Author: Jonathan Losos Page 92 of 133

Anolis lividus. Trunk anole? Trunk-ground? Trunk-crown? Photo by Jonathan Losos

Mainland anoles exhibit a great diversity in habitat use and morphology, a topic we have discussed previously on AA. For this reason, an analysis of patterns of evolution in habitat use across all anoles, not just mainland species, would be very welcome. Nicholson et al. step into the breach by presenting habitat categorizations for a large number of mainland species, as well as for most West Indian species, and then analyzing habitat evolution on their preferred phylogeny. Along the way, they coin a new term, “ecomode,” argue that the ecomorph concept is fatally flawed and should be discarded, and present a scenario for patterns of ecological diversification in both mainland and island anoles. Although I applaud the effort to understand ecological evolution in mainland anoles and welcome the attention this paper brings to an important and little-studied question, I find the conclusions unconvincing. In this post, I discuss whether the data are sufficient to create categories of habitat use and confidently assign species to them; in subsequent points I will discuss the analysis of habitat use evolution and Nicholson et al.’s critique of the ecomorph concept.

What is an “ecomode”? The term is not explicitly defined in Nicholson et al., but it appears to refer to different categories of habitat use. The problem with creating such categories and assigning species to them is two-fold. First, most anole species use a variety of different habitats. I like to say that you can find almost any anole anywhere sometimes. More specifically, most anole species use the trunks of trees, often at different heights, and most can be found on the ground occasionally. How, then, do you distinguish a trunk anole from a trunk-ground or a trunk-crown anole, or a trunk-ground from a grass-bush? Second, how can one make sure that a given species fits into a single category? Perhaps some species have a broader niche that encompasses multiple ecomodes, or perhaps a species slices up the environment in an entirely different way (e.g., a trunk-bush or twig-ground species)?

Previous workers (including me) have been able to define ecomorphs and categorize species for two reasons. First, the ecomorph categories are defined not just on the basis of habitat use, but also by reference to morphology and behavior. Indeed, the morphological differences between ecomorphs are quite clear, and they correlate strongly with habitat use and behavior. One may quibble with a few assignments (e.g., is A. opalinus a trunk-crown or trunk anole?), as I discuss in Chapter 3 of Lizards in an Evolutionary Tree, but for the most part, assignment to ecomorph category is clear-cut (including the category of “non-ecomorph” for the minority of West Indian species that fail to meet the morphology/behavior/ecology criteria of any of the ecomorph categories).

The second reason we can make these assignments is because we have quantitative data that can be statistically analyzed. By contrast, the Nicholson et al. assignments are subjective decisions based on a reading of the literature, often relying on short summaries in broad regional reviews such as Savage’s (2002) The Amphibians and Reptiles of Costa Rica and Henderson and Powell’s (2009) Natural History of West Indian Reptiles and Amphibians. Use of these summaries is problematic for two reasons. First, although some mainland species have been studied extensively and quantitatively (e.g., the work of Vitt, Fitch, and Andrews), the habitat use of many species is not well studied. As a result, evaluating some summaries can be difficult because one does not know the extent and quality of the underlying data—in some cases (not Savage or Henderson and Powell), I suspect summary statements are not based on any hard data at all, but just qualitative impressions. In addition, even when species have been studied extensively, going from an encapsulated summary of such studies to an ecomode categorization is often not straightforward. For these reasons, the Nicholson et al.’s assignments of species to specific habitat use categories in many cases may not be reliable.

West Indian Non-Ecomorph Species

I will illustrate these problems by first discussing Nicholson et al.’s treatment of West Indian non-ecomorph species. For these species, there are a number of errors resulting from trying to interpret summary information provided in overview volumes.

Read More

We’ve had a lot of great discussion about Nicholson’s et al.’s proposal to split Anolis into eight genera. To date, most of the commenters have been against the proposal; I’d like to explain why I agree with this majority view.

Anole Annals summarized the arguments for splitting Anolis several days ago. Nicholson et al. argue that the failure to divide Anolis in the past has inhibited evolutionary and systematic research:

“Systematic progress in this regard has been delayed by an extremely conservative taxonomic approach to recognizing the diversity within the group and its extraordinarily ancient historical roots.” (p.4)

“The current practice (following Poe, 2004) of treating all dactyloids as comprising a single genus underemphasizes the evolutionary diversity within the family (as currently recognized) and obfuscates major biological differences among clades. In addition, simply because of the large size of the family (nearly 400 valid species), the single genus concept can be a hindrance to scientific communication regarding evolutionary events and directions of future research.” (p.13)

These quotes suggest that research on anoles is being held back by treating the entire clade as a single genus, but where is the evidence for these claims? No examples are provided. Quite the contrary, research on anoles has flourished over the last several decades, making it a well-known group for the study of many diverse evolutionary phenomena, and much of this work has explicitly incorporated phylogenetic information. Indeed, anole evolution, considered in a phylogenetic context, has become a commonly cited textbook example of adaptive radiation, and work on anoles has become so broad and deep that one commenter at last year’s Evolution meetings noted that “I didn’t go to the Evolution meetings for three years…When I “returned” in 2011 in Norman, it was like everybody had switched to working on anoles and sticklebacks!” The Dobzhansky Prize winners at the last two Evolution meetings have conducted phylogenetically-based research on anoles, and anole workers have nabbed the Fisher Prize and four Young Investigators Prizes at the meetings in that time span. Anole research is going gang-busters, and it is hard to see how retaining the name Anolis for the entire clade has had any sort of detrimental effect. (see also comments by Eric Schaad on why taxonomic names are no longer important for conducting phylogenetically-based evolutionary studies and by Yoel Stuart on why splitting evolutionarily-interesting clades may actually impede research).

I disagree with the proposal to split Anolis into eight genera for two reasons. First, it is not possible for the Linnean classification system to fully represent phylogenetic relationships—splitting genera simply changes the information conveyed, gaining some bits of information and losing others (for more discussion on this point, see the recent post by Luke Mahler and ensuing commentary). Second, splitting Anolis will be extremely disruptive for scientific researchers and the public.

Read More