POPULATION CYCLES IN SMALL MAMMALS
273
IV. POPULATION DENSITY CHANGES
A. T E C H N I Q U E S O F E S T I M A T I N G D E N S I T Y
Progress in defining the phenomenon of population cycles has been
limited by inadequate density data. In the simplest case we recognize
only two density states: “high” density and “low” density. Next we
can obtain an index of density by the use of sampling with traps,
surveys for runways or fecal pellets, or visual sightings. Much of the
work on small rodents has utilized kill traps of various sorts to provide
an index of population density. Trap catches are a function both of
density and of activity patterns. If voles have large home ranges one
year and small home ranges the next, a trapping index will decrease
even if actual densities are the same in the two years. Individual
trappers vary greatly in the ability to set traps in good locations,
and this factor can add to the variance in trap catches. In general,
indices of density obtained by trap sampling w i l l show trends of
density changes but cannot be interpreted quantitatively.
Absolute density estimates can also be obtained by removal trapping.
This method was first developed by Leslie and Davis (1939) and independently derived by DeLury (1947). As animals are removed from
an area, the catch per unit of trapping effort will fall off and reach
zero at the point where the whole population has been removed.,If we
assume constant trappability of the whole population and no immigration, we can use linear regression techniques to estimate the size of the
population being trapped. Unfortunately, two serious problems have
plagued this approach (Smith et al., 1971). First, the probability of
capture is not constant for the whole population (Tanaka, 1960).
And second, immigration occurs once the removal trapping begins.
This forces one to try to measure the area depopulated by the kill
traps, an area which may be several times greater than the actual area
occupied by traps (Smith et al., 1971). The area affected by trapping is
difficult to determine in removal studies. A more basic limitation of
this approach is that it destroys the population we should be trying to
study, and consequently mark-and-release techniques have been
utilized for long-term studies.
Mark-and-recapture techniques permit an accurate measurement of
density. Since the pioneering work of Leslie et al. (1953), there has
been available a continuously improving series of statistical techniques
for this estimation problem (Cormack, 1968). The application of markand-recapture techniques requires an assumption of randomness of
capture of marked and unmarked voles. The randomness of capture
assumption has been tested on only a few vole populations, and in no
273
IV. POPULATION DENSITY CHANGES
A. T E C H N I Q U E S O F E S T I M A T I N G D E N S I T Y
Progress in defining the phenomenon of population cycles has been
limited by inadequate density data. In the simplest case we recognize
only two density states: “high” density and “low” density. Next we
can obtain an index of density by the use of sampling with traps,
surveys for runways or fecal pellets, or visual sightings. Much of the
work on small rodents has utilized kill traps of various sorts to provide
an index of population density. Trap catches are a function both of
density and of activity patterns. If voles have large home ranges one
year and small home ranges the next, a trapping index will decrease
even if actual densities are the same in the two years. Individual
trappers vary greatly in the ability to set traps in good locations,
and this factor can add to the variance in trap catches. In general,
indices of density obtained by trap sampling w i l l show trends of
density changes but cannot be interpreted quantitatively.
Absolute density estimates can also be obtained by removal trapping.
This method was first developed by Leslie and Davis (1939) and independently derived by DeLury (1947). As animals are removed from
an area, the catch per unit of trapping effort will fall off and reach
zero at the point where the whole population has been removed.,If we
assume constant trappability of the whole population and no immigration, we can use linear regression techniques to estimate the size of the
population being trapped. Unfortunately, two serious problems have
plagued this approach (Smith et al., 1971). First, the probability of
capture is not constant for the whole population (Tanaka, 1960).
And second, immigration occurs once the removal trapping begins.
This forces one to try to measure the area depopulated by the kill
traps, an area which may be several times greater than the actual area
occupied by traps (Smith et al., 1971). The area affected by trapping is
difficult to determine in removal studies. A more basic limitation of
this approach is that it destroys the population we should be trying to
study, and consequently mark-and-release techniques have been
utilized for long-term studies.
Mark-and-recapture techniques permit an accurate measurement of
density. Since the pioneering work of Leslie et al. (1953), there has
been available a continuously improving series of statistical techniques
for this estimation problem (Cormack, 1968). The application of markand-recapture techniques requires an assumption of randomness of
capture of marked and unmarked voles. The randomness of capture
assumption has been tested on only a few vole populations, and in no