Unifying occupancy-detection and local frequency scaling (Frescalo) models

Oliver L. Pescott 

Frescalo’s “local frequency scaling” and classical occupancy-detection models both seek to recover true species‐occurrence signals from imperfect data. In this paper, we show that the two approaches rest on the same underlying detection mathematics. Occupancy models treat each site’s repeat visits as independent detection trials and separately estimate occupancy probability and per-visit detectability. Frescalo, by contrast, pools data across ecologically defined neighbourhoods and infers a unified Poisson-process discovery rate and a temporal “time-factor” to capture trends. We demonstrate that the Bernoulli‐trial formulation of occupancy-detection can be linked to Frescalo’s Poisson framework, with occupancy and detectability represented by a single rate parameter (which approximates the product ψp when overall sampling intensity is low). This connection clarifies how Frescalo’s neighbourhood-scale and time corrections function as a coarser‐scale analogue of repeat-visit models. By casting Frescalo in occupancy modelling terms, we hope to promote further investigation into the adoption of occupancy model diagnostics, extensions and covariate tests within Frescalo analyses, improving transparency and rigour when working with less-structured biodiversity data.