POPULATION CYCLES IN SMALL MAMMALS
371
peak populations. Male mortality is thus much greater than female
mortality, and males fight vigorously for territories and show more
wounding.
The pattern of changes described by Frank (1957) for M . arvalis
should be evident in a changing sex ratio over a population fluctuation.
There has been no evidence from any of the studies on M . agrestis
(Chitty, 1952), M . ochrogaster and M . pennsylvanicus (Myers and
Krebs, 1971a) and M . californicus (Krebs, 1966) that sex ratios change
systematically over the population cycle. There is often a surplus of
females in the trappable population, but this seems to be a constant
feature of all the phases of the cycle. The absence of any sex ratio
change may mean that M . arvalis has a different pattern of social
structure to these other Microtus species.
More detailed studies are needed of the social structure of vole
and lemming populations. Some aspects of social behavior can be
studied in enclosed pens, but we badly need ways of studying field
populations in situ. Perhaps the simplest approach we can take at
this stage is to study male-male, male-female and adult-young interactions in a laboratory system, and then to use these observations as a
basis for designing field experiments. For example, we might manipulate
social organization by artifically changing the sex ratio in natural
populations. Alternatively, by cropping one sex and age group in a
population we might gain some insight on how social organization
influences population density.
An alternative series of experiments might involve the modification
of behavior with drugs. Implants of testosterone, for example, might
be used to make selected males aggressive. Nothing has been done so
far on vole and lemming populations to alter behavior with drugs.
We have said little so far about the possibility that behavioral
interactions can cause the mortality associated with declining populations. Except for the observations of Koshkina on Clethrionomys
rutilus, there is no evidence that behavioral interactions between voles
or lemmings are lethal, and it seems highly unlikely that adult animals
often die as a direct result of fighting. Some fraction of the loss of small
juveniles might be due to aggressive interactions with adults, and more
evidence is needed on this point. But we have studied fenced populations of Microtus ochrogaster and M . pennsylvanicus at very high
densities and not found any significant mortality of subadult and adult
voles (Krebs et al., 1969). If behavioral interactions are important
causes of population declines, they must act by forcing individuals to
succumb to other agents of destruction. A model for this suggestion
can be found in the muskrat (Errington, 1967) and the red grouse
(Jenkins et al., 1964). Socially inferior voles (non-aggressive animals)
371
peak populations. Male mortality is thus much greater than female
mortality, and males fight vigorously for territories and show more
wounding.
The pattern of changes described by Frank (1957) for M . arvalis
should be evident in a changing sex ratio over a population fluctuation.
There has been no evidence from any of the studies on M . agrestis
(Chitty, 1952), M . ochrogaster and M . pennsylvanicus (Myers and
Krebs, 1971a) and M . californicus (Krebs, 1966) that sex ratios change
systematically over the population cycle. There is often a surplus of
females in the trappable population, but this seems to be a constant
feature of all the phases of the cycle. The absence of any sex ratio
change may mean that M . arvalis has a different pattern of social
structure to these other Microtus species.
More detailed studies are needed of the social structure of vole
and lemming populations. Some aspects of social behavior can be
studied in enclosed pens, but we badly need ways of studying field
populations in situ. Perhaps the simplest approach we can take at
this stage is to study male-male, male-female and adult-young interactions in a laboratory system, and then to use these observations as a
basis for designing field experiments. For example, we might manipulate
social organization by artifically changing the sex ratio in natural
populations. Alternatively, by cropping one sex and age group in a
population we might gain some insight on how social organization
influences population density.
An alternative series of experiments might involve the modification
of behavior with drugs. Implants of testosterone, for example, might
be used to make selected males aggressive. Nothing has been done so
far on vole and lemming populations to alter behavior with drugs.
We have said little so far about the possibility that behavioral
interactions can cause the mortality associated with declining populations. Except for the observations of Koshkina on Clethrionomys
rutilus, there is no evidence that behavioral interactions between voles
or lemmings are lethal, and it seems highly unlikely that adult animals
often die as a direct result of fighting. Some fraction of the loss of small
juveniles might be due to aggressive interactions with adults, and more
evidence is needed on this point. But we have studied fenced populations of Microtus ochrogaster and M . pennsylvanicus at very high
densities and not found any significant mortality of subadult and adult
voles (Krebs et al., 1969). If behavioral interactions are important
causes of population declines, they must act by forcing individuals to
succumb to other agents of destruction. A model for this suggestion
can be found in the muskrat (Errington, 1967) and the red grouse
(Jenkins et al., 1964). Socially inferior voles (non-aggressive animals)