Ailanthus altissima (Mill.) Swingle is a tree native to Asia that was intentionally introduced into the United States in the late eighteenth century. Ailanthus altissima has become an invasive species that has spread throughout most of North America. Lack of effective management tactics for suppression of A. altissima has lead to alternate control methods. Investigations into using biological control with the weevil Eucryptorrhynchus brandti were initiated in 2004. Studies were conducted to understand the general biology of E. brandti, rearing efficacy, and host specificity. Eucryptorrhynchus brandti is univoltine, has a life cycle similar to the closely related species Cryptorhynchus lapathi (L.), with 6 instars, and completes development in 126 ± 6.5 d at 25°C. Efficient egg to adult rearing was accomplished by caging 12 m and 12 f for 7 days on 23 – 92 cm long billets. Males and females can be differentiated by the structure of the metathoracic sternite and 1st abdominal segment. Host specificity experiments show that E. brandti preferentially feeds on North American A. altissima when tested against 29 species from 14 families. Larval development in the rare species Leitneria floridana Chapm. was apparent. Additional studies show that A. altissima does not occur across L. floridana distribution but may have the potential to invade L. floridana sites. / Ph. D.

Peatlands are a diverse type of wetland ecosystem, characterized by high levels of soil organic matter, that provide a wide array of ecosystem services including water storage and filtration, carbon sequestration, and unique habitats. Draining peatlands degrades their resilience to future disturbances, notably including high intensity, soil-consuming fires. Peat soil fires are unique in that they can smolder vertically through the soil column, with consequences ranging from large carbon emissions to altered hydrology and dramatic shifts in vegetation communities. In this work we had two complementary objectives to understand both the drivers and impacts of smoldering fires at the Great Dismal Swamp (VA and NC, USA). First, we developed and verified a new method to model peat burn depths with readily available water level and peat hydraulic property data. Our findings suggest that drainage weakens both short- and long-term controls on peat burn depths by reducing soil moisture and by decreasing peat water holding capacity. To address the impacts of smoldering fires, we quantified the abundance of the noxious Phragmites australis in a large fire scar and the extent to which altered hydrology influenced its occurrence. We did so by leveraging satellite imagery, random forest models, LiDAR data, and water table observations. Our results suggest that P. australis is aided by a hydrologic regime generated, in part, from the combined effects of drainage and deep smoldering fires. Our conclusions from these two studies contribute to the scientific understanding of smoldering peat fires and can inform management efforts. / Master of Science / Peatlands are a diverse type of wetland ecosystem that have characteristically thick levels of organic-rich soil, known as peat. Peatlands are home to a variety of unique plants and animals, store large amounts of carbon, and provide water storage functions. Peatlands were historically drained to enable development and conversion to other land usages, which had many unintended consequences like increasing their risk to wildfires that consume soil organic matter. An intense peat fire can smolder down through the peat, with impacts ranging from large releases of carbon to changes in water levels and vegetation communities. In this work we had two objectives aimed at understanding the drivers and impacts of smoldering peat fires in the Great Dismal Swamp (GDS) (VA and NC, USA). First, we developed and verified a new method of modeling how deep peat fires burn by using readily available water level and soil property data. Our findings suggest that drainage weakens both the short- and long-term controls on peat fire burn depths by reducing soil moisture and by limiting the ability of peats to hold water. We also studied how water levels in a post-peat consuming fire environment influence the amount of the weedy Phragmites australis. We did so by using satellite imagery, elevation data, and water table observations. Results from this investigation suggest that the combined effects of drainage and deep smoldering fires help to create ideal conditions for P. australis invasion and establishment. Our findings from these two studies add to the scientific understanding of smoldering peat fires and may inform land management decisions.

Invasive species often encounter climatic conditions that differ significantly from those of their native range. These environmental shifts may trigger phenotypic responses, resulting through some combination of adaptation and plasticity, that enable the invader to persist under novel thermal regimes. In this dissertation, I examine phenotypic changes in a tropical lizard that has successful invaded a cooler temperate climate, specifically examining traits that may promote survival and reproduction in their new range. First, I examined physiological traits, as I predicted greater cold tolerance would be necessary to survival in the invasive range. I found that invasive populations tolerated lower temperatures, exhibited greater maximum sprint speeds, and had higher metabolic rates than native populations. Next, I examined how life-history traits may change in the invasive range in order to facilitate reproduction under shorter breeding and growing seasons. I found that compared to native females, invasive females had shorter interlaying intervals and produced eggs that hatched more quickly. Once I quantified changes physiological and life-history traits that may have aided in successful establishment, I executed a common garden study to determine whether changes were the result of adaptation or plasticity. I found that differences in critical thermal minimum, metabolic rate, interlaying interval, and incubation period were maintained in lab-reared offspring, while measures of sprint speed converged. My results provide evidence that life history and physiology can evolve rapidly during invasion. These findings are useful to understanding contemporary evolution, and also provide valuable insight on how species respond to environmental shifts, both during invasions and as a result of climate change. / Doctor of Philosophy / When species invade a new area, they often face different climates that make can make survival and reproduction challenging. In response, species may alter traits in order to adjust to new temperatures and conditions. In this dissertation, I examine trait changes in a tropical lizard that has successfully invaded a cooler temperate climate, specifically examining traits that may help them to survive and reproduce in their new range. First, I examined physiological traits, as I predicted greater cold tolerance would be necessary to survival in the invasive range. I found that invasive populations tolerated lower temperatures, could sprint faster, and had higher metabolism than native populations. Next, I examined how reproductive traits may change in the invasive range in order to facilitate reproduction under shorter breeding and growing seasons. I found that compared to native females, invasive females had less time between egg lays and produced eggs that hatched more quickly. Once I assessed how traits may have changed in the new range, I determined whether changes resulted from evolution or not. I found that differences in low temperature tolerance, metabolic rate, the time between egg lays, and incubation period were the result of evolution, while sprint speed did not seem to be the result of evolution. My results provide evidence that traits can evolve rapidly during invasion, allowing invasive species to persist and spread in new areas.

Invasive species are a growing issue that will compound under climate change. Rising temperatures, fluctuating precipitation and new transportation pathways will create new opportunities for invasive establishment. A direct and impactful consequence of climate change is the removal of climatic barriers to invasive survival. Species distribution modeling (SDM) for invasives must include an evaluation of future establishment potential so that managers can prioritize regions forecasted as high risk under climate change. Climatic SDMs effectively support pan-invasion risk assessments by forecasting potential invaded areas globally where climatic barriers have shifted the potential for establishment. Rarely is regional-scale climate variation considered in invasive SDMs, despite its relevance for pests that establish outside their native regional climate. Here, I apply a climatic pest risk framework to the Spotted Lanternfly grape pest (Lycorma delicatula, SLF). I assess how climate change shifts the establishment potential of SLF across important viticultural regions worldwide. I contrast an ensemble of three regional-scale SDMs to a global-scale SDM, which provided multiple predictions on how future regional climate variation might shift national SLF risk levels, impacting the global wine market. I found that the global suitable area for SLF will increase under climate change, with range expansion outpacing contractions by about 1.1 million km2. Expansions will primarily occur at present northern range edges in Europe, North America, and East Asia, and contractions will occur across the southern hemisphere. Next, 307 global viticultural regions (29% of 1,063 total sampled) will decrease in risk for SLF establishment and only 532 (50% of 1,063) will remain at any risk under climate change. Loss in SLF establishment risk under climate change followed a latitudinal gradient in the northern hemisphere. Meanwhile, only 85 known SLF populations (11% of 769 rarefied sample) will destabilize under climate change. Populations within the US and South Korean invaded ranges will remain stable with respect to climate. Our regional-scale ensemble emphasized the importance of mean winter temperature as a constraint on SLF establishment, with activity dropping sharply at -3°C. This method for regional-scale ensemble modeling should be utilized in similar invasive or climate change SDM applications to make more refined SDM predictions and to reduce uncertainty. Viticulturalists can and should use our provided tools and model framework to understand the risk of SLF establishment at their locality as climate change removes barriers to this pest’s establishment globally. / Biology

1.) Sentinel traps were used to survey for parasitoids of frugivorous drosophilids in Virginia fruit cropping systems, and determine if parasitoids were attacking invasive flies Drosophila suzukii (Matsumura) and Zaprionus indianus Gupta (Diptera: Drosophilidae) in the field. Two parasitoids of frugivorous drosophilids, Leptopilina boulardi (Barbotin, Carton, and Kelner-Pillault) (Hymenoptera: Figitidae) and Pachycrepoideus vindemiae (Rondani) (Hymenoptera: Pteromalidae), were reared, but only one P. vindemiae was reared from D. suzukii, and no parasitoids were reared from Z. indianus. Most parasitoids were reared from alternate host Drosophila melanogaster Meigen and other wild drosophilids. 2.) The ability of these parasitoids to attack D. melanogaster, D. suzukii and Z. indianus under controlled conditions was tested. Larval parasitoid L. boulardi did not develop on D. suzukii or Z. indianus, just D. melanogaster. Pupal parasitoid P. vindemiae successfully developed on all three fly species, but also increased pupal fly mortality. 3.) Olfactometry was used to ascertain if L. boulardi and P. vindemiae are selective about the type of fruit their hosts feed in. Results showed that among cherry, raspberry, blueberry, grape, and banana, L. boulardi preferred raspberry and banana to cherry, and preferred grape least, but no fruit was most preferred. Insufficient data were obtained for P. vindemiae. We conclude that parasitoids of Virginia are unlikely to provide effective biological control for D. suzukii or Z. indianus, and classical biological control should be investigated as a pest management option. Olfactometry results indicate tritrophic selectivity by Drosophila parasitoids, suggesting multiple parasitoids could be required for effective biological control. / Master of Science in Life Sciences / The spotted wing drosophila (henceforth SWD) is a globally invasive vinegar fly originating from southeast Asia, and is economically damaging to producers of small fruit, such as berries, cherries, and grapes. The African fig fly (henceforth AFF), a relative of SWD, is another recently invasive species to North America (originating from Africa) that often occurs simultaneously with SWD infestations, but its economic threat to North America is still unclear. With the economic threat posed by SWD, and potential threat posed by AFF, it is important to understand their relationships with other organisms in their environment, especially natural enemies, such as parasitoid wasps. Such information is integral for researchers to develop effective control methods, and will help determine if natural enemies can be used to our advantage as biological control agents. Biological control also helps to limit the use of chemical insecticides, mitigating the development of insecticide resistance in the pests. This project employed unique field trapping methods and laboratory bioassays to investigate the relationships of SWD and AFF with parasitoid wasps in affected fruit cropping systems in southwestern Virginia. We discovered that parasitoids of vinegar flies are present in Virginia fruit cropping systems, but they do not help to control populations of SWD and AFF. The parasitoids that are present prefer to attack other fly species, are unable to attack SWD and AFF, or do not attack in high enough numbers to have an impact on SWD or AFF populations. Biological control success is more likely to come from parasitoid species that have co-evolved with SWD and AFF in their native ranges.

付記する学位プログラム名: 社会を駆動するプラットフォーム学卓越大学院プログラム / 京都大学 / 新制・課程博士 / 博士(農学) / 甲第25317号 / 農博第2583号 / 新制||農||1104(附属図書館) / 学位論文||R6||N5489 / 京都大学大学院農学研究科森林科学専攻 / (主査)教授 北島 薫, 教授 小野田 雄介, 教授 井鷺 裕司 / 学位規則第4条第1項該当 / Doctor of Agricultural Science / Kyoto University / DGAM

Invasive species are often both ecologically and economically detrimental, particularly in freshwater ecosystems where direct and indirect impacts of invasion interact with pre-existing anthropogenic stressors to magnify consequences for native communities and habitats. Planktivorous silver carp (Hypophthalmichthys molitrix) are invasive in the Mississippi River watershed, where their daily consumption of detrital seston can total more than half their body mass, and where they outnumber all native species combined in some river reaches. Because of these astonishing abilities, silver carp can cause basal resource changes, altering the body sizes and abundances of native plankton communities upon invasion. Much research has focused on this outcome of silver carp proliferation, but investigation into other effects of their presence in invaded systems has not received the same attention. Despite the unique tendency of silver carp to experience species-specific mass-mortality events, for example, there is a dearth of peer-reviewed evaluation of the possible ecological consequences of these occurrences. To elucidate the effects of these events which can comprise hundreds of thousands of individuals, we determined how the decomposition of silver carp carcasses following mass-mortality events affected benthic invertebrate community composition and productivity using a mesocosm approach. Each mesocosm was subjected to either the presence or absence of a silver carp carcass over the length of an eight-week experiment. Carcass presence created anoxic conditions in the mesocosms and acutely increased benthic invertebrate biomass and benthic periphyton growth. While acute, changes to benthic conditions following silver carp mass-mortality events are likely to have long-term effects in both invaded freshwater systems and their surrounding terrestrial landscapes. Given the magnitude of die-offs, it is possible that these events could cause ecosystem-level alterations. Further research, therefore, is needed into the additional impacts that may follow carcass decomposition to quantify and predict outcomes associated with continued silver carp invasion.

Grass Carp (Ctenopharyngodon idella) is a species of concern in areas they have invaded, such the Laurentian Great Lakes, due to their potential to substantially reduce aquatic macrophyte coverage. Efforts to control Grass Carp in the Great Lakes have primarily been focused on two tributaries to the western basin of Lake Erie (i.e., Sandusky and Maumee rivers, Ohio) where reproduction has been documented, although other Great Lakes tributaries are known to have thermal and hydrological regimes suitable for Grass Carp spawning. Knowledge of sources contributing to the expanding population of invasive Grass Carp in the Great Lakes is key to allocating control efforts aimed at curbing further introductions, reducing natural recruitment, and limiting potential for further range expansion. A recent study demonstrated that otolith microchemistry is an effective tool for identifying the natal environment of Grass Carp in the Great Lakes. Increased captures of Grass Carp in Lake Erie and Lake Michigan indicates an ongoing need to determine which tributaries are supporting Grass Carp recruitment. Therefore, the objectives of this study were to use otolith stable oxygen isotope (δ18O) analysis to determine whether diploid and unknown ploidy Grass Carp collected from the Great Lakes during 2019-2022 were wild or of aquaculture origin, analyze Sr:Ca and Ba:Ca ratios of water samples collected from known and potential Grass Carp spawning tributaries to assess persistence of differences in water chemistry among tributaries observed in prior studies, use otolith core trace element ratios (Sr:Ca and Br:Ca) to infer natal rivers of wild Grass Carp, and estimate how many groups of aquaculture-origin Grass Carp (both diploid and triploid individuals) with distinct otolith chemistry profiles were present among fish collected during 2019-2022. Water Sr:Ca and Ba:Ca for Great Lakes tributaries were consistent with data from prior studies. Diploid and unknown ploidy Grass Carp (21%) were identified as aquaculture origin fish based on otolith core δ18O. Multiple clusters and broad ranges of otolith core Sr:Ca and Ba:Ca among aquaculture-source Grass Carp suggest multiple sources of introduced/escaped fish in the Lake Erie basin. Tributaries to the western basin of Lake Erie were identified as the primary sources of wild Grass Carp, although there was some evidence of recruitment from central or eastern basin tributaries to Lake Erie. There was no evidence of Grass Carp reproduction in the Lake Michigan basin; the one wild fish caught in the Lake Michigan basin had otolith core Sr:Ca consistent with origin in a western basin tributary of Lake Erie. Thus, efforts to control natural recruitment of Grass Carp should remain focused on tributaries to the western basin of Lake Erie, especially where reproduction has been documented. However, the relatively high percentage of aquaculture-source Grass Carp (including some fertile, diploid fish) captured during multi-agency response efforts indicates that curtailing natural recruitment, further introductions, and spread of Grass Carp is necessary for successful population control.

The dissertation consists of three essays in forestry resource management, with focuses on investigating the ecosystem collapse and invasive species control problems. The first two papers consider the shift from primary forests to cleared land and secondary forests in the context of deforestation. This process is known to lead to irreversible tipping points that lead to the loss of ecosystem services. The past literature has discussed forest rotations under stochastic prices, timber volume, and amenity values. I extend this body of work to show how stochastic processes concerning primary forests could lead to ecological collapse. Drift and volatility in these processes explain different types of long-term and short-term shocks in tropical forest systems, such as fire, drought, or climate changes, all mechanisms that can drive ecosystem function to collapse. Common examples of severe ecosystem damage include the irreversible change from tropical forests to grassy savanna, fire events, and other climate problems. However, another case of uncertainty happens when ecosystem service production of primary and secondary forests itself is stochastic, so that there is a more complicated relationship between deforestation and reaching a point where ecosystem functions collapse. I compare and contrast these two cases to determine how drift and volatility determines the timing of a tipping point in a deforestation model where primary and secondary forests, as well as agricultural land, influence ecosystem function. I examine the sensitivity of the timing of collapse in both model variants to critical market and land-use parameters. The third chapter of this dissertation explores the connection between landowners' risk preferences, invasive species spread, and optimal control efforts. This study analyzes the control effort involved in neighboring infested and uninfested municipalities, which may have differing risk preferences. In the context of an application to the spread of Emerald ash borer (EAB) in the Twin Cities, Minnesota, I develop a simulation to explore the level of control and spread in a myopic policy scenario versus that in a first-best problem, where the two municipalities may be either risk neutral or risk averse. The results suggest that heterogeneity in risk aversion across the municipalities leads to lower control efforts and a longer time to drive the probability of spread to zero. / Doctor of Philosophy / This dissertation presents three studies that aim to support policy making for sustainable forest resource management. In the first two chapters, I examine the deforestation of native tropical forests, which damages ecosystems and leads to losses that can be irreversible, even with investments in reforestation. At present, the Amazon, Indonesia, and Africa are all witnessing the catastrophic collapse of forest systems as a result of deforestation. Prior work by economists on this topic has not considered how uncertainty in native forest losses may affect the likelihood of ecosystem collapse. Native forest losses may be uncertain when policymakers cannot fully control deforestation (as is often the case in developing countries) or when complex ecosystems are not well understood. I develop a model that incorporates long-term and short-term shocks in forest systems, such as fire, drought, or climate changes, all mechanisms that can drive the primary forest stock function to ecosystem collapse. Using this model, I examine the sensitivity of the timing of collapse to critical market and land-use parameters. In the final chapter of the dissertation, I present a novel model to study how risk preferences affect the management of invasive species. The model is inspired by the Emerald Ash Borer (EAB) infestation in the Twin Cities, Minnesota, which spread from St. Paul into Minneapolis and decimated valuable urban tree cover over the past several decades. I use the model to explore how differences in risk preferences between neighboring municipalities affect the control efforts they undertake, the probability of pest spread, and the consequent economic losses.

ABSTRACT Alliaria petiolata (garlic mustard) is an aggressive non-native and invasive forb that negatively impacts native arbuscular mycorrhizal communities and inhibits or prevents the growth of native plant species. Invasive species mitigation and management strategies that use native revegetation vary in success. This study focuses on which species naturally regenerate in areas where A. petiolata has been mitigated to help inform restoration efforts. Seedling emergence of species within the seed bank of four plot types (uninvaded, invaded, chemically treated, and mechanically treated) were observed two years post restoration efforts to determine which native species are likely to persist to seedlings following management. Species abundance and percent cover of all plant species were recorded to evaluate the presence of species in addition to plant health and physiological differences. Native species abundance was significantly reduced within herbicide treated plots as compared to untreated and invaded plots (p=0.02). Plots treated with mechanical removal had the greatest percent cover of native plants as compared to all other plot types but were dominated by colonizing species which is typical of a disturbed habitat. Mechanical removal also resulted in a greater abundance and stability among functional groups of native species, than those treated with herbicide. Forb species dominated coverage of plots over other functional groups when treated with herbicide. Alliaria petiolata invasion and management methods significantly impacted forb and graminoid species, as they had significantly lower abundance in plots treated with herbicide. The results demonstrate that the method of removal as well as the presence of A. petiolata affects emergence of plant species from the seedbank. The additional disturbance of mechanical removal may alter successional trajectories following invasion. Herbicide treatment resulted in the most similar species abundance as the uninvaded reference plots, which had the lowest seedling emergence and percent cover.