SICB 2014: Dominance Interactions in Green Anoles

Anolis carolinensis from Miami. Photo by J. Losos.

Anolis carolinensis from Miami. Photo by J. Losos.

I’m often a little skeptical of studies that suggest how lab results can have implications for natural systems without actually examining the problem in nature, and studies that address the same question in both lab and field are still rare. That it investigated the same broad question in the lab, in enclosure-style experiments, and in the field is what made Trinity University student Jordan Bush’s poster remarkable.

Bush was interested in the traits associated with dominance, and began by running a tournament of agonistic interactions between a set of male Anolis carolinensis. She used a number of different Markov Chain Monte Carlo algorithms, widely used in predicting winners of sports tournaments, to convert pairwise fight outcomes into individual ranks of “fight-winning ability.” Bush found that rank could be predicted by size-corrected head length, as well as the propensity for aggressive behaviours such as  push-upping and crest-raising.

But do these same traits predict dominance in nature? Not exactly. Bush took the question to the field, and found that none of the traits that predict dominance rank in the lab are correlated with territory size in the wild. In the final piece of the puzzle, Bush constructed experimental enclosures to measure the territory sizes of males with known ranks (by conducting another dominance tournament). As predicted by the first two parts of the study, territory size was not correlated with rank.

By harnessing the power of both lab and field experiments to observations made in nature, Bush’s study will change the way behavioural ecologists think about territory formation. I’ve always assumed that winning agonistic encounters is the means by which anoles increase the sizes of their territory, but like everything in nature, it’s more complicated than that!

Leave a Reply

Your email address will not be published.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

Optionally add an image (JPEG only)