With steady growth in global commerce and intensified ship traffic worldwide, comes the increased risk of invasion by non-indigenous organisms. Annually, >7000 vessels traveled across Galveston Bay, Texas from 2005-2010. These vessels discharged ~106 million metric tons of ballast water, equivalent to ~3.4% of the total volume of the Bay. A majority of these discharging vessels originated from around the Gulf of Mexico and the Caribbean Sea. By evaluating the source and frequency of inoculations from various locations, we are striving to assess the invasibility risk to Galveston Bay by way of ballast water. We identified organisms from Galveston Bay, ballast water samples and growout experiments using molecular methods. To our knowledge, this is the first utilization of molecular methods to identify the phytoplankton community within Galveston Bay. Within Galveston Bay, we identified 15 genera of dinoflagellates, 2 of which have previously gone undetected including Takayama and Woloszynskia. Thirteen ballast water samples yielded twenty genera of Protists, Fungi or Animalia from at least ten different phyla. With more than seven genera identified, dinoflagellates were the most diverse group: including the known toxin producer Pfiesteria and Scrippsiella which has not previously been detected in Galveston Bay. The most common diatoms in the ballast water samples were Actinocyclus, Ditylum, Nitzschia, Stephanopyxis and Thalassiosirales. At the termination of the growout experiments eight genera of phytoplankton were identified including: Dinophysis, Gymnodinium, Gyrodinium, Heterocapsa, Peridinium, Scrippsiella, Chaetoceros and Nitzschia. With these findings, Galveston Bay has the potential to be both a recipient and donor region of dinoflagellates. Dinoflagellates, capable of forming harmful algal blooms leading to fish and shellfish kills, are being transported to Galveston Bay via ballast water. Our results suggest that Galveston Bay is at risk for invasive species introductions via ballast water and support the idea that a monitoring system within the ports as well as the bay should be put in place. The actions would help to maintain the current health of this ecosystem and aide in preventing a negative impact in the event of successful establishment of a non-indigenous species of phytoplankton transported to Galveston Bay via ballast water.

Nearly half of the population of the United States lives in coastal regions, and millions of visitors from across the nation and world enjoy the coasts every year. Coastal and marine areas provide for recreation, economic activities essential for the financial health of the nation, and vital ecological services. As they provide so many benefits to the U.S., it is vital to protect and preserve the coastal and ocean areas from the increasing, competing demands they are facing. In order to protect and preserve these complex systems, a comprehensive approach incorporating science, engineering, humanities, and social sciences should be taken; this approach is commonly referred to as Ecosystem-Based Management. This dissertation focuses on developing a framework that can be used to identify appropriate sub-regions in Northern Gulf of Mexico coastal and marine environments for the purposes of Ecosystem-Based Management. Through this work, the roles of three management protocols used for managing coastal areas – coastal and marine spatial planning, ecosystem-based management, and integrated ecosystem assessment – were examined individually as well as their integrations with each other. Biological, ecological, physical, human, and economic indicators for partitioning an ecosystem were developed and weighted for each management protocol using the analytic hierarchy process and expert elicitation. Using the weighted indicators, a framework for identifying sub-regions and estuarine classification system was developed. The framework and classification system were applied to five estuaries within the Northern Gulf of Mexico: Barataria, Galveston, Mobile, and Perdido Bays and Mississippi Sound. Initial results from this work show that: 1. Sub-regions can be identified as associated to each other based upon indicator data values and not upon physical location. 2. Even though the weights calculated for the management protocols vary significantly, for systems that were not highly homogeneous in indicator data values, the different weights did not produce the vastly different cluster maps expected. 3. The scale work indicates that to identify appropriate sub-regions using the developed framework, a larger grid size produces more consistent results for larger systems whereas a smaller grid size produces more consistent results for smaller systems. Recommendations for further research are also presented.