Geoscience Reference
In-Depth Information
9
Occupancy
Models
Darryl MacKenzie
9.1 Introduction
Occupancy models are a set of techniques for investigating the presence-
absence of a species while accounting for the fact that the species may be
present but goes undetected by the survey methods used at some locations.
Such methods can be useful in a wide range of ecological applications, such
as species distribution modeling, metapopulation studies, habitat modeling,
and resource selection functions, and can also be useful for monitoring pro-
grams (MacKenzie et al., 2006). They are useful not only for assessing patterns
in species occurrence, such as identifying important habitat relationships to
the current distribution, but also for understanding and predicting changes
in species distributions through time. Occupancy models have also been
extended beyond presence-absence data (or, more generally, two categories
for the status of the species at a unit) to multiple categories (e.g., absence and
presence with or without breeding; or none/some/many individuals) (Royle
and Link, 2005; Nichols et al., 2007; MacKenzie et al., 2009).
Study design is a key aspect for the correct application of occupancy mod-
els. Without reliable data, inferences from occupancy models, as from any
modeling exercise, might be unreliable. Through a careful design process,
conclusions resulting from the modeling are likely to be more accurate, and
realistic expectations about the success of the study or monitoring program
can be set.
This chapter reviews some of the main features of occupancy models from
both analysis and design perspectives. The chapter begins with a general
overview of the type of data and basic sampling requirements, followed by
introductions to models that can be used to address questions concerned
with species occurrence during a single period of time (single-season mod-
els) and changes in occurrence (multiseason models). Study design issues are
then discussed, and the chapter ends with a general discussion.
123
