Mark-recapture Distance Sampling

Mark-recapture methods and distance sampling methods are the two most widely-used methods of assessing and monitoring wildlife populations. Each has its strengths and weaknesses and for many species one or the other is more appropriate. In some situations neither method is adequate on its own and this has led to the development of "mark-recpature distance sampling methods" for such cases . These combine the strengths of both methods to deal with situations in which some of the assumptions of each method do not hold. From a distance sampling perspective, this is when the assumption of certain detection of animals at distance zero fails (indicated by histogram bars being below the black curve in the figure). From a mark-recapture perspective, this is when the assumption that animals at different distances are equally detectable.

The methods were initially developed for surveying whales and dolphins from ships or planes, but they have since been developed for many other types of survey and species. Spatial capture-recapture methods can be viewed as an extension of mark-recapture distance sampling methods.

Are the animals available to be detected?

Some species of conservation interest are not always available to be detected during a survey - even if the observer is in close proximity to the animal. For example the species of tortoise shown in the mural, the gopher tortoise, spends a considerable time in burrows. This makes it hard for observers to detect them during, for example, a visual line transect survey (see also Distance Sampling).

In the mural, the histogram bars indicate the frequencies of detections with respect to distance x from the line during a line transect survey. The black line represents the estimated detection function g(x) which intersects the y-axis at g(0)=1 using conventional distance sampling methods. The discrepancy between g(x) and the top of the histogram bars indicates that the observer missed a lot of animals and detection probabilities were overestimated. This problem can be addressed with mark-recapture distance sampling methods.

histogram and tortoise

Histogram with frequencies of detections along increasing distance from the transect line.

CREEM logo
University of St Andrews