the m arine rocky intertidal have revealed Bertness, M. D., 1984. Ribbed mussels and Spartina
97
alterniflora production in a New England salt marsh. Ecol Lugo, A. E. & S. C. Snedaker, 1974. The ecology of man
ogy 65: 1794-1807. groves. Ann. Rev. Ecol. Syst. 5: 39-64.
Bertness, M. D., 1985. Fiddler crab regulation of Spartina M aenae, W., 1968. A general account of the fauna and flora
alterniflora production on a New England salt marsh. Ecol of mangrove swamps and forests in the Indo-West-Pacific
ogy 66: 1042-1055. region. Adv. mar. Biol. 6: 73-270.
Chapman, V. J., 1976. Mangrove vegetation. J. Cramer, M aenae, W. & M. Kalk, 1962. The ecology of the man
Vaduz, Lichtenstein, 447 pp. grove swamps at Inhaca island, Mozambique. J. Ecol. 50:
Connell, J. H., 1961. Effects of competition, predation by 19-34.
Thais lapillus and other factors on natural populations of the Mattox, N. T., 1949. Studies on the biology of the edible
barnacle Balanus balanoides. Ecol. Monogr. 31: 61-104. oyster, Ostrea rhizophorae Guilding in Puerto Rico. Ecol.
Connell, J. H., 1975. Some mechanisms producing structure Mono. 19: 339-356.
in natural communities: a model and evidence from field M cLean, R. A., 1951. The Pelecypoda or bivalve mollusks of
experiments. In M. L. Cody & J. Diamond (eds), Ecology Porto Rico and the Virgin Islands. In Scientific survey of
and evolution of communities. Belknap, Cambridge, MA, Porto Rico and the Virgin Islands. N.Y. Acad. Sei. 17:
USA: 460-490. 1-183.
Ellison, A. M. & E. J. Farnsworth, 1990. The ecology of Be Menge, B. A., J. Lubchenco, S. D. Gaines & L. R. Ashkenas,
lizean mangrove-root fouling communities: I. Epibenthic 1986. A test of the Menge-Sutherland model of community
fauna are barriers to isopod attach of red mangrove roots. organization in a tropical rocky intertidal food web. Oeco-
J. exp. mar. Biol. Ecol. 142: 91-104. logia (Beri.) 71: 75-89.
Estevez, E. D. & J. L. Simon, 1975. Systematics and ecology Menge, B. A. & J. P. Sutherland, 1987. Community regula
of Sphaeroma (Crustacea: Isopoda) in the mangrove habi tion: variation in disturbance, competition, and predation
tats of Florida. In C. Walsh, S. Snedaker & H. Teas (eds), in relation to environmental stress and recruitment. Am.
Proceedings of international symposium on biology and Nat. 130: 730-757.
management of mangroves. University of Florida, Gaines Odum, W. E. & E. J. Held, 1972. Trophic analyses of an
ville, FL: 286-304. estuarine mangrove community. Bull. mar. Sei. 22: 671—
Farnsworth, E. J. & A. M. Ellison, 1991. Patterns of her- 738.
bivory in Belizean mangrove swamps. Biotropica 23: 555- Paine, R. T., 1966. Food web complexity and species diver
567. sity. Am. N at. 100: 65-75.
Golley, F., H. T. Odum & R. F. Wilson, 1962. The structure Paine, R. T., 1988. Food webs: road maps of interactions or
and metabolism of a Puerto Rican red mangrove forest in grist for theoretical development. Ecology 69: 1648-1654.
May. Ecology 43: 9-19. Perry, D. M., 1988. Effects of associated fauna on growth
Hartshorn, G., L. Nicolait, L. H artshorn, G. Bevier, R. and productivity in the red mangrove. Ecology 69: 1064-
Brightman, J. Cal, A. Cawich, W. Davidson, R. DuBois, C. 1075.
Dyer, J. Gibson, W. Hawley, J. Leonard, R. Nicolait, D. Pinto, L. & S. Wignarajah, 1980. Some ecological aspects of
Weyer, H. White & C. Wright, 1984. Belize country envi the edible oyster Crassostrea cucullata (Born) occurring in
ronmental profile. Trejos H nos Sucs. S.A., San Jose, Costa association with mangroves in Negombo lagoon, Sri Lanka.
Rica, 152 pp. Hydrobiologia 69: 11-19.
Hutchings, P. A. & H. F. Recher, 1974. The fauna of Careel Rehm, A. & H. J. Humm, 1973. Sphaeroma terebrans', a threat
Bay with comments on the ecology of mangrove and sea- to the mangroves of southwestern Florida. Science 182:
grass communities. Aust. Zool. 18: 99-128. 173-174.
Kensley, B. & M. Schotte, 1989. Guide to the marine isopod Ribi, G., 1982. Does the wood-boring isopod Sphaeroma
crustaceans of the Caribbean. Smithsonian Institution terebrans benefit red mangroves (Rhizophora manglej! Bull,
Press, Washington, D.C., 308 pp. mar. Sei. 31: 925-928.
Kjerfve, B., K. Riitzler & G. H. Kierspe, 1982. Tides at Carrie Rodriguez, C. & A. W. Stoner, 1990. The epiphyte commu
Bow Cay, Belize. In K. Riitzler & I. G. Macintyre (eds), nity of mangrove roots in a tropical estuary: distribution
The Atlantic barrier reef ecosystem at Carrie Bow Cay, and biomass. Aquat. Bot. 36: 117-126.
Belize, I. Structure and communities. Smithsonian Inst. Riitzler, K., 1969. The mangrove community, aspects of its
Press, Washington, D.C.: 47-51. structure, faunistics and ecology. In Lagunas Costeras, un
Littler, D. S., M. M. Littler, K. E. Bucher & J. E. Norris, Simposio. U N A M -U N ESC O , Mexico: 515-535.
1989. Marine plants of the Caribbean: a field guide from Riitzler, K., 1987. Recruitment of sponges and their space
Florida to Brazil. Smithsonian Inst. Press, Washington, competitors to settling plates in Twin Cay mangrove hab
D.C., 263 pp. itats. In K. Rützler (ed.), Caribbean Coral Reef Ecosys
Littler, M. M., P. R. Taylor, D. S. Littler, R. H. Sims & J. N. tems: progress report 1987. National Museum of Natural
Norris, 1985. The distribution, abundance, and primary History (Dept, of Echinoderms & Lower Invertebrates),
productivity of submerged macrophytes in a Belize barrier- Smithsonian Inst., Washington, D.C.: 23-24.
reef mangrove system. Atoll Res. Bull. 282: 1-18. Riitzler, K. & I. G. Macintyre (eds), 1982. The Atlantic bar
98
rier reef ecosystem at Carrie Bow Cay, Belize, I. Structure Sterrer, W. (ed.), 1986. M arine fauna and flora o f Bermuda.
and communities. Smithsonian Institution Press, Washing John Wiley & Sons, New York, N.Y., 742 pp.
ton, D.C., 539 pp. Stoddart, D. R., F. R. Fosberg & D. L. Spellman, 1982. Cays
Sasekumar, A., 1974. Distribution of macrofauna on a M a of the Belize barrier reef and lagoon. Atoll Res. Bull. 256:
layan mangrove shore. J. anim. Ecol. 43: 51-69. 1-76.
Sasekumar, A., 1984. M ethods for the study of mangrove Sutherland, J. P., 1980. Dynamics of the epibenthic commu
fauna. In S. C. Snedaker & J. G. Snedaker (eds), The m an nity on roots of the mangrove Rhizophora mangle at Bahia
grove ecosystem: research methods. U N ESC O , Paris, de Buche, Venezuela. Mar. Biol. 58: 75-84.
France: 145-161. Taylor, P. R., M. M. Littler & D. S. Littler, 1986. Escapes
Simberloff, D., B. J. Brown & S. Lowrie, 1978. Isopod and from herbivory in relation to the structure of mangrove is
insect root borers may benefit Florida mangroves. Science land macroalgal communities. Oecologia (Beri.) 69: 481-
201: 630-632. 490.
Smith, T. J. III, 1987. Seed predation in relation to tree dom Tomlinson, P. B., 1986. The botany of mangroves. Cambridge
inance and distribution in mangrove forests. Ecology 68: University Press, Cambridge, U .K., 413 pp.
266-273. W ada, K. & D. Wowor, 1989. Foraging on mangrove pneu-
Smith, T. J. III, H. T. Chan, C. C. M clvor & M. B. Robblee, matophores by ocypodid crabs. J. exp. mar. Biol. Ecol. 134:
1989. Comparisons of seed predation in tropical tidal for 89-100.
ests from three continents. Ecology 70: 146-151. Wahl, M., 1989. Marine epibiosis. I. Fouling and antifouling:
Snedaker, S. C., 1973. Mangroves, isopods, and the ecosys some basic aspects. Mar. Ecol. Prog. Ser. 58: 175-189.
tem. Science 183: 1036-1037. W eerdt, W. H. de, K. Riitzler & K. P; Smith, 1991. The
Sokal, R. R. & F. J. Rohlf, 1981. Biometry, second edition. Chalinidae (Porifera) of Twin Cays, Belize, and adjacent
W. H. Freeman & Co., New York, N.Y., 859 pp. waters. Proc. Biol. Soc. Wash. 104: 189-205.
Hydrobiologia 247: 99-108, 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 99
© 1992 Kluwer Academic Publishers. Printed in Belgium.
Abstract
T he vertical distribution o f m eiofauna in the sedim ents o f Avicennia marina, Bruguiera gymnorrhiza,
Ceriops tagal, Rhizophora mucronata and Sonneratia alba at G azi Bay (Kenya), is described. Seventeen
taxa were observed, with highest densities in the sedim ents o f Bruguiera (6707 ind. 10 c m ' 2), followed
by Rhizophora (3998 ind. 10 c m ~ 2), Avicennia (3442 ind. 10 c m - 2 ), Sonneratia (2889 ind. 10 c m - 2 ) and
Ceriops (1976 ind. 10 c m - 2 ). N em atodes accounted for up to 95% o f total densities; other com m on taxa
were copepods, turbellarians, oligochaetes, polychaetes, ostracods and rotifers. High densities occurred
to about 20 cm depth in the sedim ent. Especially Ceriops sedim ents show still high densities o f nem a
todes (342 ind. 10 c m - 2 ) and copepods (11 ind. 10 c m - 2 ) in the deepest layer (1 5 -2 2 cm). Particle size
and oxygen conditions were m ajor factors influencing m eiobenthic distribution; Uca burrow s h ad a m ajor
im pact on distribution and abundance o f m eiofauna.
MOMBASA
t 1 .....
X j \ r
F IS H E R M E N 'S H A R B O U R 1 : B ru g u iera s i t e
P / I S 2 : A v ic e n n ia s i t e
' 2 3 : a: C e rio p s s i t e
yiHlIlTillb \ b: R h i z o p h o r a s i t e
a 3b 3c c: S o n n e ra tia s i t e
* * * A v ice n n ia m a rin a
* * * * * SCHO OL
* * * L u m n itz e ra r a c e m o s a
* *
& R h iz o p h o ra m u c ro n a ta
* * %
* % S o n n e r a tia a lb a
* * *
* * * % m ix ed m a n g a l
* * * * *
* ■è * C o c o s -p la n ta tio n
* * *
* *
F IS H E R M E N 'S H UT
* C o rn
* * ■H*
* * * * dune
* *
* % * * pool
* * * *
* ta r ro a d
dry w e a th e r ro a d
m o to ra b le tra ck
p a th
50 0 m
Fig. 1. Study area: vegetation map of the mangrove in G azi Bay (after Gallin et al., 1989).
intervals for deeper layers. Each core was ing a m ercury therm om eter into the sedim ent. S a
preserved in a h o t (60 °C ) 4% form aldehyde linity w as m easured with a refractom eter.
solution. L aboratory treatm ent consisted of ex
tracting the m eiofauna by a com bination o f sed Results
im entation, sieving (38 /rm), centrifugation, and
staining with R ose Bengal. R oots and organic Sedim ent tem perature reached 2 4 -2 5 °C and sa
detritus were n o t separated, w hat m ade counting linity 32—35%0; a small input o f fresh w ater by the
o f the m ajor tax a very tim e consum ing. small rivers K idogoweni and M kurum u had no
m easurable influence on the m angrove. p H was
Sedim ent analysis 7.15-7.85.
Figures 2 -6 show for each station, (a) the
O ne core w as used for granulom etric analysis. m eiofauna taxa in the whole core, (b) the meio
Prelim inary drying at 100 °C was necessary to fauna taxa, excluding nem atodes, (c) depth dis
rem ove peat. A m echanical shaker was used to tributions o f nem atodes and copepods, (d) depth
p ass the sand fraction through a graded series of profiles o f m edian grain size and (e) o f m ud co n
stan d ard sieves (B uchanan, 1984). tent, (f) depth distribution o f total meiofaunal
Sedim ent tem perature w as m easured by p ush density.
depth (cm)
a m p h ip o d a
oo em b o a
.W W W W Tÿl
gastrotricha
<
c
kinorhyncha
n em atoda
w
cc without nematodes
oligochaeta CO
X
ostraco d a CO
polychaeta VI
rotifera \ \ l
tardigrada □ nematodes
turbellaria ]
■ copepods
larvae of insects
nauplii 3
S3
0 500 1,000 1,500 2,000 2,500 o 50 100 150 200 250 300 350 0 200 400 600 800
absolute density absolute density density
lenght of total core : 20 cm A B total density copepods : 215 C
Avicennia total density nematodes : 2329
depth (cm) D depth (cm) E depth (cm)
67.67% of total meiofauna
F
\\W \W
4-5
WWWWWWN
5-7.5
7.5-10
10-15
0 0.1 0.2 0.3 0.4 0.5 0.6 o 1 2 3 4 5 6 200 400 600 800 1,000 1,200
median (mm) % mud density
meiofauna from all depths : 3442
Fig. 2. Avicennia site (a): density of all meiofauna taxa in the whole core (ind./10 cm2), (b): density of the meiofauna, excluding
the nematodes (ind./IO cm2) (c): depth distribution patterns of densities of nematodes and copepods (d): depth profiles of median
grain size of the sand fraction (in mm) (e): depth profile o f the percentage of mud (% ) (f): depth distribution of the total densi
ties of meiofauna.
102
d epth (cm)
a m phipoda
calanoidea 0-1
cladocera
collem bola 1-2
copepoda \ \ \ 1
g astro p o d a 2-3
gastrotricha S3
halacaroidea 1 3-4
isopoda
kinorhyncha 4-5
nem atoda without nematodes
o ligochaeta 3 5-7.5
ostraco d a
polychaeta ] 7.5-10
rotifera
ta rdigrada
10-15
H nematodes
turbellaria
larvae of insects ■ copepods
nauplii 3
1,000 2,000 3,000 4,¿00 1,000 1,200 0 200 400 600 800 1,000 1,200 1,400
absolute density absolute density density ^
A B total density copepods : 1113
lenght of total core : 15 cm Bruguiera total density nematodes : 4205
62.69 % of total meiofauna p
depth (cm) D depth (cm) E depth (cm)
0-1 0-1
1-2 1-2
2-3 2-3
3-4 3-4
4-5 4-5
4-5
5-7.5 5-7.5
7.5-10 7.5-10
5-7.5
^ \\\\\\^ ^
7.5-10
10-15 10-15
Fig. 3. Bruguiera site (a): density of all meiofauna taxa in the whole core (ind./10 cm2), (b): density of the meiofauna, excluding
the nematodes (ind./10 cm2) (c): depth distribution patterns of densities o f nematodes and copepods (d): depth profiles of median
grain size of the sand fraction (in mm) (e): depth profile of the percentage of mud (% ) (f): depth distribution of the total densi
ties of meiofauna.
Meiofauna Taxa
Seventeen m eiofaunal taxa are recorded. N em a
Total densities todes dom inate: 77.8% (Avicennia), 53.5% (Bru
T otal densities are high in all five stations: guiera), 85.5% (Ceriops), 89.0% (Rhizophora)
6707 ind. 10 c m “ 2 in Bruguiera, 3998 ind. and 95.1% (Sonneratia). C opepod densities, p ar
10 c m “ 2 for Rhizophora, 3442 ind. 10 c m “ 2 for ticularly harpacticoids, are low in com parison
Avicennia, 2889 ind. 10 c m “ 2 in Sonneratia and with nem atodes (less than 1 % o f total density).
1976 ind. 10 c m “ 2 in Ceriops. O ther com m on taxa are turbellarians, oligo-
Except for Bruguiera, the density o f m eiofauna chaetes, polychaetes, ostracods and rotifers which
is highest near the sedim ent surface, and espe occur in high num bers in som e areas. Few gas-
cially in Sonneratia where the upper centim eter trotrichs, kinorhynchs and tardigrades are re
contains 50% o f the total (Fig. 6F). In Ceriops corded.
(Fig. 4F ) and Rhizophora (Fig. 5F), total m eio In Figs 2 -6 (A, B) no distinction is m ade be
fauna fluctuates with depth, and in Bruguiera tw een adults and larvae o f polychaetes, though
(Fig. 3F) a m axim um is reached at 4 -5 cm. In high num bers were found for the larvae in A vi
Avicennia (Fig. 2F), again, higher densities occur cennia and Bruguiera'. 93% and 78% respectively.
in the top centim eter. The other sedim ents show rem arkable low er d ata
103
depth (cm)
a m phipoda
calanoidea 0-1
clad o cera
collem bola 1-2
copepoda g
gastro p o d a 2-3
gastrotricha
halacaroidea 3-4
iso p o d a
kinorhyncha 4-5
nem ato d a without nematodes
oligochaeta 3 5-7.5
o straco d a xa
polychaeta
rotifera
S3 7.5-10 Q nematodes
tardigrada 3 10-15
■ copepods
turbellaria
larvae of Insects
\\w x
15-inf
nauplii
XX
500 1,000 1,500 0 100 200 300 400
absolute density absolute density density
lenght of total core 22.5 cm A B total density copepods : 61 C
Ceriops total density nematodes : 1709
86.49% of total meiofauna
depth (cm)
D depth (cm)
E depth (cm) F
0-1 j - j I 0-1 \\\\\\\\\\\\\\\)
1-2 , 2 -......... ‘j 1-2
XWWWWN
2-3 2-3
J 2-3 XXXXN
3-4 3-4 3-4 X \\\\\\\\\\N
4-5 4-5 4-5 X \\\\\\\\\\\\3
5-7.5 5-7.5 5-7.5 \\\\ \ \\ \\ \\ \\ \1
7.5-10 ............................................ 1 7.5-10 7.5-10 NXXXXXXI
10-15
: ...........:...... ................................................................... i 10-15 - :w
.... ........................
10-15
x \\\\\\X
lö-inf '' 1 15-inf ■ i 15-Inf.
x \ \\\\\ \ \\ \ \ \\ \ \ \ \\ \ \ 3
........................................................................ . . . .
100 200 300 4
Fig. 4. Ceriops site (a): density of all meiofauna taxa in the whole core (ind./IO cm2), (b): density of the meiofauna, excluding the
nematodes (ind./10 cm2) (c): depth distribution patterns of densities of nematodes and copepods (d): depth profiles of median grain
size of the sand fraction (in mm) (e): depth profile of the percentage o f mud (% ) (f): depth distribution of the total densities of
meiofauna.
(as an extreme 2 o f the 85 polychaetes of Sonner drop. F rom 2 -3 cm on, only nem atodes and
atia are larvae). copepods rem ain abundant, respectively 83 and
Although no attem pt is m ade to detail the dif 7% . At 7 .5 -1 0 cm copepods reach a m axim um of
ferent taxa, a lot o f epsilonem atid nem atodes are 150 ind. (abundance: 37% ), while nem atodes fol
counted, especially in Bruguiera sedim ents where low a decreasing trend (223 ind.; 55% ). Densities
they m ake up about 70% o f nem atode num bers. o f total m eiofauna quickly drop with depth
(Fig. 2f). Nevertheless there are still 161 nem a
Depth profiles todes in the sedim ent beneath 10 cm.
depth (cm)
am phipoda
calanoidea
cladocera
collembola SXXXXXXX3
co p e p o d a
g astro p o d a
gastrotricha
halacaroidea <
isopoda c
kinorhyncha
co
nem atoda cc without nematodes
o ligochaeta CC XVÑ]_______
polychaeta
rotifera
tardigrada H nematodes
turbellaria
larvae of insects H copepods
nauplii 15-inf.
1,000 2,000 3,000 o 20 40 60 80 100 120 140 o 200 400 600 800
absolute density absolute density density
lenght of total core : 17 cm A total density copepods : 83 C
Rhizophora total density nematods : 3500
depth (cm)
D depth (cm) depth (cm)
89.27 % of total meiofauna
F
0-1 - ..................I 0-1 : .......... j 0-1
wwvww wwn
1-2 I 1-2 I 1-2 \W \W \N
2-3 2-3 : _________________________ _ J 2-3 ^ x \\\\\\\%
3-4 3-4 3-4
N \\\\\\\\\\\\\\\N
4-5 4-5 4-5
wwwwwwx
5-7.5 5-7.5 5-7.5
x \\\\\\\\\\\b
7.5-10 7.5-10 7.5-10
1_________________________ : ......- . . . . - . . V J
10-15
........... I
..............................................................................
10-15 10-15
\\W \N
15-Inf
------1------L .. 1 .1 I 1 : 1 , 1 ,
15-inf
i : i !. i i
15-inf.
3
0.15 0.2 0.25 0-35 0 1 2 3 4 5 6 0 200 400 600 800 1,000
median (mm) % mud density
meiofauna from all depths : 3998
Fig. 5. Rhizophora site (a): density of all meiofauna taxa in the whole core (ind./10 cm2), (b): density of the meiofauna, exclud
ing the nematodes (ind./IO cm2) (c): depth distribution patterns of densities of nematodes and copepods (d): depth profiles of
median grain size of the send fraction (in mm) (e): depth profile of the percentage of mud (% ) (f): depth distribution of the total
densities of meiofauna.
depth (cm)
a m phipoda
calanoidea
cladocera
collem bola
copepoda \_\X_3
gastrotricha
halacaroidea
isopoda
kinorhyncha
n em atoda without nematodes
oligochaeta
o straco d a 3
polychaeta 7.5-10
rotifera
tardigrada S nematodes
turbellaria X]
larvae of insects copepods
nauplii :
1,000 2,000 3,000 100 150 200 400 600 800 1,000 1,200
500 1,500 2,500 absolute density density
absolute density
A B total density copepods : 51
lenght of total core : + 15cm Sonneratia total density nematodes : 2555
0-1 0-1
XW W W W W W W W W ^
1-2 1-2
2-3 2-3
0 0.1 0.2 0.3 0.4 0.5 0.6 0 2 3 4 5 6 400 800 1,200 1,6C
median (mm) >mud 600 1,000 1,400
density
meiofauna from all dephts : 2889
Fig. 6. Sonneratia site (a): density of all meiofauna taxa in the whole core (ind./10 cm2), (b): density of the meiofauna, excluding
the nematodes (ind./10 cm2) (c): depth distribution patterns of densities of nematodes and copepods (d): depth profiles of median
grain size of the sand fraction (in mm) (e): depth profile o f the percentage of mud (% ) (f): depth distribution of the total densi
ties of meiofauna.
shows high densities in the upper layer (1443 ind.; Sedim ent
50% o f total density), but a very quick drop
deeper in the sedim ent. C opepods only occur in Figures 2 - 6 (D , E) show the vertical profile of
the first layer (48 ind.). m edian sedim ent grain size and m ud content. The
m ean m edian particle size o f each core fluctuates
In conclusion. Although densities tended to be betw een 0.266 m m (Rhizophora) and 0.468 mm
higher in the upper layers, several taxa were dis (Avicennia); consequently all sedim ents are clas
tributed in high densities in deeper sedim ent in sified as m edium sand. A co n stan t m edian grain
tervals as well. Regarding the deepest interval size vertical profile is shown for Avicennia sedi
Sonneratia has the low est am ount o f individuals m ents (except for the upperm ost and deepest in
(2% ), followed by Avicennia (3.2% ), Rhizophora tervals), Bruguiera and Ceriops', Rhizophora and
(6.5% ), Ceriops (8 % ) and Bruguiera (11% ). Both, Sonneratia have an increasing trend in m edian
nem atodes and copepods live at sedim ent depths grain size o f the sand fraction with depth. The
deeper than the one examined. Especially Ceriops m ud percentage o f Avicennia and Sonneratia is
show s a high nem atode and copepod num ber in higher in the top centim eter in com parison with
the deepest layer (15-22.5 cm: 342 nem atodes, 11 deeper intervals. U nlike the stable m edian grain
copepods). size, m ud content show s larger variation, with
106
highest values at 0 -1 cm and 15-inf. cm (Sonne Table 1. D e n s itie s o f m e io fa u n a in m a n g ro v e s (in d ./IO c m 2).
The taxon com position o f the m eiofauna in the in the surveys o f A ustralian m angroves (H o d d a
different vegetation sites is m ostly related with and N icholas, 1985; N icholas e ta l., 1991) and
sedim ent grain size, and probably to som e extent South African m angroves (Dye, 1983b). N em a
w ith oxygen conditions. Sedim ents with higher todes seem to be less im portant in C uban m an
m ud content are m ainly inhabited by burrow ers groves (L alana-R ueda & G osselck, 1986: 35.3
such as oligochaetes, nem atodes and turbellari- and 61.1% ).
ans. T he sedim ent o f Bruguiera contains a low Alongi (1987a, b), H o d d a & N icholas (1985)
m ud percentage, and consequently shows a higher and Dye (1983a, b) noted a low copepod density
diversity o f taxa. Rhizophora and Sonneratia- (respectively 1-22 ind. 10 c m - 2 , under 10% and
sam ples, with finer sedim ents, are mainly com 3 0 -8 0 ind. 10 c m - 2 ). L alana-R ueda & G osselck
posed of nem atodes. Small m eiofauna such as (1986) in contrast found the copepods to be the
rotifers, gastrotrichs, nauplii and larvae o f poly second m ost com m on taxon in term s o f indivi
chaetes are not linked with sedim ent character dual dom inance.
istics. A pparently the activity o f currents is res U nlike m ost other studies, turbellarians were
ponsible for the distribution o f these anim als. the m ost abundant m eiobenthic group in the d ata
C opepods and epsilonem atid nem atodes do not o f Alongi (1987a), probably due to the used ex
w ithstand high silt fractions. traction technique. H o d d a & N icholas (1985)
W ith the exception o f Bruguiera no evident re only found the oligochaetes as abundant group
lationship with depth distribution is detected. The next to nem atodes and copepods, while Dye
decreasing density w ith depth in Sonneratia is (1983a, b) recorded 6.4% ciliates, followed by oli
probably governed by low oxygen conditions. gochaetes (4.5% ), turbellarians (3% ) and ki-
norchychs (2.7% ). C opepods, polychaetes, gas
trotrichs, larvae o f crustaceans and insects only
Discussion appeared sporadically. The five m ost im portant
taxa in the T oledo Lagoon were nem atodes,
Sheltered regions with m uddy sedim ents rich in copepods, T anaidacea, oligochaetes and poly
detritus generally are characterised by low diver chaetes and in the B asto Lagoon nem atodes,
sity, b u t high densities (H eip eta l., 1985; Coull, copepods, ostracods, am phipods and poly
1988; Giere eta l., 1988). In the present study chaetes (L alana-R ueda & G osselck, 1986).
m ud percentages reached 5 % and total densities Am ong the different sam pling sites, Rhizophora
up to 6707 ind. are recorded, though a high num and Sonneratia sedim ents contained the highest
ber o f taxa (17) is found. Table 1 gives a sum m ary abundance (respectively 89 and 95% ) o f nem a
o f densities recorded in several studies on the todes, but also h ad the highest m ud content (re
m eiobenthos o f m angroves (range is between 36 spectively 4 and 3% ). Low copepod percentages,
and 12000 ind. 10 c m - 2 ); values o f G azi are in both in G azi and the study areas cited above,
between the recorded values. suggest th at this taxon is m ore related to coarser
N em atodes were also the m ost ab u ndant taxon grain texture, and th at nem atodes show a broad
107
tolerance to diverse sedim ents. Interstitial groups, densities are higher in these ‘biogenetic depres
e.g., G astrotricha, T ardigrada are typically ex sions’. On the other hand abundances tend to be
cluded from the m uddy sedim ent. Sedim ents with lower due to the accum ulation o f toxic com po
a m edian particle size smaller than 125 mm are nents and the fouraging activities o f the fiddler
usually dom inated by burrow ers (Coull, 1988). crabs themselves. A study on their influence on
In m uddy sedim ents, alm ost no life persist energy flow in the m angrove ecosystem is no t yet
under a depth o f 8 cm (M cIntyre & W arwick, analysed, but a similar phenom enon is present in
1984). T he present study, however, shows that the m angroves o f northeastern A ustralia, where
high densities, especially o f nem atodes, occur be the sesarm id crab Sesarma messa has a high im
neath this level. pact on the weight loss from leaves by their leaf
In a study o f vertical distribution o f m eiofauna burying and consum ing activities (R obertson,
in m angrove sedim ents in Transkei, Dye (1983a) 1986), and in the m angroves o f Q ueensland,
found also th at m eiofauna occurs to considerable where the crab Sesarma erythrodactyla enhances
depths, although highest densities occur in the the decom position o f m angrove litter (Camilleri,
upper 10 cm o f the sedim ent. H e calculated that 1989).
when one restricts m eiofauna sam pling to the top This pilot study o f m eiobenthic com position in
5 cm o f the sedim ent only 40% o f the total den the m angroves o f G azi Bay will be followed by a
sity can be obtained. m ore intensive study investigating the quantity (in
A recent study by N icholas etal. (1991) co n the light of tidal regime and seasons) and quality
firms those findings: about 85% o f the nem atodes (identification o f the m ost im portant taxa and
occur in the top cm o f soft m ud, but 5 -7 species trophic groups) o f som e ecological concepts like
inhabit the deeper anoxic m ud dow n to 10 cm. predation, consum ption and interactions between
Low oxygen conditions are supposed to be an m acrofauna and m eiobenthos.
im portant regulator for the low densities versus
depth o f Sonneratia. Coull (1988) asserts th at ver
tical distribution patterns typically are controlled Acknowledgements
by the depth o f the redox potential discontinuity
(R P D ) level, i.e., the boundary between aerobic T he authors acknowledge the K enyan M arine
and anaerobic sedim ents, which is prim ary deter Fisheries and R esearch Institute in M om basa
m ined by the ‘super variable’ oxygen. In this con (K enya) for providing logistic facilities for the field
text W ieser et aí. (1974) claim harpacticoid cope w ork. D r A. Vanreusel and D r P. A erts are
pods are the m ost sensitive m eiobenthic taxon to thanked for the sam pling assistance. This study
decreased oxygen (this in com bination with the is financed by the Belgian N ational Science
previously discussed sedim ent factor). N em a F oundation (F K F O 2.0043.88) and the EEC
todes seem to be m ore tolerant to reducing cir project TS2-0240-C (G D F ).
cum stances (B oaden, 1975 in H o d d a & N icholas,
1985).
Fiddler crab Uca spp. are a prevalent aspect of References
the K enyan m angroves (personal observation of
Alongi, D. M., 1987a. Intertidal zonation and seasonality of
the authors). D eP atra & Levin (1989) found a
meiobenthos in tropical mangrove estuaries. Mar. Biol. 95:
fundam ental influence o f the Uca burrow s on the 447-458.
abundance and distribution o f several m eiofaunal Alongi, D. M., 1987b. Inter-estuary variations and intertidal
tax a in a N orth C arolina salt m arsh. B oth by zonation of free-living nematode communities in tropical
mangrove systems. Mar. Ecol. Prog. Ser. 40: 103-114.
passive deposition and enhancem ent o f the micro
Alongi, D. M., K. G. Boto & F. Tirendi, 1989. Effects of
environm ent, e.g., food resources, stagnation o f exported mangrove litter on bacterial productivity and dis
w ater (which provides suitable habitats for the solved organic carbon fluxes in adjacent tropical nearshore
developm ent o f young stages o f several taxa), sediments. Mar. Ecol. Prog. Ser. 56: 133-144.
108
Andersen, F. O. & B. T. Hargrave, 1984. Effects o f Spartina Heip, C., M. Vincx & G. Vranken, 1985. The ecology of
detritus enrichment on aerobic/anaerobic benthic metabo marine nematodes. Oceanogr. Mar. Biol. ann. Rev. 23:
lism in an intertidal sediment. Mar. Ecol. Prog. Ser. 16: 399-489.
161-171. Herrera, A. & R. del Valle, 1980. Características de la meio
Benner, R. & R. E. Hodson, 1985. Microbial degradation of fauna bentösica en la laguna y zona costera de Guayabal
the leachable and lignocellulosic components of leaves and en relación con el grado de contaminación. Ciencias Bio
wood from Rhizophora mangle in a tropical mangrove lógicas 5: 29-45.
swamp. Mar. Ecol. Prog. Ser. 23: 221-230. H odda, M. & W. L. Nicholas, 1985. Meiofauna associated
Boaden, P. J. S., 1975. Anaerobiosis, meiofauna and early with mangroves in the Hunter River Estuary and Fullerton
metazoan evolution. Zool. Ser. 4: 21-4. Cove, south-eastern Australia. Aust. J. mar. Freshwat. Res.
Boaden, P. J. S. & R. Seed, 1985. An introduction to Coastal 36: 41-50.
Ecology, Chapter 7, Blackie USA: Chapman & Hall, N.Y., Lalana-Rueda, R. & F. Gosselck, 1986. Investigations of the
218 pp. Benthos of Mangrove Coastal Lagoons in Southern Cuba.
Boto, K. G., J. S. Bunt & J. T. Wellington, 1984. Variations Int. Revue ges. Hydrobiol. 71: 779-794.
in mangrove forest productivity in northern Australia and Lugo, E. L. & S. C. Snedaker, 1974. The ecology of m an
Papua New Guinea. Estuar. coast. Shelf Sei. 19: 321-329. groves. Annu. Rev. Ecol. Syst. 5: 39-64.
Buchanan, J. B., 1984. Sediment analysis. In N. A. Holme & McIntyre, A. D. & R. M. Warwick, 1984. Meiofauna tech
McIntyre (eds), M ethods for the Study of Marine Benthos. niques. In N. A. Holme & McIntyre (eds), M ethods for the
IBP No 16, 2nd edn. Blackwell Scientific Publications, study of marine benthos, Blackwell Scientific Publications:
387 pp. 217-244.
Camilleri, J., 1989. Leaf choice by crustaceans in a mangrove Nicholas, W. L., J. A. Elek, A. C. Stewart & T. G. Marples,
forest in Queensland. Mar. Biol. 102: 453-459. 1991. The nematode fauna of a temperate Australian man
Coull, B. C., 1988. Ecology of the marine meiofauna. In R. P. grove mudflat; its population density, diversity and distri
Higgins & H. Thiel (eds), Introduction to the study of meio bution. Hydrobiologia 209: 13-27.
fauna: 18-38. Odum, W. E., 1988. Comparative ecology on tidal fresh water
Cundell, A. M., M. S. Brown, R. Stanford & R. Mitchell, and salt marshes. Annu. Rev. Ecol. Syst. 19: 147-176.
1979. Microbial degradation of Rhizophora mangle leaves Okemwa, E. N., R. K. R uw a& P. Polk, 1986. The autecology
immersed in the sea. Estuar. coast, mar. Sei. 9: 281-286. of the edible oyster Crassostrea cucullata Born, 1778: size
Decraemer, W. & A. Coomans, 1978. Scientific report on the related vertical distribution at Mkomani, M ombasa. Kenya
belgian expedition to the G reat Barrier Reef in 1967. Nem J. Sei. B 7: 9-14.
atodes X II. Ecological Notes on the Nematode Fauna in Robertson, A. I., 1986. Leaf-burying crabs: their influence on
and around Lizard Island. Aust. J. mar. Freshwat. Res. 29: energy flow and export from mixed mangrove forests
497-508. (Rhizophora spp.) in northeastern Australia. J. exp. mar.
DePatra, K. D. & L. A. Levin, 1989. Evidence of the passive Biol. Ecol. 102: 237-248.
deposition of meiofauna into fiddler crab burrows. J. exp. Ruwa, R. K., 1990. The effects of habitat complexities created
mar. Biol. Ecol. 125: 173-192. by mangroves on macrofaunal composition in brackishwa-
Dye, A. H., 1983a. Vertical and horizontal distribution of ter intertidal zones at the Kenyan coast. Discovery and
meiofauna in mangrove sediments in Transkei, Southern Innovation 2: 49-55.
Africa. Estuar. coast. Shelf Sei. 16: 591-598. Ruwa, R. K. & P. Polk, 1986. Short communication. Addi
Dye, A. H., 1983b. Composition and seasonal fluctuations on tional information on mangrove distribution in Kenya: some
meiofauna in a southern african mangrove estuary. Mar. observations and remarks. Kenya J. Sei. B 7: 41-45.
Biol. 73: 165-177. Steinke, T. D. & C. J. Ward, 1988. Litter production by man
Dye, A. H. & T. A. Lasiak, 1986. Microbenthos, meiobenthos groves. II. St. Lucia and Richards Bay. S. Afr. J. Bot. 54:
and fiddler crabs: trophic interactions in a typical mangrove 445-454.
sediment. Mar. Ecol. Prog. Ser. 32: 259-264. Twilley, R. R., A. E. Snedaker & C. Pattersen-Zucca, 1986.
Gallin, E., E. Coppejans & H. Beeckman, 1989. The man Litter production and turnover in basin mangrove forests in
grove vegetation o f Gazi Bay (Kenya). Bull. Soc. r. Bot. southwest Florida. Ecology 67: 670-683.
Belg. 122: 197-207. Wieser, W., J. Ott, F. Schiemer & E. Gnaiger, 1974. An
Gerlach, S. A., 1958. Die Mangroveregion tropischer Küsten ecophysiological study of some meiofauna inhabiting a
als Lebensraum. Z. Morph. Ökol. Tiere 46: 636-730. sandy beach at Bermuda. Mar. Biol. 26: 235-249.
Giere, O., A. Eleftheriou & D. J. Murison, 1988. Abiotic fac Woodroffe, C. D., 1987. Pacific Island mangroves: distribu
tors. In R. P. Higgins & H. Thiel (eds), Introduction to the tion and environmental settings. Pacific Science 41: 1-4.
Study of Meiofauna: 61-78.
Hydrobiologia 2 4 7 : 109-117, 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 109
© 1992 Kluwer Academic Publishers. Printed in Belgium.
K ey words: Crassostrea cucullata, Fourier analysis, shell form, shell length, regression analysis,
environm ental factors
A b s tra c t
V ariations o f shell form and shell length were studied for oysters growing in the m angroves o f G azi Creek,
Kenya, and related to different environm ental factors.
F o r the study o f the form, Fourier analysis w as perform ed on the circum ference o f 85 oysters. The
resulting coefficients were com pared am ong specim ens using cluster analysis. The correspondence b e
tween this classification and substrate diam eter is virtually perfect (only one m isclassification out o f 85
oysters). F o r the clusters based on height above chart datum , 9 specim ens, all on interm ediate height
levels, were misclassified. O rientation with respect to tidal current h ad 13 m isclassifications. M angrove
species seemed to influence form only m arginally, if at all.
In the study o f the size o f Crassostrea cucullata, the length o f 956 oysters, growing along tw o transects
were m easured, and correlated with several environm ental factors. O yster length w as no t related to
substrate diam eter or its orientation with respect to the m ain current. Length w as no t influenced by
density up to a cover o f 70% . F o r densities higher than 70% , there w as a fairly strong negative corre
lation (r2 = 0.634, n = 217). Length w as no t correlated with height above bottom (base o f the tree) for
heights lower th an 20 cm, while oysters growing closer to the bottom were smaller. The correlation with
height above ch art datum w as negative but very low (r2 = 0.060, n = 957). How ever, if all m easurem ents
o f oysters closer than 20 cm to the bottom , and all from a density o f m ore than 70% cover are deleted
from the d ata set, the correlation with height increased dram atically, the slope still being negative
(r2 = 0.859, n = 543).
oyster is partly determ ined by the form o f the inflowing current or facing the outflowing current.
substrate. Van Som eren & W hitehead (1961) dis T he diam eter o f the bran ch serving as a substrate
tinguished different categories o f form o f oysters, w as m easured w ith calipers to the n earest mm.
and correlated these form s with different types of T he height above bottom w as m easured to the
habitats, and w ith different types o f substrate. nearest cm.
These observations were however, qualitative. To m easure approxim ate density, we m ade ver
The first quantitative w ork w as done by the above tical transects on each o f the 10 cm vertical bands.
authors. They com pared oysters growing on dif These transects were m ade in four directions (fac
ferent artificial substrates by Fourier analysis. ing N orth, E ast, South and W est). T he approxi
Their conclusion w as th at the diam eter o f the m ate density w as expressed as the percentage of
substrate has an influence on the form o f the shell the vertical transect th at w as covered by oysters.
o f Crassostrea cucullata. H eight above chart datum w as determ ined fol
V arious studies show intraspecific differences lowing the m ethod o f O kem w a eta l. (1986). In
in size of m olluscs in relation to environm ental this m ethod, the highest level the w ater reaches
param eters (Com fort, 1946; Ebling e ta l., 1962; during the tidal cycle is m arked. The height o f this
Vermeij, 1972, 1973; R uw a & Brakel, 1981). level is taken to be the sam e as the one given in
O kem w a eta l. (1986) showed th at such differ the tide table for the sam e tidal cycle, and used
ences also exist in the m angrove oyster as an absolute reference for the rest o f the m ea
Crassostrea cucullata, growing on a rocky substra surem ents. F rom this reference, height above
tum . The present investigation w as undertaken to chart datum w as determ ined to the nearest cm.
study the effects o f different environm ental fac
tors on the size o f this oyster growing in a m an
grove ecosystem . The results o f this study m ay be Shell form
o f im m ediate im portance to the culture of
Crassostrea cucullata. Sampling
85 oysters were collected in a restricted area of
10 X 10 m. This area w as chosen for the availabil
M a t e r i a l a n d m e th o d s ity o f all environm ental conditions of interest.
‘C ollection’ w as done by m aking a photograph of
O bservations were m ade between July and Sep the oyster. These photographs were taken so th at
tem ber 1989 in a small creek near G azi, a small the oyster filled only l¡9 o f the surface o f the image,
village 49 km south o f M om basa (Kenya). Seven to avoid distortion.
species o f m angrove trees occur at G azi (G allin The sam ple taken w as stratified with respect to
eta l., 1989): Avicennia marina (F orsk.) Vierh., environm ental param eters. Only oysters th at were
Bruguiera gymnorrhiza (L.) Lam ., Ceriops tagal growing isolated (i.e. not touching any other oys
(Perrotet) C. B. R obinson, Lumnitzera racemosa ter) were included in the sample. In this way,
W illd., Rhizophora mucronata Lam ., Sonneratia crow ding did n o t interfere w ith the analysis: all
alba J. Smith and Xylocarpus granatum Koenig. oysters included in the sam ple h ad the o p p o rtu
A n elaborate description o f this creek can be nity to grow to their full size.
found in T ack (1990). The environm ental param eters studied were
(1) species o f the m angrove tree serving as a sub
strate, (2) diam eter o f the branch or pneum ato-
The environmental factors phore on which the oyster w as growing, (3) height
above chart datum and (4) orientation w ith re
O rientation with respect to the m ain tidal current spect to the m ain tidal current.
w as tak en as either perpendicular or parallel. P er F our species o f m angrove trees were present in
pendicular w as further subdivided into facing the the sam pling site. T he num ber o f oysters taken
Ill
Cluster analysis
The coefficients resulting from the Fourier tra n s
form were entered in a cluster analysis, to detect
groups o f oysters o f similar form.
W ard’s m ethod (W ard, 1963; W ard & H ook,
1963) w as used for clustering. In this m ethod, the
3- distance m easure is the squared Euclidean dis
tance. The m ethod o f aggregation is a m odifica
tion o f the weighted average m ethod; the disper
2.2 2.6 3.0 3.4
sion o f the elem ents o f a cluster around its centre
Height above datum ( m ) of gravity is taken into account w hen determ ining
Fig. 1. which clusters are to be com bined.
112
T he cluster analysis w as perform ed on a M a cating the environm ental conditions in w hich the
cintosh com puter, using the program S P S S /P C . oyster w as found, as explained in the caption to
The distances betw een different groups are given this figure. O ysters less than one unit distance
in arbitrary units. Since only relative distances apart are displayed on one line.
are im p o rtan t here, this does n o t interfere w ith the In this dendrogram , a first division o f the sam
interpretation o f the results. ple into tw o groups o f m ore or less equal size
occurred on the distance level 25. These two
groups coincided with oysters th at occurred high
Shell length above datum or low above datum . O ysters grow
ing in interm ediate levels (betw een 3.05 and
O bservations were m ade on C. cucullata growing 3.13 m above datum ) occurred in bo th groups of
on the trunks and stilt roots o f m angrove trees the cluster analysis. If the limit for discrim inating
covering tw o areas (4 x 20 m ) on opposite sides ‘H igh’ and ‘L ow ’ is set at 3.07 for oysters in the
o f the m ain channel. 20 trees were growing in our field, 9 m isclassifications occurred (i.e. oysters
study area: 12 R . mucronata, 2 A . marina and 6 th at w ere actually growing low in the tidal zone,
S. alba. but were classified w ith the oysters growing high
E ach tru n k and stilt ro o t w as divided into co n in the tidal zone by the cluster analysis). A con
secutive 10 cm vertical bands starting from the tingency table w as constructed, where one o f the
b ottom o f the tree trunk. In each vertical ban d we nom inal variables is the actual characteristic of
collected, if present, an oyster growing on the the specim en, and the other the predicted char
n orthern, the eastern, the southern and the w est acteristic based on the m em bership to a group in
ern side o f the tru n k or stilt root. the cluster analysis. This procedure perm itted the
The shell length o f all collected oysters was calculation o f a Chi square value, indicating the
m easured with calipers to the nearest 0.1m m . significance o f the association betw een observed
Length w as defined as the distance from the cen and predicted characteristics. T he Chi square
tre o f the hinge to the farthest point o f the o p p o value and its probability are given in Table 1.
site end. F o r each o f the m easurem ents, a num T able 1 also gives the value for L am bda,
ber o f environm ental variables were recorded: G oodm an and K ruskal’s m easure o f predictive
(1) species o f the m angrove tree and diam eter of ability (Everitt, 1977). This coefficient is a m ea
the bran ch serving as a substrate, (2) approxi sure o f association betw een tw o nom inal varia
m ate density o f the oysters, (3) height above the bles, and ranges from 0 to 1 (0 indicating no
bottom (base o f the tree trunk), (4) height above association, 1 a perfect association). T he signifi
ch art datum , and (5) orientation with respect to cance o f L am bda is the sam e as the significance
the m ain tidal current. o f the contingency table on w hich it is based.
A regression analysis w as perform ed to try to
T ablet. Statistics of the number of misclassifications. The
correlate length with each o f these variables sep
first three tests are based on a 2 x 2 table of predicted and
arately. These variables were taken as the m ain actual group membership. The last two lines are based on
factors to study in relation to oyster length. respectively a 16 x 4 and a 16 x 2 table of frequency of occur
rence of the different values for the tree species in the 16
different groups of Fig. 2.
Results and discussion
Environmental parameter Chi square P Lambda
Gr oup S p e c i me n Co d e s Distances
Character is t ic
O 5 IO 15 20 25
zs 1 z s i r zs 1r zs 1r zs Ha Bs 1b zs !s zs 1b —\
as 1 bs i fo fos 1s a s 1r as i § s s Hs bs !b Zs 1r — \
ZS 1 zs i b zs i s As 1s zs l a zs 1s z s 1b - / i — -—\
as 1 a s 1r bs 1r Zs l a bs i r bsHs a s 1b a s Hr as Sb s sla —-/ ¡
t! ■- \
\
a 11 a 1Hs bi I r bl Ir - \ I 1
a 1i b! i b a l Ir 4- . -- / !
z!1 i.- 1
! 1s z 1! s - 1 1
a i1 a l l a Z 1 l a a l l a fol 1b b! l a - / {
i
a Ih fol hr Z 1ha bl ha b! ha a 1ha - \ 1
i îh z 1ha z ! h s Bl hs - +-------- — \ i
alh a ! hr fol hb a 1La a 1ha Z 1La a 1ha - / 1 i
! ---------- /
zsh Zs hs a s hr - \ }
ash a s b b a s L b aLhs Z s h r bs h b fos h b 3 S hr b S Lb --------- y ¡
zsh z s h s Bs hr z s h r ZS hr z s h b — \ ! - ■ ------------ ----- /
ash bshs bs ha b s h a a s h s ashfo Z s h r a 3 hr a s hb fos h b D S hb bshs /
ash Z s ha a s h b a s hr fosha _ //
Parallel' s ; s mal l I ow a Av i c e n n i a
Perpendicular I : large ■i¡ oh b Br u gu i e r a
P e r p e n d f cu 1a r Rh i z o p h o r a
Sonneratia
Fig. 2. Results of cluster analysis. Distance units are arbitrary. Each four-character code represents one oyster. G roup charac
teristics were determined on the basis of majority, a and b are considered together, and written as ‘a’ in the group characteris
tics. Upper case characters are misclassifications (not for tree species). Oysters on one line are less than 1 unit distance apart.
Each o f the tw o above m entioned groups were of the groups o f the cluster analysis with respect
separated into tw o subgroups, on distance levels to orientation, if only tw o categories are consid
o f 9 and 14 respectively. These tw o divisions co ered (parallel or perpendicular). If the group ch a r
incided with the oysters growing on small or on acteristic is taken to be the value o f the m ajority
large diam eters o f substrate. Only one misclassi- in the group, 13 m isclassifications occurred. As
fication occurred for this criterion. Statistics for can be seen from Table 1, this is still a highly
this division (Chi square, p and L am bda) are also significant result, though the value for the index
given in Table 1. of predictive ability w as m uch lower than in
The interpretation for the tw o other environ the tw o cases discussed above. The two values
m ental param eters w as no t as straightforw ard, in for perpendicular (facing outflowing or facing in
the sense th at there w as no strict dichotom y sub flowing current) seemed to have no influence
dividing each o f the groups on levels lower than whatsoever: these values occurred together in
the ones discussed above. How ever, it w as still m ost o f the groups.
possible to assign a group characteristic to each F o r the last environm ental param eter, the spe-
114
35-
Shell length
47.5-1
x: 42.5-
42.5-
I 37.5-
•c 37.5-
o 32.5-
32.5-
ô 27.5- 27.5-
1.6 1.8 2 2.2 2.4 2.6 2.8 3
affect oyster size for densities lower than 70% strate. This w ould interfere w ith the structural
cover. strength o f the shell.
Shell lengths o f C. cucullata are small at a height The outline o f oysters growing higher above
lower th an 20 cm above the bottom as com pared chart datum is ‘w avier’ than the low er growing
to oysters growing higher. This m ay be due to the oysters: their outline h ad m ore lobes, and the
higher sedim ent load closer to the m uddy bottom . lobes were m ore pronounced. It is no t clear w hat
B ecause oysters are filter feeders, they m ay have the reason for this difference is. M aybe it is re
to expend m ore energy in filtering the w ater at lated to the tim e available to the oysters for feed
lower levels to get an equal am ount of food as do ing: the higher above chart datum the oyster is
the oysters growing at higher levels. growing, the shorter the tim e it will be subm erged.
The correlation betw een oyster length and the A wavier outline will increase the size o f the open
height above d atum is negative and very low. ing via w hich w ater can be taken in or expelled,
H ow ever, w hen we rem oved all oysters growing and hence possibly increase the filtering rate. The
lower th an 20 cm above the bottom , and all oys wavier outline o f the higher oysters could be a
ters growing on su bstrata with density greater com pensation for the shorter feeding time.
th an 70% cover, the correlation becom es very O rientation w ith respect to m ain tidal current
strong (r2 = 0.850; n = 543). This high negative seemed to have a w eaker, but still highly signifi
correlation coefficient dem onstrates th at cant correlation w ith form. O nce again, this dis
Crassostrea cucullata exhibits size-related patterns tinction is also apparent from a visual inspection
in its vertical distribution with an up-shore reduc of the outline o f the oysters (Fig. 4). O ysters
tion in shell length. growing perpendicular to the tidal current often
have lobes all around their outline, as opposed to
Conclusions oysters growing parallel to the current, th at have
lobes m ainly on one side.
The form o f the oysters is clearly related to both H eight above chart datum , and thus percentage
height above datum and diam eter o f the substrate. o f tim e im m ersed, seems to be the prim ary factor
Changes o f form associated with differences in determ ining the shell length o f the oysters. This
height above ch art datum were the m ost im por relationship is obscured by the influence o f
tan t, as reflected in the fact th at the tw o groups crow ding and proxim ity to the bottom . C om pe
were separated at the highest level o f distance. tition for space is the obvious m echanism by
Differences o f form resulting from differences of which crow ding can influence the length o f the
diam eter were smaller. The num ber o f m isclassi oyster shell. T he way in w hich proxim ity to the
fications for this last environm ental factor w as bottom is influencing oyster grow th rem ains to be
less in this study, b u t this can be attributed to the investigated, but a probable m echanism is the in
way in which the sam pling w as done (continuous terference o f high sedim ent load with the filter
sam pling for height above datum , as opposed to feeding o f the oysters.
dichotom ised sam pling for substrate diam eter). A s with the correlation betw een form and
The distinction betw een these four groups can height above chart datum , it seems logical to look
be related to the appearance o f the oysters for an explanation o f the correlation betw een shell
(Fig. 3). O ysters growing on small diam eter sub length and height in the tim e an oyster is sub
strates are m ore elongated than the ones growing m erged, and hence feeding time. O ysters growing
on large diam eter substrates. This is w hat one lower have a longer proportion o f the day avail
w ould expect: 30 mm is small com pared to the able for feeding, hence can be expected to grow
size o f the oyster. F or the oyster to grow as wide faster. If the m axim um size an oyster can reach
on this type o f substrate as on the larger diam e is dictated by age, rather th an by size itself, an
ter substrate, the oyster should grow with a large upshore reduction in size is the expected trend.
proportion o f its shell detached from the sub This study has only d em onstrated the existence
117
o f the relationship betw een environm ental param in tropical East Africa. Estuar. coast, mar. Sei. 4: 1-21.
Healy-Williams, N. & D. F. Williams, 1981. Fourier analysis
eters on one hand, and form and size of the oys
of test shape of planktonic foraminifera. Nature 289: 485-
ters on the other. A following step in this research 487.
should be the construction of a functional m odel, Kaesler, R. L. & J. A. W aters, 1972. Fourier analysis of the
to investigate the causal relationship between ostracode margin. Geol. Soc. Am. Bull. 83: 1169-1178.
m orphology and the environm ental param eters, Kesteven, G. L., 1941. The biology and cultivation of oysters
in Australia. I. Some economic aspects. Commonwealth of
and to investigate the hypotheses form ulated
Australia. Melbourne, 32 pp.
above. Kuhl, F. P. & C. R. Giardini, 1982. Elliptic Fourier features
of a closed contour. Computer Graphics and Image Pro
Acknowledgements cessing 18: 236-258.
Lewis, J. R., 1964. The ecology of rocky shores. H odder &
Stoughton, London, 323 pp.
This research w as conducted in the fram ew ork of Okemwa, E. R., R. K. Ruwa & P. Polk, 1986. The autecol-
the Kenyan-Belgian Project in M arine Sciences, ogyofthe edible oyster Crassostrea cucullata Bom, 1778: size-
at the K enya M arine and Fisheries R esearch In related vertical distribution at Mkomani, Mombasa. Kenya
stitute (K M F R I) in M om basa. W e wish to thank J. Sei. Ser. B 7: 9-14.
Ricker, W. E., 1973. Linear regressions in fishery research. J.
b oth M r. S. O. Alíela and Dr. E. Okem w a, re
Fish. Res. Bd Can. 30: 409-434.
spectively form er and current director, o f this in Ruwa, R. K., 1984. Invertebrate faunal zonation on rocky
stitute. W e also th an k M r. R. K R uw a for his shores around M ombasa, Kenya. Kenya J. Sei. Ser. B 5:
cooperation. 49-65.
Ruwa, R. K. & W. H. Brakel, 1981. Tidal periodicity and
size-related variations in the zonation o f the gastropod
References Nerita plicata on East African rocky shore. Kenya J. Sei.
Ser. B 2: 61-67.
Tack, J. F., 1990. Ecomorfologie van de mangroveoester
Anstey, R. L. & D. A. Delmet, 1972. Genetic meaning of
Crassostrea cucullata (Born, 1778). Unpublished M. Sc.
zooecial chamber shapes. Science 177: 1000-1002.
thesis, Free university of Brussels, 99 pp.
Anstey, R. L. & D. A. Delmet, 1973. Fourier analysis of zoo
Taylor, J. D., 1968. Coral reef associated invertebrate com
ecial shapes in fossil tubular bryozoans. Geol. Soc. Am.
munities (mainly molluscan) around Mahe, Seychelles. Phil.
Bull. 84: 1753-1764.
Trans, r. Soc., Lond. Ser. B 254: 129-206.
Chelazzi, G. & M. Vannini, 1980. Zonation of intertidal mol
Taylor, J. D., 1970. Intertidal zonation at Aldabra atoll. Phil.
luscs on rocky shores of southern Somali. Estuar. coast,
Trans, r. Soc., Lond. Ser. B 260: 173-213.
mar. Sei. 10: 569-583.
Trevaillon, A., R. R. C. Edwards & J. H. Steel, 1970. Dy
Christopher, R. A. & J. A. W aters, 1974. Fourier series as a
namics o f a benthic bivalve. In J. H. Steel (ed.), Marine
quantitative descriptor of miospore shape. J. Paleontol. 48:
Food Chains. University of California Press, Los Angeles:
697-709.
285-295.
Comfort, A., 1946. Patella vulgata L. relation of habitat to
Van Someren, V. D. & J. P. Whitehead, 1961. An investiga
form. J. Conch. 22: 216-217.
tion of the biology and culture of an East African oyster,
Day, J. H., 1974. A guide to marine life of South African
shores. Balkema. Cape Town, 300 pp. Crassostrea cucullata. Coi. Off. Fish. Pub. 14: 1-36.
Vermeij, G. J., 1972. Intraspecific shore-level size gradients in
Ebling, F. J., J. F. Sloane & H. M. Davies, 1962. The ecol
intertidal molluscs. Ecology 53: 693-700.
ogy of Lough Ine. XII. The distribution and characteristics
Vermeij, G. J., 1973. Morphological patterns in high intertidal
of Patella species. J. anim. Ecol. 31: 457-470.
gastropods: adaptive strategies and their limitations. Mar.
Ehrlich, R. & B. Weinberg, 1970. An exact method for char
Biol. 20: 319-346.
acterisation of grain shape. J. Sed. Petrol. 40: 205-212.
W ard, J. H., 1963. Hierarchical grouping to optimise an ob
Everitt, B. S., 1977. The analysis of contingency tables. Chap
jective function. J. am. Stat. Ass. 58: 236-244.
man and Haii. London, 128 pp.
Ward, J. H. & M. E. Hook, 1963. Application of an hierar
Ferson, S., F. J. Rohlf & R. K. Koehn, 1985. Measuring shape
chical grouping procedure to a problem of grouping pro
variation of twodimensional outlines. Syst. Zool. 34: 5 9-
files. Educational and Psychological Measurement 23: 6 9-
68.
81.
Gallin, E., E. Coppejans & H. Beeckman, 1989. The man
Younker, J. L. & R. Ehrlich, 1977. Fourier biometrics: har
grove vegetation of Gazi Bay (Kenya). Bull. r. Soc. Belg.
monic amplitudes as multivariate shape descriptors. Syst.
122: 197-207.
Zool. 26: 336-342.
Hartnoll, R. G., 1976. The ecology of some rocky shores
.
Hydrobiologia 247: 119-120, 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 119
© 1992 Kluwer Academic Publishers. Printed in Belgium.
M elckzedeck K. W. O sore
Kenya Marine and Fisheries Research Institute, P.O. B ox 81651, M ombasa
Abstract
G azi Creek is situated at some 40 km south o f M om basa (K enya 4° 25' S, 39° 50' E) sam pling stations
are located in the creek m outh (stn 1), in the inner creek (stn 3) and interm ediate (stn 2). Sam pling is
done twice a m onth; it starts from stn 1 through stn 2 up to stn 3. A 335 ¡im M esh size net is tow ed
in n ear surface w ater for 5 m inutes and the collected sam ple preserved in 5% form aldehyde. H ydro-
graphic param eters are recorded during sampling.
The w ork at G azi is undertaken to survey the Z ooplankton com position both qualitatively and
quantitatively. 22 im portant taxa have been studied to determ ine their seasonal variation in abundance
and distribution. Z ooplankton population are highest in M arch (374 anim als m _3). The abundance
gradually falls during the S.E. m onsoon period (M ay -S ep tem b er) to the lowest value in August (30
anim als m _ 3) (see Fig. 1). C opepoda is the m ost abundant taxon throughout (4 8 .5 -9 2 .4 % ) (see Table 1).
T his group is found throughout the creek. N evertheless, it is evident th at the creek m outh has a higher
diversity o f copepoda com pared to the inner creek. U sing the M argalef Index a consistently higher value
h as been observed in stn 1 com pared to stn 3 (see Table 2). The surface w ater tem perature decreases
during the S.E. m onsoon (28.0 to 25.5 °C ) and it rises during the N .E . m onsoon (29.0 to 35.5 °C). The
tren d in tem perature variation corresponds broadly w ith Zooplankton abundance (see Fig. 1), suggest
ing th at Z ooplankton thrive best in w arm er w ater. High Z ooplankton counts were also observed around
M ay (326 anim als m 3) during the long rains; probably due to high am ounts o f nutrients input. G en
erally, average m onthly pH values vary only slightly but the p H up the creek is alm ost always lower than
at the creek m outh. Salinity is quite co n stan t at 35%0.
5 100
Mar Apr May Jun Jul Aug Sep Oct Nov Dec
M o n th s
Fig. 1. Mean monthly total Zooplankton variation and mean monthly surface water temperature for the three stations (average)
at Gazi during M arch to December 1990.
120
Table 1. Percentage composition of important Zooplankton groups as average of the three stations at Gazi during M arch to
December 1990.
Group 1990
Copepoda 83.36 79.10 62.56 76.49 75.05 76.85 48.49 92.42 _ 84.26
Medusae 0.18 0.16 0.31 0.09 0.00 0.00 0.00 0.04 - 0.00
Brachyura zoea 0.62 11.51 18.85 1.16 5.04 2.50 26.87 0.91 - 2.02
Caridea 0.28 0.06 1.91 0.68 3.86 2.26 6.33 0.76 - 0.35
Chaetognatha 0.17 0.86 2.84 0.42 3.24 0.99 0.71 1.19 - 0.75
O stracoda 0.13 0.58 0.16 0.09 0.28 0.82 0.63 0.19 - 0.27
Mollusca 0.29 1.55 1.68 5.08 4.86 8.41 3.10 0.70 - 1.33
Fish larvae 0.11 0.00 0.19 0.16 0.77 0.38 1.05 0.18 - 0.23
Nauplii 8.59 1.34 0.28 0.16 0.95 4.65 1.83 0.99 - 6.62
Amphipods 0.21 0.43 0.33 1.75 0.68 1.68 1.31 0.38 - 0.48
Appendicularia 0.29 2.08 0.85 0.43 0.68 0.00 0.00 0.63 - 0.11
Polychaeta 0.13 0.84 0.58 0.25 0.94 1.37 0.63 0.16 - 0.45
Isopoda 0.11 0.00 0.24 0.93 0.27 1.03 0.63 0.34 - 0.71
Fish eggs 0.10 0.41 8.06 0.43 1.92 0.82 6.14 0.18 - 1.18
Cumacea 0.16 0.06 0.00 0.34 0.50 0.72 0.26 0.25 - 0.38
Stomatopoda 0.02 0.00 0.06 0.00 0.00 0.27 0.71 0.16 - 0.00
Penaeidae 0.02 0.00 0.06 0.00 0.21 0.03 0.00 0.05 - 0.18
Cirripede nauplii 0.02 0.34 0.00 0.00 0.00 0.03 0.00 0.05 - 0.15
Brachyura megalopa 0.00 0.10 0.54 0.09 0.30 1.03 0.47 0.11 - 0.29
Siphonophora 0.00 0.06 0.64 0.09 0.18 0.00 0.00 0.00 - 0.00
Euphausiacea 0.00 0.00 1.19 0.00 0.00 0.00 0.00 0.00 - 0.00
Sergestidae 0.00 0.00 0.0 0.00 0.09 0.00 0.00 0.00 - 0.00
Others 0.05 0.32 0.00 0.00 0.00 0.03 0.54 0.18 - 1.50
Month Diversity
T. J. D one
Australian Institute o f Marine Science, PM B 3, Townsville M C , Queensland, 4810, Australia
Abstract
M any coral reefs around the w orld have degraded to a degree th at their present intrinsic value and utility
are greatly reduced: (m ass coral m ortality followed by algal invasions; local depletions o f reef fisheries;
deficit o f reef accretion com pared to physical and biological erosion). Though we can som etim es iden
tify proxim al causes (outbreaks o f coral predators and eroders; over-fishing; habitat destruction), we do
n o t have a good understanding o f how population, com m unity and ecosystem structure and function
differ in degraded from un-degraded reefs. T he deficiencies in our understanding limit our ability to
interpret the long-term significance o f reef degradation, and therefore to develop scientifically based plans
for conservation and m anagem ent o f reefs.
A particular course o f action, or lack o f action, based on uncritical acceptance o f any o f the various
views o f tem poral variability can lead to further deterioration o f specific reefs. N one o f these views -
th at reefs are either inherently robust, inherently fragile, or inherently resilient - is true over all tim e-space
scales. This presentation reviews various m odels and case studies which suggest th at reefs can be
knocked precipitously or move slowly from one phase (coral-dom inated) to another (coral-depleted
an d /o r algal dom inated). T ransitions in the other direction (‘recovery’) involve changes (e.g. succession)
in populations and com m unities (of all reef-associated biota, no t ju st sessile benthos), and in reef function
(e.g. com m unity m etabolism , trophodynam ics) which are o f great intrinsic interest but only poorly u n
derstood.
cessive reefs in a n o rth /so u th sequence over the the reef and its resource value to hum an popula
decade 1980-90, causing serious dam age to 65% tions (D ahl, 1974). A m ore protracted switch (de
o f reefs surveyed in a 500 km section o f the cades to centuries) has m ajor im plications for reef
20 0 0 k m long G B R (M oran e ta l., 1988). By the m aintenance and growth, w henever calcium ca r
time those in the south were attacked, there had bonate losses to physical and biological erosion
been substantial coral recovery in the north (D one are not replaced by calcifying organism s such as
e t a l , in prep.). corals and coralline algae.
In this presentation, I discuss w hat degrada
tion m eans. W hat are its sym ptom s and w hat is
its significance in relation to ‘norm al’ tem poral Ecological and human consequences of phase
variability in coral reef ecosystem s? change
The one universal sym ptom o f coral reef deg
radation is m ass coral m ortality followed by an T he im plications th at extensive changes from
invasion o f coral skeletons by algae. This is the coral to algal dom inance have for the re e fs trophic
‘phase shift’ referred to in the title o f this presen structure and resource value to hum an popula
tation. A ssociated with such a change, there m ay tions have received little system atic study. M ost
be significant depletions o f reef fisheries, and a or all the increased algal production may be dis
deficit o f reef accretion com pared to physical and lodged by storm s, thence being exported from the
biological erosion. Thus the reef is degraded in reef (C arpenter, 1990a) or entering benthic m i
term s of its physical attractiveness, its fisheries crobial com m unities on the reef (M oriarty et al.,
yield and its ability to retain its integrity as a 1985; H ansen eta l., 1987). In poorly flushed
breakw ater protecting adjacent shorelines. areas, accum ulation o f algal detritus could lead to
Clearly, the potential social and economic impli excess biological oxygen dem and and anoxic con
cations o f reef degradation will depend on how ditions inimical to survival o f some higher reef
long the reef rem ains degraded. biota (Johannes & Betzer, 1975). N o t surpris
The following tw o quotations define some of ingly, the w orst reported com m unity calcification
the im portant ecological issues and points o f dis rates on reef flats were recorded on coral/algal
agreem ent about the significance o f phase shifts reef flats dom inated by m acroalgae - viz. at
in coral reefs. Referring to the entire G reat B ar Tulear, M adagascar, and M oorea (Pichón &
rier R eef system (som e 2900 reefs stretched over M orrisey, 1985; Sournia, 1976; both cited in K in
2000 km ), D avies (1988) says: sey, 1985).
Increased m acroalgal production m ay affect
‘Its uniqueness i s related to its robustness
harvestable secondary production o f herbivorous
and its ability always to regenerate after ca ta s
fish and invertebrates, and their fish predators.
tro p h es’ (italics added).
F ish production can be increased, either as an
Referring to algal invasions o f previously coral increase in population size (C arpenter, 1990b) or
dom inated areas o f individual reefs, H atcher et al. fish grow th rates, in situations w here food was
(1989) state: previously limiting. H ow ever it m ay be unaffected
in at least tw o circum stances: first, where food
‘Evidence is accum ulating th at coral reef
chains are based m uch m ore on turfs an d /o r p e
benthic com m unity structure can assum e at
lagic Zooplankton w ashed onto the re e fs w ind
least tw o stable forms: one coral-dom inated
w ard m argin (H am ner et al., 1988), than on m ac
and one m acroalgae-dom inated’.
roalgae: and second, w here fish com m unity
A switch to m acroalgal dom inance which lasts structure is strongly determ ined by recruitm ent
years to decades constitutes a fundam ental and com paratively weakly determ ined by food
change in the benthic com m unity structure o f the availability (D oherty & W illiams, 1989). In either
reef with im plications for the trophic structure of case, the significance o f the loss o f live coral may
123
Macroalgae
Algal turfs ?High
on standing nutrients
coral skeletons
? Low
nutrients
Coral larvae
Major disturbance
? High
herbiv. settlement
Intermediate growth
disturbance
com petition with other benthic organism s (e.g. The reefs becam e colonised by a non-reef-building
corals) by inhibiting their grow th or killing them ; assem blage dom inated by m acroalgae (Dictyo
(4) clearing or m odification o f large areas o f sub sphaeria cavernosa) and a variety o f suspension
stratum in a m anner, or at a time which favoured and filter feeding organism s (Sm ith et al., op. cit.).
colonisation by a formerly rare alga. H um ans may In 1977-78, the sewage w as diverted aw ay from
be directly involved, either by affecting am bient the bay and by 1982 there was some decline in
nutrient or sedim ent loads, or by changing fish Dictyosphaeria and some increase both in num
and invertebrate com m unity structure by fishing. bers o f colonies and percent cover o f corals (M a-
ragos eta l., 1985). This example dem onstrates
the capacity o f corals to re-establish populations
Coral to algal phase transitions and the extent of once a m ajor perturbation is removed.
coral recovery
directly affected the reefs and their fish popula recovery o f hydrodynam ics and o f other popula
tions. tions (viz. urchins and dam selfish) have also been
There is still m uch to be learnt about the degree recognised. In the period 1971-1981, the m ac
to which overfishing can change the abundance roalgae Boodlea siamensis (C hlorophyta), Sargas
and com position o f herbivorous fish, predators of sum sp. and Turbinaria ornata (P haeophyta) ex
herbivorous fish or invertebrates, let alone dow n panded their ranges from the fringing reef to the
stream effects on benthic com m unity structure coral dom inated inner and outer barrier reef
(H ay, 1984; M cC lanahan & M uthiga, 1989). (Payri & N aim , 1982). This took place during a
M unro & Smith (1984) attributed m ajor changes period o f dredging o f the coral reef for building
in total fish catch and com position in the C arib m aterial (G abrie eta l., 1985) and rapid develop
bean, including changes in the proportion o f her m ent o f agriculture and tourist accom m odation
bivorous fish, to sustained, intense and relatively adjacent to the shore (Salvat, 1987). It had p er
indiscrim inate fishing. In principle, fishing-in sisted until 1989 (D one et al., 1991) at which time
duced changes in fish com m unity structure m ay m uch o f the substratum occupied by m acroalgae
be expected to have significant effects on benthic com prised standing skeletons o f corals killed by
com m unities. H ow ever fishing pressure is thought Acanthaster planci around 1982 (B ouchon, 1985;
unlikely to affect reef fish com m unity structure in F aure, 1989).
systems with strong recruitm ent lim itation o f fish It is very easy to point the finger o f blam e at the
populations (M unro & Williams, 1985; D oherty coastal developm ent and dredging, but difficult to
& Williams, 1989). prove the connections. T here w as no detectable
elevation o f nutrient levels in the strongly flushed
lagoonal and inner barrier reef w aters in a set of
L a Saline Reef, Reunion sam ples collected in 1976 (R icard, 1982). Levels
o f nitrate + nitrite and reactive phosphate were
Overfishing is also thought to have played a part m uch less than occur in m oderately flushed la
in the degradation o f the reef at L a Saline in goons o f som e G B R reefs 50 km and m ore from
Reunion. A t th at reef, chronic pollution and over shore (F urnas eta l., 1990). This suggests th at if
fishing have been blam ed for a m arked coral-algal nutrient enhancem ent is necessary to sustain the
shift. There h as been a proliferation o f filam en algal biom ass at M oorea, it m ay occur in pulses,
tous and fleshy algae Vaughaniella and Gracilaria such as during dredging o f storm s (resuspension
crassa and the bioeroding sponge Cliona o f reef sedim ents, perhaps enriched by septic tank
inconstans in previously coral-dom inated areas runoff and groundw ater), and wet season runoff
(C uet eta l., 1988). The shift to algal dom inance o f silt-laden w ater, (which retains its integrity to
is believed to result partly from chronic nutrient the outer reef passes (R. Steger, pers. com m .).
pollution o f groundw ater, but also from under Alternatively, or in addition, there m ay be high
grazing by urchins, which were rare, and by fish, nutrient fluxes, despite low am bient concentra
also rare as a result o f daily netting o f the reef. tions, as a result o f very rapid uptake by benthic
The algae killed corals by overgrowth, and it was algae and phytoplankton (A tkinson, 1988).
predicted th at it would also inhibit settlem ent of C oral recolonisation at M oorea w as very
coral larvae. patchy by 1987-1989 (D one eta l., 1991). It was
high on the outside o f the northern barrier reef
and low on the inner barrier reef. In order to
Moorea, French Polynesia understand coral recovery on the inner barrier
reef, one needs to u nderstand som ething o f the
A t M oorea (French Polynesia), nutrients and hydrodynam ics o f the area and the n atural history
overfishing have again been im plicated in reef and population dynam ics o f some key species,
degradation. In addition, the im portance to coral notably sea urchins and dam selfish. O n some
127
areas o f the reef platform , sand abrasion is a gassum holdfasts during their seasonal (sum m er)
m ajor problem for small corals. Only m assive decline (D one & N avin, op. cit.).
Porites corals were abundant in 1987-9, both b e High m acroalgal biom ass on G B R nearshore
cause they were no t eaten by the starfish, and reefs m ay be indicative o f high am bient nutrient
because they can colonise the sand-blasted areas concentrations an d /o r low levels o f herbivory
by fragm entation (Highsm ith, 1980). (R uss, 1984; W illiams & H atcher, 1983). A case
P oor coral settlem ent and survival were also has been m ade for eutrophication o f G B R lagoon
n o ted in patches with dense urchin populations w aters (Bell eta l., 1989), but it is also possible
(especially Diadema setosum and Echinometra th at current standing crops o f m acroalgae may
mathaei). H ow ever there are refuges from urchin not require an anthropogenic subsidy o f nutri
grazing. These are dam selfish (Stegastes nigricans) ents. In a shallow ( < 10 m) nearshore G B R sta
territories which cover about 40% o f the habitat. tion, soluble nutrients (nitrate + nitrite, reactive
C orals (notably Acropora and Pocillopora spp.) phosphate) were strongly correlated with previ
were m uch m ore abundant and diverse inside than ous day’s w ind speed, suggesting the nutrients
outside the territories, probably in large p art due were released from bottom sedim ents disturbed
to S. nigricans’ active exclusion o f urchins and by waves (W alker & O ’D onnell, 1981). H ow ever
herbivorous fish likely to graze off small corals. the degree to which runoff and sewage outfalls
This finding is precisely the opposite to the ob have enriched the sedim ent pore-w ater has not
servations on other dam selfish species (e.g. Potts, been investigated, even though W alker and
1977; Lobel, 1980; K aufm an, 1977), where there O ’D onnell’s station is within a few kilom etres of
is a net detrim ental effect on coral survival inside the sewage outfall o f the largest city adjacent to
territories. the G reat B arrier Reef.
Nearshore and fringing reefs, Great Barrier R e e f Offshore reefs, Great Barrier R e e f
N earshore and island fringing reefs on the G reat O n the G reat B arrier Reef, coral to m acroalgal
B arrier R eef (G B R ) are dom inated over m uch of phase shifts have been caused by m ajor episodic
their shallow, w ave-sw ept slopes and flats by disturbances such as Acanthaster planci and cy
brow n m acroalgae, m ainly Sargassum spp. The clones. A . planci infestations killed close to 100%
significance o f this prolific grow th is unclear. In o f a high cover o f corals over large areas early in
the C aribbean, where Sargassum-dom inated as the 1980s (E ndean & C am eron, 1985; Done,
semblages occupy a similar habitat, heavy wave- 1985). Algal biom ass per hectare increased greatly
action is thought to favour Sargassum through as a result o f increased area for settlem ent (viz.
exclusion o f grazing sea urchins and fish (Adey skeletons o f dead corals). U nfortunately, there
et al., 1977). W ithin recent years to decades, Sar has been no large scale system atic study o f algal
gassum dom inated algal assem blages have both com m unity dynam ics following A. planci dam age.
invaded large areas (hectares) o f G re at B arrier H ow ever it appears th at fish grazing, which usu
R eef fringing reefs previously occupied by living ally keeps algal biom ass low on these reefs (R uss,
coral, and have them selves been displaced by co r 1984; W illiams & H atcher, 1983) w as no longer
als in other areas (T. J. D one & K. F. N avin, able to do so, and th at there w as an increase in
unpubl.). The annual life cycle o f Sargassum gives abundance o f fleshy m acrophytes following the
it an advantage in the w ave-sw ept rubble banks, outbreak (D one, pers. obs.).
but puts it at a disadvantage when the alga is in N evertheless, there has been significant rever
direct com petition for space against corals. F ast sion tow ards coral dom inance in the 6 -1 0 years
lateral grow th o f encrusting Montipora and ta b u since the outbreak. P ercent coral cover and col
late Acropora corals enable them to overgrow Sar ony size frequency distributions determ ined in
128
1990 (D one et al., in prep.) are consistent with re e fs geological history if the killed areas are
good settlem ent and grow th o f corals each year prom ptly recolonised’ (B uddem eier & Hopley,
since the A . planci outbreak, and a high coral 1988).
cover is predicted by the year 2000 (D one et a l,
1988). H ow ever on the inner lagoon o f one reef This is a perceptive statem ent, both acknow ledg
(only 20 0 -3 0 0 m from the well colonised outer ing th at ‘prom pt recolonisation’ will no t neces
slope), there h ad been a very poor recovery (D one sarily follow disturbance, and at the sam e time
eta l., in prep.), possibly due to low densities of counselling conservationists, biologists and m an
planulae arriving in the lagoon, poor settlem ent, agers to seek a longer-term view o f the im plica
an d /o r post-settlem ent survival (see B abcock, tions o f reef change. In som e circum stances, cor
1988). als o f any type are very slow to recolonise
dam aged areas, there being little recovery after a
decade or m ore, e.g. following coral mining at
Sum m ary o f examples M aldives (Brown & D unne, 1988), and following
coral death at M oorea (D one e ta l., 1991). H ow
O f all the examples, the outer slopes o f offshore ever there are other exam ples where dam aged
G B R reefs m ay provide the best ecological cir areas have been prom ptly recolonised by reef
cum stances for coral recovery. There is an ade building corals (P earson, 1981; Colgan, 1987;
quate supply o f planktonic larvae from upstream D one eta l., 1988), and it is apparently on the
reefs and com m unities (Oliver & Willis, 1987; basis o f the latter type o f exam ple th at som e au
Williams et al., 1984), and sufficiently good early thors have been lured to suggest th at corals are
coral grow th and survival throughout the first m ost likely to quickly (decades) establish dom i
6 -1 0 years following the disturbance. By co n nance in any suitable surface which becom es
trast, som e reefs or sections o f reefs m ay rarely be available (e.g. H opley & Kinsey, 1988 - writing
reached by dispersive coral larvae due to their in reference to inundation o f coral reef tops with
rem oteness from source reefs (G lynn, 1985), rising sea level).
or unfavourable reef-scale circulation patterns H ow ever, the definition o f th at point in time
(Black, 1988; cf. Sam m arco & A ndrew s, 1988). when recovery m ay be regarded as ‘com plete’ v ar
In several other examples (e.g. Discovery Bay; ies. R estitution o f pre-disturbance percent coral
M oorea) it w as shown th at it m ay be necessary cover is the m ost cited (see Pearson, 1981) but
to u n d erstan d som ething o f the natural history least stringent criterion o f recovery. The m ost
and population dynam ics o f other reef biota (e.g. stringent - ‘total recovery’ by both ecological and
urchins and dam selfish), the direct effect o f fish geological stan d ard s - w ould be the restitution of
ing on fish com m unity structure, and indirect ef prior diversity, mix o f colony grow th form s (if not
fects on benthic com m unity structure. species), colony size frequency distributions and
structural extent and com plexity o f reef fram e
w ork (after Johannes, 1975). These can take con
Temporal perspectives on phase change siderably longer to recover th an percent coral
coverage (see Fig. 1). F o r example, D one (1988a)
W hile degradation m ay seriously affect the im me and E ndean et al. (1988) suggest th at recovery o f
diate am enity o f a coral reef, it m ay not be sig size-frequency distributions is partially killed and
nificant when viewed from a geological perspec injured populations o f slow growing corals (genus
tive. Porites and other m assive corals) can take a cen
tury or m ore in the sam e habitats in which per
‘Events th at loom large in the eyes o f contem cent coral cover has recovered in the first 1-2
porary observes....m ay be undetectable in the decades (D one e ta l., 1988).
re e fs sedim entary record and trivial in the
129
Problems and scale of management perience and from the literature - is ju st the be
ginning. The uncertainties associated with inter
O ne m anagem ent option which is widely accept preting w hat has happened to a specific reef in the
able is the creation o f m arine preservation areas past, and predicting how it will behave in the
including coral reefs. A reas which are protected future, m ust also be confronted. On the time and
from hum an exploitation can act as reservoirs for space scales relevant to coral reef m anagem ent, it
unprotected reefs dow nstream , where species are is not appropriate to m ake decisions based on a
depleted by fishing, collecting etc. H ow ever the presum ption th at a specific reef or system o f reefs
sam e current systems which carry the larvae of has ‘ability always to regenerate after ca ta stro
corals and fish also carry pollutants, sedim ents, phes’ (D avies, 1988). N either is it correct to as
and the larvae o f destructive organism s such as sume th at a degraded reef will rem ain so forever,
Acanthaster planci and Drupella spp. (coral-eating or th at those reefs which do regenerate will do so
gastropods). It is therefore prudent from a reef in time scales which are convenient to hum an
conservation perspective to create as m any re users o f the reef.
serves as possible, and to m ake them as large as This presentation has touched lightly on the
possible. Experience on the G reat Barrier R eef is m any issues - physical, chem ical and biological
th at m ajor coral kills can take place on reefs scat - which m ust be addressed in trying to under
tered over hundreds o f kilom etres o f the reef tract, stand the causes and consequences o f coral to
equivalent to the length o f the entire Kenyan algal phase shifts in coral reefs. Clearly, a multi
coast. disciplinary approach is desirable for in-depth
understanding o f all complex natural systems.
T he initiatives o f U N E S C O - C O M A R A F in
Conclusions creating ‘The Regional Project for Training and
R esearch on A frican C oastal System s’, and the
While the sym ptom s o f w idespread coral death U niversity o f N airobi in organizing this Sym po
and algal invasion are usually obvious, the causes sium are both welcome and timely.
are frequently m ore a m atter for conjecture than
direct observation. W here a cause is definitely
identified - say, overfishing, siltation caused by Acknowledgements
dredging or land clearing, sewage pollution, toxic
w astes - it m ay be in a country’s best im m ediate I am grateful to the sponsors, executive com m it
econom ic interests to prevent th at im pact. As one tee and co-ordinators of the International Sym
example, H odgson & D ixon (1988) found the rev posium on the Ecology o f M angrove and Related
enue loss to dive tourism caused by siltation o f a Ecosystem s, 2 4 -3 0 Septem ber 1990, M om basa,
Philippine reef area w ould be greater than reve Kenya, for providing the forum for this presen
nue gained by the logging w atersheds w hich would tation. Els M artens’ tireless efforts throughout all
cause the siltation in the first place. As a second stages o f organisation and publication are espe
example, D one (1988b) suggested in relation to cially appreciated. C om m ents on the m anuscript
M auritian fringing reefs, th at one m ust weigh the by Jerem y W oodley and an anonym ous reviewer
cost o f protective m anagem ent against the poten were m uch appreciated. This is A ustralian Insti
tial for reef erosion and dam age to the shore and tute o f M arine Science C ontribution No. 536.
structures.
A scientist offering advice to governm ent on
References
the best course o f action in relation to coral reef
m anagem ent walks a precarious path. Knowledge Adey, W. H., P. J. Adey, R. Burke & L. Kaufman, 1977. The
o f general reef character, history, structure and Holocene reef systems of Eastern Martinique, French West
processes - as can be gleaned from personal ex Indies. Atoll Res. Bull. 218: 1-40.
130
Atkinson, M. J., 1988. Are coral reefs nutrient limited? Proc. reductionist dream or expansionist vision. Proc. 6th int.
6th int. Coral Reef Symp. 1: 157-166. Coral Reef Symp. 1: 9-17.
Babcock, R. C., 1988. Fine-scale spatial and temporal pat Doherty, P. J. & D. McB. Williams, 1989. The replenishment
terns in coral settlement. Proc. 6th int. Coral Reef Symp. 2: of coral reef fish populations. Oceanogr. mar. biol. ann.
635-639. Rev. 26: 487-551.
Bazzaz, F. A., 1983. Characteristics of populations in relation Dollar, S. J. & R. W. Grigg, 1981. Impact of akaolin clay spill
to disturbance in natural and man-modified ecosystems. In on a coral reef in Hawaii. Mar. Biol. 65: 269-276.
Ft. A. Mooney & M. G odron (eds), Disturbance and Ec Done, T. J., 1985. Effects of two Acanthaster outbreaks on
osystems. Components of Response. Springer-Verlag, Ber coral community structure: the meaning of devastation.
lin: 258-275. Proc. 5th int. Coral Reef Symp. 5: 315-320.
Bell, J. & R. Galzin, 1984. Influence of live coral cover on Done, T. J., 1988a. Simulation of recovery of pre-disturbance
coral reef fish communities. Mar. Ecol. Prog. Ser. 15: 265— size structure in populations of Porites. Mar. Biol. 100:
274. 51-61.
Bell, P. R. F., P. F. Greenfield, D. Hawker & D. Connell, Done, T. J., 1988b. An assessment of coral reefs in Mauritius.
1989. The impact of waste discharge on coral reef systems. COM ARAF Ser. Doc. 2.
Wat. Sei. Tech. 21: 121-130. Done, T. J., K. Osborne & K. F. Navin, 1988. Recovery of
Black, K. P., 1988. The relationship of reef hydrodynamics to coral communities post-Acanthaster: Progress and pros
variations in numbers o f planktonic larvae on and around pects. Proc. 6th int. Coral Reef Symp. 2: 137-142.
coral reefs. Proc. 6th int. Coral Reef Symp. 2: 125-130. Done, T. J., P. K; Dayton, A. E. Dayton & R. Steger, 1991.
Bouchon, C., 1985. Quantitative study of scleractinian coral Regional and local variability in recovery of shallow coral
communities of Tiahura Reef, (Moorea, French Polynesia). communities: Moorea, French Polynesia and central Great
Proc. 5th int. Coral Reef Symp. 6: 279-290. Barrier Reef. Coral Reefs 9: 183-192.
Brown, B. E., 1987. Worldwide death of corals: natural cyclic Endean, R. & A. M. Cameron, 1985. Ecocatastrophe on the
events or man-made pollution? Mar. Pollut. Bull. 18: 9-13. G reat Barrier Reef. Proc. 5th int. Coral Reef Symp. 5: 309-
Brown, B. E., 1988. The environmental impact of coral min 314.
ing on coral reefs in the Maldives. Env. Cons. 15: 159-165. Faure, G., 1989. Degradation of coral reefs at M oorea Island
Buddemeier, R. W. & D. Hopley, 1988. Turn-on and turn (French Polynesia) by Acanthaster planci. J. Coast. Res. 5:
offs: causes and mechanisms o f the initiation and termina 295-306.
tion of coral reef growth. Proc. 6th int. Coral Reef Symp. 1: Furnas, M. J., A. W. Mitchell, M. Gilmartin & N. Relevante,
253-261. 1990. Phytoplankton biomass and primary production in
Carpenter, R. C., 1990a. Mass mortality of Diadema semi-enclosed reef lagoons of the Central G reat Barrier
antillarum. 1. Long-term effects on sea urchin population- Reef, Australia. Coral Reefs 9: 1-10.
dynamics and coral reef algal communities. Mar. Biol. 104: Gabrie, C., M. Porcher & M. M asson, 1985. Dredging in
67-77. French Polynesian reefs: towards a general policy o f re
Carpenter, R. C., 1990b. Mass mortality of Diadema source exploitation and site development. Proc. 5th int.
antillarum. 2. Effects on population densities and grazing Coral Reef Symp. 4: 271-277.
intensity of parrotfishes and surgeonfishes. Mar. Biol. 104: Glynn, P. W., 1985. Corallivore population sizes and feeding
79-86. effects following el Niño (1982-1983) associated coral mor
Colgan, M., 1987. Coral reef recovery of G uam (Micronesia) tality in Panama. Proc. 5th int. Coral Reef Symp. 1: 183—
after catastrophic predation by Acanthaster planci. Ecology 188.
68: 1592-1605. Hamner, W. M., M. S. Jones, J. H. Carleton, I. R. Hauri &
Connell, J. H., 1978. Diversity in tropical rainforests and coral D. McB. Williams, 1988. Zooplankton, planktivorous fish,
reefs. Science 199: 1302-1310. and water currents on a windward reef face: G reat Barrier
Connell, J. H. & M. J. Keogh, 1985. Disturbance and patch Reef, Australia. Bull. mar. Sei. 42: 459-479.
dynamics of subtidal marine animals on hard substrata. In Hansen, J. A., D. M. Alongi, D. J. W. Moriarty & P. C. Pol
S. T. A. Pickett & P. S. White (eds), The Ecology of N at lard, 1987. The dynamics of benthic microbial communities
ural Disturbance and Patch Dynamics. Academic Press at Davies Reef, central G reat Barrier Reef. Coral Reefs 6:
Inc., Orlando: 125-151. 63-70.
Cuet, P., O. Naim, G. Faure & J. Y. Conan, 1988. Nutrient- Hatcher, B. G., 1984. A maritime accident provides evidence
rich groundwater impact on benthic communities on la Sa for alternate stable states in benthic communities on coral
line reef (Reunion Island, Indian Ocean): Preliminary re reefs. Cora! Reefs 4: 199-204.
sults. Proc 6th int. Coral Reef Symp. 2: 207-212. H atcher, B. G., R. E. Johannes & A. I. Robertson, 1989. Re
Dahl, A. L., 1974. The structure and dynamics of algae in the view of research relevant to the conservation o f shallow
coral reef ecosystem. Proc. 2nd int. Coral Reef Symp. 1: water tropical marine ecosystems. Oceanogr. mar. biol. ann.
21-25. Rev. 27: 337-414.
Davies, P. J., 1988. Evolution of the Great Barrier Reef - Hodgson, G. & J. A. Dixon, 1988. Logging versus fisheries
131
and tourism in Palawan. Occasional Papers of the East- McClanahan, T. R. & N. A. Muthiga, 1988. Changes in Ken
West Environment and Policy Institute, 7. yan coral reef community structure and function due to
Hopley, D. & D. W. Kinsey, 1988. The effects of a rapid exploitation. Hydrobiologia 166: 269-276.
short-term sea-level rise on the G reat Barrier Reef. In G. I. McClanahan, T. R. & N. A. Muthiga, 1989. Patterns of
Pearman (ed.), Greenhouse. Planning for climatic change. predation on a sea urchin, Echinometra mathaei (de Blain
E. J. BriH, Leiden: 189-201. ville), on Kenyan coral reefs. J. exp. mar. Biol. Ecol. 126:
Hay, M., 1984. Patterns of fish and urchin grazing on Car 77-94.
ibbean coral reefs: are previous results typical? Ecology 65: Moran, P. J., R. H. Bradbury & R. E. Reichelt, 1988. Distri
446-454. bution of recent outbreaks of the crown o f thorns starfish
Highsmith, R. C., 1980. Passive colonization and asexual col (Acanthaster planci) along the G reat Barrier Reef: 1985—
ony multiplication in the massive coral Porites lutea Ed 1986. Coral Reefs 7: 125-137.
wards. J. exp. mar. Biol. Ecol. 47: 55-67. Moriarty, D. J. W , P. C. Pollard, W. G. Hunt, C. M. Mori-
Hixon, M. A., 1983. Fish grazing and community structure of arty, T. Wassenberg, 1985. Productivity of bacteria and
reef corals and algae: a synthesis of recent studies. In M. L. microalgae and the effect o f grazing by holothurians on a
Reaka (ed.), The ecology of deep and shallow coral reefs. coral reef flat. Mar. Biol. 85: 293-300.
NOAA, Rockville, Maryland. 1: 79-87. Munro, J. L. & D. McB. Williams, 1985. Assessment and
Hughes, T. P., 1989. Community structure and diversity of management of coral reef fisheries: biological, environmen
coral reefs: the role of history. Ecology 70: 275-279. tal and socioeconomic aspects. Proc. 5th int. Coral Reef
Hughes, T. P., D. C. Reed & M. J. Boyle, 1987. Herbivory on Symp. 4: 543-578.
coral reefs: community structure following mass mortalities Munro, J. L. & I. R. Smith, 1984. Management strategies in
of sea urchins. J. exp. mar. Biol. Ecol. 113: 39-59. multispecies complexes in artisanal fisheries. Proc. Gulf
Johannes, R. E., 1975. Pollution and degradation of coral reef Caribb. Fish. Inst. 36: 127-141.
communities. In E. J. Ferguson-W ood & R. E. Johannes Muthiga, N. A. & T. R. McClanahan, 1987. Population
(eds), Tropical marine pollution, Elsevier Scientific Publ., changes of a sea urchin (Echinometra mathaei) on an ex
New York. Elsevier Oceanography Series 12: 15-51. ploited fringing reef. Afr. J. Ecol. 25: 1-8.
Johannes, R. E. & S. B. Betzer, 1975. Introduction: Marine Oliver, J. K. & B. L. Willis, 1987. Coral-spawn slicks in the
communities respond differently to pollution in the tropics Great Barrier Reef: preliminary observations. Mar. Biol.
than at higher latitudes. In E. J. Ferguson-Wood & R. E. 94: 521-529.
Johannes (eds), Tropical marine pollution, Elsevier Scien Payri, P. E. & O. Naim, 1982. Variations entre 1971 et 1980
tific Publ., New York. Elsevier Oceanography Series 12: de la biomass et de la composition des populations de
1- 12 . macroalgues sur le récif corallien de Tiahura (Ile de Moorea,
Kaufman, L., 1977. The threespot damselfish: effects on Polynésie Française) Cryptogamie: Algologie 3: 229-240.
benthic biota of Caribbean coral reefs. Proc. 3rd int. Coral Pearson, R. G., 1981. Recovery and recolonisation of coral
Reef Symp. 1: 559-564. reefs. Mar. Ecol. Prog. Ser. 4: 105-122.
Kinsey, D. W., 1985. Metabolism, calcification and carbon Pichón, M. & J. Morrissey, 1985. Premieres mesures de bilan
production. I System level studies. Proc. 2nd int. Coral métabolique dans l’ecosysteme recifal de Tulear (M adagas
Reef Symp. 4: 505-626. car). C.r. Acad. Sei. Paris 300: 99-101.
Lapointe, B. E., 1989. Caribbean coral reefs: Are they becom Potts, D. C., 1977. Suppression of coral populations by fila
ing algal reefs? Sea Frontiers 35: 82-91. mentous algae within damselfish territories. J. exp. mar.
Larkum, A. W. D., 1983. The primary productivity o f plant Biol. Ecol. 28: 207-216.
communities on coral reefs. In D. J. Barnes (ed.), Perspec Ricard, M., 1982. Mean features of phytoplankton and pri
tives on Coral Reefs. The Australian Institute of Marine mary productivity of M oorea Island (French Polynesia).
Science, Brian Clouston Publisher, Manuka: 221-239. Abstract. Proc. 4th int. Coral Reef Symp. 1: 453.
Lighty, R. G., 1982. Fleshy-algal domination of a modern Russ, G., 1984. The distribution and abundance of herbivo
Bahamian barrier reef: example of an alternate climax reef rous grazing fishes in the central G reat Barrier Reef. I.
community. A bstract - Proc. 4th int. Coral Reef Symp. 1: Levels of variability across the entire continental shelf. Mar.
722. Ecol. Prog. Ser. 20: 23-34.
Littler, M. M. & D. S. Littler, 1985. Factors controlling rela Russ, R. G. & A. C. Alcala, 1989. Effects of intense fishing
tive dominance of primary producers on biotic reefs. Proc. pressure on an assemblage of coral reef fishes. Mar. Ecol.
5th int. Coral Reef Symp. 4: 35-39. Prog. Ser. 65: 13-27.
Lobel, P. S., 1980. Herbivory by damselfish and their role in Sale, P. F., 1988. W hat coral reefs can tell us about ecology.
coral reef community ecology. Bull. mar. Sei. 30: 273-289. Proc. 6th int. Coral Reef Symp. 1: 19-27.
Maragos, J. E., C. Evans & P. Holthus, 1985. Reef corals in Salvat, B. (ed.), 1987. Human Im pacts on Coral reefs: Facts
Kaneohe Bay six years before and after termination of sew and Recommendations. Antenne Museum Ecole Pratique
erage discharges. Proc. 5th int. Coral Reef Symp. 4: 189— des Hautes Etudes, French Polynesia, 253 pp.
194. Sammarco, P. W. & J. C. Andrews, 1988. Localised dispersal
132
and recruitment in G reat Barrier Reef corals: the Helix phosphate and silicate in Cleveland Bay, northern Queen
Experiment. Science 239: 1422-1424. sland. Aust. J. mar. Freshwat. Res. 32: 877-887.
Samoilys, M. A., 1988. Abundance and species richness of Williams, D. McB. & A. I. Hatcher, 1983. Structure of fish
coral reef fish on the Kenyan coast: the effects of protec communities on outer slopes of inshore, mid-shelf and outer
tive management and fishing. Proc. 2nd int. Coral Reef shelf reefs of the G reat Barrier Reef. Mar. Ecol. Prog. Ser.
Symp. 2: 261-266. 10: 239-250.
Smith, S. V., W. J. Kimmener, E. A. Laws, R. E. Brock & Williams, D. McB., E. Wolanski & J. C. Andrews, 1984.
T. W. Walsh, 1981. Kaneohe Bay sewage diversion exper Transport mechanisms and the potential movement of
iment: perspectives on ecosystem responses to nutritional planktonic larvae in the central region of the G reat Barrier
perturbation. Pac. Sei. 35: 279-395. Reef. Coral Reefs 3: 229-236.
Sournia, A., 1976. Oxygen metabolism of a fringing reef in Woodley, J. D., 1979. The effects of trap-fishing on reef com
French Polynesia. Helgol. wiss. Meeres. 28: 401-410. munities in Jamaica. Proc. Is. Mar. Labs. Carib. 13: 27.
Steneck, R. S., 1988. Herbivory on coral reefs: a synthesis. Woodley, J. D. & 19 others, 1981. Hurricane Allen’s impact
Proc. 6th int. Coral Reef Symp. 1: 37-49. on Jam aican coral reefs. Science 214: 749-755.
Walker, T. & G. O’Donnell, 1981. Observations on nitrate,
Hydrobiologia 247: 133- 138, 1992 .
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 133
© 1992 Kluwer Academic Publishers. Printed in Belgium.
J. D. W oodley
Discovery B ay Marine Laboratory, P.O. B ox 35, Discovery Bay, Jamaica, West Indies
Abstract
W hen the coral reefs o f north Jam aica were described and studied in the 1950’s, 60’s and 70’s, their
shallow zones were dom inated by dense thickets o f Acropora palmata and A . cervicornis. T hose stands
were dem olished by hurricanes in 1980 and 1988, but these seemed to be unusual events.
Published records o f storm tracks indicate that, in the last 120 years, about 35 hurricanes have passed
within 364 km (200 nautical miles) o f Discovery Bay, Jam aica. In the absence o f inform ation on their
intensity, an approxim ate m easure o f their im pact is the reciprocal o f their closest distance. R ecent
experience suggests th at passage within about 65 km, w hether to the north or the south, generates waves
capable o f destroying Acropora stands. There have been 13 years, out o f the last 120, in which hurri
canes have passed th at close to J am aica, with a m edian interval o f 6.5 years. Large colonies o f A . palmata
(over 1 m high) might take about 12 years to develop. The total num ber o f years in which the reefs have
been free o f m ajor disturbance for longer than 12 years is only 33, and 24 o f them (1956-80) were in
the unusually long interval o f 36 years, from 1944 to 1980. The luxuriant Acropora stands o f the clas
sic descriptions m ay therefore be atypical; one extrem e o f a variable condition.
storm s. Thus, in relation to geological processes around - 5 to - 1 0 m, m assive corals were m uch
o f reef grow th or sedim entation, on a tim e-scale com m oner. In 1980, the huge w aves generated by
of hundreds or th o usands o f years, hurricanes H urricane Allen sm ashed the Acropora thickets
can be regarded as a continuous force. O ne could and m ost o f their fragm ents died (K now lton et al.,
recognize different intensities o f th at force due to 1981). The Acropora zones were transform ed into
differences in hurricane frequency in space or over rubble like th at which G oreau h ad seen on the
tim e (e.g., H u b b ard, 1988). On shorter time south coast. Allen w as the strongest C aribbean
scales, the occurrence o f hurricanes is irregular. storm of the century (at the time), and the first to
Their influence on processes m easured on a scale affect the n o rth coast for m any years (W oodley
o f the sam e order as the interval betw een hurri et al., 1981). It was generally believed th at it was
canes, such as the grow th o f corals, m ay be bet only a m atter o f tim e before the n o rth coast reefs
ter understood in term s o f the time elapsed since returned to their form er luxuriance. H ow ever, not
the previous storm , rather than their average fre only w as the recovery o f Acropora populations
quency (Kjerfve eta l., 1986). unexpectedly slow (K now lton eta l., 1990), but
C orals differ in their growth strategies; th at is only eight years later they suffered another violent
to say, in their growth rates, recruitm ent rates, disturbance, from H urricane G ilbert. The physi
com petitive ability, fragility and regenerative cal im pact o f this storm w as at least as great as
powers. As a result, they differ in their population th at o f Allen, but the dam age to branching corals
responses to storm im pacts. Montastraea w as m uch less spectacular because they h ad not
annularis (Ellis & Solander) is a m assive, slow- h ad tim e to regain their previous abundance
growing fram e-builder on C aribbean reefs, resis (W oodley, 1989).
tan t to storm s, but susceptible to overgrowth by T here are good records o f hurricanes in the
faster-grow ing branching corals such as Acropora C aribbean, and it is possible to determ ine their
palmata (L am arck) and A . cervicornis (Lam arck). p attern o f occurrence with reasonable accuracy
These are strong enough to resist m oderate wave for the last century or so. Im p o rtan t sources are
energies and often dom inate shallow reef zones a set o f m aps showing tropical cyclones o f the
th at are nam ed after them (G oreau, 1959). O n the N o rth A tlantic, 1871-1980 (N eum ann e ta l.,
n o rth coast o f Jam aica, G oreau described the 1981), and the very detailed Jam aica W eather
breaker zone as ‘populated alm ost exclusively by R eports w hich began in 1881. In this paper, I use
huge tree-like colonies o f Acropora palm ata', and those sources to analyse the tracks o f hurricanes
the upper p art o f the seaw ard slope as dom inated th at m ight have dam aged reefs at Discovery Bay,
by ‘im m ense beds o f staghorn coral., Acropora in the centre o f the north coast o f Jam aica, where
cervicornis'. South coast reefs had been hit by m any reef studies have been m ade.
H urricane C harlie in 1951; the Acropora spp. were
less dom inant and there w as m ore rubble. G oreau
suggested, correctly but cautiously, th at ‘storm s Methods
can play an im portant part in altering the basic
p attern o f species dom inance and zonal succes Table 1 presents d ata on 39 storm s th at passed
sion in the upper parts o f a climax re ef. closer than 364 km (200 nautical miles) to D is
Lor another 20 years, the two Acropora species covery Bay since 1870. Their initial selection was
rem ained conspicuous dom inants on the north m ade from the m aps o f N eum ann et al. (1981).
coast o f Jam aica. A . palmata form ed a continu O n the earliest m aps, tropical storm s (central
ous thicket, up to 2 m high, from the surface to winds less than 3 3 m s _1 [7 4 m p h ]) and hurri
- 6 m depth, while A cervicornis covered the shal canes are n o t distinguished. L rom 1886 to 1898,
low terrace to a depth of about - 2 2 m, punctu storm s know n to have reached hurricane inten
ated by M . annularis and other m assive corals. At sity at some tim e are distinguished from the oth
norm al wave base, in the M ixed or B uttress Z one ers, while from 1899 onw ards the various stages
135
o f storm developm ent are indicated along each four storm s might no t have reached hurricane
track. F o r the present w ork, the intensity cate intensity in this area. The sam e can be said for the
gory o f storm s passing close to Jam aica from storm s o f 1889, 1894, 1895, 1909 and 1910 (T a
1880 onw ards w as determ ined from the Jam aica ble 1). F urther evidence on the category o f these
W eather R eports. These R eports were also used storm s might be obtained from Jam aican (or
for m ore accurate definition o f hurricane tracks C uban) new spapers, but none is m entioned in
where they differed from the sm all-scale m aps of H all’s historical review (1916) o f hurricanes and
N eum ann et al. The R eports on hurricanes are other phenom ena observed in Jam aica.
often rem arkably detailed; especially those w rit The height of the waves radiating from the
ten by M axwell H aii, am ateur founder o f the eye o f a hurricane decreases with distance
w eather service in 1881 and G overnm ent M ete (B retschneider & Tam aye, 1976). A crude com
orologist until at least 1919. H e collated the re parative indicator o f their destructive pow er is the
ports o f other am ateur observers across the island reciprocal o f th at distance. If no other inform a
and interpreted them with great skill and enthu tion is available on the intensity o f different hur
siasm. ricanes, their im pacts can be com pared in this
The adjusted tracks o f the 35 rem aining storm s way, as a first approxim ation. Table 1 lists the
th at were, or might have been, hurricanes when track o f each storm , its direction (aspect) from
they passed Jam aica, are illustrated for 12 suc D iscovery Bay w hen closest, its distance at that
cessive decades in Fig. 1. In the first decade, no point (r), and the reciprocal o f th at distance (mul
records were available for 1870, and three o f the tiplied by 1000). T hose reciprocals are plotted in
Fig. 2, to an arbitrary m axim um of 33 (that co r
responds to a hurricane centre passing 30 km
-20!
away, w hich is close enough to cause catastrophic
lo a k m dam age, so there is little point in discrim inating
it from even closer encounters).
1 8 7 0 -7 9 75 W 1 9 3 0 -3 9 '
Table 1. D ata on storms that might have reached hurricane intensity within 364 km (200 nautical miles) of Discovery Bay, Ja
maica since 1870. Direction (aspect) and distance (r) from Discovery Bay, and the reciprocal o f that distance (multiplied by 1,000)
are shown. Storm data were taken from Neumann etal. (1981) and modified from other sources as indicated. The Jam aica Weather
Reports (JWR) were published twice; first as monthly and occasional bulletins, then in collected volumes entitled Jam aica Me
teorological Observations every 6 -8 years. Tropical storm (TS).
N o r did th at happen during any o f the distant P o rt Royal Cays, but little on the north coast (as
hurricanes from 1954-1966, w hen T. F. G oreau indicated by G oreau, 1959), w here m axim um
w as working in J am aica, or he would surely have w inds were only 22 ms “ 1 (50 mph).
docum ented it. H urricane Charlie (1951) passed Assum ing th at all these hurricanes were o f
only 90 km south o f Discovery Bay. It caused equal size and strength (which they were not), the
m uch dam age in K ingston and to the reefs o f the threshold distance for the destruction o f Acropora
137
30
25
*. 20
i-
o
o
O 15
10
ical order these are 1655-1689 (34 years), 1786— No. 465. In Jam aica Meteorological Observations 5. Ja
maica Government Printing Office, Kingston, 1918.
1812 (26 years), 1844-1874 (30 years) and 1917—
Hubbard, D. K., 1988. Controls of modern and fossil reef
1933 (16 years). The next period w ould be 1951- development: common ground for biological and geological
1980 (29 years) or, using the criterion of research. Proc. Sixth Int. Coral Reef Symp. Townsville 1:
‘traversing Jam aica’, 1944-1988 (44 years). 243-252.
This analysis has left out any consideration of Kinzie, R. A., 1973. The zonation of W est Indian gorgonians.
Bull mar. Sei. 23: 93-155.
differences in m agnitude between the hurricanes,
Kjerfve, B., K. E. Magill, J. W. Porter & J. D. Woodley, 1986.
largely because such inform ation is not readily Hindcasting of hurricane characteristics and observed
available. It w ould be possible to gather d ata storm damage on a fringing reef, Jamaica, West Indies. J.
about the im portant storm s, from the Jam aica mar. Res. 44: 119-148.
W eather R eports and from contem porary new s Knowlton, N., J. C. Lang & B. D. Keller, 1990. Case-study
of natural population collapse: post-hurricane predation on
papers, to h indcast their im pacts on corals. Kjer-
Jam aican staghorn corals. Smithsonian Contributions to
fve et al. (1986) calculated th at turbulent flows of the Marine Sciences 31, 25 pp. Smithsonian Institution
2 -4 ms " 1 would bring dow n A . palmata colonies, Press, Washington D.C.
depending w hether or n o t they were bio-eroded. Knowlton, N., J. C. Lang, M. C. Rooney & P. Clifford, 1981.
Equivalent values for A . cervicornis were 0 .3 5 - Evidence for delayed mortality in hurricane-damaged Ja
maican staghorn corals. Nature 294: 251-252.
1.6 ms “ \ Only when th at analysis has been m ade
Neumann, C. J., G. W. Cry, E. L. Caso & B. R. Jarvinen,
shall we know if the classic descriptions o f J a 1981. Tropical Cyclones of the N orth Atlantic Ocean,
m aican reefs are indeed atypical. 1871-1980. U.S. Departm ent of Commerce, NOAA.
Taylor, G. & Staff, 1981. Annual d ata and verification tabu
lation Atlantic tropical cyclones 1980. National Hurricane
Center, NOAA tech. Mem. NW S N H C 15.
R e fere n ce s Woodley, J. D., 1989. The effects o f Hurricane Gilbert on
coral reefs at Discovery Bay. Appendix 9 in P. R. Bacon
Bretschneider, C. L. & E. Tamaye, 1976. Hurricane wind and (ed.), Assessment of the economic impacts of Hurricane
wave forecasting techniques. Look Lab/Hawaii 6: 1-29. Gilbert on coastal and marine resources in Jamaica. U N E P
Goreau, T. F., 1959. The ecology of Jam aican coral reefs I. Regional Seas Reports and Studies 110.
Species composition and zonation. Ecology 40: 67-90. Woodley, J. D., E. A. Chornesky, P. A. Clifford, J. B. C.
Haii, M., 1889. Introduction. Jam aica Meteorological obser Jackson, L. S. Kaufman, N. Knowlton, J. C. Lang, M. P.
vations 1. Jam aica Government Printing Establishment, Pearson, J. W. Porter, M. C. Rooney, K. W. Rylaarsdam,
Kingston. V. J. Tunnicliffe, C. M. Wahle, J. L. Wulff, A. S. G. Curtis,
Haii, M., 1911. Notes on the storms which passed over or M. D. Dallmeyer, B. P. Jupp, M. A. R. Koehl, J. E. Neigel
near Jamaica, between 1902 and 1910 inclusive. Jam aica & E. M. Sides, 1981. Hurricane Allen’s impact on Jam ai
Meteorological Observations 4. Jam aica Government can coral reefs. Science 214: 749-755.
Printing Office, Kingston. Woodley, J. D. & E. Robinson, 1977. Field Guide to the
Haii, M., 1916. Notes of hurricanes, earthquakes, and other M odern and Ancient Reefs of J amaica. Atlantic Reef Com
physical occurrences in Jamaica. Jam aica Weather Report mittee, Miami.
■ÎSSÎ2
Hydrobiologia 247: 1 3 9 -1 4 0 , 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 139
© 1992 Kluwer Academic Publishers. Printed in Belgium.
Short note
M angroves and seagrass beds are know n to function as nursery for juveniles o f reef fishes. How ever,
the fish fauna o f these often linked ecosystem s is m ostly studied separately due to m ethodological
problem s. Therefore a selected num ber o f fish species w as studied both in Lac-lagoon (Bonaire, N eth
erlands Antilles) and the coral reef around Bonaire.
Lac-lagoon is the largest lagoon o f B onaire w ith an area o f about 8 km 2. It is a protected w etland
according to the R am sar convention. It consists o f a shallow basin with clear nutrient-poor w ater
bordered by m angroves and m udflats. T he bottom o f the lagoon is m ainly covered by seagrasses, es
pecially Thalassia testudinum B anks ex König. T he investigations were focussed on the im portance of
the Lac-lagoon for the fishes on the reefs. T he reefs form the m ain attraction for the tourists visiting
Bonaire.
Table 1. Importance of various biotopes for juveniles of selected fish species and the importance of Lac-lagoon nursery for the
fish fauna of Lac-lagoon itself and the reef.
3 Caesar grunt - - - - - -
4 Smallmouth grunt - - XX - - -
5 Black margate - - - - - -
6 Gray snapper XXX XX - - XXX (x )
7 Schoolmaster XXX XX - - XX X
8 Mahogany snapper - X XX - - X
11 Sergeant major (x ) - XX - - X
14 Doctorfish (X ) XXX X - - X
15 Blue tang (x ) X X X - X
- not important
() doubtful
X minor importance
XX medium importance
XXX major importance
140
Sixteen species of reef fishes were selected based on easy identification and econom ic value (attrac
tive for diving tourists, fishery) for a study o f abundance, distribution and size classes. D uring M a y -
N ovem ber 1981 abundance and size classes were estim ated by m eans o f a visual census technique;
estim ations in the seagrass beds were checked by catches w ith drop nets.
Six ‘biotopes’ have been studied, viz. seagrass beds (T . testudinum) and m angroves in the lagoon and
four depth zones on the reef. The m ajor results are sum m arized in Table 1.
M utual com parison indicates th at the seagrass beds form the m ost im portant biotope for the juve
nile stages o f five fish species, viz. F rench grunt (Haemulon flavolineatum (D esm arest)), B luestriped grunt
(H. sciurus (Shaw )), Yellowtail snapper (Ocyurus chrysurus (Bloch)), Stoplight p arro t fish (Sparisoma
viride (B onnaterre)) and D o cto r fish (Acanthurus chirurgus (Bloch)), while m angroves are im portant for
four other species, viz. Schoolm aster (Lutjanus apodus (W albaum )), G ray snapper (L. griseus (L.)), G reat
b arracu d a (Sphyraena barracuda (B onnaterre)) and Foureye butterfly (Chaetodon capistratus (L.)).
F o r five species the lagoon seemed to be less im portant, viz. Sm allm outh grunt (Haemulon
chrysargyreum G ünther), M ahogany snapper (Lutjanus mahogoni (Cuvier)), Sergeant m ajor (.Abudefduf
saxatilis (L.)), O cean surgeon (Acanthurus bahianus C astelnau) and Blue tang (A. coeruleus Bloch &
Schneider): their juveniles mainly use the shallow reef. O f tw o species, C aesar grunt (Haemulon
carbonarium Poey) and Black m argrate (Anisotremus surinamensis (Bloch)), o f w hich adults were only
found on the reef, too few juveniles were found to draw conclusions.
'IZShi.L
Hydrobiologia 247: 1 4 1 -1 6 1 , 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 141
© 1992 Kluwer Academic Publishers. Printed in Belgium.
Eric W olanski
Australian Institute o f Marine Science, P .M .B . No. 3, Townsville M .C ., Qld. 4810, Australia
Abstract
M angrove sw am ps help control the tidal hydrodynam ics o f m any tropical estuaries. They generate an
asym m etry o f the tidal currents in both the tidal creeks and the m angrove swam ps. This results in
self-scouring o f the tidal channels. M angrove land reclam ation results in siltation o f the channel. M an
grove sw am ps control the flushing rates o f the estuaries through the lateral trapping effect. Lateral
trapping leads to the aggregation o f m angrove litter along slick lines. E vapotranspiration plays a role in
the hot dry season by forming a salinity m axim um zone w hich isolates the estuary from the coastal w aters
for several m onths o f the year. In the absence o f runoff, évapotranspiration in the hot dry season gen
erates an inverse estuarine circulation which can trap high salinity m angrove w ater, and m angrove de
tritus, along the bottom o f a m angrove creek. This bottom layer can becom e anaerobic. G roundw ater
flow appears to play a key role in the nutrient budget o f m angrove creeks, exporting salt left behind by
évapotranspiration, and inhibiting runoff after rainfall. P articulates and dissolved nutrients outwelled
from m angrove sw am ps to coastal w aters are retained in a coastal boundary layer. This coastal boundary
layer w ater can be trap p ed along the shore for long periods if the coast is straight and mangrove-fringed
and the coastal w aters are shallow. H eadlands inhibit coastal trapping because they enhance mixing.
N utrient-rich coastal boundary layer w aters m ay be ejected offshore as tidal jets peeling off headlands
and locally enriching offshore waters.
(b )
t' 1 9 m };t'
100 m
U p str e a m H a lf w a y
Fig. 1. (a) M ap of Coral Creek, a mangrove creek in tropical Australia, with elevation in m. (b) shows two typical cross-sections
in the tidal creek.
the creek. They thus divided the system in its two where V¡ is the volum e o f w ater in cell i, QtJ is the
basic com ponent, firstly the tidal creek where ac discharge (in m 3 s _1) betw een cells i and j , and
celeration and inertia effects are im portant, and q¡ is the discharge from cell i direct to the tidal
secondly the m angrove swam p where they are creek. q¡ = 0 if cell i does no t touch the tidal creek.
not. F o r m odeling purposes, the swam p w as di Inertia and acceleration effects are assum ed to be
vided in a num ber of cells o f irregular shape and negligible in the m angrove swam p where the m o
size fitted to the topography o f the substrate m entum equation reduces to a simple balance be
(Fig. 3a). The w ater circulation in the swam p was tw een the frictional slope and the free surface
calculated from the continuity equation, one such slope,
equation for each cell i,
dq
143
m om entum :
ICARDW ELL
ÔQ 8 dri
~ T + T (2 2M ) + gA - Í + gA Sf + A = 0 (4)
Point at dx dx
Fig. 3. (a) shows the distribution of cells in the mangrove swamp. (b) shows the location of the cross-sections.
o f the depth-averaged currents, including the tidal over a tidal cycle there exists in the fringing m an
asym m etry, the peak ebb tidal currents being groves a net circulation tow ards the m outh. The
m uch stronger th an peak flood currents. T he lat im plications o f this circulation in the aggregation
ter peak at 1 . 2 m s ‘ \ the form er at 1.6 m s “ 1. o f m angrove litter are discussed later.
This tidal asym m etry results in self-scouring o f
the tidal creek (W olanski et al., 1990). N um erical
experim ents suggest th at destruction o f the veg Groundwater flow
etation or reclam ation o f m angrove land, reduces
both the peak tidal currents and the asym m etry o f In the presence o f strong tidal currents such as at
the tidal currents, thereby resulting in siltation of C oral Creek, groundw ater flow probably does not
the channel. contribute m uch to the hydrodynam ics o f the sys
Figure 5 show s the distribution o f the predicted tem. In other cases, groundw ater flow m ay be
peak ebb and flood tidal currents in the m angrove im portant. Ovalle et al. (1990) have suggested that
swam p. This figures illustrates another im portant the net outflow from the small Itacuruca m an
asym m etry of the tidal currents. The flood tidal grove sw am p in Brazil is due to groundw ater in
currents are small, typically o f order 0.03 m s - 1 , flow in the swam p.
and are oriented at about 90 ° to the banks o f the W olanski (unpublished data) and M az d a et al.
creek. On the other hand, the ebb tidal currents (1990a) have observed the initial stages o f the
are up to 3 times faster and are oriented at typ flooding o f the sw am p at rising tide w hen a thin
ically 20 to 30° from the creek. This asym m etry sheet o f w ater advances in the sw am p from the
in the tidal currents w as also confirmed in the creek. This sheet is blocked by the high vegetation
field through m easurem ents o f currents and cur density. The first few mm o f flooding w ater at a
rent directions at one point in the swam p. H ence, given point are no t due to w ater from th at sheet
145
200
O 150
100
1-
50 o.
•i
CD
ä O 0
-50
-100
-150
0 6 12 18 24
Tim e (h)
Fig. 4. Time series plot over one tidal cycle of the observed currents at different depths below the surface (from near the surface
to near the bottom) at the mouth of Coral Creek; the thick line shows the depth-averaged predicted velocity time series. The thin
line is the sea level time series.
spilling over obstructions on the ground, but are crab burrow s, since below th at point the ground
observed to com e from the ground through the w ater was anaerobic.
num erous holes due to crabs and rotting vegeta The im portance o f crab burrow s in enhancing
tion. groundw ater flow is apparent in C oral Creek and
W olanski & G ardiner (1981) have proposed in the Itacuruca swam p, where rainfall infiltrates
th at this groundw ater flow is an im portant flush readily in the soil and produces no local runoff,
ing m echanism preventing the accum ulation of and in the K long N gao m angrove swam p in T hai
salt left behind after évapotranspiration. land where crab burrow s are few and direct run
M azd a et al. (1990) have carried out a study of off is observed (W attayakorn eta l., 1990).
groundw ater flow in the 200 m long Bashita-
M inato m angrove sw am p in Japan. After storm s,
the tidal creek is occasionally ponded by a sill at Longitudinal diffusion in mangrove creeks
the m outh. They found th at when the creek is
ponded the w ater level in the creek decreases by Salinity in m angrove creeks increases with in
up to 10 cm d a y - \ about 15 times larger than the creasing distance from the m outh in the hot dry
evaporation rate. This observation suggests that season when runoff is negligible. The salinity in
the creek w as draining out through groundw ater crease is due to évapotranspiration and is typi
flow. The flow appeared to be limited to the upper cally o f order 0.5 to 1 p.p.t. for 5 km long m an
90 cm o f the substrate, i.e. the m axim um depth of grove creeks such as C oral Creek and D ickson
146
20 cm s'
Velocity
03
CD
03
(a ) F lo o d ti d e
C
o
'</)
<o
0L i j2 km
(b) E b b ti d e
Fig. 5. Predicted peak tidal currents in the mangrove swamp of Coral Creek at flood and ebb tides.
Inlet in A ustralia (W olanski et al., 1980; R idd estuarine circulation exists on either side o f the
et al., 1990) and the Klong N gao m angrove creek salinity m axim um zone. In tropical A ustralia,
in T hailand (W attayakom et al., 1990). This sa
linity gradient can introduce an inverse estuarine
circulation as is sketched in Fig. 6a. This circu
lation becom es hydrodynam ic ally im portant and E s tu a ry
m easurable w hen both vertical and horizontal sa
linity gradients are large, the existence o f a verti S alin ity m a x im u m z o n e
(c ) S ite A D is ta n c e S ite B
C oral C reek
S ite B
m angrove
S ite A sw a m p
o 7 2
2 km
(b ) 1
E
Q_
CD
a
0
7
o
5
T im e (h o u r s )
Fig. 7. Longitudinal distribution of salinity in Coral Creek in the hot dry season.
S a li n i t y ( % » )
S ea Head
D is o lv e d O x y g e n (m g M )
a> u
a
1000 m
Fig. 8. (a) shows the salinity distribution in the Dickson Inlet mangrove creek in the hot dry season, at low tide, (b) shows the
corresponding distribution of dissolved oxygen, also at low tide.
such an unusual estuarine circulation can last In the presence o f an inverse estuarine circu-
several m onths (W olanski, 1986; R idd eta!., lation, high salinity m angrove w ater is then
1988). trapped in the creek at the bottom and is not
148
gleet the presence o f the fringing m angrove no surface w ater for a fraction (1-a) of the tidal
swam p. In fact this m angrove system is the rea cycle o f period 2T. They predicted
son for the existence o f a large longitudinal eddy
diffusion coefficient in m angrove creeks. The pre A su 2a 2 T
B = ---------+ --------------- (13)
vailing m echanism is lateral trapping, sketched in (1 + £) 48 (1 + £)
Fig. 9. This m echanism w as first studied by
O kubo (1973) who m odeled dispersion in a tidal
estuary with a lateral em baym ent. A tagged w ater E quation (13) yield B = 10-40 m 2 s - 1 for C oral
m ass m oves upstream at flood tide, bu t a fraction Creek, D ickson Inlet and the K long N gao estu
o f th at m ass ends trapped in the em baym ent. At ary, values th at agree qualitatively well with o b
ebb tide, this m ass returns to the tidal creek and servations. Equation (13) w as recently modified
mixes with untagged w ater. Mixing is thus greatly by R idd eta l. (1990) to take into account the
enhanced and, following O kubo (1973), decrease in peak tidal velocity from the m outh to
the upper reaches o f the m angrove creek. Smaller
A values o f the eddy diffusion coefficient are pre
B = -------+ ------------------- ---------------- (12)
1 + s 2 k (1 + e)2 (1 + e + o/k) dicted in the upper reaches o f the creek particu
larly in the upperm ost 200 m or so.
Lateral trapping in m angrove sw am ps thus ap
where u0 is the peak tidal current, k~ 1 is the pear to be the dom inant effect determ ining their
characteristic exchange time betw een the em bay flushing rate. T he residence time in mangrove
m ent and the estuary, eris the tidal frequency and creeks is o f order
£ is the ratio o f volum e in the em baym ent to that
in the estuary. T0 = L 2/B (14)
W olanski & R idd (1986) m odified the O kubo
(1973) m odel to rem ove tw o assum ptions th at are where L is the length o f the mangrove-fringed
invalid in a m angrove creek, nam ely th at the creek. F or C oral Creek and the K long N gao m an
depths o f w ater in the estuary and in the swam p grove creek, it results T0 = 6 days, suggesting that
are the sam e and are m uch larger than the tidal the upper reaches o f m angrove creeks are flushed
range, and th at the exchange rate between w ater slowly.
in the swam p and in the estuary are constant in W olanski eta l. (1990) also applied equations
time. They divided the m angrove creek in a rect (13) and (14) to estim ate the values o f B and T0
angular m ain channel where tidal currents occur, in H inchinbrook C hannel. They found T0 = 50
and a sw am p seen as the lateral em baym ent. The days. This implies long-term trapping in this giant
elevation o f the swam p substrate w as taken to be m angrove swam p. In fact they found experim en
higher than m ean sea level so th at the swam p has tal evidence for this long-term trapping in the
flushing o f brackish w ater from the channel fol
lowing cessation o f runoff. The salinity deficit fol
L a te ra l e m b a y m e n t
lowed an exponential decay in time, with an e-fold
time scale o f 54 days, in surprisingly good agree
m ent with the predictions.
E s tu a ry
- ----------------------- —--------- .--------- U ncles eta l. (1990) studied the dynam ics o f a
Fig. 9. Sketch of the lateral trapping phenomenon in a tidal lärge, m angrove-fringed, estuary in M alaysia,
estuary with an embayment, adapted from Okubo (1973). This estuary receives a large fairly steady fresh-
150
I«-tidal river -*(«.---------- mangrove swamp W attayakorn etal. (1990) show ed th at the
2S small m angrove creek o f K long N gao in T hailand
low salinity w ater stratifies and destratifies at a few hours period
isohalines
(a ) H ig h tid e closely following fluctuations in b oth the tidal cur
rents and in the freshw ater inflow which varied
rapidly w ith local short-lived storm s in the small,
.:M steep w atershed draining into th at swam p. This
'N>m ix in g ^ /S low salinity w ater
unsteadiness precluded a reliable estim ate o f the
(b ) L o w tid e
outwelling, if any, in the wet season.
In tropical A ustralia, river runoff can be very
Fig. 10. Cross-channel distribution of salinity in a mangrove-
intense bu t short-lived. Salt w ater can be totally
fringed estuary at high and low tide following recovery of the
system from a major river flood. flushed out o f the estuary during floods. As soon
as the flood is over, salt w ater creeps back in the
estuary. If the estuary is m angrove-fringed, such
w ater inflow. They found th at the salinity strati as the W enlock River in A ustralia (W olanski &
fication is strong during neap tides but th at the R idd, 1986), the intruding salt w ater intrudes
system destratifies during spring tides. W ater under the fresh w ater rem aining near the surface.
flooding the m angroves is thus fresher at neap As the tide rises and salt w ater fills the estuary,
tides and saltier at spring tides. This m ay intro the freshw ater is pushed back laterally into the
duce a cycle, with a period o f 2 weeks, in the fringing m angrove sw am ps (Fig. 10). This fresh
productivity o f the fringing m angrove forest. w ater is then trapped there, with m inim al mixing
Fig. 11. Photograph showing an oceanographic front separating swamp water from creek waters, and made visible by a foam line.
151
with salt w ater in the estuary. As the tide falls, this floating m angrove leaves further inland in the
freshw ater returns to the estuary but stays trapped swam p. A t ebb tide, the currents are stronger,
as a narrow river plum e in shallow w aters along break the surface film and export floating leaves
the length o f the estuary banks (Fig. 10). In the to the creek. The leaves are then aggregated in
W enlock River estuary, this front m aintained its long straight lines (foam lines) m arking the small-
integrity throughout the ebb tide. H ence there was scale oceanographic fronts separating m angrove
minim al mixing betw een fresh w ater in the m an swam p w ater from creek w ater (Fig. 11). These
grove swam p and sea w ater in the estuary. leaves can be outwelled from the m angrove creek
The lateral trapping m echanism prevails also in to coastal w aters where they can som etimes be
m angrove creeks and is occasionally m ade readily found, still aligned in a straight line, several km
visible by the presence o f an oceanographic front from the swam p in calm weather.
forming at the start o f the ebb tide. This front, H ence, the fate o f m angrove detritus cannot be
shown in Fig. 11, separates swam p w ater from estim ated using classical advection-diffusion
creek w ater. In small m angrove creeks, this front m odels since these m odels predict th at the detri
disappears in less than an hour but it indicates tus would be dispersed in a cloud o f dim ension
th at the creek is n o t laterally well-mixed. increasing in time. The form ation o f small scale
The front forming between creek and swam p barotropic and baroclinie fronts in m angrove
w ater is im portant as it aggregates floating detri sw am ps has not been studied. These front are
tus, such as m angrove leaves. Visual observations im portant since they result for a depth-averaged
reveal th at at C oral Creek the flood tidal currents m odel in an apparent negative eddy diffusion as
are too w eak to break the organic film at the the m angrove seeds and leaves are n o t dispersed
surface o f the w ater in the m angrove swam p. but are instead aggregated. This m ay have im por
Floating m angrove leaves are often trap p ed in tan t biological consequences. F or instance one
this film and the flood tide does not carry the often sees small fish aggregating under the leaves
H ead D is ta n c e (km ) S ea
32,
33.5
Q_
CD
Q
/'
High tid e
^ __ ^ 33
y? .?
2
C'fiK. 1
floating in a sm all-scale front in a m angrove creek. The presence o f a sill or narrow opening in
W olanski & H am ner (1988) docum ented several creases trapping. F o r instance, the m angrove
sm all-scale m echanism s aggregating floating coral creek o f C hw aka Bay in Z an zib ar is nearly p o n
eggs. T hese m echanism s are due to the three- ded for a m easurable portion o f the tidal cycle by
dim ensional circulation in topographically con a small coral lim estone sill at the m outh. Strong
trolled fronts. M ost o f the examples illustrated for tidal mixing over shallow w ater o f the sill results
reefs can be applied to m angrove creeks. F o r in in the form ation o f a front separating creek w a
stance, leaves aggregate dow nstream o f a sharp ters from oceanic w aters (Fig. 13). T he sill thus
bend in the m angrove creek even in barotropic inhibits mixing betw een creek w aters and oceanic
conditions as a result o f the secondary circulation w aters, and ensures long-term trapping in the
in a zone o f strong flow curvature, in the same m angrove-fringed tidal creek (W olanski, 1989).
m anner as coral eggs are aggregated in zones of
strong flow curvatures near coral reefs. Axial con
vergence (Sim pson & Turrell, 1986), due to the Trapping in the coastal zone
transverse tran sp o rt in estuaries, also appears to
be a process m aintaining the floating leaves and Straight coastline
m angrove seeds in a foamline, especially n ear the
m outh o f m angrove creeks. M aterial outwelled from m angrove sw am ps m ay
A com bination o f lateral trapping and buoy support offshore productivity but, if the coastal
ancy effects can result in very long residence time w aters are shallow, this enrichm ent is restricted
o f brackish w ater in m angrove creeks. Figure 12 to either a narrow coastal zone or specific points
shows the longitudinal distribution o f salinity in near headlands.
m angrove-fringed D ickson Inlet at both low and W hen runoff is im portant, the outwelled m ate
high tide. B rackish w ater, a rem nant o f the ru n rial rem ains in the river plum es. In areas where
off 10 days earlier, w as still present near the sur river runoff is always im portant, such as on the
face in the creek at low tide. A t high tide, the bulk m angrove-fringed coast o f T hailand facing the
o f th at w ater w as trapped in the swam p. Because A ndam an Sea near B urm a, longshore currents
buoyancy effects inhibit vertical mixing, the ensure th at these river plum es usually rem ain
brackish w ater is basically simply moving back coastally trap p ed together with the outwelled
and forth betw een the swam p at high tide and the m angrove detritus and nutrients (e.g. W attaya-
creek and low tide, but mixes very little with creek korn eta l., 1990). Any nutrient enrichm ent from
w ater and is thus only very slowly exported from the m angrove is then limited to coastal waters.
the system. Sim ilar coastal trapping o f steady river plum es
for hundreds o f km has been reported in N icara
C h w a k a B ay M a n g ro v e frin g e d tidal c re e k gua (M urray & Young, 1985) and in South A m er
o
ica (Bouyne & Roux, 1978).
30.5
.29.5' In tropical A ustralia, river runoff is short-lived.
Sill
34.5 F reshw ater is injected into coastal w aters in a
2
pulse-like event. O nce runoff ceases, tidal cur
rents m aintain vertical hom ogeneity in shallow
S alin ity (% o )
coastal w aters, and the brackish w aters behave
m ore or less like a passive tracer. T he fate o f this
4
tracer, i.e. the time it takes to be flushed out from
coastal w aters, provides a good indication o f the
dynam ics o f the coastal zone. Tw o examples are
Fig. 13. Longitudinal salinity distribution in the Chwaka bay illustrated below: the G ulf o f C arpentaria and the
mangrove swamp in Zanzibar, in the dry season, July 1988. coast o f the G reat B arrier Reef.
153
Figure 14 shows the distribution o f salinity in 3 times as wide as th at shown in Fig. 16 (W olan
the G u lf o f C arpentaria at 3 m onthly interval in ski & R idd, 1990). The non-linear interaction be
1976. The survey in A p ril-M a y corresponds to tween tidal currents and wind-driven currents is
the end o f the wet season. It shows th at the an essential factor in the dynam ics o f the coastal
brackish w ater w as coastally trapped and did not boundary layer.
mix readily with offshore waters. Five m onths
after, in A ugust-S eptem ber, runoff was negligible
but brackish w ater w as still found in the shallow Rugged coastline
coastal w aters (depth less than 15 m or so). This
finding suggests th at brackish w aters is trapped The east coast o f the G ulf o f C arpentaria is fairly
for several m onths in a coastal boundary layer, straight, with no m ajor headlands, and extensive
and th at cross-shelf mixing between offshore w a areas o f shallow w aters, the 20 m isobath being
ters and coastal boundary layer w ater is very slow located 20 km offshore. In contrast, the G reat
(W olanski & R idd, 1990). A satellite image in the Barrier R eef continental shelf has num erous
visible b ands taken in calm w eather (no wind headlands (Fig. 17) and the w aters are deeper, the
stirring o f bottom sedim ent) in the very dry sea 20 m contour line nearly touching these head
son o f N ovem ber, 1986 (zero runoff hence no lands. The salinity distribution in coastal w aters
river plum e) also shows a band o f w ater o f dif w as m easured after the flooding o f the Burdekin
ferent colour in the dry season (negligible runoff) River and other smaller stream s in January 1981.
along the co ast o f the G ulf of C arpentaria, with This distribution is shown in Fig. 17 for 2 6-27
a front running along the 15-20 m isobath January 1981, about 3 days past the peak o f the
(Fig. 15). The presence o f a coastal boundary flood and for 3 - 4 February 1981, 10 days after
layer m ay be explained by the effects o f friction in the flood. N ote th at in the first survey the plume
shallow w ater. Figure 16 shows the predicted m oved northw ard but w as continuous over
m ean Eulerian circulation under the action o f both 200 km and coastally trapped with the coast on
the tides and wind in the tradew ind season. This its left-hand-side, as dictated by Coriolis forces,
shows a narrow coastal jet extending to the 15 to though offshore the current was southw ard
20 m isobath w ith a return flow in deeper water. (W olanski & van Senden, 1983). In the second
T he tides alone drive a negligible net circulation survey, the plum e w as broken up in patches.
even in shallow coastal w aters, while wind alone Brackish w ater rem ained preferentially in the wide
in the absence o f tides drives a coastal jet about bays sheltered by headlands.
G u lf o f C arpentaria
23 April -1 1 M ay 2 7 J u n e - 2 July
Fig. 14. Salinity distribution in the G ulf of Carpentaria at about three-monthly interval in 1976. These data were kindly provided
by P. Rothlisberg.
154
o f C a r p e n ta r ia ®
¡34
B arro n R iver
C a irn si
Jo h n sto n e
R iv e r i
Coral Sea
c u r r e n t m e te r
tr a n s e c t
'» » » •B rito m a rt R e e f
•“i -m* -
2 6 - 2 7 Jan u ary , <s?’ 3 - 4 F e b ru a ry ,
T ow nsville,
B urdekin
^Hiverw
Fig. 17. Surface salinity distribution on the Great Barrier Reef continental shelf between the mouth of the Burdekin River and
Cairns, following flooding of the Burdekin River and other smaller streams, in 1981.
— *• 5 0 c m s _1
S la c k h ig h w a te r
Gay Head
2 km
Fig. 18. Observations of the velocity field and evidence for a strong eddy behind G ay Head, M assachusetts, USA, adapted from
Geyer & Signell (1990).
156
Fig. 20. Numerically predicted dispersion around a headland with reversing tidal currents in the absence of any net currents. Two
patches of particles are released at the beginning of the eastward flowing tide, one just off the tip of the headland and the other
some distance offshore (a). (b) shows the distribution after 6 tidal cycles, and (c) after 12 tidal cycles. Adapted from Signell &
Geyer (1990).
157
x
H (x ) Hf
h (x)
) -- v e r tic a l
e n tra in m e n t
F lo o d ti d e E b b tid e
Fig. 21. Upwelling in a tidal jet at (a) flood tide and (b) ebb tide.
these flow separation dynam ics, such as the The inclusion, or no inclusion, o f separation
‘phase eddy’ m odel, are unreliable because they flow effects in num erical m odels, is not ju st a
result in sm ooth distributions from gradient-type question o f academ ic interest only. Indeed, the
diffusion since they do no t sim ulate straining and num erical m odels used at present to sim ulate the
folding o f vortex lines. dispersal of crow n-of-thorn starfish in the G reat
(a )
Surface Surface
Bottom
Barrier Reef, a problem o f great practical and A ustralia (Fig. 23). T he prim ary reason for this is
m anagem ent im portance, neglect these dynam ics that, because o f friction, neither headland eddies
and predict sm ooth distributions and long reten nor the longshore currents found further offshore
tion times near reefs (e.g. Black, 1988). Such are generally able to propagate in these very shal
m odel predictions are contrary to expectations if low, sheltered, w aters (W olanski & R idd, 1990).
straining and folding o f vortex lines were included T rapping is reinforced w hen the em baym ents
in the models. are m angrove-fringed. Such is the case for in
A nother im portant mixing m echanism is th at stance at the northern region o f H inchinbrook
o f tidal jet-vortex pair system, the dynam ics o f C hannel where the w est coast is m angrove-fringed
w hich is also intim ately linked with flow separa and very shallow (Fig. 24a). W ater moves back
tion effects (Awaji eta l., 1980; Awaji, 1982; van and forth with the tides betw een the m angrove
Senden, 1985; W olanski et al., 1988). Awaji et al. sw am p and the coastal boundary layer in shallow
(1980) and Awaji (1982) found th at the pattern of
tran sp o rt and dispersion varies greatly over short
length scales and, exactly as in the case o f island
Bowlini
wakes, does n o t display the sm earing tendency of G ro a n /
a typical diffusive process.
Usually m angrove-fringed coastal w aters are
sufficiently shallow that, through friction (Ozsoy,
1977), tidal jets do no t prevail at ebb tides at the
m outh o f tidal creeks (W olanski & R idd, 1990).
; M o d e l s h o re lin e
The frictional slope term balances the surface
A c t u a l s h o re lin e
slope term in the continuity equation, and thus
the flow resem bles potential flow. However, there
exist som e m angrove creeks, e.g. T udor Creek in
M om basa, Kenya, th at face deep water. Com plex
m ilii
mm
f t t i t l í
and biologically im portant flow patterns probably B ow ling G r e e n B ay ' tidal w e tla n d s;
CARDWELL
Oyster Point)
O y s te r
Scraggy
V Point
Fig. 24. (a) Map of the northern region of Hinchinbrook Channel with depths in m. (b) and (c) show the plume, at low and high
tide, of a tracer after 15 days of continuous discharge at point A in the upper reaches of a mangrove creek in calm weather.
w aters on the w est coast. The coastal boundary such areas m ay simply have left another nearby
layer w aters being very shallow, there is little creek at the previous tidal cycle and is chemically
cross-shelf mixing between coastal boundary different from offshore water. Thus a m angrove
layer w aters and w aters further offshore in the creek at the upstream side o f a coastal boundary
deeper parts of the channel. W olanski et al. (1990) layer will have an apparent lower bulk return co
have m odeled the w ater circulation in this area. efficient than one at the dow nstream end.
The m angrove sw am ps were considered to be U nder the action o f w ind and tides, coastal
inter-tidal storage areas. The m odel results were boundary layer w aters, enriched in nutrient by
verified against field observations o f currents. A outwelling from the tidal w etlands, can be ejected
tracer was released continuously for 15 days at offshore as tidal jets peeling off capes and h ead
point A in the m odel, in the upper reaches o f a lands (B ow m an, 1988; W olanski & R idd, 1990).
m angrove creek, and followed through time. Fig This process m ay explain the presence o f baitfish
ure 24 (b and c) shows the tracer’s plum e at both and billfish schools found a few km offshore from
low tide and high tide after 15 days in calm C ape Bowling G reen (Williams, 1988).
w eather. T he co ntam inant is found in shallow
coastal w aters at low tide, and the bulk o f this References
contam inant is found trapped in the m angrove
sw am p at high tide. Awaji, T. A., N. Imasato & H. Kunishi, 1980. Tidal exchange
T he presence o f a coastal boundary layer in through a strait: A numerical experiment using a simple
shallow, m angrove-fringed, coastal w aters, pre model basin. J. Phys. Oceanogr. 10: 1499-1508.
Awaji, T., 1982. Water mixing in a tidal current and the ef
sents problem s in interpreting outwelling studies
fect ot turbulence on tidal exchange through a strait. J.
based on a budget o f m aterial exported over a Phys. Oceanogr. 12: 501-514.
tidal cycle by a m angrove creek (e.g. B oto & Bunt, Black, K. P., 1988. The relationship of reef hydrodynamics to
1981) since w ater entering a m angrove creek in variation in number of planktonic larvae on and around
160
coral reefs. Proc. 6th. Int. Coral Reef Symposium, Towns Ovalle, A. R. C., C. E. Rezende, L. D. Lacerda & C. A. R.
ville, Vol. 2. Silva, 1990. Factors affecting the hydrochemistry of a man
Black, K. P. & S. L. Gay, 1987. Eddy formation in unsteady grove tidal creek, Sepetiba Bay, Brazil. Estuar. coast. Shelf
flows. J. geophys. Res. 92: 9514-9522. Sei.: 639-650.
Boto, K. & J. S. Bunt, 1981. Tidal export o f particulate or Ozsoy, E., 1977. Flow and mass transport in the vicinity of
ganic matter from a northern Australian mangrove system. tidal inlets. Technical Rep. U FL/CO EL/TR-036, Coastal
Estuar. coast. Shelf Sei. 13: 247-255. and Oceanographic Engineering Department, University of
Bowden, K. F. & P. Hamilton, 1975. Some experiments with Florida.
a numerical model of circulation and mixing in a tidal es Pattiaratchi, C., A. Jam es & M. Collins, 1987. Island wakes
tuary. Estuar. coast. Mar. Sei. 3: 281-301. and headland eddies: a comparison between remotely
Bowman, M. J., 1988. Estuarine fronts. In B. Kjerfve (ed.), sensed data and laboratory experiments. J. geophys. Res.
Hydrodynamics of estuaries, Vol. 1, CRC Press, Boca 92: 783-794.
Raton, Florida, 85-132. Petryk, S. & G. Bosmajan, 1975. An analysis a flow through
Bouyne, P. & J. D. Roux, 1978. Remarques sur la circulation vegetation. J. Hydraul. Div., Am. Soc. Civil Eng. 101: 871-
d ’eaux turbides d’origine amazonienne le long de la cote des 874.
Guyanes. C.r. Acad. Sei. Paris, 287, Serie D.: 203-205. Pingree, R. D. & L. M addock, 1979. The tidal physics of
Burke, R. W. & K. H. Stolzenbach, 1983. Free surface flow headland flows and offshore tidal bank formation. Mar.
through salt m arsh grass. Massachusetts Institute o f Tech Geol. 32: 269-289.
nology Sea G rant College Program, Publication No. Ridd, P. V., M. W. Sandstrom & E. Wolanski, 1988. Out
M ITSG 83-16, Cambridge, MA., 252 pp. welling from tropical tidal salt flats. Estuar. coast. Shelf Sei.
Deleersnijder, E., E. Wolanski & A. Norro, 1989. Numerical 26: 243-253.
simulation of the three-dimensional tidal circulation in an Ridd, P. V., E. Wolanski & Y. Mazda, 1990. Longitudinal
island’s wake. In G. M. Carlomagno & C. A. Brebbia (eds), diffusion in mangrove-fringed tidal creeks. Estuar. coast.
Computer and experiments in fluid flow. Computational Shelf Sei. 31: 541-554.
Publications, Springer-Verlag, Berlin: 355-381. Robertson, A. I., D. M. Alongi, P. Daniel & K. G. Boto,
Elder, J. W., 1959. The dispersion of marked fluid in turbu 1988. How much mangrove detritus enters the G reat Bar
lent shear flow. J. Fluid Mech. 5: 544-560. rier reef lagoon? Proc. 6th. Inter. Symp. Coral Reefs,
Falconer, R. A., E. Wolanski & L. M ardapitta-Hadjipandeli, Townsville, Aug. 1988.
1986. Modelling tidal circulation in an island’s wake. J. Signell, R. P., 1989. Tidal dynamics and dispersion around
Waterway, Port, Coastal and Ocean Engineering, ASCE, coastal headlands. Ph.D. thesis, MIT/W HOI-89-38.
112: 234-254. Signell, R. P. & W. R. Geyer, 1990. Numerical simulation of
Fischer, H. B., E. J. List, R. C. Y. Koh, J. Imberger & N. H. tidal dispersion around a coastal headland. In R. T. Cheng
Brooks, 1979. Mixing in inland and coastal waters. Aca (ed.), Residual currents and long-term transport in estuar
demic Press, New York, 484 pp. ies and bays, Lecture notes on Coastal and Estuarine Stud
Geyer, W. R. & R. P. Signell, 1990. A reassessment of the role ies, Springer-Verlag: 210-222.
of tidal dispersion in estuaries and bays. Estuaries, in press. Simpson, J. H. & I. D. James, 1986. Convergent fronts in the
Geyer, W. R. & R. P. Signell, 1990. Measurements of tidal circulation on tidal estuaries. In D. A. Wolfe (ed.), Estua
flow around a headland with a shipboard acoustic Doppler rine variability. Academic Press, London.
current profiler. J. geophys. Res. 95: 3159-3197. Taylor, G. I., 1921. Diffusion by continuous movements. Proc.
Mazda, Y., Y. Sato, S. Swamoto, H. Yokochi & E. Wolanski, London Math. Soc., Ser. A, 20: 196-211.
1990. Links between physical, chemical and biological pro Tee, K. T., 1976. Tide-induced residual current, a 2-D non
cesses in Bashita-Minato, a mangrove swamp in Japan. linear numerical tidal model. J. Mar. Res. 34: 603-628.
Estuar. coast. Shelf Sei., in press. Thompson, R. O. R. Y. & T. J. Golding, 1981. Tidally in
M azda, Y., H. Yokochi & Y. Sato, 1990. The behaviour of duced upwelling by the G reat Barrier Reef. J. geophys.
groundwater in a mangrove area and the influence on the Res. 86: 6517-6521.
properties of water and bottom mud. Estuar. coast. Shelf Uncles, R., J. E. Ong & W. K. Gong, 1990. Observations and
Sei. 31: 621-638. analysis of a stratification-destratification event in a tropi
Murray, S. P. & M. Young, 1985. The nearshore current along cal estuary. Estuar. coast. Shelf Sei. 31: 651-666.
ahigh-rainfall, tradewind coast - Nicaragua. Estuar. Coast. van Senden, D., 1985. Structure o f tidal jets under variable
Shelf Sei. 21: 687-699. source conditions. Proc. 21st Congress I.A .H.R., Mel
Nokes, R. I., 1986. Problems in turbulent dispersion. Ph.D. bourne, 3C, pp. A I/3-A I/8.
Thesis, Dept. Civil Eng., Univ. of Canterbury, W attayakorn, G., E. W olanski & B. Kjerfve, 1990. Mixing,
Christchurch, New Zealand, 229 pp. trapping and outwelling in the Klong Ngao mangrove
Okubo, A., 1973. Effect of shoreline irregularities on stream- swamp, Thailand. Estuar. coast. Shelf Sei., in press.
wise dispersion in estuaries and other embayments. Neth. Williams, D. McB., 1988. Significance o f coastal resources to
J. Sea Res. 6: 213-224. sailfish and blackmarlin in N E Australia: an on-going re
161
search program. Proc. International Billfish Symposium II, bays. Lecture notes on Coastal and Estuarine Studies,
Kailua-Kona, Hawaii, Aug. 1-5, 1988. Springer-Verlag: 165-183.
Wolanski, E., 1986. An evaporation-driven salinity maximum Wolanski, E. & D. van Senden, 1983. Mixing of Burdekin
zone in Australian tropical estuaries. Estuar. coast. Shelf River flood waters in the Great Barrier Reef. Austr. J. mar.
Sei. 22: 415-424. Freshwat. Res. 34: 49-63.
Wolanski, E., 1989. Measurements and modeling of the water Wolanski, E., D. Burrage & B. King, 1989. Trapping and
circulation in mangrove swamps. UNESCO-CO M ARA F dispersion o f coral eggs around Bowden Reef, G reat Bar
Serie Documentaire No. 3, 43 pp. rier Reef, following mass coral spawning. Contin. Shelf
Wolanski, E., 1989. Circulation anomalies in tropical Austra Res. 9: 479-496.
lian estuaries. In B. J. Kjerfve (ed.), Hydrodynamics of Wolanski, E., J. Imberger & M. L. Heron, 1984. Island wakes
estuaries. Vol. II, CRC Press, Boca Raton: 53-59. in shallow coastal waters. J. geophys. Res. 89: 10553—
Wolanski, E. & R. Gardiner, 1981. Flushing of salt from 10569.
mangroves. Aust. J. mar. Freshwat. Res. 32: 681-683. Wolanski, E., M. Jones & J. S. Bunt, 1980. Hydrodynamics
Wolanski, E. & W. M. Hamner, 1988. Topographically con of a tidal creek-mangrove swamp system. Aust. J. mar.
trolled fronts in the ocean and their biological influence. Freshwat. Res. 31: 431-450.
Science 241: 177-181. Wolanski, E., E. Drew, K. Abei & J. O’Brien, 1988. Tidal jets,
Wolanski, E. & B. King, 1990. Flushing of Bowden Reef la nutrient upwelling, and their influence on the productivity
goon. Estuar. coast. Shelf Sei. 31: 789-804. of the alga Halimeda in the Ribbon Reefs, G reat Barrier
Wolanski, E. & P. Ridd, 1986. Tidal mixing and trapping in Reef. Estuar. coast. Shelf Sei. 26: 169-201.
mangrove swamps. Estuar. coast. Shelf Sei. 23: 759-771. Wolanski, E., Y. M azda, B. King & S. Gay, 1990. Dynam
Wolanski, E. & P. Ridd, 1990. Mixing and trapping in Aus ics, flushing and trapping in Hinchinbrook Channel, a giant
tralian tropical coastal waters. In R. T. Cheng (ed.), Re mangrove swamp, Australia. Estuar. coast. Shelf Sei. 31:
sidual currents and long-term transport in estuaries and 555-580.
!
Hydrobiologia 247: 1 6 3 -1 7 1 , 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 163
© 1992 Kluwer Academic Publishers. Printed in Belgium.
K ey words: m angrove estuary, salt transport, estuarine stratification, dispersion, vertical shear
tran sp o rt
Abstract
The dependence o f salinity stratification on the vertical structure o f the longitudinal salinity gradient is
investigated by solving the equations o f longitudinal m om entum and salt conservation. The results are
used to interpret m easurem ents m ade throughout 31 tidal cycles in the M erbok Estuary, M alaysia, as
p art o f an ecological study o f a tropical m angrove estuary. The solutions show th at a substantial part
o f the salinity stratification was caused by vertical shear in the currents coupled with advection of the
longitudinal salinity gradient. This m echanism led to m axim um stratification around low w ater and
m inim um stratification around high w ater. O bserved intratidal variations in stratification at neap tides
greatly exceeded m odelled values when a depth-independent, longitudinal salinity gradient w as assum ed.
Sim ulations m ade using a longitudinal salinity gradient o f simple depth-dependent, power-law form
implied th at this difference arose m ainly because the near-surface, longitudinal salinity gradient in the
field w as significantly larger than th at near the bed during neap tides and at times o f high run-off. The
longitudinal, dispersive tran sp o rt o f salt, fresh w ater and solutes due to vertical shear is shown to be
im portant during periods o f stratification at neap tides and during high run-off. The relevance o f strat
ification to m angrove estuaries is discussed.
correct trends and tidal averages but failed to phasize the im portance o f stratification and the
reproduce sufficiently strong intratidal fluctua physical processes affecting it, as well as com ple
tions in stratification. In this paper we consider m ent recent w ork on this topic (N unes et al., 1989;
intratidal processes in m ore detail and show th at S im pson et al., 1990; U ncles é ta l., 1990; Sim p
it is necessary to specify the vertical profile o f the son & Sharpies (in press)).
longitudinal density gradient if the intratidal fluc
tuations are to be reproduced correctly.
This article briefly reviews observed d ata for Observations
the M erbok Estuary, followed by an outline o f the
m athem atical m odel and its solutions. T he orig Study area and measurements
inal aspects are the treatm ent o f a schem atic form
for the vertical profile o f longitudinal density gra T he M erbok Estuary (Fig. 1) is situated in the
dient and the analysis o f intratidal fluctuations in state o f K edah, M alaysia (5° 40' N , 1 0 0 °2 5 ' E).
stratification. It is n o t our objective to derive ‘best The tidal influence extends for 35 km. The width
fits’ betw een m odel and data, but rather to em ranges from about 2 km at the m outh to about
#4 5 V
Tide
ga u g e
Sungai M erb ok U iC
Sungai
Petani
STRAITS
Xi R e c la im e d a rea
MALACCA
S. Muda
Fig. 1. The Merbok Estuary, northwest Malaysia, showing the deep-channel measurement site and the location of the tide-gauge.
165
20 m tow ards the upper reaches o f the estuary, m ean depth. The depth-averaged tidal currents
w ith depths generally varying from 15 to 3 m. The (ebb positive) are show n in Fig. 2A against time
m ean annual freshw ater discharge is estim ated to in lunar hours. The R M S current speeds ranged
be 20 m 3 s " l . from about 0.2 m s -1 during the neap tide to
M easurem ents were m ade throughout 31 tidal 0.7 m s “ 1 during springs. F reshw ater runoff d ata
cycles during June 1987 as p art o f an ecological are also show n in Fig. 2A (plotted as the dashed
study o f a tropical m angrove estuary. F our sta line at 10 times actual value for clarity). R unoff
tions were w orked near the m outh o f the M erbok reached a m axim um o f about 15 m 3 s 1 during
from 0730 h on 2 June until 1100 h on 18 June the neap tides.
1987. C urrent velocities were m easured every The surface and bed salinity d ata are shown in
lunar hour at five depths using calibrated current Fig. 2B. A considerable reduction in surface sa
vanes (Kjerfve & M edeiros, 1989). W ater sam linity occurred during the neap tide, w hereas the
ples were collected and analysed for salinity at m id-depth salinity (not show n) and bed salinity
three depths (surface, m id-depth and bed) every rem ained high. This reduction was due to weak
alternate lunar hour and at one depth (m id-depth) vertical mixing during the neap-tide period and
every other lunar hour. W e will present d ata from increased runoff (Fig. 2A). The stratification (the
the deep-channel (15 m) station in order to ex difference in salinity between bottom and surface
tend the analyses presented in U ncles et al. (1990) w aters) becam e very small over the spring tide
and to provide additional insights into the phys due to increased mixing and show ed pronounced
ical processes at w ork in b oth m angrove and tem intratidal variability over the period.
perate estuaries.
co
V
CO lo o 30
S
c/i
20
100
0 10 0 200 300 400 0 100 20 0 300 400
Fig. 2. Currents, salinity and runoff data against time in lunar hours. (A): Estimated freshwater runoff (dashed curve, multiplied
by 10 for clarity, m 3 s “ 1) and depth-averaged tidal currents (cm s “ 1, ebb positive); (B): Surface and near-bed (dashed curve)
salinity (ppt).
166
S
M
CO
>"
H 0 .5
O. s
Pm u
Eh
CL,
CL,
m id-depth) were little affected by runoff and tidal persion coefficient is about 450 m 2 s ~ 1 during
state over the period (Fig. 2B), m uch o f the strong strong stratification at neap tides and less than
longitudinal salinity gradient at 100 h m ust have about 50 m 2 s ” 1 during spring tides. Therefore,
been confined to the upper half o f the w ater col during neap tides, the vertical shear flux is im
umn. p o rtan t for flushing fresh w ater and solutes from
the M erbok E stuary to its coastal w aters and for
im porting salt water.
Vertical shear salt transport
The vertical shear dispersive flux o f salt is de I n tr a tid a l s tra tific a tio n m o d el
noted by F and is defined as (U ncles et al., 1985;
M urray & Siripong, 1978; D ronkers & van de Basic equations
Kreeke, 1986; H unkins, 1981):
A vertical mixing m odel has been presented in
F=jtrw) U ncles et al. (1990). The m odel includes the com
with U' = U - U and S ' = S - S and where the peting influences o f tidal mixing and buoyancy.
overbar is a depth average. An estim ate o f the T he longitudinal acceleration o f currents is the
dispersion coefficient due to vertical shear is: result o f an im balance betw een vertical eddy vis
cosity and horizontal pressure gradients. There is
D = - <F}/dx (S > no wind stress at the surface and a quadratic
w here < ) denotes a tidal average. friction law applies at 1 m above the bed. Longi
The tidally-averaged salt flux due to vertical tudinal advection is ignored, so th at the only n o n
shear, < F > , is always directed up-estuary but is linear aspects o f the m odel are the shear and
small during spring tides (Fig. 3B). Dividing the stratification dependence o f the eddy viscosity,
tidally averaged salt flux by the depth-averaged and the quadratic dependence o f frictional drag
longitudinal salinity gradient (Fig. 3A) gives an on near-bed velocity.
estim ate o f the dispersion coefficient, D, due to The m om entum equation can be w ritten (with,
vertical shear (plotted as the dashed line at one for example, dt = 8/dt):
tenth actual value for clarity in Fig. 3B). The dis
167
and mixing length and possess the classical depth (that is, the values assigned to ß and n in
M unk-A nderson dam ping dependence on water- Equation 2).
colum n stability (U ncles et al., 1990). Figures 4(A,B) show hypothetical, vertical p ro
files of the longitudinal density (or salinity) gra
dient, relative to its depth-averaged values, for
Results of model ß= - 1/3 and ß = - 1 in Equation 2, respectively.
The case ß = - 1 corresponds to zero longitudinal
Longitudinal density-gradient profiles gradient at the bed; /?=0 (equivalent to n = 0 )
corresponds to a longitudinal gradient which is
H aving estim ated the depth-averaged, longitudi everywhere equal to its depth-averaged value, and
nal density (salinity) gradient (Fig. 3A), the strat ß= - 1/3 is an interm ediate and m ore realistic
ification, vertical profile o f longitudinal current case. Show n in Figs 4(A,B) are the vertical p ro
and the longitudinal salt flux due to vertical shear, files corresponding to n = 5, 10 and 20, (E qua
F, can be com puted using E quations 1-6. The tion 2 ) together with vertical, dashed lines corre
results depend upon the way in w hich the longi sponding to the depth averages (or n = 0 ).
tudinal density gradient is assum ed to vary with Vertical profiles having values o f n greater than
M K
\ 0 .5 \ 0 .5
N N
0
0 5 0 5 10
RELATIVE d S / d X RELATIVE d S / d X
15
20
H H
CU 10 CU
cu CU
10
5
0
0 100 200 400 0 100 20 0 300 400
Fig. 4. The effects of a hypothetical, depth-dependent, longitudinal density (salinity) gradient. (A): Vertical profiles of dxS for
ß= - 1/3 in E quation2 and n = 5 ( - ) , 10 ( ---- ) and 20 ( - • - ) ; (B): The same except that ƒ!= - 1; (C): Observed stratification
(ppt) against time; (D): Modelled stratification (ppt) for ß= - 1/3 in Equation 2 and for n = 0 ( - • - ), 5 ( - ) and 10 ( ---- ) against
time. Note that Z = z + h in (A) and (B).
169
about 5 dem onstrate an alm ost uniform value of variations in stratification cannot be reproduced
the longitudinal density gradient in the bottom with n = 0 (ƒ?= 0 ), which corresponds to a depth-
half o f the w ater colum n (Figs 4(A,B)). This n ear independent longitudinal density gradient, al
uniform ity o f the hypothetical profiles is consis though the m agnitude of the tidally-averaged
tent with the small intratidal variations in salin stratification is reproduced quite well. W ith n = 5
ity which were observed to occur during neap and n = 10 (Fig. 4D ) the intratidal fluctuations in
tides in the bottom half o f the w ater colum n in the stratification are strong and are such th at maxi
M erbok (Fig. 2B). m um and m inim um stratification tend to occur at
low and high w ater, respectively, as observed in
the M erbok.
Calculations o f intratidal variability The m odelled stratification is greater with in
creasing n. A lm ost-fresh w ater is predicted to
Results are presented o f a calculation in which occur in the surface layers during m inim um neap
the depth-averaged, longitudinal density gradient tides when n = 10. This result is a consequence of
corresponds to the estim ated values show n in our assum ption th at the longitudinal density gra
Fig. 3A. By using the time-varying, longitudinal dient is independent o f distance along the estuary
density gradient we are also effectively taking into and is therefore no t affected by the com puted salt
account the time-varying freshw ater flow. A value flux. T he neglect o f density-gradient variations
o f ß = - 1/3 is used in Equation 2 as an example along the estuary also precludes us from m odel
o f the effects o f a hypothetical, vertical profile in ling the behaviour o f a surface, freshw ater lens. In
the longitudinal density gradient and calculations som e estuaries, such as the H aw kesbury, P arra
are m ade for values o f n = 0, 5, 10 and 20. The m atta and P o rt H acking Estuaries in N ew South
case n = 0 corresponds to a constant value o f the W ales, A ustralia, a short-lived (few hours to a
longitudinal density gradient throughout the w ater day) freshw ater spate can flow to sea as a surface
colum n (equivalent to ß = 0). In nature, one would lens and have little effect on the deeper w aters
expect b oth ß and n to vary with time and espe (W olanski, personal com m unication).
cially with the state o f the spring-neap cycle. T he m odelled, intratidal fluctuations in strati
The observed stratification (bed m inus surface fication for n = 5 and 10 at neap tides are very
salinity) is shown in Fig. 4C. The intratidal fluc similar to those observed, w hereas at spring tides
tuations are very pronounced. M axim um strati they are m uch larger. This is a result o f our as
fication tends to occur at low w ater and m inimum sum ption th at the vertical profile o f longitudinal
stratification at high w ater. D uring ebb currents, density gradient does no t change with time. In
low salinity w ater is advected dow n-estuary in the reality, enhanced vertical mixing at spring tides
faster surface layers, which tends to minim ize sa will tend to reduce variations w ith depth and thus
linity at the surface around low w ater and m ax decrease the absolute m agnitudes o f ß and n.
imize stratification. D uring the flood, vertical
shear in the tidal currents is reduced by the
density-driven circulation, so th at the tendency Relevance of stratification to mangrove systems
for stratification to occur as a result of differen
tial advection o f the longitudinal density gradient The presence o f m angrove sw am ps is m ost likely
is less. T he m odelled stratification is shown in to influence the w ater colum n stratification indi
Fig. 4D for values o f n = 0 (dash-dot line), 5 (con rectly, through an influence on m orphology and
tinuous line) and 10 (dashed line). There is very thus tidal flow and mixing and through an ability
little difference in the m odelled stratification be to trap fresher w ater and thus m odify longitudi
tween the cases for n= 10 and n = 20 , so th at only nal salinity gradients. H ow ever, the longitudinal,
the case for n = 10 is plotted in Fig. 4D. dispersive tran sp o rt o f salt and freshw ater due to
It is clear from Figs 4(C ,D ) th at the intratidal vertical shear in both density-driven and tidal
170
currents, coupled with stratification, will directly due to vertical shear occurred approxim ately d u r
influence the export o f nutrients, m angrove litter ing m axim um stratification, w hen vertical mixing
and detritus from the m angrove system to the w as very small. This is consistent with theoreti
coastal area and m ay thus affect the productivity cal analyses (Fischer etal., 1979) and with the
o f coastal fisheries. conclusion o f N unes et al. (1989) th at the spring-
W olanski & R idd (1986) have shown th at tidal neap cycle has an im portant influence on estua-
trapping in the sw am ps o f m angrove estuaries rine stratification and m ass transport. This tra n s
can enhance longitudinal dispersion o f nutrients po rt will influence the export o f nutrients,
and other m aterials by tem porarily storing w ater m angrove litter and detritus from the m angrove
in the sw am ps during the late flood and releasing system to the coastal area.
it to a different part o f the m ain flow during the
ebb. It is likely th at stratification will m odify this
effective longitudinal dispersion o f fresh w ater and Acknowledgements
nutrients by supplying fresher and m ore nutrient-
rich surface w aters to the swam p. B ecause of This project (‘M angrove M alaysia’, C entre file
stratification, it is possible th at m angroves will 3-P-82-0192) w as funded by the International
experience less stress and therefore spend less D evelopm ent and R esearch C entre (IR D C ) of
energy dealing with salt in their effort to extract C an ad a with infrastructure support from the Uni-
fresh w ater for physiological and m etabolical p ro versiti Sains M alaysia. W e are particularly grate
cesses. It is also possible, as a result o f fresher ful to D r B rian Davy, A ssociate D irector (F ish
w ater flooding the m angrove m uds during the eries) and his staff at the IR D C A sia Regional
neap stratification, th at m ore nutrient leaching C entre in Singapore for their excellent support.
and export m ay occur. W e are also grateful to the B ritish C ouncil for
providing a travel grant th at allowed one o f us (R
J U ) to participate in this aspect o f the project.
Conclusions O ur thanks go to D r W ong Chee H oong and D r
Z ubir D in as well as the 60 or so laboratory staff
(1) A substantial p art o f the salinity stratification and students o f U niversiti Sains M alaysia who
w as caused by vertical shear in the currents cou cheerfully assisted in the very arduous field and
pled with advection o f the longitudinal salinity laboratory work.
gradient. This m echanism led to m axim um strat
ification around low w ater and m inim um strati
fication around high water.
References
(2) O bserved intratidal variations at neap tides
greatly exceeded m odelled values w hen a depth-
Boto, K. G. & J. S. Bunt, 1981. Tidal export of particulate
independent, longitudinal density gradient was organic matter from a northern Australian mangrove sys
assum ed. Sim ulations m ade using a hypothetical, tem. Estuar. coast. Shelf Sei. 13: 247-255.
longitudinal density gradient o f pow er-law form Boto, K. G., J. S. Bunt & J. Wellington, 1984. Variations in
mangrove forest productivity in northern Australia and
with depth dem onstrated th at this difference arose
Papua New Guinea. Estuar. coast. Shelf Sei. 19: 321-329.
mainly because the near-surface, longitudinal Bowden, K. F. & P. Hamilton, 1975. Some experiments with
density gradient in the field w as significantly larger a numerical model of circulation and mixing in a tidal es
than values lower in the w ater colum n during tuary. Estuar. coast, mar. Sei. 3: 281-301.
neap tides and at times of high run-off. Dronkers, J. & J. van de Kreeke, 1986. Experimental deter
mination of salt intrusion mechanisms in the Volkerak Es
(3) The tidally-averaged salt tran sp o rt due to
tuary. Neth. J. Sea Res. 20: 1-19.
vertical shear was directed up-estuary and was Dyer, K. R., 1974. The salt balance in stratified estuaries.
m uch smaller during spring tides. The greatest Estuar. coast, mar. Sei. 2: 273-281.
longitudinal salt flux and dispersion coefficient Fischer, H. B., E. J. List, R. C. Y. Koh, J. Imberger & N. H.
171
Brooks, 1979. Mixing in inland and coastal waters. Aca Simpson, J. H., J. Brown, J. Mathews & G. Allen, 1990. Tidal
demic Press, New York, 483 pp. straining, density currents and stirring in the control of
Hunkins, K., 1981. Salt dispersion in the Hudson Estuary. J. estuarine stratification. Estuaries 13: 125-132.
Phys. Oceanogr. 11: 729-738. Simpson, J. H. & J. Sharpies, In press. Dynamically-active
Kjerfve, B. & C. Medeiros, 1989. Current vanes for measur models in the prediction of estuarine stratification. In D.
ing tidal currents in Estuaries. Estuar. coast. Shelf Sei. 28: Prandle (ed.), Physics of Estuaries and Coastal Seas.
87-93. Springer-Verlag, Berlin/Heideberg.
Murray, S. P. & A. Siripong, 1978. Role o f lateral gradients Uncles, R. J., R. C. A. Elliott & S. A. Weston, 1985. Disper
and longitudinal dispersion in the salt balance of a shallow sion of salt and suspended sediment in a partly mixed es
well mixed estuary. In: B. Kjerfve (ed.), Estuarine Trans tuary. Estuaries 8: 256-269.
port Processes. University of South Carolina Press: 113— Uncles, R. J., J. E. Ong & W. K. Gong, 1990. Observations
124. and analysis of a stratification-destratification event in a
Nunes Vaz, R. A., G. W. Lennon & J. R. de Silva Samaras- tropical estuary. Estuar. coast. Shelf Sei. 31: 651-665.
inghe, 1989. The negative role of turbulence in estuarine Wolanski, E. & P. V. Ridd, 1986. Tidal mixing and trapping
mass transport, Estuar. coast. Shelf Sei. 28: 361-377. in mangrove swamps. Estuar. coast. Shelf Sei. 23: 759-
Rattray, M. & J. G. Dworski, 1980. Comparison o f methods 771.
for analysis of the transverse and vertical circulation con Wolanski, E., Yoshihiro M azda, B. King & S. Gay, 1990.
tributions to the longitudinal advective salt flux in estuar Dynamics, flushing and trapping in Hinchinbrook Channel,
ies. Estuar. Coast. Mar. Sei. 11: 515-536. a giant mangrove swamp, Australia. Estuar. coast. Shelf
Ridd, P. V., M. W. Sandstrom & E. Wolanski, 1988. Out- Sei. 31: 555-579.
welling from tropical tidal salt flats. Estuar. coast. Shelf Sei. Wolanski, E. Personal Communication. Australian Institute
26: 243-253. of Marine Science, PMB No. 3, Townsville M.C.. OLD
Ridd, P., E. Wolanski & Yoshihiro Mazda, 1990. Longitudi 4810, Australia.
nal diffusion in mangrove-fringed tidal creeks. Estuar. coast.
Shelf Sei. 31: 541-554.
Hydrobiologia 247: 1 7 3 -1 7 9 , 1992.
V. Jaccarini & E. Martens (eds). The Ecology o f Mangrove and Related Ecosystems. 173
© 1992 Kluwer Academic Publishers. Printed in Belgium.
Julius Francis
Institute o f Marine Sciences, P.O. B ox 668, Zanzibar, Tanzania
K ey words: m angrove sw am ps, trapping, mixing, estuarine circulation, coastal currents, Rufiji delta
Abstract
The physical processes such as coastal currents, estuarine circulation and m onsoon w inds prevailing in
the Rufiji delta are discussed. The relationship between these processes and the occurrence o f long-term
trapping o f the river discharge and the outflow o f w aters from the m angrove sw am ps into the nearshore
zone has been observed. The trapped w aters in the nearshore zone significantly reduce the mixing be
tw een the estuarine and offshore w aters, leading to the tw o w aters having distinctive properties. The
existence o f the trapped w aters in the nearshore zone is supported by evidence from a satellite image
and aerial photograph and m easurem ents o f salinity and suspended sedim ent concentration. The tra p
ping phenom enon is discussed in the light o f its possible im plications on the ecology o f m angrove ec
osystem s. T rapping m ay explain the enhanced grow th o f the m angrove in the delta com pared to other
areas. This trapping effect may be providing m ore time for nutrient retention in the m angrove zone,
incorporation o f the decom posed leaf litter and fine sedim ents in the substrate, and settling o f fruits and
seedlings in the sw am ps, thereby enhancing the regeneration o f the m angrove.
\ KENYA
T A N Z A N IA
S im b a Uranga
D odom a
D a r -es-
S a la a m K iom boni
*rin9a /R ufiji"
' R.
0 200 km
M sala
>3A) M sa la
K iasi
K iasi
R. Rufiji 2A
M som eni
16km
.Mohoro
.Yaya
Fig. 1. M ap of Rufiji delta area; A -B Suspended sediment concentration transect line, J- Station at the mouth of tidal creek J,
and S- Station at the mouth o f Simba U ranga channel.
p rotected to the SW m onsoon winds. G enerally zone and in the m angrove sw am ps have been
however, T an zan ian coast experiences stronger established in several areas such as C hw aka Bay
w inds during the SW m onsoon com pared to the in Z an zib ar and in the G u lf o f C arpentaria and
N E m onsoon (Bryceson, 1977). in its adjacent m angrove sw am ps in A ustralia
The freshw ater discharge from the delta chan (W olanski & R idd, 1986; W olanski & R idd, 1989;
nels o f the Rufiji is trapped in the nearshore zone. W olanski, 1989).
T he prevailing w inds and current conditions are Besides the report by Iversen etal. (1984),
thought to be responsible for the existence o f this which has discussed briefly the existence o f the
trapping (Iversen et al., 1984). The existence and trapping effect in the nearshore zone off the Rufiji
the causes o f the trapping effect in the nearshore delta, there are no other report th at have dis-
175
cussed the subject. K nowledge o f the existence of paper have also been obtained from the interpre
the trapping effect and its im plications on the tation o f satellite image and aerial photograph.
m angrove ecosystem s is fundam ental to our u n
derstanding o f the processes th at might be affect
ing the ecology o f the m angrove ecosystem s. Results and discussion
This p aper intends to establish the existence
and the causes o f the trapping effect in the m an A t the m outh o f J creek, the salinity values for the
grove sw am ps and in the nearshore zone off the surface w aters were found to fluctuate slightly at
Rufiji delta, and to discuss the im plications of the tidal frequencies, being relatively high during flood
effect on the m angrove ecosystems. tide and low during the ebb tide period. The sa
linity values ranged from 10.60%o to 14.26%0 (see
Fig. 2a). The d ata collected at a depth o f m ore
Material and methods than 1 m below the surface showed th at during
the flood tide m ore saline w ater w as flowing into
The salinity m easurem ents were done along the the creek, as the values o f salinity ranged from
Sim ba U ran g a channel (see Fig. 1). The channel 18.12%0 to 30.70%o. The small variation and small
w as selected in this study because it is the m ain range in salinity values at the J creek over tidal
branch in the delta river system. The m easure cycle suggest there is little mixing between the
m ents o f suspended sedim ent concentration were w ater from the creek and the w ater in the ch an
done along the Sim ba U ranga and M sala chan nel. Also the low salinity values obtained were
nels however, in this paper only d ata from the probably due to freshw ater discharged several
latter has been used. weeks earlier during the rainy season, as for the
A W PA cm 35 conductivity m eter was used to period o f about one m onth preceding the d ata
obtain the conductivity o f the w ater. T he conduc collection there w as no rain in the areas around
tivity readings were later converted to salinity the Sim ba U ranga Island and the floods o f the
readings. The conductivity d ata were taken from Rufiji river h ad ceased about two m onths before.
tw o anchored stations. The first station at the All these facts are implying the existence o f the
m iddle o f the m outh o f the J creek (see Fig. 1). trapping o f low salinity w ater in the mangrove
The creek is about 100 m wide at the m outh and swam ps.
1 km long. This creek drains only the m angrove T he surface salinity values obtained at the
sw am p around it. The second station w as at the m outh o f the Sim ba U ranga channel showed a
m iddle o f the m outh o f the Sim ba U ranga ch an relatively large variation between the peak flood
nel. The m easurem ents at the tw o stations were tide values and the corresponding ebb tide values.
taken for four tidal cycles in four consecutive days The salinity values were low during the ebb tide
i.e. from 21 July, 1990 to 24 July, 1990. and high during the flood tide period (see Fig. 2b).
S uspended sedim ent concentration m easure This large variation can be explained by the fact
m ents were done using the P artech Suspended th at the m ore saline nearshore w aters move in
M atter M onitor. A t M sala channel the m easure land along the channel during the flood tide and
m ents were m ade in a transect running from the back to the nearshore zone during the ebb tide.
offshore to the m outh o f the channel (see Fig. 1). M easurem ents o f suspended sedim ent concen
This w as done purposely to determ ine the varia tration taken during the ebb tide at the M sala
tion o f sedim ent concentration in the estuarine, channel show ed th at there w as high concentra
nearshore, and offshore w aters. The m easure tion o f sedim ent in the nearshore w aters com
m ents were done on the ebb tide as the sedim ent pared to either the estuarine or the offshore w a
concentration on the w ater is m axim um during ters. The concentration ranged from less than
this period. At M sala channel, the d ata were col 490 mgl “ 1 for the estuarine w aters to about
lected on 5 Septem ber, 1989. Inform ation for this 1000 mgl - 1 for the nearshore w aters. F or the off
176
14-
O
o
o 10-
>
t—
z
—I
<
IO
HW LW
TIME (HOURS)
30-
25-
O
o
20 -
>
I-
z
_i 15-
<
io
10 -
HW
TIME (HOURS)
Fig. 2. Variation over a tidal cycle o f surface salinity; (a) at the mouth of tidal creek J - 21 July, 1990 and (b) at the mouth of
Simba Uranga channel - 24 July, 1990.
shore w aters zero values were dom inant (as offshore w aters. It is clear th at the separation
show n in Table 1). A parallel picture is also draw n between the nearshore and the offshore w aters is
from the aerial photograph (Fig. 3), which indi well defined in m any areas, suggesting the ab
cates th at nearshore w aters contain m ore sus sence o f intensive mixing between the tw o w ater
pended sedim ent than either the estuarine or the m asses. The suspended sedim ent concentration
177
Table 1. Suspended sediment concentration along the m easurem ents and the interpretation o f the aerial
transect A -B .
photograph indicate a trapping effect in the n ear
Time (hour) Depth (m) Concentration (mg 1 ') shore w aters off the delta.
The E ast African C oastal C urrent (EA CC )
1235 3.00 490 system m ay be responsible for causing this tra p
1235 1.00 480 ping effect. They are generally flowing northw ards
1245 1.00 500
throughout the year, being faster during the SW
1251 1.00 490
1304 0.50 720 m onsoon and slower during the N E m onsoon
1312 0.50 890 (Newell, 1959). The aerial photography shows the
1320 0.30 1000 existence o f northw ard flowing currents, which
1332 1.50 1000 could be the EA C C system. These currents are
1345 1.00 980
inhibiting the flow o f nearshore w aters into the
1256 1.00 1000
1405 1.00 50 deeper offshore w aters, consequently leading to
1416 1.00 100 the outflows from the delta channels being re
1428 1.00 150 stricted in the nearshore zone.
1440 1.00 0 The Sim ba U ranga channel being stratified
1456 1.00 0
during the period o f high discharge and partially
Fig. 3. Satellite image and aerial photograph; (a) Landsat image of the Rufiji delta with dates: 23 June, 1981, 23 June, 1981, 24
June, 1981, and 18 March, 1985 and (b) Aerial photograph o f part of the delta taken on 10 June, 1965.
178
well mixed during the low freshw ater discharge found together with broken shells. In view o f this
suggests the absence o f strong mixing processes. it seems th a t the m angrove are receiving sedi
It seems th at the m onsoon w inds an d /o r the tidal m ents from bo th terrestrial and m arine environ
currents are unable to cause intensive mixing b e m ents at different times in a year.
tween either the estuarine and the nearshore w a
ters or the nearshore and the offshore waters.
The trapping effect discussed above m ay have Conclusion
several im plications on the ecology o f the m an
grove ecosystem s as will be indicated in the fore Physical processes play a significant role in the
going discussion. The trapping effect m ay be existence o f the long-term trapping effect in the
am ong factors controlling the am ount o f sus m angrove sw am ps and in the nearshore zone.
pended particulate m atter incorporated in the T he presence o f n o rthw ard flowing E A C C system
substrate in the m angrove sw am ps relative to the and the slow mixing processes are probably the
am ount exported to the adjacent m arine ecosys m ain factors responsible for the existence o f the
tem s. O ther factors include the substrate charac trapping effect in the sw am ps and in the n ear
teristics, the frequency o f tidal inundation, the shore zone. The intensity o f trapping effect m ay
energy o f the tide, and the pattern o f production be varying w ith the m onsoon season, being m ore
and supply o f m angrove litter (W oodroffe, 1985). pronounced during SW m onsoon than during the
This is due to the fact th at trapping provide m ore N E m onsoon. This is due to the fact th at the N E
tim e for the suspended m atter to settle out. Also m onsoon are characterized by w eaker EA C C
m ore tim e is provided for the fruits and seedlings system, lower freshw ater discharge, and relatively
of m angrove to settle in the swam p and regener strong m onsoon winds.
ate instead o f being w ashed away by the tidal The enhanced grow th o f the m angrove in the
action as it is in the case where trapping does no t delta m ay be explained no t only by the fact th at
exist or exist for few days. D ue to little mixing the m angrove sw am ps receive significant am ount
w hich exists betw een the estuarine and the near o f nutrients from freshw ater discharge, but also
shore w aters, the dissolved m atter is no t flushed from the fact th a t m ost o f nutrients supplied by
from the system by tidal action, and rem ains the river and from other sources are retained in
within the sw am ps for a long time. This provides the sw am ps or in the nearshore zone for a p ro
m ore opportunity for absorption o f the nutrients longed period due trapping effect.
by the m angrove as practically the sam e w ater D ue to limited d ata available some o f the issues
w ould be passing through them over and over raised in the discussion such as the ecological
again for prolonged period. im plications o f the trapping effect on the m an
The m angroves develop best in areas which grove and the possibility o f the seasonal varia
experiences periodic deposition o f sedim ents tions in the trapping effect etc. are no t conclusive.
(C intron & Schaeffer-Novelli, 1983). In the Rufiji
delta, the m axim um sedim ent tran sp o rt occurs
during the period o f m axim um outflow (Eurocon- Acknowledgements
sult, 1980). Som e o f those sedim ents are d epos
ited in the m angrove sw am ps and the rest are It is a pleasure to th an k F R O N T IE R Expeditions
tran sp o rted to the nearshore zone where as dis for all the support they gave me during d a ta col
cussed, they are trapped. H ow ever, during the lection. T hanks also should go to M essrs. J. P.
SW m onsoon, there is also considerable deposi Shunula, A. N . N . M uzuka, Y. W. Shaghude,
tion o f sand in the m angrove sw am ps o f Kiom- and A. J. M m ochi all o f w hom helped m e in v ar
boni and Sim ba U ranga islands (pers. obser.). ious ways, especially for their valuable construc
The origin o f these sands is thought to be the tive com m ents and suggestions to the first draft
nearshore zone as they are well rounded and are of the m anuscript.
179
References MacNae, W., 1968. A general account of the fauna and flora
of mangrove swamps and forest in the Indo-W est Pacific
region. Adv. Mar. Biol. 6: 73-270.
Bryceson, I., 1977. An ecological study of the phytoplankton
Newell, D. S., 1959. The hydrography of the British East
of the coastal waters of Dar-es-Salaam. Ph.D. Thesis, U ni
African Coastal Waters. Fishery Pubis. Colon. Off. No. 12.
versity of Dar-es-Salaam.
18 pp.
Cintron, G. & Y. Schaeffer-Novelli, 1983. Mangrove Forests:
VHL., 1979. Hydrological studies in the Lower Rufiji River.
Ecology and response to natural and man-induced stres
Trondheim, Norway. 100 pp.
sors. In J. C. Ogden & E. H. Gladfelter (eds), Coral reefs,
Wolanski, E., 1989. Measurements and modelling of the water
seagrass beds, mangrove: Their interaction in the coastal
circulation in the mangroves swamps. Results of practical
zones of the Caribbean. U N ESC O report in marine science
work at the COM ARAF Regional Field Training W ork
23: 87-113.
shop on Mangrove Ecosystems and Estuarine Processes.
Euroconsult., 1980. Identification study on the ecological im
Chwaka Bay, Zanzibar. June 6-16, 1989. Regional Project
pacts of the Stiegler’s Gorge Power and Flood Control
RAF/87/038.
Project Part 2: Methodology, Definition of impact area and
Wolanski, E. & P. Ridd, 1986. Tidal mixing and trapping in
Resources inventory.
mangrove swamps. Estuar. coast, mar. Sei. 23: 759-771.
Iversen, S. A., S. Myklevoll, K. Lwiza & Y. Yonazi., 1984.
Wolanski, E. & P. Ridd, 1989. Coastal mixing and trapping
Tanzanian marine fish resources in the depth region 10-
in tropical Australia. In R. T. Cheng (ed.), Dynamics of
500 m investigated by R/V ‘Dr. Fridtjof N ansen’. In S. A.
Shallow Bays and Estuaries, (in press).
Iversen & S. Myklevoll (eds), The Proc. of the NORAD-
Woodroffe, C. D., 1985. Studies of a mangrove basin, Tuff
Tanzania Seminar to review the marine fish stocks and
Crater, New Zealand: 1. Mangrove biomass and produc
fisheries in Tanzania. Mbegani, Tanzania. March 6-8,
tion of detritus. Estuar. coast, mar. Sei. 20: 265-280.
1984: 45-83.
. .
Hydrobiologia 247: 1 8 1 -1 8 6 , 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 181
© 1992 Kluwer Academie Publishers. Printed in Belgium.
K ey words: D iurnal variation, m angrove w ater, dissolved inorganic nutrients, prim ary production
Abstract
D iurnal variation in hydrological variables and dissolved inorganic nutrients such as P O | -P, N
0 2~ -N, N 0 3 -N and N H 4+-N were studied in three interconnected biotopes including freshw ater, m a
rine and m angrove brackish w ater o f the K ak in ad a coastal zone, A ndhra Pradesh. Sam ples were col
lected at intervals o f 3 hours, for a period o f 24 hours. In the m arine environm ent salinity varied from
26%0 to 32%0 w hereas in the m angrove w aters it fluctuated from 12%„ to 20%o and in both biotopes salinity
show ed bim odal type o f oscillation. D issolved oxygen content w as high in the m angrove w aters during
day time but decreased rapidly during the night hours. In the m arine environm ent P O |_ -P concentra
tion varied from 0.345 to 1.195 pg at I - \ N 0 3~ -N from 1.03 to 6.62 p,g at 1“ 1 and N 0 2“ -N from 0.086
to 0.506 pg at 1” The highest and the low est concentrations o f P O |_ -P, N O f - N , N 0 2_ -N recorded
in the m angrove w aters were 0.790 and 0.325 pg at 1~ 1, 7.10 and 1.60 pg at 1“ 1 and 0.278 and 0.060 jug
at I “ 1, respectively. The concentration o f PC>4 _ -P, N 0 3_ -N and N O y -N were high in the freshw ater
canal, the m axim um and m inim um values being 1.110 and 0.730 pg at 1” l, 26.40 and 9.98 pg at 1“ 1 and
0.520 and 0.252 ptg at 1” 1 respectively. The concentration o f am m onia was relatively high in the m an
grove water. G ro ss and net prim ary production in the m angrove w ater w as 4 times higher than in the
m arine biotope. There w as no ‘export’ o f dissolved nutrients from the m angrove environm ent to the
adjacent m arine waters.
KAKINADA
A m angrove swam p covering an area o f approx
BAY
imately 96 hectare is present near the K akinada
tow n (16° 45 ' N and 82° 1 5 'E) in A ndhra
Pradesh. The vegetation o f the swam p consists of
Avicennia officinalis L., A . marina (F orsk.) Verh,
Excoecaria agallocha L., Acanthus ilicifolius L.
with A . officinalis as the dom inant species. The
swam p, on the southern side is bordered by the
C om m ercial C anal which receives surplus fresh
w ater from irrigated land and is characterised by
the presence o f the inflow o f freshw ater through
out the year. O n the w estern side o f the K akinada
8 2 ° 1 5 'E
Bay (average salinity 32%0) lies a m angrove
swam p. T he swam p, Com m ercial C anal and the Fig. 1. M ap of the K akinada Bay showing the mangrove
K ak in ad a Bay are interconnected as shown in swamp and sampling stations.
Fig. 1. W hen the freshw ater is minim al in the
C om m ercial C anal seaw ater from the K akinada
Bay enters it at high tide and mixes with the fresh
w ater. T he resulting brackish w ater enters through given by Strickland & P arsons (1968) w as fol
a channel branching off the left side o f the C om lowed for the estim ation o f dissolved oxygen. F or
m ercial C anal. D uring the m onsoon seasons, the analysis o f dissolved nutrients, w ater sam ples
heavy influx o f freshw ater in the Com m ercial collected from surface were brought to the labo
C anal prevents the entry o f seaw ater from the ratory within 30 min o f collection and filtered
K ak in ad a Bay and during such a period fresh through a W hatm an G F /C filter p aper o f 0.05 ¡j.m
w ater from the Com m ercial C anal directly reaches porosity. T he filtered sam ples were deep frozen
the m angrove environm ent. In each biotope one and the nutrients were analysed the next day.
sam pling site w as fixed and designated as Statio- A nalysis o f P O ^ - P , N O 3- N , N O f - N and N
n I (K ak in ad a Bay), Station II (M angrove H 4 -N w as carried out according to the m ethods
Sw am p) and Station III (Com m ercial Canal). given by P arso n s et al. (1984). T he light and dark
Single surface sam ples were collected from these bottle m ethod as described by Strickland & P a r
stations at 3 hour intervals from 0600 h, 20th A u sons (1968) w as followed for assessing the pri
gust, 1988 to 0600 h, 21st August, 1988. Salinity m ary production. W ater sam ples in duplicate
was m easured by an optical salinom eter (Am er were incubated in situ for 6 hrs from 0900 hrs to
ican O pticals, H ong Kong, Cat. N o. 10423, 1500 hrs. D ue to logistical difficulty prim ary p ro
± 1%0). F o r dissolved oxygen analysis, sam ples duction w as estim ated only at Stations I and II.
were fixed in the field using m anganous sulphate A tm ospheric tem perature w as noted at the time
and alkaline iodide solution. W inkler’s m ethod o f sampling.
183
Results Salinity
o— Mari ne (St. I)
»— S w a m p (St- II)
û— Ca n a i (Still)
40
— TO
>. 20
c
o 10
to
0
12 15 18 21 24 15 18 21 24
Time (hours) Ti me ( h o u r s )
Fig. 2. Diurnal variations in (a) tidal range (m), (b) water tem perature (°C ), (c) salinity (%0), and (d) dissolved oxygen (ml 1 ')■
184
m angrove w ater (St. II) might be due to a high ation in P O |~ -P concentration at all stations over
rate o f oxygen production by phytoplankton. the sam pling period.
Nutrients
Nitrite-nitrogen
Dissolved inorganic phosphate
At Stations I and II, N 0 2“ -N concentration v ar
D issolved inorganic phosphate at the three sta ied within the range o f 0.086 to 0.506 pg at 1“ 1
tions varied from 0.345 to 1.195 pg at 1“ 1 (St. I), and 0.060 to 0.278 pg at 1“ 1 respectively which is
0.325 to 0.790 pg at I “ 1 (St. II) and 0.730 to com paratively m uch low er than the levels re
1.110 jug at I “ 1 (St. III) (Fig. 3a). At Stations I corded at St. III. A t Sts. I and II, N 0 2“ -N was
and II P O |_ -P concentration, during the day lower during the day tim e bu t rose at night. At
time, w as low and showed no relationship with St. I, N 0 2 -N concentration during the night
tides w hereas during the night hours PO¿¡ -P hours increased the flood tide w hereas at St. II
concentration increased at the time o f high tide nitrite showed significant variation between day
and decreased during the low tide regime. F rom and night time but this did no t correlate with the
Fig. 3a it is evident th at there w as significant vari tide. A t St. III, N 0 2“ -N fluctuated from 0.252 to
• — Marine (St- I)
* — S w am p(S t. II )
a — Canal (St. III)
20
300
Ö
o>
=3. 200
Z
I
o
CO o
z 100
S tI St.II
24
04
z
0-2
CM
O
z
1-3
z
IV'
I
z
6 12 18 24 3 6 6 12 18 24 3 6
T im e (h o u rs) Time (h o u rs)
Fig. 3. Diurnal variations in PC>4 - -P, N O f-N , N 0 3~-N, N H 4+-N and gross and net primary production.
185
0.520 jug at 1 1 and showed only a m inor varia show ed no definite pattern in diel changes. At
tions from 1500 hrs to 2100 hrs. St. I and II, unlike P O |- - P , N O f - N and N
0 3“ -N , N H / -N did no t show any relationship
with tides even during night hours which may
Nitrate-nitrogen indicate th at the regeneration and utilization o f N
H 4+ -N is independent in each biotope.
In the K akinada Bay (St. I) and m angrove w aters
(St. II) N 0 3“ -N show ed a m ore or less similar
range o f variation, the value being 1.03 to 6.62 ¡ig Primary production
at 1” 1 and 1.60 to 7.10 jug at 1“ 1 respectively. At
these two stations N 0 3~ -N concentration de G ross production in the m arine (St. I) and m an
creased gradually from 0600 hrs and reached the grove w ater (St. II) w as 82.4 m g C m ~ 3 h ” 1 and
lowest value at 1500 hrs and thereafter an 262.3 mg C m -3 h - 1 respectively. N et produc
increasing trend w as noticed till 0300 hrs. Like P tion at Sts. I and II w as 3 1 .2 m g C m - 3 h -1 and
0 | “ -P and N 0 2" -N , N 0 3_ -N concentration at 136.5 mg C m _ 3 h~ 1 respectively. High prim ary
these two stations, w as independent o f the oscil production at St. II m ay attributed to the opti
lations in tides during day time. However, during m um salinity condition ( 12%0 to 20 %o) prevailing
night hours N 0 3- -N concentration increased dur in the m angrove waters.
ing the high tide period and decreased during the
low tide regime. A t St. III, N 0 3- -N concentration
during day time fluctuated erratically with the fol Discussion
lowing values: 0600 hrs: 14.86 / r g a t l - ^0 9 0 0 hrs:
26.40 ^rg at I - 1 ; 1200 hrs: 13.88 ¿rg at I ” 1; M angrove ecosystem s are tidally affected to such
1500 hrs: 25.08 ¿ig at 1—1; 1800 hrs: 9.98 /rg at I - 1 an extent th at tides are the m ain ecological fac
(Fig. 3c). At St. III there was intensive hum an tor serving as m ajor energy input (O dum , 1971).
activity, with the m ovem ent o f m echanised fish It is generally believed th at coastal w etlands in
ing boats. These bo ats leave around 0500 hrs to cluding m arshes and m angroves export large
0900 hrs and return around 1500 hrs. It might be quantities o f dissolved inorganic nutrients to the
thought th at boat activity after 0500 hrs and adjacent estuarine or coastal w aters through tidal
1500 hrs explains the peaks in N 0 3_ -N in the m ovem ent (W oodwell & W hitney, 1977; Bacon,
freshw ater canal. H owever, this is an unlikely ex 1981). In contrast Boto & W ellington (1988)
planation because one would expect similar peaks show ed th at only dissolved organic m aterials
in the concentrations o f the other nutrients which are exported from the m angroves. They esti
were not observed. m ated th at nearly 24% o f the phosphate re
quirem ent of the m angrove environm ent is sup
plied from coastal w aters. G uerrero et al. (1988)
A mmonia-nitrogen also confirm ed such a phenom enon and reported
th at no t only dissolved inorganic phosphate but
A m m onia concentration w as low at St. I, the also dissolved nitrate w as im ported to the m an
highest and lowest values being 1.120 and grove environm ent from coastal waters. The
7.128 yug at I -1 respectively. At St. II, N H 4+-N present study show s th at concentration o f P
concentration w as high, the m axim um and m in 0 |~ - P , N 0 2” -N , and N 0 3_ -N were m ore or less
imum values being 23.65 and 8.85 /tg at 1“ \ A t similar in the m angrove and bay w ater at first low
St. I, N H 4+ -N concentration decreased gradually tide.
from 0600 hrs to 1800 hrs w hereas at St. II, N A t second low tide concentrations o f all these
H ^ - N concentration showed a steady decrease nutrients were slightly higher in the bay w ater
from 0900 hrs to 2100 hrs. At St. III, N H 4+-N than th at o f the m angrove. This indicates that
186
R. M o h a n ra ju 1 & R. N a ta ra ja n 2
1Centre o f Advanced Study in Marine Biology, Annamalai University, Parangipettai-608 502, India;
2Institute fo r Ocean Management, Centre fo r Water Resources, Anna University, Madras-600 025, India
Abstract
The occurrence o f m ethanogenic bacteria in the K odiakkarai (10° 18' N; 79° 52' E) m angrove sedi
m ents, where Avicennia spp are predom inant, w as studied. Trim ethylam ine under N 2:C 0 2 (80:20% v/v)
w as used as the substrate. M ost Probable N um ber (M P N ) o f m ethanogenic bacteria w as determ ined
for a period o f one year from July 1987 to June 1988 with m onthly sampling. The m ethanogenic b ac
terial populations were found to be at the m axim um o f 1.1 x IO5 M P N g m “ 1 of wet sedim ent during
August 1987 and from February to June 1988. The bacterial num bers were found to decrease during
O ctober to D ecem ber 1987 with a m inim al value o f 3.6 x IO2 M P N g m “ 1 during D ecem ber 1987. E n
vironm ental factors were correlated with the m ethanogenic bacterial population.
IN DIA
ÍKodiakarai
S am p lin g site Swam p
(Light h o u s e
Mangrove
79 52 E
salinity; while the nutrients nitrite, nitrate, total could be the regeneration o f phosphorus in the
p hosphorus and organic carbon showed negative b ottom m uds and the subsequent release o f the
correlation (Table 2 ). sam e into the w ater column due to turbulence and
mixing caused by tides as well as the heavy winds
prevailing during this season.
Discussion N o definite trend w as observed in the sedim ent
organic carbon. H igher concentrations were ob
Generally, m axim um tem perature w as recorded served during m onsoon and postm onsoon sea
during sum m er and m inim um during the m on sons probably due to the high rate o f freshw ater
soon as observed earlier by A nbhazhagan (1988), inflow enriched with organic debris from land
K arthikeyan (1988) and S am path (1989) for the run-off.
K odiakkarai region. N o distinct seasonal pattern T he m ethanogenic bacteria are a m ajor com
w as observed for pH . The values ranged from 7.0 ponent o f the m icrobial population in the m an
to 7.5 which are com parable to previous reports grove sedim ent (R am am urthy et al., 1990). They
by A nbhazhagan (1988) & K arthikeyan (1988). are restricted with respect to the n ature o f the
W illiams (1981) attributed the variation in pH to substrate utilized for m ethane form ation. F ran k
photosynthesis by subm erged vegetation in the lin eta l. (1988) observed a tw o to three times
area. higher num ber of m ethanogens in a short Spartina
D uring sum m er and early prem onsoon m axi zone than in a tali Spartina zone, and they attrib
m um salinity values were observed. The reasons uted this to the abundance o f m ore com pact Spar
for high salinity values are due to the addition of tina roots in the short Spartina zone. B artlett et al.
harvested salt heaped from salt p ans close by and (1985) on m easuring m ethane flux from a coastal
also due to the brine from the salt pan being salt m arsh found th at tem perature increase was
drained in the vicinity of the study area. The fresh accom panied by an increase in the m ethane flux.
w ater discharge w as also low during this period. Jones & P aynter (1980) observed a decrease in
W ith the advent o f prem onsoon showers the sa the m ethanogenic population which utilized hy
linity declined slightly because o f freshw ater run- drogen as a substrate on the m ud flats away from
olf from sources like irrigation channels. The the Spartina plants, showing a population level of
n ortheast m onsoon starting in O ctober and con 3.7 x IO4 to 1.2 x IO6 cells gm~ 1 o f dry sediment.
tinuing up to D ecem ber decreased the salinity King (1988) found th at m ethane production in
values to the minimum. The sam e trend has been sedim ents o f a hypersaline pond w as associated
previously observed by A nbhazhagan (1988), with algal m ats which were apparently sources of
K arthikeyan (1988) & S am path (1989) in the K o m ethylated amines.
diakkarai m angroves and by V asantha (1989) in O bservations by Phelps & Zeikus (1985) in the
the P ichavaram m angroves. sedim ents o f Lake M endota show ed a bacterial
T he m axim um nitrite and nitrate concentra population range o f IO3—IO5 cells g m -1 o f dry
tions during the m onsoon were apparently due to sedim ent. Higher population levels were observed
the influence o f seasonal freshw ater discharges during sum m er and lower ones during winter. The
(Carlucci e ta l., 1970). Higher concentrations of change in population level w as due to the fall in
nitrite th an nitrate could be attributed to bacterial tem perature and acetate availability. The present
reduction o f nitrate and oxidation o f am m onia by study o f m ethanogenic bacterial population from
nitrifying bacteria. D uring sum m er, the total the m angrove sedim ents show ed higher bacterial
phosphorus concentration increased gradually num bers during the sum m er and prem onsoon and
and reached a m axim um during the m onsoon lower values during the m onsoonal m onths. The
owing to enrichm ent by m onsoonal floods as difference in population levels is m ainly due to the
pointed out by C handran & R am am oorthi (1984) decline of salinity due to freshw ater influx, tem
for the P ichavaram m angroves. A nother reason perature changes and changes in pH and Eh.
192
Trim ethylam ine w as an im portant precursor Chapman, V. J., 1976. Mangrove vegetation. Cramer Vaduz.
Chandran, R. & K. Ramamoorthi, 1984. Hydrobiological
(3 1 .5 % -6 1 .1 % ) in the estuarine intertidal sedi
studies in the gradient zone of the Vellar estuary. I. Physico
m ents o f Lowes Cove, M aine (King et al., 1983). chemical parameters. Mahasagar-Bull. N at. Inst. Ocean-
The im portance o f this com pound as a m ethane ogr. 17: 69-77.
precursor w as due to the abundance o f decom Daniels, 1., R. Sparling & G. D. Sprott, 1984. The bioener-
posed plant m aterial in the sedim ent system. getic of methanogenesis. Biochim. Biophys. Acta 768: 113—
163.
T M A has been confirm ed as a non-com petitive
el Wakeel, S. K. & J. P. Riley, 1956. The determination of
substrate for m ethane production (King, 1984) by
organic carhon in marine muds. J. du. Coun. 22: 180-183.
sulphate reducers. In sedim ents choline and be- Franklin, M. J., W. J. Wiebe & W. B. Whiman, 1988. Popu
taine undergo initial attack by ferm entative b ac lations of methanogenic bacteria in a Georgia salt marsh.
teria rather th an sulphate reducers. This results in Appl. Envir. Microbiol. 54: 1151-1157.
the production o f trim ethylam ine and acetate Heukelikian, H. & B. Heinemann, 1939. Studies on the meth
ane bacteria. I. Development of a method for enumeration.
which are then respectively channelled to m eth- Sew. Wks. J. 11: 426-435.
anogenesis and sulphate reducers. Hungate, R. E., 1950. The anaerobic mesophilic celluloytic
In conclusion the presence o f m ethanogenic bacteria. Bact. Rev. 14: 1-49.
b acteria in m angrove sedim ent appears to be in Jones, W. J. & M. J. B. Paynter, 1980. Populations o f meth
fluenced by environm ental param eters such as ane producing bacteria and in vitro methanogenesis in salt
m arsh and estuarine sediments. Appl. Envir. Microbiol. 39:
tem perature, p H , Eh, salinity and other nutrients. 864-871.
Karthikeyan, E., 1988. Organochlorine pesticides in the trop
ical estuary (The Vellar) mangrove (Pichavaram) and
Acknowledgement swamp (Kodiakkarai) in the east coast of India. Ph.D.
Thesis, Annamalai Univ., 151 pp.
King, G. M., 1984. On the metabolism o f trimethylamine,
W e wish to th an k the D epartm ent o f N on- choline and glycinebetaine by sulfate reducing and metha
conventional Energy Sources, G ovt, o f India, new nogenic bacteria in marine sediments. Appl. Envir. M icro
D elhi for funding this study. biol. 48: 719-725.
King, G. M., 1988. Methanogenesis from methylated amines
in a hypersaline algal mat. Appl. Envir. Microbiol. 54: 130—
136.
References King, G. M., M. J. Klug & D. R. Lovely, 1983. Metabolism
of acetate, methanol and methylated amines in intertidal
sediments o f Lowes Cove, Maine. Appl. Envir. Microbiol.
Anbhazhagan, P., 1988. Hydrobiology and benthic ecology of
45: 1848-1853.
Kodiakkarai coastal sanctuary (Southeast coast of India).
Macy, J. M., J. E. Schnellen & R. E. Hungate, 1972. Use of
Ph.D. Thesis, Annamalai Univ., India, pp. 208.
syringe method for anaerobiosis. J. Clin. Nutr. 25: 1318—
Balch, W. E., C. E. Fox, L. J. Magrum, C. R. Woese & R. S.
1323.
Wolfe, 1979. Methanogens: Re-evaluation of a unique bi
M ah, R. A., D. M. Ward, L. Baresi & T. L. Glass, 1977.
ological group. Microbiol. Rev. 43: 260-296.
Biogenesis of methane. Ann. Rev. Microbiol. 31: 309-341.
Bartlett, K. B., R. C. Hariss & D. I. Sebacher, 1985. M eth
Mathrani, I. M. & D. R. Boone, 1985. Isolation and charac
ane flux from coastal salt marshes. J. Geo-phy. Res. 90:
5710-5720. terization of a moderately halophilic methanogen from a
solar saltern. Appl. Envir. Microbiol. 50: 140-143.
Belyaev, S. S. & Z. I. Finkelshtein, 1973. M ethod o f count
McCrady, A. H., 1918. Tables for rapid interpretation of fer
ing methane producing bacteria in media containing organic
mentative tube results. Can. J. Publ. Health. 9: 1153-1158.
substrates. Mikrobiologiya 42: 1102-1107.
Miller, T. L. & M. J. Wolin, 1974. A serum bottle modifica
Bock, A. & O. Kandier, 1985. Antiobiotic sensitivity of ar-
chaebacteria. In: C. R. Woese & R. S. Wolfe (eds), The tion of the Hungate technique for cultivating obligate anaer
obes. Appl. Microbiol. 27: 985-987.
Bacteria, Vol. 8. Academic Press, New York: 525-544.
Murphy, J. & J. P. Riley, 1962. A modified single solution
Bryant, M. P., 1972. Commentary on the Hungate technique
for culture of anaerobic bacteria. Am. J. Clin. Nutr. 25: method for the determination o f phosphate in natural wa
1324-1328. ters. Analyt. Chem. Acta. 12: 162-176.
Carlucci, A. F., E. O. Hertwig & P. M. Bowes, 1970. Biolog Pedersen, D. & G. S. Sayler, 1981. Methanogenesis in fresh
ical production of nitrite in seawater. Mar. Biol. 7: 161— water sediments, inherent variability and effects of environ
166. mental contaminants. Can. J. Microbiol. 27: 198-205.
193
Phelps, J. & J. G. Zeikus, 1985. Effect of fall turnover an Vasantha, K., 1989. Studies on hydrobiology and decompo
terminal carbon metabolism in Lake M endota sediments. sition of macrophytes in Portonovo marine environments
Appl. Envir. Microbiol. 50: 1285-1291. (Southeast coast of India). Ph.D. Thesis. Annamalai Univ.,
Ramamurthy, T., R. Mohanraju & R. N atarajan, 1990. Dis 132 pp.
tribution and ecology of methanogenic bacteria in man Walsh, G. E., 1974. Mangroves: a review In: R. Reinhold &
grove sediments of Pichavaram, east coast of India. Ind. J. W. Queen (eds), Ecology of Halophytes. Academic Press,
Mar. Sei. 19: 269-273. New York, 174 pp.
Rochford, D. J., 1951. Studies in Australian estuarine hydrog William, W. D., 1981. The limnology of saline lakes in W est
raphy I. Introductory and comparative features. Aust. J. ern Victoria. A Review of some recent studies. Hydrobio-
mar. Freshwat. Res. 2: 1-116. logia 82: 233-259.
Sampath, K., 1989. Studies on the ecology of shore birds Wolin, E. A., M. J. Wolin & R. S. Wolfe, 1963. Formation of
(Aves: Charadriformes). Ph.D. Thesis, Annamalai Univ., methane by bacterial extracts. J. Biol. Chem. 238: 2882—
202 pp. 2886.
Siebert, M. L. & W. H. J. Hattingh, 1967. Estimation of Yancey, P. H., M. E. Clark, S. C. H and, B. D. Bowlus &
methane producing bacterial numbers by the most proba G. N. Somero, 1982. Living with water stress: evolution of
ble number (M PN ) technique. Wat. Res. 1: 13-19. osmolyte system. Sei. 217: 1214-1222.
Strickland, J. D. H. & T. R. Parsons, 1972. A practical hand Zeikus, J. G. & M. Winfrey, 1976. Temperature limitation of
book of sea water analysis. Bull. Fish. Res. Bd Can. 167: methanogenesis in aquatic sediments. Appl. Envir. Micro
310. biol. 31: 99-107.
Hydrobiotogia 247 : 195- 207 , 1992.
V. Jaccarini & E. Martens (eds), The Ecology of Mangrove and Related Ecosystems. 195
© 1992 Kluwer Academic Publishers. Printed in Belgium.
Abstract
M angrove inlets and creeks in Selangor, M alaysia are the habitat for 119 species o f fish and 9 species
o f praw ns. The m ajority o f fish and all praw ns sam pled in the inlets were juveniles. The com m on fish
species in the inlets in term s of weight were Arius sagor, Ambassis gymnocephalus, Liza subviridis, Toxotes
jaculator, Sphyraena barracuda and Lates calcarifer. P raw ns were represented by juvenile Penaeus
penicillatus, P. merguiensis, P. indicus, Metapenaeus brevicornis and M . affinis. Sam ples from enclosure
trap s set on m udflats during ebbing w ater captured 37 species o f fish and 11 species o f praw ns. The role
o f m angroves as nursery and feeding grounds for fish and praw ns is reviewed in the light of recent work
in Selangor. It is apparent th at m angroves support fisheries by providing habitat and food.
'fR7n^
■ KUALA SELANGOR STUDY
AREA
\* fe---
V. Í * KAMPONG SUNGAI JANGGUT
V- .
KLANS
RIVER
KETAM IS'.
KLANG IS:
■ENGAH,
IS.'i^
CAREY ISLAND
LE G E N D
S a n d b anks exposed at
e x t r e m e lo w t i d e
EM Mangroves
• D e p t h in f a t h o m s
Bund
Sampling sites
Fig. 1. The study area indicating sampling sites in Angsa Bank, Klang Strait and mangrove inlets on the Selangor coast. The
enclosure trap sampling sites are situated north and south of the Buloh River.
197
A total o f 119 species o f fish and 9 species o f Angsa bank and Klang straits
praw ns were found in the creeks and inlets o f
these m angrove sw am ps (Table 1). T he creeks A total o f 102 species o f fish and 15 species of
were dom inated by schooling fish species belong praw ns w ere found in the area (Table 1). The fish
ing to the families A m bassidae, Engraulidae, C lu com m unity w as dom inated by m em bers o f the
peidae, Leiognathidae, Eleotridae and M ugilidae families Sciaenidae, Cynoglossidae, Engraulidae,
(Table 2). The praw ns were mainly penaeids, C lupeidae and Leiognathidae. The praw n com-
198
Table I . Fish and prawn species found in the inshore waters of the Selangor coast, Malaysia, c = creeks/inlets; m = mudflats; n = near inshore waters;
f = far inshore waters.
Fish species
Table 1. (Continued)
Prawn species
Fish 119 37 58 92
Prawns 9 11 8 15
Table 2. Percentage abundance by number of major fish species caught by the bag net in two mangrove creeks in S. Sementa Kecil
and S. Perepat Kecil from November 1987 to May 1988, ( + ) means present in small numbers.
Species Site
m unity w as dom inated by tw o penaeid species, were im m ature. Their carapace lengths ranged
Parapenaeopsis sculptilis and P. hungerfordi. from 7 to 28 m m CL. The percentage o f im m ature
praw ns in the shallow coastal w aters o f the Klang
S exual maturity Straits and A ngsa B ank were 8 and 38% respec
tively.
Ninety-seven percent o f the fish sam pled in the
inlets (including creeks) were im m ature. This was
in co n trast to 80% over the m udflats and 60% in D is c u s s io n
the nearshore w aters. O f the total fish population
in the far inshore area, only 29% were im m ature T he com m unity o f fish found in m angrove inlets
(Table 7). All the praw ns sam pled in the inlets in Selangor resem bles th a t o f similar habitats
Table 3. Percentage abundance by numbers o f the major fish species caught by gili nets in the mangrove inlet S. Sementa Kecil
from November 1987 to May 1988, ( + ) means present in small numbers.
Table 4. Percentage abundance by numbers of the major and prawn species caught by beam trawl in the mangrove inlet S . Sementa
Kecil from Januray 1988 to August 1989. ( + ) present in small numbers.
Fish
Ambassis gymnocephalus 18.46 4.02 89.90 56.18 8.33 3.31 34.94
Thryssa hamiltoni 22.81 1.21 16.38 12.14 1.24 14.53
Ilisha megaloptera 0.43 0.22 5.17 39.52 3.72 2.48
Thryssa kammalensis 43.32 1.31 3.66 + 23.01
Leiognathus brevirostris 1.25 3.66 6.19 26.86 1.90
Glossogobius giuris 10.07 0.76 0.10 0.93 5.24 13.22 1.28
Pomadasys hasta 17.79 0.31 0.06 0.93 3.57 5.37 1.01
Scatophagus argus 9.73 0.52 2.55 4.24 4.76 2.07 1.98
Chelonodon fluviatilis 1.68 3.68 0.38 2.95 2.62 10.74 2.73
Leiognathus daura 18.31 2.45 0.14 10.08
Prawn
Penaeus penicillatus 21.21 23.89 20.74 36.15 43.14 22.81 28.65
Metapenaeus brevicornis 20.61 46.02 61.17 17.79 1.96 15.79 30.01
Penaeus merguiensis 9.70 17.70 6.91 31.40 31.37 43.86 21.93
Penaeus indicus 48.48 2.65 2.66 8.97 11.76 1.75 13.53
Metapenaeus mutatus 7.08 5.32 1.85 11.76 10.53 3.88
Macrobrachium equidens 2.65 2.66 1.58 3.51 1.68
Penaeus monodon 0.26 1.75 0.21
Parapenaeopsis sculptilis 0.53 0.10
elsewhere (Little e ta l., 1988; Pinto, 1988; R ob eta l., 1984). P redation o f benthic m angrove in
ertson & D uke, 1987). Six species of small fish of vertebrates including sesarm id crabs (Chiroman
low and m edium econom ic value com prised 70% thes spp.), M etaplax spp., bivalves, gastropods
o f a total o f 21670 specim ens (119 species) ex and sipunculids by catfish eel and catfish is com
am ined. The m ost abundant, Ambassis gymno m on (S asekum ar et al., 1984). In addition, m an
cephalus, is n o t eaten by m an because o f its small grove invertebrates liberate larvae into coastal
size. H ow ever, others like the Sardinella sp., w aters th at becom e food for planktivores. M a
Thryssa spp. and Stolephorus sp. are consum ed. cintosh (1979) estim ated th at two species o f fid
Although the catfish-eel, Plotosus canius, was dler crabs ( Uca spp.) in K uala Selangor produced
rarely captured in the gili nets or the beam trawl, 200000 m egalopae m - 2 y r - 1 . C rab m egalopae
it constitutes an im portant fishery and is norm ally are im portant food for several species o f ancho
landed by eel trap or hook and line (Leh, u npub vies (Engraulidae) and Ambassis spp. (pers. obs.).
lished). The praw ns in the m angrove inlets are repre
Fish occurring in creeks feed on benthic food sented largely by juvenile penaeids as has been
resources including crabs, penaeids, insects, observed in Q ueensland (R obertson & Duke,
m angrove detritus, diatom s, m acroalgae, and on 1987). Several species o f praw ns which include
pelagic fauna like Acetes and copepods (Leh & the predom inant species o f Penaeus merguiensis,
S asekum ar, 1991). The mullets, Liza malinoptera P. penicillatus and P. indicus have been shown to
and L. subviridis consum e significant quantities of use these inlets as nursery areas (C hong eta l.,
m angrove detritus (Chong, 1977; Sasekum ar 1990). Metapenaeus brevicornis, M . affinis and
202
Table 5. Percentage abundance by numbers of the major fish and prawns species caught by the enclosure net in intertidal mudflats,
north and south of the river the River Buloh, Selangor, Peninsular Malaysia, from January 1988 to August 1989.
Fish
Macrones gulio 16.82 46.40 63.22 26.50
Cynoglossus macrolepidotus 10.28 14.70 15.24 40.22 13.65
Arius tonggol 20.56 8.21 28.81 9.75
Liza subviridis 14.02 2.45 7.22 23.69 6.99
Pomadasys hasta 3.27 6.34 11.23 20.84 6.72
Arius caelatus 20.56 20.56 5.92
Ambassis gymnocephalus 1.73 15.24 16.97 4.64
Johnius coiter 1.64 11.38 3.74 16.76 6.72
Arius sagor 14.71 14.71 3.70
Thryssa kammalensis 0.29 9.89 10.18 2.62
Prawn
Palaemon styliferus 45.18 75.82 121 70.34
Metapenaeus brevicornis 41.03 3.07 3.62 47.72 4.36
Metapenaeus mutatus 7.69 16.67 12.86 37.22 13.23
Penaeus indicus 25.64 7.02 32.66 1.43
Parapenaeopsis coromandelica 12.82 17.54 0.65 31.01 3.03
Penaeus merguiensis 7.69 4.39 3.68 15.76 3.86
Macrobrachium equidens 5.13 2.19 2.91 10.23 2.87
Penaeus penicillatus 3.51 3.51 0.44
Parapenaeopsis gracillima 0.44 0.44 0.06
Metapenaeus lysianassa 0.26 0.26 0.22
Parapenaeopsis hardwickii 0.19 0.19 0.17
Penaeus merguiensis occurred together both in the able at low tide as the w aterw ays do no t have sea
inlets and the adjacent m udflats. However, there grass cover or other refugia. Therefore, high food
is evidence o f spatial distribution o f species within abundance is m ore likely the m ain benefit to ju
the inlet. The three Penaeus species are restricted venile praw ns. Turbidity (including food avail
mainly to the upper and m iddle reaches o f the ability and tem perature) is know n to influence the
inlet, w hereas the tw o Metapenaeus species are m ovem ent and distribution o f juvenile fish in es
found in the lower reaches and m udflats (Figs 2a tuarine w aters (B laber & Blaber, 1980; C yrus &
& 2b). Spatial partitioning o f juvenile praw ns in Blaber, 1987). Similar studies have n o t been done
m angrove estuaries has also been observed in with praw ns though high turbidity could be im
E ast Africa (de Freitas, 1986). p o rtan t in reducing predation.
The presence o f juvenile praw ns in m angrove Penaeid praw ns are opportunistic generalists
inlets m ay be attributed to three m ain factors: in their food habits, being able to consum e a v a
shelter from predators, high food abundance and riety o f anim al and plant m atter and bacteria
turbidity. Although the catfish eel and ariid ca t (C hong & Sasekum ar, 1981; Leh & Sasekum ar,
fish are com m on in the inlets, praw ns could take 1984; M oriarty, 1976).
refuge from these predators am ongst the prop The abundance o f food m aterial on the surface
roots at high tide. How ever, praw ns are vulner o f m ud or sand-m ud substrates associated with
203
oo oo
■tí" \¿ prawn 3886 37.8 31.4 30.8
o
Qh xi <N oX
a
o
'S
3 OO OO Os OO
the m angrove environm ent provides a conducive
U environm ent for the growth o f praw ns. O n aver
age there are IO6 m eiofauna m -2 (1 -2 g m -2 dry
Table 6. Biomass and numbers of fish and prauwns captured in the enclosure trap subsamples (10%
40
BSSSSSSSS8
35 LOWER
30
S 25
i 20
15
IO
i
9
1
IO
ISSÛ
11
12
13
14
15 17
18
19
20
21
22 23 25 27
C A R A P A C E LENGTH (mm)
Penaeus penicillatus
60
LOWER
50 - C" ' I
M/FLATS
40 mmmM
MODLE
ffiHWtfWWa
30 UPPER
20
io -
IO 12 14 16 18 20 22 24 26 28
11 13 15 17 19 21 23 25 27
CARAPACE LENGTH (mm)
Penaeus merguiensis
70
60
50
40
30
20
IO
IO 12 14 16 18 20 22 24 26 28
11 13 15 17 19 21 23 25 27
CARAPACE le n g th <mm) Penaeus indicus
Fig. 2a & 2b. Size frequency distribution of six prawn species by location in the upper, middle and lower reaches of the man
grove inlet (Sungai Sementa Kecil) and the adjacent mudflat.
205
140
gggggaggft
LOWER
120
c u : .1
M/FLATS
FREOflCVfê)
100
MIDDLE
80
tMMWBB
UPPER
60
40
20
13 15 17
a io 12 16
C A R A P A C E LENGTH ( m m )
Metapenaeus brevicornis
70
LOWER
60 -
M/FLAT5
50
MODLE
FREQUENCY'(?^
40 IM B H T O
LFFER
30
20
IO
i 11 13 15 17 10 21 23
10 12 14 16 IS 20 22
C A R A PA C E LENGTH ( m m )
Metapenaeus mutatus
LOWER
M
M ODLE
löUUÖUPPPB
UPPER
Macrobrachium spp.
C A R A PA C E LENGTH ( m m )
Fig. 2b.
206
ers. Penaeus merguiensis, P. penicillatus, P. indi a tropical Northern Australian estuary: their occurrence in
cus, M etapenaeus brevicornis and M . affinis m ay be the adjoining sea and estuarine dependence. Estuar. coast.
Shelf Sei. 29: 509-531.
m ore dependent on m angrove estuaries than other Boto, K. G. & J. S. Bunt, 1981. Tidal export of particulate
species. T hus it is possible th at the species com organic matter from a N orthern Australian mangrove sys
position o f offshore catches o f praw ns in M alay tem. Estuar. coast. Shelf Sei. 13: 247-255.
sia m ay change with the reclam ation or destruc Chong, V. C., 1977. Studies on the small grey mullet Liza
malinoptera. J. Fish. Biol. 11: 293-308.
tion o f large tracts o f m angroves.
Chong, V. C., 1980. The biology of the white prawn Penaeus
Fish and praw ns generally occur in greater merguiensis de M an (Crustacea: Penaeidae) in the Pulau
num bers and biom ass in m angrove habitats than Angsa Klang Strait waters (Straits of Malacca). M. Sc.
in adjacent near shore habitats (Thayer eta l., thesis. D epartm ent of Zoology, University of Malaya,
1987; R obertson & D uke, 1987; Blaber eta l., K uala Lumpur. 141 pp.
Chong, V. C. & A. Sasekumar, 1981. Food and feeding habits
1989; Chong eta l., 1990). The abundance o f a
of the white prawn Penaeus merguiensis. Mar. Ecol. Prog.
variety o f food item s and the turbid w aters appear Series 5: 185-191.
to be the m ain contributing factors. Although ju Chong, V. C., A. Sasekumar, M. U. Leh & R. D ’Cruz, 1990.
venile praw ns m ay not derive nutrition directly The fish and prawn communities of a Malaysian coastal
from m angrove detritus (S toner & Zim m erm an, mangrove system, with comparisons to adjacent mudflats
1988), it is ap p aren t th at significant quantities of and inshore waters. Estuar. coast. Shelf Sei. 31: 703-722.
Coull, B. C. & S. S. Bell, 1979. Perspectives in marine meio-
m angrove carbon are being assim ilated through faunal ecology. In R. J. Livingston (ed.), Ecological Pro
secondary sources (Rodelli eta l., 1984). W e cesses in Coastal and Marine Systems. Plenum Press, New
therefore conclude th at m angroves provide sig York: 189-216.
nificant support for fish and praw n com m unities. Cyrus, D. P. & S. J. M. Blaber, 1987. The influence o f tur
bidity on juvenile marine fishes in estuaries. Part I. Field
studies at Lake St. Lucia on the southeastern coast of
Africa. J. exp. mar. Biol. Ecol. 109: 53-70.
Acknowledgements de Freitas, A. J., 1986. Selection of nursery areas by six
Southeast African Penaeidae. Estuar. coast. Shelf Sei. 23:
W e th an k Jalal M o khtar and Jam ri T ohid for 901-908.
field assistance, and C heong Nge, G . H . Wee, Hatcher, B. G., R. E. Johannes & A. I. Robertson, 1989. Re
view o f research relevant to the conservation of shallow
P. K. G eeth a and Siti N uraini for assistance in
tropical marine ecosystems. Oceanogr. Mar. Biol. annu.
the laboratory. W e also thank D r. S usan Lim for rev. 27: 337-414.
her critical com m ents on the m anuscript. O ur Jansson, B. O., A. D. McIntyre, S. W. Nixon, M. M. Pam-
studies were supported by research grants from atmat, B. Zeitzschel & J. J. Zijlstra, 1988. Coastal-offshore
A ID A B (A ustralia) and R & D funds from the interactions - an evaluation of presented evidence. In B. O.
Jansson (ed.), Coastal-Offshore Ecosystem Interactions.
U niversity o f M alaya.
Springer-Verlag: 357-363.
Kesteven, G. L., 1960. M anual of field methods in fisheries
biology. FAO M anuals in Fisheries Science. No. 1, FAO,
References Rome. 152 pp.
Leh, M. U. & A. Sasekumar, 1984. Feeding ecology of prawns
in shallow waters adjoining mangrove shores. In E. Soe-
Alongi, D. M., 1986. Quantitative estimates of benthic
padmo et al. (eds), Proc. Asian Symp. Mangrove Environ
protozoa in tropical marine systems using silica gel: a
ment - Res. & Management: 331-353.
comparison of methods. Estuar. coast. Shelf Sei. 23: 443-
Leh, M. U. & A. Sasekumar, 1991. Ingression of fish into
450.
Alongi, D. M., 1988. Bacterial productivity and microbial bio mangrove creeks. In Alcala (ed.), Proc. Asean-Australian
mass in tropical mangrove sediments. Microb. Ecol. 15: ‘Living Resources in Coastal Areas’ Project. 30 Ja n -1 Feb.
59-79. 1989. Manila: 495-501.
Blaber, S. J. M. & T. J. Blaber, 1980. Factors affecting the Little, C., P. J. Reay & S. J. Grove, 1988. The fish commu
distribution of juvenile estuarine and inshore fish. J. Fish. nity of an East African mangrove creek. J. Fish. Biol. 32:
Biol. 17: 143-162. 729-747.
Blaber, S. J. M., D. T. Brewer & J. P. Salini, 1989. Species Macintosh, D. J., 1979. Predation of fiddler crabs ( Uca spp.)
composition and biomass of fishes in different habitats of in estuarine mangroves. In Srivastava, P. B. etal. (eds),
207
Proc. Symp. mangrove and estuarine vegetation in South Rodelli, M. R., J. N. Gearing, P. J. Gearing, N. Marshall &
east Asia. University of Agriculture, Malaysia: 101-110. A. Sasekumar, 1984. Stable isotope ratio as a tracer of
Maenae, W., 1974. Mangrove forests and fisheries. FAO, mangrove carbon in Malaysian ecosystems. Oecologia 61:
Rome. 35 pp. 326-333.
Moriarity, D. J. M., 1976. Quantitative studies on bacteria Sasekumar, A., 1981. The ecology of meiofauna on a Malayan
and algae in the food of the mullet Mugil cephalus L. and the mangrove shore. Ph. D. thesis. Departm ent of Zoology,
prawn Metapenaeus bennettae (Racek & Dali). J. exp. mar. University of Malaya, K uala Lumpur. 183 pp.
Biol. Ecol. 22: 131-143. Sasekumar, A., T. L. Ong & K. L. Thong, 1984. Predation of
Pinto, L., 1988. Population dynamics and community struc mangrove fauna by marine fishes. In E. Soepadmo et al.
ture of fish in the mangroves of Pagbilao, Phillipines. J. (eds), Proc. Asian Symp. Mangrove Environment - Res.
Fish. Biol. 33 (Supplement A): 35-43. and Management: 378-384.
Robertson, A. I., 1986. Leaf-burying crabs: their influence on Stoner, A. L. & R. J. Zimmerman, 1988. Food pathways as
energy flow and export from mixed mangrove forests sociated with penaeid shrimps in a mangrove-fringed estu
{Rhizophora spp.) in northeastern Australia. J. exp. mar. ary. Fish. Bull. 86: 543-551.
Biol. Ecol. 102: 237-248. Thayer, G. W., D. R. Colby & W. F. Hettler, Jr., 1987. Uti
Robertson, A. I. & N. C. Duke, 1987. Mangroves as nursery lization of red mangrove prop habitat by fishes in South
sites: comparisons of the abundance and species com Florida. Mar. Ecol. Prog. Series 35: 25-38.
position of fish and crustaceans in mangroves and other Turner, R. E., 1977. Intertidal vegetation and commercial
nearshore habitats in tropical Australia. Mar. Biol. 96: yields of penaeid shrimp. Trans, am. Fish. Soc. 106: 411-
193-205. 416.
■ -
Hydrobiologia 247 : 209 - 214 , 1992.
V. Jaccarini & E. Martens (eds), The Ecology of Mangrove and Related Ecosystems. 209
© 1992 Kluwer Academic Publishers. Printed in Belgium.
Joseph R asow o
M oi University, Zoology Department, B ox 3900, Eldoret, Kenya
Abstract
P ond culture o f brackishw ater praw ns and shrim ps has recently generated m uch interest in Kenya. The
m angrove areas are the target zones for the construction o f these ponds. W ith the increasing aw areness
o f the unique ecological role played by the m angroves, there is an urgent need to stop the conversion
o f m angrove sw am ps into aquaculture ponds. To develop pond aquaculture w ithout destroying the
m angroves, a shift from tide-fed to pum p-fed pond system s is recom m ended in order to divert the farming
from the m angroves to higher grounds. ‘M angrove-friendly’ m ariculture practices like pen, cage and raft
culture are discussed. M ethods o f efficiently utilising the already destroyed m angrove areas are consid
ered.
The m ariculture subsector has been lagging be orated in controlled areas to produce brine for
hind and it was in 1978 th at a pilot project w as crystallising salt. A typical saltw orks in K enya
set up to test the technical and econom ic viability consists o f a series o f ponds w here the brine is
o f brackishw ater aquaculture o f shrim ps on the transferred as it gets progressively m ore saline.
K enya coast. The production levels o f an average As m any as six transfers are involved from the
o f 525 kg h a - 1 y r _ 1 has created considerable in first reservoir through a series o f evaporation
terest am ong private sector entrepreneurs and n a ponds to a serving pond and finally to the crys-
tional policym akers resulting in plans being draw n tallisers (Fig. 1). The first reservoir ponds which
up for large scale com m ercial operations. generally occupy 7 -1 5 % o f the saltw orks total
This p aper proposes a cautious approach to area are indeed ready-m ade extensive ponds suit
the envisaged m ariculture developm ent which able for fish and shrim p farming. The sea w ater
w ould involve the integration o f aquaculture in the reservoir is always freshly draw n and has
with the m anagem ent o f m angroves and the effi a salinity which does not differ m uch from th at of
cient utilisation o f already reclaim ed m angrove the ocean since the pum ps are constantly in o p
areas. eration. U sing the estim ates o f Sam m y & Tyler
(1986) th a t some 65 tonnes o f sea w ater are
needed to produce one tonne o f salt, a solar salt
Efficient utilisation of already reclaimed man producer could culture plankton feeding anim als
grove lands
Irrespective o f locality, the solar salt m aking pro- p¡g I Generalized layout of a saltwork in Kenya (Source:
cess is fairly uniform , with sea w ater being evap- Yap & Landoy, 1986).
2 1 1
like oysters, m ussels and milkfish on the plankton Integrated solar salt - A rtem ia culture
content o f this volum e o f pum ped sea water.
T hese reservoir ponds could also be used as ex As the sea w ater gets progressively m ore saline
tensive shrim p ponds in which the shrim ps could w ith passage through a salt field there is a n atu
subsist and grow on the naturally occurring food ral m ortality o f organism s because o f the increas
in the ponds. Considering th at the physical facil ing salinity. T hose organism s that survive tend to
ities, civil w orks, and equipm ent required is com proliferate through lack o f com petition. O ne of
m on with th at for the salt industry and th at there these organism s is the brine shrim p Artemia which
is an existing pool o f m anpow er with actual ex thrives in the evaporator ponds where the brine
perience in all aspects o f pond construction and attains a concentration o f 60-3 0 0 % (Fig. 2). In
m aintenance, the establishm ent o f an aquaculture the low salinity ponds (1 0 0 -1 5 0 g salt I - 1 ) the
facility in the first reservoir m akes good econom ic Artemia reproduces by ovoviviparity (free-swim
sense. T he necessity and cost o f pum ping sea ming larvae) while at higher salinities
w ater are inherent in the operation of a salt field ( > 1 5 0 g I - 1 ) the anim al responds to the sub-
hence no extra costs w ould be incurred in obtain optim al conditions by producing cysts (ovipari-
ing the plankton biom ass. ty). Jones et al. (1981) have shown th at Artemia
S e a w a te r in ta k e w in d d ir e c tio n
100 g/l.
o voviviparous
ovoviv i p arou s
50 g- s a Its / 1 Artemia
Artemia
■(Continuous re c ru itm e n t)
dyin g A r t e m i a
g r a v i t y f l o w of w a te r
+ Ar t emi a
Fig. 2. Schematic diagram of solar salt operation with natural occurrence of Artemia (Source: Sorgeloos, 1983).
212
plays a fundam ental role in the salt m aking p ro Aquaculture in the mangroves
cess. The continuously non-selective filter feeding
Artemia keeps under control the algal bloom s Som e types o f aquaculture activities can be p ra c
which usually prevent early precipitation o f gyp ticed in m angrove areas w ithout adverse effects
sum and in extreme situations even ham per salt on the m angrove ecosystem . Pen, cage, raft, long
crystallisation. T h q Artemia m etabolites or the de line, and tray culture are good alternatives to pond
caying Artem ia also provide a suitable substrate culture in the m angroves although the m ethods
for the halophilic m icrobes o f the genus Halobac- require biological, technological and infrastruc
terium, the presence o f which assures a red col tural inputs. O yster culture is currently being
oration o f the cyrstallisers. D ark red crystallisers practiced in one creek in K enya w here a simple
assures m uch better heat absorption and thus and inexpensive culture system using a com bina
faster salt crystallisation than in a situation where tion o f rack and raft m ethod for grow th and roof
the white salt at the bottom o f the pond reflects ing tiles for spat collection has given very im pres
m ost o f the light (Sorgeloos, 1983; Jones e ta l., sive results w ith the local species Crassostrea
1981). In crustacean hatcheries Artemia is used as cucullata. K enya possesses 22 extensive m angrove
a practical and suitable larval food source either creeks covering 52 000 h a in surface area, thus the
in the form o f nauplii th at can easily be hatched potential for growing oysters is vast provided that
from the dry cysts or as adults (biom ass) th at can a m arket can be established. Som e species of
be harvested from evaporator ponds. Artemia Tilapia, e.g., T. spilurus and T. mossambica, show
biom ass w ould alleviate the feed supply problem rem arkable tolerance to salinity. T he feasibility of
in the ponds while the Artemia cysts which can tilapia cage culture within the m any sheltered tidal
com m and as high as $50 per kilogram in the world creeks should be studied. T he culture o f brack-
m arket could be a source o f foreign exchange ishw ater fish like rabbit fish (Siganus spp.), m ul
earnings. Artem ia culture in the coastal saltfarm s lets, and milkfish in floating cages or fish pens
has already been investigated and found viable should be encouraged. B w athondi (1986) reports
(R asow o & R adui, 1986) and only needs m an successful attem pts at cage culture o f Siganus
agement to enhance cyst and biom ass p roduc canaliculatus in neighbouring T anzania, though
tion. the identity o f the species in question is probably
S. sutor (F ischer & W hitehead, 1974).
Singh, T., 1987. The use of mangrove-managed areas for Turner, R. E., 1977. Intertidal vegetation and commercial
aquaculture. Workshop on the conversion of Mangrove yields of penaeid shrimp. Trans, am. Fish. Soc., 106:411-6.
areas to Aquaculture. U N D P/U N E SC O Regional Project Yap, G. W. & R. J. Landoy, 1986. Development of Coastal
RAS/79/002. Aquaculture. Report on a survey of the coastal areas of
Sorgeloos, P., 1983. Potential of the mass production of brine Kenya for shrimp farm development. KEN/80/018, Fish
shrimp Artemia. Journal of the Society for Underwater eries Department, FAO, Rome.
Technology. Spring 1983, 27-29.
A
Hydrobiologia 247: 215-222, 1992.
V. Jaccarini & E. Martens (edsj, The Ecology o f Mangrove and Related Ecosystems. 215
© 1992 Kluwer Academic Publishers. Printed in Belgium.
The effect of oocytic atresia on fecundity estim ates o f the rabbit fish
S i g a n u s s u t o r (Pisces: Siganidae) of Kenyan marine inshore waters
Abstract
In the strongly group-synchronized oocyte developm ent o f Siganus sutor (Valenciennes, 1835) the group
o f oocytes to be released in the following spawning, is identified. The sm allest size o f oocyte belonging
to this group w as identified by the presence o f cytoplasm ic vacuoles in oocytes in histological sections.
These vacuolated oocytes corresponded to oocytes o f 150 p m diam eter obtained by treatm ent with
G ilson’s fixative. T he m ean num ber o f such oocytes in stage 4 (late developing) ovaries w as found to
be 638000. T he p roportion o f these oocytes rem oved by atresia before spawning w as determ ined on
histological sections to be 5 %. T he corrected estim ate o f m ean fecundity w as thus 606 000 oocytes per
spawning.
The siganids contribute 50% o f the total artisanal were released by teasing. T he debris o f the ova
catch o f the reef fisheries o f K enya (N zioka, rian walls w as then rem oved and the rem aining
1984). The m angrove-lined creeks provide n urs fluid containing the oocytes poured into 11 b ea
ery grounds for these economically im portant fish kers. T he oocytes were w ashed in tap w ater which
(G rove eta l., 1986). w as changed several times by decanting after al
lowing the oocytes to settle. E ach tim e the w ater
w as decanted, the supernatant w as checked under
Materials and methods a dissecting m icroscope for any small oocytes.
F o r counting, the clean oocytes were poured
Fish were caught using traditional bottom traps into a 11 beaker and w ater m ade up to a know n
in shallow w ater ( < 14 m depth) in the vicinity of volume. To ensure the even distribution o f o o
M om basa, Kenya. Sam pling was done fortnightly cytes in the w ater colum n, a plastic ruler w as used
around neap tides throughout 1985, with an o c to stir the oocytes vigorously with a to and fro
casional additional sam ple taken on spring tides. m otion. A subsam ple w as taken after 10 strokes
The total num ber o f fish caught was 904. All fish o f the ruler by a Labsystem finelet 1 ml pipette.
were sexed and m aturity stages assigned accord O ne aliquot usually gave sufficient num bers o f the
ing to a m odified N ikolsky (1963) scheme, which large and small oocytes to yield satisfactory
h as been described elsewhere (N tiba & Jaccarini, counts and diam eter distributions.
1990). In this scheme six m aturity stages are dis T he oocytes were pipetted into a solid w atch
tinguished, viz. ( 1) virgin, (2 a) developing virgin, glass and their diam eter m easured along a hori
(2b) resting and recovering, (3) early developing, zontal axis, regardless o f their shape, under a
(4) late developing, (5) ripe, som etim es running, stan d ard m icroscope using a calibrated eye-piece
( 6 ) spent. graticule using a total m agnification o f x 40. The
accuracy o f the subsam pling m ethod w as tested
by taking seven replicates and calculating the co
Histological techniques
efficient o f variation, w hich w as found to be 6.2 %.
T he total num ber (N ) o f oocytes in any size class
F o r histology, portions o f ovaries were either fixed
in the ovary w as calculated as follows:-
for 24 h in Sm ith’s form al dichrom ate, w ashed in
running tap w ater for another 24 h, and stored in
V W
10% form alin, or fixed and stored in B ouin’s fluid. N =— nx— , where
The stored m aterial w as dehydrated in graded yi wl
alcohols, cleared in xylene and em bedded either
n = num ber o f oocytes o f a given size class
in paraffin wax or in ester wax (Steedm an, 1960).
in the subsam ple;
Sections were cut at 4 -1 0 fim and stained in iron
haem atoxylin and eosin. V = volum e o f sam ple;
3 .0 -
2 .9 -
1 . 2-
1.1-
1. 0 -
to
3 °-9 "
c/>
•2 0 .8 -
o
o
cn
c
0 .7 -
‘o.
o
§Ü 0 . 6-
-a
'S 0 .5 -
0.4-
0.3-
the yolk is resorbed and the oocytes collapse. am eter frequency distributions from portions of
A ssociated with such oocytes is a m ultitude of the sam e ovaries fixed in B ouin’s and in G ilson’s
squam ous follicular cells and m any blood vessels fluid separately supports this interpretation. Fig
w hich invade the atretic oocytes. These observa ure 5 shows the proportion o f atretic oocytes in
tions are in close agreem ent with those o f Cyrus different size classes in 10 ovaries in histological
& B laber (1984) working with Gerres species. sections o f m aterial fixed in Bouin’s and the size
Figure 4 gives the percentage frequency of frequency distribution for the sam e 10 ovaries
atretic oocytes for each m aturity stage o f the fixed in G ilson’s fluid. It is clear th at the batch of
ovary. It is apparent from this figure th at the de the interm ediate-sized oocytes (210-330 fim),
gree o f oocyte atresia is low est in stage 1 ( < 20 % ), present in the histological sections of stage 4 ova
rises to a m inor peak o f 45% in stage 3, and a ries, is alm ost com pletely m issing from the oocyte
m ajor peak o f 58% in stage 6 . Stages 4 and 5 population derived from G ilson’s treated m ate
have a rate o f atresia o f 27 and 32% , respectively. rial. As seen from histological sections, atresia
While some stages 4 and 5 ovaries treated with peaks precisely in the oocyte-size class missing
G ilson’s fluid h ad the interm ediate-sized oocytes from G ilson’s treated m aterial.
m issing altogether, this was never the case with It follows th at the estim ate o f 638000 as the
histological sections o f stage 4 ovaries, which al m ean fecundity o f S. sutor in K enyan w aters de
ways show ed developing oocytes o f all sizes in rived from stage 4 ovaries m ust be corrected by
cluding an appreciable proportion o f oocytes o f the proportion o f oocytes lost through atresia by
interm ediate size. This apparent discrepancy can
be explained by the observation th at m any o f the
interm ediate-sized oocytes (as well as some
smaller and larger ones) were atretic and h ad a
disintegrated chorion and were already invaded
20-
by a num ber o f squam ous follicular cells. One
would expect these oocytes to disintegrate under
G ilson’s treatm ent. A com parison of oocyte di-
cn
IQ-
3 40
cr
the time they reach stage 5, i.e., before they are (1977) reported a m ean per female o f 300 000 eggs
spaw ned. F rom Fig. 4, it is seen th at this am ounts w ith a m ean diam eter o f 0.5 mm. This wide range
to 5% over the stage 4 level. Applying th at cor agrees with our findings in w hich the fecundity
rection factor the estim ated resultant fecundity ranges from 200000 in a fish o f 18.0 cm to well
becom es 606000 oocytes per spawning. over 3 million in one o f 25.5 cm SL. A similar
Though there are tw o spawning peaks per year situation w as found for Trachurus trachurus o f the
in the population o f S. sutor (N tiba & Jaccarini, N orth Sea (M acer, op. cit.). M acer found a cor
1990), it is n o t clear from the evidence available relation betw een fecundity and the cube o f body-
w hether individual fish spaw n once or twice a length (P < 0.001). T hough we also see such a
year. O n the latter assum ption, the m ean cor correlation (P < 0.05) in S. sutor inspection o f re
rected fecundity estim ate w ould be about 1.2 mil siduals show s th at the linear regression m odel
lion. w as no t a good fit. B ut F isher’s z transform shows
no significant difference (z = 1.31) betw een M ac-
er’s and our correlation coefficients. So m ore d ata
Discussion
are required to test the apparent discrepancy with
Fecundity the m odel. A possible explanation o f our findings
is th at S. sutor is short-lived, e.g. the largest fish
T he strongly bim odal size frequency distribution in a sam ple o f over 900 w as less than tw o years
o f oocytes in stage 4 and 5 ovaries o f S. sutor old (N tiba & Jaccarini, 1988). O ne perhaps might
suggests th at oocyte developm ent in this species not expect a well-defined trend in fecundity to be
is group-synchronized. In this type o f develop established in the short tim e available.
m ent tw o populations o f oocytes are distin Som e w orkers use already running fish ob
guished: a m ore advanced, fairly synchronous, tained from com m ercial fishermen (H asse e ta l.,
population o f large oocytes, and a less advanced 1977; D e Souza, 1988). B ut the uncertainties in
heterogeneous population o f small oocytes from troduced by this procedure are obvious, since
w hich the larger ones are recruited (W allace & such fish m ay have already shed some eggs.
Selm an, 1981). Taking the fecundity for the im It is certain th at S. sutor populations spaw n
m ediately following spawning season to be the twice a year in E ast A frica w aters (N tiba & Ja c
total num ber of all developing oocytes in the ovary carini, 1990 and this Sym posium ). This is indi
at stage 4 m inus w hat will be lost to atresia by cated by the tem poral variation in the weights o f
stage 5, the estim ated m ean for S. sutor in K en the gonads, the seasonal appearance o f juveniles,
yan coastal w aters is 606000. D e S ouza (1988) the tem poral changes in the relative condition fac
obtained a total fecundity o f 700000 eggs for tor, K n, and especially by the seasonal occurrence
S. sutor at the K enya coast by counting only of spent males and females in the catches. This is
yolked eggs from stage 4 and 5 ovaries treated in agreem ent with results o f other w orkers study
w ith G ilson’s fluid. H is counts therefore involved ing different species o f siganids in Palau, Sin
only non-atretic oocytes. D e S ouza’s estim ate gapore, and the Philippines (Lam , 1974; K am i &
agrees num erically fairly well with ours, but this Ikehara, 1976; H asse eta l., 1977). H ow ever
is h ard to explain since we counted both yolked w hether the individual S. sutor spaw n once or
oocytes and the earlier vacuolated ones to obtain twice a year it is im possible to decide on the
our estim ate. M onacop (1937, cited in Lam , 1974) present evidence.
working with S. canaliculatus, a species closely
related to S. sutor, reports 300000-500000 o o
cytes in its ovary. In the ovaries o f artificially Functional significance o f atresia
spaw ned S. canaliculatus individual variation
ranging from 50000 to over 2 million eggs were In group-synchronous ovaries, a variety o f oocyte
found (Bryan eta l., 1975), while H asse etal. recruitm ent strategies exist (W allace & Selman,
221
340um
Fig. 6. Sections from stages 4 (a) and 5 (b) ovaries of Siganus sutor. Iron haematoxylin and eosin. AO: atretic oocyte; RO: resting
oocyte; LPP: late protoplasmic phase; EVP: early vitellogenic phase; LVP: late vitellogenic phase; CV: cytoplasmic vacuoles; EM:
egg membrane; YG: yolk granules.
1981). R ecruitm ent can be (a) directly from oo- oocytes, while at stage 5 there are two distinct
gonia, at the end o f the gonadotropin-independent m odes separated by a gap. These results suggest
stage, (b) from pre-vitellogenic to vitellogenic th at during developm ent from stages 4 to 5, the
stages, or (c) from oocytes which have term inated group o f largest oocytes is augm ented in num bers
the vitellogenic stage. In S. sutor recruitm ent by the later developm ent of the smaller ones. This
seems to be at the end o f the gonadotropin- is further supported by the histological appear
independent stage, since oocytes in this stage are ance of stages 4 and 5 ovaries (Figs 6 a and b) in
always present in the ovary. This type o f oocyte which the form er present a wide variety o f oocyte
recruitm ent w as show n in the herring, Clupea stages, from early resting ones, through vacu
harengus (Bowers & Holliday, 1961). C onsidering olated oocytes o f a wide range o f sizes, to ad
S. sutor (Fig. 1), it is clear th at at stage 4 there is vanced vitellogenic oocytes; while stage 5 ovaries
a considerable num ber o f the interm ediate-sized present an alm ost uniform distribution o f ad
222
vanced vitellogenic oocytes, with som e scattered Grove, S. J., M. C. Little & P. J. Reay, 1986. Tudor Creek
groups o f m uch smaller resting oocytes situated in Mombasa. The early life-history stages o f fish and prawns
1985. Overseas Development Administration, Research
the interstices. Project R 3888, U K, 133 pp.
A tresia o f oocytes is prevalent in both pre- and Guraya, S. S., R. K aur & P. K. Saxena, 1975. Morphology of
post-spaw ning stages o f S. sutor. While in the lat the ovarian changes during the reproductive cycle of the
ter atresia aids in the rem oval o f unw anted m a fish, Mystus tengara (Ham.). A cta anat. 91: 222-260.
terial from the ovary and helps to re-utilize valu Hasse, J. J., B. B. M adraisau & J. P. McVey, 1977. Some
aspects of the life history of Siganus canaliculatus (Park)
able m aterial, the functional significance o f p re (Pisces: Siganidae) in Palau. Micronesica 13: 297-312.
spaw ning atresia is not so clear. W allace & Kami, H. T. & 1.1. Ikehara, 1976. Notes on the annual juve
Selm an (1981) suggest th at poor nutrition and nile siganid harvest in Guam. Micronesica 12: 323-325.
horm onal im balances, especially o f the gonadot Lam, T. J., 1974. Siganids: their biology and maricultural po
ropins, m ay cause involution o f developing oo tential. Aquaculture 3: 325-354.
Macer, C. T., 1974. The reproductive biology of the horse
cytes in the pre-spaw ning phase. The ovaries m ay
mackerel, Trachurus trachurus (L.) in the N orth Sea and
function as storage organs. The occurrence o f English Channel. J. Fish Biol. 6: 415-438.
atresia at all oocyte m aturation stages m ay sug Mommsen, T. P. & P. J. Walsh, 1988. Vitellogenesis and oo
gest th at this storage function is a finely-tuned cyte assembly. In W. S. H oar & D. J. Randall (eds), Fish
process and th at this fish m ay be responding to Physiology 11A. Academic Press, New York: 374-406.
M onaco, P. J., E. M. Rasch & J. S. Balsano, 1978. Cytolog-
one or m ore physiological and environm ental cues ical evidence of temporal differences during asynchronous
(M om m sen & W alsh, 1988). ovarian maturation of bisexual and unisexual fishes of the
genus Poecilia. J. Fish Biol. 13: 33-44.
M onacop, P., 1937. The artificial fertilization of dangit,
Acknowledgements Amphacanthus oramin (Bloch & Schneider) (Probably = Si
ganus canaliculatus). Philip. J. Sei. 62: 229-237.
W e th ank M r. Alíela, D irector, K enya M arine Nikolsky, J. R., 1963. Ekologiya Ryb. Moscow: Vysshaya
Schkola Press, 368 pp.
and Fisheries R esearch Institute, M om basa, for
Ntiba, M. J., 1986. The biology of the Kenya reef fish of the
use o f facilities. This w ork w as supported by a genus Siganus. M.Sc. Thesis, University o f Nairobi,
G erm an D A A D research grant and another grant xviii + 143 pp.
from the N ational C ouncil for Science and Tech Ntiba, M. J. & V. Jaccarini, 1988. Age and growth param e
nology, K enya. W e th an k M r. K igondu and M r. ters of Siganus sutor o f Kenya marine inshore waters de
Oswago, D epartm ent o f Zoology, U niversity of rived from numbers of otolith microbands and fish lengths.
J. Fish Biol. 33: 465-470.
N airobi, for help with the illustrations. W e also Ntiba, M. J. & V. Jaccarini, 1990. G onad maturation and
th an k the B otany D epartm ent for help with the spawning times of Siganus sutor off the Kenya coast: evi
m icrographs. dence for definite spawning seasons in a tropical fish. J.
Fish Biol. 37: 315-325.
Nzioka, R. M., 1984. The evaluation of marine fisheries re
References sources in Kenya. In S .A . Iversen & S. Myklevoll (eds),
Proceedings of the NORAD -K enya seminar to review the
Bowers, A. B. & F. G. T. Holliday, 1961. Histological changes marine fish stocks and fisheries in Kenya. M ombasa,
in the gonad associated with the reproductive cycle of the Kenya, 13-15 M arch 1984. Institute of Marine Research,
herring (Clupea harengus L.). Mar. Res. 5: 1-16. Bergen, Norway: 27-32.
Bryan, P. G., B. B. M adraisau & J. P. McVey, 1975. Steedman, H. F., 1960. Section cutting in microscopy. Ox
Hormone-induced and natural spawning of captive Siganus ford: Blackwell, 172 pp.
canaliculatus (Pisces: Siganidae) year round. Micronesica Wallace, R. A. & K. Selman, 1981. Cellular and dynamic
11: 199-204. aspects of oocyte growth in teleosts. Am. Zool. 21: 325-
Cyrus, D. P. & S. J. M. Blaber, 1984. The reproductive biol 342.
ogy o f Gerres in Natal estuaries. J. Fish Biol. 24: 491-504. Yamamoto, K., 1956. Studies on the formation of fish eggs.
De Souza, T. F., 1988. Reproduction, length-weight relation 1. Annual cycle in the development of the ovarian eggs in
ship and condition factor in Siganus sutor (Valenciennes, the flounder, Liopsetta obscura. J. Fac. Sei. Hokkaido U ni
1835) (Pisces: Siganidae) from the Kenyan waters of the versity Ser. 6, Zoology 12: 352-373.
western Indian Ocean. Kenya J. Sei. Ser. B. 9: 89-101.
Hydrobiologia 247 : 2 2 3 -2 2 7 , 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 223
© 1992 Kluwer Academic Publishers. Printed in Belgium.
Abstract
C om parative studies on the biom ass and carrageenan production o f tw o strains of Eucheuma denticulatum
and tw o strains o f Kappaphycus alvarezii were m ade to assess the seasonality in their production ca
pacities.
The high and similar refined carrageenan (R C ) yields (43 -5 3 % o f dry w t.) o f the four strains in the
first cropping season (Ju n e -O cto b er) coincided w ith their high biom ass production with plants averaging
from 1.1 to 1.8 kg each at harvest. The poor R C yields (2 1 -3 3 % ) recorded in the second cropping
(O c to b er-F eb ru ary ) coincided with their season o f low biom ass (av. wt: 0.34 to 1.0 kg). The four strains,
however, recorded contrasting perform ance in the third cropping season (F ebruary-July) with the two
E. denticulatum strains recording high R C yields (43 and 42.5% ) together with high biom ass (av. wt: 1.5
and 1.6 kg) in co n trast to the low R C yields (30 and 39% ) and low biom ass (av. wt. 0.21 and 0.28 kg)
o f the tw o K. alvarezii strains. R ecords for semi-refined carrageenan (S R C ) yields in the second and third
cropping seasons were quite consistent and similar for the four strains (4 2 -5 5 % ), except in the second
cropping where the tw o K. alvarezii strains recorded low SRC. These differences in production poten
tials highlight the need for cropping m anagem ent o f the four strains to im prove their cropping perform
ance.
year, the seedstocks selected for the succeeding the fourth week harvest were analyzed for semi
cropping season are not at all acclim ated to this refined (S R C ) and refined carrageenan (R C )
season resulting in low production yields o f pure yields. T he extraction o f S R C w as done by M a
carrageenan o f crops. rine Colloids Philippines, Inc. in C ebu City. RC
The green (G S P 1 ) and the brow n (B SP1) w as processed by the alcohol precipitation
strains o f Eucheuma denticulatum (B urm an) C ol m ethod (Towle, 1973).
lins et H ervey and the green (G S T 1) and the The d ata on standing crop and carrageenan
brow n (BST1) strains o f Kappaphycus alvarezii yields o f the four strains from the different cro p
(D oty) D oty are com m only farm ed throughout ping periods were analyzed and com pared.
the year. E. denticulatum produces iota carrag
eenan and is called the ‘spinosum ’ o f com m erce
while K. alvarezii produces kappa carrageenan Results
and is the ‘cottonii’ o f com m erce. W e report here
the seasonal differences in biom ass production N o significant differences in the am ount o f RC
and carrageenan yields (semi-refined and refined) betw een strains were observed in the first cro p
of the four strains from three different cropping ping (Ju n e -O cto b er) (Table la). T he R C yields of
seasons with the view to illucidating the season the four strains were within the narrow range of
ality in their production potentials as basis for 4 3 -5 3 % o f dry weight and were the highest
their cropping m anagem ent. am ong values recorded in the three cropping sea
sons. T he biom ass o f the plants w as also the
highest recorded as all strains show ed good
M aterials and methods growth.
The second cropping (O c t-F e b ) w as a period
This research w as conducted at the A ntonio-A gro o f poor grow th for all strains as the strains suf
Seaw eed F arm on D anajon Reef, N orthern fered the ‘ice-ice’ disease, with the exception of
Bohol, Philippines (10° 1 6 ' N. L at; 124° 3 ' E. the brow n strain o f K. alvarezii (B ST1) which re
Long.) from M arch 1983 to Septem ber 1984. corded relatively good grow th (attaining m ore
T he fixed bottom m onoline m ethod (T rono & than 1000 g m axim um average biom ass). The
Fortes, 1988) w as used in the culture o f the four m axim um average biom ass attained by the other
strains. The experim ental plots for each strain strains w as barely 500 g.
consisted o f 20 m onolines, each m onoline planted Differences betw een strains in am ount o f car
to 20 plants/cuttings, the individual cuttings rageenan yields in the second cropping were o b
weighing 100 g. served. T he E. denticulatum strains (BSP1 and
F o ur croppings were m ade, the first was from G S P 1 ) yielded significantly higher am ount of
13 June to 5 O ctober 1983; the second from 12 S R C than did the K. alvarezii strains (Table lb).
O ctober 1983 to February 1984; the third from 7 T he am ount o f R C appeared to be significantly
April to 20 July 1984. The fourth cropping from different am ong the four strains although the green
5 August to Septem ber 1984 was destroyed by a strains (G SP1 and G S T 1) seemed to yield m ore
typhoon. The growing period lasted from 12 to 18 R C than did the brow n strains (BSP1 and B ST1)
weeks. (Table la). The yields o f R C relative to SR C from
The sam pling for biom ass w as done one week the four strains during this cropping season dif
after planting and weekly thereafter. Sam pling fered, i.e., the R C yields o f the K. alvarezii strains
w as done by random ly harvesting the whole were alm ost as m uch as the S R C in co n trast to
m onoline from experim ental plots o f each strain. those o f E. denticulatum strains w hich were con
The average biom ass (fresh) o f the plants in the siderably less th an th at o f the SRC.
m onoline w as recorded. The harvests (biom ass) In the third cropping (F e b -Ju l), the R C yields
were sun dried. The dried m aterials starting with o f the E. denticulatum strains were significantly
225
Table 1. Comparison of the RC (Table la ) and SRC (Table lb) yields (± std. deviation), in per cent of dry weight, o f the two
strains of Eucheuma (BSP1 and G SP1) and Kappaphycus (BST1 and GST1). Results of the ANOVA-Duncan’s Multiple Range
Test are shown on the left (for comparisons between cropping seasons within a strain) and on the right (for comparisons between
strains within a cropping season) of the values; means with the same letter are not significantly different from each other (p < 0.05).
For Table lb, the t-Test was used to compare the yields of a strain from the two cropping seasons; s = significant, ns = not sig
nificant (p< 0.05).
Table la
Ju n -O ct O ct-F eb Feb-Jul
Table I b
O ct-F eb F eb-Jul
higher th an those o f the K. alvarezii strains weeks) w as required to attain m axim um biom ass
(Table la). All strains, however, yielded fairly the in the third cropping com pared to the first crop
sam e am ount o f SRC but the am ount o f R C from ping (7 weeks). In contrast, high biom ass pro
the K . alvarezii strains w as significantly less than duction am ong the K. alvarezii strains w as ob
th at o f the SRC while the am ount o f R C from the served only in the first cropping periods. Their
E. denticulatum strains w as alm ost as m uch as R C yields appeared to follow the trend in their
their SRC. biom ass production, e.g., high during the first and
The four strains appeared to follow a com m on low during the second and the third croppings.
pattern of intercropping differences in the yields
o f RC, i.e., highest in the first cropping, lowest in
the second, and interm ediate in the third Discussion
(Table la). These differences appeared significant
am ong the E. denticulatum strains while they did The low quality o f farm produce (low R C yields)
n o t appear substantial in the K . alvarezii strains has always been attributed to poor post-harvest
in the third cropping. processes. D espite efforts to induce farm ers to
Differences in biom ass production o f the dif im prove the quality o f their produce such as
ferent strains during the three cropping seasons through graded pricing based on the physical
were also observed. The E. denticulatum strains quality, processors still receive batches o f sea
showed a tren d similar to th at o f their R C yields: weeds with low R C yields. The results o f this
high in the first and third croppings but low in the study show th at the state o f the crop at the time
second cropping although a longer period (16 o f harvest, i.e., w hether healthy or not, appeared
226
to be an im portant factor affecting the SR C and from good cropping period but the R C yields were
R C yields. The am ount o f yields appeared to vary significantly lower. A large portion o f the carra
with the grow th perform ance o f the strains, as geenan in diseased or sickly crops apparently is
indicated by biom ass production. High R C yields degraded to other products and cannot be recov
o f the four strains were recorded in the J u n e - ered by alcohol precipitation during processing of
O ctober cropping period when grow th w as good R C. But during S R C processing, this portion is
(high standing crop) and significantly low yields apparently retained. The presence o f these de
in the O cto b er-F eb ru ary cropping period when graded products m ay reduce the gel strength and
grow th w as poor. Thus, the observed intercrop viscosity o f the carrageenan. This im portant
ping differences in the yields o f R C appear to take problem needs clarification.
exception to the general observation th at rapid The results o f this study show th at the produc
grow th o f carrageenophytes, as a result o f ele tion potentials o f the different farm ed strains v ar
vated nutrient levels is accom panied by less ca r ied with species and with cropping seasons. The
rageenan production (D aw es et al., 1974a, b; M c- production in the first cropping period w as no t at
C andless & Craigie, 1979; C hopin e ta l., 1990). all affected by choice o f species or strains as all
The lack o f agreem ent between the results o f the show ed good perform ance (high biom ass and
present studies and those o f the other w orkers carrageenan yields). In the second cropping pe
might have been caused by differences in am bient riod, however, the green strains o f both species
environm ental conditions. D aw es et al. (1974a, b) appeared to record better crop perform ance than
reported th at m axim um grow th and m inim um did the brow n strains. In the last cropping, both
carrageenan yields occurred in period o f low tem strains o f E. denticulatum recorded good grow th
perature. T rono & O hno (1989) reported high com pared to those o f K . alvarezii. This inform a
growth and biom ass during periods o f relatively tion can be used as basis for cropping m anage
high tem perature regimes. m ent o f different strains to im prove the p ro d u c
Differences in the RC yields and biom ass p ro tivity o f the farm s.
duction observed am ong the four strains appeared
to be species-specific. T he two E. denticulatum
strains showed similarly seasonality in biom ass Acknowledgements
and R C yields as did the two strains o f K. alva
rezii. T rono & O hno (loc. cit.) reported the same This study w as supported by the International
species-specificity in the seasonality o f biom ass D evelopm ent and R esearch C enter (ID R C ) -
production of the different strains. But differences C an ad a and the Philippine Council for A quatic
in R C yields between the brow n and green strains and M arine R esearch and D evelopm ent (PC A M -
o f the tw o species were also noted during the R D ).
second cropping periods with the green strains of
the two species recording higher R C yields than
their brow n counterparts.
The distinctly higher ratio o f SR C to R C yields References
o f crops during cropping seasons o f low biom ass
Chopin, T., M. D. Hanisak, F. E. Koehn, J. Mollion & S.
production and lower ratio during cropping sea
Moreau, 1990. Studies on carrageenans and effects of sea
sons of high biom ass production are very ap p a r water phosphorus concentration on carrageenan content
ent. The incidence o f w hat is know n as the ‘ice- and growth of Agardhiella subulata (C. Agardh) Kraft and
ice’ disease is preceded by the paling o f the color, Wynne (Rhodophyceae, Solieriaceae). J. appl. Phycol. 2:
the loss o f the glossy surface and thinning o f the 3-16.
Dawes, C. J., A. C. Mathieson & D. P. Cheney, 1974a. Eco
branches, and occurrence o f epiphytes and other logical studies of Floridean Eucheuma (Rhodophyta, G i
weeds. The SRC yields during this period rem ain gartinales) I. Seasonal growth and reproduction. Bull. mar.
approxim ately similar to th at recorded in crops Sei. 24: 235-273.
227
Dawes, C. J., J. M. Lawrence, D. P. Cheney & A. C. Mathie- their Derivatives. Academic Press, New York/London: 8 3 -
son, 1974b. Ecological studies of Floridean Eucheuma 113.
(Rhodophyta, Gigartinales) III. Seasonal variation of car Trono, G. C. Jr. & E. G. Fortes, 1988. Philippine Seaweeds.
rageenan, total carbohydrate, protein and lipid. Bull. mar. National Bookstore, Metro Manila. 330 pp.
Sei. 24: 286-299. Trono, G. C. Jr. & M. Ohno, 1989. Seasonality in the biomass
M cCandless, E. L. & J. S. Craigie, 1979. Sulfated polysac production of the Eucheuma strains in Northern Bohol,
charides in red and brown algae. Ann. Rev. PI. Physiol. 30: Philippines. In I. Umezaki (ed.), Scientific Survey of M a
41-53. rine Algae and their Resources in the Philippine Islands.
Towle, Gordon, A., 1973. Carrageenan. In R. L. Whistler & Monbushio International Scientific Research Program,
J. N. BeMiller (eds), Industrial Gums: Polysaccharides and Japan.
■- ü i
H
Aiyj w p i M i^iii^li
T, iw fc a y in w w w ^ 'i ,it o j
35 ??
Hydrobiologia 247: 2 2 9 -2 3 4 , 1992.
V. Jaccarini & E. Martens (eds). The Ecology o f Mangrove and Related Ecosystems. 229
© 1992 Kluwer Academie Publishers. Printed in Belgium.
Abstract
The distribution and growth form s o f the m angroves on Z anzibar island (U nguja and Pem ba) were
investigated. P em ba island, with approxim ately 12000 h a o f m angroves, h ad nine species; and U nguja
island, with 6000 ha, had eight. Average cropping intensities in P em ba island varied between 350-1937
cut plants per h a while in U nguja these were between 777-3567 cut plants per ha. In Pem ba, m angrove
forests were relatively less heavily exploited com pared to those in U nguja, where clear felling was
com m on. C harcoal, lime and salt production are the m ajor activities which involve heavy utilization o f
m angrove wood.
The fishery resources (shellfish and finfish) associated with the m angrove vegetation were also inves
tigated by examining the com position o f fish catches at landing stations in m angrove as well as n o n
m angrove zones of fishing districts. A survey on the fauna associated with m angrove forests was also
conducted. A num ber o f the local com m ercial fish species were found to be associated with the m an
grove vegetation. The fishes o f the genus Lethrinus and the rabbit fish Siganus were found to constitute
the bulk o f the fish caught from w aters adjacent to m angroves using m ovable traps. The m ollusc Pyrazus
sp., which is com m only used by the local people for food and bait, was found only in the m angroves.
39* 15' E 39* 30’ E 35' 40' 39* 4 5 'E 50’ 55’
las Kigomasha
Ras Kiuyu
Mangrove areas
surveyed.
WSifà Mangrove areas not
surveyed during the
study. Njao Is, ¡heweni
5* OCfS-
Bumbwini 05‘-
5*00' S- PEM BA
IS LA N D
10 -
UNGUJA ISLAND
£> Chake Chake Ba)
15 -
Chwaka
I Bay
Zanzibar Town
20-
Kisakaka« Mkoanl
8*15-5-
(Unguja Pete>
Kiwani
S Bay iMuungoni
Mangroves
study site. The inform ation collected at each site
included a list o f m angrove species, an estim ate The distribution o f the species of m angroves to
o f cut m angrove trees in random sam ple plots of gether with estim ates o f their percentage abun
5 m X 5 m, percent coverage by species, diam eters dance and cutting intensities at the study sites is
and average lengths o f random ly sam pled m an given in Table 1.
grove poles, and the fauna associated with the Prior to these studies, eight species o f m an
m angroves. Inform ation regarding the local uses groves were recorded for T an zan ia ( S emesi, 1986)
of the various m angrove species w as gathered and these were found both on the m ainland and
from local residents through interviews. The spe the islands. Xylocarpus moluccensis (Lam ) Roem
cies com position o f the fish caught by artisanal w as identified during these studies and has so far
fishermen w as obtained by identification o f fish been recorded only in P em ba islands (J.P .S .) and
species at landing sites and calculation o f the per T anga area on the m ainland (Sem esi, 1988). The
centage com position o f the species for each fish other species include Bruguiera gymnorrhiza (L)
ery type e.g. dema trap, towe trap, hand lines etc. Lam , Rhizophora mucronata Lam ., Ceriops tagal
231
Table 1. In fo rm a tio n o n m a n g ro v e s o n U n g u ja a n d P e m b a , Z a n z ib a r.
Total mangrove areas (ha)c _ 344 460 320 3240 520 1288 1324 1128
Cutting (N r/ha)d - 2317 1463 777 n.a. 3567 1937 350 335
a F = firewood, C = charcoal, P = poles, B = boat building, D = dye, L = leather tanning, T = fish traps, M = medicine.
b Underlined % value refer to species m ost affected by cutting at the location indicated.
c From Griffith (1949).
d Average of three plots in each area.
(P err.) C.B. R obinson, Avicennia marina (F orsk) O n P em ba island (See Fig. 2) the m angrove
Vierh, Lumnitzera racemosa Willd, Xylocarpus forests are similarly found on the w estern side.
granatum K oen, Sonneratia alba J. Smith, and They stretch southw ard from N jao island north
Heritiera littoralis D ryand in Ait. The m ost com o f W ete tow n through C hakechake to M koani.
m on species found on the sites are R. mucronata, M ore th an three quarters o f the m angroves lie in
B. gymnorrhiza, and S. alba. this region. O ther com paratively less extensive
A bout tw o thirds o f the m angrove forest stands but im portant stands are found at A dam son bay
on Z an zib ar are located in Pem ba island while on the east coast, and in the bay betw een R as
one third is found in U nguja island. The areas K iuyu and R as K igom asha on the northern part
with m angroves in U nguja (See Fig. 1) are lo o f the island. The total area covered by m angrove
cated m ostly to the w est and include M enai bay vegetation here has been estim ated (Griffith, 1949)
(U nguja U kuu, Pete and M uungoni, Kiwani bay at 12000 ha. This figure will also be revised after
(K isak asak a) and M akoba bay (M uw anda and interpretation o f m ore recent data.
Bumbwini). On the eastern side o f U nguja island The various local use o f the m angrove species
there is only one stand o f m angroves, found are enum erated in Table 1. The m ost destructive
around C hw aka bay. T here are also small stands o f these are firewood and charcoal production
at M aruhubi and C hukw ani to the north and because the trees are cut indiscrim inately result
south o f Z an zib ar tow n respectively. The U nguja ing in wide gaps in the area. Several affected areas
m angrove stands cover about 6000 ha. However, were seen at U nguja U kuu, K isakasaka, M u
m ore accurate d ata about the area and cover of w anda, around C hw aka bay (U nguja) and to
m angroves would be available after the interpre some extent at M icheweni (Pem ba) where cutting
tation o f m ore recent aerial photographs. w as done for lime production. Salt production by
232
boiling seaw ater, although less practised now has e.g. L . racemosa, C. tagal and the m angrove as
caused dam age to som e areas, e.g. the M uw anda sociate, Pemphis acidua, the roots o f R . mucronata
area along M ak o b a bay (U nguja). This area w as are the m ost com m only used.
know n for salt production and collection o f ho- F or m edicinal purpose, only X . granatum and
lothurians (bêche de mer) for export in the early H. littoralis are used. A pow der obtained from the
seventies, both activities involve boiling trea t seed coat from both species is dissolved and orally
m ent, thus requiring firewood. adm inistered for the treatm ent o f abdom inal ail
The use o f m angrove tim ber is also im portant m ents, however, overdose o f this concoction can
especially in U nguja U kuu, C hw aka, K isakasaka, be dangerous.
and to some extent at M ichew eni and The utilization o f m angroves in Z an zib ar is
C hakechake (as show n in Table 1). Pole p roduc thus carried out at as shown in various intensities
tion w as qualitatively estim ated to be m ore in (Table 1). Som e o f the stands are under m ore
tense in U nguja th an in Pem ba, as judged by the serious threat than others. F rom interviews with
am ount of poles seen staked at various points on villagers it w as apparent th at in som e places, in
the islands. This is probably due to a higher de discrim inate m angrove cutting had already re
m and for construction m aterials because o f the sulted in serious exposure o f the coastline to ero
relatively higher population density on U nguja in sion. T he M uw anda village, for example, was
co n trast to Pem ba. T he larger area o f m angroves reported to have been shifting landw ard away
in P em ba and the low cutting pressure from the from the approaching ocean. Ruins o f the older
smaller hum an population probably explains the settlem ents th at stand surrounded by w ater could
superior grow th form in term s o f tree stem thick be seen as evidence o f loss o f land to the sea. This
ness, observable in m ost stands (as show n in T a phenom enon has been partly due to the cleared
ble 1). m angrove vegetation.
L eather tanning is another use o f the m angrove.
The b ark o f R . mucronata is stripped off the trees, Associated fishery resource
pounded and m ixed with lime, and soaked in
m etal or concrete troughs into w hich anim al skins D uring the survey o f the fauna associated with
are im m ersed and left to soften for one to tw o m angroves, the m ost com m on fish encountered
weeks. T he skins are then ready for use in the were Lethrinus harak (F orskal) p arro t fishes,
m anufacture o f shoes (makubadhi), in small local Chanos chanos (F orskal) and fishes o f the genus
shoe factories. L eather tanning w as com m on at Siganus. C rustaceans included Scylla serrata
U zi island in U nguja; and at M icheweni, W ete (Forskal), herm it crabs, and praw ns; and m ol
and C hakechake in Pem ba. D ebarked trees die luscs included Pyrazus sp., Pinctada sp., Modiolus
on drying. The problem is serious because the sp., Crassostrea sp.
trees affected are the m ature seed producing The association o f fish with m angrove vegeta
plants. tion has been pointed elsewhere (A hm ad, 1966),
T he m aking o f fixed stake trap s utilizing the and the potential o f several tropical fisheries have
roots o f R . mucronata is another destructive prac been know n to correspond to the respective area
tice. It rem oves the supporting roots, som etim es o f m angrove vegetation (M acN ae, 1974). It was
o f an entire plant leading to the collapse o f the evident also from the interviews conducted d u r
trees. In the m aking o f the trap the stilt roots of ing the study th at a considerable am ount o f co n
the m angrove are cut and split into thin trips ventional as well as non-conventional fishing
which are then woven together in a long line. takes place either within or in the vicinity o f the
T hese long lines are then stuck erect in the m ud m angrove stands. N otable am ong the non-
or sand in m id-shore at low tide to trap fish m ov conventional fishery is the collection o f m olluscs.
ing out with the w ater on the next low tide. Al Evidence for the m agnitude o f edible m ollusc col
though some other m angrove species are also used lected w as provided by the large heaps o f em pty
233
Table 2. Species composition of fish landings at three stations on Unguja island, Zanzibar (Tarbit, 1976).
shells left at various points on the beach after the m entioned above, instead, the coralline fishes o f
rem oval o f the m eat. Villagers reported th at m ost the genera Scolopsis and Pampeneus dom inated
o f the molluscs are used to supplem ent protein the catches.
especially for those families th at cannot afford C intron & Schaeffer-Novelli (1983) observed
finfish. A small proportion is used as bait in hook th at due to the presence o f juvenile organism s and
and line fishing. M olluscs are also collected for their m igration in m angroves as well as seagrass
their opercula especially those from the genus beds, these regions m ake excellent fishing grounds
M urex. T hese opercula cost as m uch as U S $ 40 even for larger predatory fishes such as red snap
per kilogram. A t C hakechake for instance the pers, yellow fin tuna, sharks and other highly ca r
trade in this article w as reported to have been nivorous species th at forage these regions. On
quite attractive and th at the products were col Z an zib ar islands, 62% and 82% o f all the fixed
lected and exported to M om basa, Kenya. stake trap fishery units operate on the w est coast
Three fishery types (based on the type o f fish o f U nguja and Pem ba, respectively. O f the m ov
ing gear) were found to be associated with the able traps, 77% and 85% operate on the respec
m angroves and adjacent habitats. These included tive w est coasts o f U nguja and Pem ba. Likewise,
fixed stake trap s (uzio), m ovable traps (madema m ore troll line fishery units (89.5 % in U nguja and
and towe) and hand line fisheries (H oekstra et a i, 76.6% in Pem ba), a gear which is used to catch
1990). A com parison o f trap fishery landings at large pelagic fish, operate on the w est coasts where
three stations (M alindi, C hw aka and M kokotoni) extensive m angrove stands are found (H oekstra
indicates th at m ore fish are caught from m an eta l., 1990).
grove-lined areas than elsewhere (Tarbit, 1976).
Im p o rtan t fish species belonging to the families
Lethrinidae, Siganidae and Scaridae are caught Conclusion
in greater num bers in m angrove-lined area (as
show n in Table 2). How ever, fish landed at M a M angrove sw am ps support several types o f fish
lindi (Z anzibar Town), which are caught m ostly eries on Z an zib ar islands and indeed worldwide.
from non-m angrove lined areas, were found to M angroves, also provide direct resources to the
contain less o f the m angrove associated groups coastal com m unities and play an im portant role
234
in protecting the coastline from erosion. H ow Cintron, G. & Y. Schaeffer-Novelli, 1983. Mangrove forests:
Ecology and response to natural and man induced stres
ever, it is n o t realized th at over-exploitation o f the
sors. U N ESC O Report in Marine Science # 23 ROST-
m angrove forests leads to a decline in the fisheries LAC. 133 pp.
resources and severe environm ental dam age. Griffith, L. A., 1949. Reconnaissance report on the forest
There is therefore an urgent need for the intro problems o f Zanzibar protectorate. Govt. Printer, Zanzi
duction o f effective m anagem ent m easures that bar. 21 pp.
Hoekstra, M., M. A. K. Ngoile, N. S. Jiddawi & C. Rotteglia,
w ould guarantee sustained utilization o f the m an
1990. Census of the marine fishing units of Zanzibar in 1989
groves. F A O /U N D P: RAF/87/008/DR/60/99/E. 119 pp.
MacNae, W., 1974. Mangrove forest and fisheries. Indian
Ocean Program Publication Number 34, Rome, Indian
Acknowledgement Ocean Fishery Commission. 35 pp.
Semesi, A. K., 1988. Status and utilization of mangrove along
W e wish to express our deep appreciation to the the coast of Tanga, Tanzania. In Mainoya, J. R. (ed.), Pro
C om m ission o f L ands and Environm ent (C oLE), ceedings of workshop on Ecology and Bioproductivity of
Z an zib ar, for suggesting and providing funding to marine coastal waters of Eastern African, D ar es Salaam,
support the m angrove survey on U nguja and Tanzania, 18-20 January 1988. Faculty o f Science, U ni
versity of D ar es Salaam: 174-181.
P em ba island through F IN N ID A ’S financial as
Semesi, A. K., 1986. Zonation and vegetation structure of
sistance. W e would also like to thank all those mangrove communities in Tanzania. In Mainoya, J. R. and
who have contributed in one way or another in P. R. Siegel (eds), Proceedings of W orkshop on ‘Save the
the preparation o f this paper. mangrove ecosystems in Tanzania’. D ar es Salaam, Tan
zania, 21-22 February 1986. Faculty of Science, University
of D ar es Salaam: 15-35.
References Tarbit, J., 1976. Demersal Fisheries Research Terminal Re
port 1974-76. EAM FRO Annual Report 1976, EAM FRO,
Ahmad, N., 1966. Fish and fisheries of the Sundarbans. In Zanzibar.
Scientific problems of humid tropical zone deltas and their
implications. Proceedings of the D accar Symposium
February/M arch 1964, U N ESC O Paris: 271-276.
Hydrobiologia 247: 2 3 5 -2 4 1 , 1992.
V. Jaccarini & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 235
© 1992 Kluwer Academic Publishers. Printed in Belgium.
K ey words: assessm ent, Caribbean, ecosystem interactions, m anagem ent, m angroves, site variability
Abstract
M angrove ecosystem s occupy different locations on C aribbean island coasts, ranging from open bays
(fringe m angals) to totally enclosed salt ponds and salinas. On géom orphologie ally active coastlines, such
as south Jam aica, systems are at varying degrees o f m aturity and productivity. Furtherm ore, because
o f system variability, the interactions betw een m angroves and associated m arine systems, such as coral
reefs and seagrass beds, are developed to different degrees.
Com m unity structure and productivity of a range o f m angals on different islands o f the G reater and
Lesser Antilles are discussed. Forcing functions and levels o f interaction with the m arine environm ent
are identified.
The rational choice o f m anagem ent options m ust be based on the range o f goods and services p ro
vided by the different system s; and a good understanding o f their ecology is essential w hen choosing sites
for protection, w aste disposal, landfill, m arina developm ent and fisheries enhancem ent. Exam ples are
given from current studies in Jam aica, St. Lucia, the B ritish Virgin Islands and Trinidad, o f a flexible
m anagem ent response to m angrove ecosystem diversity.
types, w hich have different ecological attributes. and lagoon developm ent has influenced m angrove
F urtherm ore, B acon (1986)pointed out th at these distribution and successional processes; so that
m angals were associated with a variety o f coastal there are pioneering, m ature and senile m angal
ecosystem s w hose geographic characteristics d e com m unities (B acon, 1986). Site aspect also in
term ined their ecology and productivity. This fluences the degree o f hurricane im pact, which is
p aper examines the influence o f site location on a further factor controlling m angrove forest struc
the ecological and econom ic attributes o f m angals ture and productivity (Alleng, 1990).
in selected C aribbean islands and discusses how M angrove areas studied in St. Lucia show ed a
these features can be incorporated into the m an similar site variability on both east and w est
agem ent decision-m aking process. coasts o f this geom orphologically different island
(see Fig. 2). T he generally small m angrove stands
occupied open bay fringes or were isolated from
Description of sites studied the sea behind beach barriers which breach only
in tim es o f heavy flooding. Site characteristics
M angrove systems were surveyed on a 50 km influence plant distribution, tidal exchanges, sa
stretch o f coastline near K ingston, south Jam aica; linity range, groundw ater and riverine inputs,
these ranged from open bay fringe m angal to en sedim ent grain size and com position and nutrient
closed salina with scrub m angroves (see Fig. 1). status.
O n this geom orphologically active coast, barrier
Hunts Bay
Riverine
M a n g ro v e s Kingston
Port Royal
Lagoon fringe
Gt . S a l t Pond
Ephemeral hypersaiine pond
C ab a r ita
Complex fringe/
basin wetland M anatee Bay Lousy Bay Pond
Open bay fringe Closed salina
y West Harbour
Lagoon fringe
Jam aica
E s p e r a ne e
H b r.
dH Sand bar
Mangr ove s
Bo i s
D'Or a n g e
Bay
Choc
Bay
Ma n
Kote'
Fig. 2. Selected mangrove sites in St. Lucia. Scale for a -c , & f= 1:10000; Scale for d & e = 1:25000.
238
The ecological and socioeconom ic values o f these The danger o f generalizing about m angrove m an
tw o groups o f m angals were assessed using a agem ent is apparent when this assessm ent m eth
m ethodology adapted from A dam us (1983). The odology perm its the ranking o f system s according
m ethod assigns a rating to attributes and func to their respective values. All the sites were p o s
tions o f these w etlands in order to obtain a ra n k itive for some values, but the ranking identified
ing th a t can be used for m anagem ent decision areas o f priority for conservation while singling
m aking (see Table 1). T he rating system used w as out sites with lower value th at might be available
simple, scoring High (3), M edium (2), Low (1) for use as m arinas, dum p sites or landfill. All
and N one (0), because o f the relative paucity o f isolated and sem i-isolated sites scored low values
d ata for some categories. Field observation and for wildlife habitat and fisheries, and the potential
m apping by the senior author and reports on the for m ariculture, educational and recreational use
Jam aican sites (Chow, 1989; M cC ain & Bacon, (see the W H , EC & R E categories in Table 1).
1989; Alleng, 1990) and St. Lucian sites (Porte- Em baym ent and Lagoon-type sites also received
cop & B enito-Espinal, 1985) provided detailed the highest rankings when this evaluation system
infirm ation on areal extent, plant distribution and w as used in the B ritish Virgin Islands (Blok-
use o f the habitats by wildlife, but assignm ents Meeuwig, 1990); in T ortola, for example, top pri
u nder the socioeconom ic categories (such as, im ority sites for conservation were P araquita (la
portance to fisheries, education value) were based goon fringe), W itches Brew (lagoon fringe) and
largely on inform ed opinion. H odges Creek (em baym ent fringe).
Table 1. Assessment of ecological & socioeconomic values of selected mangrove sites in Jam aica & St. Lucia.
A. Jam aica
Kingston Hbr. 3 3 2 3 3 3 2 2 2 2.3
West Hbr. 3 3 2 3 3 3 2 2 2 2.3
G SP/Flashes 3 3 2 2 3 3 2 2 2 2.2
Cabarita 3 2 2 3 2 2 2 2 2 2.0
M anatee Bay 3 1 2 1 3 2 1 1 1 1.5
Long Pond 2 2 2 1 3 1 1 0 1 1.3
Lousy Bay 1 1 1 1 1 1 1 0 0 0.7
B. St. Lucia
Savannes Bay 3 1 2 1 3 2 2 2 2 1.8
Esperance Hbr. 3 2 2 1 3 2 2 1 1 1.7
Marigot Hbr. 2 1 2 1 2 2 2 1 1 1.4
M an Kote 3 2 3 1 0 1 1 1 1 1.3
Bois d’Orange 2 2 2 1 0 1 1 0 1 1.0
Choc 1 1 1 1 3 1 1 0 0 0.9
O pen m angals perm it m ovem ent o f m arine and H owever, it still acts as a source o f nitrite and
estuarine organism s in and out of the system and phosphorus to the inshore w aters w hen there is
export o f organic m atter. Ogden & G ladfelter positive net discharge only during the wet season
(1983) noted th at interaction betw een m angals, and particularly after storm s (M cC ain & Bacon,
seagrass beds and coral reefs was a com m on fea 1989). This situation is similar to the N ariva m an
ture o f C aribbean coasts, but did n o t com m ent on grove lagoon, Trinidad, where discharge to the
the degree o f variability which exists in such link Atlantic O cean occurs only during tw o to three
ages. The nature o f the association between three m onths o f the wet season, when the effluent
o f the south coast Jam aica m angals and neigh stream can carry up to 100 t org.C d a y ” 1 from
bouring m arine ecosystem s is show n in Fig. 3, to the 90 km 2 w etland (B acon, 1981). Such pulses of
illustrate this point. The com m unities at these lo nutrient-rich, brackish w ater have a seasonal, or
cations are fringe m angroves dom inated by interm ittent im pact on neighbouring m arine h ab
Rhizophora. In P o rt Royal they occupy approxi itats.
m ately 6 ha, w ith m ean canopy height 4.6 m (Al As W est H arb o u r lacks riverine inputs, has
leng, 1990), while in the G reat Salt P ond/F lashes prevailing onshore w inds, a w eak circulation p a t
area they occupy approxim ately 25 ha, with m ean tern, and experiences a high evaporation rate be
canopy height 5.4 m. Fringe m angroves in W est cause it is located in one o f the driest areas of
H arb o u r occupy approxim ately 260 ha, with Jam aica, a net outflow o f w ater containing dis
m ean canopy height at 6.0 m. A part from varia solved and particulate organic m atter has no t been
tions in areal extent, other structural param eters detected. This m angal probably has little direct
are within the sam e order o f m agnitude and ac influence on the ecology o f the adjacent P ortland
cord well with the description o f fringe m angroves Bight, bu t functions as a net accum ulator o f o r
given by Lugo & Snedaker (1974). ganic m atter. Such lagoonal system s which act as
How ever, the geographic setting and hydrog sinks for organic m atter occur in several p arts of
raphy influence the fate o f the organic m atter and the Jam aican coast, but are poorly described.
nutrients generated within the three systems, with M angrove production appears to be available to
im plications for the nature o f their linkages with consum ers only if they enter the system, hence
contiguous ecosystem s. Site geography indicates W est H a rb o u r’s im portance as fish habitat and
th at the im pacts o f exported m aterials are prob nursery environm ent and its potential for mari-
ably in inverse relation to the area o f m angroves culture.
at each site. P o rt Royal m angroves are function T he relative roles o f the three fringe mangals
ally part o f the K ingston H arb o u r system, which show n in Fig. 3 have been assessed as indicated
has strong fluviatile influences, w ind-driven and in Table 1 using differences in com m unity struc
tidal circulation patterns. K ingston H arb o u r is a ture and ecological forcing functions (Lugo &
m ajor exporter o f nutrients and organic m atter to S nedaker, 1974) and degree o f site variability
the inshore environm ent and its effluent strongly (B acon, 1986). This allows these ecologically and
influences the neighbouring reefs and seagrass economically im portant systems to be rated for
beds (H ead & H endry, 1986; B acon et al., 1989). m anagem ent purposes.
Furtherm ore, this system is an open, continuously W hat has no t been done is to include in the
exporting system. In contrast, the G re at Salt assessm ent m ethodology criteria which express
P o n d /F lash es m angroves developed around an the value o f the m angals to associated m arine
hypersaline p o n d closed by an ephem eral sand habitats. Although the value o f the m angal itself
bar. A lagoon form ed briefly during the wet sea has been assessed, the relationship between p ar
son when flood w ater breached the barrier per ticular m angrove sites and the continued health
m itting entry o f sea water. N ow th at there is a and productivity o f an island’s fisheries, seagrass
perm anent artificial channel stable, m ore m arine beds or coral reefs is a critical factor in coastal
conditions have been established in the pond. resources m anagem ent. If this dim ension can be
240
& K i n g s t o n H a r b o u r
<¿r Port R o y a l
iling
S a lt Pond P o rt R o y a l
& Cays
FI a s h e s
M a n g r o v e s
5 km
M C o ra I reefs
il S e a g ra s s
beds
P o r tla n d
P r e V a i I i n9
Bight
w i nd
West
Harbour
3 km
Fig. 3. Associations of three Jam aican mangals with neighbouring coastal ecosystems.
241
D. Van Speybroeck
University o f Nairobi, Botany Department, P.O. B ox 30197, Nairobi, Kenya
A b s tra c t
In June 1990 a research project sponsored by D A N ID A and A W F w as carried out by botany students
o f N airobi U niversity to investigate the regeneration strategy o f m angroves at G azi bay and M ida creek.
Statistical analysis o f 449 quadrats (5 x 5 m ) sam pled along 35 line transects in 4 m angrove forests
showed th at m angrove seedlings follow the sam e distribution pattern in the intertidal zone as their parent
trees. In other w ords, m angrove seedlings mainly develop within a well-defined species specific zone.
T hese distribution zones for the various m angrove trees and their seedlings are defined in term s of el
evation above the m ean low w ater level o f spring tides.
T hrough assigning m angrove seedlings in the intertidal zone to the categories (i) ‘fixed’ or ‘not-fixed’,
and (ii) ‘covered’ or ‘not-covered’, evidence w as found th at propagule dispersal followed both the
‘self-planting’ theory and the ‘stranding’ theory. The ‘self-planting’ theory appeared to be the m ajor
m echanism o f propagule dispersal in ‘undisturbed’ m angrove forests, whilst the ‘stranding theory’ proved
to be predom inant in colonizing ‘over-exploited’ and cleared m angrove forests.
It is concluded th at re-afforestation o f m angrove seedlings in the intertidal zone will be m ost successful
w hen the seedlings are planted in their specific distribution zones under fixed conditions.
R e s u lts
M ida c re e k
Zonation o f mangrove trees and mangrove seedlings
Indian O cean
4°S Figure 2 (G azi bay) and Fig. 3 (M ida creek) sum
m arize the distribution in the intertidal zone o f 7
M om basa
species o f m angrove trees and their respective
seedlings in term s o f height above M LW S.
G azi bay
km 100 Inspection o f Fig. 2 shows th at m angrove trees
T a n za n ia and m angrove seedlings clearly grow in species
specific distribution zones. M edian elevation val
Fig. 1. Map of Kenya situating the study sites at Gazi bay
(226 vegetation quadrats) and Mida creek (223 vegetation
ues (in cm ) for adult trees (tr.) and seedlings (sd.)
quadrats). for the respective m angrove species are:
Sonneratia alba (tr. = 160, sd. = 148), Rhizophora
along each line transect. In each quadrat the m an mucronata (tr. = 212, sd. = 198), Bruguiera
grove vegetation w as described according to the gymnorrhiza (tr. = 212, sd. = 209), Ceriops tagal
principles o f the floristic-sociological approach (tr. = 245, sd. = 240), Avicennia marina (tr. = 258,
(W esthoff & van der M aarel, 1978). Species cover sd. = 256), Lumnitzera racemosa (tr. = 271,
and abundance were estim ated by the ordinal sd. = 299) and Xylocarpus granatum (tr. = 282,
scale o f van der M aarel (1979). All the m angrove sd. = 270). Figure 2 suggests the existence of
seedlings ( = young trees having at least 4 leaves, m onospecific m angrove belts along the intertidal
and less th an 1 m high) present in the quadrats slope, starting with Sonneratia alba at the sea
were counted and assigned to the categories fixed w ard side, bu t it nevertheless allows mixed m an
or not-fixed and covered or not-covered. grove stands as well, since there is a m ajor over
The elevation in cm above M LW S w as m ea lap in distribution pattern betw een Rhizophora
sured for each quadrat, according to the line level mucronata and Bruguiera gymnorrhiza, between
m ethod (D aw es, 1981). Heights above M LW S Ceriops tagal and Avicennia marina and between
were com puted using the tide table (June 1990) Lumnitzera racemosa and Xylocarpus granatum.
for Kilindini harbour (M om basa). Figure 3 show s the box plots for the distribu
G raphical representation o f d ata distributions tion o f m angrove trees and m angrove seedlings at
in Figs 1 and 2 follows the box plot display M ida creek. The sam pled m angrove forests at
(C ham bers eta l., 1983). A box plot sum m arizes M ida creek are described as used m angrove for
the distribution o f a param eter by m eans o f 5 ests. Lumnitzera racemosa and Xylocarpus
specific values from the d ata set: the m edian granatum, although present at M ida creek, did
( = the 50 percentile), the upper and lower quartile not figure in our transects. Only 6 Sonneratia alba
(the 75 and 25 percentiles) and the upper and trees were present in the line transects with a
246
Table 1. Percentage of seedlings growing under ‘fixed’ and ‘covered’ conditions as observed at G azi bay and M ida creek. G azi 1
refers to an ‘undisturbed’ mangrove forest, M ida 1 and M ida 2 to ‘used’ mangrove forests and G azi 2 to an ‘over-exploited’
mangrove forest. The percentage of seedlings growing under ‘not-fixed’ and ‘not-covered’ conditions are not listed since they add
up to 100% with the ‘fixed’ and ‘covered’ seedlings respectively. The table summarizes growing conditions o f 9479 seedlings in
the four line transects.
400.
Gazi bay H i mangrove tree EU mangrove seedling
E 350
o
03
o
300
o
O o
XX
E o
3 250- o
CO
-O
03
> 200
o 1
-
? ±
n o
CO
£ 150-1
03
0)
s: loo.!
TV s i o o
Fig. 2. Box plots showing the position above MLWS of mangrove trees and mangrove seedlings at G azi bay. The figure includes
data from 79 quadrats from an undisturbed mangrove forest (G azi 1) and 147 quadrats from an over-exploited mangrove forest
(Gazi 2).
m edian elevation above M LW S o f 101 cm , but A lthough no direct evidence is given in the
no Sonneratia alba seedlings were found. F rom present paper, no difference w as observed in the
Fig. 3 it is seen as well th at m angrove trees and zonation pattern o f m angrove trees and seedlings
their seedlings tend to grow in species specific in undisturbed, used and over-exploited m an
distribution zones. grove forests.
247
400-
Mida creek |ü mangrove tree □ mangrove seedling
E 350-
o
(D
> 300- o
o
E
3 o
250- o
co
X3 i
<D
> 200 -
O
_Q
CO
— 150-
D)
X
CD
100- TT T TTT
50- — i--------1-------- 1--------1-------- 1-------- r
Sonneratia Rhizophora Bruguiera Ceriops Avicennia Lumnitzera Xylocarpus
alba mucronata gymnorrhiza tagal marina racemosa granatum
Fig. 3. Box plots showing the position above MLWS of mangrove trees and mangrove seedlings at M ida creek. The figure in
cludes data from respectively 103 quadrats (M ida 1) and 120 quadrats (M ida 2) from used mangrove forests.
100 h eig h t a b o v e d atu m (cm ) 400 forests occupy the intertidal zone between 100
and 370 cm above M LW S. Although there is a
Sonneratia alba
substantial overlap betw een the distribution p a t
terns in the intertidal zone for the various m an
Rhizophora mucronata grove species, a clear zonation pattern is appar
ent and easily recognized in the field. If present,
Bruguiera gymnorrhiza
Sonneratia alba form s the outerm ost forest m ar
Ceriops tagal gin tow ards the open w ater, followed by pure
stands o f Rhizophora mucronata or mixed stands
Avicennia marina of R. mucronata and Bruguiera gymnorrhiza, then
followed by pure or mixed stands o f Ceriops tagal
Lumnitzera racemosa
and Avicennia marina, and with at the m ost land
Xylocarpus granatum w ard m argin o f the forest stands o f Lumnitzera
racemosa and a very small margin, if still present,
Fig. 4. Profile diagram showing the overall zonation pattern of Xylocarpus granatum. Seedling developm ent is
of mangrove trees (full lines) and mangrove seedlings (dashed
restricted to the species specific distribution zones
lines) at Gazi bay and M ida creek. Thick lines are defined by
the upper and lower quartiles, whilst thin lines are defined by along the intertidal slope.
the upper and lower extremes. Albeit no t the case in our transects, it should
be noticed th at Avicennia marina m ay replace
Figure 4 shows the profile diagram o f m angrove Sonneratia alba at the seaw ard side o f m angrove
trees and m angrove seedlings along the intertidal stands in small channels and creeks (G raham ,
slope at G azi bay and M ida creek. M angrove 1929; M acN ae & K alk, 1962).
248
100
CD
a> " u n d is tu r b e d "
- - ■© " o v e r -e x p lo ite d "
T3
" u n d is tu r b e d "
Sonn, fíhiz. Brug. Ceri. Avie. Sonn. Rhiz. Brug. Ceri. Avic.
Fig. 6. Percentage seedlings growing under fixed conditions Fig. 7. Percentage seedlings growing under covered condi
in an over-exploited (Gazi 2; dashed line) and an undisturbed tions in an over-exploited (Gazi 2; dashed line) and an un
mangrove forest (Gazi 1, M ida 1 and M ida 2; full line). M an disturbed mangrove forest (Gazi 1, M ida 1 and M ida 2; full
grove species are listed according to their position on the line). Mangrove species are listed according to their position
intertidal slope, with Sonneratia alba (Sonn.) closest to the low on the intertidal slope, with Sonneratia alba (Sonn.) closest to
water line, followed by Rhizophora mucronata (Rhiz.), the low water line, followed by Rhizophora mucronata (Rhiz.),
Bruguiera gymnorrhiza (Brug.), Ceriops tagal (Ceri.) and Bruguiera gymnorrhiza (Brug.), Ceriops tagal (Ceri.) and
Avicennia marina (Avic.). Avicennia manna (Avic.). The Ceriops zone in the over-,
exploited mangrove forest behaves as in an undisturbed man
covered situation. W hen we also take into co n grove forest (it still has an overall tree cover of 70 % ), resulting
sideration the uprooting effect o f the wave action, in an untypical high percentage covered seedlings for an over-
exploited mangrove forest.
we m ay assum e th at there is a positive correlation
betw een the intensity o f the wave action and the
num ber o f seedlings growing under fixed condi ests (Table 1, Fig. 7), further supports the idea of
tions. Figure 6 shows th at tow ards the seaw ard coexistence o f the ‘stranding’ and the ‘self
side o f the m angrove forests (where exposure to planting’ theory. Seedlings in undisturbed m an
wave action is m axim al) nearly 90 % o f the m an grove forests how ever are principally covered, in
grove seedlings are fixed. In undisturbed m an dicating a preponderance o f the ‘self-planting’
grove forests this percentage decreases with theory over the ‘stranding’ theory, whilst seed
higher positions in the intertidal gradient, coin lings in over-exploited m angrove forests are prin
ciding with a gradient o f decreasing wave action. cipally not-covered, supporting here the idea of a
Figure 6 shows as well th at at the very landw ard preponderance o f the ‘stranding’ theory.
site o f the undisturbed m angrove forest (where
exposure to wave action is m inim al) only 25% of
the seedlings grow under fixed conditions. In the C o n c lu s io n
over-exploited m angrove forests on the other
han d the percentage o f fixed seedlings does not Analysis o f 449 vegetation quadrats along 36 line
decrease with higher positions in the intertidal transects in m angrove forests at G azi bay and
gradient and even never drops beyond 75% . This M ida creek show ed th at establishm ent o f m an
can be explained (i) by an inadequate protection grove propagules, once dispersed, takes place in
o f the intertidal slope from wave action (more species specific distribution zones in the intertidal
than 75 % o f the m angrove trees have been felled) zone, which we defined in term s o f height above
and (ii) by the suprem acy o f the ‘stranding’ the M LW S. The resulting zonation o f m angrove
ory over the ‘self-planting’ theory in an over- seedlings starts with Sonneratia alba closest to
exploited m angrove forest. the low w ater level, respectively followed by
Inspection o f the percentage covered seedlings Rhizophora mucronata, Bruguiera gymnorrhiza,
in undisturbed and over-exploited m angrove for- Ceriops tagal, Avicennia marina, Lumnitzera
250
racemosa and Xylocarpus granatum. This zonation their overall support, and to D r. M . J. N tiba, M r.
pattern perfectly corresponds with the zonation A. O. O luoch, M r. C. A. A ketch, M r. J. M ur aya,
pattern displayed by their parent trees. This clear M r. S. M athenge and 50 second year students for
m angrove zonation pattern nevertheless allows their help with the field w ork. W e th an k the D a n
the existence o f mixed stands o f m angrove trees ish International D evelopm ent Agency (D A N I
in the intertidal zone. D A ) at N airobi, K enya, for financing this study,
Analysis o f growing conditions o f seedlings in as well as the A frican Wildlife F oundation of
undisturbed and over-exploited m angrove forests W ashington D. C., A frican operations, for their
supported the idea th at propagule dispersal fol contribution.
lows b oth the ‘self-planting’ and the ‘stranding’
theory. The ‘self-planting’ theory being the m ajor
m echanism o f propagule dispersal in undisturbed R e fe re n c e s
Kokwaro, J. O., 1985. The distribution and economic impor ishwater intertidal zones at the Kenyan coast. Discovery
tance of the mangrove forests in Kenya. J. E. Afr. Nat. and Innovation 2: 49-55.
Hist. Soc. 75: 1-12. Ruwa, R. K., 1990b. Impact on mangroves and crabs in creek
MacNae, W. & M. Kalk, 1962. The ecology of the mangrove and fringe mangrove vegetation due to disturbances at the
swamps at Inhaca island, Moçambique. J. Ecol. 50: 19-34. Kenyan coast. Paper presented at the International Sym
Moorjani, S. A., 1977. The ecology of marine algae of the posium on Mangroves and related Ecosystems, 1990, M o
Kenya coast. University of Nairobi, Ph.D. Thesis. 285 pp. mbasa, Kenya.
Orlóci, L. & N. C. Kenkel, 1985. Introduction to data anal Tomlinson, P. B., 1986. The botany of mangroves. Cambridge
ysis. Statistical Ecology monographs 1. International Co university Press Tropical Biology Series. 413 pp.
operative Publishing House, Maryland, USA. 339 pp. Van Speybroeck, D., 1990. Field course to the Kenya coast.
Rabinowitz, D., 1978a. Dispersal properties of mangrove June 1990. Report on the activities. University of Nairobi,
propagules. Biotropica 10: 47-57. Botany Department, internal report. 65 pp.
Rabinowitz, D., 1978b. Early growth of mangrove seedlings van der Maarel, E., 1979. Transformations of cover-
in Panama, and an hypothesis concerning the relationship abundance values in phytosociology and its effects on com
of dispersal and zonation. J. Biogeogr. 5: 113-133. munity similarity. Vegetatio 39: 97-114.
Ruwa, R. K. & P. Polk, 1986. Additional information on Walter, H. & M. Steiner, 1936. The ecology of East African
mangrove distribution in Kenya: some observations and mangroves. Zeitschrift für Botanik 30.
remarks. Kenya Journal of Science and Technology Series Westhoff, V. & E. van der Maarel, 1978. The Braun-Blanquet
B 7 (2): 41-45. approach. In: Whittaker, R. H. (ed.), Classification of plant
Ruwa, R. K., 1990a. The effects of habitat complexities cre communities. D r W. Junk Publishers, The Hague: 287-
ated by mangroves on macrofaunal composition in brack- 399.
| t i ^ J * # ){ É | ¡ '» y -» "W ;* i 3
|Ífj^fl^ij^(p^lÍ'Í^ÍB
ntíjoAlk è A
Hydrobiologia 247: 2 5 3 -2 5 9 , 1992.
F. Jaccaríni & E. Martens (eds), The Ecology o f Mangrove and Related Ecosystems. 253
© 1992 Kluwer Academic Publishers. Printed in Belgium.
K ey words: M angrove ecosystem , species com position, zonation, com m unity structure, past
m anagem ent practices
A b s tra c t
M uthupet m angrove ecosystem (10° 25' N and 79° 30' E) is located at the southern end o f the Cauvery
delta on the east coast o f India. The swam p is characterized by the presence o f only 4 species o f ex
clusive m angroves nam ely Avicennia marina, Excoecaria agallocha, Aegiceras corniculatum and Acanthus
ilicifolius.
C om m unity structure analysis carried out at four random ly selected quadrats show ed th at the rela
tive density, relative dom inance, relative frequency and im portance value o f A . marina w as relatively
higher than for the other three species. H ence, M uthupet m angrove ecosystem is characterized by the
mono-specific dom inance of A . marina. Earlier palynological studies indicate th at species like Rhizophora
were present in this swam p some 200 years ago. It is believed th at the p ast m anagem ent practices to
gether with the present poor scientific policy o f the decision m akers and harsh environm ent conditions
m ay be considered as a few factors responsible for the disappearance o f Rhizophora from this mangrove
swamp.
four were established in the pure thickets form ed namely, P rosopis zone, S uaeda zone, m angrove
by exclusive m angroves and the rem aining one zone and m ud flat zone (Fig. 1). The P rosopis
w as laid in the non-m angrove zone. F rom each zone is present on the landw ard side o f the lagoon
q u ad rat the following d ata w ere collected: i) n- and its length and breadth were approxim ately 1
um ber o f species o f plants ii) num ber o f individ to 1.2 km and 30 to 50 m respectively. This zone
uals o f each species iii) circum ference o f the trees w as com pletely occupied by the non-m angrove
at b reast height. The circum ference readings were species Prosopis spicigera L. which is a thorny
divided by n to obtain the diam eter at breast which w oody plant. A few individuals o f E. agallocha
were used for basal area com putation, using the were also present on the banks o f the small tidal
following formula. creeks of the P rosopis zone. N ext to the P rosopis
g(m2) = 0.00007854 dbh2, where ‘g’ is the basal zone, a well m arked Suaeda zone w as present.
area in m 2, dbh is the diam eter at breast height. This zone w as occupied by the halophyte, Suaeda
The relative frequency (R F), relative dom i maritima is an undershrub. T he m angrove zone
nance (R D o), relative density (R D e) and im por w as located next to the S uaeda zone and was
tance value (IV) were calculated using the stan restricted to the edges o f the brackish w ater la
dardized m ethodology (C intron & Novelli, 1984). goon. The length o f the m angrove zone varied
The im portance value o f a given species is the from 0.2 km to 1.0 km and the breadth w as ap
sum o f R F, R D o and R D e o f th at species. proxim ately 0.05 km to 1.5 km. All the four ex
clusive species o f the m angroves were present in
this zone.
Past management practices
Excoecaria agallocha 32 - 8 24 - -
(6.25) (25) (75)
Opuntia monocantha 14 14 - - - -
(2.73) (100)
Total 512
Suaeda maritima 94 9 — — — —
(23.74) (100)
Total 396
Total 252
Excoecaria agallocha 62 - 18 37 7 -
(18.45) (29.03) (59.68) (11.29)
Total 336
Aegiceras corniculatum 24 20 4
(10.13) (83.33) (16.67)
237
which constituted 10.13% o f the total individu five quadrats are given in T able 2. In Q uadrat I,
als. P. spicigera was the dom inant species w hose rela
The com m unity structure indices such as rela tive dom inance, relative density, relative fre
tive dom inance, relative frequency, relative den quency and im portance values were 93.54% ,
sity and im portance value o f each species o f the 91.01% , 33.34% and 217.89 respectively. In all
257
Table 2. Basal area (m2), relative domainance (% ) relative density (% ) relative frequency (% ) and importance value of flora of
Muthupet mangrove swamp.
Q uadrat I
Prosopis spicigera 0.7321 93.54 91.01 33.34 217.89
Excoecaria agallocha 0.0506 6.36 6.25 33.33 46.94
Opuntia monocantha - - 2.74 33.33 -
Q uadrat II
Avciennia marina 1.8324 100 75.26 50 226.26
Suaeda martima - - 23.74 40 -
Quadrat III
A. marina 1.2346 100 72.41 50 222.41
S. maritima - - 27.59 50
Quadrat IV
A. marina 1.9080 87.35 81.55 50 218.90
E. agallocha 0.2766 12.65 18.45 50 81.01
Q uadrat V
A. marina 1.5388 93.70 89.87 50 233.57
Aegiceras
corniculatum 0.1035 6.30 10.13 50 66.43
- indicates value could not be calculated as the stem diameter was less than 3 cm.
other q u adrats A . marina was the dom inant spe the non-m angrove species P. spicigera w hereas
cies. The relative dom inance value o f A . marina in A. marina show ed monospecific dom inance in the
Q uadrats II and III w as 100% w hereas in Q ua m angrove zone.
d rat IV and V its value was 87.35% and 93.70%
respectively. The relative density o f A . marina in Past management practices
Q u ad rat II to V w as 75.26% , 72.40% , 81.55%
and 89.87% respectively. The im portance value A ccording to C hengappa (1918), the M uthupet
o f A . marina was also high in all the last four m angrove forest w as under the control o f the
q u adrats the values being 226.20 (Q .II) 222.41 C hatram D epartm ent from 1853 to 1912. The
(Q .III) 218.90 (Q.IV) and 233.57 (Q.V). Thus the C hatram D epartm ent, a branch o f the then D is
com m unity structure analysis indicates th a t the trict B oard o f Tanjore, looked after the num erous
landw ard portion o f the swam p is dom inated by C hatram s (free boarding and lodging houses)
258
which were constructed by the R aja o f T anjore bearers. The average yield w as 25 tonnes/
for the convenience o f pilgrims and travellers. To hectares.
m aintain these C hatram s, the C hatram D ep art
m ent collected revenue from M uthupet m angrove
forest. They harvested the m angrove w ood and D is c u s s io n
sold it in the m arket as tim ber and firewood. The
revenue thus realized was spent to m aintain the The species com position o f M uthupet m angrove
C hatram s. In 1912, the Revenue D epartm ent took forest is very poor w hen com pared to the num ber
charge and during this year the First W orking o f exclusive m angrove species present in the Pit-
Plan for T anjore m angroves w as proposed and chavaram m angrove forest which is located ju st
advocated. A ccording to this plan the coppice 140 km north o f the M uthupet m angrove. U n
system o f clear felling with 12 years rotation was controlled and unregulated exploitation by the
introduced. The revenue collected from selling the authorities o f the C hatram D epartm ent prior to
m angrove w as given to the C hatram D epartm ent. 1912 and the unscientific m anagem ent policies
In 1923, once again the C hatram D epartm ent followed by the Revenue as well as C hatram D e
took over the responsibility o f m anaging the m an partm ents after 1912 and some environm ental
grove forest and followed a 20 years rotation with factors like poor inundation and dam age due to
clearfelling as advised by the F orest D epartm ent cyclonic storm seem to be the reason for the poor
(V enkatesan, 1966). The C hatram D epartm ent species diversity o f the M uthupet m angrove for
exploited the m angrove resources o f M uthupet till est. The available records show, th at prior to 1912,
1936. F rom 1923 to 1936, half o f the revenue the w oody plant species o f M uthupet m angrove
obtained from selling m angrove products was were exploited throughout the year for sales as
paid to the Revenue D epartm ent and the rem ain tim ber as well as for fuelwood by the C hatram
ing half was spent to m aintain the C hatram s. In D epartm ent (C hengappa, 1918). After 1912, a
February 1937, M uthupet m angrove forest was coppice m ethod with 12 years rotation period w as
declared as reserve forest by the G overnm ent and followed for the harvest o f m angrove wood. A c
accordingly the m angrove forest w as handed over cording to this m ethod an area o f m angrove forest
to the F o rest D epartm ent o f the M ad ras Presi w as allotted for clear felling by the R evenue D e
dency (G azette, 1937). After 1937, the F orest D e partm ent which w as then brought to the notice of
partm ent m anaged the M uthupet m angroves fol the C hatram D epartm ent. Since authorities of
lowing the 20 years rotation system. B ased on the both the R evenue and C hatram D epartm ents
frequency and dom inance o f Avicennia which was were not aw are o f the im portance o f leaving seed
found to be a coppicing species, 50 trees per acre bearers in the clear felling area they rem oved all
were left as seedbearers. A ccording to the study the trees which resulted in the poor regeneration
o f V enkatesan (1966) it w as found th at the above of the m angrove plants in th at particular area
ro tation pattern w as followed w ith the added p re (Ahm ed, 1937). To correct the above m istake,
scription o f retaining all advanced grow th below A hm ed (1937) introduced another system accord
1.8 m (6 feet) in height. ing to which all the trees, except 50 seed bearers/
N o detailed inform ation is available on the yield hectare were harvested. H e also changed the 12
and revenue obtained by the above m anagem ent years clear felling rotation to 20 years. A hm ed
bodies from M uthupet m angrove Forest. H ow (1937) also noticed the poor regeneration o f the
ever, V enkatesan (1966) m entioned th at from m angrove plants and prescribed artificial regen
1935 to 1948, the average yield w as 16 tonnes/ eration o f A . marina by direct sowing. But this
hectare in T anjore district. In a neighbouring practice w as w ithdraw n after a few trials due to
m angrove forest area, C uddalore, South Division unsatisfactory results. H e also suggested the dig
o f Tam il N adu, a cycling pattern o f 40 years w as ging up o f channels in areas where the ground
followed leaving 38 m other trees/acre as seed level was raised due to fresh accretion o f sand
259
and where Avicennia w as showing signs o f death Blasco, F., 1975. Mangroves of India. Inst. Fr. Pondicherry,
Trav. Sect. Sei. Tech., Vol. 14. 180 pp.
due to insufficient inundation. This suggestion
Burbridge, P. R., 1984. Management and Planning of M an
w as discontinued after a few attem pts due to high grove Resources of India. Proc. As. Symp. Mangr. Env. -
cost involved in it. The coppice system with 20 Res. Manag., 27-42.
years rotation period w as followed by the F orest Chengappa, D., 1918. Working Plan for the Tanjore Forest
D epartm ent till 1970. The m anagem ent policy Division. Revenue Department, District Board of Tanjore,
78 pp.
suggested by C hengappa (1918) w as unscientific
Cintron, G. & Y. S. Novelli, 1984. Methods for studying man
because it was based on the wrong assum ption grove structure. In: Snedaker, S. C. & Snedaker J. G. (eds),
th at all m angrove plant species could regenerate The Mangrove Ecosystem: Research Methods. M ono
from the ro o t stock. Blasco (1975) showed th at graphs on Oceanographic Methodology. U N ESCO , Paris,
only the species such as Avicennia and Excoecaria 91-113.
Gazette, 1937. Government of the Presidency of M adras,
have the capacity to regenerate from ro o t stocks.
M adras, 318 pp.
The palynological studies carried out by T issot Gamble, G. S. & C. E. S. Fischer, 1957. Flora of the Presi
(1987) indicate th at the species o f Rhizophora and dency of M adras Vol. I—III, Government of India, Cal
Sonneratia were present in the M uthupet m an cutta, 1388 pp.
grove ecosystem some 200 years ago but they Prain, D., 1905. Flora of the Sunderbans. Ree. Bot. Sur. India
2: 237-370.
have disappeared com pletely now. The uncon
Saenger, P., E. J. Hegerl & J. D. S. Davis, 1983. Global sta
trolled and unregulated exploitation carried out tus of mangrove ecosystems. Commission of Ecology paper
by the authorities o f the C hatram and Revenue No. 3 IU C N , Gland, 38 pp.
D epartm ents, some unscientific policies intro Sidhu, S .S ., 1963. Studies on the mangroves of India. I-East
duced in 1920 and thereafter as well as the poor G odavari region. Indian Forester 89: 337-351.
Tissot, C., 1987. Recent evolution of mangrove vegetation in
attention paid tow ards the replantation of the
the Cauvery Delta: A palynological study. J. mar. biol. Ass.
m angrove plants seem to be some o f the causes India 29: 16-22.
responsible for the present poor species diversity Untawale, A. G., 1984. Present status of mangrove along the
o f the M uthupet m angrove forest. west coast of India. Proc. As. Symp. Mangr. Env. - Res.
Manag: 57-74.
Untawale, A. G., 1987. Conservation of Indian mangroves -
A national perspective. In: Rao, T. S. S. (ed.), Contribution
R e fere n ce s
in Marine Sciences. National Institute of Oceanography
Goa: 85-104.
Ahmed, S., 1937. Working plan for the Tanjore Forest Divi Venkatesan, K. R., 1966. The mangroves of M adras state.
sion Revenue Department, District Board of Tanjore, Indian Forester 92: 27-34.
82 pp.
m m & k
w w e w iii
L.a1,^uL j-i
WWNPWFWR
. ¿ 4 J XjjJ'.fiï'ïi'lî^â,
lifîillW friiA«1<iái ip i
^ swmfwmwytf-""iwm
■
Hydrobiologia 2 4 7 : 2 6 1 - 2 6 6 , 1992
V. Jaccarini & E. M artens (eds.), The Ecology o f Mangrove and Related Ecosystems. 261
Index