Using spatial patterns to infer disease processes in a multi-host, multi-pathogen system

Background/Question/Methods

Like communities of free-living organisms, the distribution, abundance, and composition of pathogen communities are determined by interactions among ecological and evolutionary processes that occur at multiple spatial scales. Large scale processes, such as climate variation and seasonality, determine the potential overlap of host, vector, and pathogen distributions. In addition, local variability in host susceptibility and vector transmission patterns determine the within- and between-host distribution of pathogens. The complexity of such systems can make it difficult to disentangle the relative importance of processes driving pathogen dynamics in natural communities. To address this gap, we analyzed the spatial patterns of four aphid-vectored plant virus species (Barley and Cereal yellow dwarf viruses) in a spatially nested grassland experiment in California, USA. We used variance components analysis to assess the relative contribution of processes at the host population, host metapopulation, and regional scales to variation in pathogen occurrence and the number of pathogen species per host individual (within-host diversity). Then, we used autocorrelations and cross-correlations to examine the spatial patterns of pathogens at the host population scale. These allowed us to infer patterns of transmission, the nature of between-pathogen interactions, and the role of environmental variability in determining pathogen occurrence and within-host diversity.

Results/Conclusions

We found that most variation in infection prevalence occurred at the host population scale, indicating the importance of local processes, such as spatial variation in susceptibility and vector transmission, in driving pathogen occurrence. In contrast, most of the variation in within-host diversity was explained by the residuals rather than processes at any particular scale. These results confirm previous findings of differential drivers of Barley and Cereal yellow dwarf virus prevalence and within-host diversity. Using autocorrelations to characterize the distribution of each virus species at the host population scale, we found little evidence for aggregated transmission. Cross-correlations between virus species indicated frequent co-occurrences within the same host individual or neighboring hosts. This appears to result in spatial “hotspots” of virus diversity. Overlap of virus species’ distributions at the local scale may be contributing to the high rates of co-infection observed in the field. Manipulation of the grass hosts’ nutrient availability did not have a large effect on the distribution of single virus species. However, it eliminated or decreased the strength and distance of positive associations between virus species. Therefore, resources shown to be important to the prevalence of viruses may also be influencing their interactions.