Overlap of habitat selection does not predict spatial
distributions of contact
By comparing the contact-RSF model with a paired (overlap)
individual-RSF model, we were able to reject our null hypothesis and
found that landscape features impacted habitat selection and contact
differently. These results indicate that interactions among unmeasured
factors such as social behaviors, resource value and competition
structure contact and conspecific interactions in more complex ways than
purely selection of habitat.
The mismatch between habitat selection behavior and the spatial
distribution of contacts has ramifications for our interpretation of
contact based ecological processes. In disease systems, accurately
predicting the spatial distribution of host contacts is crucial for
assessing the risk of spillovers and estimating disease spread. Previous
research often used the distribution of host population habitat
selection as a proxy for contact to estimate transmission hotspots. For
example, in the chronic wasting disease system, overlapping areas that
are shared by deer are considered high probability areas for contact and
disease transmission, and were used to inform disease control and
surveillance priorities (Walter et al. 2011). In predator-prey systems,
the prediction of predation risk for prey species, such as the
probability of encounter, attack, and kill, is often estimated based on
the perception of risk by prey in their resource selection or by
directly assessing landscape characteristics where predation occurs
(Atwood et al. 2009). However, because landscape features may influence
habitat selection and contact differently, estimation of spatial
features influencing contact would be more accurate if done directly.