Bottomland hardwood forests provide forage and other important resources for wintering and breeding waterfowl in the Southeastern United States. My objective was to conduct an initial investigation of possible influences of habitat features, flood events, and human disturbance on relative abundances of wintering ducks in Delta National Forest (DNF), Mississippi. I surveyed 65 wetlands in DNF 17 times from November 2012 to March 2013 and analyzed abundance data from wood ducks (Aix sponsa), mallards (Anas platyrhynchos), gadwalls (Anas strepera), and hooded mergansers (Lophodytes cucullatus). Waterfowl abundance varied by species but generally increased during major flood events and on wetlands having an approximately equal interspersion of scrub-shrub and open water consistent with the “hemi-marsh” concept. I recommend partial removal of scrub-shrub from selected wetlands to promote emergent plant communities and increased duck use and experimental evaluation of waterfowl responses to management of wetlands and human activities during winter in DNF.

The objective of this study was to examine some mechanisms of bottomland hardwood afforestation and their impacts on above- and belowground carbon sequestration. Six combinations of bottomland hardwood species and two levels each of fertilizer and herbicide were applied in a completely randomized design on two sites in the LMAV. Survival, ground line diameter and total height were monitored for two growing seasons. Soil carbon and nitrogen to a depth of one meter, herbaceous biomass, and tree biomass were sampled in the first and second years of establishment. Species mixes, fertilizer, and herbicide application significantly affected survival, growth, above- and belowground tree biomass carbon after two years of establishment. Survival was generally average, while growth for most species mixes was below expectation. Species mixes E (green ash/oak mix) and F (NRCS species mix) had the highest tree vegetation carbon both above- and belowground. Soil carbon and nitrogen were not significantly affected by any treatments.

Riparian forests are integral components of the landscape. The inherent biogeochemical processes that occur in such forests provide numerous benefits to wildlife and society. Maintaining good water quality is a major benefit from riparian forests and therefore, the maintenance, creation, or restoration of riparian forests is indispensable. This study was designed to broaden current knowledge of the complex, interrelated biogeochemical processes and determine indices for riparian forest restoration based on the various carbon pools/fluxes that may represent restoration success. This study was implemented on the Savannah River Site, an Environmental Research Park, where several riparian forests are recovering from thermal disturbances. The streams in these forests were subjected to thermal discharges that increased flows and resulted in removal of soil and a decline in the amount of woody vegetation. Two of these riparian forests are at different ages post-disturbance and represent different stages of recovery, which provides an exceptional opportunity to study successional processes in riparian forests and enhance restoration efforts. Linear transects perpendicular to the main stream channels were established in 2 recovering riparian forests of different ages (two areas in Pen Branch ~ 8 years post-disturbance; Fourmile Branch ~ 12 years post disturbance) and an undisturbed (thermally) more mature riparian forest (Meyer's Branch ~ 60 years). Along these transects quantitative data were obtained on above and belowground carbon pools and fluxes. Carbon pools exhibited a close correlation with riparian forest development. Biomass and carbon pools increased with increasing riparian forest stand age. The importance of the herbaceous carbon pool declined relative to the total above ground biomass, and the root carbon pool increased with forest age/succession. In general, net primary production (NPP) in young riparian forests (~8-10 years) rapidly approached and even exceeded NPP of more mature riparian forests. Once the herbaceous stage of succession was surpassed, the litterfall component of NPP plays a greater role riparian forests. As a woody overstory became established (after ~ 8-10 years), annual litterfall rates as a function of NPP were independent of forest age. Establishment of woody species occurred ~8 to 10 years after thermal disturbance and litterfall amount in young riparian forests rapidly became comparable to mature riparian forests. Lateral litter movement from the riparian forest toward the stream was less than the amount of litter (carbon pool) deposited from upstream into the riparian forest during a flood event. Overall lateral litter movement supplied less energy to the stream system than vertical inputs. A decline in riparian forest floor biomass was observed with increasing riparian forest development. However, a difference in foliar forest floor percent carbon lended itself to a minimal increase in the forest floor carbon pool with increasing riparian forest age. Woody debris in riparian forests comprised a relatively small carbon pool compared to tree and soil carbon pools. The species composition of litter appeared to be more of an overriding factor influencing decomposition rates than forest age. The influence of litter quality was evident in the decomposition rates of the different litter composites used in this study. In all 4 sites the litter composite from the mature riparian forest decomposed significantly more than the litter composites from the younger riparian forests. The fairly rapid decomposition of red maple (Acer rubrum L.), which was one of the main components in the mature riparian forest litter composite, influenced the greater decomposition rate. The litter composites from the younger riparian forests were similar and both included more decomposition resistant litter types, specifically waxmyrtle (Myrica cerifera L.) and alder (Alnus serrulata (Ait.) Willd.). Decomposition rates did not differ between the individual successional stages. Riparian forests are intimately associated with their hydroperiod. During flood events the riparian forest receives inputs of organic matter and sediment, and the amount of deposition may decrease along a distance gradient from the main stream channel. The differential amount of inputs could affect forest productivity. However, in these riparian forests, a distance gradient effect was not observed. Trends in herbaceous biomass were evident along a microtopographic moisture gradient. The ridge and swale microtopography prevalent in the younger riparian forests counteracted a distance from the stream channel gradient effect across the riparian forest. This study provided knowledge of how carbon pools and fluxes change with riparian forest recovery from disturbance as well as through different seral stages. Implementing the findings of this study will enhance restoration evaluation efforts to ensure that these areas continue to provide the numerous benefits gleamed from them. / Ph. D.

Reforestation of bottomland hardwood (BLH) forests has occurred within the Lower Mississippi Alluvial Valley (LMAV), USA, to support a wide range of ecosystem services, but especially wildlife habitat enhancement. As ecosystem restoration efforts proceed in BLH ecosystems, managers and policymakers are seeking criteria to evaluate wildlife habitat enhancement goals. Specialist wildlife that evolved within forest ecosystems can be sensitive to the composition, structure, and function of an ecosystem in relation to the system's natural or historical range of variation and thereby serve as indicators of habitat quality. The swamp rabbit (Sylvilagus aquaticus) is a specialist of BLH forests throughout the LMAV and therefore may be an appropriate indicator species for this ecosystem. In Chapter 1, I reviewed peer-reviewed literature to evaluate the utility of swamp rabbits as an indicator species according to three commonly-used criteria: habitat factors defining swamp rabbit relationships to BLH forests, the importance of swamp rabbit habitat to other wildlife, and the efficiency of swamp rabbit monitoring. I concluded that the swamp rabbit is a suitable indicator of wildlife habitat quality in BLH ecosystems in the LMAV because they evolved and remain endemic to the ecosystem, use habitat that integrates desirable characteristics that positively influence wildlife biodiversity, and are easy to monitor routinely.

Bottomland hardwood forests are highly productive ecosystems which perform many important ecological services. Unfortunately, many bottomland hardwood forests have been degraded or lost. Accurate land cover mapping is crucial for management decisions affecting these disappearing systems. SPOT-5 imagery from 2005 was combined with Light Detection and Ranging (LiDAR) data from 2006 and several ancillary datasets to map a portion of the bottomland hardwood system found in the Sulphur River Basin of Northeast Texas. Pixel-based classification techniques, rulebased classification techniques, and object-based classification techniques were used to distinguish nine land cover types in the area. The rule-based classification (84.41% overall accuracy) outperformed the other classification methods because it more effectively incorporated the LiDAR and ancillary datasets when needed. This output was compared to previous classifications from 1974, 1984, 1991, and 1997 to determine abundance trends in the area’s bottomland hardwood forests. The classifications from 1974-1991 were conducted using identical class definitions and input imagery (Landsat MSS 60m), and the direct comparison demonstrates an overall declining trend in bottomland hardwood abundance. The trend levels off in 1997 when medium resolution imagery was first utilized (Landsat TM 30m) and the 2005 classification also shows an increase in bottomland hardwood from 1997 to 2005, when SPOT-5 10m imagery was used. However, when the classifications are re-sampled to the same resolution (60m), the percent area of bottomland hardwood consistently decreases from 1974-2005. Additional investigation of object-oriented classification proved useful. A major shortcoming of object-based classification is limited justification regarding the selection of segmentation parameters. Often, segmentation parameters are arbitrarily defined using general guidelines or are determined through a large number of parameter combinations. This research justifies the selection of segmentation parameters through a process that utilizes landscape metrics and statistical techniques to determine ideal segmentation parameters. The classification resulting from these parameters outperforms the classification resulting from arbitrary parameters by approximately three to six percent in terms of overall accuracy, demonstrating that landscape metrics can be successfully linked to segmentation parameters in order to create image objects that more closely resemble real-world objects and result in a more accurate final classification.

Bottomland hardwood forests present a unique opportunity to examine the ecological response to hydrologically altered river and floodplain systems. These ecosystems, under natural conditions, are floodplain forests directly linked to the river. However, with major alterations to the river-floodplain system, bottomland hardwood forests can now provide insight into the loss of river connectivity and history of the flood pulse. This study reviewed the age distribution and growth of Quercus lyrata Walt. and the anatomical response of Quercus spp. to high magnitude floods in the 19th and 20th century in an old-growth Mississippi River floodplain forest in southeastern Missouri. Tree-ring samples collected from 43 trees at Big Oak Tree State Park contain physiological signatures (`flood rings') associated with growing season floods in the 20th century, and similar signatures are present in years corresponding to historical floods of the 19th century. The duration of 20th century Mississippi River growing season floods were examined to compare the occurrence of flood rings and the duration and intensity of floods on the Mississippi River at New Madrid, Missouri. Patterns in the occurrence of flood rings are developed as a response to alterations to the Mississippi River, river floodplain, and park hydrologic system.

The Lower Mississippi Alluvial Valley (LMAV) was once the largest floodplain in the United States. Flooding regimes in the LMAV have strong impacts on the soil biogeochemical processes, nutrient cycling, forest species distribution, agricultural production, and wildlife habitat. This study characterized the LMAV flooding regimes using the 8-day-composite Moderate Resolution Imaging Spectroradiometer (MODIS) surface reflectance product from 2001 - 2018. The results showed significant spatial variations in the annual flooding duration in the LMAV, and the flooding area showed remarkable seasonal variations in the northern and central LMAV with the peak flooding area in winter and early spring. Digital Elevation Model (DEM) and slope were identified as the two major factors in determining the spatial pattern of flooding regimes. The LMAV flooding regimes dataset provides a scientific basis for the governments to design forest, agriculture, and wildlife management policies to enhance ecosystem services. Landowners also can use this information to make decisions for cropland retirement and tree plantation.

While there is a clear link between riparian forests and freshwater organisms, floodplain forests are seldom investigated due to difficulties in sampling structurally complex and periodically inundated habitat. This lack of research has led to large knowledge gaps that hinder our understanding of the conservation value of these unique, complex ecosystems for inland fisheries. Therefore, I aimed to determine how bottomland hardwood forests influence fish taxonomic, functional diversity and food web structure. I hypothesized that fish taxonomic and functional diversity are driven by forest complexity and the aquatic food web structure is driven by terrestrial carbon sources, specifically forest vegetation. Results indicated a higher taxonomic diversity and functional richness in the floodplain forest and that this forest type provides thermal refugia for fish assemblages. Contrary to my prediction, phytomicrobenthos were a primary carbon production source driving some or all of the aquatic food web in a complex floodplain–river system.

Bottomland hardwood forests (BHF) are a disappearing habitat of importance to numerous migratory and non-migratory birds of conservation concern. Thus, understanding variables of bottomland hardwood forests that affect avian assemblage patterns are of great interest. I examined factors influencing avian assemblage and guild patterns in BHF of Noxubee National Wildlife Refuge in east-central Mississippi by conducting winter and spring fixed-radius avian point counts and vegetation surveys from 2002 ? 2004. The goals of this project were to 1) determine effects of greentree reservoir (GTR) management on breeding and wintering non-game bird community structure, 2) test hypotheses regarding plant structural complexity and avian assemblage patterns, and 3) examine the effects of landscapes on local bird assemblages within BHF. Greentree reservoirs have subtle effects on avian communities. Sites within GTRs, sites in BHF adjacent to GTRs, and random, unimpounded BHF sites differed significantly in breeding bird diversity, richness, and evenness. However, the patterns exhibited were a decreasing continuum of these variables with sites peripheral to GTRs demonstrating intermediate values. No differences were found among abundance, conservation status, or habitat specialization of breeding birds. Guild parameters did not differ except for a decreasing continuum in species richness among unimpounded BHF, BHF to GTRs, and GTR sites. The canopy nesting and ground gleaning foraging guilds were the only guilds found to differ among sites. Patterns could indicate that given the inherent structural variation of BHF, GTRs are within the expected range of variation, or GTR management could be indirectly affecting the surrounding landscape. GTR management did not affect community parameters for wintering birds, and only the bark gleaning foraging guild varied among sites (higher in GTRs). These results reflect the ubiquitous nature of wintering birds in east-central Mississippi forests. There was direct support for the structural vertical heterogeneity hypothesis; however, regression analyses of principal component scores derived from sixty structural descriptor metrics suggest that avian communities and guilds are more directly influenced by overall local plant structural complexity, as predicted by the spatial heterogeneity hypothesis. Avian community and guild parameters did not differ among the three landscape contexts defined by 1,000 and 1,500m buffers.

acase@tulane.edu