This study examined relationships between simulated logging disturbances and growth and spread rates of roadside cogongrass patches into adjacent forest stands. Effects of varying levels of imazapyr and surfactant were evaluated to assess cogongrass control. Determination of the optimal cogongrass treatment regime along with the monetary impacts of cogongrass on Land Expectation Values (LEV) are analyzed and discussed. Growth and spread rates of roadside cogongrass plots one growing season after simulated logging disturbance were significantly higher than undisturbed control plots. Rhizome biomass and level of disturbance were primary causal factors. Arsenal AC at 3.5 oz, Accord Concentrate at 14.7 oz, and 11.8 oz of surfactant applied in a volume of 35 gallons per acre provided higher levels of cogongrass control (76%) two years after treatment. The monetary cost of cogongrass was $69 per acre in after tax LEV with assumption of no decreases in seedling mortality or growth and yield.

Cogongrass management generally requires multiple herbicide applications, however, success is limited if not integrated with other methods. Experiments were conducted to evaluate the use of cover cropping systems with herbicides on cogongrass control. Field studies determined that sequential glyphosate applications in the summer were necessary to achieve 80% or greater control, but a single application could be effective if weather conditions allowed early planting and good cover crop establishment of Roundup Ready soybeans. Studies also indicated that the use of ALS-resistant Italian ryegrass and white clover crop combinations showed no effect, but imazapyr applications made in May or June provided 80% or higher control by October. Greenhouse experiments showed that delayed planting at least 1 month after imazapyr preemergence applications from 70 to 280 g ae ha-1, significantly reduced emergence failure, height and biomass reductions of legumes used for revegetation.

Biological invasions are a significant area of research due to perceived negative environmental and economic impacts. The study of biological invasions has identified three broad components involved in successful invasions: propagule pressure, abiotic and biotic conditions. Propagule pressure is the product of introduced propagules and the frequency of those introductions, and is considered a driver in all stages of invasion. Data to quantify propagule pressure is often unavailable, and therefore, this research approached genetic information to produce estimates of propagule pressure in the successful invasion of cogongrass (Imperata cylindrica) in the United States. The following research utilized molecular methodologies to estimate genetic diversity and to infer historical introductions. Population genetic analyses were conducted for the purpose of estimating extant population-level genetic diversity at multiple-scales to first address documented cogongrass introduction(s) into this country, followed by an exploration of substantial range expansion into seven states across the region. Lastly, this research explicitly seeks evidence to support interspecific hybridization between cogongrass and a co-occurring congeneric (Imperata brasiliensis) having occurred. Aggressive range expansion by cogongrass has been attributed to such a hybridization event, but has not yet been specifically tested. Historical accounts of foreign introduction of cogongrass propagules identify two separate introductions of distinct source material made into Mississippi and Alabama. Localized population genetic analysis of these two states found substantial genetic variability within and among cogongrass populations, and supported a two-introduction scenario of distinct genetic source propagules establishing and subsequently intermixing. Enlarging the geographic scale of study incorporated five additional U.S. states currently experiencing and managing cogongrass invasion. Considerable genetic variability was found within and among the seven states surveyed. It was found that range expansion was unequal across the range, and that the most distant states tested were not genetically isolated from source populations, suggesting a possible anthropogenic role. Focusing solely on Florida where two congeners overlap ranges, morphology, and ecology, this population genetic analysis failed to detect significant evidence to support interspecific hybridization. Collectively, these investigations explored genome-level dynamics during invasion by a noteworthy invasive grass in an effort to better understand the process of biological invasions.

The purpose of this research was to increase understanding of ecology and control of cogongrass (Imperata cylindrica (L.) Beauv.) to assist with management of this invasive grass. To address different aspects of cogongrass management, I examined factors that affect spread of cogongrass, effects of cogongrass on native plant communities, and use of three native species to suppress cogongrass. Relative susceptibility of pine-bluestem and pine-shrub vegetation associations to vegetative encroachment and seed dispersal of cogongrass were evaluated. Vegetative encroachment into burned and unburned areas of these two vegetation associations also was measured. Effects of infestation size on vegetative growth of cogongrass and relationships between military activity and establishment and growth of cogongrass on military firing points were investigated. I also compared frequency of cogongrass infestation and vegetative growth rates for unpaved roads with different levels of traffic and maintenance. Native plant richness, diversity, and abundance were compared between cogongrass infestations and in uninfested adjacent areas for longleaf pine-bluestem and ruderal, cleared areas. Morella cerifera (L.) Small, Chamaecrista fasciculata (Michx.) Greene, and Panicum virgatum L. were tested for their ability to suppress cogongrass re-growth after treatment with a herbicide. Mean vegetative encroachment of cogongrass was < 2 m/yr for both habitat types regardless of burning. Vegetative encroachment was more than double in burned plots compared to unburned plots. Spikelets of cogongrass consistently dispersed farther into pine-bluestem (x¯ =17.3 m) forests compared to pine-shrub forests (x¯ = 9.4 m). Vegetative encroachment was not affected by size of infestation for cogongrass patches on firing points (x¯ < 1 m/yr) (P > 0.643). In areas of soil disturbance from military equipment, vegetative expansion rates of 7-10 m/yr were recorded. There was a positive relationship between military troop use and cogongrass establishment for one of the 2 years of the study (P = 0.023). Growth and establishment of cogongrass did not vary for unpaved roads with differing levels of maintenance and traffic (P > 0.173). Species diversity and abundance of herbaceous vegetation was less in cogongrass infestations compared to uninfested adjacent areas. Complete suppression of cogongrass was not evident for any of three native species tested.

Of concern in this research were the ecological parameters associated with the establishment of a model invasive plant species, Imperata cylindrica, across a scale of ecological organization. Specifically, the study addressed the species’ ability to: differentially respond to abiotic and biotic constraints during seedling establishment, exhibit a novel underground competitive interference mechanism, and alter the decomposition dynamics in newly invaded ecosystems. Finally, the last portion of the research was centered around creating a predictive habitat model that will provide information on the most important variables responsible for creating habitat for this species. The population level seedling study indicated that soil characteristics and light availability play a significant role in seedling establishment. There were large trends in biomass allocation attributable to soil type with seedlings performing best in high nutrient soils representative of the Mississippi Alluvial Valley physiographic region. I. cylindrica seedlings also showed a positive response to increased seedling density during the initial stages of seedling establishment. The community level research examining a hypothesized novel interference mechanism deployed by I. cylindrica showed a significant and robust pattern of I. cylindrica damaging its own belowground tissue more often than that of its surrounding neighbors. Therefore, it is highly unlikely that I. cylindrica gains a competitive advantage by exposing the native plant assemblage to pathogen invasion (via ruptured tissue) as the plant would expose itself to these pathogens (to which it is evolutionarily naive) at much higher volumes. The ecosystem level examination of this globally important invasive species indicated that I. cylindrica invasion into native systems will significantly accelerate ambient rates of decomposition. Furthermore, fungal community composition in invaded areas was drastically altered as well as bacterial community functional activity in relation to several key enzymes responsible for the decomposition of plant tissue which were produced more abundantly in invaded areas.The landscape-scale analyses and modeling work validated decades of anecdotal evidence and indicated that anthropogenic disturbance factors associated with road maintenance and construction (soil disturbance and vegetation removal) are the principal factors responsible for creating habitat suitable for invasion by this species.