As part of the Long-term Ecological Productivity (LTEP) study,we are monitoring the response of ground-dwelling vertebrates to three levels of coarse woody debris treatments - no treatments (control), high CWD retention, and low CWD retention. Three replicates (i.e., blocks) of the three treatments are sampled for ground-dwelling vertebrates. We sampled ground-dwelling vertebrates in the LTEP plots (located in the Isolation Block section of the Willamette National Forest) receiving the mid-seral, high and low coarse woody debris treatments and respective control plots for a total of nine plots (Table 1). Stands were naturally regenerated 80-100 yrs ago. The low CWD treatment was essentially a clearcut with the removal of large downed wood. The high CWD treatment was a clearcut but with 20% of the original basal area felled and left on site. Pre-treatment sampling on all plots was conducted October-November 1995. Another year of pre-treatment sampling was possible on replicate II and III plots during October-November 1996. Plots in replicate I were harvested after the 1995 vertebrate sampling and were sampled Fall 1996 to measure post-treatment effects. However, felling of residual stems did not occur in the high CWD treatment plot. Sampling was conducted in 1997 after harvest of all treatment plots. Again, however, the high CWD treatments were not fully implemented; residual stems remained on these plots until 1998. Sampling was conducted in 1999-2000 and after all treatments were fully implemented.
Four 100-m long transects were established in each stand, with trap stations spaced 10 m apart along a transect. Transects were distributed to adequately sample a stand without traversing the centrally located mensuration plots. Transects were equidistant from the stand perimeter and the perimeter of the mensuration plot. Bearing and distance from the corners of mensuration plots to transects, and between transects was recorded for future reference. Stations were flagged with blue and white striped flagging on branches. Stations were additionally marked with wire flagging during the Fall 1996 sampling effort.
Two types of traps were used to sample small mammals and amphibians owing to trap selectivity of species (McComb et al. 1993). We used pitfall traps to sample shrews and amphibians. We established one pitfall trap (double-deep #10 can) within one meter of each trap-station center. The top of traps was flush with ground level. Holes were punched in the bottom of the traps for drainage of rain water. To minimize trap deaths, we placed a small amount of moss on the bottom of the traps to absorb moisture and polyfiber batting on top of the moss for insolation. Additionally, we placed a pint-size juice carton filled with polyfiber batting in the traps and provided a small portion of bait. Bait consisted of a mixture of peanut butter, rolled oats, and sunflower seeds. Additionally, rodent chow and suet were placed in pitfall traps as a food source for shrews. When in use, a plastic funnel made from a 2-lb margarine tub was placed in the neck of the pitfall trap to prevent escape of captures. Between trapping periods, pitfall traps were filled in with sticks or rocks to prevent animals from entering. Sherman live-traps (7.62 x 8.89 x 22.86 cm) were used to primarily record rodents. These traps were also baited and filled as much as possible with polyfiber batting, and placed inside empty one-half gallon milk cartons to shelter captures from rain. During a trapping period, we placed two traps within two meters of a trap station, one on each side of the transect. Sherman traps were removed when a plot was not being trapped.
All three treatments within a replicate block were simultaneously trapped for two consecutive weeks, with some exceptions. Pitfall and Sherman traps were trapped for eight consecutive nights. Time-constraint searches were additionally conducted for amphibians during the 1995 trapping season. This sampling was also attempted in spring of 1997 in addition to destructive sampling in a non- experimental stand to evaluate the effectiveness of this technique. Given the very low observation rate (ca. 1 observation/2 hrs of searching), this technique was not used in subsequent years.
For each capture, we recorded station number and trap type, body mass, body and tail length, and sex and reproductive condition where possible. Captures were ear tagged with Monel #1 ear tags except for species of shrews with too small of external ears and amphibians which were toe clipped. Tag or toe number were recorded for each capture as was capture status (i.e., new, recapture). Animals were released at the site of capture. We attempted to revive all torpid animals by placing them in the pocket of field vests containing a chemical heating pad and polyfiber batting. When possible, we fed animals sugar water with an eye dropper to aid in their recovery. If individuals were not revived within several hours, they were taken to the crew's quarters and provided ample food and water and insolation. Surviving animals were released as soon as possible at the site of capture. All dead captures were frozen and made available to the OSU Fish and Wildlife Department's study skin collection. We also recorded numbers of disturbed or closed traps for calculating comparable measures of captures per trap night among plots and among trapping years.
See http://andrewsforest.oregonstate.edu/data/studies/we026/we026_supplemental_information.pdf for table 1 LTEP mid-seral and control plots monitored for treatment effects on relative abundances of ground-dwelling vertebrates and table 2 Maximum and undisturbed trap nights for each trap type by subplot.
To date (1999), 12 species of mammals and four species of amphibians were recorded over the three years of trapping. The California red-backed vole, deer mouse, shrew species (mostly Trowbridge shrew), Townsend's chipmunk, and to some extent the northern flying squirrel were the more commonly captured mammal species across all stands. Based on ANOVA, treatment differences were significant (P<0.05) for three of the six species examined. Numbers of individuals of the red-backed vole and the northern flying squirrel were lower on CWD treatments than on control plots. These trends reflect the loss of overstory cover. The strong association of this vole with downed-woody debris and the observed decrease over the treated stands suggests that CWD levels on the latter are insufficient to make up for the loss of overstory cover. The deer mouse was the only species to show a positive response to the CWD treatments. This trend fits this species' preference for shrubby, open areas. CWD treatments had no effect on shrew and chipmunk numbers. This likely reflects the amount of debris remaining on treated sites and the general habitat needs (i.e., overstory or ground cover) of these species. Further assessments are proposed after the 1999 sampling effort.
LITERATURE CITED
Aubry, K. B., L. L. C. Jones, and P. A. Hall. 1988. Use of woody debris by plethodontid salamanders in Douglas-fir in Washington, pages 32-37 in Szaro et al., eds. Management of amphibians, reptiles and small mammals in North America, proceedings of the symposium. USDA Forest Service General Technical Report RM-166.
McComb, W.C., C. L. Chambers, and M. Newton. 1993. Small mammal and amphibian communities and habitat associations in red alder stands, central Oregon Coast Range. Northwest Science 67: 181-208.
