Environmental Engineering Reference
In-Depth Information
Day
12
8
4
0
12
Night
8
4
0
0
20
Distance (m)
40
Fig. 2.5
Duikers such as this blue duiker, caught in a snare in Equatorial
Guinea, are an extremely common target for bushmeat hunters in west and
central Africa. Unfortunately, it's very difficult to estimate their abundance
using distance sampling because they are good at moving away from people
before being detected, as shown by the humped day time detection function.
The result is better using torchlight by night, but this is much harder work for
the surveyors. Photo © Noelle Kumpel.
to get round the problem of evasive movements during the day, all observations
within 4 m of the transect were discarded prior to analysis (a technique known as
left-truncation). However, despite left-truncation, duiker density estimated at
night was around three times higher, and presumably more accurate, than the
daytime estimate (estimates and 95% confidence intervals, daytime: 36 km
2
[24-56]; night-time: 101 km
2
[68-150]). As is usually the case, left-truncation
was unable to fix the problem of evasion.
This technique is generally applied to animals that are difficult to see in the wild, but
which can somehow be recognised as having previously been observed when
encountered. Usually this is achieved by catching, marking and recapturing indi-
viduals. By collecting information on the individuals that are captured and recap-
tured over two or more different occasions, it is possible to estimate the proportion
of the population that has been captured, and so calculate the total population size
(Box 2.4). In essence, the method works as follows. On the first of two capture
occasions, individuals are caught and marked in some way, then released to mix with
the remaining individuals that were not caught. On the second occasion, provided
that thorough mixing of marked and unmarked individuals has taken place, the
