Environmental Engineering Reference
In-Depth Information
available. This can be a difficult judgement to make, and every situation is unique,
but it is possible to make some
general recommendations
. Figure 2.9 summarises
the most important factors that determine the kinds of techniques that are most
likely to be applicable in a given situation.
In general, the more complex the model used to estimate the proportion of the
population seen, the greater the model error, and the lower the precision will be.
On the other hand, making strong assumptions that are difficult to evaluate or sat-
isfy is likely to cause serious bias. We are therefore looking for the simplest possible
analytical approach that is practically possible without violating key assumptions.
So although complete census is conceptually the simplest possible approach, the
assumption that the entire population can be observed is so difficult to satisfy in
practice that the method is almost never appropriate. Plot sampling relaxes this
assumption by restricting search to more manageable areas, and remains relatively
simple, so is the best choice when the species is easily seen, relatively abundant, and
static. By adding a little more analytical complexity, distance sampling (line
or point transects) can be applied to less visible or more mobile species, and has a
relatively low risk of bias. It is therefore a good choice in a wide range of cases, and
Species
visibility
Extremely
good
Moderate
or good
Extremely
good
Moderate
or good
Poor
Poor
Distance
sampling
(points or
transects)
Plot
sampling
Harvest
data
Harvest
data
Poor quality
or absent
Good
quality
Poor quality
or absent
Good
quality
Offtake-
based
methods
Species
catchability
Low
High
Low
High
Technical
capacity
Mark-
recapture
(trapping)
or
distance
sampling
(trapping webs)
Low
High
Low
High
Presence /
absence
or
sign
surveys
Mark-
recapture
(camera trap
or genetic)
or
presence/
absence
(camera trap)
Fig. 2.9
A decision tree for identifying the most appropriate census method for a
given situation. This provides a general guide for the most likely approach, but
does not rule out innovative uses of apparently inappropriate methods.
