Geoscience Reference
In-Depth Information
Table 4.3
Habitat Use, Availability, and Perceived Selection for Gaur and
Banteng in Thailand During the Dry Season
Habitat Availability (%)
Habitat Use (%) and Selection
Species
Habitat Type
Home
Ranges
Home
Ranges
Area 1
Area 2
Area 1
Area 2
Gaur
Evergreen
31
40
27
28 n
46 n
46+
Mixed deciduous
56
49
61
70 +
45 n
45 -
Dry dipterocarp
13
11
13
2 -
9 n
9 n
Banteng
Evergreen
30
39
7
5 -
8 -
8 n
Mixed deciduous
57
49
76
75 +
75 +
75 n
Dry dipterocarp
13
12
17
20 n
17 n
17 n
Values are rounded-off percentages from Prayurasiddhi (1997:230).
Use was compared with availability at three scales, denoted as area 1, area 2 (a somewhat smaller area
with more intensive use by radiocollared animals), and home range boundaries. Selection was assessed
with the methods of Neu et al. (1974) and Manly et al. (1993). Note changes in perceived selection
(+ selected for, -selected against, n = no perceived selection) with changes in the area being considered.
A problem with assessing selection from use alone is that use often does
change with availability. In the eagle study, tidal flats were rarely used during
high tide because they were largely unavailable. The eagles may have preferred
tidal flats to other habitats, but use of these areas depended on availability, up
to a point. When availability exceeded some threshold they did not continue
to increase their use, and when it declined below some threshold their use of
the area ceased. Between these thresholds some relationship may exist between
use and availability. However, relationships between use and availability prob-
ably vary among habitat types, and within each habitat may depend on what
other habitats are available (figure 4.2a).
It might seem that site attribute studies avert these problems because selec-
tion is inferred not from differences between use and availability, but from dif-
ferences between used sites and other sites. However, the comparative sites are
taken proportionately from those available, so the comparison is still, in a sense,
the same as in the use-availability design. Consider a study that locates 10 nest
sites in each of two available habitat types and then compares the characteris-
tics of these to 20 randomly chosen points. If one habitat type is much more
common than the other, then the 20 randomly chosen points fall mostly within
