clipping and released at the point of capture. Species, physical condition, 

 number, sex, breeding condition, weight, age and location of capture were re- 

 corded for all captured animals. (See appendix G for an example of data sheets 

 used to record these data.) 



Snap-Trapping 



Twenty-four snap trap lines were trapped in 24 vegetation associations 

 (maps 1 and 3), Each line consisted of 25 stations having two museum special 

 traps each at 15 m (49 ft) intervals. Lines were located in stands closely 

 representative of the various community types with the aid of the DNRC plant 

 ecologist. Whenever a habitat type was not sufficiently expansive to accomo- 

 date a straight trap line, the line was split into two parallel lines separated 

 by at least 60 m (196 ft). Traps were baited with peanut butter and rolled oats. 

 Each snap-trap transect was sampled for two separate three-night periods. The 

 second sampling occurred from one to four weeks after the first. All of these 

 areas were not sampled concurrently but rather in an overlapping series begin- 

 ning in June and lasting through October. Standard body measurements (Hall 

 1962) and reproductive data were collected from snap-trapped specimens, and 

 representative study skins were prepared and deposited with DNRC. 



Pitfall Trapping 



Seven grids, each consisting of eight pitfall trapping stations placed in 

 two rows of four were placed in representative vegetation types (see map 3). A 

 metal can 17.5 cm (6.89 inches) high and 15.5 cm (6.10 inches) in diameter was 

 set flush with the ground at each station and half filled with water. Traps were 

 checked periodically through October 1977, and specimens captured identified to 

 species. 



Analysis 



Population Indices . Population indices for a-11 snap-trap lines are repre- 

 sented by the number of individuals captured per hundred trap nights (TN), and 

 for live-trap grids by the minimum number of individuals captured per four-day 

 trapping period per plot divided by plot area. As recommended by Krebs (1966), 

 no attempt was made to compute absolute population density estimates from cap- 

 ture-recapture data because of the non-random nature of the sampling technique. 



Biomass Estimates . Monthly biomass estimates were computed for live-trap 

 grids by dividing the total weight of individuals captured by the grid area, 

 which yielded a figure in grams per hectare. For comparisons among all stations, 

 the statistic grams per one hundred trap nights (gm/100 TN) was calculated. 



Niche Breadth and Niche Overlap . Habitat niche breadth was computed by 

 Mac Arthur's Formula (Pianka 1973): 



38 



