Because riparian forest strips are perceived to buffer aquatic ecosystems f
rom logging-related disturbance, they are usually not harvested. However, t
heir value as refuges for terrestrial wildlife is unknown. We conducted two
live-trapping experiments in the riparian zone adjacent to rivers in a bor
eal balsam fir (Abies balsamea) forest in Quebec. In the first experiment,
we compared late summer use, during 4 separate years, of different width ri
parian strips (20, 40, 60 m, and control [>300 m wide]), and different stan
d thinning intensities (20 m intact and 20 m thinned of; of all trees) on r
esident small mammals. We found no differences in the densities of the most
common species, Clethrionomys gapperi and Peromyscus maniculatus, among st
rip types or among years (P>0.05). We also tested for edge effects in large
strips (60 m and controls). In controls, C. gapperi was less abundant in t
he first 20 m adjacent to the river (P=0.004) while P. maniculatus was more
abundant (P=0.02) in that area. Neither species, however, showed an edge e
ffect in the 60 m-strips (P>0.10). In the second experiment, we monitored s
mall mammals over eight consecutive weeks in a 160 mx170 m quadrat enclosin
g a 20 m-thinned forest strip and st clear-cut to investigate some aspects
of the role of riparian strips at the landscape scale. During that time, Mi
crotus pennsylvanicus, which was nearly absent from our study area in the p
revious years, invaded the clear-cuts and apparently confined C. gapperi an
d P. maniculatus to forest remnants such as 20 m-wide strips. A conclusion
that emerges from this study and related studies on birds is that some spec
ies prefer larger strips or non-riparian habitats whereas others prefer nar
row strips along riparian habitats. We recommend that managers ban the all-
encompassing norms and manage for heterogeneity at different scales. Becaus
e our study was conducted at the stand scale and because it is not accompan
ied with an evaluation of the socio-economic aspects of riparian management
, we cannot determine the proper mixture of strips in the landscape. Howeve
r, our results could help managers to enhance the key-role of riparian ecos
ystems in maintaining regional biodiversity and contribute to the maintenan
ce of local biodiversity by creating refuges for terrestrial wildlife. (C)
2001 Elsevier Science B.V. All rights reserved.