Our brains are fast to picture differences between males and females, probably because these differences are everywhere in nature. Popular examples might include the manes and antlers of lions and deer, which are specific to males, or maybe the striking sexual differences in coloration in many birds (think of the classic peacock example). Besides these illustrative cases, size differences between the sexes might be even easier for us to accept as a common pattern in nature. Males are, on average, larger than females in elephants, pheasants, iguanas, and in our own species. Females, however, take the role of the larger sex in other animals like spiders, snakes, and many bird species. In any case, I think one could conclude that we are familiar with the idea of males and females being phenotypically different, and that we are especially aware of sexual differences in size.

Going a step further, I think we even have a general idea of why these differences exist, at least in the case of size differences between sexes (which I will now refer to as sexual size dimorphism, or SSD) and especially in the case where males are larger than females. “Males are larger because they need to be stronger” might be a stereotyped idea, but it is ultimately true in many cases as males usually need to physically compete with other males to get access to reproduction.

Now, if we go a step further, I think things start to get muddy. To recapitulate: (1) we understand that there are size (and other) differences between males and females and (2) we have some idea of why. But an idea that is vastly unknown to people is that SSD can vary among related species and populations. Male deer are larger than female deer, probably because of the need to be physically stronger (or appear to be so), but why are some deer populations more dimorphic than others?

That is the question at the heart of our new paper, in which we studied the variation in SSD and shape dimorphism across several populations of the green anole (Anolis carolinensis), a species that shows male-biased SSD (i.e., males are larger than females), and whose males are known to show territorial and aggressive behavior in the context of intrasexual competition. So, we know that in these lizards (1) males are larger than females and (2) we have a good idea of why, but we don’t really know how or why SSD varies among populations.

Finding patterns in variation: Rensch’s rule

Biologists have been fascinated by the evolution of SSD for a long time and have tried to determine what might cause variation in SSD among related species and populations. Thanks to all those decades of previous research, we actually know there are some patterns of SSD that appear repeatedly in nature, meaning that sometimes the variation in SSD actually follows certain rules. The most famous of these patterns is Rensch’s rule, which originally stated that male-biased SSD increases with species size. Previous works (e.g., De Lisle and Rowe, 2013) suggest that Rensch’s rule can arise when selection on male body size has been stronger than on female body size, like cases where a large size is of utter importance in male-male competition, for example.

There is some evidence pointing towards anoles following Rensch’s rule, which makes sense as most species show male-biased SSD and some degree of territoriality (e.g., Siliceo-Cantero et al., 2016). However, there is no direct evidence of the rule being followed intraspecifically in A. carolinensis, so we tested and confirmed this in the paper. Notice that Rensch’s rule is confirmed by a slope <1 in a regression of female size on male size:

Anolis carolinensis follows Rensch’s rule. The solid black line indicates a 1:1 relationship such that populations falling in that line should have, on average, males and females of the same size (none does, because males are larger than females in all populations). In this plot, a slope <1 when regressing female on male size indicates Rensch’s rule, meaning that as the average body size in a population increases, male-biased sexual size dimorphism (distance between the points and the 1:1 line) also increases. We ran two sets of analyses in the paper, with and without two particular populations from Florida (read paper for details), and in both cases Rensch’s rule was confirmed.

This is a good first step! We now know that variation in SSD in A. carolinensis follows a pattern. Now we can continue asking ourselves what processes can result in such pattern.

Mixing rules

The evolution of SSD will naturally depend on factors affecting size evolution, especially when males and females respond differently to them. Let me introduce you to another macro-pattern: Bergmann’s rule. This rule states that larger sizes should evolve at higher latitudes and/or colder climates among related species or populations as a way to optimize the temperature conservation of the body. Nonetheless, some evidence suggests that the rule should be inverted for squamates, which depend on external sources of energy to regulate their body temperature. Focusing on this latter scenario, imagine how body size increases towards the equator. Now, what if it does so differently for males and females? Specifically, what if the rate of size increase is higher for males than for females? Exactly, we end up with a latitudinal version of Rensch’s rule (see Blanckenhorn et al., 2006 for a pioneering test of this idea).

The test

In our paper we tried to disentangle the causes for the observed pattern of Rensch’s rule in A. carolinensis. One possibility is that populations of the species follow the reversed Bergmann’s rule, meaning that larger sizes are favored at lower latitudes, while simultaneously male body size is being affected by sexual selection. In this case, both male and female body size should increase towards the equator as a response to environmental selection, but males should do so at a faster rate because of the additional effect of constant sexual selection (as the example at the end of the previous paragraph). Notice that in this scenario SSD is lower at higher latitudes because there is a conflict between sexual selection, favoring larger sizes in males, and environmental selection, favoring smaller sizes (see panels C and G in the figure below).

Alternatively, environmental selection might not have any direct effect on body size, and only a latitudinal version of sexual selection might be at work. For example, sexual selection might be more intense at lower latitudes with higher temperatures, causing males to evolve larger sizes towards the equator and, consequently, higher SSD (check Tarr et al., 2019 for a nice test of this hypothesis). The only difference between this and the previous scenario should be the latitudinal patterns of female body size. In the absence of environmental selection, which should normally act on both sexes, female body size shouldn’t be related to latitude or temperature (see panels D and H in the figure below).

Two alternative hypotheses. In C and G: Sexual size dimorphism is driven by the simultaneous effect of climate-associated selection and sexual selection. In D and H: Sexual size dimorphism is driven by the sole effect of climate-associated sexual selection acting on male body size.

Our results

Although it took me some lines to reach this point, the main results are actually pretty simple. Of the two proposed scenarios, our results support the one in which the main driver of SSD evolution in A. carolinensis is latitudinally-variable sexual selection (i.e., sexual selection favors larger body size in males, but its strength varies with latitude/temperature) as female body size was not related to environmental temperature:

The origin of Rensch’s rule in Anolis carolinensis. Only male body size (black dots) changes with temperature (here PC1climate), supporting one of the hypotheses.

Interestingly, one could say in this case that Rensch’s rule is a consequence of only males following the inversed Bergmann’s rule (increase in size with temperature), although not for reasons related to temperature conservation.

Bonus: shape dimorphism

In this paper we also measured shape dimorphism considering three sets of traits: head, limbs, and pelvis/trunk. I will not go too deep into this part, but I will say that we were interested to see how shape dimorphism in these traits was related to SSD. We had some hypotheses regarding these relationships. For example, if the process behind the Rensch’s rule pattern in populations of the green anole was temperature-dependent sexual selection, we could expect shape dimorphism in head traits (which male lizards use to fight each other) to align with SSD (i.e., both SSD and head dimorphism should be positively related). We had no strong expectations regarding shape dimorphism in limb and pelvis/trunk morphology, but what we found was nonetheless interesting:

Relationships between size and shape dimorphism. Panels show sexual shape dimorphism (SSHD) in head (A), pelvis/trunk (B), and limb traits (C). Solid lines represent significant fitted models.

Head shape dimorphism was positively associated with SSD, supporting the latitudinal sexual selection hypothesis. No pattern was found for pelvis/trunk traits, and limb dimorphism was actually negatively related to SSD. Some more extended discussion of these patterns can be found in the paper, but I particularly like the inverse relationship between SSD and limb dimorphism. Maybe this points toward alternative ways green anoles use to avoid intersexual competition for microhabitat use (maybe both differences in size and differences in limb morphology allow males and females to use separate spatial niches)? As usual, only more research will tell…

Some things I learned from working on this project:

  • Macroecological/macroevolutionary rules are awesome. They seem very simple and easy to understand, but put two of them together and things get messy quite fast. Their initial simplicity, however, makes these interactions fun to work on.
  • Size and shape tell different stories. This might be almost a cliché, but looking at patterns of size alone probably means we will be overlooking other processes that are evident only by looking at shape, and vice-versa.
  • Going back to old concepts can be rewarding. I’ve been working on macroevolutionary/macroecological rules as part of my research for the last few years and lately I have noticed in the literature a renewed interest in certain old concepts like Rensch’s rule. The interesting thing is that these new waves of research on old ideas seem to uncover new and valuable knowledge, as if the original concept needed to rest for some decades before experiencing a burst of progress. I personally find this fascinating.

Thanks for reading!

References

-Blanckenhorn WU, Stillwell RC, Young KA, Fox CW, Ashton KG. 2006. When Rensch meets Bergmann: does sexual size dimorphism change systematically with latitude? Evolution 60: 2004–2011.

-De Lisle SP, Rowe L. 2013. Correlated evolution of allometry and sexual dimorphism across higher taxa. The American Naturalist 182: 630–639.

-Siliceo-Cantero HH, García A, Reynolds RG, Pacheco G, Lister BC. 2016. Dimorphism and divergence in island and mainland anoles. Biological Journal of the Linnean Society 118: 852–872.

-Tarr S, Meiri S, Hicks JJ, Algar AC. 2019. A biogeographic reversal in sexual size dimorphism along a continental temperature gradient. Ecography 42: 706–716.

-Toyama KS, Mahler DL, Goodman RM. 2022. Climate shapes patterns of sexual size and shape dimorphism across the native range of the green anole lizard, Anolis carolinensis (Squamata: Dactyloidae), Biological Journal of the Linnean Society 2022; blac136.

Latest posts by Ken Toyama (see all)