Biology Theses and Dissertations
Permanent URI for this collectionhttp://hdl.handle.net/1903/2749
Browse
4 results
Search Results
Item IMPACTS OF WEIGHTING SCHEMES AND TRANSFORMED ENVIRONMENTAL VARIABLES ON BIODIVERSITY MODELING WITH PRESENCE-ONLY DATA(2017) Pradhan, Kavya; Fitzpatrick, Matthew C; Marine-Estuarine-Environmental Sciences; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Biodiversity modeling techniques at the community- and species-level can be used to address questions in ecology, management, and conservation. I addressed aspects of community-level and specie-level models using virtual and inventoried species in North and South America. Firstly, I assessed the effectiveness of two weighting schemes in reducing impacts (if any) of five sampling routines (simulating unrepresentative sampling in presence-only data) on the model performance of Generalized dissimilarity model (GDM). Unrepresentative sampling lowers model performance, but weighting species can reduce this negative impact to a certain extent. However, PO data severely impacts GDM’s ability to detect the relative contribution of environmental gradients. Secondly, I examined the potential of (GDM) transformed environmental variables in improving the performance of Maxent models (presence-only) along with the influence of range size, sample size, and species dependence type. Transformed environmental variables improved model performance, especially when used with small-ranged species and/or low sample sizes.Item NOVEL APPROACHES TO STUDYING BIODIVERSITY IN REMOTE AREAS: DISTRIBUTION OF LICHENS AND PENGUINS ACROSS THE ANTARCTIC PENINSULA(2013) Casanovas, Paula Victoria; Fagan, William F; Lynch, Heather J; Biology; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Biodiversity inventories are a critical resource, providing baseline information for assessing environmental changes over time. In many cases, the underlying datasets are generated by "opportunistic" sampling efforts or they are consolidated from diverse datasets collected for different purposes. These datasets are typically patchy and incomplete, requiring the use of sophisticated statistical analyses. The Antarctic Peninsula (AP) is one of those areas where direct observation of species distribution is difficult; it is also an area that in recent decades has been experiencing important environmental changes, which influence population and ecosystem dynamics. I addressed biogeographical questions in the AP archipelago, using remote sensing and opportunistic data sets for two very different groups of organisms: lichens and penguins. Although taxonomically different, both groups are key components of the AP terrestrial ecosystem, and share the need to couple biodiversity surveys with modeling to understand species distribution and abundance patterns in large areas of remote wilderness. The results of this dissertation work are interesting to polar biologists, because evidence suggests that the input of nutrients by seabirds can significantly impact floral diversity and abundance in nutrient-poor polar communities. The datasets and protocols for data collection and analyses generated in this project are valuable in themselves for the scientific community. They could be used as the basis for a valuable and practicable monitoring program and procedures for the evaluation of the data derived from it. In the Antarctic Peninsula in particular, this information will aid in the delineation and management of protected areas, as well as in the evaluation of the impacts of climate change and human visitation to the most traveled locations. Furthermore, this research provided an example of how an approach that integrates the use of existing remote-sensing products with independent ongoing field sampling efforts, "citizen scientist" data collection, and historical datasets can yield low-cost, high-benefit studies that can be useful both to understand how species respond to their environment, and to help environmental managers to predict and cope with imminent changes due to global warming.Item Diversity, Invasibility, and Resource Use in Marine Fouling Communities of San Francisco Bay(2011) Altman, Safra; Reaka, Marjorie L; Hines, Anson H; Behavior, Ecology, Evolution and Systematics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Invasive species threaten the biodiversity of estuaries worldwide. To examine the relationships between biodiversity, invasibility, and invasion success, I conducted field surveys and experiments in San Francisco Bay marine fouling communities, including 1) surveys to estimate alpha, gamma, and beta diversity of native, non-native and cryptogenic components of the community; 2) experiments to assess the influence of diversity and resource availability on short-term recruitment of novel non-indigenous species (NIS) into test communities and subsequent community development over time; and 3) an experiment to explore the role of facilitative interactions of NIS in the diversity-invasibility relationship. Surveys (10-24 sites) showed that non-native alpha diversity was significantly greater than native or cryptogenic alpha diversity, beta diversity was significantly greater for native and cryptogenic species than for NIS, and gamma diversity was similar for NIS and native species. These results indicate that native species had high turn over from site to site while NIS were spread throughout the Bay. Experiments showed that on short time scales (2-4 weeks), the effect of initial diversity on the density of recruitment of NIS was significant and negative, with no effect of resource level (increased open space). Changes in community composition over time (2-24 weeks) also indicated significant inverse relationships between percent cover of NIS and diversity of the initial community with no evidence of a resource effect. Abundant NIS occupied less space in communities with higher initial diversity. However, the same NIS occupied (i.e., had invaded) all experimental communities regardless of starting diversity. Additional experiments revealed that recruitment to secondary substrates did not vary significantly with invasive species diversity or resource availability. When total recruitment to primary and secondary substrates were combined, there was no longer a significant relationship between diversity and recruitment. Analysis of secondary settlement patterns revealed that some NIS, such as Bugula neritina, were facilitating recruitment and settlement of additional NIS. In contrast, other species, such as Clathria prolifera and Botryllus schlosseri, inhibited secondary settlement of NIS. The influence of diversity and primary resource availability on secondary settlement did not appear to affect settlement on facilitative species, but reduced settlement on inhibitive species.Item Effects of the thermal effluent from C.P. Crane Generating Station on submersed aquatic macrophyte communities in the Saltpeter-Dundee Creek system(2007-08-13) Beser, Todd M.; Engelhardt, Katia; Marine-Estuarine-Environmental Sciences; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)While water quality is often cited as the main factor that controls the distribution of submersed aquatic macrophytes (SAM) in the Chesapeake Bay, additional factors associated with physical and/or biological disturbances also affect the distribution. At local scales, such as in Saltpeter Creek, a tributary to the Gunpowder River, the thermal effluent from C.P. Crane Power Plant may be an important environmental gradient. I mapped the temperature signature of the effluent in Saltpeter Creek and intensively sampled the plant community structure to investigate the ecological similarity of SAM communities within and across different thermal regimes. I also conducted growth chamber experiments to study how different species and populations sampled from different temperature regimes respond to a controlled temperature gradient. Analyses show that although significant differences in water temperature exist across the study site, differences in temperature do not appear to significantly drive the plant community composition of the system.