Current Research Projects
Assessing Species-Habitat Relationships in Priority Open Pine Ecological Systems in MississippiAmber Owen, senior wildlife, fisheries and aquaculture major, surveyed six species in imperiled open pine ecosystems. In the Southeastern U.S., indicator species in imperiled open pine ecosystems provide wildlife managers and biologists insight into habitat conditions. In this study, Owen surveyed six indicator species: Bachman's sparrow (Peucaea aestivalis), Brown-headed nuthatch (Sitta pusilla), Northern bobwhite (Colinus virginiana), Red-cockaded woodpecker (Leuconotopicus borealis), Pine warbler (Setophaga pinus), and Prairie warbler (Setophaga discolor) in three management units containing upland pine forest in Sam D. Hamilton Noxubee Wildlife Refuge, Mississippi. These species prefer open understory layers in a predominately pine system and the presence of some forbs and grasses. Owen used empirical point transect bird and forest structure surveys across a random sample of 161points during June 2016 to create species-habitat association models to further knowledge of how priority species interact with their habitat. She then used Poisson regression models in an information theoretic approach to assess target species relationships with forest structure variables. Pine and Prairie warblers were the most detected target species across management units. Of the target species, sample size allowed for habitat association models for Pine warbler, Prairie warbler and a guild of pine-grassland birds that included Blue Grosbeak (Passerina caeruea) Bachman's sparrow, Northern Bobwhite, and Field Sparrow (Spizella pusilla). The model for Pine warbler showed a preference for pine basal area. Prairie Warbler models showed no noticeable preference for one habitat component with relative abundance influenced by components of pine basal area and overstory, hardwood midstory, and shrub and herbaceous ground cover. For the grassland guild, the preference was for midstory and understory.
The Gulf KillifishJacob Jones, a senior wildlife, fisheries and aquaculture major, is studying the gulf killifish. The Gulf killifish (Fundulus grandis) is a small, euryhaline fish species occurring in estuarine areas along the northern Gulf of Mexico coast. This species is capable of rapidly moving between fresh and salt water and as a result is quite hardy to environmental changes, leading to its popularity as a baitfish for sportfish species such as Red Drum (Sciaenops ocellatus) and Spotted Seatrout (Cynoscion nebulosus). This remarkable trait also facilitates its culture in low salinity conditions, even freshwater ponds. However, several culture bottlenecks remain before Gulf Killifish can be developed at the commercial level. The primary bottleneck is the collection and incubation of eggs. In nature, females deposit eggs episodically on vegetation during new moon phases of tidal cycles. Recent research has shown that eggs can be collected in spawning mats suspended in tanks and incubated in a moist aerial environment, such as an incubator. Limitations are in difficulties associated with the separation of eggs from collection mats, and the space required to incubate eggs. Simplification of the process of egg collection and incubation is needed prior to commercialization. Therefore, this experiment compared egg deposition and fertilization by Gulf Killifish and the difficulty of collecting eggs using three different collection devices: spawning mats, disk collectors, and nylon mesh collectors. Adult Gulf Killifish were distributed in three commercial-scale (4,000-L) recirculating tank systems at 9 ppt. All three types of egg collection devices were used in each tank, with eggs collected every two days, quantifying collection time, egg abundance, and fertilization percentage. It was determined that the mesh collector was the most efficient device to use for collecting eggs from Gulf Killifish.
Plot Size and Prediction Model Form Effects on Stand Diameter Distribution Recovery MethodsJosh Bankston, a forestry student, looked at various impacts on the accuracy of growth and yield models. Diameter distribution information of a forest stand provides insights in to the stand's value. In whole stand forest growth and yield systems, the future diameter distribution of a stand gets predicted by use of models that predict the stand's diameter moments and/or percentiles in conjunction with a mathematical system to recover the diameter distribution from the predicted moments and/or percentiles. For industry applications, it is important to know which combination of plot size, recovery method used, and moment/percentile equation form yields the most precise prediction of future stand diameter distribution. Being able to more accurately predict future stand diameter distribution can give better financial insight and aid in management making decisions associated with sampled forest stand. This research examined how various plot sizes which data were recorded from, recovery methods used to obtain predicted diameter distribution, and how different forms of moment/percentile prediction equations affect the precision of model development. Data used for this analysis were collected from 202 quarter-acre measurement plots that were located within 56 loblolly pine stands across Alabama, Arkansas, Louisiana, and Mississippi. The data were collected by the Mississippi State University Loblolly Pine Cooperative between 1981 and 1989. A total of 336 measurement observations were accounted for. Results indicated that precision is generally lower for smaller plot sizes. Changes in precision rank for the different recovery methods may occur as plot size increases. This result falls in line with our hypothesis of larger plot sizes yielding more precise results. Trends also show that distribution recovery methods are also less accurate in thinned stands. This trend is most likely due to disruption of the normal distribution when a stand undergoes a thinning operation. The quarter-acre plot size was shown to yield the most accurate results, but the slight increase in precision may not prove to be worth the extra money when a smaller plot size can be used and yield similar results.
Prescribed Fire in Upland Oak Forests: Impacts on Canopy Cover, Organic Layer, and Seedling DensityJosh Byers, senior Natural Resources and Environmental Conservation major, studied how canopy cover and depth of the soil organic layer could affect seedling density. Upland hardwood forests are beginning to shift from predominantly oak forests, to more competitive, shade tolerant species due to lack of fire in the ecosystem. Prescribed fire should help eliminate the seedlings of shade tolerant species that are out-competing oak seedlings, as well as increase understory light and reduce the soil organic layer. After analyzing the results of my experiment, out of the six sites, we concluded that: Single fire had minimal impact on canopy cover, multiple fires reduced canopy cover by 10%, both single and multiple fires reduced SOL depth, and finally, single fires had no impact on oak densities compared to unburned plots, but increased red maple densities. The analysis of the results suggests that with fire, oaks cannot compete with shade tolerant species.
Improving growth and yield estimates from growth index ratio method of stand table projection in red oak-sweetgum bottomland hardwood standsForestry senior Leah Leonard, investigated how well Stand Table
Projection (STP) methods of forest growth and yield prediction performed in red
oak-sweetgum bottomland hardwoods stands. Data used in the project were
collected from 1981-2005. Two approaches of STP were investigated: the traditional
STP on diameter classes and an alternative approach of implementing STP on
individual trees. STP was implemented under assumption that tree survival was
predicted without error and also under survival prediction system that contained
error (a mortality model). Under the assumption that tree survival was predicted without error, STP was generally unbiased for stands in the age range 35-70 years but over predicted yields and stumpage value for stands below 35 years and under predicted yield in stands older than 70 years. When tree survival was predicted with error (using a mortality model), STP generally over predicted yields and stumpage value across all ages. Comparing traditional STP to STP on individual trees, there was no clear advantage of carrying out STP on individual trees.
Viability of CavitiesSamantha Rushing, wildlife, fisheries and aquaculture major, studied the key characteristics of cavity trees most important to bats and what actions should be taken to increase cavity tree numbers and preserve these habitat features. Using field surveys, information on physical characteristics of known cavity trees were collected and compared to information collected on the same cavity trees 10-13 years ago. The research found that occupancy of cavity trees by bats is highly dependent on the size of the tree, but not all trees with basal cavities likely support bats. Bats were found in larger diameter trees with larger cavity area, those promoting stable microclimates through thicker cavity walls. The likelihood that trees survived to a size large enough to support these cavities is relative high (likelihood of tree remaining through the period between surveys was 86%). Therefore, preservation strategies should be put in place to allow for cavity trees to grow to a suitable size, thus increasing availability of these trees to bats.