This week’s BEACON Researchers at Work blog post is by MSU postdoc Aaron Wagner.
The evolution of sociality is one of the most fascinating and productive topics in evolutionary biology. Though it is often very useful to look to social species to understand the current function, costs, benefits, and circumstances of cooperative relationships, current observable social behaviors may not reflect the selective pressures that led to their initial evolution. Instead, the observed benefits of a social strategy may be a consequence of that strategy, not the force that initially favored its evolution. For example, social grouping in carnivores is often explained as a way to increase hunting success via pack hunting. However, in many carnivores, social group size is larger than hunting group size (e.g. lions Panthera leo or spotted hyenas Crocuta crocuta), so any foraging benefits may be an effect, not a cause, of grouping.
Simple spatial groups can form through a passive process when the costs of tolerating others drop below the costs of sharing space (such as a territory). The difference between these costs is largely determined by availability of resources: when the resources in an animal’s range exceeds its needs, there is little or no cost to sharing the range, tolerance can develop, and spatial groups can form even in the absence of any benefits. Once simple groups have formed, natural selection can then act to promote the evolution of cooperative strategies and the realization of any associated social benefits. This broad construct (excess resources -> tolerance -> aggregation -> stability -> social benefits) sets the stage for developing and testing predictions about the role of resource availability and distributions in the evolution of sociality. Because simple groups must exist prior to the evolution of more complex social organizations, a fundamental part of my research focuses on uncovering the conditions that permit, promote, or preclude the evolution of simple grouping strategies and the persistence of groups.
Within the carnivores, sociality is extremely rare: 85% of carnivore species are considered solitary. For many good reasons, the vast majority of carnivore behavioral ecology research has focused on the social minority. However, several recent studies of species from the solitary majority have unveiled intriguing variations in grouping strategies and behaviors across populations. In particular, temporary groups sometimes form in these ‘solitary’ animals when resource conditions can support them. Within these groups, cooperation and social behaviors are often limited and primitive and, in the absence of active benefits to grouping, the groups themselves are often unstable. Among these ‘incipiently social’ species is the striped hyena (Hyaena hyaena), which I had the pleasure of studying for many years in Kenya.
Before we began working with striped hyenas, they were routinely described as being strictly solitary. However, we discovered that this was not always the case. At our first site, while they almost never interacted, males were willing to share ranges while females were strictly solitary. At the second site where resources were more plentiful, males were strictly solitary but some females shared ranges with other females and often interacted with each other’s cubs at dens. Studies like this, where differences in non- or minimally-cooperative grouping strategies can be compared with differences in resource conditions, provide glimpses of the earliest stages of social evolution and hints about the constraints that were “breached” to permit the evolution of far more complex forms of sociality, like that found in the highly gregarious (and über complex) spotted hyena.
While our work with the striped hyena was enlightening (and an experience I wouldn’t trade for most anything), significant questions remain. For instance: Are permissive conditions sufficient to maintain group stability? Is group stability necessary and sufficient for the evolution of sociality? And what types of modifications in resource conditions explain variations in social strategies, including the conditions that favor immigration (moving to a new group) over philopatry (staying in the same group you were born in) as a means of group formation and maintenance? While it may be possible to address these questions via additional field studies, I have taken a very different approach… following a fairly dizzying left turn into Charles Ofria’s Digital Evolution Lab.
What first brought me to the ‘Devolab’ was an encounter with a description of the digital evolution platform Avida that Charles, his colleagues, and students have developed for and applied toward studying an impressive array of fundamental questions in evolutionary biology. When I first read about Avida, I was instantly convinced that this platform was ideal for addressing questions about the pressures and patterns underlying the evolution of tolerance, group formation, and social cooperation. In a broad sense, Avida seemed perfect for evolving ‘digital hyenas’ or, alternatively, for uncovering the conditions that lead to the evolution of carnivore-like grouping, proto-social, and social behaviors.
While cooperative behaviors have evolved in ‘avidian’ populations in the past, this occurred under conditions in which grouping was a given. That is, organisms were placed into groups, they were not ‘asked’ to form their own groups first. This was also not done in a spatial context whereby groups have physical ranges or territories in the digital world encompassing locally accessible patches of resources. My work seeks to examine exactly that process: starting with a solitary population, what distributions of resources drive the population to evolve ranging behaviors and to either defend that range as a territory or to tolerate one another’s presence?
While it is obvious to us now, having come from working with a large and complex animal that has already evolved to do it, what we did not anticipate was that in order for the organisms to ever evolve such ranging and grouping behaviors, they would also need to evolve the skills to intelligently move and navigate through their environment. Hyenas already do that… it’s something taken for granted in the field and never thought about. Since coming to BEACON, I think about it a lot! Because having intelligent and flexible navigators is so critical for addressing our original goals, much of our efforts have focused on looking at what aspects of the environment drive organisms toward evolving navigation in the first place (compared to not moving at all, or just happily running around like crazed chickens… both of which they are often quite content to do) and toward evolving use and control of sensors.
As it turns out, starting with a simple, non-moving, and blind ancestor and evolving intelligent navigation in an open ended environment like Avida is far from simple. However, we have succeeded in doing just that. We began with various environments in which the organisms were required to navigate out to a food resource and return to the nest on which they were born. The three biggest challenges that we had to overcome here were to uncover 1) the specific characteristics of the environment that would pressure the avidians to travel far from the nest, 2) the characteristics that would create pressures to evolve away from random movement, and 3) the conditions that would prevent the evolution of fixed ‘blind’ strategies. For the last one, the solution was to put the food resources in motion. For the first two, competition resulting from resource depletion and crowding does the trick: organisms suffer if they all try to feed from the same food resource, because the food runs out. Thus, they are driven to find resources farther from the nest, but the evolution of intelligent sensor use becomes more critical the farther out the organisms travel.
Avidians evolved abilities to detect distant food resources (small greys) moving in fixed orbits, navigate to and feed from them, and return home to reproduce at a central nest (large grey). Organism colors reflect the food resource the organisms are seeking or have fed from. Food resources appear speckled as the organisms deplete them. In this experiment, unlike the food resources, the central nest is not visible from long distances. The organisms here have compensated for this by using the (visible) closest orbiting resource as a landmark from which they search for the nest (note the streams of organisms heading to that lowest orbit resource, but that it is not being depleted). The noisy clouds around the food resources are a partial consequence of crowding… there are fewer cells on the resources than organisms trying to feed from them.
Once the avidians demonstrated an evolved ability to control and use their sensors and to navigate across large distances, we added a new twist to their lives: we allowed avidians to eat other avidians. In other words, we allowed for the evolution of predators from the prey population. With the same sensor capacities as before, avidians almost immediately evolved successful predation strategies in stable predator-prey populations.
Prey (greens) ‘blooms’ occur in response to seasonal shifts in resource (black) abundance and location. Greyed resources are out of season and are below the minimum the prey need to feed. Evolved from the prey, not introduced, predators (red) have adapted their sensors for use in locating prey and rapidly swarm in response to prey blooms.
Our next step is to tie these successes together with other non-spatial work in which avidians evolved abilities for territory defense as a response to local resource availability and competition. The primary hurdle here will be to evolve territory establishment and fidelity by the predators. In the process, we will ultimately be able to test the suite of hypotheses we originally targeted. Namely, under what resource conditions do predators establish ranges and alter their levels of tolerance toward conspecifics. Undoubtedly, as always, we will be learning as much about the intricacies of evolution from the development process itself as we do from the results of the final experiments…but that’s half the fun of it all!
For more information about Aaron’s work, you can contact him at apwagner at msu dot edu, or visit his website.