Category: New Research Page 28 of 66

In a recent paper in The American Naturalist, Martha Muñoz, Johanna Wegener, and Adam Algar noted an interesting pattern in two clades of Caribbean anoles evolving independently on Cuba and Hispaniola: high elevation species tended to have smaller body sizes than lower elevation species*.

Figure 1: Body size-elevation relationships in Hispaniolan (cybotes clade) and Cuban (sagrei clade) anoles. The x-axis is elevation in meters (on the log scale). The y-axis is SVL, or snout-vent length, a measure of size. The colors represent individual species within each clade; grey represents A. cybotes and A. sagrei on Hispaniola and Cuba, respectively.

Having found that the two groups converged independently on a similar evolutionary pattern, the authors wanted to know: was the underlying evolutionary progression also the same?

To answer this question, the authors took advantage of the fact that the two clades harbored multiple species. By measuring body size-elevation patterns within each species, and then asking how those patterns combined with interspecific patterns to create the overall body size-elevation cline (SEC) observed across all species, Muñoz & Co. could discern subtle differences between clades in the evolutionary trajectory towards convergence. For example, one clade might build its overall size-elevation cline by having the same SEC relationship present in each species, with species also sorting themselves by elevation and size (Model H1). Whereas another clade might build its size-elevation cline just through interspecific differences in size and elevation, without an SEC relationship within species (Model H2).

Figure 2: Two models, of eight that the authors proposed, which might explain how body size-elevation clines evolve. Within-species clines are represented by different colored/dashed lines. Across-species clines are best visualized by drawing an imaginary line through average size and elevation of each species . In H1, each species has the same size-elevation relationship (i.e., the negative slope) and is found at different elevations. This creates a size-elevation relationship that depends on both intra- and interspecific patterns. In H2, each species has no size-elevation relationship (i.e., the flat slope) but is found at different elevations. Here, the size-elevation relationship is driven purely by interspecific differences in elevation and size.

The authors developed eight models for how elevation and size might be related within species and across species. They tested which of those eight models best explained variation in the relationship between size and elevation within species and clades, while accounting for spatial autocorrelation among collection localities and differences in elevational range among species . They then compared best models across clades to see whether convergence was reached by similar or different evolutionary pathways.

What did they find?

On Hispaniola, each species tended not to show any significant intraspecific SEC relationships: note the flat slopes of the dashed lines in the left panel of Figure 3. Instead, much of the overall SEC comes from interspecific differences in size and elevation, consistent with Model H2 (Figure 2).

On Cuba, in contrast, the authors found some significant within-species SEC relationships–the solid lines in the right panel of Figure 3–but found that interspecific differences in size and elevation explained very little of the clade’s overall SEC (Figure 4)**.

The model most consistent with the Cuban data.

Thus the authors answered their question: “Although the precise mechanisms underlying inverse size[-elevation] clines remains  unknown, it is clear that they were constructed in different ways on Cuba and Hispaniola.” In other words, the two clades show a pattern of convergence to small size, but they took different routes of intra- and interspecific evolution to get there. It reminds me of Yogi Berra’s response when asked directions to his house: “When you get to the fork in the road, take it!”

CITATION: M.M. Munoz, J.E. Wegener, and A.C. Algar. 2014. Untangling Intra- and Interspecific Effects on Body Size Clines Reveals Divergent Processes Structuring Convergent Patterns in Anolis lizards. The American Naturalist 184: 636-646.

* This pattern was measured from 16 Anolis species: nine in the sagrei clade (Cuba) and seven in the cybotes clade (Hispaniola). The finding of small body size at high elevations is the inverse expectation of Bergmann’s rule. Bergmann’s rule, as originally conceived, states that endothermic species living in colder climates should be larger (or have a larger surface area to volume ratio), all else equal, to conserve heat. As lizards are ectothermic, one would expect an inverse Bergmann. Perhaps we could call the inverse cline Nnamgerb’s rule? It does have a certain charm to it, no?

** I wonder if the authors might chime in in the comments section. What does it mean that a size-elevation cline wasn’t found on Cuba when using the mean size and elevation of each population (Fig 5 below), but it was found when species identity was ignored (Fig 1 above)? Is this an example of Simpson’s paradox?

If the biology of Anolis lizards is a puzzle, then a new paper by Yoel Stuart, Todd Campbell and colleagues is a crucial piece. It’s a puzzle piece that not only contains a wealth of information when held up on its own, but also brings clarity to a broader picture of anole biology when fitted into place.

Anolis carolinensis on small spoil-islands in Florida are the subject of Stuart et al. (2014)

A tight relationship between microhabitat and morphology characterizes variation across Anolis species in the Caribbean.  Anole biologists have long suspected that negative interactions, such as competition, are responsible for driving different species into different microhabitats, with subsequent morphological  adaptation to these microhabitats over evolutionary time. But pinning down interspecific interactions as the cause of evolutionary divergence in microhabitat and morphology has been difficult.

Why is establishing this causality challenging?  Upon observing a pattern of consistent differences between populations of a species that occur in sympatry and allopatry with an interacting species, it seems logical to attribute this pattern to the presence of the interacting species. But many processes other than an evolutionary response to negative interspecific interactions can generate such a pattern–environments may differ between sympatric and allopatric populations in a way that drives the observed divergence, individuals from sympatric populations may all be similar only because they  are closely related to each other, the divergence may be a consequence of phenotypic plasticity, or most dishearteningly, the whole pattern may simply be due to chance.

Ruling out these alternatives seems a gargantuan undertaking. Indeed, as Stuart and Losos (2013) point out, in a review that serves as a nice companion piece to this study, only a small fraction of studies describing patterns of divergence between sympatric and allopatric populations tackle the problem of eliminating these alternatives and can thus conclude with confidence that interspecific interactions cause the divergence they observe. But Stuart et al. (2014) take on the challenge.

Like recent research by Helmus et al. (2014) that exploits human-mediated anole dispersal to test classic principles of island biogeography, Stuart et al.’s (2014) research is rooted firmly in the Anthropocene. Occupying centrestage is the interaction between Anolis carolinensis, native to the United States, and Anolis sagrei, a relatively recent invader. The stage itself comprises small man-made spoil islands in Florida, created in the 1950s. When Todd Campbell began this study in the 1990s, A. carolinensis occurred on many of these little islands. Campbell introduced A. sagrei to three islands, and watched how, over the next three years, A. sagrei numbers rose steadily and A. carolinensis shifted higher into the trees on invaded islands, while continuing to perch at lower heights on nearby un-invaded islands.

Lead authors Yoel Stuart and Todd Campbell boating between spoil islands in FL

This rapid shift in microhabitat spurred Stuart and Campbell to return to the islands 15 years later (with a team of field assistants, of whom I was one!) to ask if A. carolinensis on invaded islands had subsequently diverged morphologically from conspecifics on un-invaded islands. By this time, A. sagrei had spread widely. Nevertheless, they found five un-invaded islands. A. carolinensis still perched lower on these un-invaded islands than on nearby invaded islands.

Across Caribbean anoles, species perching higher up on trees have larger toepads and more lamellae on these toepads than do species perching closer to the ground. Recapitulating this interspecific difference, Stuart et al. (2014) found that A. carolinensis on invaded islands had evolved larger toepads and more lamellae than lizards on un-invaded islands in about 20 generations, rapidly establishing a pattern of character displacement. But is this pattern caused by the presence of A. sagrei?

It seems almost criminal to squish into one paragraph everything that Stuart et al. (2014) did to rule out alternative explanations for the pattern of divergence. They reared hatchlings from invaded and un-invaded islands to rule out phenotypic plasticity as a cause for divergence, sequenced a mind-bogglingly large number of SNP loci to establish that A. carolinensis on invaded islands were not closely related to each other, and conducted intensive habitat surveys to rule out environmental differences between invaded and un-invaded islands. This mountain of work supports the idea that the presence of A. sagrei has driven the evolutionary divergence among sympatric and allopatric populations of A. carolinensis. It’s this mountain of work that makes Stuart et al. (2014) a tremendously satisfying paper. We now have a much firmer basis from which to suggest that interspecific interactions have driven patterns of ecomorphological diversification across Caribbean anoles.

But I personally think that this study’s most exciting implications arise from it defining more clearly a part of the anole biology puzzle that still remains relatively empty, namely our understanding of within-population, among-individual variation in microhabitat use and morphology, and the consequences of this variation for behavioural interactions. This summer I came across an A. carolinensis and A. sagrei perched together thus:

A. carolinensis perched below A. sagrei on the University of Florida campus in Gainesville.

These particular lizards couldn’t care less for Stuart et al.’s (2014) findings–clearly, the effect demonstrated in this study is a population-level effect. But this leaves us with a gap between behavioural interactions and eco-evolutionary dynamics–how exactly do we transition from individual A. carolinensis that are content to perch below A. sagrei to a population-level shift in A. carolinensis perch height in the presence of A. sagrei? Reassuringly, the divergence that Stuart et al. (2014) document is so rapid that this question becomes tractable–their results  emphasize an opportunity to integrate behavioural timescale with eco-evolutionary timescales. We can now examine individual interspecific behavioural interactions  among anoles, safe in the knowledge that ecological and evolutionary responses are not far behind.

Editor’s Note (October 28, 2014): Yoel Stuart provides the first perspon perspective on the study on eco-evolutionary dynamics

Editor’s Note II (November 3, 2014): The most thorough press coverage of this paper was in the Orlando Sentinel which as an added bonus had two animated talking anoles explaining the results.

Editor’s Note III (November 4, 2014): Yoel Stuart provides a more in-depth description of the study on the Howard Hughes Medical Institute’s The Conversation

In that classic of American cinema, Rocky III, Rocky Balboa (Sylvester Stallone) employed a particularly cunning strategy during the climactic fight with the younger, stronger Clubber Lang (Mr. T): he used his face to repeatedly absorb all of Clubber’s most powerful blows until Clubber grew very tired. Rocky’s strategy worked, and Clubber, fatigued from what seemed like hours of savagely beating Rocky in the head, ultimately succumbed to one of the relatively few punches Rocky managed to land.

Repeat until World Champion

Now one might suspect that Rocky III’s inspiring message of never giving up being punched in the face would have few adherents in the animal world, and this is indeed what we find. In most cases of male-male combat, combatants are reluctant to enter into escalated physical altercations because the risk of injury to themselves is too high. Instead, males of many animal species have evolved ritualized aggressive signals or displays aimed at intimidating their opponents into withdrawing, and will turn to violence only as a last resort when all else has failed. But some species have adopted the spirit of Rocky’s strategy, if not the letter, and rely on persistence to outlast as opposed to outfight their opponents.

A new study by Wilczynski et al. shows that Anolis carolinensis (the undisputed greatest study organism in the world) may use persistence as part of its fighting strategy as well. Adult male green anoles establish dominance hierarchies initially through aggressive interactions, and the outcomes of these interactions are affected by a variety of behavioural, physiological and morphological factors, many of which are likely reflected in the pattern and intensity of their ritualized aggressive displays. Wilczynski et al. set up staged aggressive interactions between pairs of adult males in the laboratory and tested whether males that responded faster or for longer to behavioural challenges were more likely to win fights. They also noted the colour state, as well as the presence of post-orbital eyespots, of winners and losers, both of which have been the subject of previous discussion on Anole Annals. The authors found that for the measured types of display, future dominant individuals generally displayed more frequently, and continued to display for longer than future subordinate individuals, whereas the effects of latency to display on competitive outcomes is less clear. With regard to colour, despite some intriguing trends, there were no significant differences between dominants and subordinates in any aspect of post-orbital eyespot expression. However, future dominant individuals did remain bright green for longer throughout the interactions than did future subordinates, supporting earlier suggestions that dark brown colouration is linked to subordinate social status and/or stress.

While persistence is a key component of contest behaviour in many animal species, the apparent importance of persistence in display duration in particular is especially interesting within the context of lizard displays. For example, duration of sagittal compression has previously been suggested as a handicap display in Uta stansburiana lizards, and previous studies have also suggested that persistence, perhaps related to accumulation of metabolic costs (paper here), might also dictate male contest outcomes in green anoles. Despite the wealth of knowledge regarding male green anole displays, studies such as Wilcynski et al.’s show that we still have much to learn regarding the behavioural aspects of male combat in this species, not to mention the likely relationships between behaviour and physiology.

Rocky III was unjustly spurned by the Academy of Motion Picture Arts and Sciences in 1983, not even receiving a nomination in the category of best picture (Ghandi won that year for some reason). Even more outrageous, it didn’t win the Best Original Song category it was nominated in! (Would anyone seriously argue that “Up Where We Belong” is a better song than “Eye of the Tiger”? Because it isn’t, and you are wrong). In retrospect, the reason for this travesty is clear: persistence is an important part of animal fighting strategies, and Rocky III was actually a nature documentary.

“I caught an anole lizard and tossed it ten feet or so out into the water. To my dismay, it popped to the surface, swam expertly back to the shelter of the trees, and climbed up a mangrove trunk. Well, I continued, suppose a full hurricane blew an anole so far away on open water it couldn’t get back. Our little experiment shows that it could swim to the nearest islet if it were not too far away.”

So wrote E.O. Wilson (1995 p. 271, Warner Books, NY) in his autobiography, Naturalist, reflecting on his island defaunation work with Daniel Simberloff. From his ‘little experiment’ (I can hear animal care committees cringing), Wilson postulated that anoles could, if they had to, disperse from island to island across open water. Whether anoles can cross water, however, isn’t that important. Rather, what’s important is that they rarely do. Anoles’ status as a symbol of island biogeography and adaptive radiation is largely due to the fact that isolation and the resulting low gene flow among islands set the stage for in situ speciation and adaptive radiation. In fact, much of what we (and island biogeography in general) owe to anoles, we owe because they don’t swim so well. And they don’t colonize new islands very often.

Or rather, they didn’t.

A new paper in Nature by Matt Helmus, with AA stalwarts Luke Mahler and Jonathan Losos, shows how human-mediated dispersal of anoles among Caribbean islands is reorganizing anole biogeography in a very predictable way. I suspect many who have worked on anole island biogeography, me included, have considered what to do about recent introductions and have often, like me, dropped them out of a dateset with the goal of trying to discern the ‘natural’ pattern. Helmus et al., however, saw the spate of recent anole introductions across the Caribbean as an opportunity, rather than a nuisance. Their great leap came from realizing that this reorganization of anole Caribbean biogeography should be predictable from the basic tenets of island biogeography theory.

Based on MacArthur and Wilson’s equilibrium theory, adaptive radiation theory and drawing on Losos and colleagues’ past work from 1993 and 2000, Helmus et al. predicted three patterns: 1) Species-impoverished islands (for their size) should have more exotics than more saturated islands, 2) The phylogenetic diversity of islands should increase due to exotic establishment, and 3) Human-mediated introductions should degrade richness–(geographic) isolation relationships. In short, they found evidence consistent with all of these patterns. Furthermore, they showed that economics that has replaced distance as the key determinant of island isolation. Needless to say, these are very exciting results that have supplied a key test, at biogeographic scales, of some classic theory*. It’s a must read.

This paper is also important because it shows how ‘blue skies’, curiosity-driven science can help us understand and, most importantly, predict how human activity will impact ecological systems. Did MacArthur and Wilson know, more than half a century ago, that their work would predict how increasing globalization and trade embargoes would affect modern biodiversity? I doubt it (cue someone pointing out in the Comments the exact line in the ETIB where they do predict this). However, regardless of whether they knew it at the time, this is exactly what their theory has done. As Helmus et al. state (p. 545): “Our results support the theory that it is the influence of geographic area and isolation on … speciation and colonization that fundamentally determine island biodiversity”. However, as they crucially find, what we now need to do is rethink how we define ‘isolation’. We can’t leave ourselves out of the equation any more. It’s economics, not geography, that matters now. Thus, not only does Helmus et al.’s paper test a long-standing theory, but it provides a clear example of the importance of fundamental scientific theory for understanding and predicting ecological dynamics in the ‘Anthropocene’.

In conclusion, the observation that humans are moving anoles — and other taxa — around faster than they could make their own way will come as a surprise to no one. But finding that the subsequent reorganization of life can be predicted by island biogeographic theory is fantastic (it should be pretty clear by this point that I like this paper. A lot). So if you haven’t read the paper, you should. I know it’s a terrible cliché to call a study ‘elegant’. So I won’t. I’ll call it damn elegant.

*I can’t help but mention that Helmus et al.’s findings were mostly based on good old-fashion OLS regression and ANOVA, and visualized using simple scatterplots – No fancy-shmancy statistical machismo here (phylogenetics aside). Just a clear set of predictions that could be parsimoniously tested. Chapeau.

Editor’s note #1: nice summaries of this paper have been written by Ed Yong’s Phenomena: Not Exactly Rocket Science blog and by Emily Singer in the new online Science magazine Quanta.

Editor’s note #2: The paper grabbed the cover of Nature.

This, in turn, joins a long list of recent science journal covers sporting an anole: