Environmental Engineering Reference
In-Depth Information
40
20
0
0
50
100
150
Harvesting days per year
Fig. 2.6
Even relatively light harvesting pressure can dramatically reduce the
abundance of crayfish in forest streams in Madagascar. These results were
derived from intensive mark-recapture estimates of abundance.
Source
: Jones et al. (2005).
of models were tested, including the possibilities of size-dependent capture
probability, and of different first and subsequent capture probabilities. Rather
than selecting a single 'best' model, capture probabilities and population sizes were
estimated by averaging across all candidate models, weighting model contribu-
tions by the strength of evidence in the data (program MARK includes a model
averaging facility to do this). The results indicated no discernible relationship
between harvesting pressure and abundance for smaller crayfish, but a very strong
reduction in the abundance of large crayfish in sites that were harvested to any
significant extent (Figure 2.6).
For larger animals, practical difficulties and risks to the animals usually pro-
hibit capture on a sufficiently large scale to make mark-recapture a viable option.
However, Skalski
et al
. (2005a) got round this problem by marking elk
Cervus
elaphus
with paint-ball guns fired from helicopters in a remote part of Washington
State, USA. In a single day, 55 elk were marked in a discrete population defined
by watersheds. During a two hour flight four days later, 54 elk were seen, of which
36 were marked. This provides the necessary information for a simple two-sample
Lincoln-Petersen estimate of 82 elk in the population. However, this result is
likely to be an underestimate caused by violated model assumptions. At the time
of the survey (late March), elk form herds which would have been quite stable
over the short period between marking and resighting, and there was probably
therefore too little mixing of marked and unmarked individuals. Added to this,
larger herds tend to be easier to find, resulting in individual heterogeneity in
resighting probability. Because the marking method does not allow recognition of
individuals, it is impossible to detect and correct for this heterogeneity, so it is
essential to try to avoid the problem through careful survey design. In this case,
Skalski
et al
. (2005a) recommend carrying out surveys in the summer, when herds
are more fluid, smaller and less variable in size. This runs counter to previous prac-
tice favouring winter surveys, when large herds allow many individuals to be
marked in a short space of time.
