(1) determine species present in low-elevation conifer forests of Olympic National Park, (2) evaluate performance of closed-model population estimators of species abundance, (3) evaluate spatial and temporal variation in indices of abundance derived from constant-effort trapping grids, and (4) make recommendations for future monitoring of small mammal population trends.
A secondary objective of this study was to develop an inventory of species distribution patterns and presence/absence in various habitats throughout Olympic National Park. We were not able to sustain that sampling effort beyond the first year. We report preliminary results of that 1-year effort here, and make recommendations for conducting more extensive inventories throughout the park.
For the preliminary inventory effort, we established 4 index sampling sites in Douglas fir forests of the Elwha Valley, Sitka spruce forests of the Hoh Valley, and 4 hardwood forests in the Hoh and Elwha Valleys (2 in each valley). Index sites within each valley were randomly selected from Landsat maps of forested vegetation in Olympic National Park.
For logistical reasons, all sampling was conducted within 400 m of a road or trail and not more than 2 miles up a hiking trail from the end of a road.
SMALL MAMMAL TRAPPING METHODS Trapping Grids:--Each trapping grid encompassed 2.25 ha and consisted of a 4 x 25 array of trap stations spaced 15 m apart (100 trap-stations total). We placed three live-traps at each station, generally a small Sherman trap (Sherman Co., Tallahassee, Fl., 5.1 x 6.4 x 16.5 cm), a large Sherman trap (7.6 x 8.9 x 22.9 cm), and a Tomahawk trap (15.2 x 15.2 x 48.3 cm) (occasionally we used two large or two small Sherman traps within a site). We used Tomahawk traps only during the last two years of this study. We placed Sherman traps in the best microhabitats available within about 3 m of each station point (i.e., logs, debris, or other habitat structures). We alternated the placement of Tomahawk traps on the ground versus anchoring them about 1.5 m up the bole of a large tree or snag within 5-7 m of the sampling point center. We baited Sherman live traps with 1:1 mixture of oat groats and sunflower seeds and a piece of apple. We baited Tomahawk traps on the ground with alfalfa to target snowshoe hares, and those anchored on trees with pieces of apple and a mixture of strawberry jam, peanut butter, and oatmeal to target two species of squirrels (Northern flying squirrel and Douglas squirrel, all Latin names of mammals are provided in Appendix C). We covered each of the Tomahawk traps with a waxed cardboard milk carton to provide protection from sun and rain, and placed hydrophobic polyester batting within each trap for bedding. We set the traps on each grid each summer for two 4-night trapping sessions conducted within 11 nights (i.e., a total of 8 trapping occasions, with traps closed for 2-3 nights). We trapped between late July-early September each year to avoid periods of high immigration (Carey and Johnson 1995).
Index Sites: -- Each of the Index trapping sites used for inventory purposes consisted of a 2 x 8 array of trap stations spaced 15 m apart (16 stations total). We placed two Sherman traps at each station, generally one large and one small, baited and equipped the same as on the large grids. We did not set any Tomahawk traps at the index sites. We set and checked traps on each of the index sites for 7 consecutive nights.
Handling Procedures:--Field technicians baited and set the traps on the first day of each trapping session, and checked the traps daily each morning to record captures, record sprung but empty traps, adjust treadle sensitivity, and clean and re-bait traps as needed. We determined and recorded species (except shrews; see below), gender, age (juvenile versus adult), weight, and capture status (new capture versus recapture) of each animal caught. We did not distinguish among species of shrews captured in 1998 and 1999, during which years all shrews were recorded as Sorex spp. In 2000, we collected, labeled, and froze all captured shrews and submitted them to the Burke Memorial Museum at the University of Washington for species, sex, and age determination. Excepting shrews, we distinguished males from females by examining genitalia and juveniles from adults based on pelage characteristics, breeding condition, and body size. When it was difficult to determine juvenile from adult mice or voles, we distinguished on the basis of body weight (juveniles <15 g; Wilson and Carey 2000). We uniquely marked each captured animal prior to release (except shrews) using hair dye to mark hares, squirrels, and chipmunks, or by clipping a unique sequence of toes to mark mice and voles. Additionally we hole-punched a different ear of mice captured each year to distinguish mice first captured in a previous year.
ENVIRONMENTAL SAMPLING We described environmental characteristics of each trapping grid to permit the general comparison of sampling sites. At every second trapping station of each large trapping grid (e.g., 50 stations total), we described the prevailing forest association following Henderson et al. (1989), and recorded slope and aspect using a Silva Ranger compass and a clinometer. We calculated the mean annual precipitation at each sampling site by extrapolating data collected at various weather stations throughout Olympic National Park based on modeled relationships of precipitation to elevation, aspect, and forest type (Hoffman pers. comm.).
We recorded locations of the four corner points of each large trapping grid and the starting point of each index site using a Global Positioning System (Appendix A). Further, we described vegetation and habitat characteristics at each trapping station of both the large trapping grids and the index sites. We measured and recorded distance, diameter at breast height, and species of the nearest live tree in each of four sampling quadrants defined by the cardinal directions from each trap station (Mueller-Dombois and Ellenberg 1974). We ocularly estimated overstory canopy cover within a radius of approximately 30 m from the point center. We ocularly estimated vertical projections of cover of three understory vegetation classes, including shrubs, ferns, and forbs, within a 2.5 m radius of each trapping station. Lastly we enumerated the number of downed logs in each of five diameter size classes: 10-23 cm, 23-53 cm. 53-81 cm, 81-122 cm, >122 cm, within a 2.5-m radius of each trapping station.