Eco-evolutionary dynamics of sexual selection: A trait-based perspective

As the readers of the Eco-Evo Evo-Eco blog know, wide-spread appreciation for the reciprocal feedbacks between evolution and ecology has begun to emerge. While many questions remain, evidence from various systems indicate that phenotypic evolution can indeed influence ecological phenomena. The majority of work on this topic – as judged by Andrew’s recent book – focuses on the ecological feedbacks generated by traits evolving because of their role in capturing energy and avoiding predation; think alewife gill arches and walking stick camouflage.

However, over the last few years I’ve been thinking about the eco-evolutionary dynamics of a different suite of traits - those involved with competition for mates? This curiosity began during my dissertation research on sexual signaling in Bahamian mosquitofish. Basically, I wondered how the evolution of sexually selected traits might feedback to influence ecology. So, I began thinking about how to apply an eco-evolutionary framework to this system. After a year or two of failing to come up with anything compelling, I began reading more generally about the evolutionary ecology of sexual selection (Figure 1). Essentially, I wanted to know whether traits evolved by sexual selection were ecologically consequential. And if so, which ones?

Figure 1. I began writing this paper during a short break in the Scottish Highlands (pictured above). I soon discovered that there was no place better suited for thinking about the role of sexual selection in ecology. This beautiful landscape is populated by animals well known for having strong ecological effects derived, at least in part, from sexually selected traits: Red grouse, Red deer, Soay sheep, and Common lizards, to name a few. But despite being the subject of so much relevant work (Behavioral Ecology, Sexual Selection, Demography, and Sexual Conflict), Scotland is far from exceptional in this regard. Indeed, the more I read, the more I envisioned the potential for sexual selection to have substantial ecological importance anywhere. Nevertheless, that April in the heath-covered Monadhliath mountains had a significant impact on me. It not only helped frame the eventual paper, it reshaped how I think about the evolutionary ecology of sexual selection.

My main goal was to understand how the evolution of sexually selected traits might generate ecological feedbacks. This deep dive led to a surprisingly long list of examples. Sadly, very few of these examples met what I would describe as evidence of an eco-evolutionary feedback. Interestingly, this wasn’t because sexually selected traits are understudied, nor is it because they’re ecologically irrelevant. I could speculate as to why this is the case, but I honestly don’t know for sure. Yet, what is clear, is that a wide range of traits evolve by sexual selection, and that sexually selected traits are not only ecologically relevant, they’re often quite potent.

In a recent review in The Quarterly Review of Biology (Giery and Layman 2019), I highlight some examples – especially ones that I thought could represent the range of feedbacks that occur at population, community, and ecosystem levels. I had two primary goals. First, I wanted to get ecologists thinking more about the role of sexual selection in ecology. Second, I wanted to encourage evolutionary ecologists to think more about the ecological consequences of sexual trait evolution. The resulting review, perhaps more of a literature survey, is long and fairly speculative in places. Nevertheless, I think it satisfied these goals. And if you’re thinking about what traits might be relevant to your own system, I encourage you to check it out (link).

The purpose of this blog post is to broadcast some of the insights that I gained while putting the paper together. So, I’ve structured this post in three parts. First, I review what I believe is the most well-developed outlook on the role of sexual selection in eco-evolutionary thinking. Second, I share a few thoughts on the role of sexually selected traits to ecology. And third, I end with some recommendations for future work on the topic.

I. Sexual Selection and The Evolution of Population Fitness

Condition-dependent mating success: Currently, there’s a lot of excitement about the effect of sexual selection on ecology. Much of the current work on eco-evolutionary dynamics of sexual selection deals with the evolution of population fitness via condition-dependent mate choice. In general, these researchers are trying to understand whether sexual selection can ameliorate various genetic loads (usually mutation load), when mating success is biased toward the healthiest individuals. For example, if individuals with deleterious genetic variation have lower mating fitness (because they compete poorly for mating opportunities), then sexual selection might promote local adaptation and lead to an increase in population growth rate. Specifically, does condition-dependent mating purge mutation loads if genic capture underlies mating success? This is an exciting, standing question and I hope to see more studies addressing it using natural systems.

Antagonistic sexual selection and regulation of intralocus sexual conflict: For quite a few years now there is an increasing focus on the effects of sexually selected traits on population fitness. In many cases, the evolution of sexual traits actually adds a genetic load via intralocus sexual conflict. Under this scenario, incomplete evolution of mechanisms allowing for optimal expression of sexual phenotypes can subsequently influence population fitness through what has been termed a gender load, or a sexual dimorphism load. These sex-specific fitness costs arise because the benefits of sexual ornaments are sex-specific while the viability and/or fecundity costs are not. While the theory is relatively mature, empirical research is just beginning to take an overtly ecological perspective. The eco-evolutionary dynamics of intralocus sexual conflict presents many exciting opportunities for those interested in the topics already discussed.

II. The Ecological Consequences of Sexual Traits?

As discussed above, eco-evolutionary models of sexual selection have focused on examining whether sexual selection exacerbates or ameliorates genetic loads. But what I wanted to know was a bit more elementary – do sexually selected traits mediate ecological interactions and processes. It turns out that yes, traits that evolve by increasing mating fitness are ecologically relevant. It’s also worth noting that this is a recurring sentiment among evolutionary ecologists. 

Collias seems to think that sexually selected aggression might have interspecific effects: “The role of aggressive behavior at the level of the individual, social group, and species have been suggested. It remains to discuss the relationship of aggressive behavior to the balance of density ratios of different species, in the relationship commonly known as the ‘web of life.’” Collias 1944

Chitty suggests that density-dependent evolution of behavior (male-male aggression in particular) is worth testing if you want to understand population dynamics: “…given field evidence of unexplained reductions in breeding success and survival, we can be reasonably sure that … powerful intraspecific processes of some kind are at work in the wild.” “It seems, then, that behavior, physiology, and genetics must be of increasing concern to population ecologists, who have probably spent too long already on purely descriptive studies.” Chitty 1967

Trivers knows that male fitness depends on priority access to females, not viability: “In species with little or no male parental investment, selection usually favors male adaptations that lead to increased mortality. Male competition in such species can only be analyzed in detail when the distribution of females in space and time is properly described.” Trivers 1972

And Clutton-Brock et al. seem to know that researchers are routinely overlooking the role of sexual selection in ecology: “During the last two decades, ecology and evolutionary biology have again converged, and the importance of setting explanations of ecological phenomena within the framework provided by natural selection is now generally accepted. What is less often appreciated…is that Darwin’s theory of sexual selection has direct relevance to ecology, too.” Clutton-Brock et al., 1982

Nearly 40 years since Clutton-Brock’s quote, interest in the evolutionary ecology of sexual selection remains strong. Discussion about the coevolution of female preference and male sexual phenotypes remains quite lively, and the role of sexual selection in speciation continues to generate interest. But while calls for greater attention to the ecological consequences of sexual selection have echoed through the literature for decades, these recurring references seem siloed, disconnected, and generally ineffective at stimulating and/or sustaining interest. This, I believe, is evident in the paucity of work showing ecological feedbacks of sexual trait evolution.

Obviously, reviewing the evolutionary ecology of sexual selection is a colossal task – reinforcing my respect for Malte Andersson’s wonderfully inclusive review – which still stands the test of time (crack that book to any page and you’ll find interesting biology). Anyways, out of some practical concerns, I narrowed the focus of the survey to studies of vertebrates under natural or semi-natural conditions. After reading widely to collect cases in which sexually selected traits have strong and well-documented effects on interspecific interactions, population dynamics, and ecosystem-type stuff, a few generalities emerged.

Generality 1: Not all sexually selected traits are ecologically potent. This is seen when traits are chunked up among a few different categories (e.g., behavior, intrasexually selected weapons, ornaments and displays). One could argue for a different categorization, but I think this works ok.

Figure 2. Sexually selected behaviors such as male-male aggression, infanticide, sexual harassment, and coercive copulation are ecologically relevant. These traits, and their effects on demography, are relatively well studied in rodents (e.g., Myodes) and carnivores such as brown bears (Ursus arctos). However, some terrific experimental work in common lizards (Zootoca vivipara) and red grouse (Lagopus lagopus scotica) show that sexually selected behaviors are taxonomically widespread and can have large effects on population dynamics. Photos by Hanna Knutson, Philip McErlean, Jan Rose, and Mark Hope.

Sexually Selected Behavior: Evidence that infanticide, male-male aggression, sexual harassment, and coercive mating are ecologically important appeared throughout my search and were routinely among the most ecologically impactful sexually selected traits. Along with various morphological and physiological adaptations evolved by sexual selection, behaviors seemed to be the most important. Indeed, the direct and indirect effects of those behaviors have a strong influence on population dynamics for a wide variety of vertebrates. And while sexually selected behaviors can influence interspecific interactions (i.e., reproductive interference), the ecological effects of behavior tended to be most apparent when looking at intraspecific interactions. For example, sexually selected infanticide seems to be important for immature survivorship and female fecundity – an obvious finding in retrospect (Figure 2).

Figure 3. Sexual selection by male-male combat often leads to the evolution of weapons such as tusks, antlers, and horns. It can also drive the divergent evolution of traits that have other functions. This is perhaps best seen in the case of sexual selection on male bite force. In many lizards, males have enlarged heads and greater bite force. This often translates to divergent diets with males consuming taxa that are larger and require more power to subdue and handle. For example, broad-headed skink (Plestiodon laticeps) males consume much larger prey on average and across a broader range. Photo by Edward Prenzler.

Intrasexually Selected Weapons: Enlarged teeth and large body size were also important. Unlike the behaviors discussed above, morphological adaptations evolved by male-male combat seemed to be more relevant for interspecific interactions. For example, larger heads in many male lizards translate to an ability to capture and process larger, harder prey. Corresponding differences in diet appears routinely between the sexes in lizards and a variety of other vertebrates, especially carnivores (Figure 3).

Figure 4. Sexually selected display behaviors in frugivorous birds generates clumped patterns of seeds and seedlings. Dispersal follows two modes. First, a typical endozoochorous pathway where seeds are dispersed post consumption. Second, fruits and seeds are moved and aggregated without being consumed. The former case is exemplified by lekking birds that forage widely yet spend disproportionate time at display sites. Defecating at leks aggregates seeds and generates seedling patches. Examples include the long-wattled umbrellabirds (Cephalopterus penduliger), cock-of-the-rock (Rupicola peruviana), and white-bearded manakin (Manacus manacus). An example of the other mode is seen in bowerbirds, in which males also disperse and aggregate seeds, but in an interaction not mediated by endozoochory. That is, they don't eat the fruits. Instead, males of many bowerbird species gather forest materials to construct bowers, complex structures subject to female choice. As seen in the spotted bowerbird (Ptilonorhynchus maculatus), fruits aggregated at (and disposed near) bowers from the surrounding forest subsequently germinate; altering nearby plant communities (Madden et al. 2012). The photo depicts a great bowerbird (Ptilonorhynchus nuchalis), a congener which does not include fruit or seeds but maintains a similar bower structure. Photo credits from left to right: Nick Athanas, Ricardo Sanchez, Sergey Pisarevskiy, and Julie Burgher.

Ornaments, signals, and display behaviors: Interestingly, while there has been a lot of research on the evolution of signals and ornaments, there appears to be very little linking the evolution of these traits to ecology. For example, you might expect limitation of pigments such as carotenoids to drive compensatory increases in foraging on carotenoid-rich foods. But from what I could find, there’s little indication that this is widespread. Indeed, the only case I could find was a study showing that male hihi - a bird expressing yellow, carotenoid-containing nuptial plumage - seem to consume more carotenoid-rich fruit than females. In general, it seems that the evolution of signals and ornaments has a rather limited influence on ecological systems (happy to be proven wrong here). An exception might be made in cases in which display behaviors such as lek formation influences other species. For example, in several species of lek-forming birds a tendency to center activity around display sites seems to lead to an aggregated pattern of seed dispersal that could potentially influence community structure (Figure 4). Nevertheless, ornaments and signals themselves don’t seem particularly consequential.

Generality 2: Sexual selection can have indirect effects on trait-mediated ecology. This point is a bit more complicated, but the essence of this generality is that sexual selection shapes a range of traits not typically imagined to be under sexual selection. In part this is because they’re not – at least not directly. This point is perhaps best illustrated by the extreme sexual difference in ecology seen between males and females. While much of the divergent life history and ecology of vertebrates can be attributed to selection on fecundity and/or viability, signatures of sexual selection on the evolution of sexual dimorphism is often quite clear. For example, take the divergent diets of red deer. Sexual size dimorphism in red deer is well known and presumably evolves by male-male competition – larger males have higher mating fitness. But males and females also eat different foods – why? One hypothesis is that sexual selection for large body size indirectly influences the digestive physiology of ungulates. This hypothesis, based on the Jarman-Bell principle, was originally developed to explain the tendency for larger ungulates to consume less-nutritious plant material, has been adapted to intraspecific variation as well. Indeed, the larger body size of males confers an ability to digest and gain energy from rougher forage. Therefore, sexual selection has an indirect effect on the trophic ecology of sexually size dimorphic critters (Figure 5).

Figure 5. Indirect effects of sexually selected traits are common. For example, a positive allometric relationships between body size and digestive physiology in ungulates and some other large herbivores such as elephants contribute to sex- specific diets and sexual segregation. Photo by Jan Rose.

Essentially, the indirect ecological effects of sexual selection result from selection on functionally integrated traits. That is, selection on body size influences digestive physiology, predation risk, thermal performance, locomotion performance, etc. In turn, these traits have their own ecological effects. Accounting for the relative influence of sexual selection on functionally integrated phenotypes is a challenge. Nevertheless, such indirect effects of sexual selection appear to be quite strong in some cases.

III. Take Home
This review was really about getting others to try and view the evolutionary ecology of natural populations through the lens of sexual selection. Ultimately, if we want to integrate sexual selection into and eco-evolutionary framework there is a lot of work left to do. First, evidence for ecological feedbacks of sexual trait evolution are generally lacking despite evidence that sexually selected traits can evolve on contemporary timeframes (Svensson 2019). Second, we need to spend more time thinking about the role of sexual selection in intraspecific ecological diversity (De Lisle 2019, Fryxell et al. 2019). And third, development of eco-evolutionary frameworks that integrate sexual selection will require advancement of fitness-based perspectives as well as trait-based ones.

I don't see conceptual or practical impediments to this integration of sexual selection and ecology. How then, might we promote a fuller view of evolutionary ecology? I suppose the quote below offers some insight:

“Even if one strongly believes in the action of natural selection it is exceedingly difficult as Darwin has pointed out, to keep it always firmly in mind. Neglect of natural selection in ecological thinking is, therefore understandable though regret[t]able. However, its deliberate exclusion…would seem to be exceedingly unwise.” Orians 1962

Here, Orians reminds ecologists that natural selection is ubiquitous and that the wise ecologist would be in error to infer the function of the natural world without its due consideration. As we know, the farsighted sentiment Orians articulates is now rather familiar. Indeed, it seems to me that most recognize natural selection as pervasive and relevant for  ecological inference. Now, I don't think that keeping evolutionary processes such as sexual selection in mind is exceedingly difficult. While it's probably not the most common way ecologists think about the world, many ecologists do pay keen attention to sexually selected traits. Nevertheless, I wouldn't hesitate to say that a broader inclusion of sexual selection in ecological thinking would be welcome.

Some recent papers worth a read:
De Lisle, S. P. 2019. Understanding the evolution of ecological sex differences: integrating
      character displacement and the Darwin-Bateman paradigm. EcoEvoRxiv:         

      https://doi.org/10.32942/osf.io/e2ahg. (link)

Fryxell, D. C., D. E. Weiler, M. T. Kinnison, and E. P. Palkovacs. 2019. Eco-evolutionary  

      dynamics of sexual dimorphism. Trends in Ecology & Evolution xx:1–4. (link)

Giery, S. T., and C. A. Layman. 2019. Ecological consequences of sexually selected traits:  

      an eco-evolutionary perspective. The Quarterly Review of Biology 94:29–74. (link)

Svensson, E. I. 2019. Eco-evolutionary dynamics of sexual selection and sexual conflict. 

      Functional Ecology 33:60–72. (link)

Dissecting phases of speciation in stick-insects

I have worked on many different study systems over the years, including killer whales, livebearing fishes (family Poeciliidae) and Timema stick insects (for more detail please see my homepage). Originally, I started my research career with a Diploma in Biology in Germany (equivalent to a Masters in other countries), with my thesis work focusing on whistle communication in diverging killer whale populations around Vancouver Island in British Columbia, Canada. However, I soon realized that there are only so many questions you can ask using a study system that largely precludes running controlled experiments. Thus, for my PhD thesis and subsequent postdoctoral work, I focused on population divergence and speciation in livebearing fishes (Poeciliidae) living along various environmental gradients (e.g., gradients of predation, toxicity, and access to light). It was during this time that I realized that my main interest was not so much in one particular study system, but rather in discovering the mechanisms that create, maintain, and sometimes constrain, biodiversity. That interest eventually led me to add yet another study system to the mix in 2012: the system of Timema stick insects I am writing about here. More specifically, I wanted to use that particular system to study the potential role chemical communication might have on population divergence and speciation.

As the above paragraph suggests, each of my study systems has their own system-specific peculiarities [e.g., cultural differences seem to play a prominent role in driving population divergence in killer whales (Riesch et al. 2012), but not in stick insects or livebearing fishes]. However, they also have a lot in common. For example, selection from predation is integral to both, the livebearing-fish and stick-insect systems I study, while foraging specialization plays a prominent role in population divergence of both, stick insects and killer whales. Thus, the three systems simply constitute different examples of how ecologically-based divergent selection drives population divergence and ultimately (ecological) speciation.

A rain shower moving through the chaparral near Santa Barbara, California. Timema are often found in this biome of dense thickets and thorny bushes.

The idea that speciation can be thought of as a continuum is yet another concept that dates back at least to Charles Darwin’s world-changing On the origin of species. The concept in its modern form posits that pairs of populations move along a continuum between panmixis on one extreme end and complete reproductive isolation on the other. Progress can be towards speciation or towards collapse, the latter showcased by studies on speciation reversal in European whitefish Coregonus spp. (Vonlanthen et al. 2012), three-spined stickleback Gasterosteus aculeatus (Tayler et al. 2006), and cichlid fishes (Seehausen et al. 1997).

This concept of a speciation continuum has gained traction again in recent years (e.g., Hendry et al. 2009). Consequently, studies across closely related taxa at different phases of speciation are beginning to illuminate the processes and genetic changes underlying the formation of new species (Seehausen et al. 2014). It is well-known, of course, that speciation involves genetic differentiation, and that, in the absence of gene flow, genome-wide differentiation can readily build up by selection and drift. If speciation is to happen in the face of gene flow, however, the picture gets more complex. According to the genic model of speciation (Wu 2001), speciation is initiated by a few genetic regions that become resistent to gene flow before others. This results in a localized pattern of genetic differentiation, which becomes more genome-wide as speciation progresses.

In a recent study just published in the April issue of Nature Ecology and Evolution (http://www.nature.com/articles/s41559-017-0082), we took a closer look at the transitions between phases of genomic differentiation during speciation of Timema stick insects. Like other studies on the speciation continuum (including my other study systems), we were faced with a key problem: speciation is often slow enough that we cannot simply follow a single lineage through time to see in real-time how the process unfolds. The solution then is to take as many different snapshots of the process from different pairs of natural populations as possible, and to then start to reconstruct a bigger picture of what might be happening across different moments in time. This is exactly what we did using data from >100 populations of 11 species of Timema stick insects. Our work suggests that speciation can be initiated by few genetic changes associated with natural selection on few loci, but the overall process is multi-faceted and involves mate choice and genome-wide differentiation.

A male Timema cristinae on one of its host plants (genus Ceanothus).
Photo: Moritz Muschick

This study is the culmination of almost 30 years of research into this system, and consists of data collected between 1996 and 2014, including >1000 re-sequenced whole genomes. In fact, research in this system began when two of our coauthors, Cristina Sandoval (University of California in Santa Barbara, USA) and Bernie Crespi (Simon Fraser University, Canada), recognized this group harbours variation in phases of speciation. Patrik Nosil then entered the system in 2000 and eventually wrote his PhD on it, using experiments to estimate reproductive isolation. An important component of the current paper was the chemical ecology of stick insects. This part of the project was born ~2009, emerging out of initial discussions between Patrik and I, with additional input from Bernie Crespi and Gerhard Gries at Simon Fraser University. Fast-forward to the year 2012, where the alignment of different projects getting funded finally enabled us to team up at the University of Sheffield in the UK (key components were a European Research Council Grant to study the genomics of speciation to Patrik, a Human Frontier Science Program Postdoctoral Fellowship to study the role of chemical communication in speciation to myself, and a burgeoning collaboration with Zach Gompert, a statistical population geneticist from Utah State University).

The emphasis of the chemical ecological aspect of the project was on cuticular hydrocarbons (CHCs), the oily/waxy chemicals on the cuticle of insects that can function to prevent desiccation and physical injury (Drijfhout et al. 2013 in Behavioral and Chemical Ecology, pp. 91-114), but that have also been repeatedly implicated as integral to mate choice (e.g., Blows and Allan 1998; Chung et al. 2014). For stick insects, we found that populations that differed more strongly in their CHC profiles also had higher degrees of sexual isolation and stronger genome-wide differentiation. We confirmed the causal role of CHCs in mate choice by means of a perfuming experiment.

Evaporating hexane samples in Santa Barbara, California, as part of the perfuming experiment on Timema mate choice.

Preparing another CHC-hexane sample for analysis with the gas chromatograph in the Gries lab at Simon Fraser University in 2014. Photo: Sean McCann

When combining this CHC-data with other phenotypic data and genomic analyses, we uncovered that, consistent with early phases of genic speciation, colour-pattern loci that confer camouflage to particular host plants reside in localised genetic regions of accentuated differentiation between populations experiencing gene flow. Transitions to genome-wide differentiation are also observed with gene flow, but appear to have little to do directly with differentiation in color. Rather, genome-wide differentiation is associated with divergence in CHCs, which we show to be polygenic, modestly heritable traits. Thus, intermediate phases of speciation are not associated with growth of a few peaks or ‘islands’ in the genome. Finally, we show that complete reproductive isolation was associated with a conspicuous increase in the overall degree of genomic differentiation. Thus, although speciation is perhaps continuous, this does not mean it always proceeds in a strictly uniform fashion (this component was led by our collaborators Moritz Muschick, now a postdoctoral fellow at EAWAG in Switzerland, and Victor Soria-Carrasco, who is now a Leverhulme Early Career Fellow at the University of Sheffield). Overall, the results suggest that substantial progress towards speciation may involve the alignment of multi-faceted aspects of differentiation. We suspect similar conclusions may apply to other systems where strong reproductive isolation involves many traits and evolves in a polygenic fashion.

A female Timema bartmani, cryptic against the needles of white fir. Photo: Moritz Muschick.

In conclusion, although many questions remain unanswered, it seems clear that speciation in this group involves more than divergence in cryptic coloration, and the results point to mating isolation and other reproductive barriers, as well as geographic separation, as being important. Thus, the striking example of crypsis in Timema that has been the focus of many previous studies (e.g., Sandoval 1994; Nosil and Crespi 2006; Comeault et al. 2015) represents only one aspect of the multi-faceted speciation process.