where K 1 is the carrying capacity of the environment for individuals of deermice when
alone, N 1 is the number of deermice, N 2 is the number of voles, and 20.75is the con-
version factor, and standardizes the species in terms of their metabolic rates.
The body weight (W) of voles is about two times that of deermice, and the basal
metabolic rate (M) is taken as M=W0.75(see Section 4.5.2). Using various com-
binations of N 1 and N 2 an average estimate ofa=0.06 was obtained.
Properly designed removal experiments are difficult to carry out for practical
reasons, so it is not surprising that they have not yet been performed with large
mammals.An easier approach uses natural absences or combinations of species to observe re-
sponses that would be predicted from interspecific competition. For example, mal-
lard ducks (Anas platyrhynchos) breed on oligotrophic (low nutrient) lakes in Sweden
(Pehrsson 1984). Some of the lakes contained fish while others did not. In lakes with
fish, the density of mallards was lower, mean invertebrate food size was lower, and
emerging insects were significantly smaller. In an experiment where ducklings were
released, their intake rate was higher on lakes without fish (Table 9.2). These results
imply competition between ducks and fish.
Another type of natural experiment is illustrated by the distributions of two ger-
billine rodent species in Israel (Abramsky and Sellah 1982). One species, Gerbillus
allenbyi, lives in coastal sand dunes and is bounded in the north by Mt Carmel. In
the same region the other species, Meriones tristrami, is restricted to non-sandy
habitats. In the coastal area north of Mt Carmel, M.tristramioccurs alone and
inhabits several soil types including the sand dunes. Abramsky and Sellah suggested
that M.tristramicolonized from the north and was able to bypass Mt Carmel, whereas
G.allenbyicolonized from the south and could not pass the Mt Carmel barrier. In
the region of overlap, south of the barrier, interspecific competition had excluded M.
tristramifrom the sand dunes. They tested this hypothesis by removing G.allenbyi
from habitats where the two species overlapped, and found that there was no
significant increase in M.tristrami. They concluded that there was no present-day
competition occurring. Instead they suggested that competition in the past had
resulted in a shift in habitat choice so that there was no longer any detectable
competition.
Islands are sometimes used to look at the distributions of overlapping species, because
on some islands a species can occur alone while on others it overlaps with related
species. The theory of interspecific competition would predict that when alone a species
would expand the range of habitats it uses (a process we call competitive release),
while on islands where there are several species the range of habitats contractsCOMPETITION AND FACILITATION BETWEEN SPECIES 141
Year Lakes without fish Lakes with fishMean dry weight 1977 119.8 (21.0) 45.3 (13.7)**
(May–June) 1978 159.0 ( 9.9) 26.5 ( 4.8)**
Duckling feeding 1977 12.4 ( 0.6) 9.5 ( 0.5)***
(food items/min) 1978 20.4 ( 5.1) 7.9 ( 0.7)****P<0.01; ***P<0.001.
From Pehrsson (1984).Table 9.2Mean dry
weight of subaquatic
invertebrates available
to mallard ducklings
and the rate of duckling
food intake in Calnes
with and without fish
in Sweden.
9.3.2Natural
experiments