Site-occupancy modeling is widely used in ecology but its application is still limited in paleoecology, where incomplete detection is routine. Here, we make extensive expansions to an earlier multispecies occupancy model used to estimate the dynamics of relative species abundance in fossil communities. These expansions include incorporating counts of individuals at sites, explicitly allowing for the inclusion of specimens assignable to genus- but not species-level, a situation common in paleontology, and modelling regional presence/absence. We provide simulations to check the performance of this new model, as well as simulations to quantify the benefits of using individual count data versus subsample occupancy data and model estimates versus face-value (raw) estimates, respectively. We also provide an empirical case study using occupancy data from a community of marine benthic colonial animals preserved in the Pleistocene of New Zealand. We find that the new model performs well, especially when it comes to recovering relative abundance dynamics and that it is well worth the effort to both collect individual count data and to include individuals unidentified to species-level in the site-occupancy modelling framework. This extended model can be widely applied in paleoecological settings and is necessary when both the average and uncertainty values of relative abundance dynamics need to be robustly estimated.

The number of individuals of species within communities varies, but estimating abundance, given incomplete and biased sampling, is challenging. Here, we describe a new occupancy model in a hierarchical Bayesian framework with random effects, where multi-species occupancy and detection are modeled as a means to estimate relative species abundance and relative population densities. The modelling framework is suited for occupancy data for temporal samples of fossil communities with repeated sampling including multiple species with similar preservation potential. We demonstrate our modelling framework using a fossil community of benthic organisms to estimate changing relative species abundance dynamics and relative population densities of focal species in 9 (geological) time-intervals over 2.3 million years. We also explored potential explanatory factors (paleoenvironmental proxies) and temporal autocorrelation that could provide extra information on unsampled time-intervals. The modelling framework is applicable across a wide range of questions on species-level dynamics in (palaeo)ecological community settings.

The occupancy and relative abundance of species are temporally varying. Estimating these, given incomplete and biased sampling is challenging, not least for fossilized organisms, where preservation is an additional issue. Here, we describe a relative abundance-focused multi-species occupancy model (TRAMPOline) in a hierarchical Bayesian framework. We designed our model on the basis of the need to understand the dynamics of several focal species over 2.3 million years, by drawing on additional information provided by non-focal species observed in the same fossilized community. We expanded our model by adding random effects of species and time intervals (geological formations) and explored potential explanatory factors (paleoenvironmental proxies) and temporal autocorrelation that could provide extra information on unsampled geological time intervals. Our new model, set in an occupancy modeling framework widely used in ecology but little applied in paleoecology, is applicable across a wide range of questions on species-level dynamics in contemporary and palaeoecological community settings.