The evolution and plasticity of social networks traits

Social interactions, such as mating, fighting, and cooperating or competing for resources, are very important for many aspects of animals' lives and for how they cope with and respond to environmental change. Therefore, understanding how they vary within generations (plasticity) and might change across generations (evolution) is of paramount importance. However, predicting these changes is difficult as social interactions depend on the phenotypes and genotypes of each of the interacting individuals - meaning any change is much more complex to predict than for traits solely expressed by individuals. In this project I will develop an experimental system to combine social network analysis with core concepts in evolutionary biology to understand how social traits respond to changing climates. This will also allow me to test ideas of multilevel selection - which have remained highly controversial for years but yet have rarely been directly tested with experimental data. Developing this system will thereafter allow me to test numerous key hypotheses in the ecology and evolution of social interactions in an experimental setting.

To develop the system and answer important questions around environmental change and trait variation, I will use groups of the gregarious Argentinian wood roach (Blaptica dubia) combined with automated methods of data collection and analysis to conduct two experiments. In the first I will measure how social interactions change with increasing aridity within a generation. I will quantify social phenotypes at the individual, sub-group, and group level using the range of measures available from social network analysis. I will then test how these social network phenotypes change as individuals are exposed to increasingly dry environments (reducing from 50% humidity to 20% humidity gradually using a climate-controlled incubator), using dynamic social network models to quantify plastic change.

In the second experiment I will estimate how the fitness consequences of social network phenotypes at the individual, sub-group, and group level change depending on whether the groups are kept at 20, 35, or 50% humidity. This will allow me to estimate selection on social behaviour and therefore predict its evolution. I will therefore be able to understand how social interactions and the social structures they create might respond to environmental change both within-generations and across-generations.

I will use passive integrated transponder tags to continuously record individuals associating at shelters, allowing me to quantify regular social associations and so infer social networks. Aridity is the ideal environmental variable to consider as it is predicted to change with climate change and is known to affect the costs and benefits of grouping, especially in arthropods. Gregarious cockroaches such as B. dubia are excellent study organisms for questions such as these, as they readily form groups, engage in collective behaviour as well as competitive interactions, produce offspring which can be counted easily, are particularly sensitive to changes in humidity, and in some species their formation of aggregations has important consequences for human health.

This project will provide key information on how social behaviours, in the form of social network traits, will change both within- and across-generations, helping us better understand the consequences of climate change for animal populations. I will also identify how the strength of multilevel selection varies with the environment; providing experimental data to address the long-running debate about when levels of selection above that of the individual contribute to evolutionary change. This project will boost the integration of social network traits with ideas at the forefront of evolutionary biology and provide a springboard for larger collaborative projects and grant applications.