To examine trends in small mammal abundance, snap-trapping was conducted annually 1993- 1997 on 25 permanent transects during early March. Transects were located in major habitat types (n=7) and were 500 m long with stations located 20 m apart along the transect. One rat and 1 mouse Victor snap-trap (Woodstream Corp., Lititz, Pennsylvania, USA) was placed at each station. Traps were prebaited with a mixture of oatmeal and peanut butter for 3 days before being set for 4 days. Captured animals were identified to species with date, sex, weight, trap number, trap type (rat or mouse) and transect number recorded. Small mammals were classified as either: rats [cotton rats, marsh rice rat (Oryzomys palustris)], mice [white-footed, cotton mouse (Peromyscus spp.), golden mouse (Ochrotomys nuttalli), eastern harvest mouse (Reithrodontomys humulis), house mouse (Mus musculus)] or shrews [(least shrew (Cryptotis parva) and southern short-tailed shrew (Blarina brevicauda)].
Live-trapping of small mammals was initiated in February 1997 to gather habitat-specific small mammal density information on TWMA and GP. Although snap-trapping provided trend information on small mammal communities, we decided to conduct intensive live-trapping to estimate small mammal density within each macrohabitat type. Small mammals were trapped on 100-m × 100-m grid on TWMA (n = 28) and GP (n = 10) using Sherman live-traps. Each grid contained 25 traps, only 1 trap was placed at each station within the grid. Winter sampling was conducted during February-March, while summer sampling occurred during June-August in major habitat types. Multiple grids were sampled for each habitat type to address concerns over site-specific abundance of small mammals. All stand types were sampled with a minimum of 2 grids. Captured small mammals were uniquely toe-clipped, sexed, aged and released. Trapping sessions on each grid were conducted for 4 consecutive nights.