Community-level occupancy interpretation

[Emily Blackwell]

Hi everyone! Thank you Dr. Doser for setting up this community, and for creating a great, user-friendly tool for multi-species occupancy models. I have completed an occupancy analysis in spOccupancy for my Master's thesis and am presenting my defense soon, but this more of a general question about multi-species occupancy modeling. In my study, I interpret results for community-level occupancy and I also used the output of the model to perform a species richness analysis. It has been brought up to me that interpreting both community occupancy and species richness might be a bit confusing, so I wanted to check my understanding of community-level occupancy. Community occupancy is the presence/absence of the entire community of species included in the model, correct? I found that one of my covariates had a significant negative relationship with community-level occupancy. Is it correct to interpret this as "this entire community is less likely to be present at sites with high values of this covariate?" I would love to hear the community's thoughts on this. Thank you!

[Jeffrey Doser]

Hi Emily,

Thanks for the kind words about the package! Your interpretation is close, but it is not completely correct. The community-level occupancy mean parameters in an occupancy model are best interpreted as the average effect of the covariate across the community. So, if there is a significant negative effect of a covariate at the community-level, I would interpret this as something along the lines of "on average across all species in the community, it is less-likely for a species to be present at high values of the covariate and more likely for it to be present at low values of the covariate." So, it is not so much the presence/absence of the entire community of species, but rather looking at the average patterns across the community. However, even with a significant community-level average effect, there can still be species that don't show that same pattern, in which case you could look at the species-specific effects. The community-level variance parameters give an indication as to how much variation there is the relationship among the species in the community from the overall mean. The higher the variances are, the more variation there is across species in the community. The smaller the variance, the less variation there is across species (and thus they would all be very close to the community-level mean). Figure 1 in our manuscript on spAbundance tries to show this idea. That figure is for a multi-species distance sampling model, but the concept is analogous to the case with multi-species occupancy models.

Kind regards,

Jeff