Print PageUnderstanding the spatial dynamics of animal populations requires knowledge of both individual movements and the interrelation of movements among individuals. While much behavioral research has tackled mechanisms of collective movement in herds and schools, the interrelationships among animal movements at broad temporal and spatial scales have been largely neglected. When individuals’ movements are non-independent, quantifying spatiotemporal dynamics at the population level will require accounting for the relationships among moving individuals.
The lack of attention paid to non-independent movements is at least partly due to a dearth of tools for quantifying such relationships. To fill this gap, Mueller et al. (2011) introduced a movement coordination index, but their index was highly biased at small sample sizes and nonlinearly related to the underlying correlation among individuals. Here, we introduce a new movement coordination index, C, based on the average intercorrelation among individuals, that ranges from zero for uncoordinated movements to one for perfectly coordinated movements.
Results/Conclusions
Under an assumption of multivariate normality, we present explicit formulae for the expected value, bias, and variance of C under repeated sampling. These results show that C has very small bias even for small numbers of individuals and of movement steps over which those individuals were observed. Finally, we develop a null hypothesis test that can be used to identify a statistically significant amount of non-independence among moving individuals. We highlight the utility of our index using data on large-scale movements of Mongolian gazelles (Procapra gutturosa) and khulan (Equus hemionus hemionus). For gazelles we find no statistically significant coordination among individuals (C=0.106, p=0.183), while in contrast, khulan exhibit more coordination than can be explained by chance alone (C=0.229, p=0.009).
phone 202-833-8773