Estimating abundance from repeated presence-absence data or point counts

Royle and Nichols showed that variation in detection probability across sites carries information about local abundance, because sites with more individuals are more likely to yield a detection. Their model links the per-individual detection probability to site-level abundance, allowing abundance to be estimated from simple repeated detection-nondetection data without individual identification. For camera-trap studies of unmarked or hard-to-identify cats, this offered a way to infer relative abundance from occupancy-style data. The Royle-Nichols model became a widely used bridge between occupancy and abundance estimation and inspired later N-mixture and unmarked-density approaches. It remains an important option when individual recognition is impossible but more than mere presence is desired from detection histories.