I monitored 60 radio-collared and tagged pen-raised white-tailed deer (Odocoileus virginianus) to ascertain mortality rates on three areas in Louisiana from April 2001 - April 2002. Two of 60 were not used in the analysis because mortality occurred from complications with darting and an affixed radio-collar. Study deer were exposed to legal hunting mortality only if they emigrated from the three study areas. Twenty deer were released to the wild on the Louisiana State University Deer Study Area (LSUDSA) (4,810 ha), where 9 of 20 (45%) suffered mortality. Harvest accounted for 56%, vehicles 22%, and other 22% of the mortality on the LSUDSA. Two groups of 20 were released into 2 high fence enclosures, (275 ha and 608 ha), where 4 of 20 (20%) and 1 of 18 (6%) suffered mortality, respectively. Other mortality accounted for 100% of the losses in the high fence enclosures. Mortality rates were significantly lower in penned deer than in the released deer (Chi-Square 8.33 2df, P<0.05). These results suggest that high fence enclosures reduce emigration and hence deaths caused by hunting and vehicles and may reduce overall mortality. These results also suggest that a higher harvestable surplus may result from high fencing.

Benthic intertidal bivalves play an essential role in estuarine ecosystems by contributing to habitat provision, water filtration, and promoting productivity. As such, ecosystem level changes that impact population distributions and persistence of local bivalve populations may have large ecosystem level consequences, making it important to better understand the population distribution and ecology of native bivalve populations. Gametogenesis, recruitment, growth, mortality, size structure and density of the ribbed mussel, Geukensia demissa, were examined across a salinity gradient in southeastern Louisiana. In summer 2012, 100-m transects were placed at interior and edge marsh plots to collect data on mussel density, and vegetation data at duplicate sites in upper (salinity ~4), central (salinity ~8) and lower (salinity ~15) Barataria Bay, LA. Caged growth and recruitment plots were established adjacent to each transect, in April, and growth, mortality and recruitment of individually marked mussels within plots were recorded in November 2012. Mussels were randomly sampled monthly from low (~ 5) and high (~25) salinity marsh sites, and histologically processed to determine the seasonal progression of gametogenesis. Mussel densities were greatest within mesohaline marsh (66.6 + 18 m-2), J. roemerianus vegetation (191.2 + 42.7 m-2) and plots experiencing 20-60% annual flooding rates (46.7 + 13.8 m-2). Mussel recruitment, growth, size and survival were significantly higher at mid and high salinity marsh edge sites as compared to all interior and low salinity sites. Peak gametogenic ripeness occurred between April and September, positively correlated with temperature, and coincidental with seasonal shifts in salinity. The observed patterns of density, growth and mortality in Barataria Bay may reflect detrital food resource availability due to local site flooding rates, vegetation community distribution along the salinity gradient, and reduced predation at higher salinity edge sites.

The Atchafalaya River Basin (ARB), Louisiana supports a diverse centrarchid assemblage, characterized by abundant populations of largemouth bass Micropterus salmoides, black crappie Pomoxis nigromaculatus, warmouth Lepomis gulosus, bluegill Lepomis macrochirus, redspotted sunfish Lepomis miniatus, longear sunfish Lepomis megalotis and redear sunfish Lepomis microlophus. This dynamic floodplain ecosystem is comprised of a mosaic of macrohabitats, including natural bayous, shallow lakes, dead-end and open pipeline canals. I conducted an extensive feeding ecology study to determine the influence of these macrohabitats on foraging activity through stomach content and stable isotope analyses. I collected 2,036 centrarchids with electrofishing efforts in the summers of 2011 and 2012. Canonical correspondence analysis of prey items revealed evidence of resource partitioning within the assemblage, as suggested by limited dietary overlap and significant separation of diets among the seven species (F = 8.6516, P = 0.01). Multivariate analysis of stable nitrogen (δ15N) isotopes revealed evidence of trophic specialization (F = 24.29, P = 0.0001) within the assemblage, despite moderate overlap in foraging habitat suggested by carbon (δ13C) signatures (F = 2.76, P = 0.0147). Results of the generalized linear mixed model (F = 1.06, P = 0.3907) and multivariate analyses of stable isotopes (P = 0.86) indicated that macrohabitat did not have an overall effect on diets or isotope values. Traditional dietary indices, such as index of relative importance values, indicated specialization in prey preference for each species, suggesting a limited degree of competition between centrarchids. In summary, despite physicochemical differences among macrohabitats, these results suggest that prey selection was species specific and independent of macrohabitat type.Condition (relative weight) was also investigated to discern the potential influence of macrohabitat and physicochemical properties. Associations were species specific, indicating disparate manners of habitat utilization that may influence niche separation, although parameters such as dissolved oxygen differential were positively associated with multiple species. Although water quality and physical habitat characteristics differed among macrohabitats, trends in centrarchid condition suggest that while habitats within the ecosystem differ regarding their overall suitability as sunfish foraging habitat, all of the centrarchid species are generally able to flourish within the permanent waterbodies of the ARB floodplain.

Nest success of upland nesting ducks is the primary driver of duck population growth in the Prairie Pothole Region. Nest success is greatly influenced by nest predation and the amount of available nesting cover on the landscape. The decline in acres enrolled in the Conservation Reserve Program (CRP) in this region has negatively impacted the amount of available nesting cover, making nesting cover sparse and confined to small patches where predation rates are potentially elevated. I evaluated the efficacy of seasonal predator reduction on increasing nest success on low-grassland density (>10% grassland cover), 93 km2 landscapes in two different habitat types used by nesting birds, large fields and roadside ditches. Ditches were sampled because they are a major cover source in low-grassland density landscapes. I monitored 1,899 nests during the 2010-2012 breeding seasons. Predator reduction had a significant influence in large fields as nest success was 1.6 times greater in large fields on trapped sites (44% nest success) than on control sites (27% nest success). Predator reduction, however, did not significantly increase nest success in roadside ditches (13% nest success on trapped sites, 12% nest success on control sites). A large majority of monitored nests were located in large fields resulting in the overall effect of predator reduction significantly increasing nest success by 13%. These results indicate that predator reduction is an effective intensive management technique in low-grassland density landscapes and can be used as a management tool in a post-CRP era. Future research should evaluate different trapping techniques in efforts to increase nest success in the roadside ditches.

This project examined the ecologically and economically valuable eastern oyster (Crassotrea virginica) in Louisiana with two field experiments. Little information has been gathered on the performance of local Louisiana oyster stocks and no study has focused on a comparison of stocks among variable estuarine conditions. Additionally, the use of alternative grow-out methods and intensive cage aquaculture has never been evaluated in Louisiana. For the first study, a dermo-resistant stock of oysters (LSU-OYS: OBOY) was compared to three wild oyster stocks along a salinity gradient. The objectives of this study were to determine the optimal oyster stock(s) and the ideal grow-out condition(s) for intensive oyster aquaculture production. The second experiment compared the efficiencies of three commercially used, off-bottom culture systems. The objective of this study was to suggest which off-bottom grow-out method(s) is most suitable for use in Louisiana estuaries. The results of the stock comparison suggested that the selected dermo-resistant stock had greater mortality than two of the three wild stocks in all the environmental conditions tested. Specifically, wild stocks taken from low salinity areas had greater performance in test areas with low salinities while wild stocks collected from high salinity areas had greater performance at high salinity sites. The results of the grow-out method comparison revealed that an adjustable long line system (ALS) was the most suitable culture system in both high and low salinity conditions, specifically due to overall higher survival, improved growth in shell height, and reduced effort in labor and handling time. For the first time, the performance of four oyster stocks and three intensive oyster culture methods were quantified, suggesting superior stocks, grow-out conditions, and culture systems for augmenting wild production and increasing total production in the Louisiana oyster industry.

Agriculture and moist-soil management are important management techniques used on wildlife refuges to provide adequate energy for migrant and wintering waterbirds. However, in arid systems, the presence and accumulation of soluble salts throughout the soil profile can limit total biomass production of wetland plants and agronomic crops and thus jeopardize meeting waterbird energy needs. It is unknown how moist-soil management and traditional agriculture practices influence the accumulation and distribution of soluble salts of soil profiles. In this study of an arid wetland ecosystem, I determine: 1) the effect of long-term, distinct surface hydrologic regimes associated with moist-soil management and agricultural production on salt accumulation; and 2) the specific effects of rototillage and irrigation frequency on salinity concentrations and plant biomass in moist-soil impoundments. My study was conducted at Bosque del Apache National Wildlife Refuge near San Antonio, New Mexico. In May 2012, prior to the growing season, I collected one meter deep soil cores from both moist-soil impoundments and agricultural fields; cores were analyzed in 10 cm segments for soluble salt concentrations. I implemented a split-plot experiment to evaluate salinity concentrations in moist-soil impoundments between rototilled and no-till soils under a 9 and 14 day irrigation frequency. Soil salinity was measured in May and August of 2011 and 2012 and plant biomass in August. My findings suggest that agricultural fields contain significantly higher concentrations of soluble salts in deeper portions of the profile. This may be attributed to the lack of leaching afforded by summer agricultural irrigations as little connectivity to the groundwater and groundwater salinity was detected during groundwater monitoring. In contrast, periodic flooding in winter and summer flood irrigations in moist-soil impoundments may serve as leaching events and created a more dynamic groundwater hydrograph. This seasonal wetland hydroperiod may facilitate lower soil profile salinities but further research is needed to evaluate its successful use in agriculture fields to lower soil salinities. Few differences in soil salinity were detected between tillage and irrigation treatments within moist-soil impoundments. However plant above ground biomass of annual wetland grasses was greater in rototilled soils. This is most likely attributed to the effects of physical disturbance that stimulates germination rather than differences in soil salinity, however greater aboveground biomass does not necessarily equate to higher seed or tuber production.

The armoring of river banks with riprap can have detrimental effects on lotic ecosystems due to the subsequent alteration of hydrologic regimes; however, evidence suggests that riprap can also increase aquatic diversity in degraded systems. The goal of my study was to determine what impacts riprapped banks have on fish and macroinvertebrate assemblages in the Pearl River, which has a history of anthropogenic degradation. I collected fishes with an electrofishing boat from armored and natural banks at five regions during fall 2011, winter 2012, and summer 2012. I also collected macroinvertebrates with introduced substrates in fall 2011. Richness was analyzed with rarefaction curves, whereas seasonal abundance and evenness were analyzed with ANOVA, and differences in assemblage structure were assessed with PERMANOVA. Fish and macroinvertebrate richness never varied between bank types. However, fish evenness and abundance were higher at riprapped banks during the summer, and fish assemblage composition varied during base flows in the summer and fall. Macroinvertebrate and assemblage structure also varied between bank types in the fall. My results imply that, at base flows, riprapped banks in the Pearl River support unique aquatic assemblages, possibly due to differences in heterogeneous habitat availability. On August 9, 2011 waste material from a paper mill in Bogalusa, LA was accidentally discharged into the Pearl River, causing anoxic conditions that resulted in complete fish extirpation downstream of the spills source. ANOVA indicated that, by October 2011, fish species richness and evenness at sites ~10 km downstream of the spill did not differ from sites in undisturbed areas, although richness and evenness at sites ~40 km downstream were still significantly lower at this time. However, by January, richness and evenness at all disturbed and undisturbed sites were similar. PERMANOVA indicated that, despite similarities in richness and evenness, fish assemblage composition at sites ~10 km downstream of the spill remained significantly different from undisturbed areas until January 2012, while sites ~40 km downstream of the spill were still significantly different by the studys end in July 2012. These results suggest a gradual recovery, with colonization rates related to the proximity of source populations.

Gopher tortoise (Gopherus polyphemus) populations have declined with longleaf pine (Pinus palustris) ecosystems across their historic range. The influence of gopher tortoise burrows on co-inhabiting plants and animals necessitates an understanding of how landscape features and management practices influence gopher tortoise presence, absence, and abandonment. In this study, naïve gopher tortoise burrow encounter rates from a line transect distance sampling (LTDS) pilot study were used for two methods of modeling gopher tortoise habitat. In Chapter 1, naïve encounter rates were tested for a linear correlation to a HSI model created from three ranked geographic information system (GIS) landscape variables. Initial results showed a positive linear correlation (all P < 0.0001, 0.55 < r< to 0.70) but a second test using only transects with observed burrows resulted in the loss of all correlations (all P > 0.05, r values ranged from 0.17 to 0.42). However, logistic regression analysis revealed the HSI model was able to predict burrow presence along transects (P = 0.0003). In Chapter 2, microhabitat variables and five GIS landscape variables were reduced into seven correlated principal components (PCs). According to a generalized linear (logit) model three PCs were significantly associated to active and abandoned borrows. Active burrows were positively associated to: 1) sandhill habitats, longleaf pine canopy, Lakeland soils, high elevations, xeric oak midstory, and wiregrass (Aristida beyrichiana) presence (overall P = 0.003; active P = 0.008); and 2) grassland habitats, little to no canopy, and increased herbaceous ground cover (overall P = 0.0042; active P = 0.0052). Active burrows were negatively associated to mesic flatwoods, Scranton soils, mixed pine canopy, high basal areas, and increased percent tree canopy (overall P = 0.003; Active P = 0.008). Abandoned burrows were positively associated to xeric hammocks, xeric hardwood canopy, mesic hardwoods midstory, increased canopy cover, increased litter ground cover, and increased mean years between burns (overall P = 0.0448; abandoned P = 0.0137). The relationship between fire suppression and burrow abandonment is widely accepted but poorly documented, and the poor resolution of this fire layer accentuates the importance of this detected relationship.

The goal of this study was to isolate biological, natural, and silvicultural factors that encourage the production of epicormic branches on bottomland red oak saw-timber trees and to use this information to help landowners and forest managers reduce epicormic branches in stands dominated by bottomland red oaks. Quercus nigra, Q. pagoda, Q. phellos, and Q. texana, the most common commercial bottomland red oaks in Louisiana, were evaluated in this study. Both qualitative and quantitative variables were assessed in this study. The following attributes were collected from each sample tree and primary competitor: site class, basal area, crown class DBH, presence or absence of epicormic branches, soil series, flooding regime, total tree height, height to base of live crown, live crown ratio, localized disturbance, and distance to nearest competitor. A total of 768 trees were evaluated, of which 384 trees displayed epicormic branching and 384 were without epicormic branches. Sample trees of all four species were equally distributed across four crown classes (dominant, co-dominant, intermediate, and suppressed). Equal numbers of all four oak species were collected to maintain homogeneity within the dataset. Logistic and Poisson regressions were used to analyze the data. An á level of 0.1 was considered effective for significance in this exploratory research. Logistic regression analysis yielded three variables significant in relation to the probability of epicormic branch occurrence. Variables significantly affecting the probability of epicormic branch production were total height of the sample tree, distance from the sample tree to the primary competitor, and crown class of the nearest primary competitor. The Poisson regression analysis was used to evaluate the number of epicormic branches a tree might produce. In this analysis dbh of the sample tree, crown class of the primary competitor, disturbance, flooding regime, and live crown ratio of sample trees were significant in relation to the number of epicormic branches produced. Tree, site, and environmental characteristics isolated as being important to epicormic branch production produced a means for evaluating the potential development of epicormic branches. These factors in concert provide the forester or land manager with a simple means of enhancing bottomland red oak value. The driving mechanism for sudden appearance of epicormic branches seems to be related to reduced tree vigor as affected by competition (or relative vigor based on relative height to competitors) and site stress factors. Taller trees with less competition from similar sized trees have a reduced probability of epicormic branches. Or stated in another way, trees with close competitors or nearly equal height are more likely to produce epicormic branches. Dominant trees have a reduced likelihood of developing epicormic branches if other stress factors are not at play. Intermediate and suppressed crown class trees should be removed during early thinnings of red oaks stands to avoid or reduce epicormic branch production. In concert, factors reducing competition and decreasing tree vigor tend to reduce the probability of epicormic branch production on bottomland red oaks.

With an estimated population loss of at least 5% annually and a 100-year history of decline, Rusty Blackbirds (Euphagus carolinus) are one of the fastest declining bird species in North America. Determining cause of decline is important, both for conserving the species and for identifying threats to the wooded wetland ecosystems they use on their boreal breeding-grounds and their wintering-grounds in the southeastern United States. One hypothesis is that loss of wintering ground habitat, possibly in conjunction with competition or disease, is causing Rusty Blackbird decline. To determine contribution of wintering ground conditions, it is important to understand the behaviors and habitat requirements that make these birds prone to decline. My research objectives were to develop survey strategies for detecting and quantifying Rusty Blackbird presence, to examine inter- and intraspecific associations for potential competition, and to determine habitat requirements at spatial scales appropriate to foraging movements. The present study uses data from 550 survey occasions and 163 independently collected birder observations to examine survey methods and seasonal trends over two winters. Occupancy modeling was used to investigate flocking behavior and habitat associations (at 25 m and 100 m scales) at 74 unique sites (naïve occupancy = 0.82, average occupancy by survey round = 0.51). Results indicate that Rusty Blackbirds may be less dependent on forested habitat than previously thought, but show strong a strong relationship to availability of wet ground. Range-wide alterations in hydrological processes, due to drainage or flood control, could lead to decreased quality or availability of shallow-water habitat and aquatic food resources. Rusty Blackbirds frequently flock with other blackbird species, and show similarities in use of open habitats. Competition cannot be ruled out and could worsen with use of degraded habitat. Overall results suggest that wintering behavior and habitat changes are likely contributors to long-term and continuing Rusty Blackbird decline.