You are on page 1of 13

CASE STUDY

Wetland Cover Types and Plant Community


Changes in Response to Cattail-Control
Activities in the Palo Verde Marsh, Costa Rica
Florencia A. Trama, Federico L. Rizo-Patrón, Anjali Kumar, Eugenio González, Daniel Somma
and Michael B. McCoy C.

ABSTRACT
The Palo Verde marsh restoration was performed in Costa Rica to reduce invasive Typha domingensis stands. The invasion
of this cattail in the 1980s decreased the habitat quality, and thousands of waterbirds that used this seasonal marsh were
not returning. In 2002, the Organization for Tropical Studies decided to manage the marsh using a tractor with angle-iron-
paddle wheels to crush cattail. We evaluated the land cover changes in response to the prescribed management activities
through GIS and vegetation-assessment analysis. Crushing activities started in |uly 2002. We defined three management
plots: A and B) with crushing activities for different lengths of time and C) the control plot with passive management.
We evaluated wetland cover changes, richness, and dominance of plant species in the three plots. We found that owing
to management activities 1 ) cattail cover area decreased from 35% to 9% (Plot A) and from 62% to 7% (Plot B), while
in Plot C it remained almost constant (always greater than 60%); 2) the area of total live vegetation including cattail
decreased, giving way to open water and bare ground; 3) plant species richness was higher in the crushed plots than in
the control plot; 4) cattail dominance decreased and more species were sharing the marsh area in managed plots than
in the control plot; and 5) cattail cover was negatively related to species richness. We recommend continuing the control
of cattail by crushing in order to rehabilitate the marsh and conserve its biodiversity.
Keywords: aquatic vegetation responses, cattail {Typha domingensis), Costa Rica, Palo Verde Marsh, restoration

I n the neotropics, the richest and


most diverse biogeographical region
in the world, only eight countries have
Other species were registered in a single
count (McCoy 1996, McCoy and
Rodriguez 1994, Sánchez et al. 1985,
and 1980, when the total precipitation
records were lower than the annual
mean values (OTS and IUCN 2005,
implemented measures to restore or Vaughan et al. 1996). Vaughan et al. 1982). Also, the gov-
rehabilitate threatened wetlands Since the early 1500s, when the ernment had granted water conces-
(Ramsar 2005). Costa Ricas Palo Spanish arrived in the area, this wet- sions to large-scale rice farmers from
Verde is a neotropical marsh that has land has contained cattle (Vaughan et theTempisque River (adjacent to Palo
been very important for more than al. 1996). It was part of a huge cattle Verde Marsh), decreasing natural flow
60 species of resident and migratory hacienda until the mid-1970s, when and altering the river's hydrological
waterbirds (Boza 1981, Boza and it was expropriated and declared as relationship with the marsh (OTS and
Mendoza 1981, Meza 1991) (Figure Palo Verde National Wildlife Refuge. UICN 2005).
1). In 1979, more than 5,000 wood In 1978, it was named Palo Verde
storks [Mycteria americana), 30,000 Biological Reserve by the Costa Rican Cattail Invasion and
black-bellied whistling ducks {Den- government (McCoy and Rodriguez
Effects on Plant
drocygna autumnalis), 25,000 blue- 1994). In 1980, cattle were removed
winged teals {Anas discors), 700 roseate (McCoy 1996) and a road was con-
Communities
spoonbills (Ajaia ajaja), and many structed, interrupting the flow of two Cattail (Typha spp.) has colonized
natural streams, one that supplied many wetlands, and its proliferation
Ecological Restoration Vol. 27, No. 3, 2009 water from adjacent hills and another is recognized as a potential threat to
ISSN 1S22-4740 E-ISSN 1543-4079 that allowed tidal flow into Palo Verde wetland ecosystems (Keller 2000,
©2009 by the Board of Regents of the
Marsh (Calvo and Arias 2004). A very Smith and Newman 2001, Waters
University of V^/isconsin System.
dry period occurred between 1975 and Shay 1991). Cattails can be very

278 I* September 2009 ECOLOGICAL RESTORATION 27:3


wetlands invaded by cattail (Bancroft
etal. 2002, Cominetal. 2001, Chau-
han and Gopal 2001, Shy et al. 1998).
In an attempt to control the cat-
tail invasion, the Costa Rican Wildlife
Service signed a five-year contract in
1986 with a local rancher to rein-
troduce 1,000 cattle into the marsh.
However, this management was not
effective because cattle were not
densely enough concentrated into wet,
cattail-infested areas, and these cattle
Palo Vanta HaHkinii Parti
were brought in from upland areas and
were not accustomed to marshy sites
(McCoy and Rodriguez 1994). From
1987 to 1990, initial research and
experimental management led to an
effective technique to control cattail in
Palo Verde Marsh (McCoy and Rodri-
guez 1994).Afarm tractor with angle-
Figure 1. Location of Palo Verde National Park, Palo Verde Marsh and cattail {Typho domingeníis)
treatment plots in Costa Rica; A) crushing; B) crushing and burning; C) control.
iron paddle wheels was used to crush
cattail under water to kill it (Figure
2). After opening 30 ha between 1989
aggressive, expanding their coverage Management Efforts and 1990, 20,000 black-bellied whis-
outward around 9-10 m/y by rhi- to Restore Palo Verde tling ducks and 12,000 blue-winged
zomes, contributing to both biotic Marsh: 1985-2002 teal were registered in a single count in
and abiotic alterations (Houlahan and the managed area (McCoy and Rodri-
Findlay 2004, Keller 2000, McGoy Owing to the management and hydro- guez 1994). Cattail crushing contin-
and Rodriguez 1994). Biotic changes logical alterations in Palo Verde Marsh ued until 1997, when 70 ha had been
include 1) a decrease in richness and from 1975 to 1985, the wetland opened up (McCoy and Rodriguez
diversity of plant species due to com- changed ftom a heterogeneous land- 1994). During this period, no plant
petition; 2) a rapid accumulation scape with open water, low grasses, diversity studies were carried out in
of biomass and leaf litter that limits and floating and low emergent vegeta- the marsh. However, by 2000, with
periphyton development; 3) a decrease tion with some tall emergent patches this opening in the marsh and the
of physical living space for organisms of cattail {T. domingensis) to a mono- accompanying lesser wildfire occur-
across trophic levels; and 4) rapidly typic stand of cattail (Castillo and rence, two tree species, palo verde and
invaded shallow water and decreas- Guzman 2004, McCoy 1994, McCoy michiguiste {Pithecellobium lanceola-
ing waterfowl habitat. Abiotic changes and Rodriguez 1994, Carbonell 1998, tum), increased their coverage to 267
include the alteration of dissolved Vaughan et al. 1996). In 1985, cat- ha (Castillo and Guzman 2004).
oxygen, light, depth, and hydrology tail invaded more than 55 percent of
Because of its importance as an
(Ball 1990, Gronk and Fennessy 2001, the wetland (around 692 ha). This
international site for waterbirds,
Ghiang et al. 2000, Keller 2000, Miao invasion, combined with a very dry
PVNP was registered as a Ramsar Site
et al. 2001, Miao et al. 2000, Newman period, led to several dry-season wild-
(internationally important wetland)
et al. 1998, Carbonell 1998, Smith fires during the 1980s that eliminated
other species, especially palo verde in 1991, and it was included on the
and Newman 2001, Sojda and Solberg
1993). trees [Parkinsonia aculeata), and pro- Montreux list of Ramsar wetlands that
mulgated a homogeneous stand of cat- need management activities in 1993
Moreover, cattail liberates pheno- tall . Cattails overtook almost all open (Quiros et al. 1991). In 1998, tbe
lic components that have inhibitory water and bare ground, causing a large Costa Rican government authorized
effects on the growth and propaga- reduction in waterbird species richness active management in the wetlands
tion of other plant species, and these and abundance in the marsh. In 1988, of PVNP (Presidential Decree No.
detrimental efltcts are not reversible as only 3,000 black-bellied whistling 27345). In 2001, the Organization
long as cattail remains (Gallardo et al. ducks and 500 blue-winged teals were for Tropical Studies (OTS) and the
1999, Gallardo et al. 1998). For these observed in the marsh, numbers which Ministery of Environment and Energy
reasons, plant species richness will be coincide with observations for other (MINAE) through its National Wet-
higher when cattail is managed. lands Program developed a plan to

September 2009 ECOLOGICAL RESTORATION 27:3 279


alligator-flag (7Aíí/íí2^^«¿c«¿2í/2). Some
submerged plant species, Cuadalupe
waternymph [Najas guadalupensis)
and prickly hornwort [Ceratophyllum
muricatum), can be found in deep and
open water areas. However, palo verde
trees, semiaquatics like yerba de jicotea
[Ludwigia erecta), and shrubs like
catsclaw mimosa {Mimosa pigra) and
other legumes (Fabaceae) can be dis-
tributed in shallow-water areas. When
the dry season begins, the aquatic and
semiaquatic plants begin to die, but
some seeds, bulbs, or rhizomes remain
in the propagule bank for the next wet
season (Chauhan and Gopal 2001,
Middleton et al. 1991). However,
cattail remains in the lagoon all year
long, since soil under the plants retains
Figure 2. Tractor with paddie wheeis used to crush cattail in February 2004. Photo by Federico L moisture during the dry season (Crow
Rjzo-Patrón
2002, Carbonell 1998).
Both hydroiogical rehabilitation
recover waterbird habitat by restor- S5°2l"W). Palo Verde Marsh is a sea- and vegetation management activities
ing 50 ha using the cattail-crushing sonal marsh located inside the PVNP, started in July 2002. The two streams
method in addition to cattle grazing covering approximately 1,207 ha that were interrupted in the 1980s
(McCoy and Rodriguez 1994, OTS (Castillo and Cuzmán 2004) (Figure were rehabilitated to reestablish the
2002). In 2002, an additional 250 1). The seasonality of this marsh is wetland hydrology, and water levels
ha of the marsh were included in this the result of rainwater and run-off were monitored by OTS (Jiménez et
management plan. from bordering hills and drying by al. 2003, Trama 2005). Three 80 ha
To evaluate plant responses to the evapotranspiration. The interacting management plots were established.
crushing management activities, we dynamics determine the volume and In Plot A, cattail was crushed without
developed the following two objec- depth of the marsh, in this case vary- controlled fires from July 2002 to the
tives: to analyze the wetland vegetation ing from 0 m to 1.5 m in the rainy end of this study. Prior to this man-
cover changes in Palo Verde Marsh season and between 0 m and 0.4 m agement plan, all of the 1987-2002
using CIS and to analyze the commu- in the dry season until drying up experimental management activities
nity-level plant diversity within two completely between March and April were carried out in 50 ha of this plot.
managed plots, as well as a control plot (Vaughan et al. 1982, Vaughan et al. In Plot B, cattail was managed with
free from active management. We pre- 1996). In the dry season (March and crushing and controlled flre from Sep-
dicted that I) management activities April), extraordinary tides can flood tember 2002 to the end of this study.
will reduce invasive cattail cover, and part of the marsh, and the water can Prior to this study, there were no man-
diverse types of wetland vegetation enter and flood some zones in the last agement activities in this plot. In plots
will develop; 2) greater plant diversity stage of the dry season (Vaughan et A and B, three different tractors were
will be found in managed plots com- al. 1982, Vaughan et al. 1996). Thus used to crush cattail underwater twice
pared to the control plot; and 3) the toward the beginning of the rainy a year for 3 years In early wet season
dominance of cattail will be reduced season, the floating aquatic vegeta- and early dry season, when the water
in managed plots. tion, mainly dotleaf waterlily [Nym- levels were below 35-40 cm. Areas
phaea ampld), water dead and awake with cattail were crushed two or three
{Neptunia plena), and water hyacinth times depending on density until cat-
Methodology
{Eichhornia crassipes), starts to grow tail was totally crushed under water. In
in open-water areas. Other areas of plot B, two controlled fires occurred in
Study Area and Design the marsh are dominated by emergent the dry season to reduce cattail stands
Palo Verde National Park encompasses vegetation, including sedges (Cyper- that remained unmanaged. Plot C was
20,000 ha and is situated in the Are- aceae) and grasses (Poaceae), which the control that did not receive any
nal-Tempisque Conservation Area in form large stands. Still other areas are active management during or before
northwestern Costa Rica (10''2r N, invaded by cattail and talia or bent this study (Figure 1).

280 September 2009 ECOLOGICAL RESTORATION 27:3


Monitoring both a dry and a wet season, using six Results
25 m X 2 m transects (300 m^ total)
We carried out a photo-interpretation
for each plot every month. Plant spe- CIS Analysis
analysis using ArcView 3-3 (ESRI,
cies were identified using Crow (2002)
Redlands CA) to estimate wetland In 1998, baseline wetland cover in
and the Palo Verde Biological Station
cover changes in all plots before the three plots consisted of four cover
herbarium collection. Each plant spe-
and during management activities. types: palo verde, open water, cattail,
cies was counted along each transect,
We analyzed two aerial photographs and other vegetation. In December
and its percent cover was estimated.
and one satellite image: 1) a Terra 2002, six cover types were differenti-
To facilitate the estimation of percent
orthophotograph (resolution 30 m) ated: palo verde, open water, cattails,
cover, we used a 2 m x 2 m quadrat
acquired November 1998, late wet crushed cattails, emergent vegetation,
subdivided into four sections. For each
season/early dry season; 2) a three- and floating vegetation. By March
transect, we used 12.5 quadrats (75 per
band Ikonos image (resolution 4 m) 2003 in late dry season, another wet-
management plot). We estimated the
acquired December 2002, late wet land cover class was identified in the
total cover of every plant species in all
season/early dry season, five months crushed areas: exposed ground (Figure
quadrats along each transect. A value
after management activities began; and 3). During the first five months of
of 1% was given to those species that
3) an infrared aerial photograph (reso- crushing activities, cattail cover was
covered 1 % or less of the area sampled.
lution 2 m) from the CARTA project reduced from 34.7% (baseline) to
Clonal reproduction of many wetland
of March 2003, late dry season, nine 23.9% in plot A and from 61.5% to
plant species prevented calculation of
months after management. 52.2% in plot B, while in the control
abundance (Crow 2002).
We delineated eight wetland cover plot it increased by 5.1 %. Afterward,
classes depending on the season Data Analysis from December 2002 to March 2003
and management plot (Crow 2002, cattail cover decreased by 25.5% in
Hernández 1990, Hernández and We carried out a descriptive GIS plot A and 54.5% in plot B. How-
Gómez 1993, Trama 2005): 1) cattail analysis comparing the wetland cover ever, plot C, the unmanaged plot, was
stands; 2) crushed cattail; 3) emer- classes before the restoration initia- always more than 60% covered by
gent plants, including grasses, sedges, tive started (baseline at 1998) and cattail.
legumes, and spurges (Euphorbia- five months (2002) and eight months
A change in the area occupied by
ceae); 4) floating vegetation, includ- (2003) after management activities
TLVA due to cattail-crushing manage-
ing water hyacinths (Pontederiaceae), began. The estimated coverages were
ment in plots A and B was observed
wateriilies (Nymphaceae), duck- compared between years and manage-
during 2002-2003 (Figure 3). In
weeds (Lemnaceae), and tape-grasses ment plots. We calculated the Total
1998, TLVA, including cattail, was
(Hydrocharitaceae); 5) open water Live Vegetation Area (TLVA), includ-
95.1% of plot A, 99.9% of plot B, and
areas; 6) palo verde trees; 7) exposed ing cattail and other vegetation, as well
98.7% of plot C. By December 2002,
ground, delineated during dry season as other wetland cover classes as a total
TLVA decreased to 85.4% in plot A,
and including dry or wet mud, typi- for each year or plot analysis.
and to 67.6% in plot B, but still cov-
cally free of aquatic vegetation, but The six transects were considered
ered 99.9% of the control plot (C).
in some cases covered by some dry- replicates for each plot. The mean spe-
In March 2003, TLVA was 28.2% of
season plant species; and 8) other veg- cies richness was compared between
plot A, 22.6% for plot B, and 97.9%
etation cover, including emergent and the three plots by one-way ANOVA
of the control plot. The remaining
floating vegetation for the 1998 aerial (LSD interval for equal samples) using
area in all cases was occupied by open
photograph. Statgraphics 3.1 (Manugistics, Rock-
water, exposed ground, and managed
ville MD). For the seasonal distribu-
We defined a minimum mapping sectors with crushed cattail.
tion of species, we tested interactions
unit (mmu) of 30 m for the 1998 During management activities,
between dates and all three plots.
photograph based on the resolution emergent and floating vegetation
Using the vegetation sampling data,
of imagery used. The mtnu defined could be delineated, and several sea-
we estimated the percentage of TLVA
for the 2002 and 2003 analyses was 5 sonal changes were observed in the
and other wetland covers. Dominance
m, making the two years directly com- plots. In 1998, open water occupied
of cattail and other plant species was
parable to each other. Accuracy assess- 4.9% of plot A and 0.1% of plot B.
compared between the three plots.
ment of classified areas was performed In December 2002 after five months
We defined a plant as dominant when
using data provided by managers and of tnanagement, the percentage of
it covered more than 10% of the
by ground truthing in the marsh. open water increased to 9.5% and
sampled area within a plot on at least
We sampled vegetation 11 times 16.1% in plots A and B, respectively.
one sampling date. Finally, a linear
in the three management plots from Additionally, we observed that in the
regression analysis determined if cat-
August 2003 to July 2004, including dry season (from December 2002 to
tail cover influenced species richness.

September 2009 ECOLOGICAL RESTORATION 27:3 ** 281


1998

2002

I I Floating Vegetation
•I Crushed Cattail
n3 Exposed Ground
^1 Emergent Vegetation
^1 Open Water
• Palo Verde
^1 Other Vegetation
im Cattail

2003

900 1600 Meters

Figure 3. Wetland cover by treatment plot, before (1998) and during (2002, 2003) management activities.

282 September 2009 ECOLOGICAL RESTORATION 27:3


Table 1. Plant species recorded in each treatment plot: A) plot managed since 1987 with cattail crushing; B) plot
managed since 2002 with cattail crushing and burning; and C) control plot without active management.
Plot
Family Species

Alismataceae Echinodorus paniculatus X


Araceae Pistia stratiotes X
Asteraceae Edipta prostrata X
Boraginaceae Heliotropium sp. X
Cannaceae Canna glauca X
Ceratophyllaceae Cerataphyllum muricatum X
Convoivulaceae Aniseia martinicensis X
Ipomoea carnea X
Cucurbitaceae Luffa operculata X
Cyperaceae Cyperus articulatus X
C. digitatus X
C tria X
C. tenerrimus X
Eleocharis elegans
E. interstincta X
E. mutata X
Fimbristylis spadicea (= F. thermalis)
Oxycaryum cúbense X
Elatinaceae Sergio capensis X
Euphorbiaceae Caperonia castaneifolia X
Crotón argenteus X
Euphorbia sp. X
Fabaceae Aeschynomene sensitiva X
Mimosa pigra X
Neptunio natons X
Parkinsonia aculeato X
Sesbonio emerus X
Hydrocharitaceae Limnobium iaevigotum X
Nojas guodalupensis X
Hydrophyllaceae i-lydraieo elatior
Lemnaceae Lemna aequinoctioHs X
Wolffieila welwitschii X
Lentibulariaceae Utricuiaria foliosa X
U. gibbo X
Malvaceae Molochro fasciota X
Molvostrum americanum X
Marantaceae Tholia geniculota X
Marsileaceae Morsiieo deflexa
Nymphaeaceae Nymphoeo omazonum X
Nymphaeo ampio X
Onagraceae Ludwigio erecto X
Ludwigio indinota X
Parkeriaceae Cerotopteris pteridoides X
Poaceae Echinochloa colono X
Hymenachne omptexicaulis X
Leersia hexondra
Paspolidium geminatum X
Pospalum repens X
Pospolum sp. X
Polygonaceae Polygonum hispidum
P. segetum X
Pontederiaceae Eichbornio crassipes X

September 2009 ECOLOGICAL RESTORATION 27:3 283


Table 1., continued

Plot
Family Species

Heteranthera limosa
Portulacaceae Portulaca sp.
Rubiaceae Borrerio assurgens
Azolla mexicana
Salviniaceae Salvinia minima
Bacopa axillaris
Scrophulariaceae B. repens
Physalis angulota
Solanaceae Solanum campechiense
Typha domingensis
Typhaceae Phyla nodiflora
Verbenaceae Unidentified 1
Malvacea Unidentified 2
Verbenacea Unidentified 3
Asteraceae Unidentified 4
Unidentified 5
Unidentified 6

March 2003), che area of open water ( f = 3 . 3 1 , df= 20,197,;'< 0.0001) Discussion and
decreased owing to desiccation, to (Figure 4). Recommendations
8.5% in plot A and 0.4% in plot B. According to our vegetation tran-
However, in plot C, open water was Cattail in tropical and some temper-
sects, cattail cover was always less than
always around 2% cover because that ate wetlands is considered an invasive
20% in plot A, less than 40% in plot
part of the marsh is more than one plant that reduces habitat availability
B, and greater than 70% in plot C
meter deep and does not dry up until for wildlife and plant communities
during this study [F = 884.55, df =
around April. Also, the floating veg- (Farnsworth and Meyerson 2003,
2,197,p < O.OOOl) (Figure 4). Total
etation cover decreased from 24% to Crace and Wetzel 1981, Houlahan
live vegetation area varied seasonally
7.6% in plot A and 0.9% to 0% in and Findlay 2004, Lee et al. 1996).
in the managed plots but remained
plot B owing to the normal seasonal Monotypic cattail stands reduce over-
almost constant in the control plot
change of the marsh. We did not all habitat value, affecting some species
(> 85%) (Figure 5). Additionally, in
detect floating vegetation in plot C. of wildlife such as waterfowl, water-
the dry season (January to May), the
At the same time, exposed ground was birds, fish, amphibians, and inver-
percentage of TLVA was lower in the
appearing in the dry season when open tebrates (Lee et al. 1996, Sojda and
managed plots. Moreover, we found
water areas and floating vegetation was Solbergl993).
that dominance was shared among
decreasing, covering 32.9% in plot A more species in the managed plots The results of this study show that
and 17.8% in plot B. Exposed ground than in the control plot. In plots A crushing activities have significantly
was not detected on the control plot in and B, we found cattail and tnore reduced cattail stands and improved
any of the analysis years, as this cover than 50 other plant species with habitat for other plant species. In 1998,
appears only after open water areas dry greater than 10% cover, whereas in the difference in cover classes between
up (Figure 3). plot C, cattail was clearly the domi- plot A and plots B and C is striking.
nant species, and less than 20% of By then, plot A had already experi-
the total area was shared with 19 enced more than ten years of experi-
Vegetation Sampling species (Figure 6). Furthermore, the mental active management (McCoy
Ihe total plant species richness of the regression analysis showed a nega- and Rodriguez 1994). Plot A had less
marsh was 69; we found 59 species tive relationship between cattail cover cattail cover and the rest of the plot
in plot A, 61 in plot B, and 20 in and plant species richness by transect was occupied by other vegetation and
plot C (Table 1). The mean species (Figure 7). Fire did not have a sig- open water. However, in plots B and C,
richness was higher in plots A (13.2) nificative effect on cattail in plot B. cattail was present over more than 60%
and B (11.8) than in plot C (4.2) Cattail started growing again 30 days of the area owing to its invasiveness
{F= 104.94, df = 2.197,/- < 0.0001). after the burn. In the wet season, the and the lack of management. Between
Additionally, mean species richness tractor had to pass over the burned 2002 and 2003, plots A and B showed
varied temporally during this study area to reduce this cover. an increase in plant cover types and

284 September 2009 ECOLOGICAL RESTORATION 27:3


dormant, and exposed ground replaces
all the other areas except cattail. This
is because cattail does not vary season-
ally like other vegetation in a seasonal
Rot marsh (Crow 2002). By decreasing
I — A cattail cover and, consequently, the
er percentage of TLVA, management
B
activities have permitted the reestab-
- C lishment of the wetland heterogeneity.
CO
ö
o r \.H. I From the beginning of cattail-
crushing activities, the vegetation
assessment showed that plant species
richness varied both between treat-
ments and from seasonal changes
c o c o
during the year. However, species rich-
99 ness was always higher in the managed
plots when compared to the control
plot. In plots A and B, there were
SarrpJingcËte many dominant species compared to
the one dominant plant (cattail) in
the control plot. Additionally, 60%
110 of the variation seen can be explained
by the relationship between cattail per-
90 cent cover and species richness. This
implies that management is necessary
Rat to improve and maintain habitat in
— A Palo Verde Marsh.
if 50 We assume that the return of the
B
m wetland heterogeneity is related to the
C recovery of flora, fauna, and ecologi-
Ü '*' cal processes (Ruiz-Jaen and Mitchell
2005, Young 2000). In complemen-
10 tary assessments, greater species rich-
ness and numbers of waterbirds were
observed in plots A and B than in plot
C. Of 62 bird species recorded in the
study, 54 were registered in plot A, 52
in plot B, and 22 in plot C, with only
Ssrrplingdate 12 in the cattail cover. Moreover, more
than 35,000 black-bellied whistling
Figure 4. Seasonal variation in mean plant species richness (top) and percent cover of cattail
(Jypha domingensii) (bottom) in three experimental plots: A) plot managed since 1987 with cat-
ducks and about 9,000 blue-winged
tail crushing; B) plot managed since 2002 with cattail crushing and burning; and C) control plot teals were observed in the managed
without active management. plots in a single count (Trama 2005).
Open water and exposed ground were
plant species diversity compared with activities is related to larger and more the two wetland covers preferred by
the control plot, which demonstrates numerous areas of shallow water or waterfowl and other waterbirds,
that this particular type of manage- exposed ground in the dry season, which have used the wetland during
ment is successfial. The small changes conditions that are normal in this sea- the crushing activities (Colweli and
in wetland cover in the control plot sonal marsh (McCoy and Rodriguez Taft 2000, Green 1996, Trama 2005).
observed during the study could be due 1994). After cattail was crushed in Additionally, more and different taxa
to variations in resolution. In spite of plots A and B, these areas were occu- of invertebrates have been recorded in
this, control plot C always had more pied by open water or other wetland managed plots than in the control plot
than 60% of cattail coverage. cover classes. During dry-season desic- (Rizo-Patrón et al. 2009).
The reduction of total live vegeta- cation, the open water decreases, the However, after almost 20 years of
tion area (TLVA) due to the crushing seeds of floating vegetation become invasion by cattail, this plant is part of

September2009 ECOLOGICAL RESTORATION 27:3 285


PIOtA
grow back very fast (McGoy 1994).
Instead of being eradicated, most of
the cattail stands were reduced, so the
risk of teinvasion remained (McCoy
and Rodriguez 1994, SER 2004). In
cases where water levels were low ( 15
cm) during crushing, some cattail
resprouted. We recommend manu-
ally cutting this sparse regrowth to
reduce the need for future crushing
in the same area. Otherwise, cattail
returns quickly to original densities. It
PlotB is necessary to continue the manage-
ment every year to guarantee cattail
is controlled. Also, it is important to
note that management activities with
the tractor were directly associated
with more open areas free of cattail.
TLVA These areas were colonized by other
Other habitats plant species that behaved like normal
seasonal plant communites. In this
case, seasonality was responsible for
the reduction of floating vegetation
and open water and their replacement
by exposed ground from December
2002 to March 2003 (Trama 2005).
Fire as a management tool was not
successful because it occurred in the
dry season. Cattail started to grow
very fast and there was no water in
the marsh to allow the tractor to enter
and control it. Additionally, fire was
very bad for wildlife that use cattail
stands as a refuge in the dry season
(Castañeda 2003).

The key to conservation of inland


wetland biodiversity is to discourage
the spread of community dominants
Sampling date
(Houlahan and Findlay 2004). Thus it
Figure 5. Percentage of Total Live Vegetation Area (TLVA) and other cover types (inciuding is necessary to continue with manage-
crushed cattail, open water, and exposed ground) during this study: A) plot managed since ment activities to maintain the Palo
1987 with cattail crushing; B) piot managed since 2002 with cattail crushing and burning; and Verde wetland habitat by controlling
C) control plot without active management.
invasive species. Crushing with the
tractor is the first step and should be
the Palo Verde marsh and several spe- Management intensity varied complemented by grazing to control
cies of birds, reptiles, and invertebrates depending on availability of the cat- other species that appear after cat-
use this cover, mainly for resting and tail-crushing tractor and water levels. tail elimination, such as West Indian
protection from predators (Castañeda The tractor could not enter when marsh grass {Hymenachne amplexi-
2003, Trama 2005, Vaughan et al. water was deeper than 45 cm. In some caulis) and Egyptian panicgrass {Pas-
1996). Therefore, leaving cattail paths areas, the tractor had to pass two or palidium geminatum), both of which
or corridors in some areas and near the three times to complete the cattail produce an effect similar to cattail as
fence line has been suggested to help crushing. Because cattail is rhizoma- far as decreasing habitat quality for
supply the seasonal requirements of tous, if it is not eradicated within 15 waterfowl. Grasses can also be a very
wildlife (Vaughan et al. 1996). days of management activities, it can dangerous fuel source for wildfire

286 September 2009 ECOLOGICAL RESTORATION 27:3


Plot A
(Calvo and Arias 2004). The results
and implications of these activities
are still being analj'zed. However, it is
believed that the reestablishment of
the two previously rerouted streams
•*T'
\.\ )
HymenactiiTv I
N»pliinienaians
will permit the wetland to fill with
II I ' Nymphaoe amBiormm rainwater more quickly. Cattail and
(B39 TttaKa geniculata
40-
^ B LimnotMum faanoatum aquatic grasses cannot survive as easily
• • O'ycaryum cubens«
t^^-^ Eictionia crasspos if the wetland is filling quickly with
ODwr 59 epeom
rainwater and therefore will likely
establish in fewer areas. With crushing
activities and hydrological restoration,
it is predicted that Palo Verde Marsh
PlolB
will be rehabilitated to a self-regulat-
ing ecosystem with conditions similar
to those prior to the cattail invasion.
Although this study only considered
the management responses in Palo
Verde Marsh, similar responses could
be obtained in other wetlands invaded
by cattail. Inside PVNP, La Bocana
Wetland also has problems with cattail
invasion (Rizo-Patrón 2004), while
Varillal and Piedra Blanca wetlands
have problems with catsclaw acacia
PlotC {Mimosapigra). We recommend evalu-
ating possible management activities
in these other wetlands to recover and
conserve these habitats and their bio-
diversity. After the management done
in Palo Verde, thousands of ducks and
other waterbirds were observed at the
wetland, and the historic abundance
of individuals recorded before the
cattail invasion for some species has
recovered (Trama 2005). In this way,
including several sites in a restoration
strategy to recover wetlands inside and
around the national park would ensure
the conservation of these habitats and
Sampling date
biodiversity at a landscape level.
Habitat improvement to benefit
Figure 6. The relative cover of the most dominant plants during this study: A) plot managed since wildlife was the main objective of the
1987 with cattail crushing; B) plot managed since 2002 with cattail crushing and burning; and
C) control plot without active management. restoration initiative in Palo Verde.
Few data exist on waterbirds' pref-
erences for cover class areas in Palo
in the dry season (Barboza 2005, is usually too low to produce an Verde Marsh; however, a more het-
Vaughan et al. 1996). Palo Verde impact on vegetation, unless restricted erogeneous wetland and more shal-
National Park administration should by fencing. Recent research shows that low-water or exposed-ground areas
improve grazing supervision and take proper cattle grazing with fencing does could imply more feeding and resting
advantage of the cattle present at the control cattail and grasses and increase habitats for more than 60 waterbird
park. Cattle should he concentrated in plant diversity (M. McCoy, pers. obs.). species (McCoy 1996, Trama 2005).
smaller areas with fencing and rotated Additionally, some activities to We recommend studies on waterbirds'
between fenced parcels. Cattle density restore hydrology have occurred preferences for each vegetation type.

September 2009 ECOLOGICAL RESTORATION 27:3 287


100 —L bv¿ g • —
Cronk, J.K. and M.S. Fennessy. 2001. \^et-
Ban D o
- X. ^& D^\n u• B
a §
\ -
- land Plants: Biology and Ecobgy. Boca
- \.^~N. Raton FL: CRC Press.
50 • \ -
• ^
v Crow, G.E. 2002. Plantas Acuáticas del
\ Q \
Parque Nacional Palo Verde. Heredia,

N X.
8
60 -

-
N
\
X n
G ^
Costa Rica: INBio.
Farnsworth, E.J. and L.A. Meyerson. 2003.
\
Comparative ecophysiology of four wet-
attail

AQ D
D •
\ ^ 5 - land plant species along a continuum
o \
X ofinvasiveness. U^f¿íní¿ 23:750—762.
20 -
° \ . • Gallardo, M.T, J. Ascher, M.J. Collier. B.B.
\

\ Martin and D.E Martin. 1999. Effect


Q D
N •
-

of cattail ( Typha domingensis) extracts,


0 0 -^ -
- , . . . 1
leachates, and selected phenolic com-
10 15 20 25 30 pounds on rates of oxygen production
Plant species richness by salvinia {Salvinia minima). Journal of
Aquatic Plant Management 37:80-82.
Figure 7. Relationship between cattail cover and plant species richness, (f = 268.65, df = 1,112, Gallardo, M . T , B.B. Martin and D.F.
R* = 0.60, p < 0.0001). Martin. 1998. Inhibition of water
fern Salvinia minima hy cattail ( lypha
domingensis) extracts and by 2-chloro-
Acknowledgments Boza, M.A. and R. Mendoza. 1981. The
phenol and salicylaldehyde. Journal of
Our sincerest thanks to Organization for National Parks of Costa Rica. Madrid:
Chemical Ecology 24:1483-1490.
Tropical Studies (OTS), the Instituto en INCAFO.
Grace, J.B. and R.G. Wetzet. 1981. Hab-
Conservación y Manejo de Vida Silvestre Calvo, J.C.A. and O.R. Arias. 2004. Res- itat partitioning and competitive dis-
ICOMVIS (formerly PRMVS), and the tauración hidrológica del humedal Palo
placement in cattails: Experimental
Ministerio de Ambiente y Energía (MINAE) Vtiáe. Ambientico 129:7-8.
field studies of intensity of competi-
for making possible this project. To Ducks Carbonell. M. 1998. Sitio Ramsar Parque tion. American Naturalist 118:463-474.
Unlimited, the United States Fish and Wild- Nacional Palo Verde, Costa Rica. Misión
Green, A.J. 1996. Analyses of globally
life Service (USFWS). Costa Rica-United Ramsar de Asesoramiento Report
threatened anatidae in relation to
States Foundation (CRUSA), The Ramsar No. 39. Gland, Switzerland: Ramsar
threats, distribution, migration patterns,
Convention, AVINA. and IDEA WILD Convention Office.
and habitat use. Conservation Biology
for supporting the Palo Verde restoration Castañeda, E. 2003. Informe del impacto 10:1435-1445.
project. Also, we would like to thank the de las quemas en la laguna Palo Verde.
Hernández, D.A.E. 1990. Cambios anu-
OTS administration that helped co get the Guanacaste, Costa Rica: Organization
ales en la composición y distribución
research and collecting permit (No. 30856). for Tropical Studies.
de la vegetación acuática en el humedal
Thanks go to my friend Montserrat Car- Castillo, M. andJ.A. Guzman. 2004. Cam- estacional de Palo Verde, Costa Rica.
bonell, who has always encouraged me with bios en cobertura vegetal en Palo Verde
Heredia, Costa Rica: Organization for
her advice; to Mauricio Castillo and José según SIG. Ambientico 129:4-6.
Tropical Studies.
Guzman for technical support with GIS; Chauhan, M. and B. Gopal. 2001. Biodi- Hernández, E.D. and J. Gómez. 1993. La
and a very special acknowledgment to Ken versity and management of Keoíadeo
Flora Acuática del Humedal Palo Verde.
Kriese and Dawn Browne for their editing National Park (India)—A wetland of
Heredia, Costa Rica: EUNA.
and advice. international importance. Pages 217-
Houlahan, J.E. and C.S. Findlay. 2004.
256 in B. Gopal, W.J. junk and J.A.
Effect of invasive plant species on tem-
Davis (eds). Biodiversity in Wetlamh:
perate wetland plant diversity. Conser-
References Assessment, Function and Conservation,
vation Biology \%:\\^l-\ 138.
Ball, J.R 1990. Influence of subsequent vol. 2. Leiden, Netherlands: Backhuys
Jiménez, J.A., E. González and J. Calvo.
Hooding depth on cattail control by Publishers.
2003. Casos humedal Palo Verde y La
burning and movj'mg. Journal of Aquatic Chiang, C , G.B. Craft, D.W. Rogers and Bocana. Heredia, Costa Rica: Organi-
Plant Management 28:32-36. C.J. Richardson. 2000. Effects of 4
zation for Tropical Studies.
Bancroft, G.T., D.E. Gawlik and K. years of nitrogen and phosphorus addi-
Keller, B.E.M. 2000. Plant diversity in
Rutchey. 2002. Distribution of wading tions on Everglades plant communities.
Lythrum, Phragmites and Typha Marshes,
birds relative co vegetation and water Aquatic Botany 68:61-70.
Massachusetts. Wetlands Ecology and
depths in the northern Everglades of Colwel!, M.A. and O.W. Taft. 2000. Water-
Management 8:391-401.
Florida, USA. Waterhiräs 25:265-277. bird communities in managed wetlands
Lee, M.A., R. Conrow, K. Fallón, K. Ponzio,
Barboza, G.G. 2005. Sosrenibilidad del pas- of varying water depth. Waterbirds H\
L. Scafidi and K. Snydcr. 1996. 'Die
toreo en un humedal tropical: El caso 45-55.
natural history and management of
del Parque Nacional Palo Verde, Costa Comin, F.A., J.A. Romero, O. Hernan-
Typha: A review of the literature.
Rica. MS thesis. Universidad Nacional dez and M. Menendez. 2001. Restora-
Florida Department of Water Resources.
de Costa Rica. tion of wedands from abandoned rice
McCoy, M. 1994. Seasonal, freshwater
Boza, M.A. 1981. Aves sobresalientes de fields for nurrienr removal, and biolog-
marshes in the tropics: A case where
Palo Verde. Biocenosis 3(l-2):23-25. ical and community landscape diversity.
cattle grazing is not bad. Pages 352-353
Restoration Ecology 9:207-208.

288 September 2009 ECOLOGICAL RESTORATION 27:3


in G. MefFe and C. Carroll (eds), Princi- Quirós, G., M. Solano and J. Gamboa. de manejo y desarrollo. Heredia, Costa
ples of Conservation Biology Sunderland 1991. Ficha informativa de Ramsar del Rica: EUNA.
MA: Sinauer Associates. Parque Nacional Palo Verde. Gland. Vaughan, C . M. McCoy. J. Fallas. H.
__. 1996. Tlie seasonal, freshwater marsh Switz.erland: Ramsar. Chaves, G. Barbosa et al. 1996. Plan de
at Palo Verde National Park. Pages 133- Ramsar. 2005. Regional overview of the manejo y desarrollo del Parque Nacio-
137 in A.J. Hails (ed). Wetlané, Biodi- implementation of the Convention nal Palo Verde y Reserva Biológica
versity and the Ramsar Convention: The and its strategic plan 2003-2008: Neo- Lomas Barbudal. Heredia, Costa Rica:
Role of the Convention on Wetlands in tropics. 9th Meeting of the Conference of Universidad Nacional.
the Conservation and Wise Use of Bio- the Contracting Parties to the Convention Waters, 1. and J.M. Shay. 1991. Effective
diversity. Gland, Switzerland: Ramsar on Wetlands (Ramsar, Iran, 1971), Kam- water depth on population parame-
Convenrion Bureau. pala, Uganda, 8-15 November 2005. ters of a Typha glauca stand. Canadian
McCoy, M.B. and J.M. Rodriguez. 1994. Ramsar COP9 Doc. 12. www.ramsar journal of Botany 70:349-351.
Cattail ( Typha dominguenis) eradication .org/cop9/cop9_docl2_e.htm Young, T.P. 2000. Restoration ecology and
methods in the restoration of a tropical, Rizo-Patrón, F.L. 2004. Humedal La conservation biology. Biological Conser-
seasonal, freshwater marsh. Pages 469- Bocana precisa restauración. Ambien- vation 92:73-78.
482 in W.J. Mltsch (ed). Global Wet- tico 129:9-10.
lands Old World and New. Amsterdam, Rizo-Patrón, EL., EA. Trama and E.
Netherlands: Elsevier. González. 2009. Responses of the
Meza, T.A. 1991. Refugio Nacional de invertebrate community to the restora-
Fauna Silvestre Rafael Lucas Caballero tion activities at the Palo Verde Marsh. Florencia Trama, Centro Neotropical de
(Palo Verde). Herencia 3(l-2):68-76. Unpublished manuscript. Entrenamiento en Humedales—Perú,
Miao, S.L., P.V. McCormick, S. Newman Ruiz-Jaen, M.C. and TA. Mitchell. 2005. fr. Puerto Inca (ex calle 8), # 174 Depto
and S. RajagopaJan. 2001. Interactive Restoration success: How is it being 302, Urbanización Los Olivos-Suero,
etfects of seed availability, water depth, measured? Restoration Ecology 13: Lima 33, Perú, +511 273-3558.
and phosphorus enrichment on cat- 569-577. ftrama@centroneotropical.org
tail colonization in an Everglades wet- Sánchez, J., J.M. Rodriguez and C. Salas.
land. Wetlands Ecology and Management 19S5. Distribución, ciclos reproductivos Federico L. Rizo-Patrón, Centro Neo-
9:39-47. y aspectos ecológicos de aves acuáticas. tropical de Entrenamiento en Humedales
Miao, S.L., S. Newman and EH. Sldar. Pages 83-102 in E. Guier (ed), Investig- - Perú. Jr. Puerto Inca (ex calle 8). # 174
2000, Effects of habitat nutrients and aciones Sobre la Fauna Silvestre de Costa
Depto 302, Urbanización Los Olivos-
seed sources on growth and expansion Rica. San José: Universidad Estatal a
Surco. Lima 33, Perú. +511 273-3558,
of Typha domingensis. Aquatic Botany Distancia.
frizopatron@centroneotropical.org
68:297-293. Shy, E.. E. Beckerman, T Oron and E.
Middleton, B.A., A.G. Van de Valk, D.H, Frankenberg. 1998. Repopulation and
Mason, R.L. Williams and C.B. Davis. colonization by birds in the Agmon Anjali Kumar, Council for International
1991. Vegetation dynamics and seed wetland, Israel. Wetlands Ecology and Educational Exchange, Study Center,
banks of a monsoonal wetland over- Management 6:159-167. Monteverde, Puntarenas, Costa Rica,
grown with Paspalum distichum L. in Smith, S. and S. Newman. 2001. Growth kumar.anja@gmail.com
northern India. Aquatic Botany 40: of southern cattail {Typha domingen-
239-259. sis Pers.) seedlings in response to rife/ Eugenio González, Director, Soltis
Newman, S., J. Schuette, J.B. Grace, K. related soil transformations in the north- Research and Education Center in Costa
Rutchey, T. Fontaine et al. 1998. Fac- ern Florida Everglades. Wetlands. 21: Rica, San Isidro de Penas Blancas, Apdo
tors influencing cartail abundance in 363-369. 80-^417, Fortuna San Carlos, Costa
the northern Everglades, Aquatic Botany Society for Ecological Restoration (SER). Rica and Apartado 676-2050, San Pedro,
60:265-280. 2004. The SER International Primer on Costa Rica. +506 8325-9898, egonzaUz
Organization for Tropical Studies (OTS). Ecological Restoration. Tucson: SER and @ipomail. tamu. edu
2002. Conservation of neotropical www.ser.org
migrants and their habitat in the Palo Sojda, R.S. and K.L Solberg. 1993. Manage- Danielf. Somma, Administración
Verde marsh, Costa Rica: Restoration ment and control of cattails. Waterfowl de Parques Nacionales de Argentina,
and management of a Montreux Record management handbook. U.S. Fish and Av. Santa Fe 690, 1059 Capital Fed-
Ramsar site. Wildlife Service Leaflet 13.4.13. digital eral, Argentina. +5411 4311-0303,
Organization for Tropical Studies and Inter- commons.unl.edu/icwdmwfni/33/
djsomma(Sryahoo.com.ar, dsomma@apn
national Union for Conservation of Trama, EA. 2005. Manejo activo y restaura-
.gov.ar
Nature (OTS and lUCN). 2005. Car- ción del humedal Palo Verde: Cambios
acterización hidrológica del Río Temp- en las coberturas de vegetación y respu-
isque: Base para la definición de caudales Michael B. McCoy CoUon, Escuela de
esta de las aves acuáticas. MS thesis,
ambientales, capitulo III. Estudios bási- Universidad Nacional de Costa Rica. Ciencias Ambientales, Universidad Nacio-
cos preliminares para la determinación Vaughan, C . G. Canessa, M. McCoy, M. nal, Heredia, Costa Rica, +506 8844-
del caudal ambiental del Ríolempisque. Rodriguez, J. Bravo et al. 1982. Refugio 5276, mmccoycolton@yahoo.com
OTS-IUCN Project Report. San José: de Fauna Silvestre Rafael Lucas Castro
OTS-IUCN. Rodríguez Caballero (Palo Verde). Plan

September 2009 ECOLOGICAL RESTORATION 27:3 289

You might also like