Using occupancy analysis to select focal species for adaptive management

Background/Question/Methods

Most imperiled species are rare or elusive and difficult to detect which makes it challenging to gather data to estimate their response to habitat restoration. Management based on focal species that are indicators of habitat requirements of rare species may be a useful alternative for monitoring the success of habitat restoration until populations rebound. Focal species have commonly been used to indicate ecosystem health or to evaluate the effects of changes in landscapes or management.  However, selection of focal species is often criticized when qualitative methods are used, or when single species are selected to represent diversity at landscape scales.  We present a quantitative multispecies method for selection of focal species on 11 sites in Alabama.  We developed occupancy models based on literature review and expert opinion to determine relationships between use by each species to land cover, vegetative structure, and landscape characteristics. We estimated relative sensitivity of each species to covariates from the log-odds of the effect of standardized site covariates on probability of occupancy.  We selected focal species based on their relative sensitivity to changes in site covariates that would be affected by restoration of habitat characteristics preferred by rare species.

Results/Conclusions

We conducted 1944 surveys at 714 sites on 11 study areas from the southern Appalachians to the lower coastal plain in Alabama. We developed 1782 single season a priori models of occupancy for 191 species detected on at least 5 survey occasions. We identified focal species for 4 general habitat types and 5 structural characteristics preferred by rare species in Alabama.  For example, for open pine habitat we selected 2 focal species that demonstrated positive sensitivity to changes in pine and ground cover, and a negative sensitivity to canopy cover. For all-age bottomland hardwood habitat, we selected 3 birds and 3 herptile focal species that had positive sensitivity to changes in bottomland hardwood and canopy cover, and/or duff.  Although, there are numerous methods used to select focal species, the method presented provides an innovative and quantitative science-based approach useful for monitoring habitat restoration and informing management decisions when target species are absent or difficult to survey.