POPULATION CYCLES IN SMALL MAMMALS
311
in the increase and peak phases, and are high in the decline phase and
also in the phase of low numbers. Juvenile losses are high in the peak
phase and in the decline phase. Prenatal mortality does not seem to
vary systematically during the population cycle.
c. D I S P E R S A L
Population densities can change because of variations in birth,
death, or dispersal rates, and almost all population studies on lemmings
and voles have been concerned only with b i r t h and deaths. The
simplest dynamic assumption is that immigration cancels emigration
and the population changes are solely a function of birth and death
rates. This simple view would be adequate if there were no spatial
heterogeneity in nature and no marginal habitats for small rodents
(Anderson, 1970).
The importance of dispersal in population regulation of voles W&B
first shown by studies on enclosed populations. Clarke (1955) showed
that Microtus agrestis populations in large cement cages (67 m2) would
increase to numbers far in excess of those ever found in natural areas.
He obtained a population “high” at 58 individuals, which is equivalent
to 3500 per acre (8657per ha), about ten times higher than ever occurs in
nature. Van Wijngaarden (1960) obtained densities of M . arvalis up to
7.25 voles per m2 in 100 m2 pens, which is equivalent to 29 300 voles
per acre (72 500 per ha), approximately 100 times higher than natural.
The same results were reported by Louch (1956) for M . pennsylvanicus,
Frank (1953) for M . arvalis, and Houlihan (1963) for M . californicus.
These studies on enclosed vole populations are difficult to interpret
because animals are maintained on artificial food with little or no
predation, and dispersal is prevented. The next step was to study
an enclosed population in a natural habitat with a normal complement
of predators and natural forage. Krebs et aZ. (1969) studied populations
of Microtus pennsylvanicus and M . ochrogmter enclosed in a two-acre
(0.8 ha) grassland surrounded by a wire fence extending two feet above
ground. These populations behaved in the same way as the confined
laboratory populations-they increased to abnormally high densities
and then decimated the natural forage (Fig. 25). Since severe overgrazing is rarely seen in natural grasslands, we concluded that by
preventing dispersal we had destroyed the ability of the population to
regulate its density at a level below that of gross starvation.
Few studies have been made of vole populations in large enclosures
in the field. Gentry (1968) observed that M . pinetorum in a two-acre
enclosure reached densities far above those in natural habitats. He
311
in the increase and peak phases, and are high in the decline phase and
also in the phase of low numbers. Juvenile losses are high in the peak
phase and in the decline phase. Prenatal mortality does not seem to
vary systematically during the population cycle.
c. D I S P E R S A L
Population densities can change because of variations in birth,
death, or dispersal rates, and almost all population studies on lemmings
and voles have been concerned only with b i r t h and deaths. The
simplest dynamic assumption is that immigration cancels emigration
and the population changes are solely a function of birth and death
rates. This simple view would be adequate if there were no spatial
heterogeneity in nature and no marginal habitats for small rodents
(Anderson, 1970).
The importance of dispersal in population regulation of voles W&B
first shown by studies on enclosed populations. Clarke (1955) showed
that Microtus agrestis populations in large cement cages (67 m2) would
increase to numbers far in excess of those ever found in natural areas.
He obtained a population “high” at 58 individuals, which is equivalent
to 3500 per acre (8657per ha), about ten times higher than ever occurs in
nature. Van Wijngaarden (1960) obtained densities of M . arvalis up to
7.25 voles per m2 in 100 m2 pens, which is equivalent to 29 300 voles
per acre (72 500 per ha), approximately 100 times higher than natural.
The same results were reported by Louch (1956) for M . pennsylvanicus,
Frank (1953) for M . arvalis, and Houlihan (1963) for M . californicus.
These studies on enclosed vole populations are difficult to interpret
because animals are maintained on artificial food with little or no
predation, and dispersal is prevented. The next step was to study
an enclosed population in a natural habitat with a normal complement
of predators and natural forage. Krebs et aZ. (1969) studied populations
of Microtus pennsylvanicus and M . ochrogmter enclosed in a two-acre
(0.8 ha) grassland surrounded by a wire fence extending two feet above
ground. These populations behaved in the same way as the confined
laboratory populations-they increased to abnormally high densities
and then decimated the natural forage (Fig. 25). Since severe overgrazing is rarely seen in natural grasslands, we concluded that by
preventing dispersal we had destroyed the ability of the population to
regulate its density at a level below that of gross starvation.
Few studies have been made of vole populations in large enclosures
in the field. Gentry (1968) observed that M . pinetorum in a two-acre
enclosure reached densities far above those in natural habitats. He