Carrot Rock and the Endemic Anolis ernestwilliamsi

Carrot Rock, a small protrusion of British Virgin Island, links the southern end of Peter Island to the edge of the shelf constituting the Puerto Rico Bank. This <1.3 hectare, steeply sloped island is home to two endemic squamate species: the Carrot Rock Skink (Mabuya macleani) and Ernest Williams’ anole (Anolis ernestwilliamsi). This is a somewhat surprising situation, given the proximity of Carrot Rock to Peter Island (400m) and its recent connection to the latter by a breaking shoal (water depths are but 2-3 m between the two). Hence, separation of Carrot Rock was likely recent, occurring as early as the end of the Wisconsin Glaciation (~8000 yrs ago) or at nearly any point more recently, likely within the last 3000 years (suggested by Mayer and Lazell 2000).

Carrot Rock, British Virgin Islands. This 1.3 hectare island is steeply sloped, with an elevation of ~25 m asl and a very steep aspect on all sides. There are no landing areas and the island must be reached by swimming. Obtaining a beachhead and summiting require exertion and great care.

Carrot Rock, British Virgin Islands. This 1.3 hectare island is steeply sloped, with an elevation of ~25 m asl and a very steep aspect on all sides. There are no landing areas and the island must be reached by swimming. Obtaining a beachhead and summiting require exertion and great care.

Nevertheless, morphological distinction has resulted in the specific epithets for these lizard species. The Carrot Rock Skink was described by frequent AA contributors Greg Mayer and Skip Lazell (Mayer and Lazell 2000) based on unique coloration and color pattern. The species was recognized in Blair Hedges and Caitlin Conn’s tome on West Indian skinks (Hedges and Conn 2012)–indeed, they used the node subtending M. macleani and other Virgin Island species as a calibration point. Recent analysis (Pinto-Sánchez et al. 2015) has suggested this species (along with other Virgin Island species), is (are) minimally divergent from the widespread M. sloanii complex. As the species was described based on morphology and appears to exhibit little genetic variation owing to a recent separation, species delimitation based on molecular data will surely point to collapsing these species and hence this latter finding is unsurprising.

Carrot rock is dominated by seagrape (Cocoloba uvifera) and the vine Stigmophyllon periplocifolium, with two large branching Pilosocereus royenii cacti on the crown. The majority of the anoles occur on the windward slope, where a few Cocoloba are sheltered enough to grow to heights of 1-3 meters.

Carrot rock is dominated by seagrape (Cocoloba uvifera) and the vine Stigmophyllon periplocifolium, with two large branching Pilosocereus royenii cacti on the crown. The majority of the anoles occur on the windward slope, where a few Cocoloba are sheltered enough to grow to heights of 1-3 meters.

Anolis ernestwilliamsi is very much a close relative of the widespread A. cristatellus. The endemic species is notable (and specifically recognized) largely for its increased lamellae number, color pattern, and apparently larger body size (Lazell 1983). It was described, again, by Skip, who is likely one of the few of us to have visited the island (and certainly the most frequent visitor). This description was published in Ernest Williams’ festschrift (Rhodin and Miyata 1983), in which, by my count, A. ernestwilliamsi is one of four nominate species named in honor of Ernest. As with the Carrot Rock Skink, molecular evidence suggests that A. ernestwilliamsi is minimally, or perhaps not at all, distinct from the widespread relative (A. cristatellus). Mitochondrial genetic analyses (Strickland et al., in review) demonstrate that A. ernestwilliamsi is nearly identical to many Puerto Rico Bank A. cristatellus haplotypes, suggesting a very recent maternal common ancestor (not surprising). Nuclear DNA has not yet, to my knowledge, been studied, likely owing to a lack of suitable (or available) DNA samples from the island. Concomitantly, several recent studies have demonstrated rapid evolution of key morphological traits in both Anolis sagrei (Stuart el al. 2014) and A. cristatellus (Winchell et al. 2016), including lamellae number, in response to presumed shifts in selection associated with either competitor species (Stuart et al. 2014) or non-natural substrate use (Winchell et al. 2016).

Female Anolis ernestwilliamsi. In a 1.5 hour survey around 1200h I counted fewer than 12 females.

Female Anolis ernestwilliamsi. In a 1.5 hour survey around 1200h, I counted fewer than 12 females.

Turning back to Carrot Rock itself, we might suspect that selection differs on this small island, and that selection would act rapidly in the face of the (presumably; Lazell 2005) small effective population size. This shifting of phenotype, owing to either plasticity or underlying allelic shifts, represents the processes of genetic drift and selection acting on a small population. This is an expected scenario, but leads to the question of how we like to recognize lizard species. As I teach my Zoology students, and as we all know, this is a tricky question. Anolis ernestwilliamsi is phenotypically distinguishable from other populations of A. cristatellus (Lazell, 1983). Some (myself included) might argue that this limited morphological distinctiveness is insufficiently diagnostic of speciation given the lack of genetic distinctiveness and the overall degree of morphological variation in the species. Nonetheless, some (Dmi’el et al., 1997) have examined whether the population of A. ernestwilliamsi is behaviorally and physiologically adapted to an arid and exposed habitat, implying an adaptive evolutionary response resulting in phenotypic evolution despite very recent separation and genetic similarity. That these authors found a similar physiological response (evaporative water loss rates) and that Carrot Rock is really not ecologically different from Peter Island (or most of the coastal portions of the BVI), further support the idea that the population is not terribly distinct.

Male Anolis ernestwilliamsi. In a 1.5 hour survey around 1200h I counted only 3 adult males.

Male Anolis ernestwilliamsi. In a 1.5 hour survey around 1200h, I counted only 3 adult males.

With all of this in mind, and having recently been to Carrot Rock, I remain skeptical regarding the prospects for continued recognition of A. ernestwilliamsi, despite the desire to see Ernest continue to have an Anolis namesake. Nevertheless, this should not (and indeed, didn’t/doesn’t) diminish the joy of seeing this population grasp tenaciously to existence on this speck of beautiful land.

 

 

References
Dmi’el et al., 1997. Biotropica 29:111-116.
Hedges, S.B. and C. Conn. 2012. Zootaxa 3288
Lazell, J. 1983. In: Rhodin and Miyata.
Lazell, J. 2005. Island: fact and Theory in Nature. University of California Press.
Mayer, G.C. and J. Lazell. 2000. Proceedings of the Biological Society of Washington 113:871-886.
Pinto-Sánchez N.R., et al. 2015. Molecular Phylogenetics and Evolution 93:188-211.
Rhodin, A.G.J. and K. Miyata. 1983. Museum of Comparative Zoology, Harvard University.
Stuart, Y.E., et al. 2014. Science 346:463-466.
Winchell, K.M., et al. 2016. Evolution 70:1009-1022.
[disclosure, I am an author on some of the papers mentioned in this article]

About Graham Reynolds

Graham is an Assistant Professor at the University of North Carolina Asheville. His research focuses on Caribbean herpetology- specifically anoles and boas.

3 thoughts on “Carrot Rock and the Endemic Anolis ernestwilliamsi

  1. Simply looking at male meiotic chromosomes might hint at a likely answer…because if ernestwilliamsi is different from cristatellus, they might have trouble producing viable offspring.

    A. gundlachi, the three grass anoles, and cooki all have male 2n=29; squash a gonad and you will see 13 bivalents and one (sex) trivalent.

    A. cristatellus (and scriptus) have one fewer pair of chromosomes.

    So … squash a few gonads.

    If ernestwilliamsi has the “typical” male 2n=29, they might hybridize with other cristatellus , but it’s likely that the f-1 hybrids would be sterile because of screwed-up meiosis.

    Of course if it has the “typical” cristatellus karyotype, male 2n=27..it tells you nothing about whether or not ernestwilliamsi is a “good” species.

    I’ll admit, I’m generalizing from the single case I worked on where something like this occurred (Gorman et al. Reproductive failure among hybridizing Anolis lizards of Trinidad. Systematic Zoology 20: 1-18 (1971)

  2. I hope that I’m not being overly technical here, but you’re a little off on the timing of the Wisconsin Glaciation. The International Commission on Stratigraphy places the end of the Wisconsinan (and thus the beginning of the Holocene) at 11,700 years ago.

  3. Thanks George and William! If A. e’e’w’si is not a species then it is just a spectacularly divergent and absolutely distinct, apparently reproductively isolated, independent evolutionary lineage. Got a category for that? Skip

Leave a Reply

Your email address will not be published. Required fields are marked *

Optionally add an image (JPEG only)