by efraser » Mon Jun 10, 2019 12:01 pm
We are using a multi-season analysis in Presence to model variation in detectability (p) of red squirrels over the annual cycle and have been reading some of the detection probability literature. Detection probability is often described as having three components: probability of presence (related to abundance); probability of availability; and probability of detection given availability (e.g. Hostetter et al. 2019). Are the "p" estimates modelled by Presence considered to be the product of these three probabilities? If so, could you recommend the key references that describe this relationship? Also, given that we are doing a multi-season analysis, does the fact that we are modelling both colonization and extinction rates between seasons possibly mitigate the effect of changing abundance (probability of presence) on changing estimates of detectability?