You are on page 1of 16

P r o c e e d i n g s of t h e Third I n t e r n a t i o n a l Symposium o n L i t t o r i n i d B i o l o g y , 9-24 (1992)

J. G r a h a m e , P . J . M i l l & D . G. R e i d (Eds)
C o p y r i g h t - The M a l a c o l o g i c a l Society of L o n d o n 1 9 9 2

T H E P H Y L O G E N E T I C RELATIONSHIPS OF T E N A T L A N T I C
LITTORINIDS ASSESSED BY A L L O Z Y M E ELECTROPHORESIS

THIERRY B A C K E L J A U and THIERRY W A R M O E S


1 2

^Koninklijk B e l g i s c h I n s t i t t i u t v o o r N a t u u r w e t e n s c h a p p e n , Afdeling Malacologie, Vautierstraat 29, B-1040


B r u s s e l , B e l g i u m . ^ U n i v e r s i t e i t A n t w e r p e n (RUCA), L a b o r a t o r i u m v o o r Algemene Dierkunde,
Groenenborgerlaan 171, B-2020 Antwerpen, Belgium.

ABSTRACT bias the interpretation of the results, the


generic name L i t t o r i n a will be used in its
Ten Atlantic littorinid species were surveyed widest sense throughout this paper.
electrophoretically at nine enzyme loci in order to Until now, the classification and
test some phylogenetic hypotheses. Three measures phylogenetic relationships of the Littorinidae
of genetic distance were assessed (Nei, Cavalli- were inferred from shell characters, radula,
Sforza & Edwards and Prevosti) and two tree-making operculum and anatomical features, mainly of
methods were compared (UPGMA and distance the reproductive system. Such data were used
Wagner). It is shown that tree topologies depended by Reid (1986, 1989, 1990), who constructed a
on the tree-making method, the genetic distance most parsimonious cladogram, which served as
measure and the rooting procedure, but not on the basis for the (infra)generic division of the
order of data input. Because of the presumed family. However, a cladogram is an
evolutionary rate heterogeneity among the taxa hypothesis of phylogenetic relationships, which
investigated, it is suggested that a rate-independent must be tested by analysing complementary
tree-making method, combined with outgroup data sets (Reid, 1989, 1990).
rooting, should give a more reliable phylogeny Electrophoretic comparisons of allozymes in
estimate. Trees obtained in this way confirm the several littorinid species have supported some
clustering of L i t t o r i n a s t r i a t a with L . l i t t o r e a , L . of Reid's (1989, 1990) phylogenetic
s a x a t i l i s s . l . and L . o b t u s a t a s . l . They also support interpretations. This was the case for the
the currently accepted relationships between the last relationships between L . angulifera and the L .
three species, even though they cluster L . a r c a n a ziczac species complex (Janson, 1985), as well
with L . n i g r o l i n e a t a , which is at variance with the as for the relationships between L . n e r i t o i d e s ,
results of previous electrophoretic analyses. In L . s a x a t i l i s s . l . , L . o b t u s a t a s . l . and L . l i t t o r e a
opposition to the L i t t o r i n a species, there is a (Warmoes, 1986; Ward, 1990).
bifurcation containing N o d i l i t t o r i n a and L i t t o r a r i a . In a similar way, i.e. by using
Finally, the distant relationship of M e l a r h a p h e electrophoretic allozyme comparisons, the
n e r i t o i d e s to the L i t t o r i n a species is confirmed. present work aims at providing additional data
on the phylogenetic relationships between 10
Atlantic littorinids (Table 1).
INTRODUCTION
The systematics of several Atlantic littorinid MATERIALS AND METHODS
species are still much debated. This is for
example the case for the generic position of A total of 257 specimens from 22 populations
L i t t o r i n a n e r i t o i d e s (Linnaeus, 1758), L . s t r i a t a representing 10 species were analysed by means of
King & Broderip, 1832, L . p u n c t a t a (Gmelin, vertical polyacrylamide gel electrophoresis (PAGE).
1791) and L . cingulifera Durtker, 1845, as well A sample and species list is presented in Table 2.
as for the subgeneric relationships of L . Specimens were transported alive to the laboratory
s a x a t i l i s (Olivi, 1792) s . l . and L . o b t u s a t a where they were stored at -20°C or at -80°C.
(Linnaeus, 1758) s . l . A summary of the most Individual body homogenates were prepared by
important alternative opinions on these subjects thawing frozen animals, crushing their shells and
is given in Table 1. Actually, only one of the dissecting the soft parts in distilled water. Usually,
species listed has never provoked systematic shell fragments, penes, radulae, opercula and
controversy, viz. L . l i t t o r e a (Linnaeus, 1758), sometimes complete oviducts were separated and put
the type species of the genus. In order not to into 70% ethanol for storage as reference material.
10 T. B A C K E U A U & T. WARMOES

Table 1. Comparison of different viewpoints on the systematics of some Atlantic littorinid


species. References: (1) Bequaert (1943); (2) Fischer (1967); (3) Rosewater (1970); (4)
Sacchi (1974); (5) Bandel (1974); (6) Rosewater (1981); (7) Bandel & Kadolsky (1986); (8)
Reid (1989). *indicates species complex.

Ref. striata punctata neritoides cingulifera saxatilis* obtusata* littorea

1 _ _ Littorina Littorina Littorina Littorina


- - (Melarhaphe) (Littorivaga) (Neritrema) (Littorina)

2 Littorina _ Littorina _

(Melarhaphe) - (Melarhaphe) - - - -

3 Nodilittorina Littorina Littorina Littorina Littorina Littorina Littorina


(Granulilittorina)(Austrolittorina) (Melarhaphe) (Littoraria) (Littorina) (Littorina) (Littorina)

4 Littorina Littorina Littorina _ Littorina Littorina Littorina


(Melarhaphe) (Melarhaphe) (Melarhaphe) - (-) (-) (-)

5 Littorina Littorina Littorina - Littorina Littorina Littorina


(-) (-) (Melarhaphe) - (Neritrema?) (Neritrema?) (-)

6 Nodilittorina Littorina Littorina Littorina Littorina _ _

(Liralittorina) (Austrolittorina) (Melarhaphe) (Littoraria) (Littorina) - -

7 Nodilittorina Nodilittorina Melarhaphe - Littorina Littorina Littorina


(Liralittorina) , (Nodilittorina) (Melarhaphe) - (-) (-) (-)

8 Littorina Nodilittorina Melarhaphe Littoraria Littorina Littorina Littorina


(Liralittorina) (Echinolittorina) (-) (Littoraria) (Neritrema) (Neritrema) (Littorina)

All remaining body parts were rinsed in distilled 240 min with the TBE buffer. The temperature was
water, blotted on Kleenex paper, weighed and kept below 5 °C.
transferred into a 20% (w/v) aqueous sucrose Enzyme staining recipes were adapted from
solution (5/tl solution per mg tissue, except for L . Nyman (1970), Shaw & Prasad (1970) and Harris &
n e r i t o i d e s for which the ratio was 8(tl solution per Hopkinson (1976). Alleles (electromorphs) were
mg tissue). designated alphabetically according to decreasing
Crude homogenates were centrifuged for 45 min at electrophoretic mobilities (A = fastest allele = the
27200g (15000 r.p.m.) at 4°C. The supernatant was band with the most anodal position). The BIOSYS-1
stored at -80°C until required for electrophoresis. computer package of Swofford & Selander (1981)
PAGE was performed in 80x80x2.7 mm gel slabs was used to analyse the data. For reasons outlined
with a gel concentration of 6% (Backeljau, 1989). later, we pooled all conspecific individuals in single
To facilitate comparisons between gels, previously 'populations' (10 gene pools). Allele frequencies
typed specimens were included in each run. were calculated and used to derive Nei's (1978)
Electrophoreses were conducted using two unbiased genetic identities (I) and distances (D),
continuous buffer systems: Tris/Citric acid (TC) at Cavalli-Sforza & Edwards' (1967) chord distance
pH 8.0 and Tris/Boric acid/EDTA (TBE) at pH 8.9. and Wright's (1978) Prevosti distance. 1 values and
TC buffer was employed to resolve acid phosphatase Prevosti distances were employed to construct
(ACP, E . C . 3.1.3.2), fumarase ( F U M , E . C . U P G M A dendrograms. Distance Wagner trees
4.2.1.2), glucose-6-phosphate isomerase (GPI, E . C . (Farris, 1972) were derived using the chord and
5.3.1.9), glutamic-pyruvate transaminase (GPT, Prevosti metrics.
E.C. 2.6.1.2), malate dehydrogenase (MDH, E . C . Wagner trees were constructed using different
1.1.1.37), malic enzyme (ME, E . C . 1.1.1.40), orders of data input and rooting methods, because
peroxidase (PER, E . C . 1.11.1.7) and 6- these factors may affect the branching sequence of
phosphogluconate dehydrogenase (PGD, E . C . such trees (Swofford, 1981; Emberton, 1988).
1.1.44). T B E buffer was used for superoxide For each species we also provide rough estimates
dismutase (SOD, E.C. 1.15.1.1). of percentages of polymorphic loci (P), mean
Electrophoresis was started at 25V for the first 15 numbers of alleles per locus (A), mean observed
min, increased to 50V for the following 15 min and heterozygosities ( H ) and mean Nei's (1978)
0

then 150V for either 180 min with the TC buffer or unbiased expected heterozygosities ( H ) . Wee
PHYLOGENETIC RELATIONSHIPS OF LITTORINIDS 11

Table 2. Locality data on the species and populations studied. NT = total number of specimens;
N M = number of males; N F = number of females; NJ = number of juveniles; T(°C) =
temperature at which samples were kept.

Species Abbr. Locality Date NT NM . NF NJ T(°C)

littorea li Breskens, the Netherlands 19/03/1989 10 4 5 1 > -80

Dieppe, France 08/04/1989 10 5 5 - -80

saxatilis sa Breskens, the Netherlands 19/03/1989 10 2 8 - -80

Boulogne-sur-Mer, France 31/07/1988 10 3 7 - -20

Trebeurden, France 04/04/1990 1 - 1 - -80

arcana ar Roscoff, France 10/04/1986 10 4 6 - -20

Trebeurden, France 104/04/1990 17 - 17 - -80

nigrolineata i ni Roscoff, France 10/04/1986 10 5 5 - -20

Trebeurden, France 03/04/1990 18 6 12 - -80

obtusata ob Skarberget, Norway 21/07/1988 10 7 3 - -20

Roscoff, France 10/04/1986 10 5 4 1 -20

mariae ma Zeebrugge Belgium 07/10/1989 10 1 9 - -80

Narvik, Norway 19/07/1988 12 3 9 - -20

striata St Santa Cruz, Flores, Azores 10/07/1989 20 5 15 - -80

Anjos, Santa Maria, Azores 14/06/1990 20 9 11 - -80

Formigas, Azores 09/06/1990 11 2 9 - -80

punctata pu Brucoli, Sicily 31/05/1990 25 6 18 1 -80

Goree, Senegal 31/07/1990 11 3 8 - -80

Dakar, Senegal 02/08/1990 4 - 4 - -80

cingulifera ci Diembereng, Senegal 24/07/1990 10 1 9 - -80

neritoides ne San Laurenco, Santa Maria, Azores 25/06/1990 10 1 8 1 -80

Boca de Ribeirinha, Sao Miguel, 06/06/1990 9 7 2 -80


Azores
Brucoli, Sicily 31/05/1990 10 2 7 1 -80

considered a locus as being polymorphic when more eight loci).


than one allele was detected. However, for The protocols and photographs of the
comparative purposes we also mention the P values electrophoretic runs, and the remaining parts of the
with the 0.95 criterion. Because scoring A C P specimens studied, are deposited in the collections of
zymograms was sometimes difficult, all analyses the Koninklijk Belgisch Instituut voor
were conducted for two data sets: one including ACP Natuurwetenschappen (KBIN), Brussels.
(total of nine loci) and one excluding ACP (total of
12 T. B A C K E L J A U & T. WARMOES

Table 3. Overall allele frequencies at 9 enzyme loci in 10 littorinid species. Names are
abbreviated as in Table 2.
Locus St li sa ar ni ma ob pu ci ne

SOD(N) 44 18 20 23 25 4 14 29 10 14
A - 0.500 - - - - - - - -
B - - 0.450 0.239 0.140 - 0.500 - _ -
C
D
-
-
0.472
0.028
-
0.550
-
0.587
-
0.640
1.000
-
0.036
0.464
-
-
-
-
--
E - - - 0.174 0.220 - -- - - -
F - - - - - - - 1.000 -
G
H
1.000
-
-
-
-
- -
- -
-
-
- -
- 1.000
-
-
-
-
1.000
PGD(N) 51 14 12 17 19 21 15 36 10 26
A - - - - - - - -- - 0.346
B - - - - - - - - 0.654
C 0.961 -- - - _ - - - -- _

D 0.039 - - - - - -
E - - - 0.088 - - - 0.014 - -
F - 0.429 0.667 0.176 - - 0.067 - -
G
H
-
-
0.286
0.286
0.333
-
0.735
-
1.000
-
1.000
-
0.933
-
0.944
0.042
0.100
0.900
--
GPI(N) 50 20 18 25 28 22 19 36 10 26
A - 1.000 - - - - - - - -
B - - - - -- - - - 0.050 -
C - - - -- - - 0.900 -
D - - - - - - - 0.050 -
E - - - - - - 0.053 - - -
F
G
-
-
-
-
0.167
0.639
0.040
0.920
-
0.982
-
0.773
0.105
0.632
_

-
_

- -
H
I
0.010
-
-
-
-
0.167
-
0.040
_

0.018
-
0.227
_

0.211
-
_
_ _

J 0.250 - - - - - _ _ - -
K
L
-
0.010
-
-
-
-
-
- -
- - - 1.000
-
_
_
0.019
_

M
N
0.010
0.710
-
-
-
-
-
_ -
-
-
_
_
-- _

-
0.058
-
O - - - - - - - - - 0.404
P - - - - - - - - - 0.077
Q 0.010 - - - - - - - - 0.423
R - - - - - - - -- - 0.019
s - - 0.028 - - - - - -
PER(N) 47 20 12 15 17 21 10 24 10 12
A - - - - - 0.167 0.050 - - -
B - - 0.208 - - 0.833 0.950 - - -
C - 0.425 0.125 0.100 0.088 _ _ _ - -
D 0.500 . 0.575 0.542 0.900 0.912 - _ - -
E - - 0.125 - - - - - - -
F 0.500 - - - - -- - 1.000 1.000 0.167
G - - - - - -- -- - 0.500
H
I
-
-
-
-
-
-
-
- -
-
-
- -
-
-
0.083
0.208
J - - - - - - - - - 0.042
GPT(N) 42 13 9 17 18 4 6 35 10 26
A - - 0.222 - 1.000 - - - -
B 0.095 1.000 0.556 0.206 -- 1.000 1.000 - -
C
D
0.881
0.024
-
-
-
0.111
-
0.147 -
-
-
-- 0.014
0.957
0.100
0.900
-
-
E - - - - - -- - 0.029 - -
F - - - 0.529 - - - - -
G - - 0.111 0.118 - - - - - -
H - - - - - - - - - 1.000
PHYLOGENETIC RELATIONSHIPS OF LITTORINIDS 13

Table 3. (cont.)
Locus St li sa ar ni ma ob pu ci ne

MDH(N) 51 19 17 17 21 21 14 36 10 14
A 0.010 - - - - - - - - -
B _ _ _ _ _ - 0.036 - - -
C - - - - - - - - 1.000 -
D - - 0.294 - _ - - - - -
E 0.990 _ _ _
- - - -
F - - _ - - - 1.000 - 1.000
G - - 0.706 1.000 1.000 - - - - -
H - - - - _ 1.000 0.964 - - -
I - 0.105 - - - - - - - -
J - 0.895 - - - - - - - -
FUM(N) 44 17 11 18 18 8 7 35 10 26
A - - - - - - - - 0.500 -
B - 1.000 - - - - - - - -
C - - - - - - - - 0.300 -
D - - 0.273 1.000 1.000 - - - - -
E - - 0.364 - - 0.500 0.429 - - -
F 0.068 - - - - - - - - -
G 0.284 - 0.364 - - 0.500 0.571 - - -
H _ - _ - - - - 0.200 -
I 0.602 - - - - - - - - 0.058
J 0.023 - - - - - - - -
K 0.023 - - _ - - - - - 0.500
L - - - - - - - 0.986 - 0.404
M - - - - - - - 0.014 - 0.038
ME(N) 39 18 14 21 19 21 18 35 10 13
A - 0.944 - - - - - - - • -
B - 0.056 - - - - - - - -
C - - - - 0.048 0.306 - - -
D - - - - - 0.048 - - - -
E - - 0.036 0.095 0.053 0.405 0.694 - - -
F - - - - - 0.500 - - - -
G 1.000 - - - - - - - - 1.000
H - - 0.393 0.143 - - - - 0.400 -
I - - 0.571 0.762 0.895 - - - 0.600 -
J - - - - 0.053 - - - - -
K - - - - - - - 0.929 - -
L - - - - - - - 0.071 - -

ACP(N) 40 14 8 19 23 21 18 30 10 13
A - - 0.625 0.053 - 0.810 0.111 - - -
B - - - - 0.109 0.048 - - - -
C 0.100 - - 0.158 0.435 - - - - 0.038
D - - 0.313 0.158 0.457 0.143 0.889 - - -
E 0.375 - 0.579 - - - - - 0.808
F 0.475 - 0.063 0.053 - - - - - 0.154
G 0.050 - - - - - - - - -
H - - - - - - - 0.100 - -
I - 0.536 - - - - - 0.867 - -
J - 0.464 - - - 0.033 - -
K - - - - - - - - 1.000 -

RESULTS The P and He values we observed are


generally larger than previously reported for
The enzymes screened were assumed to be the same species (no comparative data
coded by nine putative loci. Allele frequencies available for L . cingulifera and L . s t r i a t a ) .
at these loci are given in Table 3. The The only exceptions to this observation are L .
measures of genetic variability assessed are p u n c t a t a and partly L . n e r i t o i d e s (see review
presented in Table 4, which shows that these by Ward, 1990). However, the discrepancies
measures did not depend strongly on whether between He and Ho, are not conspicuously
ACP was included or not. large, especially not if one considers the
14 T. B A C K E U A U & T. WARMOES

Table 4. Summary of genetic variability measures in 10 Atlantic


littorinid species. Upper figures refer to the analysis of 9 loci
(including ACP), the lower ones (in parentheses) to the analysis
without ACP. N = mean sample size per locus; A = mean
number of allele^ per locus; P = percentage polymorphic loci using
no criterion; P = percentage polymorphic loci using the 0.95
criterion; Ho = mean observed heterozygosity (direct count); He =
mean unbiased expected heterozygosity under Hardy-Weinberg
conditions.

Species N A P P* H 0
H e

striata 45.3 2.9 77.8 55.5 0.211 0.271


(46.0) (2.8) (75.0) (50.0) (0.232) (0.227)
littorea 17.0 1.9 66.7 66.7 0.253 0.282
(17.4) (1.9) (62.5) (62.5) (0.275) (0.253)
saxatilis 13.4 3.0 100.0 100.0 0.448 0.559
(14.1) (3.0) (100.0) (100.0) (0.488) (0.562)
arcana 19.1 2.8 77.8 77.8 0.156 0.338
(19.1) (2.5) (75.0) (75.0) (0.176) (0.302)
nigrolineata 20.9 1.9 55.6 44.4 0.064 0.171
(20.6) (1.8) (50.0) (37.5) (0.062) (0.117)
mariae 15.9 1.9 55.6 55.6 0.183 0.234
(15.3) (1.8) (50.0) (50.0) (0.206) (0.222)
obtusata 13.4 2.2 88.9 77.8 0.315 0.286
(12.9) (2.3) (87.5) (75.0) (0.340) (0.297)
punctata 32.9 1.9 55.6 33.3 0.035 0.066
(33.3) (1.8) (50.0) (25.0) (0.039) (0.044)
cingulifera 10.0 1.8 55.6 55.6 0.111 0.192
(10.0) (1.9) (62.5) (62.5) (0.125) (0.216)
neritoides 18.9 2.7 55.6 55.6 0.224 0.306
(19.6) (2.6) (50.0) (50.0) (0.243) (0.302)

pooling procedure we employed. our study, these two species revealed the
Tables 5-7 present different measures of smallest genetic distance (D=0.149) (Table 5).
genetic 'relatedness' between the species The genetic distances between L . o b t u s a t a
investigated. Apparently also these data are and L . m a r i a e , v i z . D=0.231 (+ACP) and
only slightly influenced by whether or not ACP D=0.156 (-ACP) (Table 5), are smaller than
was considered. The interspecific genetic the value of Moyse et a l . (1982) (D=0.501),
distances we obtained are comparatively large but larger than the value of Morris (1979)
for littorinids. For example, the mean genetic (D=0.087).
distance between species of the L . s a x a t i l i s All other genetic distances between L .
complex is D=0.275 (+ACP) or D=0.223 (- l i t t o r e a , L . s a x a t i l i s s . l . and L . o b t u s a t a s . l .
ACP) (Table 7), while the largest (single) D (Tables 5 and 7) are larger than the D values
value Ward (1990) reported in this species mentioned by Ward (1990), while they are of
aggregate was D=0.237 (L. s a x a t i l i s l L . the same magnitude as the genetic distances
n i g r o l i n e a t a ) . The largest mean distance Ward between L . angulifera and L . ziczac s . l .
(1990) observed among species in L . s a x a t i l i s (Janson, 1985).
s.l. was between L . a r c a n a and L . n i g r o l i n e a t a As there exist no comparative data with
with D=0.140 (range: 0.052-0.227). Yet, in respect to interspecific genetic distances
PHYLOGENETIC RELATIONSHIPS OF LITTORINIDS 15

Table 5. Nei's (1978) unbiased genetic identities (above diagonal) and distances (below
diagonal) between 10 Atlantic littorinid species. Upper figures are based on 9 loci (including
ACP), the lower ones (in parentheses) were obtained after excluding ACP (8 loci). Species
names are abbreviated as in Table 2. * = infinity.

Species St li sa ar ni ma ob pu ci ne

striata - 0.059 0.090 0.117 0.071 0.035 0.040 0.207 0.088 0.236
(0.063) (0.092) (0.080) (0.069) (0.038) (0.044) (0.223) (0.098) (0.193)

littorea 2.834 - 0.260 0.172 0.125 0.263 0.206 0.103 0.042 0.000
(2.766) (0.287) (0.185) (0.133) (0.288) (0.228) (0.042) (0.047) (0.000)

saxatilis 2.407 1.347 - 0.740 0.688 0.482 0.551 0.073 0.118 0.002
(2.383) (1.247) (0.793) (0.723) (0.422) (0.548) (0.081) (0.135) (0.000)

arcana 2.146 1.760 0.301 - 0.861 0.275 0.340 0.118 0.109 0.079
(2.520) (1.689) (0.231) (0.892) (0.288) (0.349) (0.128) (0.122) (0.000)

nigrolin- 2.640 2.083 0.375 0.149 - 0.259 0.342 0.119 0.086 0.002
eata (2.674) (2.017) (0.325) (0.114) (0.269) (0.311) (0.128) (0.096) (0.000)

mariae 3.344 1.334 0.729 1.290 1.353 - 0.794 0.124 0.014 0.000
(3.264) (1.244) (0.863) (1.244) (1.313) (0.856) (0.137) (0.016) (0.000)

obtusata 3.227 1.582 0.596 1.079 1.073 0.231 - 0.120 0.014 0.000
(3.131) (1.476) (0.602) (1.053) (1.168) (0.156) (0.134) (0.016) (0.000)

punctata 1.576 2.270 2.618 2.135 2.126 2.087 2.120 - 0.255 0.219
(1.500) (3.178) (2.509) (2.056) (2.052) (1.989) (2.007) (0.288) (0.242)

cingulifera 2.427 3.178 2.137 2.213 2.448 4.260 4.293 1.366 - 0.025
(2.324) (3.065) (2.002) (2.106) (2.347) (4.135) (4.153) (1.244) (0.028)

neritoides 1.442 + 6.249 2.539 6.012 * * 1.520 3.700 -


(1.645) (*) (*) (•) (*) (*) (*) (1.417) (3.570)

involving L . s t r i a t a , L . p u n c t a t a , L . cingulifera subclusters: one involving L . p u n c t a t a and L .


and L . n e r i t o i d e s , we can only note that they cingulifera and one composed of L . s t r i a t a and
yield very large D values both in comparisons L . neritoides.
with true L i t t o r i n a species, as in comparisons The Wagner trees we constructed are
amongst each other (Table 5). presented in Figs 2-4. The data from which
The UPGMA dendrograms inferred from they were inferred are given in Table 6. The
Nei's (1978) unbiased genetic identity and topologies obtained did not depend on the in-
Wright's Prevosti distance are presented in or exclusion of ACP. Fig. 2 shows that the
Fig. 1. Both measures yielded the same the branching sequence of Wagner trees is affected
branching order, while the in- or exclusion of by the choice of the genetic distance measures.
ACP did not affect the tree topology. It is also clear that the two Wagner trees in
The UPGMA dendrograms reveal three Fig. 2 are at variance with the UPGMA trees
notable features (Fig. 1): in Fig. 1.
1) L . l i t t o r e a , L . s a x a t i l i s s . l . and L . o b t u s a t a The effect of four different data input
s . l . form a single cluster in which three combinations upon the topology of a Wagner
subclusters can be distinguished, each tree was tested. Yet, each data input order
coinciding with a species or species complex. yielded the same tree topology (Fig. 3).
2) Within the L . s a x a t i l i s complex, L . a r c a n a is In Figs 2 and 4, finally, it is shown that both
clustered with L . n i g r o l i n e a t a , while L . the rooting method and the choice of outgroup
s a x a t i l i s s.s. appears to be somewhat less affect the branching sequences of Wagner
closely related. trees.
3) In opposition to the cluster containing L . By using midpoint rooting (no outgroup), we
l i t t o r e a , L . s a x a t i l i s s . l . and L . o b t u s a t a s . l . , obtained a tree, whose topology largely
there is a cluster consisting of two bifurcated corresponds to the branching pattern obtained
16 T. B A C K E L J A U & T. WARMOES

Table 6. Prevosti distance (above diagonal) and Cavalli-Sforza & Edwards' chord distance
(below diagonal) between 10 Atlantic littorinid species. Upper figures are based on 9 loci
(including ACP), the lower ones (in parentheses) were obtained after excluding A C P . Species
names are abbreviated as in Table 2.

Species St li sa ar ni ma ob pu ci ne

striata - 0.934 0.893 0.873 0.933 0.958 0.958 0.829 0.931 0.796
(0.926) (0.887) (0.923) (0.938) (0.953) (0.953) (0.808) (0.922) (0.842)

littorea 0.857 - 0.782 0.848 0.891 0.805 0.843 0.900 0.957 1.000
(0.851) (0.754) (0.829) (0.878) (0.780) (0.823) (0.959) (0.952) (1.000)

saxatilis 0.833 0.765 - 0.436 0.508 0.618 0.537 0.951 0.869 0.993
(0.835) (0.746) (0.398) (0.485) (0.667) (0.532) (0.944) (0.853) (1.000)

arcana 0.815 0.780 0.442 - 0.256 0.773 0.697 0.900 0.890 0.926
(0.850) (0.764) (0.406) (0.203) (0.769) (0.685) (0.888) (0.876) (1.000)

nigrolin- 0.855 0.817 0.516 0.397 - 0.774 0.700 0.895 0.922 0.996
eata (0.861) (0.806) (0.486) (0.360) (0.770) (0.720) (0.882) (0.913) (1.000)

mariae 0.865 0.781 0.634 0.747 0.771 - 0.301 0.895 0.989 1.000
(0.861) (0.765) (0.667) (0.750) (0.774) (0.245) (0.882) (0.988) (1.000)

obtusata 0.864 0.798 0.546 0.679 0.707 0.387 - 0.896 0.989 1.000
(0.859) (0.784) (0.560) (0.680) (0.725) (0.365) (0.883) (0.988) (1.000)

punctata 0.796 0.825 0.855 0.836 0.850 0:850 0.852 - 0.772 0.822
(0.782) (0.864) (0.849) (0.827) (0.844) (0.844) (0.846) (0.743) (0.799)

cingulifera 0.841 0.866 0.823 0.819 0.846 0.884 0.885 0.767 - 0.981
(0.833) (0.861) (0.813) (0.808) (0.839) (0.882) (0.883) (0.749) (0.979)

neritoides 0.762 0.900 0.895 0.857 0.894 0.900 0.900 0.783 0.880 -
(0.801) (0.900) (0.900) (0.900) (0.900) (0.900) (0.900) (0.767) (0.877)

Table 7. Mean genetic distances (Nei, 1978)


between L i t t o r i n a species as inferred from Table 5.
Species names are abbreviated as in Table 2.

Species comparison D(+ACP) D(-ACP)

among sa-ar-ni 0.275 0.223

among ob-ma 0.231 0.156

between l i and sa-ar-ni 1.730 1.651

between l i and ob-ma 1.458 1.360

between sa-ar-ni and ob-ma 1.020 1.041

with the U P G M A procedure. Only the (Fig. 2A), whereas in the U P G M A trees they
positions of L . striata and L . neritoides differ form a single cluster (Fig. 1).
in the two tree topologies, for in the Wagner The outgroup method revealed a different
tree these species are separated by two nodes picture. Using L . punctata+L. cingulifera as
PHYLOGENETIC RELATIONSHIPS OF LITTORINIDS 17

0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0 90 100


-. , 1 1 , 1 r i 1 t r n 1 1

si
ne
pu

li
sa
ar
ni
ma
ob

1.00 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10

si
ne
pu
ci
li
sa
ar
ni
ma
ob

Figure 1. U P G M A trees for 10 Atlantic littorinid species. A : using Nei's (1978) unbiased genetic identity
and including A C P ; B: using Wright's (1978) Prevosti distance (ACP included). Species names are
abbreviated as in Table 2.

DISCUSSION
outgroup we obtained a Wagner tree in which
the clustering of L . l i t t o r e a , L . s a x a t i l i s s . l . and Methodology ofphylogeny inference
L . o b t u s a t a s . l . is confirmed. This cluster is
joined with L . s t r i a t a , while the resulting
subtree is in turn connected with L . n e r i t o i d e s . The merits and drawbacks of various methods
Hence, in this case L . s t r i a t a and L . n e r i t o i d e s for phylogenetic tree construction from
are separated by a single node (Fig. 4A). The molecular data have been discussed and
same topology is obtained when L . p u n c t a t a compared excessively in the literature (e.g.
and L . cingulifera are used separately as Swofford, 1981; Tateno et a l , 1982; Nei et
outgroups. In either case, the remaining a t . , 1983; Tateno & Tajima, 1986; Patton &
species of the couple is placed near the root. Avise, 1983; Fitch & Atchley, 1987; Nei,
However, if L . n e r i t o i d e s is the outgroup, this 1987; Hartl et a l . , 1990; Sundberg et a l . ,
species is separated from L . s t r i a t a by two 1990; Swofford & Olsen, 1990). Yet, the
nodes, for now the L . p u n c t a t a + L . cingulifera conclusions of these studies are sometimes
cluster becomes inserted between L . n e r i t o i d e s contradictory with respect to the reliability of
and L . s t r i a t a (Fig. 4B). Thus, this tree the different algorithms or procedures.
topology contains two clusters: one One of the issues provoking much debate is
encompassing L . p u n c t a t a and L . c i n g u l i f e r a , whether a tree should be constructed under the
the other one involving L . s t r i a t a , L . l i t t o r e a , assumption of (nearly) equal evolutionary rates
L . s a x a t i l i s s . l . and L . o b t u s a t a s . l . in all branches (e.g. UPGMA) or whether it
Note that in all trees the configuration of L . should allow for a non-constant evolutionary
s a x a t i l i s s . l . and L . o b t u s a t a s . l . did not clock (e.g. distance Wagner algorithm,
change. maximum likelihood methods). Although
18 T. B A C K E L J A U & T. WARMOES

0 0.07 0.13 0.20 0.27 0.34 0.40 0.47 0.54 0.60 0.67

i f—i 1 — "i 1 1 1 1 1 r
pu

ob

0.06 0.11 0.17 0.23 0.28 0.34 0.40 0.45 0.51 0.57
I 1 1 1 1 1 1 1 1 1 i 1 1 1 1 1 1 1 1

pu

ob

Figure 2. Distance Wagner trees for 10 Atlantic littorinid species. A : using Wright's (1978) Prevosti
distance; B: using Cavalli-Sforza & Edwards' (1967) chord distance. Both trees were rooted by midpoint
rooting. ACP was included. Data input order: pu-ci-li-sa-ar-ni-ma-ob-st-ne. Species names are abbreviated
as in Table 2.

phylogenetic relationships inferred from Nei, 1987). This is particularly true when
allozyme data are usually summarized in genetic distances are small, as is the case in the
UPGMA dendrograms, Sundberg et a l . (1990) study of Sundberg et a l . (1990). Nevertheless,
concluded that at least for populations of L . Nei (1987) also ascertained that when distance
s a x a t i l i s s.s. and L . a r c a n a the most reliable estimates are subject to large stochastic errors,
trees are obtained by using algorithms which UPGMA is often superior to other tree making
do not rely on a constant evolutionary clock. methods based on distance matrices. However,
The reason for the poor performance of if rates vary greatly along branches, the
UPGMA in the study of Sundberg et a l . (1990) distance Wagner algorithm (or other rate-
may be twofold: (1) the assumption of rate independent procedures) should perform better
homogeneity may be unrealistic, even at the (Nei, 1987).
level of closely related species such as L . Whether the assumption of rate homogeneity
s a x a t i l i s and L . a r c a n a , and (2) the number of is applicable to the study of Sundberg et a l .
loci surveyed (16) may have been too small. 1990) is difficult to decide. Fuerst & Ferrell
This latter constraint is expected to produce a 1983) suggested that interspecific
tree which is subject to a large stochastic error differentiation of electromorphs is not
(Nei et a l . , 1983; Nei, 1987). In general it is equivalent to intraspecific change. Similarly,
recommended that for phylogenetic inferences Eldredge (1985) and Crother (1990) claimed
a minimum of 30 loci should be surveyed in that evolutionary change operates on different
order to have a reasonable chance of time scales, which are evidenced by two
recovering a correct tree (Nei et a l . , 1983; patterns: a within-species pattern and an
PHYLOGENETIC RELATIONSHIPS OF LITTORINIDS 19

0 0.08 0.16 0.24 0.33 0.41 0.49 0.57 0.65 0.73 0.81

n 1 1 1 r -i 1 i 1
i r—i r st

ob
pu

Figure 3. Distance Wagner tree obtained with different data input combinations of 10 Atlantic littorinid
species, using Cavalli- Sforza & Edwards' (1967) chord distance. The tree was rooted with L . n e r i t o i d e s as
outgroup. A C P was included. The data input orders were: 1) st-ne-li-sa-ar-ni-ma-ob-pu-ci, 2) st-li-sa-ar-ni-
ma-ob-pu-ci-ne, 3) pu-ci-li-sa-ar-ni-ma-ob-st-ne and 4) li-sa-ar-ni-ma-ob-st-pu-ci-ne. Species names are
abbreviated as in Table 2.

0 0.08 0.16 0.24 0.32 0.40 0.49 0.57 0.65 0.73 0.81
l 1 1 r i 1 r 1 1 1 1 1 r ~ i i

ob
. st
ne
pu

0.09 0.18 0.26 0.35 0.44 0.53 0.62 0.70 0.79 0.88
i 1 1 1 1 1 1 1
-i n — i 1 r li

ob

pu

B
Figure 4. Effect of different rooting procedures upon distance Wagner trees for 10 Atlantic littorinid species,
using Wright's (1978) Prevosti distance. A C P was included. A : outgroup = pu-ci; B: outgroup = ne.
Order of data input: li-sa-ar-ni-ma-ob-st-pu-ci-ne. Species names are abbreviated as in Table 2. For the
effect of midpoint rooting, see Fig. 2A.
20 T. B A C K E L J A U & T. WARMOES

among-species pattern. Thus intra- and relationships between L . l i t t o r e a , L . s a x a t i l i s


interspecific genetic distances would not s . l . and L . o b t u s a t a s . l . , whereas the chord
measure the same process. If this is correct, distance does not (Fig. 2). Swofford &
then it is evident that in trees in which different Berlocher (1987) also argued for the
species are compared at the population level superiority of the Prevosti distance.
there must be an a p r i o r i rate heterogeneity. Since we preferred a rate-independent tree-
This was probably the case in the study of making algorithm, we also preferred an
Sundberg et a l . (1990) and as such it may outgroup method rather than midpoint rooting,
partially explain why U P G M A performed so for this latter assumes a constant rate of
poorly. evolutionary change (Nei, 1987).
The two constraints mentioned above are Two alternative tree-making procedures
also expected to hold for our study. We appeared inappropriate in the present context.
screened even fewer loci than Sundberg et a l . The maximum likelihood method of Felsenstein
(1990), while our trees consisted of both (1981) was not employed since its underlying
closely and distantly related species. Thus, in assumptions (Brownian motion model) are
this study it seems unrealistic to assume a unrealistic (Nei, 1987). Probably the
constant evolutionary clock, even if we pooled algorithm is only applicable for problems
conspecific populations to avoid problems of involving intraspecific populations or closely
rate heterogeneity due to simultaneously related species (Swofford & Berlocher, 1987).
assessing intra- and interspecific relationships. A Hennigian analysis of electrophoretic data
Our analyses confirm the observation of (e.g. Patton & Avise, 1983; Buth, 1984;
Sundberg et a l . (1990) that for a given data set Richardson et a l . , 1986; Emberton, 1988), was
rate-dependent (UPGMA) and rate-independent not performed in view of the drawbacks
(Wagner procedure) distance matrix algorithms pointed out by Swofford & Berlocher (1987)
may produce different trees (compare Figs 1 and Sundberg et a l . (1990). Particularly the
and 2). Because of the considerations outlined small number of loci we studied and our
above, we tentatively follow Sundberg et a l . relatively small sample sizes make a Hennigian
(1990) in preferring rate-independent approach suspect.
algorithms (in this case the Wagner procedure)
when inferring littorinid phylogenies. This Electrophoretic techniques
premise is also supported by the work of
Knight et a l . (1987), who applied the Wagner Most electrophoretic studies on littorinids have
algorithm to L . s a x a t i l i s s.s. and L . a r c a n a involved starch-gel electrophoresis, while only
and obtained in this way a much better tree six relied on PAGE (Wium-Andersen, 1970;
than with U P G M A . The Wagner procedure Vuilleumier & Matteo, 1972; Berger, 1973,
has furthermore been advocated by Swofford 1977; Caugant, 1979; Warmoes, 1986). Yet,
(1981), Buth (1984), Emberton (1988) and the detectability of protein differences does not
Swofford & Olsen (1990). only depend on characteristics of the proteins
As our work indicates that the topology of a themselves, but also on the electrophoretic
Wagner tree depends on the distance measure conditions employed (gels, buffers, pH, etc.)
and rooting method used (Figs 2 and 4), we (e.g. Coyne et a l . , 1979; Ramshaw et a l . ,
must evaluate these factors too. 1979; Singh, 1979; Beaumont & Beveridge,
We used three genetic distance measures, of 1983; Nei, 1987; Backeljau, 1989; Backeljau
which only two were employed to construct & DeBruyn, 1990; Murphy et a l . , 1990).
Wagner trees. Indeed, because Nei's (1978) Thus both the choice of proteins under study
unbiased estimate is non-metric, it is not (e.g. slow versus fast evolving loci), and the
appropriate for the Wagner procedure running conditions, may influence the genetic
(Swofford & Selander, 1981; Buth, 1984). data obtained. Heterozygosity estimates, for
Nevertheless, we included Nei's (1978) example, may vary considerably between
measure in Table 5, for it allows a direct different data sets for a given taxon (e.g.
comparison with the literature on littorinids, in inclusion or exclusion of esterases) (Simon &
which this is the only genetic distance measure Archie, 1985).
hitherto used. Three of the nine enzymes surveyed have not
Cavalli-Sforza & Edwards' (1967) chord been investigated previously in littorinids
distance and Wright's (1978) Prevosti distance (PER, GPT, FUM). At the six remaining loci,
are much less widespread throughout the our data confirm the numbers of alleles
literature, even though these metrics are highly reported in the literature and where they
appropriate for phylogenetic inferences deviate from previous studies, they usually
(Swofford & Selander, 1981; Felsenstein, reveal higher numbers of alleles. At GPI in L .
1984; Emberton, 1988; Swofford & Olsen, s a x a t i l i s , L . a r c a n a and L . m a r i a e , for
1990). In the present work, we prefer the example, we observed the same numbers of
Prevosti distance over the chord distance, since alleles as Snyder & Gooch (1973), Moyse et
the former recovers the generally accepted a l . (1982), Ward & Janson (1985), Knight et
PHYLOGENETIC RELATIONSHIPS OF LITTORINIDS 21

a l . (1987) and Sundberg et a l . (1990) to name L . s a x a t i l i s s . l . , L . o b t u s a t a s . l . and L . l i t t o r e a ,


a few. At the same locus in L . n e r i t o i d e s , L . as hitherto reported, are so small (Ward, 1990)
n i g r o l i n e a t a and L . o b t u s a t a , on the contrary, and because Janson (1985) noted realistic
we found more alleles than respectively Nevo "intergeneric" genetic distances among L .
& Lavie (1989), Beardmore & Morris (1978) angulifera and L . ziczac s . l . In retrospect, it
and Moyse et a l . (1982). For L . l i t t o r e a and seems as if our hypothesis was only partially
L . p u n c t a t a we observed fewer alleles than warranted, for the inclusion of L . n e r i t o i d e s
reported in the literature. Similar observations provoked infinite genetic distances (Table 5),
can be made for the other loci studied. indicating that the phylogenetic gap between
However, electrophoresis can only this species and the remaining ones may be too
demonstrate allozyme differences (Allendorf, large to be evaluated by protein
1977; Ferguson, 1980; Richardson et a l . , electrophoresis.
1986). Hence, when we found certain loci to Finally, the use of allele frequency data in
be polymorphic, while the same loci were phylogenetic reconstructions has been
monomorphic in starch-gel studies, it is questioned by Crother (1990), who argued that
possible that this is the result of the inability of allele frequencies are temporally unstable, as
the starch-gel technique to detect differences they are easily modified by random genetic
revealed by PAGE (and v i c e v e r s a ) . drift and/or selection (Mickevich & Johnson,
These observations could account for the 1976). Allele frequencies may even vary over
relatively larger P, H and D values (or lower I the span of a few years and therefore they
values) we generally detected. Note that the cannot by definition be synapomorphic
genetic identities we observed among sibling (Crother, 1990).
species of L . s a x a t i l i s s . l . (1=0.763) and L . Although this issue is far from resolved, it
o b t u s a t a s . l . (1=0.794) (Table 5) correspond must be emphasized that allele frequencies are
well to the data provided by Thorpe (1979, only a way to weigh the presence or absence of
1982, 1983). This author suggested that in a particular allele (Swofford & Olsen, 1990).
about 97% of interspecific comparisons Nei's We have tried to alleviate possible temporal
mean genetic identity between "good" animal (and geographical) instabilities of allele
species is lower than 1=0.850, whereas in 98% frequencies by pooling all conspecific
of intraspecific comparisons the mean identity individuals in single gene pools. Frequency
is higher than 1=0.850. Moreover, the genetic differences at population levels are thus
distances we obtained between species averaged, so that local and temporal
currently assigned to different subgenera or fluctuations are expected to have little effect on
genera, are of the same magnitude as the the overall frequency. In addition, we avoided
genetic distances Janson (1985) reported as much as possible the use of loci which are
between L . angulifera and the L . ziczac suspected to be under selective control (e.g.
complex, species which are now placed in esterases and aspartate aminotransferase)
different genera (Reid, 1989). (Newkirk & Doyle, 1979; Johannesson &
Thus, starch-gel electrophoresis and PAGE Johannesson, 1989), although we did include
may produce divergent population genetic and GPI (Nevo & Lavie, 1989).
phylogenetic data. Even among studies
involving only starch gel electrophoresis, there Systematic and p h y l o g e n e t i c considerations
may be large differences with respect to the H ,
P and D values reported. This is illustrated by Our results are largely in agreement with
the data of Noy et a l . (1987) and Ward (1990) recent revisions of littorinid classification and
for L . n e r i t o i d e s or by the genetic distances relationships (Warmoes, 1986; Reid, 1986,
Morris (1979) and Moyse et a l . (1982) 1989, 1990; Ward, 1990). Only the position of
reported between L . o b t u s a t a and L . m a r i a e L . a r c a n a and the (uncertain) branching of
(reviewed by Ward, 1990). N o d i l i t t o r i n a and L i t t o r a r i a are points of minor
The value of protein electrophoresis in disagreement.
assessing phylogenetic relationships is limited L . a r c a n a is, according to Ward (1990),
by the common observation that allozyme data most closely related to L . s a x a t i l i s , whereas L .
are only reliable at intra- and interspecific n i g r o l i n e a t a would be a somewhat more
levels. Usually, such data should no longer be distantly related species. However, the
relied on for taxa beyond the genus level ovoviviparous reproduction of L . s a x a t i l i s
(except in birds), because of the high incidence appears to be the only cladistically relevant
of allele similarities caused by chance 'morphological' character from which the
convergences (Ferguson, 1980; Richardson et relationships between the three species have
a l . , 1986; Murphy et a l . , 1990). been inferred. Yet, since ovoviviparity is an
Despite these limitations we expected that autapomorphy, it is consistent with any
allozyme comparisons of littorinid genera and topology. As a consequence, Reid (1990)
subgenera would yield reliable phylogenetic arrived at an incompletely resolved trichotomy
estimates, because the genetic distances among for the L . s a x a t i l i s group. This uncertainty
22 T. B A C K E L J A U & T. WARMOES

shows that the relationships within this species are identical, while those of L . s t r i a t a are
group need further investigation. different.
The branching sequence of L . l i t t o r e a , L . Note finally that our results also confirm the
s a x a t i l i s s . l . and L , o b t u s a t a s . l . in Wagner distant relationships between M e l a r h a p h e and
trees based on Prevosti distances (Fig. 4) L i t t o r i n a (Warmoes, 1986; Reid, 1989; Ward,
corroborate the trees of Reid (1990) and Ward 1990), as can be inferred from the extremely
(1990), who used respectively morphological large genetic distances involved (Table 5).
and allozyme data. Hence, Reid's (1989)
infrageneric classification of these species
seems to be a robust one. ACKNOWLEDGEMENTS
The clustering of L . p u n c t a t a and L .
cingulifera is inconsistent with the cladogram We are much indebted to E . Dumoulin
of Reid (1989), but consistent with the tree of (Knokke), J . M . Azevedo, C. Brito, R.T.
Reid (1986). However, the latter tree was Cunha, A . M . Frias Martins, (all of the
explicitly refuted by (Reid, 1989) because of Universidade dos Azores), G. Spada (Bologna)
the limited data set on which it was based and and A . Waren (Naturhistoriska Riksmuseet,
because of the apparent high incidence of Stockholm) for collecting parts of the material.
parallel evolution in the genera N o d i l i t t o r i n a We are furthermore grateful to W . N .
and L i t t o r a r i a (Bandel & Kadolsky, 1982; Verheyen and J . L J . Hulselmans (both
Reid, 1989). Both these points may have University of Antwerp) who provided the
provoked the clustering of L . p u n c t a t a and L . necessary working facilities. J. Van Goethem
cingulifera in the present work too. (KBIN, Brussels) and an anonymous referee
The U P G M A dendrograms and the Wagner are sincerely thanked for their comments on
trees rooted at the mid point of the greatest the manuscript. H . Van Paesschen (KBIN,
patristic distance, support Fischer's (1967) Brussels) kindly prepared the figures. Travel
opinion, according to which L . s t r i a t a would grants and financial support were provided by
be a M e l a r h a p h e . The Wagner trees rooted by the Belgian National Fund for Scientific
the outgroup method, on the contrary, cluster Research and the "Koninklijk Belgisch Instituut
L . s t r i a t a with the L i t t o r i n a group and thus voor Natuurwetenschappen, Brussel". The
confirm the branching sequence proposed by Italian populations were sampled during the
Reid (1989, 1990). For reasons outlined in the "Mission Sicile, 1990" organized by P.
first section we favour these latter results. Bouchet (Museum National d'Histoire
Note that the tree topologies remained Naturelle, Paris). The Azorean material was
essentially the same irrespective of whether L . collected during the missions "Flores 89" and
n e r i t o i d e s or L . p u n c t a t a + L . cingulifera were "Santa Maria e Formigas, 1990" organized by
used as outgroups. A.M. Frias Martins (Universidade dos
Our outgroup choice was based on the Azores).
cladograms of Reid (1986, 1989, 1990). In the
present context L . n e r i t o i d e s seems more
appropriate as outgroup than the L . REFERENCES
c i n g u l i f e r a + L . p u n c t a t a cluster (or each of
these species separately), for M e l a r h a p h e is the ALLENDORF, F.W. 1977. Electromorphs or
sistergroup of the clade containing the alleles. G e n e t i c s , 87: 821-822.
remaining Littorininae, including L i t t o r a r i a , BACKELJAU, T. 1989. Electrophoresis of albumen
N o d i l i t t o r i n a and L i t t o r i n a (Reid, 1989). gland proteins as a tool to elucidate taxonomic
Indeed, whatever outgroup was used, L . s t r i a t a problems in the genus A r i o n (Gastropoda,
was always clustered with L i t t o r i n a rather than Pulmonata). J o u r n a l of M e d i c a l a n d A p p l i e d
with M e l a r h a p h e . In addition, in none of the M a l a c o l o g y , 1: 29-41.
trees was L . s t r i a t a clustered with BACKELJAU, T. & D E BRUYN, L . 1990. On the
N o d i l i t t o r i n a . Thus we are fairly confident that infrageneric systematics of the genus A r i o n
the inclusion of this species in N o d i l i t t o r i n a Ferussac, 1819 (Mollusca, Pulmonata). B u l l e t i n
cannot be maintained. Hence, our data support de I ' I n s t i t u t R o y a l des Sciences N a t u r e l l e s de
the transfer of L . s t r i a t a from M e l a r h a p h e or B e l g i q u e , B i o l o g i e , 60: 35-68.
N o d i l i t t o r i n a (Table 1) to L i t t o r i n a (Reid, BANDEL, K . 1974. Studies on Littorinidae from the
1989, 1990). Nevertheless, L . s t r i a t a appears Atlantic. V e l i g e r , 17: 92-114.
to be rather distantly related to the other BANDEL, K . & KADOSKY, D. 1982. Western
members of this genus, and is therefore Atlantic species of N o d i l i t t o r i n a (Gastropoda:
appropriately placed in a monotypic subgenus Prosobranchia): comparative morphology and its
L i r a l i t t o r i n a (Reid, 1989, 1990). This is functional, ecological, phylogenetic and
supported by Wium-Andersen (1970), who taxonomic implications. V e l i g e r , 25: 1-42.
noted that the electrophoretic profiles of BEARDMORE, J.A. & MORRIS, S.R. 1978.
haemoglobin in L . l i t t o r e a and L . s a x a t i l i s s . l . Genetic variation and species coexistence in
PHYLOGENETIC RELATIONSHIPS OF LITTORINIDS 23

Littorina. N a t o Conference Series (Marine interspecific electrophoretic differentiation and


Sciences), 2: 123-140. amino acid divergence. J o u r n a l of M o l e c u l a r
BEAUMONT, A . R . & BEVERJJDGE, C M . 1983. E v o l u t i o n , 19: 449-454.
Resolution of phosphoglucomutase isozymes in HARRIS, H . & HOPKINSON, D . A . 1976.
M y t i l u s e d u l i s L . M a r i n e B i o l o g y L e t t e r s , 4: 97¬ H a n d b o o k of enzyme electrophoresis i n human
103. genetics. Elsevier, Amsterdam.
BEQUAERT, J.C. 1943. The genus L i t t o r i n a in the HARTL, G.B., WILLING, R. & SUCHENTRUNK,
western Atlantic. J o h n s o n i a , 1 (7): 1-28. F. 1990. On the biochemical systematics of
BERGER, E . M . 1973. Gene-enzyme variation in selected mammalian taxa: empirical comparison of
three sympatric species of Littorina. B i o l o g i c a l qualitative and quantitative approaches in the
B u l l e t i n , 145: 83-90. evaluation of protein electrophoretic data.
BERGER, E . M . 1977. Gene-enzyme variation in Zeitschrifl flir zoologische Systematik und
three sympatric species of L i t t o r i n a . II. The E v o l u t i o n s f o r s c h u n g , 28: 191-216.
Roscoff populations, with a note on the origin of JANSON, K . 1985. Genetic and morphological
North American L . l i t t o r e a . B i o l o g i c a l B u l l e t i n , variation within and between populations of
153: 255-264. L i t t o r i n a angulifera from Florida. O p h e l i a , 24:
BUTH, D . G . 1984. The application of 125-134.
electrophoretic data in systematic studies. A n n u a l JOHANNESSON, K. & JOHANNESSON, B. 1989.
Review of E c o l o g y a n d Systetnatics, 15: 501-522. Differences in allele frequencies of Aat between
CAUGANT, D . 1979. Variabilite enzymatique chez high- and mid-rocky shore populations of L i t t o r i n a
quelques especes du genre L i t t o r i n a . these de s a x a t i l i s (Olivi) suggest selection in this enzyme
D o c t e u r 3eme c y c l e , Universite de Paris-Sud. locus. G e n e t i c Research, C a m b r i d g e , 54: 7-11.
CAVALLI-SFORZA, L . L . & EDWARDS, A . W . F . KNIGHT, A . J . , HUGHES, R . N . & WARD, R . D .
1967. Phylogenetic analysis: models and 1987. A striking example of the founder effect in
estimation procedures. E v o l u t i o n , 21: 550-570. the mollusc L i t t o r i n a s a x a t i l i s . B i o l o g i c a l J o u r n a l
COYNE, J . A . , EANES, W . F . , RAMSHAW, J . A . M . o f t h e L i n n e a n Society, 32: 417-426.
& KOEHN, R . K . 1979. Electrophoretic MICKEVICH, M . F . & JOHNSON, M . S . 1976.
heterogeneity of ^-glycerophosphate Congruence between morphological and allozyme
dehydrogenase among many species of data in evolutionary inference and character
D r o s o p h i l a . Systematic Z o o l o g y ; 28: 164-175. evolution. Systematic Z o o l o g y , 25: 260-270.
CROTHER, B.I. 1990. Is "some better than none" MORRIS, S.R. 1979. G e n e t i c v a r i a t i o n i n t h e genus
or do allele frequencies contain phylogenetically Littorina. PhD thesis, University of Wales.
useful information? C l a d i s t i c s , 6: 277-281. MOYSE, J., THORPE, J.P. & A L - H A M A D A N I , E .
ELDREDGE, N . 1985. Unfinished synthesis: 1982. The status of L i t t o r i n a a e s t u a r i i Jeffreys:
b i o l o g i c a l h i e r a r c h i e s a n d modem evolutionary an approach using morphology and biochemical
t h o u g h t . Oxford University Press, New York. genetics. J o u r n a l of C o n c h o l o g y , 31: 7-15.
EMBERTON, K . C . 1988. The genitalic, allozymic, MURPHY, R . W . , SITES, J.W., BUTH, D . G . &
and conchological evolution of the eastern North HAUFLER, C . H . 1990. Proteins I: Isozyme
American Triodopsinae (Gastropoda: Pulmonata: electrophoresis. In: M o l e c u l a r systematics (D.M.
Polygyridae). M a l a c o l o g i a , 28: 159-273. Hillis & C. Moritz, eds), 45-126. Sinauer
FARRIS, J.S. 1972. Estimating phylogenetic trees Associates, Sunderland.
from distance matrices. A m e r i c a n N a t u r a l i s t , 106^ NEI, M . 1978. Estimation of average heterozygosity
645-668. and genetic distance from a small number of
FELSENSTEIN, J. 1981. Evolutionary trees from individuals. G e n e t i c s , 89: 583- 590.
gene frequencies and quantitative characters: NEI, M . 1987. M o l e c u l a r e v o l u t i o n a r y genetics.
finding maximum likelihood estimates. E v o l u t i o n , Columbia University Press, New York.
35: 1229-1242. NEI, M . , TAIIMA, F . & TATENO, Y . 1983.
FELSENSTEIN, J. 1984. Distance methods for Accuracy of estimated phylogenetic trees from
inferring phytogenies: a justification. E v o l u t i o n , molecular data. II. Gene frequency data.
38: 16-24. J o u r n a l of M o l e c u l a r E v o l u t i o n , 19: 153-170.
FERGUSON, A . 1980. B i o c h e m i c a l systematics and NEVO, E. & LAVD2, B. 1989. Selection of
e v o l u t i o n . Blackie, Glasgow and London. allozyme genotypes of two species of marine
FISCHER, P . - H . 1967. Sur quelques especes gastropods (genus L i t t o r i n a ) in experiments of
noduleuses du genre L i t t o r i n a . J o u r n a l de environmental stress by nonionic detergent and
C o n c h y l i o l o g i e , 106: 8-19. crude oil-surfactant mixtures. Ginitique,
FITCH, W . M . & ATCHLEY, W . R . 1987. S e l e c t i o n et E v o l u t i o n , 21: 295-302.
Divergence in inbred strains of mice: a NEWKIRK, G.F. & DOYLE, R.W. 1979. Clinal
comparison of three different types of data. In: variation at an esterase locus in L i t t o r i n a s a x a t i l i s
Molecules a n d morphology i n evolution: conflict and L . o b t u s a t a . C a n a d i a n J o u r n a l of Genetics
o r c o m p r o m i s e ? (C. Patterson, ed.), 203-216. a n d C y t o l o g y , 21: 505-513.
Cambridge University Press, Cambridge. NOY, R., Lavie, B. & NEVO, E . 1987. The niche-
FUERST, P . A . & FERRELL, R . E . 1983. The width variation hypothesis revisited: genetic
analysis of hidden electrophoretic variation: diversity in the marine gastropods L i t t o r i n a
24 T. B A C K E L J A U & T. WARMOES
p u n c t a t a (Gmelin) andL. n e r i t o i d e s (L.). J o u r n a l Bronx, New York.
of E x p e r i m e n t a l M a r i n e B i o l o g y a n d E c o l o g y , SWOFFORD, D . L . & SELANDER, R . K . 1981.
109: 109-116. BIOSYS-1: a FORTRAN program for the
N Y M A N , O.L. 1970. Electrophoretic analysis of comprehensive analysis of electrophoretic data in
hybrids between salmon (Salmo s a l a r L.) and trout population genetics and systematics. J o u r n a l of
(Salmo t r u t t a L.). T r a n s a c t i o n s of t h e A m e r i c a n H e r e d i t y , 72: 281-283.
F i s h e r i e s Society, 99: 229-236. SWOFFORD, D . L . & BERLOCHER, S.H. 1987.
PATTON, J.C. & AVISE, J.C. 1983. A n empirical Inferring evolutionary trees from gene frequency
evaluation of qualitative Hennigian analyses of data under the principle of maximum parsimony.
protein electrophoretic data. J o u r n a l of M o l e c u l a r Systematic Z o o l o g y , 36: 293-325.
E v o l u t i o n , 19: 244-254. SWOFFORD, D . L . & OLSEN, G.J. 1990.
RAMSHAW, J.A., COYNE, J.A. & LEWONTIN, Phylogeny reconstruction. In: Molecular
R . C . 1979. The sensitivity of gel electrophoresis Systematics ( D M . Hillis & C. Moritz, eds), 411¬
as a detector of genetic variation. G e n e t i c s , 93: 501. Sinauer Associates, Sunderland.
1019-1037. TATENO, Y . , NEI, M . & TAJHMA, F. 1982.
REID, D . G . 1986. The l i t t o r i n i d m o l l u s c s of Accuracy of estimated phylogenetic trees from
m a n g r o v e f o r e s t s i n t h e Indo-Paciflc region. molecular data. I. Distantly related species.
British Museum (Natural History), London. J o u r n a l of M o l e c u l a r E v o l u t i o n , 18: 387-404.
RErD, D . G . 1989. The comparative morphology, TATENO, Y . & TAJJMA, F. 1986. Statistical
phylogeny and evolution of the gastropod family properties of molecular tree construction methods
Littorinidae. P h i l o s o p h i c a l T r a n s a c t i o n s of t h e under the neutral mutation model. J o u r n a l of
R o y a l Society of L o n d o n , B 324: 1-110. M o l e c u l a r E v o l u t i o n , 23: 354-361.
REID, D . G . 1990. A cladistic phylogeny of the THORPE, J.P. 1979. Enzyme variation and
genus L i t t o r i n a (Gastropoda): implications for taxonomy: the estimation of sampling errors in
evolution of reproductive strategies and for measurements of interspecific genetic similarity.
classification. H y d r o b i o l o g i a , 193: 1-19. B i o l o g i c a l J o u r n a l of t h e L i n n e a n Society, 11:
RICHARDSON, B . J . , BAVERSTOCK, P.R. & 369- 386.
ADAMS, M . 1986. Allozyme electrophoresis. A THORPE, J.P. 1982. The molecular clock
h a n d b o o k f o r a n i m a l systematics a n d population hypothesis: Biochemical evolution, genetic
s t r u c t u r e . Academic Press, Sydney. differentiation and systematics. A n n u a l Review of
ROSEWATER, J. 1970. The family Littorinidae in E c o l o g y a n d Systematics, 13: 139-168.
the Indo-Pacific. Part I. The subfamily THORPE, J.P. 1983. Enzyme variation, genetic
Littorininae. Indo-Paciflc M o l l u s c a , 2: 417-506. distance and evolutionary divergence in relation to
ROSEWATER, J. 1981. The family Littorinidae in levels of taxonomic separation. In: P r o t e i n
tropical West Africa. A t l a n r t d e R e p o r t , 13: 7-48. polymorphism: adaptive and taxonomic
SACCHI, C.F. 1974. Le polychromatisme des significance. (G.S. Oxford & D. Rollinson, eds).
littorines (Gastropodes, Prosobranches) d'Europe: The Systematics A s s o c i a t i o n Special Volume N o .
points de vue d'un ecologiste. M e m o i r e s de l a 24:131-152. Academic Press, London
S o c i i t i Z o o l o g i q u e de F r a n c e , 37: 61-101. VUILLEUMIER, F. & MATTEO, M . B . 1972.
SHAW, C.R. & PRASAD, R. 1970. Starch gel Esterase polymorphisms in European and
electrophoresis of enzymes - a compilation of American populations of the periwinkle, L i t t o r i n a
recipes. B i o c h e m i c a l G e n e t i c s , 4: 297-320. l i t t o r e a (Gastropoda). E x p e r i e n t i a , 28: 1241¬
SIMON, C. & ARCHIE, J. 1985. An empirical 1242.
demonstration of the lability of heterozygosity WARD, R . D . 1990. Biochemical variation in the
estimates. E v o l u t i o n , 39: 463-467. genus L i t t o r i n a (Prosobranchia: Mollusca).
SINGH, R.S. 1979. Genie heterogeneity within H y d r o b i o l o g i a , 193: 53-69.
electrophoretic "alleles" and the pattern of WARD, R . D . & JANSON, K . 1985. A genetic
variation among loci in Drosophila analysis of sympatric subpopulations of the sibling
p s e u d o o b s c u r a . G e n e t i c s , 93: 997-1018. species L i t t o r i n a s a x a t i l i s (Olivi) and L i t t o r i n a
SNYDER, T.P. & GOOCH, J.L. 1973. Genetic a r c a n a Hannaford Ellis. J o u r n a l of M o l l u s c a n
differentiation in L i t t o r i n a s a x a t i l i s (Gastropoda). Studies, 51: 86-94.
M a r i n e B i o l o g y , 22: 177-182. WARMOES, T. 1986. Een i n l e i d e n d e systematische
SUNDBERG, P., KNIGHT, A . J . , WARD, R . D . & en t a x o n o m i s c h e s t u d i e v a n het genus Littorina
JOHANNESSON, K . 1990. Estimating the (Gastropoda, Prosobranchia). Licentiaatsthesis,
phylogeny of the mollusc L i t t o r i n a s a x a t i l i s (Olivi) Universitaire Instelling Antwerpen.
from enzyme data: methodological considerations. WIUM-ANDERSEN, G. 1970. Haemoglobin and
H y d r o b i o l o g i a , 193: 29-40. protein variation in three species of L i t t o r i n a .
SWOFFORD, D . L . 1981. Oh the utility of the O p h e l i a , 8: 267-273.
distance Wagner procedure. In: Advances in WRIGHT, S. 1978. E v o l u t i o n a n d t h e g e n e t i c s of
c l a d i s t i c s : p r o c e e d i n g s of t h e f i r s t m e e t i n g of t h e n a t u r a l p o p u l a t i o n s , 4: Variability within a n d
W i l l i e H e n n i g Society. (V.A. Funk & D.R. among natural populations. University of
Brooks, eds), New York Botanical Garden, Chicago, Chicago.

You might also like