presenceabsence

Presence-absence refers to data that record whether a species is detected at a sampling unit during a survey, encoded as 1 for presence and 0 for absence. Such data can be organized into presence-absence matrices, with rows representing sampling sites and columns representing species or time periods.

Data collection: Field-based surveys, traps, camera traps, acoustic monitoring, or environmental DNA samples can provide presence-absence signals. Because detectability is imperfect, a species may be present but not detected in a given survey; therefore multiple visits or repeated sampling are often required.

Analytical approaches: Simple analyses use logistic regression or generalized linear models with a binomial error to relate occupancy to covariates. More formal frameworks separate the ecological process of occupancy from the observation process, known as occupancy models, which estimate occupancy probability psi and detection probability p using repeated surveys.

Applications and limitations: Presence-absence data are widely used to map distributions, monitor biodiversity, and assess responses to habitat change or management. They offer less information about abundance than count data and are sensitive to detectability and sampling bias. The need for repeated sampling is a practical constraint.

Relation to presence-only data: In contrast, presence-only methods rely on records of where a species has been observed without absence data, and often require background or pseudo-absence points and different modeling approaches.