Anole Annals

It’s not anoles but at least it’s about convergent evolution!

A recent study by Damien Esquerre and Scott Keogh published in Ecology Letters found that pythons and boas, the two famous constrictor snake families, have evolved convergent head shapes. The study was based on over 1,000 specimens and including most of the species. Pythons and boas that occupy the same micro-habitat or ecology (i.e. arboreal, terrestrial, semi-aquatic, semi-fossorial) look more like each other than to other snakes in their own family. This is exciting because it highlights how important ecology and adaptation is in shaping biological diversity.

Abstract:

Pythons and boas are globally distributed and distantly related radiations with remarkable phenotypic and ecological diversity. We tested whether pythons, boas and their relatives have evolved convergent phenotypes when they display similar ecology. We collected geometric morphometric data on head shape for 1073 specimens representing over 80% of species. We show that these two groups display strong and widespread convergence when they occupy equivalent ecological niches and that the history of phenotypic evolution strongly matches the history of ecological diversification, suggesting that both processes are strongly coupled. These results are consistent with replicated adaptive radiation in both groups. We argue that strong selective pressures related to habitat-use have driven this convergence. Pythons and boas provide a new model system for the study of macro-evolutionary patterns of morphological and ecological evolution and they do so at a deeper level of divergence and global scale than any well-established adaptive radiation model systems.

Reference:

Esquerré, D & J S Keogh. 2016. Parallel selective pressures drive convergent diversification of phenotypes in pythons and boas. Ecology Letters, 19(7): 800-809.

While doing some local herping for fun this weekend with a couple of friends visiting from out of town (Janson Jones of previous AA fame; 1,2,3,4,5), we happened upon this Cuban knight anole (Anolis equestris) with a fairly conspicuous hole in its dewlap. Despite this, the lizard appeared in prime condition. Other reports of strange dewlaps have been documented on AA before, such as these grey-dewlapped Puerto Rican crested anoles (A. cristatellus) and American green anoles (A. carolinensis), but has anyone ever seen any individuals with tissue missing from the core region of the dewlap (as opposed to injuries sustained on the peripheries, such as this Cuban brown anole (A. sagrei), which aren’t generally that uncommon)?

Here’s one example, from an AA post from four years ago.

Christian Perez is currently studying anoles at La Selva Biological Station in Costa Rica. Recently, he found a vine snake. Here’s his report:

As I reached towards the snake, I startled an anole (Anolis limifrons) that was hiding nearby. The snake turned rapidly, looked at the anole, and made movements synchronized with the wind to remain inconspicuous among surrounding vegetation. I stayed with them for a while, and after one failed strike, the snake successfully stalked and caught the anole. The snake took under a minute to consume the lizard, and it was very friendly after its meal.

Over at Lizards and Friends, Amy Payne from Michele Johnson’s lab reports on her first field experience studying green anoles. Fear not–they kept an eagle eye out for snakes.

On a fleeting one-night stopover in Miami last week, Anthony Geneva had the chance to pop in and say hello at Fairchild Tropical Botanical Gardens and take a morning stroll to view some of the resident anoles (see others posts about Fairchild anoles here: 1,2,3,4). While waiting to be joined by fellow local anolologist and distichus aficionado Winter Beckles (University of Miami), Anthony and I noticed some commotion by the edge of a nearby pond. Upon closer inspection, we realized that a tricolored heron (Egretta tricolor) appeared to be juggling a large anole in it’s mouth! In my morning rush, I had managed to forget not just my anole-catching noose pole, but alas, also my camera. Fortunately, Anthony was on hand to fill the David Bailey role.

After re-positioning the lizard a few times, the heron appeared to do something peculiar – it repeatedly dunked the lizard in and out of the water. This happened perhaps 5-6 times. Was this an attempt to expedite a fatality prior to consumption, or perhaps a neat trick to help lubricate such a large prey item?

In all, the process of ingestion took less than 10 seconds, following a couple of minutes of dunking and repositioning.

This observation follows a recent hot post reporting the predation of anoles by reintroduced whooping cranes (Grus americana) in Louisiana, which itself was preceded by various observations of avian-fuelled anolivory in South Florida (1, 2, 3, 4). Even more recently, while showing Thom Sanger and Bonnie Kircher around Fairchild Gardens a few weeks back, we observed a Cooper’s hawk (Accipiter cooperii), a widely-regarded bird specialist, snatch an American green anole (A. carolinensis) from the frond of a towering Royal Palm (Roystonea regia) – an event Rob Heathcote and I had observed the previous year with an adult male A. cristatellus in nearby Matheson Hammock. Unfortunately none of us were privileged with Anthony’s camera reflexes to capture any of those events.

So, why’s this interesting? (Excluding the obvious natural history enlightenment of revealing, at least personally, a previously unclear predator-prey interaction). Well, tricolored herons are a widespread breeding resident throughout much of the US Gulf states and as far south through the Caribbean to central Brazil and Peru. Therefore, the consumption of crested anoles (A. cristatellus) isn’t necessarily a novel interspecific interaction – it’s possible that this occurs in the native range of A. cristatellus, Puerto Rico, where both exist. However, although tricolored herons are natural residents of South Florida, it would be a tough sell to argue that crested anoles would be naturally on the menu. Crested anoles were first introduced to South Miami in the 1970s – the original site of introduction being a mere stone’s throw from this observation (for a review of the subsequent dispersal patterns of A. cristatellus in Miami see Kolbe et al. 2016; pdf here). So although crested anoles are being exposed to many novel biotic interactions in Miami, it seems they can’t escape some.

Have any Puerto Rico anolophiles observed this interaction before?

A smug bird.

The spectacular new anole species, Anolis eladio, named after Hispaniolan nature photographer extraordinaire Eladio Fernandez, who first discovered the species.

We have been remiss here at AA in not reporting on a recent monograph in Novitates Caribaea (the journal of the Museo Nacional de Historia Natural of the Dominican Republic) by Köhler and Hedges dividing the Hispaniolan green anoles into sixteen species, up from the previously recognized four. Specifically, Anolis chlorocyanus is split into four species, A. coelestinus into five species, and A. aliniger is subdivided into six species. Poor A. singularis remains as it is.

The analysis is based on mitochondrial DNA and morphological characters. The monograph is available online and should be consulted for the fine details. Appended below are the abstract and the heart of the methods.

Say what you may about the proliferation of new species (and word on the street is that this will not be the last word on green anole species diversity), some of the new species are spectacular in appearance and certainly there is more variation in this group than many may have realized.

Abstract:

We revise the species of green anoles (i.e., the species related to Anolis aliniger, A. chlorocyanus, and A. coelestinus) occuring on Hispaniola. Based on our analyses of morphological and molecular genetic data we recognize 16 species of green anoles, eight of which we describe as new species (A. apletolepis sp. nov., A. chlorodius sp. nov., A. divius sp. nov., A. eladioi sp. nov., A. gonavensis sp. nov., A. leucodera sp. nov., A. prasinorius sp. nov. and A. viridius sp. nov.) and three of which are raised from subspecific to species level (A. cyanostictus, A. demissus and A. pecuarius) and one is resurrected from synonymy with A. chlorocyanus (A. peynadoi). Because the six syntypes of A. chlorocyanus (MNHN 785, 787, 2007.2066–09) are conspecific with the only available syntype of A. coelestinus (i.e., MCZ 3347), we have petitioned the International Commission of Zoological Nomenclature (ICZN) to use its plenary power to set aside the type status of the syntypes of Anolis chlorocyanus and to allow the designation of a neotype in order to stabilize the current and long established usage of the names A. chlorocyanus and A. coelestinus. For each species we provide a standardized description of external morphology, color descriptions in life, color photographs in life, description and illustration of hemipenis morphology (if available), distribution maps based on the specimens examined, comments on the conservation status, and natural history notes. Finally, we provide a dichotomous key for the identification of the 16 species of green anoles occuring on Hispaniola.

And here’s how they did it:

For this study, we have examined a total of 787 specimens of green anoles from Hispaniola. Head length was measured from the tip of the snout to the anterior margin of the ear opening. Snout length was measured from the tip of the snout to the anterior border of the orbit. Head width was determined with the broad tips of the calipers aligned with the levels of posterior margin of eye and supralabial scales, respectively, with the calipers held in a vertical position relative to the head. Dorsal and ventral scales were counted at midbody along the midline. Tail height and width were measured at the point reached by the heel of the extended hind leg. Subdigital lamellae were counted on Phalanges II to IV of Toe IV of the hind limbs, and separately on distal phalanx. We considered the scale directly anterior to the circumnasal to be a prenasal. Abbreviations used are AGD (axilla–groin distance), dorsAG (number of medial dorsal scales between levels of axilla and groin), dorsHL (number of medial dorsal scales in one head length), HDT (horizontal diameter of tail), HL (head length), HW (head width), IFL (infralabials), IP (interparietal plate), SAM (scales around midbody), ShL (shank length), SL (snout length), SO (subocular scales), SPL (supralabial scales), SS (supraorbital semicircles), SVL (snout–vent length), TL (tail length),VDT (vertical diameter of tail), ventrAG (number of medial ventral scales between levels of axilla and groin), and ventrHL (number of medial ventral scales in one head length). In reporting the frequencies of character states, we used the following terminology (Köhler submitted): if a character state was present in more than 65% of the examined specimens, we coded it as “usually”; <65% but >20% “commonly”; <20% but >5% “occasionally”; and <5% “exceptionally”. The use of size categories also follows Köhler (2014): (1) small: <50 mm SVL; (2) moderate-sized: 50–60 mm SVL; (3) moderately large: 60–80 mm SVL; (4) large: 80–110 mm SVL; (5) giant: >110 mm SVL.

As lines of evidence for species delimitation, we apply a phenotypic criterion (external morphology: coloration, morphometrics, and pholidosis) and a criterion for reproductive isolation (genetic distinctness of the cytochrome B and ND2 genes). Sequences from 77 ingroup and two outgroup taxa were analyzed (a total of 2217 aligned sites). Alignments (MUSCLE) and best-fit model selection were performed in MEGA 6.06 (Tamura et al., 2013). A maximum likelihood (ML) analysis was performed using MEGA 6.06), unpartitioned, using the evolutionary model GTR + I + Γ. Gaps were treated as missing data. All parameters for the ML analyses were estimated by the program during the run. Branch support in the trees was provided by standard bootstrap analysis (2,000 replicates). A Bayesian phylogenetic analysis using MrBayes 3.2.2 (Ronquist et al., 2012) also was performed, also using the GTR + I + Γ model. The Bayesian analysis was set to two parallel runs for five million generations, sampled every 100 generations, each run employed three heated and one cold chain, with a temperature parameter of 0.10. The first 10% of samples were discarded as burn-in. Convergence was assessed by the standard deviation of split frequencies (< 0.01 in all cases).