untitled

of five species of warblers within conifer trees in the northeastern USA. They varied

in both height in the tree and use of inner or outer branches. Nudds et al. (1994)

found habitat partitioning in dabbling ducks in both Europe and North America. Species

with a high density of filtering lamellae in their bills (fine filter feeders) tended to

live in deep water with short, sparse vegetation compared with those species with

few lamellae which lived in shallow water with tall, dense vegetation.

As we have mentioned above it should not be surprising that species divide up the

resource available to them. However, Gause’s principle implies that there should be

a limit to the similarity of niches allowing coexistence of two species. Earlier stud-

ies predicted values of limiting similarity based on theoretical arguments (MacArthur

and Levins 1967). If the distance between the midpoints of species distributions along

the resource axis is dand the standard deviation of the curves (such as those in Fig.

9.7a) isw(the relative width) then limiting similarity can be predicted from the ratio

d/w. However, various assumptions, such as the curves must be similar, normally dis-

tributed, and along only one resource axis, make this approach unrealistic.

Pianka et al. (1979) asked: how much would niches overlap if resources were allo-

cated randomly among species in a community? A frequency distribution of niche

overlaps generated from randomly constructed communities is shown in Fig. 9.12.

COMPETITION AND FACILITATION BETWEEN SPECIES 147

Duiker

Bushbuck

Greater Kudu

Zebra

Impala

Rhinoceros

Reedbuck

Steinbuck

Tsessebe

Grassland Savanna Woodland

Fig. 9.10Habitat
partitioning and overlap
by ungulates in Kyle
National Park,
Zimbabwe. The width
of the boxes reflects the
degree of preference.
(After Ferrar and
Walker 1974.)

9.6.2Limiting
similarity