Is the Martinique Ground Snake Erythrolamprus Cursor Extinct?

Is the Martinique ground snake Erythrolamprus cursor extinct?

S TEPHANE C AUT and M ICHAEL J. JOWERS

Abstract The Caribbean Islands are a biodiversity hotspot throughout Martinique during the th and th centuries where anthropogenic disturbances have had a significant (Moreau de Jonnès, ). It was last observed on the impact, causing population declines and extinction of en- Martinique mainland in , when a single individual demic species. The ground snake Erythrolamprus cursor is was caught near Fort-de-France. There are two potential a dipsadid endemic to Martinique; it is categorized as reasons for its decline: people may have mistaken it for Critically Endangered on the IUCN Red List and is known the venomous lancehead Bothrops lanceolatus, which may only from museum specimens. The snake was common on have led to its eradication, and the small Indian mongoose Martinique during the th and th centuries but there Herpestes javanicus auropunctatus, an invasive predator, have been no reliable sightings since , suggesting it was introduced to the West Indies at the end of the th cen- may have gone extinct, probably as a result of the introduc- tury, resulting in declines and extirpations of reptile species tion of the small Indian mongoose Herpestes javanicus aur- (Henderson, ). opunctatus. However, the islet known as Diamond Rock, To the south-west of Martinique, c.  km from the coast, south-west of Martinique, is mongoose-free and the last re- lies a volcanic islet (spanning . ha, with a maximum ele- ported sighting of E. cursor there was in . The islet was vation of  m; Fig. a; Plate ) known as Diamond Rock. last occupied during the Napoleonic Empire (the early th The last published record of E. cursor is associated with century), is now completely protected, and is difficult to ac- this unique ecosystem (the snake was observed there in cess (it spans . ha, with a maximum elevation of  m). ; Lazell, ), and the last specimen was collected on We conducted the first extensive survey of the islet, over the islet in  (Breuil, ). Diamond Rock was occupied  days, to clarifty the status of E. cursor. Our study revealed by the English during the Napoleonic Wars (), when that unique conditions exist on Diamond Rock (i.e. aridity c.  men lived there for  months, sleeping in caves. and a distinct potential prey community) and that E. cursor Since this period of occupation few records of Diamond would have had to modify its ecology to persist on the islet. Rock have been kept but it seems to have remained uncolo- Although the rugged terrain of Diamond Rock makes it dif- nized by the mongoose. The site is now fully protected and ficult to explore, it is probable that E. cursor is now extinct. people have been banned from landing on the islet since . Fishermen and tourist boats anchor near Diamond Keywords Alien species, conservation, Diamond Rock, Rock, and there have been reports of snakes basking on Erythrolamprus cursor, extinction, Liophis cursor, shoreline rocks. However, several -day surveys for the spe- Martinique, West Indies cies failed to locate it (Breuil, ; M. Dewynter, pers. he Caribbean Islands are considered to be a biodiversity comm.). The lack of any confirmed sightings of the snake  hotspot because they harbour a high concentration of since suggests that it is probably extinct (Honegger, T  endemic species that are threatened by habitat loss and in- ). The aim of this research was therefore to carry out ’ vasive species (Myers et al., ). Of the region’s herpeto- an extensive survey to elucidate the species status on fauna, % of frogs and % of reptiles are endemic Diamond Rock. (Hedges, ). The Critically Endangered endemic dipsa- Because of the difficulty of exploring the steep slopes of did snake Erythrolamprus cursor (previously known as Diamond Rock, and given its protected status, we surveyed –  Liophis cursor, Lacépède, , Plate ) was common the islet during June as a two-person team. We used several methods to search for all potential traces of E. cursor. Firstly, eight snake-specific intercept traps (cylin- drical cages  ×  cm with cones at the extremities, STEPHANE CAUT (Corresponding author) Consejo Superior de Investigaciones  Científicas, Departamento de Etología y Conservación de la Biodiversidad, modified after Linnell et al., ) were set in various micro- Estación Biológica de Doñana, C/Americo Vespucio, s/n (Isla de la Cartuja), habitats (near rocks, tree trunks, crevasses and at the base of Sevilla E-41092, Spain, and ANIMAVEG Conservation, 58 Avenue Allende,   Villejuif F-94800, France. E-mail [email protected] stone walls; Fig. b) and were checked twice per day ( trap- nights total; Table ). Some traps were baited with live mice MICHAEL J. JOWERS Centro de Investigação em Biodiversidade e Recursos   Genéticos, Universidade do Porto, Vairão, Portugal captured on the islet ( trap-nights; Table ). Secondly, we Received  December . Revision requested  January . searched for snakes, snake eggs and sloughed skin in all Accepted  February . First published online  June . the various habitats on the islet both during the day and

Oryx, 2016, 50(3), 545–548 © 2015 Fauna & Flora International doi:10.1017/S0030605315000228 Downloaded from https://www.cambridge.org/core. IP address: 170.106.33.19, on 01 Oct 2021 at 08:18:01, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605315000228 546 S. Caut and M. J. Jowers

PLATE 1 Museum specimen (MNHM .) of the dipsadid PLATE 2 Diamond Rock. Photograph by Stephane Caut. racer Erythrolamprus cursor, endemic to Martinque. Photograph by Stephane Caut.

FIG. 1 (a) Location of Diamond Rock off the coast of Martinique; (b) map of Diamond Rock from the period of English settlement (Commodore Hood’s dispatch,  August ; Stuart & Eggleston, ), with the locations of the eight trapping sites (filled black circles).

at night ( hours per day per researcher). To examine E. We did not detect any signs of the snake. The intercept cursor’s potential trophic relationships with other species traps functioned well because we captured all the reptile spe- on Diamond Rock (predator–prey and/or competitive inter- cies that we observed during our visual surveys (Table ). actions) we compiled an inventory of the species we ob- Although some remote areas of the islet remain unexplored served during the survey. To confirm the presence of for logistical reasons, our survey covered more of Diamond invasive rodents we used INRA rat traps ( ×  ×  cm) Rock’s surface than previous surveys. We ascended steep and mouse traps ( ×  ×  cm) baited with peanut butter. slopes, whereas previous researchers focused on accessible Trap lines were established in various habitats (six sites; rocks at low elevations (Lazell, ), surveyed for only a Fig. b),  and  m apart for rats and mice, respectively. few hours, and did not explore the islet extensively. In reports For logistical reasons we could only use  traps (six rat written during the English occupation of Diamond Rock traps and eight mouse traps), which we moved between (Stuart & Eggleston, ) snakes were described as abun- sites. Trapping was conducted for  or  consecutive nights dant near the edge of the water tank (site , Fig. ). The last on each line transect (a total of  trap-nights; Table ). time E. cursor was observed on Diamond Rock was in July

TABLE 1 The number of trap-nights with rodent traps (for rat and mice traps, see text for details, combined) and with interception traps for snakes (with number of nights on which traps were baited with live mice), with the species caught in each, at the eight survey sites (see numbered locations on Fig. ) on Diamond Rock (Fig. ).

Rodent traps Interception traps No. of trap-nights No. of trap- (no. of trap-nights with Site nights Species trapped live mice bait) Species trapped 1213Mus musculus 6 (0) 2213M. musculus 21 (3) 9 Sphaerodactylus vincenti;2Anolis roquet 3211Gecarcinus ruricola 6 (0) 4 21 21 (3) 2 S. vincenti;5Gecarcinus ruricola 5 6 (0) 6212M. musculus 4 (2) 7 8 (0) 8143M. musculus 8 (0) 2 Gymnophthalmus pleei Total 119 11 M. musculus, 1 G. ruricola 80 (8) 11 S. vincenti;2A. roquet;5G. ruricola;2G. pleei

 (Lazell, ). Weather and temperature conditions Another cause of the snake’s disappearance could be the during our survey in June (including some precipitation) introduction of non-native predators. Diamond Rock was should have been favourable to observing the snake. Some granted full protection only in . During our survey seabirds (e.g. bridled tern Onychoprion anaethetus, red- we found no evidence of the presence of superpredators billed tropicbird Phaethon aethereus, brown noddy Anous on the islet (e.g. mongoose or feral domestic cat). We did stolidus) breed on Diamond Rock during this season, not capture any rats Rattus spp., and based on our experi- which could have enhanced the likelihood of spotting the ence trapping rodents on other oceanic islands we are con- snake if it had adapted to consume eggs or juvenile birds. fident there are no rats on Diamond Rock and probably Proving the absence of a species is notoriously difficult, never have been. However, mice appear to be present at and previous work on Erythrolamprus ornatus on Maria high densities. They can have devastating, irreversible and Major Island, Saint Lucia, suggests that snakes are difficult ecosystem-changing effects on islands and are known to to survey, with as few as one located per  days of searching consume invertebrates, seabird eggs and juvenile seabirds (J. Daltry, pers. comm.). Although, it seems likely that E. (Angel et al., ). On Diamond Rock mice could have af- cursor is extinct, what allowed the species to be abundant fected the snake population directly by consuming juvenile on Diamond Rock in the th century and persist until the snakes or snake eggs, or indirectly by decreasing prey avail- s? Given the aridity on the islet and its distinct prey ability. Egg clutches of E. cursor may have contained no communities E. cursor would have had to modify its ecology more than five eggs (Arlington & Henderson, ). to persist on the islet. Although the genus Erythrolamprus However, as the mouse population dates from the time of contains  species little is known about their ecology and the English occupation in , why was the snake still pre- diet, and living E. cursor have never been studied. Studies sent on the islet , years ago? of the species’ trophic ecology (Henderson & Bourgeois, Lack of data on the species’ abundance in the sis ; Henderson, ) were based on the stomach contents a significant limitation to evaluating the cause of the spe- of preserved specimens, indicating consumption of frogs cies’ extinction. There may never have been a large popu- (Eleutherodactylus spp.) and lizards (Anolis spp.) and a lation of E. cursor on Diamond Rock, or the population small proportion of insects. The closest living related spe- may have decreased during the English occupation. If E. cies, Erythrolamprus juliae (Jowers et al., ), has a similar cursor on Diamond Rock lived at the same density as E. diet, comprising mainly anurans (Schwartz & Henderson, ornatus on Maria Major (.–. individuals per ha; ; Breuil, ). There have not been any reports of J. Daltry, pers. comm.), this would imply , individuals frogs on Diamond Rock and we found no evidence of in total. Such a low-density population would probably them, and therefore E. cursor probably underwent a dietary suffer from inbreeding depression (e.g. reduced fertility, shift to prey on the species available (e.g. four reptile species; congenital defects, reduced resistance to disease) and be Table ). We found that the Martinique anole Anolis roquet vulnerable to stochastic forces such as a skewed sex was common, especially at lower elevations. Other potential ratio, natural disasters and chance fluctuations in repro- prey present at high densities included seabird chicks, ju- duction and mortality (Traill et al., ). Alternatively, venile Antillean fruit-eating bats Brachyphylla cavernarum, the disappearance of E. cursor may have resulted from and invasive house mice Mus musculus. a particular event (e.g. climatic change, species invasion)

that took place between the s and , when the HEDGES, S.B. () Biogeography of the West Indies: an overwiew. In islet became fully protected. Biogeography of the West Indies, Patterns and Perspectives (eds C. – Although the complexity of the Diamond Rock ecosys- A. Woods & F.E. Sergile), pp. . CRC Press, Boca Raton, USA. HENDERSON,R.W.() Consequences of predator introductions and tem makes it difficult to explore, we conclude that E. cursor habitat destruction on amphibians and reptiles in the post-Columbus may now be extinct, given that  days of extensive survey- West Indies. Caribbean Journal of Science, , –. ing and trapping yielded no signs of its presence. It is more HENDERSON, R.W. () Lesser Antillean snake faunas: distribution, difficult to declare a species extinct than to describe or redis- ecology, and conservation concerns. Oryx, , –.  cover species. Erythrolamprus cursor is just one example of HENDERSON, R.W. & BOURGEOIS, R.W. ( ) Notes on the diets of West Indian Liophis (Serpentes: Colubridae). Caribbean Journal of several reptile species known to survive only on one small Science, , –. island, at a low density (e.g. E. ornatus, Alsophis antiguae, HONEGGER, R.E. () List of amphibians and reptiles either known Cnemidophorus vanzoi, Ameiva polops). There is an urgent or thought to have become extinct since . Biological need to assess these species and to ensure that their habitats Conservation, , –. are free from invasive species and other potential threats. JOWERS, M.J., CAUT, S., GARCIA-MUDARRA, J.L., ALASAAD,S.& INEICH,I.() Molecular phylogenetics of the possibly extinct The invasion of oceanic islands by non-native predators Martinique ground snake. Herpetologica, , –. can lead to dramatic effects on island ecosystems and LAZELL, J.D. () Wiederentdeckung von zwei angeblich cause numerous extinctions. From a conservation perspec- ausgestorbenen Schlangenarten der westindischen Inseln. tive, the Caribbean, a biodiversity hotspot with a large num- Salamandra, , –. ber of species at risk of extinction, is one of the most LINNELL,M.A.,ENGEMAN,R.M.,PITZLER,M.E.,WATTEN,M.O., WHITEHEAD,G.F.&MILLER,R.C.()Anevaluationoftwo important areas in the Western Hemisphere. designs of stamped metal trap flaps for use in the operational trapping of brown tree snakes (Boiga irregularis). The Snake, , –. MOREAU DE JONNÈS,A.() Monographie de la Couleuvre couresse Acknowledgements des Antilles Coluber cursor de Lacépède. Journal de Physique, , –. We thank L. Juhel, I. Ineich and M. Breuil for their help in MYERS, N., MITTERMEIER, R.A., MITTERMEIER, C.G., DA FONSECA, organizing the survey, and A. Bérubé-Lampron for the docu- G.A.B. & KENT,J.() Biodiversity hotspots for conservation mentary films. We also thank Jessica Pearce-Duvet for her priorities. Nature, , –. English editing services. Appropriate ethical procedures SCHWARTZ,A.&HENDERSON, R.W. () Amphibians and Reptiles of were employed during this research. SC was granted approval the West Indies. Descriptions, Distributions, and Natural History. University Press of Florida, Gainesville, USA. to engage in animal experimentation (R-GRETA-F–) STUART,V.&EGGLESTON, G.T. () His Majesty’s Sloop-of-War: by the French Minister of Agriculture. DEAL Martinique Diamond Rock. Hale, London, UK. (J. Mailles, N. Boulard and C. Barnérias), the Conservatoire TRAILL, L.W., BRADSHAW, C.J.A. & BROOK, B.W. () Minimum du Littoral, and ONCFS provided legal support viable population size: a meta-analysis of  years of published (DEAL-) and allowed us access to Diamond Rock estimates. Biological Conservation, , –. and its facilities. This work was funded by the Consejo Superior de Investigaciones Científicas (JAE postdoctoral contract to SC) and the NGO Animal Conservation. Biographical sketches

S TEPHANE C AUT studies the ecology and conservation of threatened References species. Much of his work focuses on ecosystem functioning and the key role played by vertebrate species and/or invasive species. Using a ANGEL, A., WANLESS, R.M. & COOPER,J.() Review of impacts of multidisciplinary approach (field, laboratory, modelling, and experi- the introduced house mouse on islands in the Southern Ocean: are mental), he seeks to understand biological and trophic interactions mice equivalent to rats? Biological Invasions, , –. among species. He specializes in isotopes, using them to address ARLINGTON,J.&HENDERSON, R.W. () Communal nesting site in species- and ecosystem-level conservation challenges across a wide the snake Liophis juliae in Dominica, West Indies. Caribbean range of taxa and habitats. M ICHAEL J OWERS studies evolutionary Journal of Science, , –. biology, applying population genetics, phylogeography, and biogeog- BREUIL,M.() Histoire naturelle des amphibiens et reptiles raphy. He uses the phylogenetic species concept to assess the tax- terrestres de l’archipel Guadeloupéen: Guadeloupe, Saint-Martin, onomy of poorly described reptiles and amphibians. Most of his Saint-Barthélemy. Patrimoines Naturels No. . Institut d’Ecologie work has focused on reptiles and amphibians native to the islands et de Gestion de la Biodiversite, Paris, France. of Trinidad and Tobago but he has also worked on other organisms, BREUIL,M.() The terrestrial herpetofauna of Martinique: past, including mammals, social insects, scorpions, freshwater fish and present, future. Applied Herpetology, , –. parasites.