Investigation into the edge effect by use of capture-recapture data in a vole population

Summary

A substantial explication of the edge effect has been attempted by use of capture-recapture data for a vole population (Microtus montebelli), gathered intwo plots of 100×100 m or less during 12 days, cheked twice daily, in August 1970; the sample was quite sufficient for the aim.

The edge effect as guessed by increased catch per trap is usually suspected to ensue from range-settlers in the outside boundary strip of a plot and immigrants. But by a theoretical analysis I could attain a tentative conclusion that no increased catch per trap will occur unless any invasion takes place. Then it follows that, apart from the effect of invasion, the role of the adjoining outside settlers in the edge effect is essentially required to be studied in the light of knowledge on the truth of size and shift in home range.

The variation in range behavior for 183 adult voles, captured 6 times or more, could be grouped into eight types, of which the range-conservative type possessed 52% of the sample and the group of the type was justly utilized for giving averages of range size. Besides, it was seen from the observed frequency of types that a considerable number of immigrants onto the census plot were induced perhaps being allured by trap baits, but the majority of them proved to be assigned to the voles that have their ranges inside the assessment line ofDice; the rest referable to effective immigrants was only a few (7%). I could perceive no reason such as disproves the idea ofDice’s additional boundary strip.

Viewed from maps of ingress shift of ranges, the effect of ingress must have been greater in the outer trap rows than in the inner within the plot, so that it might well be called edge effect in general; such effect, however, is seen gradually diminishing toward the center, and hence it is almost unlikely that one should find any clear-cut intra-plot assessment lines demarcating such an inner square as quite free from edge effects.

Averages of observed range length and width (ORL and ORW), as reliable measures for the true range size, were determined from the above group of specimens; as a result, the remarkable concept of elliptic range shape was established by regarding ORL as long axis and ORW as short one, and, directly from these averages, the mean range sizes worked out at 0.04 for females and 0.09 for males in acreage which proved to be surprisingly well agreeable with those of isotope-revealed ranges for voles given byGodfrey (1954) andAmbrose (1969).

The catchability for marked voles ($\hat \pi$) was estimated by the maximum likelihood method by use ofJolly’s formulae (1965), but that for unmarked ones ($\hat p$) was made by the regression census formula; as a result it was shown that the population was clearly of π>p type and that the trap-experience that voles underwent one month or more ago can make them retain as high catchability as π.