Sperm storage and low incidence of multiple paternity in the European pond turtle, Emys orbicularis: A secure but costly strategy?

The freshwater pond turtle, Emys orbicularis, has recently suffered from population declines throughout its range, mainly due to habitat destruction. The mating strategies of this species were studied using genetic data from successive clutches within and between years. To test for the occurrence and frequency of multiple paternity and sperm storage, genetic paternity at six microsatellite markers was assessed in 114 embryos and hatchlings from single and subsequent clutches of 11 females (including clutches from the same or consecutive years). Multiple paternity was rare and only found in two out of 20 clutches from 11 females. All annual successive clutches and 58% of the clutches in the next year, were fertilized with sperm from the same male. The use of stored sperm is thus a frequent strategy in E. orbicularis. However, hatching rate, hatchling mass, and hatchling length decreased in clutches fertilized by stored sperm, suggesting sperm depletion or deterioration through time. The occurrence of stored sperm despite an associated reduced reproductive output indicated that mating and/or the fertilization process is costly to females. The low incidence of multiple paternity may simply be the residual consequence of the capacity to store viable sperm. These results provide important and innovative insights for the conservation of E. orbicularis. In threatened populations, management strategies may aim to enhance effective copulations in order to increase the reproductive output of females.     

Effects of landscape composition, habitat features, and nest distribution on predation rates of simulated turtle nests

We investigated predation of simulated turtle nests in an effort to understand how land-use patterns and the availability of nesting habitat may affect turtle recruitment in a region where human populations and associated development are increasing. Simulated nests were patterned after those created by painted turtles (Chrysemys picta), a common aquatic turtle in our study area, and distributed in four patterns (clustered and near pond, scattered and near pond, clustered and far from pond, and scattered and far from pond) around 36 ponds. Landscape composition (500-2000 m from pond perimeters) and habitats surrounding pond edges (an area extending 250 m from the shore of each pond) were then compared with rates of predation at each pond. Nest-site characteristics also were compared to the fate of individual nests. Landscape composition and habitats surrounding ponds apparently had little influence on predation rates. Nest distribution and the immediate habitat features associated with each nest did affect vulnerability to predation. Clumped nests were preyed upon at a higher rate than scattered nests, and nests close to ponds (within 50 m) were more vulnerable to predators than those created far (100-150 m) from a pond. Counter to our expectations, proximity to edge habitats (other than the shore of a pond) reduced the probability that a nest would be detected by predators. Also, nests placed near roads and suburban lawns had a reduced likelihood of predation whereas those placed in agricultural areas or disturbed sites had a greater probability of being preyed upon. Our results suggest that predation of simulated turtle nests may be a consequence of their distribution and location relative to the foraging activities of common nest predators, especially raccoons (Procyon lotor). Efforts to enhance recruitment among declining populations of turtles should consider the abundance and distribution of nesting habitat. Providing additional nesting sites away from predator foraging habitats may reduce nest predation and increase the recruitment of hatchlings into a population.     

Phylogenetic diversity of endangered and critically endangered southeast Asian softshell turtles (Trionychidae: Chitra)

The intense exploitation of turtles in Asian markets has contributed to declines in turtle populations across the continent. Three-quarters of Asia’s turtles are threatened and half are endangered. A recent workshop on the Asian turtle crisis identified taxonomic studies of widespread species as a priority for research because these low risk species may include unrecognized, narrowly distributed taxa of much higher concern. Chitra indica is a widely exploited softshell turtle (family Trionychidae) found across southern Asia. Individuals from Thailand have been described as a separate species, Chitra chitra, but this has not been universally accepted, and many sources consider Chitra monotypic. Phylogenetic analysis of sequence data from the mitochondrial ND4 gene revealed three deeply divergent, monophyletic lineages within Chitra: C. indica, C. chitra, and a third unnamed form from Myanmar. This new form is probably Critically Endangered, which highlights the importance of systematic studies in determining conservation priorities.     

Road crossing structures for amphibians and reptiles: Informing design through behavioral analysis

Road traffic causes significant amphibian and reptile mortality, which could be mitigated through the installation of road crossing structures that facilitate safe passage, but only if reptiles and amphibians are willing to use them. Through a series of behavioral choice experiments with frogs and turtles, we examined how aperture diameter, substrate type, length, and light permeability influenced individuals’ preferences for specific attributes of crossing structures, and how individuals responded to various heights of barrier fences. Snapping turtles (Chelydra serpentina), green frogs (Rana clamitans), and leopard frogs (Rana pipiens) preferred larger diameter tunnels (>0.5 m) whereas painted turtles (Chrysemys picta) preferred tunnels of intermediate (0.5-0.6 m) diameter. Green frogs preferred soil- and gravel-lined tunnels to concrete- and PVC-lined tunnels. Painted turtles showed non-random choice of different lengths of tunnel, possibly indicating some avoidance of the longest tunnel (9.1 m); although no species preferred to exit via the longest tunnels (9.1 m), members of all four species used such tunnels. Green frogs preferred tunnels with the greatest light permeability. Fences 0.6 m in height were effective barriers to green frogs, leopard frogs, and snapping turtles, whereas 0.3 m fences excluded painted turtles. We conclude that tunnels > 0.5 m in diameter lined with soil or gravel and accompanied by 0.6-0.9 m high guide fencing would best facilitate road crossing for these and likely other frog and turtle species.     

Incidental capture of marine turtles by the Italian trawl fishery in the north Adriatic Sea

In the north Adriatic Sea on-board observations on midwater and bottom trawlers were carried out during years 1999 and 2000. Results indicate that the north Adriatic Sea, and especially the north-east part, is a very important foraging and overwintering habitat for loggerhead turtles (Caretta caretta). Since the distribution of fishing effort in the area is not known, the total turtle catch by the Italian bottom trawl fleet was conservatively estimated from the lower catch rate observed in the south-west sub-area: 4273 turtle captures/year (95% C.I.=2186-8546). The actual total catch in the whole north Adriatic is likely to be much higher, due both to an unknown Italian trawling effort in the south-west part with multi-gear vessels and to the Italian and Croatian trawling efforts in the north-east part, where catch rate was 15 times greater than in the south-west part. In the south-west sub-area 9.4% of captured turtles were dead and potential mortality (assuming that all comatose turtles would die too) was 43.8%. Hence, trawling in the north Adriatic is likely to represent a serious threat to the populations, and possible conservation measures are discussed.     

Long-term conservation efforts contribute to positive green turtle Chelonia mydas nesting trend at Tortuguero, Costa Rica

Worldwide, green turtle Chelonia mydas populations have declined and the species is classified as globally endangered. Tortuguero, Costa Rica, hosts the largest remaining green turtle rookery in the Atlantic basin. Tortuguero green turtles have been hunted since pre-Columbian times. Monitoring and conservation of the green turtle population began in 1955. The long-term efforts provide an excellent opportunity to evaluate the success of sea turtle conservation action and policies. Nest counts conducted 1971-2003 were analyzed to: (1) determine the nesting trend, (2) estimate rookery size and (3) identify events and policy decisions influencing the trend. A nonparametric regression model indicates a 417% increase in nesting over the study period. Rookery size was defined as the mean number of nests 1999-2003 and estimated at 104,411 nests year⁻¹, corresponding to 17,402-37,290 nesting females year⁻¹. A comparison with 34 index populations verifies Tortuguero as one of the two largest green turtle rookeries worldwide. Events and policy decisions in Costa Rica, Nicaragua, and Panama that comprise the main nesting, feeding and mating grounds for the Tortuguero population are likely to have had the greatest influence on green turtle survivorship. Conservation efforts and policies catalyzing increased hatchling production and decreased adult and juvenile mortality since 1963 have contributed to the positive nesting trend. The trend demonstrates that long-term conservation efforts can reverse nesting declines and offers hope that adequate management can result in recuperation of endangered sea turtle species.     

Pelagic longlines: A threat to sea turtles in the Equatorial Eastern Atlantic

Pelagic longlines are widely known to interact with several species of sea turtles, and there is an increasing concern about the by-catch of turtles in commercial fisheries and its impact on their populations. However, information on sea turtle by-catch in the South Atlantic Ocean is scarce, and there are no quantitative by-catch data available on olive ridleys for the Equatorial Eastern Atlantic. In this paper we analyze data collected by observers on board an Uruguayan long-liner targeting swordfish in two areas in the Equatorial Eastern Atlantic: off the Gulf of Guinea and north of Saint Helenaa Island. Specimens of Lepidochelys olivacea and Dermochelys coriacea were hooked or entangled in 26 longline sets. All registered interactions with olive ridleys took place off the The Gulf of Guinea, with captures ranging from 1 to 3 specimens in a single set. The captured specimens, though not measured directly, appeared to be juveniles. In addition, the examination of the stomach contents of one female mako shark showed dermal scutes, vertebrae and the complete head of a sea turtle identified as L. olivacea, allowing us to estimate its curved carapace length. In contrast, adult specimens of D. coriacea were caught in the two fishing areas. The capture of 10 individuals in a single set was recorded. Due to the high rate of sea turtle by-catch observed off the The Gulf of Guinea (1.02 ind/1000 hooks) conservation programs in the area should take into consideration the possible existence of a developmental and feeding area in this zone. Accordingly, longline fisheries in this area should be monitored and mitigation measures put in place to avoid or minimize damage to the pelagic phase of African populations of sea turtles.     

Using habitat models to determine conservation priorities for pond-breeding amphibians in a privately-owned landscape of northern Idaho, USA

As rural landscapes experience increasing levels of residential development, the persistence of species that rely on them will depend on informed planning and management decisions. An understanding of habitat requirements is essential for setting priorities and developing landscape-level plans for the survival of these species. In many temperate rural landscapes, artificially created ponds may be the only wetlands available for aquatic reproduction by amphibians. The introduction of non-native fish into these ponds reduces survival and prohibits successful reproduction of many native pond-breeding amphibians. We surveyed 105 randomly selected wetlands in a primarily privately-owned, rural landscape in north Idaho, USA, for pond-breeding amphibian larvae in 2004 and 2005. We used an information theoretic multimodel inference and an algorithmic (random forests) approach to model habitat for each species based on local and landscape characteristics. We also used a mail survey to quantify how landowners value fish in their wetlands and their plans for future wetland development and fish stocking. Sixty-seven percent of pond owners reported that fishing in their pond was at least slightly important to them and 36% of owners indicated that they were at least 50% likely to add fish to their ponds in the next 5-10 years. Landscape change predictions for this area indicate that forests will become more open due to thinning; habitat models indicated that this is likely to be detrimental to long-toed salamanders and beneficial to Pacific treefrogs. Habitat models also indicate that Columbia spotted frog breeding sites consist of wetlands on flat ground with high solar insolation and that this species is sensitive to nearby development, indicating that as this landscape becomes further developed, this species may require habitat protection for persistence.     

Aerial surveying of the world’s largest leatherback turtle rookery: A more effective methodology for large-scale monitoring

For many marine megavertebrate species it is challenging to derive population estimates and knowledge on habitat use needed to inform conservation planning. For marine turtles, the logistics required to undertake comprehensive ground-based censuses, across wide spatial and temporal scales, are often insurmountable. This frequently leads to an approach where a limited number of index nesting beaches are monitored in great detail by foot. In this study we use nationwide aerial surveying interfaced with ground assessments across three seasons of leatherback turtle nesting in Gabon (Equatorial West Africa), highlighting the importance of a synoptic approach to marine turtle monitoring. These surveys allow the first complete population assessment of this nesting aggregation to be made, identifying it as the world’s largest for the species (36,185-126,480 clutches, approximating to 5865-20,499 breeding females per annum and a total estimate of 15,730 to 41,373 breeding females). Our approach also serendipitously provides insights into the spatial appropriateness of Gabon’s protected areas network, for example (mean ± 1SD) 79 ± 6% (range 67-86%) of leatherback turtle activities recorded during aerial surveys (n = 8) occurred within protected areas (345 km, 58%, of surveyed coastline). We identify and discuss sources of potential error in estimating total nesting effort from aerial surveying techniques and show that interannual variation in nesting is considerable, which has implications for the detection of statistically significant changes in population size. Despite its relative costliness per day, aerial surveying can play an important role in providing estimates of relative population abundance of large vertebrates dispersed over extensive areas. Furthermore, it can provide data on habitat use and deliver real-time information on the spatial efficacy of protected area networks.     

Habitat characteristics of adult frosted elfins (Callophrys irus) in sandplain communities of southeastern Massachusetts, USA

Changes to land use and disturbance frequency threaten disturbance-dependent Lepidoptera within sandplain habitats of the northeastern United States. The frosted elfin (Callophrys irus) is a rare and declining monophagous butterfly that is found in xeric open habitats maintained by disturbance. We surveyed potential habitat for adult frosted elfins at four sites containing frosted elfin populations in southeastern Massachusetts, United States. Based on the survey data, we used kernel density estimation to establish separate adult frosted elfin density classes, and then used regression tree analysis to describe the relationship between density and habitat features. Adult frosted elfin density was greatest when the host plant, wild indigo (Baptisia tinctoria), density was >2.6 plants/m² and tree canopy cover was <29%. Frosted elfin density was inversely related to tree cover and declined when the density of wild indigo was <2.6 plants/m² and shrub cover was ?16%. Even small quantities of non-native shrub cover negatively affected elfin densities. This effect was more pronounced when native herbaceous cover was <36%. Our results indicate that management for frosted elfins should aim to increase both wild indigo density and native herbaceous cover and limit native tree and shrub cover in open sandplain habitats. Elimination of non-native shrub cover is also recommended because of the negative effects of even low non-native shrub cover on frosted elfin densities. The maintenance of patches of early successional sandplain habitat with the combination of low tree and shrub cover, high host plant densities, and the absence of non-native shrubs appears essential for frosted elfin persistence, but may also be beneficial for a number of other rare sandplain insects and plant species.     

Identifying road mortality threat at multiple spatial scales for semi-aquatic turtles

Prioritizing sites for localized mitigation measures, and forecasting the effect of interventions on an endangered population, requires an understanding of the spatial scales at which threat processes operate. Road mortality is among the greatest threats to semi-terrestrial freshwater turtles due to the group’s life-history traits. Declining throughout much of their range, spotted (Clemmys guttata) and Blanding’s turtles (Emydoidea blandingii) are exposed to high road densities and traffic volumes in the northeastern United States. We examine the distribution of roadkill risk for spotted and Blanding’s turtles at three spatial scales. Tortuosity during upland movements was used to predict road-crossing locations at the single-movement scale. A gravity model of wetland-to-wetland interactions was then developed to identify road mortality hot spots at a broader road segment scale. Finally, road-crossing risk was assessed at the scale of focal areas that support distinct populations, using a population viability analysis to evaluate the consequences of road mortality on resident populations. The observed spatial variability of road mortality risk was high for single road crossing movements, limiting the effectiveness of static mitigation measures conducted at this scale. At the broader road segment scale, road mortality hotspots were evident. The demographic risk associated with roads varied widely among discrete populations, with probabilities of extinction over 100-year projections reaching 5.1% for spotted turtles, and 58.8% for Blanding’s turtles. We conclude that conservation interventions are most likely to be effective in mitigating the effects of road mortality when implemented at the road segment and population scales.     

Partitioning of molecular variation at local spatial scales in the vulnerable neotropical freshwater turtle, Hydromedusa maximiliani (Testudines, Chelidae): implications for the conservation of aquatic organisms in natural hierarchical systems

Hydromedusa maximiliani is a vulnerable freshwater turtle endemic to mountainous regions of the Atlantic rainforest in southeastern Brazil. Random amplified polymorphic DNA (RAPD) were surveyed with the purpose of assessing the genetic structure and determining the partitioning of molecular variation in H. maximiliani across the natural spatial hierarchical scale of its habitat. The goal of the study was to integrate ecological data with estimates of molecular genetics diversity to develop strategies for the conservation of this freshwater turtle. Specimens were sampled from rivers and streams across three drainages. Nine of the 80 primers used generated 27 scoreable bands of which 10 (37%) were polymorphic and produced 16 RAPD phenotypes. Significant heterogeneity was found in the distribution of RAPD molecular phenotypes across the three drainages. Analysis of molecular variance for molecular phenotypes showed that the heterogeneity had a spatial structure since a significant amount (22%) of the total variance was attributable to variation among rivers and streams. Since the genetic variation of this turtle seems to be structured according to the natural hierarchical system of rivers and streams within drainages, it is suggested that local populations should be considered as separate management units.     

The case of the missing marmots: Are metapopulation dynamics or range-wide declines responsible?

In the mid-1990s, anecdotal reports of Olympic marmot (Marmota olympus) disappearances from historically occupied locations suggested that the species might be declining. Concern was heightened by the precipitous decline of the Vancouver Island marmot (Marmota vancouverensis), coupled with reports that climate change was affecting other high-elevation species. However, it was unclear whether the Olympic marmot was declining or undergoing natural extinctions and recolonizations; distinguishing between normal metapopulation processes and population declines in naturally fragmented species can be difficult. From 2002-2006, we used multiple approaches to evaluate the population status of the Olympic marmot. We surveyed sites for which there were records indicating regular occupancy in the later half of the 20th century and we conducted range-wide surveys of open high-elevation habitat to establish current and recent distribution. We used these targeted and general habitat surveys to identify locations and regions that have undergone extinctions or colonizations in the past 1-4 decades. Simultaneously, we conducted detailed demographic studies, using marked and radio-tagged marmots, to estimate the observed and projected current population growth rate at nine locations. The habitat surveys indicate that local extinctions have been wide-spread, while no recolonizations were detected. Abundance at most intensive study sites declined from 2002-2006 and the demographic data indicate that these local declines are ongoing. Adult female survival in particular is considerably lower than it was historically. The spatial pattern of the extinctions is inconsistent with observed metapopulation dynamics in other marmot species and, together with very low observed dispersal rates, indicates that population is not at equilibrium.     

Impacts of urbanization on Florida Key deer behavior and population dynamics

Rapid human population growth and urbanization have had a negative impact on species biodiversity. As competition for resources between man and wildlife continues, it is important to understand the effects of urbanization on species. Endangered Key deer (Odocoileus virginianus clavium) are endemic to the Florida Keys which have undergone rapid human population growth and development over the past 30 years. Our study objectives were to evaluate the impacts of urban development on Key deer habitat use, population dynamics, behavior, and body mass. We used data from two comprehensive studies on Key deer spanning 30 years to evaluate these changes. Our results suggest that Key deer have become more urbanized, using urban areas more today than they did 30 years ago. Contrary to our predictions, survival was higher for more urban deer than for less urban deer. Problems still exist with mortality factors heavily impacting some portions of the deer population including lower survival associated with less urban male deer. Analysis of Key deer body mass also was converse to our predictions as deer weights appear to have increased over time. Collectively, our results suggest that over the past 30 years Key deer have become more urbanized and that deer plasticity has allowed them to adapt and persist in an urbanizing environment. However, the future ability of Key deer to persist in an environment with continued urban development is unknown. At some threshold, urban development would become unsustainable and unlike other forms of habitat change or environmental disturbances, urban development is in most cases irreversible requiring careful planning in habitat conservation strategies.     

Widespread occurrence of an emerging pathogen in amphibian communities of the Venezuelan Andes

Many recent amphibian declines have been associated with chytridiomycosis, a cutaneous disease caused by the chytrid fungus Batrachochytrium dendrobatidis, but increasing evidence suggests that this pathogen may coexist with some species without causing declines. In the Venezuelan Andes, the disappearance of three anuran species during the late eighties was attributed to B. dendrobatidis. Recently, this pathogen was found to be prevalent in this region on the introduced American bullfrog, Lithobates catesbeianus. As a first step toward assessing the risk of amphibian communities to B. dendrobatidis in this region, we conducted a broad survey across multiple habitat types and an altitudinal gradient spanning over 2000 m. We diagnosed 649 frogs from 17 species using real time and conventional PCR assays, and recorded relevant abiotic characteristics of host habitats. Infection was detected in 10 native species of pond, stream and terrestrial habitats from 80-2600 m, representing nine new host records. L. catesbeianus was the most important reservoir with 79.9% of individuals infected and an average of 2299 zoospores. Among native frogs, Dendropsophus meridensis, an endangered species sympatric with L. catesbeianus, showed the highest infection prevalence and mean zoospore load (26.7%; 2749 zoospores). We did not detect clinical signs of disease in infected hosts; however, species such as D. meridensis may be at risk if environmental stress exacerbates vulnerability or pathogen loads. While surveillance is an effective strategy to identify highly exposed species and habitats, we need to understand species-specific responses to B. dendrobatidis to stratify risk in amphibian communities.     

European Bison habitat in the Carpathian Mountains

European Bison (Bison bonasus) barely escaped extinction in the early 20th century and now only occur in small isolated herds scattered across Central and Eastern Europe. The species’ survival in the wild depends on identifying suitable habitat for establishing bison metapopulations via reintroductions of new herds. We assessed European Bison habitat across the Carpathian Mountains, a stronghold of European Bison and one of the only places where a viable bison metapopulation may be possible. We used maximum entropy models to analyze herd range maps and habitat use data from radio-collared bison to identify key habitat variables and map European Bison habitat across the entire Carpathian ecoregion (210,000 km²). Forest cover (primarily core and perforated forests) and variables linked to human disturbance best predict bison habitat suitability. Bison show no clear preference for particular forest types but prefer managed grasslands over fallow and abandoned fields. Several large, suitable, but currently unoccupied habitat patches exist, particularly in the eastern Carpathians. This available suitable habitat suggests that European Bison have an opportunity to establish a viable Carpathian metapopulation, especially if recent trends of declining human pressure and reforestation of abandoned farmland continue. Our results also confirm the suitability of a proposed Romanian reintroduction site. Establishing the first European Bison metapopulation would be a milestone in efforts to conserve this species in the wild and demonstrate a significant and hopeful step towards conserving large grazers and their ecological roles in human-dominated landscapes across the globe.     

Threat and response: A decade of decline in a regionally endangered rainforest palm affected by fire and introduced animals

The wild population of the palm Ptychosperma macarthurii near Darwin, in monsoonal northern Australia, is regionally endangered and provides a focus to illustrate a range of issues pertinent to conservation of rainforest habitat. Surveys in 1990 found that several populations exhibited a polarised size class structure typified by large adults and small juvenile plants. Over the following decade, in the absence of wildfire and in a period of reduced disturbance from introduced buffalo (Bubalus bubalis), cattle (Bos indicus) and pig (Sus scrofa), sufficient small juvenile plants survived and grew so as to infill the intermediate size classes. Three stage (bifid, juvenile and adult) transition matrix models characterised the population as declining under all observed conditions (intrinsic rate of increase: unburnt + few animals 0.9850; unburnt + many animals 0.9584; burnt <1 year 0.8737; burnt 1-2 years 0.9146; burnt >2 years 0.9937). In the absence of fire, simulations conducted to explore management options revealed a positive rate of increase with exclusion of introduced animals. With only partial introduced animal control or supplementation with juvenile plants, the median rate of increase remained negative. The regional population is at risk by more frequent and more intense fire due to the invasion of exotic grass species and land use changes in the catchment which result in an increased drying of the rainforest habitat. Ongoing decline is the most likely outcome in the absence of effective management intervention.     

Metapopulation dynamics and conservation of the marsh fritillary butterfly: Population viability analysis and management options for a critically endangered species in Western Europe

This study investigates the dynamics and viability of a marsh fritillary butterfly Euphydryas aurinia metapopulation in a Belgian successional landscape. Based on capture-mark-recapture and winter nest census data, we first estimated demography (survival and recruitment rates, population size, density dependence) and dispersal parameters (emigration rate, effect of patch connectivity on dispersal, mortality during dispersal). Then using RAMAS/GIS platform, we parameterised a population viability analysis (PVA) model with these parameters to simulate the future of this metapopulation under different scenarios.The metapopulation does not seem viable even if natural reforestation is controlled by adequate management. In its present state, the patch system is not able to sustain enough individuals: due to the large temporal fluctuations in demographic parameters, a carrying capacity far higher than currently would be necessary to limit extinction risk to 1%, suggesting the existence of an extinction debt for the species in Belgium. The situation of E. aurinia appears much worse compared to two other fritillary species threatened in Belgium, for which similar PVA are available. It is therefore urgent to increase the carrying capacity of the patch system. How and where it is achieved are of secondary importance for the gain in viability: improvement of habitat quality through restoration, or increase of habitat quantity via enlargement of existing patches and/or creation of new habitat in the matrix. A regime of management based on regular re-opening and maintenance of habitat patches may be the only guarantee of long-term persistence for this critically endangered species in Belgium.     

The effect of habitat fragmentation on communities of mutualists: Amazonian ants and their host plants

The consequences of fragmentation for communities of mutualist partners are for the most part unknown; moreover, most studies addressing this issue have been conducted on plant-pollinator communities. We evaluated how the experimental fragmentation of lowland Amazonian rain forest influenced a community of ant-plant mutualists. We inventoried a total of 1057 myrmecophytes in four fragments and four continuous forest sites; the twelve plant species recorded were occupied by 33 ant morphospecies, of which 11 were obligate plant inhabitants. Neither plant species richness, ant species richness, nor total ant-plant density were significantly lower in forest fragments. However, eight of the plant species, including three of the four most common, had higher mean densities in continuous forest than fragments. Of these four species, only one (Cordia nodosa) had significantly different colonization rates between habitat types, with higher colonization rates of plants in fragments. This may be because the Azteca species it is associated with increases in abundance in forest isolates. Although our results suggest that communities of ant-plant mutualists are likely to persist in fragmented tropical landscapes 25 years after fragment isolation, most species are rare and populations sizes in fragments are extremely low. Environmental and demographic stochasticity could therefore limit long-term population viability. We suggest future studies focus on evaluating how fragmentation has altered herbivore pressure and the dispersal of ants and plants to fragments, since the interaction of these factors is likely to have the greatest impact on long-term patterns of population persistence.