Trait diversity and environmental variation: sex matters

Submitted by editor on 9 August 2016. Get the paper!

by Adam C. Algar
In this paper, we focused on how phenotypic diversity within a species changes across a diverse set of environments. We were particularly interested in how ‘harsh’ environments affect phenotypic diversity. Specifically, we tested two hypotheses: whether the position of populations in ‘morphospace’ shifts between environments, which we’d expect if different trait values were favoured in different environments, and whether some phenotypes are filtered out of harsh environments, leading to lower phenotypic diversity, and a smaller volume of occupied morphospace.
Of course, ecologists have been searching for the answers to these questions for decades, but what makes our work different is that we suspected that males and females may respond differently to changes in environmental conditions through space. This suspicion arose because sexual size dimorphism is common across animals (obviously, since we can usually tell males and females apart), males and females are thought to be affected by different selection pressures, and there is evidence that males often vary more, at least in size, than females (a loose version of ‘Rensch’s Rule’, though it’s a rule made to be broken). However, whether there are sexual differences in the biogeography of traits, or in how phenotypic diversity arises across environmental gradients, is unknown. Sex-specific responses to environmental variation could be important because if one sex doesn’t differentiate across environments, then this could constrain or slow ecological diversification, compared to when both species are affected by selection.
To test these hypotheses, we needed an abundant species found in a wide range environments that didn’t require sampling across a continent (though, in retrospect that would have been a lot of fun . . .). We found our answer in the generalist lizard Gallotia galloti on Tenerife, the largest island in the Canarian archipelago. Tenerife famously contains a diverse array of habitats, from low-elevation arid coastal scrub, to cool pine forest, to its harsh, high-elevation moonscape (snow in winter, raging sun and aridity in summer). This diversity arises from the island’s sub-tropical latitude, elevation gradient (the Teide volcano reaches >3000m), and the influence of the trade winds, which give very different characters to the north and south aspects of the island. Across all of these environments you can find G. galloti, a medium-sized, sexually dimorphic lacertid lizard. The genus Gallotia is endemic to the Canary Islands, and G. galloti is found only on La Palma and Tenerife. As a plus, it’s very abundant and easy to catch!
After a lot of driving around Tenerife, measuring a lot of animals, drinking a lot of coffee — and the odd beer (not while driving) — and a lot of null modelling, we had our answers: it seems that populations both shift in morphospace between environments and that harsh environments filter out a large portion of phenotypic diversity. At least for males. In contrast, females barely responded phenotypically to environmental differences at all.  This finding raises a couple of key further questions. Firstly, why? Well, at this stage, we don’t know, but we suspect that it has to do with the differential affect of sexual selection on males, who need to be large with big heads to win fights against other males. More broadly, however, such sex-specific responses could have implications for how confident we can be in our knowledge of trait-environment relationships, and spatial variation in functional trait diversity. Sex differences have been mostly ignored in such studies; we suspect that most have focused exclusively on males (I have no data to back that up, but it would be interesting to trawl through the literature and check . . .), which means we may be overestimating the effects of environmental gradients on functional trait, or phenotypic, diversity, or at least only looking at half the story.

Watch Adam talk about this paper here