Animal behavior is diverse and variable, has different relevance at different time scales, and it is influenced by many interacting factors. How can we measure behavior in a meaningful way when it is so complicated? What kind of information is generated from individuals or groups of organisms, and what can we learn about animal systems from emergent behavioral patterns? How do organisms perceive and interpret information in their environment? By the time I began graduate school one thing had become clear to me: animals were complex systems. My research seeks to understand macaque monkey behavior at the perceptual, individual, and group-level using a framework and tools from complexity science.
As an undergraduate at Southwestern University I became interested in animal behavior in a class that explored the sensory systems of animals, which encouraged me to think about life beyond the human experience. Over evolutionary time, animals have developed a collection of sensory receptors which vary quite drastically across species and allow an animal to extract information from physical phenomena in the environment: lateral line systems, magneto-reception, echolocation, detection of UV light. The vastly different sensory-perceptive ‘self-worlds’ of animals create drastically different experiences and behaviors, and I found this fascinating. Post-graduation I began working on a research team at a zoo, primarily working on a nation-wide project that aimed to assess the welfare of zoo elephants. In the wild, animals are exposed to many complex patterns - branching plant life, rough textures of terrain, a diversity of sounds and smells, unpredictable weather and interactions with other organisms - but captivity can sometimes put animals in less complex environments with more predicable daily routines. While many advocate for using more naturalistic enclosures to improve welfare, I realized little work has been done to determine precisely which qualities about a ‘naturalistic’ environment are discernibly different to specific animals and important for their welfare. Given the animal’s sensory self-world, what is an animal capable of perceiving, what patterns are animals predisposed to notice, and what environmental patterns might elicit desirable physiological or behavioral responses?
Monkey Psychophysics: I conduct experiments with captive rhesus macaques using eye-tracking technology to investigate how they interact with images of varying complexity. As a proof-of-concept study, I presented monkeys with simulated black and white contours or ‘coastlines’ of varying fractal dimension, to see if a) monkeys are interested enough in fractals to look at them, and b) if so, if they distinguish between different fractal dimensions, measured by attentional biases. It turns out, both of these things are true. Future projects will investigate the actual scanning paths of macaque vision across these images, whether or not macaques will choose to spend more time in environments with more complex visual stimuli, and attentional biases to images that vary in other complexity metrics. By understanding which features of patterns animals attend to and what information they may process from their visual environments, we can design more stimulating environments.
Behavior Pattern Complexity: Another challenge with captivity is encouraging species-typical behavior - it is not uncommon to see large captive mammals exhibiting behavioral stereotypy, or repetitive behaviors such as pacing (essentially loss of behavioral complexity to a periodic state) that are generally indicative of boredom, stress, or anticipation. It is difficult to measure global phenomena such as ‘health’ or ‘stress’ because these states can rarely be reduced to one measurable factor, but considering emergent structure in sequences of behavior may offer important information about the state of an animal system. Not only does the measurement of behavioral organization offer a framework to investigate pathology, it creates an avenue to explore many other concepts in animal behavior such as an animal’s ability to adjust its behavior in response to a changing environment. I began investigating the temporal structure of animal behavior patterns by collecting time series data of wild Japanese macaque activity on Koshima Island, Japan. How does one collect time series data of monkey activity? By camping on a remote island of monkeys, tracking individuals around the forest all day, and recording their patterns of behavior!
With this data I’ve begun using complexity time series analyses to quantify their behavior patterns. By using fractal analyses and computational mechanics on binary sequences that represent monkey locomotion (moving or not), we’ve begun exploring what pattern properties of monkey movement vary across monkeys, their behavioral state, and the terrain they are traveling across. I will be returning to Japan in the winter of 2017 to further examine the extent of inter-individual differences and intra-individual variability in behavioral organization using GPS and accelerometer biologging collars to collect longer sequences at finer time-scale that include both temporal and spatial data. By identifying what information exists in biologged movement data, and which pattern properties vary with variables of interest, it is my hope to extract untapped information about humans or animals from the vast amount of movement data collected through biologging technology and self-tracking devices.
Mating network spillover: Finally, I am also interested in a systems approach to understanding animal groups, as social species can have socially rich and complex societies. Like humans, monkey groups can have social overthrows or societal collapse, where other individuals try to overthrow the dominate family (like monkey Game of Thrones). Because this is something we wish to avoid from a captive management perspective, my home lab conducts social network studies with large groups of captive rhesus macaques to understand the relationship between different network structures and group stability. Primate societies can be described by multiplex networks constructed from many different types of relationships including aggression, grooming, and status signaling. I am investigating how group dynamics change at the onset of mating season, which is known for an increase in competition, injury, and of course a slew of mating behaviors that only occur during the 3-4 months out of the year. How does the commencement of this additional layer of macaque society influence other affiliative and agonistic interactions? There is presumably increased uncertainty and shifts in power in dyads due to mating relationships, and it creates a fascinating natural experiment of seasonal network spillover. The study of macaque societal stability offers an interesting model system for understanding hierarchy formation, multiplex social networks, and ultimately human societal stability.
I believe complexity science is changing the way we view and study behavioral sciences as it offers quantitative means of explaining qualitative changes in behavior that have traditionally been hard to measure. It is my hope that my research will help us to understand how animals interact with patterns in their environments, to describe information in patterns of animal activity, and to appreciate the interdependencies of group dynamics.