Effects of Strip-cropping on Small Mammal Population Dynamics in Soybean Agroecosystems

The present study examined the effects of strip-cropping and harvesting practices on small mammal population dynamics in soybean agroecosystems. Small mammals were live-trapped in four treatments (three replicates each): soybean monoculture, soybean-clover, soybean-buckwheat, and soybean-corn. Peromyscus maniculatus was found in all four treatment types, whereas Mus musculus resided mainly in the soybean-corn treatment. Peromyscus population densities were significantly greater in the soybean monoculture during the week preceding harvest than in the soybean-clover strip-cropped treatment. Peromyscus population densities immediately increased following harvesting practices, then declined. Short-term changes in density were attributed to seed accessibility; long-term changes appeared to be in response to reduced crop cover resulting in increased predation. Populations of Mus were unaffected by harvest practices. Interestingly, more Peromyscus dispersed from strip-cropped treatments than from the monoculture (control) treatment. Female deer mice were found to have larger mean home ranges in the corn strip-cropped treatment than in the monoculture or buckwheat strip-cropped treatment suggesting an impact of spatial resource patterning on small mammal population dynamics. OHIO J. SCI. 94 (4): 94-98, 1994 INTRODUCTION Recently much emphasis has been placed on an approach to agriculture more sustainable than that used in the past (e.g., NRC 1989, Barrett et al. 1990, Parr et al. 1990). Strip-cropping is a type of inter-cropping that represents an alternative approach to conventional agriculture. Although several studies have focused on the effects of strip-cropping on insect population dynamics (Kemp and Barrett 1989, Pavuk and Barrett 1993) and insect movement behavior (Risch 1981, Bohlen and Barrett 1990), no studies have been conducted to investigate the effects of strip-cropping on the population dynamics of small mammals. The present study was designed to investigate the effects of strip-cropping, including harvesting practices, on small mammal population dynamics and dispersal behavior within replicated monoculture (control) and strip-cropped soybean agroecosystems in southwestern Ohio. MATERIALS AND METHODS Study Site The study was conducted at the Miami University Ecology Research Center located near Oxford, OH. The study area consisted of twelve 0.4-ha (75 m x 60 m each) agroecosystem plots (see Kemp and Barrett 1989 for aerial view of study site). Soybeans (Glycine max var. Williams 82) were planted 2-4 May 1992. Three plots were planted as soybean monocultures to serve as controls and the remaining nine plots were strip-cropped: three plots of soybean-corn, three plots of soybean-clover, and three of soybean-buckwheat. Corn (Zea mayes Hybrid County Mark 707) was planted on 4 May 1992, red clover (Trifoliumpratense) was planted on 11 May 1991, and 'Manuscript received 21 June 1993 and in revised form 18 April 1994 (#93-19). buckwheat (Fagopyrum esculentum) was planted on 11 June 1992. Each strip-cropped treatment contained 14 alternating strips (six rows per strip) each 6 m in width. The 12 plots were aligned in two rows of six (Fig. 1). Ten meters separated each experimental plot, whereas 15 m separated all plots from surrounding habitats. Census Procedures Two hundred eighty-eight Sherman live traps (24 traps/plot) were uniformly distributed among treatments. Twenty-four traps were placed in the eight center strips in each strip-cropped plot. An equal number of traps were positioned at equivalent sites in each soybean monoculture plot. The distance between each trap within each strip was 15.7 m. Trapping was conducted at least twice weekly from 12 September-20 November 1992 and consisted of 1,440 trap nights per treatment equally distributed before and after harvesting. Trapping was initiated at the beginning of seed (crop) maturation. Each plot was disturbed by harvesting practices from 12-25 October. Traps, baited with peanut butter and supplied with cotton for nesting material, were set between 1900-2000 hours and checked the following morning between 07000800 hours. Traps were locked open between trapping dates, thus effecting a pre-baiting trapping regime (Smith et al. 1975). Captured individuals were identified to species, marked by toe clipping, weighed to the nearest 1.0 g, sexed, and the reproductive condition recorded. Reproductive condition for males was determined by position of testes (scrotal/non-scrotal) and for females by vaginal perforation (perforate/non-perforate), pregnant, and/or lactating. Population density was estimated by the calenclar-ofcatches method (Petrusewicz and Andrzejewski 1962). Home range was determined by the inclusive boundary strip method (Stickel 1954). Dispersal behavior was OHIO JOURNAL OF SCIENCE C. K. WILLIAMS ET AL. 95