Information content of signals and cues
To understand the complex links between behavior, physiology, and ecology, it is critical that the behavioral code be deciphered using appropriate statistical tools. Behavioral and physiological responses are often highly complex, comprising multiple components, which vary among individuals and among contexts. A primary goal of the Behavioral Complexity Lab is to quantify complex behaviors and physiological patterns and illuminate cryptic biological phenomena that may have impacts on populations. To this end, we employ field experiments, fine-scale time series approaches, and information theory to decipher complexities of animal communication, components of which are linked to individual fitness. Past and current projects include:
- Detection of neighbor-stranger recognition using complex interaction networks. We show that researchers’ ability to detect simple behavioral phenomena is improved when behavioral networks are considered. These results suggest that local response networks are critical, yet underappreciated, components of environmental context and that they may drive interpretation of many experimental playback studies (in preparation for Animal Behaviour).
- Context-dependent information transfer in birds. In collaboration with Santiago Escobar (former field assistant), we used song wrens (Cyphorhinus phaeocephalus) and fine-scale time series analysis to show that information transfer (bit-rate) was maximized during simulated intrusions by unfamiliar individuals. This pattern was undetectable using data grouped into pre-stimulus and stimulus time blocks (in preparation for Animal Behaviour).
- Quantification of signal display complexity. With Carla Vanderbilt and Emily DuVal, we developed an entropy-based quantification of complexity of coordinated male-male courtship displays of lance-tailed manakins (Chiroxiphia lanceolata) (Vanderbilt et al. 2015), providing a basis for understanding the impact of coordinated male-male displays on differential reproductive success. In as yet unpublished work, we used a similar entropy framework to quantify complexity of male multi-modal courtship (vibratory and visual) displays in 46 species of jumping spiders (Habronattus spp.) to test hypotheses about the origin of multi-modal signals.
- Eavesdropping and community structuring in army ant following birds. In collaboration with Henry Pollock, Ari Martinez, and Corey Tarwater, we have explored how social information in army ant-following birds impacts structuring of mixed-species foraging aggregations. We have found that eavesdropping (exploitation of acoustic cues) is not only a learned behavior but is also asymmetric between species in the community; that is, there is a distinct social information cascade whereby species respond only to particular species. This work is currently being expanded to explore these phenomena as a function of community composition at differe sites across the Isthmus of Panama.