Biology subjects

Silk, M. J.

Publications and source records attributed to Silk, M. J..

8 recordsLinked to original sources

bioRxiv2025

The Dominance of Geometric Graph Models in Animal Social Networks

O_LIDetecting patterns in animal social behaviour and movement is complicated by the diversity of ecological, evolutionary, environmental, and biological drivers of these behaviours such as migration, foraging, assortative mixing, socio-ecological factors, and human influence. C_LIO_LIAddressing these complexities requires a multidisciplinary approach. Recent advances in network analysis and machine learning offer powerful tools for examining and interpreting complex network structures, aiding in the identification and quantification of movement patterns and the prediction of behavioural changes. C_LIO_LIHere we use a comparative approach leveraging network analysis and machine learning techniques to assess commonalities in standard theoretical social structures governing networks across the animal kingdom. We investigate how these theoretical structures explain social organization at different scales, from entire populations to smaller groups. By leveraging interpretable machine learning techniques, we examine the predictive power of species and network construction techniques in predicting structural features of animal social networks. C_LIO_LIWe found that geometric graphs are the frequently predicted network model across the animal kingdom. These graphs represent both spatial and social processes and are formed by positioning individuals uniformly on a 2D plane, with links established based on proximity within a specified distance. In particular, geometric graphs demonstrate structural similarities with interaction types and data collection methods. For example, we found that this graph model had strong structural similarities with networks derived from physical contact and spatial proximity data. Networks with small-world properties, in contrast, were rare across all interaction types and collection methods. Additionally, the occurrence of these networks is influenced by the identity of species and sampling duration. Although incorporating species identity into the classification model did not improve the accuracy of the prediction, it enabled us to account for the varying dependencies of biological characteristics on specific behaviours. C_LIO_LIThis study highlights the value of predictive modelling for uncovering ecological drivers of animal network structures. By focusing on standard theoretical models that are well established in network science, we connect animal social networks to broader theoretical findings, while recognizing that more tailored models combining multiple generative models or network properties may offer deeper insights into network structures. We also emphasize the importance of considering how the specific methods used to build networks for each taxon could influence the biological inferences that can be drawn from those networks. C_LI

ecology↗

bioRxiv2025

Environment dependent benefits of sociality in Soay sheep

An individuals social connections have strong effects on fitness. Despite this, there is pronounced among-individual variation in social behaviour. This variation may be maintained if different types of social connections have environment-dependent fitness benefits, but this has rarely been tested. We applied network analysis to 37 years of social association data in Soay sheep to test this hypothesis. Our results show that both relationship quantity (having many association partners) and relationship quality (having strong associations) are linked with survival. Crucially, the relative importance of relationship quantity and quality depends on overwinter weather, with harsh conditions favouring quantity over quality, likely due to thermoregulatory benefits. Relationship quality, meanwhile, was most beneficial under benign weather conditions, possibly due to enhanced familiarity with conspecifics. These findings advance our understanding of social benefits in fission-fusion societies and show that changing environmental conditions may be an important mechanism maintaining variation in social behaviour in natural populations.

animal behavior and cognition↗

bioRxiv2025

The Role of Pace of Life in Animal Social Networks

The fast-slow continuum is a key axis of variation in life history strategies that captures the evolutionary trade-off between allocating in lifespan versus reproduction. It has been suggested that social behaviour, and therefore network structure, may vary across the fast-slow continuum, but formal theory remains scarce. We develop a novel mathematical model to examine how the rate of demographic replacement may influence emergent social network structures in natural populations. We additionally consider variation in social preferences and modes of social inheritance. Our key finding is that more rapid demographic replacement can substantially constrain the structure of dynamic social networks. For species with longer generation times, network structures are primarily determined by social preferences, while for shorter generation times they are primarily determined by mechanisms of social inheritance. By considering how demographic replacement can constrain social network organisation, our work provides important insights into social evolutionary ecology.

ecology↗

bioRxiv2024

Social ageing can protect against infectious disease in a group-living primate

The benefits of social living are well established, but sociality also comes with costs, including infectious disease risk. This cost-benefit ratio of sociality is expected to change across individuals lifespans, which may drive changes in social behaviour with age. To explore this idea, we combine data from a group-living primate for which social ageing has been described with epidemiological models to show that having lower social connectedness when older can protect against the costs of a hypothetical, directly transmitted endemic pathogen. Assuming no age differences in epidemiological characteristics (susceptibility to, severity, and duration of infection), older individuals suffered lower infection costs, which was explained largely because they were less connected in their social networks than younger individuals. This benefit of social ageing depended on epidemiological characteristics and was greatest when infection severity increased with age. When infection duration increased with age, social ageing was beneficial only when pathogen transmissibility was low. Older individuals benefited most from having a lower frequency of interactions (strength) and network embeddedness (closeness) and benefited less from having fewer social partners (degree). Our study provides a first examination of the epidemiology of social ageing, demonstrating the potential for pathogens to influence evolutionary dynamics of social ageing in natural populations.

animal behavior and cognition↗

bioRxiv2021

Group size and modularity interact to shape the spread of infection and information through animal societies

Social interactions between animals can provide many benefits, including the ability to gain useful environmental information through social learning. However, these social contacts can also facilitate the transmission of infectious diseases through a population. Animals engaging in social interactions must therefore face a trade-off between the potential informational benefits and the risk of acquiring disease. In order to understand how this trade-off can influence animal sociality, it is necessary to quantify the effects of different social structures on individuals likelihood of acquiring information versus infection Theoretical models have suggested that modular social networks, associated with the formation of groups or sub-groups, can slow spread of infection by trapping it within particular groups. However these social structures will not necessarily impact the spread of information in the same way if its transmission is considered as a "complex contagion", e.g. through individuals copying the majority (conformist learning). Here we use simulation models to demonstrate that modular networks can promote the spread of information relative to the spread of infection, but only when the network is fragmented and group sizes are small. We show that the difference in transmission between information and disease is maximised for more well-connected social networks when the likelihood of transmission is intermediate. Our results have important implications for understanding the selective pressures operating on the social structure of animal societies, revealing that highly fragmented networks such as those formed in fission-fusion social groups and multilevel societies can be effective in modulating the infection-information trade-off for individuals within them. Significance statementRisk of infection is commonly regarded as one of the costs of animal social behaviours, while the potential for acquiring useful information is seen as a benefit. Balancing this risk of infection with the potential to gain useful information is one of the key trade-offs facing animals that engage in social interactions. In order to better understand this trade-off, it is necessary to quantify how different social structures can promote access to useful information while minimising risk of infection. We used simulations of disease and information spread to examine how group sizes and social network fragmentation influences both these transmission processes. Our models find that more subdivided networks slow the spread of disease far more than infection, but only group sizes are small. Our results demonstrate that showing that fragmented social structures can be more effective in balancing the infection-information trade-off for individuals within them.

animal behavior and cognition↗

bioRxiv2020

Quantifying the influence of space on social group structure

When studying social behaviour, it can be important to determine whether the behaviour being recorded is actually driven by the social preferences of individuals. Many studies of animal social networks therefore attempt to disentangle social preferences from spatial preferences or restrictions. As such, there are a large number of techniques with which to test whether results from network analysis can be explained by random interactions, or interactions driven by similarities in space use. Selecting which of these methods to use will require determining to what extent space might influence social structure. Here we present a simple method (Social Spatial Community Assignment Test) to quantify the similarity between social and spatial group structure. We then apply this method to both simulated and empirical data of social interactions to demonstrate that it can successfully tease apart social and spatial explanations for groups. We first show that it can resolve the relative importance of space and social preferences in three simulated datasets in which interaction patterns are driven purely by space use, purely by social preferences or a mixture of the two. We then apply it to empirical data from a long-term study of free-ranging house mice. We find that while social structure is similar to spatial structure, there is still evidence for individuals possessing social preferences, with the importance of these preferences fluctuating between seasons. Our method provides a robust way of assessing the overlap between spatial and social structure, which will be invaluable to researchers when investigating the underlying drivers of social structure in wild populations.

animal behavior and cognition↗

bioRxiv2020

Calculating effect sizes in animal social network analysis

O_LIBecause of the nature of social interaction or association data, when testing hypotheses using social network data it is common for network studies to rely on permutations to control for confounding variables, and to not also control for them in the fitted statistical model. This can be a problem because it does not adjust for any bias in effect sizes generated by these confounding effects, and thus the effect sizes are not informative in the presence of counfouding variables. C_LIO_LIWe implemented two network simulation examples and analysed an empirical data set to demonstrate how relying solely on permutations to control for confounding variables can result in highly biased effect size estimates of animal social preferences that are uninformative when quantifying differences in behaviour. C_LIO_LIUsing these simulations, we show that this can sometimes even lead to effect sizes that have the wrong sign and are thus the effect size is not biologically interpretable. We demonstrate how this problem can be addressed by controlling for confounding variables in the statistical dyadic or nodal model. C_LIO_LIWe recommend this approach should be adopted as standard practice in the statistical analysis of animal social network data. C_LI

animal behavior and cognition↗

bioRxiv2020

Common datastream permutations of animal social network data are not appropriate for hypothesis testing using regression models

O_LISocial network methods have become a key tool for describing, modelling, and testing hypotheses about the social structures of animals. However, due to the non-independence of network data and the presence of confounds, specialized statistical techniques are often needed to test hypotheses in these networks. Datastream permutations, originally developed to test the null hypothesis of random social structure, have become a popular tool for testing a wide array of null hypotheses. In particular, they have been used to test whether exogenous factors are related to network structure by interfacing these permutations with regression models. C_LIO_LIHere, we show that these datastream permutations typically do not represent the null hypothesis of interest to researchers interfacing animal social network analysis with regression modelling, and use simulations to demonstrate the potential pitfalls of using this methodology. C_LIO_LIOur simulations show that utilizing common datastream permutations to test the coefficients of regression models can lead to extremely high type I (false-positive) error rates (> 30%) in the presence of non-random social structure. The magnitude of this problem is primarily dependent on the degree of non-randomness within the social structure and the intensity of sampling C_LIO_LIWe strongly recommend against utilizing datastream permutations to test regression models in animal social networks. We suggest that a potential solution may be found in regarding the problems of non-independence of network data and unreliability of observations as separate problems with distinct solutions. C_LI

animal behavior and cognition↗