untitled

species pairs with small overlaps than would be predicted by chance, implying that

interspecific competition was causing niche segregation.

As we have seen, species usually differ from each other by choosing different

resources such as food types, habitats, etc. We call this choice habitat selection. One

approach to measuring the competition coefficients has been to look at the variation

of species density in different habitats. First, the variation in density due to habitats,

and other resources, is estimated by statistical procedures such as multiple regres-

sion (Crowell and Pimm 1976). Then the remaining variance should be attributable

to interspecific competition with another identifiable species. An example of this

approach is given by Hallett (1982) in a study of 10 desert rodent species in New

Mexico. He measured habitat variables related to the common plants such as num-

ber of individuals, plant height, distance to nearest plant from trap, and percent cover.

Regression analysis was used to partition the variance in capture frequency at trap

stations due to habitat variables and competitors. Competition was observed within

one group of three species, Perognathus intermedius,Perognathus penicillatus, and

Peromyscus eremicus(Table 9.3). Although the competitive effects differed from year

to year, they were not random. Also the inhibitory effects were not symmetrical: thus

P.eremicusalways had a greater effect on the other two species than the reverse, and

similarly P.intermediushad a greater effect on P.penicillatusthan vice versa.

COMPETITION AND FACILITATION BETWEEN SPECIES 149

9.6.3Habitat

selection from

field data

25

0

100

50

0

850

300

200

100

0

0 0.5 1.0

Overlap in diet

Australia

North America

Kalahari

Number of interspecific pairs

Fig. 9.13Distributions
of observed overlap
in the diets of desert
lizards. Dietary overlap
is higher in Kalahari
lizards, where termites
comprised 41% of the
diet. (After Pianka
et al. 1979.)