Home > Research > Research Results > Research Results 2018 > A method for estimating species richness and densities by distinguishing imperfect detection and temporary movement

Update:January 25, 2018

Main content starts here.

A method for estimating species richness and densities bydistinguishing imperfect detection and temporary movement


Article title

Community distance sampling models allowing for imperfect detection and temporary emigration

Author (affiliation)

Yuichi Yamaura (a)、J. Andrew Royle (b)

 (a) Department of Forest Vegetation, FFPRI, Tsukuba, Ibaraki, Japan.
 (b) U.S. Geological Survey, Patuxent Wildlife Research Center, Laurel, Maryland, USA.

Publication Journal

Ecosphere, 8(12), December 2017, DOI: 10.1002/ecs2.2028( External link )

Content introduction

In field surveys, it is difficult to detect all animals; thus, in many cases, individuals remain undetected.
Over recent years, we have been proposing models that estimate the number of species and individuals (i.e., community models) for incompletely sampled survey data.
However, the home range of animals is often larger than the survey area, and animals are not necessarily present in the study area at the time of survey.
Under these circumstances, animals cannot be detected in the survey area.
Previous models that do not account for movements in and out of the survey area deal with individuals outside of the survey area as those undetected in the survey area. Therefore, such models can overestimate the density of animals.

Therefore, in this study, we developed a community model that distinguishes between imperfect detection and temporary movement out of the survey area.
A distance sampling is a method (Note) that records the distance from the investigator to the animal when an animal is found in the study area.
By repeating distance sampling multiple times, it is possible to estimate the population density of each species as well as species richness and densities of the animal community, considering incomplete detections and temporary movements separately.

When this model was applied to survey data on birds, the probability of animal movement outside of the survey area was as small as 0.14 in skylarks and as large as 0.98 in blue-and-white flycatchers. Therefore, probabilities of temporary movements largely depend on the species. If temporary movements are ignored, bird densities and species richness can be overestimated.
To estimate the species richness and animal densities in field surveys, accounting for individuals and species with a large home range has remained a long-standing challenge.
By distinguishing detection rates and temporary movement rates, it becomes possible to quantify animal diversity more accurately, thereby improving our understanding of the significance of biodiversity in nature.


Note: The distance sampling is a method to record the distance from the investigator to the detected individual.
We model the process by which the discovery rate drops as distance between the detected individual and investigator increases. By considering the existence of undetected individuals, we can estimate the population density of animals in the study area.
Because the model can estimate detection rates and densities in one survey, it can be applied not only to birds and terrestrial mammals but also to marine mammals.


A previous community model.A new community model


Photo: Black woodpecker

Photo. Black woodpecker feeding on a stump at a logging site in a forest plantation.

Woodpeckers, including the black woodpecker and great spotted woodpecker, are known for their wide home ranges.

In conventional models that do not distinguish between the detection rate and temporary movement rate, the density of these bird species is overestimated, resulting in an overestimation of species numbers and densities.

It is expected that one can appropriately estimate species numbers and densities by distinguishing detection and temporary movement rates via repeated distance sampling in a survey area.