THE ECOLOGY OF AN INVASIVE TOXIGENIC PROTIST

Humans have facilitated a spread of nonnative biota across the Earth. As nonnative species, their interactions with native species, their surrounding community, and their environment are novel, and thereby provide unique opportunities for studying community assembly and the maintenance of community diversity. The study of nonnative species is also important because some experience population explosions and spread rapidly throughout a new habitat and cause negative environmental and economic impacts. Such nonnative species are often referred to as invasive species.Freshwater systems are among the most impacted environments by invasive species because they are some of the most disturbed ecosystems on the planet. Disturbances to freshwater ecosystems include impoundments in riverine networks and increased anthropogenic nutrient pollution (i.e., cultural eutrophication). These disturbances facilitate the establishment and spread of invasive species by changing resource availability and altering species interactions, which creates openings for invasive species.Another result of these disturbances in freshwater ecosystems is the worldwide proliferation and establishment of harmful algal blooms (HABs). Harmful algae are planktonic microbes from multiple taxonomic groups that can produce toxins. Similar to invasive species, these species are infamous for being able to dominate their ecosystems at high densities and produce devastating ecosystem and economic effects. Much of the research on harmful algae ecology focuses on the prevention and control of bloom formation or toxin production, but fundamental questions of community ecology regarding dispersal, distribution, and ultimately establishment of these microbial invasives have received little consideration.This lack of attention may be due to the fact that microbial species in general are neglected in the invasion ecology literature. Most concepts and hypotheses in invasion ecology have been derived from the study of macrobial multi-cellular organisms. That microbes have historically not been considered within the field of invasive ecology is at least partially explained by one of the prevailing hypotheses of microbial community ecology: “everything is everywhere, but, the environment selects.” This hypothesis suggests that dispersal is not limiting for microbes and thus microbes may not be invasive at all, but rather present in a given environment at low densities at any given time. One implication of this hypothesis for microbial invasion ecology is that our ability to discern a microbial invasion is limited by our ability to detect it. However, another implication of the “everything is everywhere, but, the environment selects” hypothesis is that the environment is the primary determinant of whether a given microbe is abundant enough to be detected (i.e., microbial establishment). Because relatively high detection limits (which are continually being reduced as molecular technology advances) and a lack of traditional taxonomy has severely limited the study of microbial invasives, there are numerous questions that need to be addressed in order to understand the factors governing invasion and successful establishment of microbial species.The harmful algal species, Prymnesium parvum provides an excellent opportunity for ecologists to gain insight into the factors responsible for the successful establishment of invasive microbes. Prymnesium parvum is a toxigenic protist in the class Prymnesiophyceae. It was first characterized from marine systems, but has since invaded freshwater systems worldwide. It is notorious for its negative impacts on ecosystems, chief among them, the killing of fish. Fish affected by P. parvum toxins exhibit hemorrhaging of the gills, and P. parvum blooms typically cause massive fish kills. Prymnesium parvum blooms are also considered to be Ecosystem Disruptive Algal Blooms (EDABs) characterized by their adverse direct effects on fishes and herbivorous invertebrate grazers, as well as their indirect effects on nutrient and food web dynamics, which create feedbacks enabling bloom persistence. In North America, the first record of a P. parvum bloom is from a fish kill in 1985 in the Pecos River system of southern Texas, USA. During the subsequent two decades, P. parvum has expanded its range and bloomed and caused fish kills in reservoirs and rivers throughout much of the southern United States from California to Florida, and as far north as Wyoming and West Virginia. Numerous studies, in particular laboratory experiments, have been conducted into the autecology of P. parvum that suggest attributes that might make it a good invader, although many of these have not been put into a natural context. Similarly, much is known about the distribution of its blooms, although the factors responsible for invasion of a system and subsequent bloom development are not well understood and have not been directly addressed.In 2004, P. parvum invaded and bloomed in Lake Texoma, OK-TX, USA causing a massive fish kill in many of the shallow areas of the reservoir. Including this first bloom, P. parvum has bloomed during winter in seven of the nine winters through 2012. This created a unique opportunity for me to use P. parvum to investigate questions in invasion ecology using a microbial invader, specifically questions about microbial establishment. In my dissertation, I address factors important to understanding the success of P. parvum invading new systems and the consequences of a P. parvum bloom on the fish community. In my first chapter, I modified a recently developed method for detecting P. parvum using quantitative polymerase chain reaction (qPCR) and tested its efficacy as an alternative to microscopy for P. parvum detection and enumeration in a long-term monitoring program in Lake Texoma. Abundance estimates of P. parvum were similar for both methods, but I detected P. parvum at multiple sites using qPCR where it previously had gone undetected by microscopy. Using qPCR, I substantially reduced processing time, increased detection limit and reduced error in P. parvum abundance estimates compared to microscopy. Thus, qPCR is an effective tool for detecting and monitoring P. parvum, particularly at pre-bloom densities, and should likewise prove useful in monitoring programs for the other HAB species for which qPCR methods have been developed.In chapter two, I sampled fish near and offshore over an annual cycle encompassing a P. parvum EDAB event in two coves (i.e., a bloom site and a reference site) of Lake Texoma. My objective was to document the processes of extirpation and recovery of a fish assemblage to the disturbance of an EDAB event. Prymnesium parvum bloomed in one cove from mid-December 2008 until May 2009, eliminating all fish during this period. Fish toxicity bioassays indicated no substantial differences in susceptibility across fish species to P. parvum toxins. Fish rapidly recolonized the bloom site in May 2009 after the P. parvum bloom diminished. Fish assemblages were resilient to the P. parvum EDAB, recovering to previous abundance, richness, and composition within six months. My results suggest that the reservoir-wide fish metacommunity enabled a rapid recovery of local fish assemblages following a spatially heterogeneous EDAB event.In chapter three, I used a four-year data set from an ongoing monitoring program in Lake Texoma (OK-TX) to construct a predictive model relating P. parvum presence or absence to environmental parameters at a local scale. I then tested this model at the regional scale in conjunction with environmental sampling to predict presence and absence of P. parvum in the watershed of the Red River, one of two tributaries to Lake Texoma and a neighboring watershed, the Canadian River, as well as a few sites in the Arkansas River watershed. Based on three environmental factors, specific conductance, total nitrogen, and total nitrogen : total phosphorus ratio, my predictive model accurately classified P. parvum as present or absent in Lake Texoma for 74% of the samples. Applying this model to the adjacent watersheds also showed strong predictive power, correctly classifying 87% of the sites sampled within the Red River watershed and 81% of the sites sampled in the Canadian River watershed. Sites where the model predicted P. parvum but none was detected may be particularly vulnerable to P. parvum establishment and should be more closely monitored for future invasion success. Misclassifications by the model of sites in which P. parvum was detected suggests that dispersal has occurred, but that the environmental conditions were not conducive to population establishment. Indeed, at these sites, P. parvum abundances, when detected, were low. While this study cannot rule out dispersal limitation as a major factor involved in the biogeography of P. parvum, my results do indicate that the establishment and spread of this harmful algal species appears to be limited by environmental conditions in the invaded habitat.In chapter four, I experimentally assessed the roles of community resistance and propagule pressure (i.e., the number of invaders entering a habitat or the frequency of invasions to a habitat) on the establishment success of P. parvum. One of the seminal hypotheses in the field of invasion ecology is that more diverse communities should be more resistant to invasion by exotic or non-native species. Similarly, propagule pressure is thought to facilitate an invasive species' establishment success by increasing the ability of an invading population to absorb the challenges of its new environment. Propagule pressure and community resistance to invasion are predicted to interact to affect the probability of community invasion and species establishment in a new community. I show experimentally that regardless of community diversity, establishment success by the microbial invader, Prymnesium parvum in an environmentally-compatible habitat, is determined by propagule pressure.Currently, there are two theoretical possibilities that might explain the rapid range expansion of P. parvum. One possibility is that P. parvum is an invasive species that has dispersed to and established in new ecosystems. The second possibility is that this range expansion is driven by changes in the environment, particularly relating to salinity as affected by climate change and water resource overexploitation, and that P. parvum has always been present in these systems (i.e., “everything is everywhere, but, the environment selects”). My dissertation adds to the literature by showing that indeed both dispersal and environmental selection can play a role in P. parvum establishment.The results from my dissertation are only a starting point for continuing to ask questions about the importance of dispersal and the environment to the outcomes of microbial invasion. Further investigation into the roles of environmental filtering, and the possibility of prior presence of P. parvum should prove to be fruitful avenues of future research. Furthermore, understanding how the environment and P. parvum affect community composition should allow us to gain insight into how P. parvum interacts with other microbes. Ultimately, all of this information is important not only to our understanding of P. parvum specifically, but to our understanding of microbial invasions and harmful algal blooms, in general, and thus will contribute to our ability to mitigate the impacts of HABs and EDABs