Geoscience Reference
In-Depth Information
where the items of interest are either present or the number exceeds some
threshold. Similarly, two-phase adaptive sampling involves taking samples
from different geographical parts of a population (strata) and then deciding
where to do more sampling based on what is found in the initial sampling.
The idea is then that the second round of sampling can be in those parts
of the population where more sampling will be most beneficial. These and
the other methods considered in Chapter 3 are generally concerned with
estimating either the total abundance or density per unit area of animals or
plants in a region.
Chapter 4, “Line Transect Sampling,” is mainly concerned with the situ-
ation when one or more observers travel along a line in a region of interest
and record the number of objects of interest seen (again animals or plants)
and the distance of these objects from the line. Typically, this may involve
following a line on the ground, flying in an aircraft and observing objects
on the ground, or moving along a path in a boat over a sea or lake and
recording the objects seen. It is generally assumed that the probability of
detecting an object depends on its distance from the line, with a low prob-
ability of detection for objects far away. The data are then used to estimate
a detection function and then estimate the total number of objects within
a certain distance from the line and the density of objects per unit area
around the line. This method of sampling is usually used when alternative
methods such as randomly sampling quadrats in the region of interest are
not practical for some reason.
Chapter 5, “Removal and Change-in-Ratio Methods,” discusses two meth-
ods for estimating the size of animal populations. With the removal method,
an animal population is sampled a number of times, and the animals caught
are removed or possibly marked so that if they are caught again it will be
known that they were seen before so that they are effectively “removed.”
Assuming a closed population (i.e., a population in which animals are not
entering and leaving between samples), the number of animals left in the
population will decrease every time a sample is taken, so that the number
of animals available for removal will decrease with time and will be zero
if enough samples are taken. At that time, the population size will consist
of the total number of animals removed in the earlier samples. However,
it is not necessary to remove all of the animals to obtain an estimate of the
total population size; Chapter 5 describes how the results from several sam-
ples can be used to estimate the total number of animals in a population
even though some of these animals have not been seen. The change in ratio
method is similar but involves a population in which there are two or more
recognizable types of animal in a population, such as males and females or
juveniles and adults. Then, a sample is taken, and a fixed number of one of
the animal types is removed or possibly marked so that a marked animal is
considered removed. If a second sample is taken, then it becomes possible to
estimate the population sizes at the time of the first sample and at the time
of the second sample (which is just the first sample size minus the number of
