Is family grouping sufficient for frequency-dependent disease transmission?

Background/Question/Methods

Behavioral decisions made by individuals can generate ecological processes operating at the scales of populations and communities.  In some cases, relatively simple behavioral decisions can be represented in equations that describe processes at the level of whole populations, but such cases are limited and cannot capture the complete range of possible behaviors.  Individual-based models (IBMs) provide an intuitive and flexible framework for scaling up the mechanistic implications of individual decisions on processes at higher levels of organization.  Social grouping has been proposed as a cause of frequency-dependent transmission, because interactions among group members occur relatively independent of overall population density.  Except in herding animals, however, the population-level pattern of transmission depends on interactions between groups.  We are using IBMs to explore mechanisms whereby social interactions among host individuals could cause transmission of disease to be density-dependent, frequency-dependent, or somewhere in between.  We compare findings of a highly detailed, spatially explicit model of disease spread (direct and indirect) within and among matrilineal groups of white-tailed deer (DLD:  "Deer-Landscape-Disease") with those of a more flexible but less detailed IBM that incorporated social need beyond the family group (i.e., any animal left alone always joins a new group).

Results/Conclusions

Whether transmission was direct or indirect and whether the landscape was fragmented or contiguous, transmission in DLD was strongly density dependent despite contacts being highly compartmentalized (within group contact rates much higher than between different groups).  Thus, matrilineal family grouping alone is not sufficient to generate frequency dependent disease transmission.  However, we found that extending social need beyond the matrilineal group can shift transmission dynamics toward the frequency-dependent end of the spectrum, particularly for long-lived hosts with strong social structure and when transmission within groups is frequency dependent.  These results lead us to urge caution in assuming that sociality implies frequency dependent transmission, help inform empirical research, and lend clarity to mechanisms that can feasibly produce hypothesized patterns of transmission.