Afforesting Agricultural Lands in the Mississippi Alluvial Valley (usa): Effects of Silvicultural Methods on Understory Plant Diversity

-To compare methods for bottomland hardwood reforestation on marginal farmlands in the Mississippi Alluvial Valley, four afforestation treatments (natural colonization, sown oak acorns, planted oak seedlings, cottonwood–oak interplant) were established in 1995 on former soybean cropland. Natural, sown, and planted-oak plots were not managed after establishment. Interplant plots received intensive management including two seasons of weed-control disking between planted cottonwoods, after which oaks were interplanted. Previous work found that forest canopy development was accelerated by interplanting; however, the best methods for establishing trees could have different effects on forest community diversity. Multi-year data on understory plant composition were analyzed to determine if less intensive methods promoted greater diversity. Ground-layer vegetation was sampled annually from 1996 to 1998, and again in 2006. Only total biomass was affected by afforestation technique, with the greatest declines in the interplant treatment. Changes in all species composition measures were a function of successional time. Although diversity did not vary substantially with reforestation method, lack of hydrologic restoration favored an understory flora more typical of moist old-fields than natural floodplain forests. INTRODUCTION Bottomland hardwood forests once covered 10 million ha in the Lower Mississippi River Alluvial Valley (LMAV). By the 1980s, large flood-control projects and land clearing for agriculture had reduced forest extent by roughly 75 percent (Haynes 2004). This historic forest loss has been addressed by various reforestation efforts in the past few decades (King and Keeland 1999, King and others 2006, Schoenholtz and others 2001). Reforestation methods (termed afforestation when converting from agriculture or other non-forest land use) have varied from completely passive to intensive, depending on land ownership and management objectives. Active, low-cost methods were favored as a way to establish desired species (mainly ‘hard-mast’ oaks, Quercus spp.) and overcome the dispersal limitations of passive colonization. However, the mixed results from active low-intensity techniques raised questions about potential tradeoffs between satisfying habitat/diversity objectives versus enhancing economic returns (such as timber yield) on private lands (Haynes 2004, Stanturf and others 2001, Twedt and Wilson 2002). A long-term experiment was established in 1995 to compare four afforestation methods ranging from passive to intensive, so that ecological and economic trade-offs could be assessed at operational scales (Gardiner and others 2008, Schweitzer and Stanturf 1999). The methods were natural tree colonization, establishing oak species by direct-seeding or by planting, and interplanting oak seedlings with a fast-growing early-succession tree species. A specific goal was to evaluate if the interplant method could accelerate forest development for both timber and habitat values. Results from this experiment and analogous studies indicated that interplanting favors rapid development of tree height, vertical structure and canopy closure, whereas less intensive methods may allow for greater tree diversity (Stanturf and others 2009, Twedt 2004, Twedt and Wilson 2002). Most of the afforestation research has focused on forest structure or overstory tree diversity. Understory plant composition is an aspect that has been evaluated only infrequently. Groundlayer biomass was assessed in the long-term experiment as a source of competition for planted tree seedlings (Stanturf and others 2009), but this layer is also a component of community diversity and contributes to habitat values. Intensive afforestation methods, while favorable for tree development, could have negative effects on ground-layer plant diversity. In this paper, we analyze data from the longterm experiment to determine if alternative methods led to differences in understory plant composition. STUDY SITE AND METHODS The long-term experimental site is located in Sharkey County, MS, on a tract that was in soybean cultivation until fall 1994. The soils are