×

You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Title
Remote sensing for resources development and environmental management
Author
Damen, M. C. J.

ies
ities
ies
ommunities
naged)
ged)
ss
il legend is
ation exists
fs therefore
r individual
lecessary is
communities
he basis of
iwances being
he survey's
hotographic
checked for
d identified
frequently
bracken.
idual legend
erpretation
he tall herb
uracy level
was better
a result of
of the 1981
983. A good
'lace in the
terpretation
the 1984
rrect. The
n 1979 and
■ r of errors
using air
population
successful.
entrated on
bers, bird
¡rsity using
such as area
diversity,
i woodland
vestigate if
sed in these
techniques can be obtained from air
photographs, and therefore, used to predict
bird carrying capacity. A bird census was
conducted in Saltswells Wood in 1983 by
Harrison and Normond. These local
ornithologists identified the bird species
which were breeding, their numbers and the
approximate location of their territories.
This field data was used as the control in
this analysis.
One of the simplist and most common
variables used for prediction is area of
woodland. This relationship was first
suggested by Arrhenius 1921. Woolhouse 1983
and Moore and Hooper 1975 have both conducted
studies in British woodlands suggesting a
linear relationship between the number of
bird species present and the area of
woodland.
CL ;
if) /
LL ;
O
O
z
AREA
Hypothetical species-area curve
S=cA z
AREA (ha)
Species-area curve on a logarithmic
scale lnS=clnA z
from Woolhouse 1981
FIG 3.
The area of the classifed habitat types
was the major variable to be investigated.
The area information was obtained by using a
Hewlett Packard digitizer connected to a BBC
micro computer using a digitising program.
The first stage was to identiy a list of
bird species which were likely to be found
breeding in Saltswells wood and relate their
nesting and feeding requirements to the
classification categories identified.
Fifty-seven species commonly found in
woodlands of a similar size and found
breeding in the West Midland were earmarked.
These birds were allocated one or more of the
classification categories using the 'Handbook
of British Birds' (Witherby et al 1965). For
example a Blackbird is likely to be found
breeding and feeding in categories
34/43/44/45, that is, in deciduous shrubby
habitats. Some bird species are limited to
a single category such as the Wood Warbler.
This species was only allocated closed
deciduous woodland - 31. Other species need
areas of woodland and grassland to fulfil all
their requirements. The system allowed for
this, for example the Common Crow needs
32/33+51/52/53/58, that is, open deciduous
woodland for nesting and grassland areas for
feeding. As some species have similar
requirements 32 habitat patterns were
identified for the 57 species.
The equations relating bird species numbers
to areas of woodland suggested by Woolhouse
and Moore and Hooper were investigated.
Different habitat combinations and their
corresponding area totals were used in these
equations to identify which, if any, habitat
combination produced a result similar to 27 -
the actual number of bird species found
breeding in the area by the control study.
For results see Table 1.
1 Woolhouse equation =
In S = 0.227 In A + 2.632
2 Moore and Hooper equation =
In S = 0.271 In A +- 0.26
Where In S = Natural log of the number of
species
In A = Natural log of area
Table l.The relationship between woodland
area and the number of bird species.
Estimated number of species
Area (ha)
Eq. 1
*1
Eq. 2
*2
Reserve
40.2
32.2
+5.2
2.1
-24.9
Hab. 31
14.4
25.66
-1.34
1.6
-25.4
Hab. 34
7.0
21.59
-5.4
1.23
-25.7
Hab.31+34
21.4
27.84
+ 0.84
1.76
-25.2
Hab. 43,44,
45,
10.65
23.8
-3.2
1.46
-25.5
Hab.31,34,
30.56
+3.56
1.97
-25.0
43,44,45
32.1
* = the difference between the actual number
of species,27 and the predicted number.
There is a large discrepancy between the
number of bird species predicted by the two
equations. The Moore and Hooper equation
obviously under estimates the number of
species expected in this area therefore only
the figures produced by the Woolhouse
equation were considered.
Of the different combinations of habitat
units the total area of 31 and 34 provides
the most accurate estimate of the number of
bird species. Two apparent conclusions can
be drawn.