Species occupancy modelling for detection data collected along a transect

The proportion of sampling sites occupied by a species is a concept of interest in

ecology and biodiversity conservation. Occupancy surveys based on collecting detec-

tion data along transects have become increasingly popular to monitor some species.

To date, the analysis of such data has been carried out by discretizing the data, divid-

ing the transects into discrete segments. Here we propose alternative occupancy models

which describe the detection process as a continuous point process. These models pro-

vide a more natural description of the data and eliminate the need to divide transects

into segments, which can be arbitrary and may lead to increased bias in the estimator of

occupancy or increased chances of obtaining estimates on the boundary of the param-

eter space. We present a model that assumes independence between detections and an

alternative model that describes the detection process as a Markov modulated Poisson

process to account for potential clustering in the detections. The utility of these models

is illustrated with the analysis of data from a recent survey of the Sumatran tiger Pan-

thera tigris sumatrae. The models can also be applied to surveys that collect continuous

data in time, such as those based on the use of camera-trap devices. Supplementary ma-

terials for this article are available online.