Colonial nesting seabirds are threatened by habitat loss and degradation, human disturbance, predation, and climate change. Several species of conservation concern concentrate high percentages of their total U.S. populations in Louisiana breeding colonies. We studied seabirds, including Royal Terns, Sandwich Terns, and Black Skimmers, nesting on Isles Dernieres barrier islands along the Gulf coast of Louisiana. Two of the four islands in this chain host extensive seabird colonies and two do not. We used an experimental approach to test the hypothesis that large terns and skimmers are prevented from nesting on Trinity Island, the largest of the Isles Dernieres, by lack of social stimuli. Decoys and call broadcast attracted Royal Terns to visit experimental sites, but they did not nest. Sandwich Tern and Black Skimmer visits to the sites were not significantly affected by the social stimuli; however, isolated nesting attempts imply interest. Lack of colony establishment in response to the experiment indicates that social factors alone are not responsible for the lack of nesting by these species on Trinity Island. Scent station transects revealed the presence of raccoons, rats, and coyotes on two non-colony islands, and no mammalian predators on two colony islands, suggesting that seabirds avoid predator-infested areas. Least Terns were an exception, nesting on islands with mammalian predators. In 2008 and 2009, we monitored 53 and 80 Least Tern nests on Trinity Island and modeled nest success using logistic exposure. A subset of nests was protected by fences in each year (n= 3 in 2008, n= 19 in 2009). For unprotected nests, model-estimated nest success was 20% in 2008 and 53% in 2009. Fenced nest success was 83% and 49% in 2008 and 2009, respectively. We believe the increase in nest success between years reflects effects of Hurricanes Gustav and Ike on predator populations on Trinity Island. Rats and raccoons declined in surveys and anecdotal field observations. The impact of mammalian predators on this Least Tern population supports the hypothesis that predation limits seabird colonization of Trinity Island.

In Louisiana, the extensive loss and fragmentation of coastal marshes has prompted inquires into the impacts these processes may have on estuarine-dependant nekton. To date, research on nekton response to marsh loss and fragmentation has been limited to landscape-level studies which focus on the relationship between nekton productivity and the availability of marsh edge. These studies have relied on the assumption that marsh edges provide the same level of support to nekton regardless of the degree of surrounding marsh loss or fragmentation. This study tested this assumption by investigating the impacts of marsh loss and fragmentation on marsh-edge characteristics and their associated nekton assemblages. The effects of marsh loss at the 1 km2 scale were examined by stratifying three brackish marsh management units located in the Chenier Plain of western Louisiana into three 1 km2 treatment squares, each representing one of three levels of marsh loss: Low (10%-35% water), Medium (40%-65% water), and High (70-95% water). Within each treatment square, nekton assemblage (density, diversity, body condition) and marsh-edge characteristics (water quality, submerged aquatic vegetation biomass, sediment organic matter, and emergent stem density) were sampled concurrently at six randomly established sampling points during the spring and fall of 2008. Variables representing marsh loss and fragmentation were also quantified within 1 ha squares centered on each sampling point. Relationships between nekton assemblage characteristics and environmental variables were explored with a canonical correspondence analysis. Data analysis revealed differences in sediment organic matter, ha scale percent marsh, and ha scale fragmentation in the spring sampling season. The passage of Hurricane Ike between sampling seasons may explain why only ha scale percent marsh differed between treatments in the fall. Despite differences in these environmental variables between treatments, nekton assemblage characteristics were not found to differ between treatments. This may be partially explained by the lack of strong relationships between nekton assemblage characteristics and environmental variables as indicated by the canonical correspondence analysis. The results of this study do not indicate that nekton support provided by marsh edges is influenced by the degree of marsh loss at the 1 km2 scale.

Since 1982, the Louisiana Department of Wildlife and Fisheries (LDWF) has stocked Florida-strain largemouth bass (FLMB; Micropterus salmoides floridanus) to incorporate Florida alleles into native populations (NLMB; M. s. salmoides) and enhance recreational fishing opportunities. I collected habitat data and largemouth bass samples from 12 LDWF stocked reservoirs and disconnected oxbow lakes to examine the relationships between reservoir characteristics, genetic identity, relative weight (Wr), liver somatic index (LSI), parasite loads, and back-calculated length-at-age. I examined the relationships between water body and genetic identity with Wr, LSI, and back-calculated length-at-age by ANOVA. Parasite loads were analyzed with ordinal regression. Principle component analysis was used to reduce the dataset into a smaller number of principal components and group the study lakes as habitat types. I then used ANOVA to assess whether Wr, LSI, or length-at-age were influenced by genetic strain or habitat type as represented by the principal components and logistic regression to determine whether principal component influenced the abundance of NLMB or non-native largemouth bass. Although similar across bass strains, Wr, LSI, and length-at-age varied by lake (all P<0.0001). Parasite loads were not significantly different between years or among water bodies or genetic strain. Based on the ANOVA results, shallow, nutrient rich lakes with high chlorophyll a concentrations and large littoral zones appeared to promote high Wr and non-native largemouth bass. Conversely, northern largemouth bass were more prevalent, and Wr was lower, in lakes with dense vegetation. Aquatic macrophytes may reduce foraging efficiency of larger, piscivorous fish while increasing survival of young-of-the-year largemouth bass. Reservoirs with little vegetative cover may reduce natural recruitment of resident largemouth bass allowing rapid genetic introgression. These findings may improve our understanding of the effects of FLMB stocking activities and provide the opportunity to modify stocking protocols to better achieve the management goals of the largemouth bass program.

Marsh loss is a problem in many areas around the world. In Louisianas coastal marshes, where Spartina patens is the most common plant, restoration and management seek to slow wetland loss rates that average approximately 77.4 km2/year. To combat the problem, scientists and managers require tools to determine local causes and evaluate the effectiveness of management techniques. Current methods for identifying factors that limit productivity in marshes are too time-consuming or expensive for wide-spread, regular use. Critical values of elemental concentrations in plant tissue are widely used to diagnose mineral deficiencies and toxicities in agricultural crops, however. I used the chemical composition of leaf tissue from S. patens grown under controlled conditions to develop critical values of C:N ratio and concentrations of Na, Mn, and Ca to identify N limitation, salinity stress, and flooding stress, respectively. I tested these critical values and identified seasonal changes in leaf tissue chemistry in a field experiment where all three limiting factors fluctuated naturally. I also compared the leaf tissue chemistry of S. patens and Spartina alterniflora, the second most common plant in Louisianas coastal marshes, to facilitate comparisons between the species. Finally, I investigated the effects of flooding stress on above- and belowground biomass of S. patens. Aboveground biomass was reduced by low N-uptake, high salinity, and high flooding. Belowground biomass was reduced by increased flooding. In leaf tissue, C:N decreased with increasing N-availability and [Na] increased with increasing salinity. [Mn] and [Ca] in leaf tissue decreased with increased flooding. In the field study, C:N increased seasonally. Biomass was most highly correlated with [Na] in spring and with both [Na] and C:N in summer. In fall, leaf tissue composition appeared unrelated to biomass. Managers should take seasonal differences in leaf tissue composition and nutritional requirements into account when diagnosing the causes of limited production and when creating management plans. Leaf tissue should be collected in summer to diagnose limiting factors. To have the most impact, flooding should be used in late spring or early summer to reduce salinity and increase N-availability.

Seasonal overbank flooding in systems like the Atchafalaya River Basin (ARB) provides the opportunity for fishes in the mainstem to access off-channel areas on the floodplain. Typically, newly inundated floodplain habitats allow adult fishes to add biomass, avoid predation, and potentially, to reproduce. However, in systems like the ARB, the timing, duration, and magnitude of flooding infrequently coincides with known reproductive periods of many fishes assumed to be floodplain-dependent. To quantify the level of floodplain-exploitative fish reproduction in the ARB, I collected larval and juvenile fish with a variety of sampling gear that allowed estimates in both ultra-shallow (< 2-m) and continuously-inundated habitats (headwater lakes, canals, and bayous). A suite of water quality parameters, river stage, flow, and hours of daylight were used to gauge the influence of environmental phenomena on age-0 abundance during both inundation and drawdown. The results of the 19-month study suggest that many taxa do not rely on the floodplain to ensure high survivorship. Interestingly, the reproductive ecologies of many ARB fishes appeared to be largely independent of widespread connectivity. Although an increasing hydrograph appeared to enhance reproductive output, the interannual timing and intensity of spawning showed limited variability. Larval densities were also contrasted with the microcrustacean zooplankton (copepods and cladocerans) population to assess if a potential food limitation existed in the weeks and months following hatching. During the study, increased zooplankter abundance was typically preceded by elevated river-floodplain connectivity. Conversely, as floodwaters receded during the summer, zooplankton abundance declined to lowest levels observed during the study. Overall, there was limited synchronous overlap between the hatchlings of most fish taxa and their zooplankter prey. This could have potentially resulted in starvation and reduced annual recruitment. Yet, my analysis of the factors that regulate larval fish abundance in the ARB suggest that the density of zooplankton was highly significant although high numbers of larvae and zooplankton rarely coincided. Finally, I compared the intraday (morning vs. afternoon) density and mean length of larval fish at fixed sample sites. The results suggest that once-daily ichthyoplankton collections may fail to provide accurate density and length measurements for young fish populations.

Greater structural complexity is often associated with more diverse and abundant species assemblages. Biogenic reefs formed by the eastern oyster (Crassostrea virginica) are structurally complex in nature and have been recognized for their potential habitat value in estuarine systems along the Atlantic and Gulf of Mexico coasts. To determine how the structural complexity of newly created oyster reefs may influence the abundance and distribution of species, three objectives were established. First, to examine spatial and temporal patterns of nekton use at newly created oyster reefs, as well as the impact of wave exposure, six paired oyster reef and mud-bottom treatments at low and medium wave energy shorelines were sampled quarterly, from June 2009 to March 2010, at Caillou (Sister) Lake, Louisiana, using gill nets, seine, and substrate trays. Transient species showed seasonal shifts with no evidence of habitat preference. Resident species were consistently more abundant at oyster reefs than mud-bottom treatments. There were no patterns in nekton use that could be directly attributed to wave exposure. Second, to determine how changes within the structural complexity of newly created oyster reefs may influence nekton use, oyster reef treatments of various complexities were created and sampled using a drop sampler. The presence of oyster reefs per se was the most important factor determining nekton assemblages; newly created oyster reefs provided habitat for nekton assemblages, but there was little difference between reef treatments. Lastly, to determine how oyster reefs mediate predator foraging success, treatments of various complexities were created and trials executed in a laboratory setting using wild red drum (Scianops occelatus) and grass shrimp (Palaemonetes pugio). Foraging success was negatively correlated to the structural complexity of oyster reefs, indicating there may be a point above which increased complexity no longer increases the refuge value of the reef. These results show that oyster reefs may support a high abundance and diversity of resident nekton, but that after structure is introduced, further increasing structural complexity does not automatically increase species abundance and diversity, or the amount of refugia provided.

Seeding survival has been a continuing problem since the start of the commercial pine plantation in 1950s. The first-year survival of bare-root loblolly pine seedlings at intensively prepared sites in Louisiana has reached a survival plateau of 75 to 85 % with an average of almost 80 %. The major hypothesis of this research was that the survival plateau is a function of the interaction between the frequency distribution of seedling quality and the frequency distribution of microsite quality. This study examined bare-root seedlings and microsite variation, and analyzed the possible options to increase the first-year seedling survival. The study was approached with simulation and field studies. In simulation study, twenty hypothetical seedling and microsite quality distributions were paired in a manner that simulated 400 plantings. In field study, caliper, stem height, shoot-root ratio, leaf area, and xylem pressure potential were measured for a bale of nursery seedlings and the quality distribution was computed from the seedling volume. Similarly, the microsite variables soil penetration, bed height, moisture content, total mineral nitrogen, and texture were measured and the quality distribution of 8 Weyerhaeuser planting sites was generated from the height increment of associated seedlings. The distributions were combined to predict the first year survival from the assumptions about proportional survival for each pairing. The simulation results provided initial support to the hypothesis that consistent survival results from random pairing of initial seedling and site quality distributions. The average caliper was 4.22 mm for the seedling sample obtained from a local nursery. The sample contained at least 31 % inferior quality seedlings and, the planting sites contained 21 % adverse microsites. Analysis showed that the significant proportion of inferior seedlings and adverse microsites would result in lower average survival based on assumed survival matrix. The elimination of seedlings below 5 mm caliper of the nursery stock increased the survival to 90 % at the cost of 40.9 ¢ per seedling, an increase of 37 ¢ per seedling.

This research assessed the current conditions of water quality in a shallow subtropical lake, influenced by a highly developed urban environment. Specifically, the research focused on the dynamics of phosphorus, nitrogen, and dissolved oxygen, as well as the effects of stormwater runoff on lake water quality. Furthermore, the research quantified gross primary production, net primary production, and respiration in order to discern seasonal variation in lake metabolism. A water quality monitoring platform with multi-parameter probes was deployed in the center of University Lake on the Louisiana State University campus. A series of lake water quality parameters including dissolved oxygen (DO) concentration, temperature, pH, specific conductivity, and cyanobacteria were recorded continuously at a 15-min interval from July 2008 to October 2009. In addition, water samples were collected monthly as well as after major rain events during the study period to determine changes in total phosphorus (TP), nitrate (NO3-), nitrite (NO2-), and total Kjeldahl nitrogen (TKN) concentrations. Results show a monthly average concentration of 0.286 mg/L TP (min max: 0.167 - 0.621 mg/L), 0.053 mg/L NO3-N (below detection 0.24 mg/L), 0.045 mg/L NO2-N (below detection 0.012 mg/L), and 2.5 mg/L TKN (below detection 5.12 mg/L). Mean storm event loading of 28.1 kg (7.5 47.8 kg) caused an immediate increase in total phosphorus within the water column by 14.1%. This resulted in a mean TP lake concentration of 0.383 mg/L. A similar trend was observed in a lesser degree in NO3-N and TKN. Phosphorus loads in the lake were correlated with runoff volume (r2 = 0.71), suggesting runoff volume is the most important factor effecting lake TP concentration after a storm event. Analysis of metabolism rates found a mean annual gross primary productivity value of 4.41 g O2/m2/day, a mean annual net primary production value of 2.13 g O2/m2/day, and a mean annual respiration value of 5.90g O2/m2/day. Annually, 1610 g O2/m2 were produced while the annual sum of respiration was 2150 g O2/m2. Respiration rates were mostly equal to or greater than productivity rates throughout the year, indicating that this shallow subtropical urban lake was net heterotrophic throughout most of the year.

Few studies have examined the effectiveness of timber harvesting Best Management Practices (BMPs) in water quality protection of widely-spread, low-gradient, and highly intermittent headwaters on the Gulf Coastal Plain. Also, a spatial disparity exists between the plot-scale water quality benefits afforded by BMP implementation and the watershed-scale with which most watershed stewardship programs are managed. In this thesis research, paired-watershed and Before-After-Control-Impact (BACI) designs were utilized to quantify plot- and watershed-scale changes in streamflow, as well as baseflow and stormflow Total Suspended Sediment (TSS) concentration and yield for 27 months after clearcut harvesting with and without BMPs in a low-gradient, forested, 3rd-order watershed of north-central Louisiana. Based on analyses of post-harvest baseflow, stormflow, stage-discharge relationships, TSS concentration, and sediment yield, low-intensity (2-8% disturbance of sub-watershed drainage area), clearcut harvesting adjacent to streams and with BMP implementation did not impact streamflow or sediment transport at the plot- or watershed-scale. No difference was found between treatment periods for monthly baseflow discharge measurements or in peak water level response to storm events. Flow duration curve analysis showed that baseflow decreased during the post-harvest period, possibly due to differences in the timing of precipitation between treatment periods. Changes in the stage-discharge relationship were observed downstream of harvesting without BMPs, indicating harvest-induced changes to stream geomorphology. Baseflow and stormflow TSS concentration (mg L-1) and yield (kg ha-1 mo.-1) were similar between treatment periods and were on the lower end of published results for Coastal Plain sites. Post-harvest TSS yield increased downstream of harvesting without BMP implementation when high flow events were included in yield calculations. These results indicate that current Louisiana BMPs for timber harvesting are effective in mitigating sediment runoff at the plot- and watershed-scale for conditions similar to the monitored sites, which include an abundance of beaver/debris dams and highly intermittent streamflow. These natural conditions may have further improved sediment reduction from BMP implementation through ponding and reduction of flow rate and carrying capacity. The potential exists for future studies to determine the intermittency and beaver dam impacts to streamflow and sediment transport as forest disturbance increases throughout the watershed.

Research was conducted to implement remote setting technology for the commercial- scale production of oysters on alternative cultch material in Louisiana. Genetic research has created an enhanced oyster stock for commercial cultivation. Since the Louisiana oyster industry relies on natural seed production for product, they are unable to utilize these new advances and are never guaranteed a reliable source of oyster seed. Remote setting is one way of producing a consistent supply of genetically improved oysters. Cultch material for oyster settlement is a limiting resource. Several alternatives to clamshell have been explored. The objectives of this study are: (1) test larval setting success (penetration) using three cultch types in aerated and non-aerated treatments; (2) characterize triploid oyster seed growth and abundance among three cultch types on a commercial oyster lease; (3) characterize ploidy dilution of planted triploid oyster seed by natural diploids among three cultch types on an oyster lease; and (4) document the costs associated with seed bedding. Limestone had significantly less spat set than both whole and crushed oyster shell; spat set significantly less at bottom-depth and mid-depth in non-aerated compared to aerated treatments (objective 1). This shows the importance of aeration for spat settlement distribution among depths. Oyster growth was significantly less on limestone compared to both whole and crushed shell (objective 2). Spat on limestone were limited in two dimensional growing space. Of the fifty individuals sampled from each cultch type for percent triploidy, 2%, 4% and 8% were triploid for crushed shell, limestone and whole shell, respectively; all cultch types showed at least 92% diploid dilution (objective 3). These results are based on one site and season, therefore cannot be used to make conclusions about overall practicality for oyster production in Louisiana. The cost associated with seed bedding is $6.00 per barrel (objective 4). Whole oyster shell had significantly more initial spat set and growth than limestone. There was no significant difference in oyster abundance among all cultch types.