Jamaican boas

Population Genetics and Conservation of the Jamaican boa

We initiated in 2006 a conservation genetics project on the Jamaican Boa (or "Yellow Snake", or "Nanka", Epicrates subflavus) in collaboration with Christophe Remy (Vivarium of Tournai, Belgium), coordinator of the captive breeding program in Europe. The captive breeding program was initiated in the 1970s at the Durrell Wildlife Conservation Trust (Jersey Zoo, UK).

Our responsibilities in the project included:

1. •The isolation and characterization of microsatellite loci;
   

2. •Use of these molecular markers on wild individuals to:
   

3. ‣Measure genetic variability of remnant natural population(s),
   

4. ‣characterize the population structure of the species in Jamaica,
   

5. •Use of these molecular markers on captive individuals (kept in >14 institutions across Europe) for:
   

6. ‣Assessing reproductive success among breeders & identify cases of multiple paternity,
   

7. ‣Measuring overall genetic variability of the captive-bred population,
   

8. ‣Suggesting modifications to the breeding program (to maximize genetic diversity),
   

9. ‣Improve the breeding program by retaining the population structure observed in the wild,
   

10. ‣Allocate breeders to natural populations and identify the most relevant potential releasing site of offspring.
    

This work would not have been possible without the efficient collaboration of scientists, privates, and institutions in Jamaica. Our special thanks to:

1. •NEPA (National Environment and Planning Agency), including Yvettes Strong (head of the "Biodiversity Branch"), Canute Tyndale, Ricardo Miller, and Andrea Donaldson;
   

2. •Susan Koenig and Michael Schwartz (Windsor Research Center, Cockpit Country, Jamaica);
   

3. •Wendy Lee (Seven Oaks Sanctuary for Wildlife, Runaway Bay, Jamaica);
   

4. •Orlando Robinson (Hope Zoo, Kingston, Jamaica).
   

Please, consult the text below as well as the following full publications for references and much additional information on our conservation genetics work on Jamaican boas:

1. ✓Tzika A. C., Koenig S., Miller R., Garcia G., Remy C. & M. C. Milinkovitch
   Population structure of an endemic vulnerable species, the Jamaican boa (Epicrates subflavus)
   Molecular Ecology 17, 533-544 (2008)
   

2. ✓Tzika A. C., Remy C., Gibson R. & M. C. Milinkovitch
   Molecular Genetic Analysis of a Captive-Breeding Program:  The Vulnerable Endemic Jamaican Yellow Boa
   Conservation Genetics 10:69–77 (2009)
   

Persecutions of the Jamaican boa

The Jamaican (or ‘Yellow’) Boa (Epicrates subflavus) is endemic to Jamaica and is the island’s largest native terrestrial predator. Although it remained abundant from the arrival of the first Europeans in Jamaica in the 1500s until the 19th century, the condition of the species deteriorated dramatically in the following years. Rats reached Jamaica with the earliest European ships and rapidly developed into a large population causing highly significant destruction in sugar cane fields (then the main commercial product of the island). In an attempt to control this pest, several exogenous species were deliberately introduced in the island: a carnivorous ant from Cuba (Formica omnivora), the cane toad (Bufo marinus), and the European ferret (Mustela putorius). These efforts failed to control the Rattus populations. Subsequently, nine small Asian mongoose (Herpestes javanicus) individuals were brought to Jamaica in 1872 to initiate a population that was expected to ‘extirpate the whole race of the vermin’. Ten years later, the result was disastrous: mongooses had indeed a significant impact on the rat population but also devastated multiple native avian and reptilian species. Similarly, populations of the Jamaican boa have been declining dramatically ever since the introduction of mongooses. Anecdotal observations further suggest that Jamaican boas may be vulnerable to the toxic bufogenins secreted by cane toad. Introduced species are not the only threat to Jamaican boas. Indeed, despite the fact that the snake species may actually be beneficial (rats are a component of the boa’s diet in forest edge/farm habitat), most Jamaican people have a traditional antipathy to reptiles in general and to snakes in particular, such that snakes are usually killed on sight. The snakes are mistakenly considered venomous and are sometimes viewed as evil creatures in this often strongly religious culture. In addition, human activities contribute to the deterioration of the local flora and fauna. Only 8% of the island’s land cover is considered minimally disturbed, while an additional 23% encompasses ‘disturbed’ forest degraded by various types of human activities, ranging from small-scale agriculture to open-pit bauxite mining. Such activities, along with conversion of forest for urban development and tourism, have greatly fragmented the remaining natural habitat. In-situ conservation of the Jamaican boa is seriously hindered by the lack of information on demographic and ecological parameters as well as by a poor understanding of the population(s) distribution and structure. Given the secretive nature of the Jamaican boa and the difficulties to perform fieldwork in its natural habitat, the latest in situ surveys were performed in the 1980s and 1990s on the basis of questionnaires rather than on direct observations. These studies confirmed the localized and patchy distribution of the species and its disappearance from multiple localities. Epicrates subflavus is considered today a ‘Vulnerable’ species by IUCN (Red List).

Molecular analysis of the Jamaican boa captive breeding program

Given the signs of decline of the natural population(s), a captive breeding programme was initiated in 1976 at the Durrell Wildlife Conservation Trust (Jersey, UK) and was rationalized in 2002 by incorporating pedigree information into a European Endangered species Programme (EEP) involving 14 member institutions of the European Association of Zoos and Aquaria (EAZA; www.eaza.net). Recently, using nuclear and mitochondrial DNA markers, we identified parental allocation errors and ambiguities in the EEP studbook, and assessed the genetic diversity and levels of inbreeding of the current captive population. Combining measures of relatedness derived from multilocus genotypes with practical parameters such as age of animals and localization of host institutions, we proposed a scheme of mating groups that would produce minimal inbreeding in the captive population of the Jamaican boa (Tzika et al. 2008). Check HERE the results of this study.

March 2006 sampling campaign

A sampling campaign was organized across the island in March 2006.
A report of this field trip is available here in PDF format. A sample photo gallery is available HERE (All pictures are copyrighted by Michel Milinkovitch 2006, unless specified otherwise).

Animation showing the different sites that we visited during our sampling campaign.

This campaign was also the opportunity to perform education activities in collaboration with NEPA, the Windsor Research Center, the Hope zoo, and the Seven Oaks Sanctuary for Wildlife. We designed a poster for explaining this project to the public in Jamaica as well as to the public of the European zoos participating to the captive breeding program. The poster can be downloaded here as: low resolution PDF (for screen viewing) and high resolution PDF (for printing).

Molecular genetic analysis of the yellow boa natural populations in Jamaica

Using nine nuclear microsatellite loci and a fragment of the mitochondrial cytochrome b gene, we present the first molecular genetic analyses focusing on the structure and diversity of the natural population(s) of the Jamaican boa. Our results provide insights and guidance for an objective management of the species through both in-situ and ex-situ approaches.

In March 2006, we collected samples (blood, scale clips, and shed skin) in Jamaica from 46 individuals, some of which having been maintained (but not bred) in captivity by zoos, conservation institutions, or private individuals. We also incorporated samples from (i) 36 individuals previously collected on the island in 1995, and (ii) 5 of the wild-born individuals that have been used as the founders of the captive breeding programme. Importantly, all 87 individuals were wild-born and captured opportunistically during the last three decades, such that it is very unlikely that they are particularly closely related. We processed all 87 samples for molecular analysis, but reliable information on their geographical origin was available for only 38 sampled individuals.

We constructed a microsatellite-enriched genomic DNA library for the Jamaican boa, and nine polymorphic nuclear microsatellite loci were selected. Using a previously available E. subflavus cytochrome b sequence, we designed primers for the amplification of a 647-bp fragment.

We used Bayesian model-based methods (“Structure v2.1” and “Baps v4.14”) for inferring population structure and assigning individuals to populations. We used multiple network estimation methods (statistical parsimony as implemented in “tcs”; the median-joining network approach as implemented

in “network v2.0”; and the union of most parsimonious trees method as implemented in “combinetrees v1.0”) to estimate genealogical relationships among individuals based on their cytochrome b sequences. See our Genealogical Network Inference page for a discussion on the relative merits and limitations of the different network estimation approaches.

We estimated genetic diversity and population genetic parameters using the programs “Convert”, “fstat v2.9.3”, “Arlequin v3.1”, and “Spagedi v1.2”.

To test the monophyly of the E. subflavus species and root the haplotypic network, we extended the cytochrome b data set by including the new haplotypes generated from the samples presented here, and we performed Bayesian as well as metaGA phylogeny inferences. The metaGA (see the phylogeny inference methods we developed) is an evolutionary computation heuristics (i.e. implementing a set of operators that mimic processes of biological evolution) that vastly improves the speed and efficiency with which maximum-likelihood trees are found and yields a probability index for each branch.

Results: (1) population structure & network estimation

As the exact geographical origin of many (58%) of the sampled Jamaican boa individuals is unknown or uncertain (many of the individuals sampled in zoos or at privates were not associated to reliable information on their exact location of capture), the “Structure” analysis was particularly relevant because it allows clustering of the individuals without the need for a priori geographical information.

Figure a on the right shows the average and variance of log-likelihood among five independent runs for one to six populations, and Figure b shows the corresponding distribution of #K values. Both statistics suggest that a clustering of individuals in three groups best fits the data. Most individuals are consistently assigned to a single group with a high probability (color figure a). Visual inspection of the population assignment values indicates that only five individuals (indicated with an asterisk in color figure a) out of 87 have variable group assignments among the five independent runs (K = 3). When, a posteriori to the “Structure” analysis, we examined the localities some of the sampled individuals are known to originate

from, we noticed that all individuals assigned to the first and second clusters (blue and red) were known to originate from localities at the western and central parts of the island, whereas all individuals known to come from St Thomas parish (at the most eastern part of Jamaica) are assigned to the third cluster (yellow). The “Baps” admixture analyses based on the three partitions inferred by “Structure” provide a good support for that clustering: only five individuals (indicated with circles in color figure a) are associated with significant admixture (P < 0.05), meaning that they are assigned to a cluster with a probability significantly lower than those of the other individuals assigned to the same cluster. Three of these five individuals showed assignment problems with “Structure” as well.

The 87 sequences of the 647-bp cytochrome b fragment define 16 polymorphic sites and 12 distinct haplotypes. The TCS, median-joining, and UMP approaches all inferred the haplotypic network shown in color figure c. Two haplogroups are well differentiated and are separated by four to six fixed mutations (depending on the resolution of loops in the network). Haplogroup I (color figure c) includes seven haplotypes belonging to individual snakes that are (i) known to originate exclusively from the western and central portions of the island (color figure b), and (ii) assigned to groups 1 (blue) and 2 (red) (as inferred from the corresponding microsatellite data; color figure a). One exception is haplotype 10; although most individuals bearing it are assigned to group 2 (blue) on the basis of their multilocus microsatellite genotype, two individuals exhibiting that haplotype are assigned to group 3 (yellow). Haplogroup II, is even more geographically heterogeneous; although most individuals are assigned to group 3 (yellow) and originate from St Thomas parish at the extreme east of the island, haplotypes H4 + H5 and H3 occur in individuals originating from more western portions of the island and assigned to groups 1 (blue) and 2 (red) on the basis of their multilocus microsatellite genotypes. In other words, although the major split in the haplotypic network is generally consistent with the microsatellite-based group assignment (haplogroup I includes haplotypes occurring mostly in individuals assigned to group 1, whereas haplogroup II includes haplotypes occurring mostly in individuals assigned to group 3), a few individuals exhibit a mixture of an ‘eastern’ genotype and a ‘western-central’ haplotype or vice versa. The biological relevance of this pattern is examined in the discussion section.

Genetic diversity statistics, both for nuclear microsatellite and mitochondrial cytochrome b data, as well as FST analyses, are described in details in our full publication.

Results: (2) Phylogeny inference

Bayesian and metaGA analyses indicate that the Jamaican boa forms a monophyletic species (clade supported by a posterior probability of 77% and a metaGA branch support value of 95%, see Figure below). From our analyses, it was not possible to confirm that Epicrates fordi, a boa endemic to Hispaniola, is indeed the sister taxon to the Jamaican boa because both posterior probability and metaGA branch support values for the placement of that lineage are below 50%. Within the Jamaican

boa clade, haplotypes are separated into reciprocally monophyletic ‘haplogroup I’ and ‘haplogroup II’, although posterior probabilities supporting the monophyly of the former is higher than that of the latter. In other words, the haplotypic network presented in color figure c     above is most probably rooted either on the branch corresponding to the fixed differences between the two haplogroups, or on a branch within haplogroup II.

Discussion

As, respectively, about 40% and 20–25% of all vertebrate species are considered vulnerable and endangered, a comprehensive assessment of population structure is one critical component for their effective conservation. Endemic island species are at even greater risk, as 95% of reptilian species extinctions in the last 400 years have taken place on islands. Epicrates subflavus is a clear representative of such ‘high risk’ species. This endemic Jamaican snake is very poorly known, not only in terms of its population structure but also of its basic ecology. Given the rapid decline of the Jamaican boa and the constant deterioration of its natural habitat, it is of vital importance to gather objective demographic and population structure data for developing efficient in situ and ex situ conservation programmes. Here, we used both nuclear and mitochondrial DNA markers for the first assessment of the natural population structure of the Jamaican boa. Our analyses of nine nuclear microsatellite loci indicate the existence of three significantly differentiated groups, whereas our network inference analyses of the mitochondrial cytochrome b locus generated two significantly differentiated haplogroups. The two types of analyses underline an Eastern vs. (Western + Central) pattern of differentiation (group 1 vs. 3, and haplogroup I vs. II). On the other hand, individuals assigned to group 2 on the basis of their multilocus microsatellite genotypes are dispersed in the two haplogroups. Various causal factors that can explain the partial discrepancy between the patterns inferred from microsatellite genotypes and that obtained using mitochondrial haplotypes are associated with the different modes of inheritance for the two types of markers. For example, contrasting patterns between nuclear and mitochondrial DNA in terms of F-statistics and population assignment can be interpreted as evidence for male-biased dispersal that homogenizes allele frequencies among populations at biparentally, but not at maternally, inherited genetic markers. Although, no reliable information is available on the spatial ecology and home range of E. subflavus, we can extrapolate information available for a closely related species, Epicrates inornatus. Although the home range does not differ significantly between the sexes in this endemic Puerto Rican species, the mean daily and monthly movements recorded for males are greater than those of females. Such a mating system, where males perform active searches of receptive females, and females are philopatric, has been described for other snakes species within and outside the family Boidae.

Still, high pairwise FST values among populations and the occurrence of a mostly western-central haplogroup vs. a mostly eastern haplogroup, indicate that dispersal patterns (sex specific or not) are insufficient to disrupt this very significant differentiation in the distribution of genetic diversity of the Jamaican boa.

This pattern is in close agreement with findings originating from molecular and morphological analyses of other Jamaican species, such as frogs, terrestrial crabs, millipedes, and beetles. The occurrence of such a recurrent (across taxonomic groups) pattern of genetic differentiation correlates with the geological history of Jamaica. In the late Oligocene or early Miocene (15–20 million years ago), the island, which had been fully submerged for 33 million years, re-emerged through uplift of the Caribbean plate. The first emergent land areas were thick limestone (carbonate) sequences in the north-central region of the Clarendon Block (see Fig. on the right). Uplift next followed (and continues today) for the eastern Blue Mountain Block, formed of Cretaceous and Lower Eocene clastics and limestone, with an associated subsidence of its southern St Thomas Belt along the Plaintain Garden Fault. As the Blue Mountain Block was uplifted, it may have been initially separated from the central Clarendon Block by an older structural feature, the volcanic island-arc of the Wagwater Trough, which developed 65–50 million years ago but which was overlaid with limestone during Jamaica’s long submergence. Extreme western Jamaica, the Hanover Block, was likely a third emerging island, separated from the Clarendon Block by the Montpelier-New Market Trough. As uplift continued during the Lower Miocene and Pliocene (10–5 million years ago), the ‘islands’ of these three blocks coalesced into the present-day form of Jamaica. Of relevance for the Jamaican Boa population, the intrusion of the clastic Blue Mountains between the carbonate John Crow Mountains in the extreme east and the Port Royal Mountains located on the western flank of the Blue Mountains may constitute a major barrier to dispersal, either because of its high elevations (maximum 2256 m) and associated cool, misty climate, or because boas prefer habitats associated with a limestone substrate for as-yet-unknown reasons. In addition to geological events, variations in sea level in the late Pliocene and early Pleistocene period probably played a significant role in the separation between Eastern and Western Jamaica fauna.

Utility of these results

Jamaican reptilian and amphibian species have been poorly studied in comparison with those in other Caribbean islands. As far as the Jamaican boa is concerned, very limited data are available on the species ecology, population densities, habitat usage, and potential morphological differentiation among populations. Given the direct and indirect threats that these animals face, mainly resulting from human activities (persecution, predation by introduced species, natural habitat destruction and fragmentation), the molecular analyses we report here could be particularly important for delineating management units. Significant mtDNA divergence (possibly with reciprocal monophyly) among populations, significant differentiation at nuclear loci, as well as adaptive distinctiveness should all be taken into account for the definition of management units.  Our analysis based on multiple molecular genetic markers is the first attempt to uncover the levels of variability and differentiation of the natural populations of the Jamaican boa. We identify an Eastern vs. (Western + Central) pattern of differentiation demonstrating the importance of managing the Blue and John Crow Mountains separately from the populations of the West-central dry, mesic, and wet limestone forests. These latter habitats do, however, require comprehensive management at the landscape level not only to ensure protection of the full historic ecological range of habitats occupied by boas but also to ensure connectivity among populations across the landscape for the west-central haplogroup I. At present, scant attention has been given to managing remnant forest within a landscape context of ecosystem (and genetic) connectivity; instead, fragments are typically managed as isolated units. Furthermore, and unfortunately, coastal areas are being extensively destroyed and left with increasingly small and isolated patches of forest amidst a sea of development for high-density tourism and urban sprawl. Of concern is also the increasing threat of bauxite mining to Cockpit Country, the largest block of contiguous wet limestone forest on Jamaica and a recognized stronghold for the Jamaican boa. Open-pit bauxite mining of valleys not only results in an absolute reduction in forest cover but also severely isolates forests left on hillsides. The extensive road network built to access the mining pits further facilitates encroachment by humans for illegal timber extraction, increasing the risk of fatal encounters for boas. The reduction in population size and spatial extent of the Cockpit Country boa population could have profound consequences for the long-term viability of this species in Jamaica.

Continue with page 2

Back to top