Effects of mahogany (Swietenia macrophylla) logging on small mammal communities, habitat structure, and seed predation in the southeastern Amazon Basin

Vast areas of the Amazon have been subjected to low-intensity selective logging for high value timber species such as mahogany (Swietenia macrophylla). Although the ecological consequences of such logging are often assumed to be minimal, few studies have directly tested this assumption. In this paper we examine the effects of various intensities of mahogany logging on small mammal communities, habitat structure, and seed predation in the southeastern Amazon Basin. We sampled small mammal communities using live traps at two logged and two unlogged sites. Within each site, we established five 1 ha sampling grids that sampled the range of logging damage (in the case of logged sites) or mahogany densities (in the case of unlogged sites). Grids were sampled with 10 trap stations on the ground, 10 in the understory (at 1-3 m in height), and five in the canopy (at approximately 10 m in height) during three, 10-night trapping sessions over the course of the 10-month study. We measured a suite of forest structure and small mammal food resource (insect and fruit abundances) variables at each station and mahogany and peanut seed predation rates were measured at five locations per grid. We captured a total of 817 individuals of eight marsupial and 17 rodent species in 29,500 trap nights. Neither mahogany density across unlogged grids nor logging damage across logged grids were significantly correlated with variation in any of the measured habitat variables, seed predation rates, or measures of small mammal resource abundance. However, site-specific means for several variables (volume of downed wood, number of vines per tree, number of vines, number of herbaceous stems, and understory density) were higher in logged than unlogged sites, but significantly so in only two cases (number of vines per tree, volume of downed wood). When pooled, the logged grids had higher small mammal abundances and species richness than unlogged grids; however, predation of mahogany seeds and peanuts did not differ significantly between the two site types. When the two logged sites were analyzed separately, the effects of logging were found to be very different at the two sites. One of the logged sites showed few effects of logging and had similar forest structure and small mammal communities to the unlogged sites, whereas the other logged site demonstrated effects consistent with previous reports on logging, including increased small mammal abundances, increased seed predation, and decreases in seedling and sapling densities as a function of increasing levels of logging damage. This study provides evidence that the ecological consequences of mahogany logging are strongly dependent on local site features; however, the overall effects of the logging appeared to be small, suggesting that areas that have been subjected to only low-levels of logging can still make valuable contributions to conservation.