Deer Browsing, Forest Edge Effects, and Vegetation Dynamics Following Bottomland Forest Restoration Kenneth J. Ruzicka, 1,2 John W. Groninger, 1,3 and James J. Zaczek 1 Abstract Vegetation composition and forest stand development are frequently mediated by browsing herbivores. These rela- tionships have received little attention in a forest restora- tion context even though White-tailed deer (Odocoileus virginianus) is likely to influence these agriculture, forest, and restored ecosystem mosaic landscapes. Tree species composition, herbaceous vegetation, and deer browsing patterns were assessed 5 and 7 years following bottomland hardwood forest restoration on a 526-ha site in the Cache River watershed in southern Illinois, United States. Light- seeded tree species (Fraxinus pennsylvanica, Acer negun- do, Liquidambar styraciflua, and Platanus occidentalis) of volunteer origin dominated the woody vegetation com- ponent, with especially high stocking and density near ex- isting forest cover and potential seed sources. At more distant locations, presumably planted Quercus spp. were more likely to dominate and were the only tree species found in 15% of plots in year 7. Quercus stocking increased over the course of the study, constituting 7% of trees during year 7. Deer herbivory was associated with reduced stem height and disproportionately impacted seedlings of Quercus palustris and Celtis spp. Our results suggest that deer browsing influences forest stand compo- sition and density as a function of distance from the near- est forest edge. Herbaceous vegetation had little impact on early stand development. Continued spread of the exotic and invasive Lonicera japonica and potential mor- tality of F. pennsylvanica due to an anticipated Emerald ash borer (Agrilus planipennis) epidemic, combined with low stand density and delayed canopy closure, may result in persistent overstory gaps and compromise long-term restoration success. Key words: afforestation, forested wetland, insect infesta- tion, oak, silviculture, stand dynamics. Introduction Forest ecosystem regeneration in eastern North America is strongly influenced by biotic agents including browsing ungulates, and exotic plants and insects. In particular, the White-tailed deer (Odocoileus virginianus) has influenced stand composition and development through browsing and is presently a keystone species in some ecosystems (Russell et al. 2001; Cote et al. 2004). In contrast to natu- ral systems, studies of bottomland forest restoration have emphasized challenges to achieving adequate initial sur- vival arising from physical constraints, including flooding, changes in soil properties, and biotic stresses, particularly herbaceous competition and rodent herbivory (King & Keeland 1999). Also, some researchers have reported early successes in the use of tree shelters in a restoration context (Stange & Shea 1998; Sweeney et al. 2002; Ponder 2003; Jacobs et al. 2004). However, deer browsing effects on subsequent stand development have received little attention in an afforestation context, but see Opperman and Merenlender (2000). Within this region, a combination of passive and active restoration techniques is used to reestablish bottomland hardwood forest cover where row cropping has recently been abandoned (Stanturf et al. 2001; Twedt 2004; Groninger 2005). Typically, heavy-seeded oak (Quercus) species are planted uniformly across the restoration site. Light-seeded invader species, including Maple (Acer spp.), Green ash (Fraxinus pennsylvanica), and Sycamore (Platanus occidentalis), arrive from nearby forest stands and provide additional stocking (Lockhart et al. 2006). Within the first few years following initiation of restora- tion, these new stands are characterized by a gradient of decreasing tree density and stocking from established for- ested edge to restoration site interior (Allen 1997; Kruse & Groninger 2003; Battaglia et al. 2008). Unlike young natural stands characterized by very high stem density (Hosner & Minckler 1960; Nelson et al. 2008), herbaceous communities become dominant on these formerly row cropped sites where woody species do not initially regen- erate, influencing successional pathways and delaying forest canopy development (Bazzaz 1968, 1975; Twedt 2004). Significant deer browsing impacts on early stand development have been documented in regenerating natu- ral forests dominated by rapidly growing stump sprouts (Kittredge & Ashton 1995; Horsley et al. 2003). Within postagricultural landscapes undergoing forest restoration, abundant row crop food stocks are replaced by sparsely 1 Department of Forestry, Center for Ecology, Southern Illinois University, Carbondale, IL 62901-4411, U.S.A. 2 Present address: ENSR, Warrenville, IL 60555-3998, U.S.A. 3 Address correspondence to J. W. Groninger, email grnonige@siu.edu Ó 2009 Society for Ecological Restoration International doi: 10.1111/j.1526-100X.2008.00503.x 702 Restoration Ecology Vol. 18, No. 5, pp. 702–710 SEPTEMBER 2010