Larval bay anchovy (Anchoa mitchilli) habitat quality in Chesapeake Bay was predicted using an individual-based model applied to spatially-detailed field data from Rilling and Houde (1999). Habitat quality was predicted using the ratio of instantaneous mortality rate to instantaneous growth rate. Model predictions of habitat quality were compared to field estimates of habitat quality derived from the spatially-detailed field data. Three sets of one-day simulations were performed to estimate larval growth and mortality rates throughout Chesapeake Bay during June and during July 1993. Field-based simulations used field data to estimate the model inputs of water temperature, zooplankton densities, and the densities and sizes of bay anchovy larvae and gelatinous predators (Mnemiopsis leidyi and Chrysaora quinquecirrha). Standardized larvae simulations used the same field data, but standardized larval sizes and densities throughout the Bay. A third set of simulations was performed to determine the relative importance of six factors in determining the bay-wide spatial variation in predicted growth and mortality rates. Model predictions from the field-based simulations produced spatial patterns of habitat quality in the Bay that sometimes conflicted with the otolith-based predictions of Rilling and Houde. Field estimates of anchovy egg and larvae abundances were generally high in regions predicted to have low M/G ratios, but low in regions with low otolith-estimated M/G ratios. The standardized larvae simulations generally supported the conclusions of the field-based simulations. The effect of habitat quality on larval production was evaluated using the predicted mortality rates from the two sets of simulations. Initial larval abundances dominated the percent of survivors projected 20-days into the future that a region would produce, but when larvae were standardized across the bay, differences in habitat quality among regions was important in determining the relative contributions of survivors by region. Initial larval length and zooplankton densities were the most important factors determining the spatial variation in growth rate, while predator density was most important for mortality rate. Future research should focus on field and laboratory data collection to resolve the discrepancy between model-predicted and otolith-estimated M/G ratios.

This study examined various practical and theoretical aspects of disturbance in a coastal wetland marsh in southern Louisiana. A literature review approached disturbance ecology from both practical and theoretical perspectives and assessed its applicability to developing broad predictive models. However, specific knowledge of environmental variables, competitive relationships, and the interactive effects of multiple disturbances are required for meaningful usage of these models. The Lacassine National Wildlife Refuge (LNWR) proved to be an ideal laboratory to test various aspects of ecological disturbance theory. I found that the primary disturbances affecting the LNWR have been hurricanes, droughts, water-level manipulations, prescribed burning, oil and gas recovery activities, grazing by Myocastor coypus (nutria), and managed cattle grazing. The 1990's application of three-dimensional (3-D) seismic technology used in the oil and gas recovery business challenged landowners, government regulators, and industry to develop ways to recover these resources without damaging surface features. I developed a conservative estimate that an area exceeding 2.5 times the area of Louisiana's coastal wetlands was covered by overlapping seismic surveys in southern Louisiana from 1997 through 2002, equal to 22.5 km²/year. I provided a general overview of 3-D seismic survey programs, potential adverse impacts, and management and restoration strategies. I also conducted a field study at the LNWR on vegetation in control and treatment transects before, and for two years after, a 3-D survey. I found vegetative cover and the amount of dead plant biomass were significantly lower in treatment plots, but live biomass was not different in treatment and control plots. Species richness was higher in treatment plots compared to control plots, but the live biomass and cover of the dominant species (Panicum hemitomon) was lower. The live biomass and cover of Eleocharis spp., a colonizing species, was greater in treatment plots compared to control plots. There was no significant effect of equipment type or traffic level within treatment plots for total live cover, total live biomass, or total dead biomass. Clear trends of the disturbance effects across disturbance types and habitats were not revealed. Furthermore, extrapolating the effects of a disturbance using the available general concepts should be done with caution because of the overriding influence of the site and species on disturbance effects.

The habitat value of petroleum platforms for red snapper, Lutjanus campechanus, is poorly understood. However, it is widely recognized by both scientists and fishermen that the presence of platforms in the northern Gulf of Mexico (GOM) has affected the distribution of red snapper by the addition of hard substrate habitat. I evaluated the habitat value of standing and toppled platforms by monitoring the fidelity of red snapper to these structures with acoustic telemetry. In May 2003, 125 red snapper were captured with hook and line at several platforms in a 35-km2 portion of the South Timbalier oil and gas lease blocks, 50 km south of Port Fourchon, LA. Following anaesthetization with MS-222, an individually coded acoustic pinger was surgically implanted into the peritoneal cavity of each fish. After a short recovery period the red snapper were released at five platforms in the study area. Presences of individual snapper were recorded with omnidirectional acoustic receivers attached to seven platforms, and to one artificial reef, a toppled platform. Red snapper exhibited little movement between platforms in the study area. However, logistic regression showed a high initial fidelity to release location which subsequently decreased over time, thus site fidelity was found to be high in the short-term, but much lower in the long-term. This result differs from previous studies on red snapper fidelity that reported high fidelity over longer time spans. Red snapper recaptured outside of the study area showed little uniform directional movement. Estimates of instantaneous fishing mortality on this population were higher than those predicted by the most recent stock assessment. A Fourier analysis revealed a diel pattern of movement away from the structures at night, most likely for offsite foraging. Knowledge of red snapper fidelity to petroleum platforms will lead to more effective management of this species by clarifying both the specific function of these structures as habitat and their importance to the red snapper population in the GOM.

Habitat degradation of saltmarsh ecosystems reduces the suitability of the marsh as permanent environments for resident species and seasonal nursery grounds for transient fauna. This study was conducted in Terrebonne-Timbalier bays near Cocodrie, Louisiana. Fishes and macroinvertebrates were collected using a drop sampler in marsh-edge and open-water habitat types, in three locations (i.e., pond, channel, and bay), and four seasons (Fall of 2000, Winter of 2000-2001, and Spring and Summer of 2001). The specific objectives of this study were: (1) to determine the difference in composition and structure of fish and macroinvertebrate communities, (2) to establish the food habits of the fish community, and (3) to determine the dependence on Spartina alterniflora of fishes and macroinvertebrates through the use of stable isotope techniques. Densities of resident and transient species including darter goby, naked goby, and brown shrimp were detectably different among habitat types, seasons, and locations. The naked goby consumed higher abundances of harpacticoid copepods along the marsh edge than in the open water. Bay anchovy and naked goby had detectable differences in prey utilization among seasons and ingestion was associated with changes of environmental variables. Despite the presence of detritus in the stomachs of several fish species, it was relatively rare. Naked goby was the only species that had detectable differences in carbon stable isotopes between mash-edge and open-water habitat types. Significant seasonal variations were identified in the carbon isotopic values of naked goby, daggerblade grass shrimp, and blue crab, and in the nitrogen isotopic values of brown shrimp. In the field experiments with naked goby, differences in prey utilization and in carbon or nitrogen isotope values were not detectable among habitat types. The marsh edge is essential for food and refuge for estuarine nekton, but macroinvertebrates densities may be more sensitive to marsh loss than fish densities. Nevertheless, marsh loss may have an effect on the feeding ecology of juvenile fishes that rely primarily on benthic prey along the marsh edge. The relative contribution of Spartina alterniflora to support the nekton community was less than 35% in the study area.

Pore geometry and topology are important determinants of sediment physical properties, such as porosity and permeability. They also influence processes that occur in the sediment, such as acoustic propagation, attenuation, and dispersion, single- and multi-phase fluid flow, and hydrodynamic dispersion. This study uses images to evaluate pore geometry and topology of ooid (subspherical particles) and siliclastic (angular quartz) sand that was collected from the marine environment south of Bimni Bahamas and Ft. Walton Beach, FL, respectively. Image analysis techniques and predictive tools enable insight into the relationships among sediment pore geometry, topology, and physical properties for these differently shaped sands. High frequency acoustics utilize short wavelength signals to evaluate sediments. Correspondingly short length scales are then needed for sedimentary property predictions, which is possible with planar and volumetric image analysis of sand. This data was compared to data obtained by direct large scale measurements (e.g., water weight loss, constant head permeability) were made. Mean porosity differed by as much as 0.04 and mean permeability showed good agreement and differed by a factor of 2. Given that the image analysis predictions were made from much smaller samples (~equivalent to the length scale of the high acoustic frequencies used) than the bulk samples, a sediment characterization at acoustically relevant length scales is possible. It was also demonstrated that for these homogeneous sands (i.e., ooids and quartz) two-dimensional pore geometry and topology are quite similar to three-dimensional pore geometry and topology (i.e., pore connectivity). Additionally it was determined that pore network models typically overestimate the topology and therefore, in order to match image and bulk predictions of sediment properties, these models must underestimate the conductance of individual pore throats (i.e., conductive element in sand). Typically pore throats are depicted as straight cylinders. Image data suggests that pore throats are better represented by biconical shapes where conductance is as much as 3 times higher than conductance within the straight cylinders. These findings indicate that increased realism in pore throat shape (higher conductivity) and in topology (fewer pore throats) may significantly influence network model evaluations of fluid flow or acoustic propagation in marine sand.

The Barataria-Terrebonne estuary has been eroding at a rate of up to 103.6 km2 yr-1 for several decades. If the current rate of loss is not reduced, an additional 2,550 km2 of coastal wetlands will be lost by the year 2050. Currently, stormwater in Terrebonne Parish is pumped into canals, ultimately discharging into the Gulf of Mexico. An opportunity exists to use this stormwater for wetland restoration; however, the ecological impacts of stormwater diversions on wetlands are unknown. The objectives of this project were to 1) to investigate the seed banks of a degraded marsh to determine if a viable seed source exists, 2) to gather baseline soil chemistry of a degraded marsh prior to receiving stormwater input, and 3) to compare the soil P chemistry, metal concentrations and accumulation rates, and sedimentation rates for different wetlands receiving stormwater input. An existing degraded marsh that is scheduled to receive stormwater input in the fall of 2004 was selected for the baseline study. These results will be compared to conditions after a new stormwater pump becomes operational. Coastal wetlands that have been receiving stormwater for <10 years and 30 years were selected to carry out the sediment chemistry study. A total of 370 stems germinated from the seed banks from the vegetated areas and a total of 2 stems were counted in the seed banks from the mudflat. These results suggest that replanting will be necessary for establishment of vegetation in the mudflat areas. There was a significant increase (p<0.05) in P concentrations in soil that had been receiving stormwater for 30 years when compared to soil that had been receiving for stormwater for <10 years, which may lead to alteration of plant communities as seen in the Everglades. Sediment profiles indicated a significant difference in Fe, Al, Zn, and Pb concentrations when comparing the age of the pump (P<0.05). However, soil metal concentrations were found below EPA toxicity thresholds even in coastal wetlands receiving stormwater input for 30 years. The results from this study will be used as a foundation for future studies of stormwater input to coastal Louisiana wetlands.

The western Louisiana inner shelf along the Chenier Plain coast has experienced fluid mud deposition in response to increased fine sediment supplied by the Atchafalaya River since ~1950s. The goal of this project is to assess the characteristics comprising the sedimentary strata and stratigraphic architecture that result from mud deposition on the inner shelf. Recent time-series cores collected from the inner shelf along the Chenier plain coast show the region is subject to transient fluid mud deposition, leading to high long-term accumulation rates. Sediment cores were collected in May 2001, March and May 2002 from the inner shelf landward of the 10 m isobath, 100 km west of Atchafalaya Bay. Porosity, granulometry, X-radiography, and 7Be, 210Pb and 137Cs geochronology indicate the presence of high-porosity event layers 2-25 cm thick composed of clay with basal silt laminations. These event layers appear to concentrate around a depocenter located 95-110 km west of the Atchafalaya River, landward of the 7 m isobath, but can be ephemeral features on a seasonal time scale. The combination of high-energy benthic hydrodynamics and sufficient fine sediment can result in cross-shelf gravity-driven flows (on very low slopes) that can blanket hundreds of square kilometers to thicknesses exceeding 10 cm. The sedimentary fabric that results from gravity-driven flows consists of a stacked pattern of predominantly fine-grained, fining upward packages. The resulting morphology of the shelf may be a clinoform, with maximum deposition occurring on the foreset (convex upward) region. The observations from the western Louisiana inner shelf (mud/silt couplets that comprise the fine-scale stratigraphy of the region, ephemeral 7Be deposits, and clinoform morphology exhibiting erosional features in sidescan and chirp data) are consistent with the wave-enhanced gravity-driven flow model. These results indicate that wave-enhanced gravity driven flows may be an important component in the dispersal of Atchafalaya River sediment.

Knowledge of mangrove litter dynamics is crucial to an understanding of the energetic links between mangrove ecosystems and nearby estuaries and coastal waters. Previous research into the role played by macrodetritivores in Neotropical mangrove litter processing has been contradictory. This study used leaf tethering to examine the effects of macrodetritivores on initial rates of mangrove litter degradation in South Florida, USA. Leaf litter dynamics experiments were run in both natural and restored mangrove forests to assess functionality of the restoration projects. Although less important than in some other parts of the world (e.g., the Indo-Pacific), macrodetritivores played a significant role in increasing in situ leaf degradation within mangrove forests on both east and west coasts of Florida. In contrast to Indo-Pacific forests, gastropods were the primary macrodetritivores usually observed feeding on abscised mangrove leaves in South Florida. During leaf tethering trials, macrodetritivores (the gastropod Melampus coffeus and grapsid crabs) attacked between 1.7 and 29.6% of deployed leaves at different sites and accounted for 24.0% more leaf mass loss compared to non-attacked leaves. Macrodetritivores increased leaf litter degradation in several different mangrove forest types and under different environmental conditions within this study. In addition, macrodetritivore degradation rates varied among site histories (natural versus restored) and macrodetritivore characteristics. Although macrodetritivore community populations were greater in restored forests, litter cycling processes were similar to nearby reference forests. Litter and environmental characteristics were also examined to help understand macrodetritivore impacts on litter cycling. Litter fall rates varied from 0.4 ± 0.1 to 3.2 ± 0.3 g m-2 d-1 at the several sites. Standing stocks of litter on the forest floor varied widely within and among sites and ranged from 29 g m-2 to 559 g m-2 with an overall mean of 269 g m-2. Differences in litter fall and litter standing crop across sites and seasons were partly attributable to environmental variables (i.e., temperature, porewater salinity, forest structure, and light). The restored mangrove forests observed during this study appear to have gained some natural functionality of leaf litter dynamics as compared to nearby reference forests.

In this study, I quantified freshwater and nutrient inputs in the Breton Sound estuary which is receiving freshwater reintroduction in an effort to restore deteriorating wetlands. Almost all wetlands of the Mississippi deltaic plain are isolated from riverine input due to flood control levees along the Mississippi River. This has altered water and nutrient budgets and is a primary cause of the massive wetland loss in the delta. Maintenance of the delta depends on a healthy, functioning ecosystem which includes riverine input. The Breton Sound estuary is located southeast of New Orleans and until recently was hydrologically isolated from direct riverine input. In 1992, a freshwater diversion became operational at Caernarvon, LA that re-introduces freshwater, nutrients, and sediments from the Mississippi River into the estuary. Several inputs and losses were calculated for three annual (2000, 2001, and 2002) water budgets including precipitation (PPN), potential evapotranspiration (PET), the diversion, stormwater pumps, and groundwater. The inputs of ammonium (NH₄-N), nitrate (NO₃-N), total nitrogen (TN) and total phosphorus (TP) were determined for each of the water sources. There was a different precipitation pattern for each of the years for which water and nutrient budgets were calculated. Precipitation contributed 48-57% of freshwater input while the diversion structure accounted for 33-48%. The net input of fresh groundwater was 3 to 4 orders of magnitude less than diversion input and precipitation. Atmospheric deposition was the largest contributor of NH₄-N accounting for 62-72% of the total NH₄ input followed by the diversion (total annual NH₄-N input was 1.39x10⁵ to 1.96x10⁵ kg). NO₃-N input to the estuary was an order of magnitude greater than NH₄-N input. The diversion was the greatest source of nitrate to the study area (7.78x10⁵ to 1.64x10⁶ kg) contributing 77-88% of total nitrate input. The diversion contributed 1.26x10⁶ to 2.10x10⁶ kg of TN, representing 77-79% of TN input. The diversion contributed 81-98% of TP input and was an order of magnitude greater than precipitation and stormwater pumps combined. Annual loading rates of NH₄-N and NO₃-N were 0.16-0.22 and 1.6-2.2 gNm^-2y^-1, respectively. TN ranged from 1.9-3.2 gNm^-2y^-1 and TP ranged from 0.17-0.29 gPm^-2y^-1.

Wetlands improve water quality through sedimentation and the uptake of excess nutrients. As human population increases in the coastal zone, wetlands receive greater nutrient inputs. These additional nutrients may accelerate microbial activity, leading to faster decomposition rates. This decomposition could exceed belowground organic matter production, resulting in a net reduction in soil organic matter accumulation and vertical marsh accretion. The effects of nutrient enrichment on belowground organic matter decomposition in subtropical marshes have received little attention. As such, this research examined the effects of four levels of nitrogen combined with two levels of phosphorus enrichment on belowground decomposition through the use of cotton strip and litter bag assays in a Sagittaria lancifolia dominated marsh in Madisonville, Louisiana. Litter bags contained S. lancifolia root or shoot tissues; roots were of uniform tissue quality while shoots were from unenriched or enriched soils. Soil nitrogen and phosphorus applications both significantly increased belowground decomposition rates of cotton strips. The effect of tissue quality on shoot decomposition was dependent on nitrogen soil enrichment level. At low nitrogen enrichment levels, low quality shoot tissues decomposed more slowly than high quality tissues; this relationship was reversed at high nitrogen soil enrichments. Also, the effect of phosphorus enrichment on shoot decomposition was dependent on the level of nitrogen enrichment. Phosphorus soil enrichment only increased decomposition at the high nitrogen levels. Similarly, phosphorus enrichment combined with moderate nitrogen enrichments raised the decomposition rate of labile root tissue components. However, neither nitrogen nor phosphorus enrichments affected the decomposition rate of recalcitrant root components. Cellulose decomposition was positively correlated with interstitial pH. Shoot decomposition and the recalcitrant root decomposition rate also positively correlated with interstitial pH. This research demonstrated that nitrogen and phosphorus soil enrichments affect the decomposition of roots, shoots, and cotton strips, though in different ways. Variations in the nutrient and carbon quality of the individual tissues, as well as abiotic factors such as pH, modify the effects of soil nutrient enrichments on the decomposition of different tissues in the study marsh.