Unmanned aerial vehicles (UAVs) with high-resolution cameras are common in today’s society. Industries, such as the forestry industry, use drones to get a fast overview of tree populations. More advanced sensors, such as near-infrared light or depth data, can increase the amount of information that UAV images provide, providing information about the forest, such as; tree quantity or forest health. However, the fast-expanding field of deep learning could help expand the information acquired using only RGB cameras. Three deep learning models, FasterR-CNN, RetinaNet, and YOLOR were compared to investigate this. It was also investigated if initializing the models using transfer learning from the MS COCO dataset could increase the performance of the models. The dataset used was Swedish Forest Agency (2021): Forest Damages-Spruce Bark Beetle 1.0 National Forest Data Lab and drone images provided by IT-Bolaget Per & Per. The deep learning models were to detect five different tree species; spruce, pine, birch, aspen, and others. The results show potential for the usage of deep learning to detect tree species in images from UAVs. / Obemannade drönare med högupplösta kameror är vanliga i dagens samhälle. Branscher, så som skogsindustrin, kan använda sig av sådana drönare för att få en snabb översikt över ett skogsområde.Mer avancerade sensorer, som använder nära-infrarött ljus eller djupdata, kan öka mängden information som drönarna kan samla in, information såsom; trädmängd eller data om skogens hälsa. Det snabbt växande området djup-maskinlärning kan dock hjälpa till att utöka informationen som kan extraheras vid användning av endast RGB-kameror. Tre modeller för djupinlärning, Faster R-CNN, RetinaNet och YOLOR, jämfördes för att undersöka detta. Det undersöktes också om initiering med för-tränade vikter, med överföringsinlärning från datasetet MS COCO, skulle kunna öka modellernas prestanda. Datasetet som användes var Skogsstyrelsen (2021): Skogsskador-Granbarkborre1.0 Nationell Forest Data Lab samt drönarbilder tillhandahållna av IT-Bolaget Per & Per. Det tredjupinlärnings-modellerna skulle detektera fem olika trädarter: gran, tall, björk, asp, och övrigt.Resultaten visar potential för användning av djupinlärning för att upptäcka trädarter i bilder från drönare.

For this dissertation, I measured how plant communities in Appalachian forests responded to disturbances caused by forest management activities. I had two primary objectives including (1) testing theories of biological diversity and invasions by nonnative species; and (2) providing empirical data that will help guide the sustainable use of forest resources. This work is part of the Southern Appalachian Silviculture and Biodiversity (SASAB) experiment that was established in the early 1990s to investigate ecosystem responses to a gradient of timber harvesting disturbances. Ranging from undisturbed controls to silvicultural clearcuts, the disturbance gradient is replicated at sites located throughout the Appalachian mountains of Virginia and West Virginia. The plant community was sampled across a wide range of spatial scales (2 hectares to 1 m2) using a nested sampling design and was also sampled at three times including pre-disturbance, one year post-disturbance, and ten year post-disturbance. For one element of the study I tested modern theories of biological invasions and investigated how the forest disturbance gradient interacted with species diversity to influence invasion by nonnative species (Chapter 2). Contrary to popular ecological theories of biotic resistance, the most diverse sites tended to be more easily invaded following intense canopy disturbance. Interestingly, none of the sites in this study were dominated by invasive plants, which led me to ask whether dominant tree species in forests provided resistance to nonnative plant establishment and growth through the quality of the litter they produce. I also asked how might animals that are known to alter litter layers interact with dominant tree species to influence plant invasions. Therefore, I conducted an experiment investigating how changes in litter from dominant tree species and invasions by nonnative earthworms might influence invasibility of forests using forest floor mesocosms (Chapter 3). I found that plant invasion was inhibited by native oak litter even when earthworms were present, suggesting that oak forests may resist plant invasions via oak tree litter. In contrast, plant invasion was greater under invasive tree litter and earthworm activity tended to facilitate invasive plant success only under invasive tree litter. I was also interested in understanding how disturbance might alter relationships between local and regional diversity. The long-term data of the forest disturbance experiment allowed me to investigate how local species richness is mediated by regional species richness after disturbance and during forest community development (Chapter 4). Local richness depended strongly on regional richness only after disturbance via colonization of species, but this relationship changed during forest aggradation. These results suggest that regional species pools are important to maintain local diversity following disturbance, but that local interactions (through canopy closure of dominant trees) exert control over species diversity during community reorganization. Lastly, I tested current theories on how diversity influences compositional stability after disturbance (Chapter 5). Disturbance consistently resulted in decreased compositional stability, but diversity was associated with stability in complex ways, which depended on how stability was measured and at what scale. Species-rich areas were in some instances less stable; in other instances areas with intermediate levels of diversity were more stable. These results suggest that disturbance causes shifts in species composition via colonization, but the ways in which diversity of sites influences compositional stability is complex and depends on methods used and the scales of observation. Taken together, these results suggest that disturbance influences invasibility, species saturation, and compositional stability of ecological communities. These properties change immediately following disturbance, and during forest development and canopy closure. Data from this project were useful in testing existing theories of community ecology, and may ultimately prove useful for forest managers as they decide how to protect biodiversity while planning for other uses of forest resources. Overall, these results suggest that colonization of species is the primary process driving plant community patterns in Appalachian forests following disturbance. / Ph. D.

The rate of extinction of plant and animal species has accelerated at an alarming rate throughout the 20th century. This depletion has resulted in greater legislative control over wildlife within the United States, and the eventual passage, in 1973, of the Endangered Species Act (ESA). The ESA establishes regulations designed to protect species which are threatened with extinction. The ESA has evolved out of past legal statutes and jurisprudence concerning the management and protection of wildlife, but the ESA has created controversy since inception. The latest criticisms concern its lack of success in meeting stated policy goals of preserving species richness and recovering listed species. As a result, a new movement has developed which seeks to change the traditional species-by-species approach of the ESA to an approach which concentrates on entire ecosystems. Ecosystem approach proponents sight various advantages, one of which is cost-effectiveness. This thesis analyzes the development of wildlife protection within the United States, and the Endangered Species Act in particular. The purpose of this analysis is to examine both the legislative powers granted for wildlife protection, and the feasibility of the ESA to encompass ecosystem protection. Modifications to the ESA are proposed. The Clinch River Valley, in Southwest Virginia and Northeast Tennessee, is used as a case study for a cost-effectiveness analysis of an ecosystem approach and species-by-species approach. Costs generated through Fish and Wildlife Service recovery plans are used in conjunction with generated ecosystem plan recovery costs. The results of the cost-effectiveness analysis show that an ecosystem approach is more cost-effective in meeting the recovery goals within the 0inch River Valley. The ESA can be modified to incorporate an ecosystem approach. Such modification can protect species richness and allow for a priority ranking system for protection. / M.S.

The objective of the present study was to acquire a better understanding of parameters controlling the species generation and transport from compartment fires. The experiments were performed in a half-scale ISO 9705 compartment and a 6.1 m long hallway connected in a head-on configuration. The buoyancy driven propane fire was provided by a burner and a continuous gaseous fuel supply system. All the measurements were obtained during the steady state of the fire. The ventilation conditions were fixed and three different fire source elevations were studied for heat release rates ranging from 20 kW to 150kW. The species yields were obtained from performing detailed mapping measurements at the compartment and hallway exit planes. The measurements included local specie mole fractions of oxygen, carbon dioxide, carbon monoxide and unburned hydrocarbons. The local temperature and the local pressure (for local gas velocity calculations) were also measured. In addition, visual observations of the flow dynamic were performed through a window and the vents to give useful insights and lead to a better understanding of the combustion process. The data obtained from the species generation study was analyzed using previously developed methods. The method based on equivalence ratio was presented and determined inappropriate for the present study where the global equivalence ratio was not equal to the plume equivalence ratio due to the complexity of the fire dynamic taking place. The method consisting of correlating the species yields based on the combustion within the compartment as a function of a non dimensional heat release rate allowed qualitative conclusions to be made. The non-dimensional heat release rate was based on the fuel load and the geometric parameters of the compartment. This methodology revealed similarities in the species production between the three fire source elevations investigated. A correlation of the data was obtained based on experimental data. The transport of species to remote locations was studied for the three fire source elevations and fixed ventilation conditions. Species mole fractions and yields were obtained at the compartment exit plane (compartment/hallway interface) and at the hallway exit plane. The results were compared for various heat release rates and showed differences for some scenarios attributed to mixing along the hallway and oxidation reactions outside the compartment. / Master of Science

The study of the process of speciation is instrumental to understanding the species diversity observed today. Diverging populations are intriguing, because speciation has not reached an endpoint, yet the process that may eventually lead to distinct species can be studied. Systems that contain many putative species and/or parallel divergences, such as many species flocks and species pairs, are extraordinary examples of divergence and therefore are critical to the understanding of the speciation process. A "miniature" species flock of pupfish (Cyprinodon variegatus) discovered in lakes on San Salvador Island, Bahamas has evolved in less than 6 000 years, and is, therefore, important to the study of the pace of evolutionary processes. The San Salvador Island pupfish flock is composed of a normal form, which resembles coastal C. variegatus, and bulldog and bozo morphs, which diverge ecologically and morphologically from the normal morph. In Chapter 1, I sequenced the mtDNA control region and used haplotype frequency analyses to assess the level of differentiation between sympatric normals and bulldogs sampled from Osprey Lake and Little Lake on San Salvador Island. The bozo morph was too rare to include in the study. I also included samples of normals that occur in lakes without bulldog and bozo morphs to assess any differences between lakes on the island. All haplotype frequency comparisons for sympatric normals and bulldogs were highly significant, which suggests these morphs are distinct populations in sympatry and, therefore, have characteristics of biological species. Further, an estimation of Time for Speciation supports geological data that suggest this fauna is very young (6 000 years). The San Salvador Island pupfish species flock is, therefore, the youngest known species flock and presents an important model system for the study of how morphological and ecological divergence can promote speciation in Cyprinodon. In Chapter 2, I first compared the San Salvador Island pupfishes to other Bahamian C. variegatus populations to assess the level of inter- and intra-island pupfish population differentiation in the Bahamas. The mtDNA control region was sequenced for bulldogs and normals from San Salvador Island and normals sampled from New Providence and Exuma Islands. San Salvador Island bulldogs were found to be distinct from all normal populations sampled, and comparisons of shared haplotypes suggest they originated on San Salvador Island rather than any of the other islands sampled. This was intriguing, because a "bulldog-like" morph has recently been observed in a lake on New Providence Island, which suggests parallel divergences may be occurring throughout the Bahamas. I also sequenced the mtDNA cytochrome b gene to assess the phylogeography of C. variegatus. Populations were sampled from the Bahamas and the east coast of North America, and the results suggest the Bahamas were only recently colonized by the Southern coastal lineage of C. variegatus. A distinct Northern lineage of C. variegatus, which may warrant species designation, was also supported by the cytochrome b data. Overall, the results supported a San Salvador Island origin for the Little Lake and Osprey Lake bulldog morphs, and also suggest the Bahamian C. variegatus populations are very young. / Master of Science

Species that become naturalized in a new geographical range, subsequently multiply and spread, and persist to the detriment of resident communities, are known as alien invasive species. Two aspects of species invasion – spread and ecological impact – were examined using Lantana camara L. (henceforth lantana) as the study system, specifically in the context of a seasonally dry tropical forest ecosystem of the Mudumalai Wildlife Sanctuary and National Park. Lantana is a thicket-forming woody shrub of South and central American origin, which is now widespread across the tropics. The thesis is divided broadly into two parts -the first part examines the influences of environmental factors on the distribution and spread of lantana while the second part focuses on the effects of lantana on the distribution, survival and growth of native woody species. Much of the work presented in this thesis was conducted within a 50 ha permanent plot (the Mudumalai Forest Dynamics Plot, MFDP hereon) in Mudumalai, chiefly because the history of invasion by lantana has been recorded here since 1989. The influence of changing resources on lantana invasion was explored at two scales -small spatial but fine temporal scale in the MFDP and at the level of the landscape. Available data on an 18 year chronosequence of changes in the qualitative density of lantana from the MFDP and field studies between the years 2009 and 2010 were used to determine the environmental correlates of lantana spread in time and space. It was found that biotic factors such as the presence of the shrub Helicteres isora and abiotic factors such as proximity to drainages and the combination of fire and drought promoted the intensification of lantana invasion in time while proximity to streams, higher total annual rainfall and low fire frequency contributed to lantana invasion at the landscape level. The impacts of lantana on the seedlings of native woody species were assessed in 10m x 10m plots within the MFDP. An initial enumeration of 60 such plots revealed that animaldispersed, dry forest habitat preferring species were most affected by the presence of dense lantana. A follow-up study comprising of growth and survival measurements made on 1105 seedlings over two years (2008-2010) further confirmed that dry forest preferring species were most affected by the presence of dense lantana and that this response at the community level was most likely influenced by the most abundantly sampled species in this habitat preference guild – Catunaregam spinosa. In conclusion, while the environmental correlates of lantana most likely promoted its invasion, only certain guilds of native species seemed to be negatively affected by the presence of lantana.

Models of climate change predict that temperature will increase during the 21th century and the largest warming will take place at high northern latitudes. In addition to warming, predictions for northern Europe include increased annual precipitation and a higher proportion of the precipitation during winter falling as rain instead of snow. These changes will substantially alter the hydrology of rivers and streams and change the conditions for riverine communities. The warming is also expected to result in species adjusting their geographic ranges to stay within their climatic tolerances. Riparian zones and wetlands are areas where excess water determines the community composition. It is therefore likely that these systems will be highly responsive to alterations in precipitation and temperature patterns. In this thesis we have tested the predicted responses of riparian vegetation to climate-driven hydrologic change with a six year long transplant experiment (I). Turfs of vegetation were moved to a new elevation with shorter or longer flood durations. The results demonstrate that riparian species will respond to hydrologic changes, and that without rare events such as unusually large floods or droughts, full adjustment to the new hydrological regime may take at least 10 years. Moreover, we quantified potential effects of a changed hydrology on riparian plant species richness (II) and individual species responses (III) under different climate scenarios along the Vindel River in northern Sweden. Despite relatively small changes in hydrology, the results imply that many species will become less frequent than today, with stochastic extinctions along some reaches. Climate change may threaten riparian vegetation along some of the last pristine or near-natural river ecosystems in Europe. More extensive loss of species than predicted for the Vindel River is expected along rivers in the southern boreal zone, where snow-melt fed hydrographs are expected to be largely replaced by rain-fed ones. With a seed sowing experiment, we tested the differences in invasibility between open wetlands, forested wetlands and riparian zones (IV). All six species introduced were able to germinate and survive in all habitats and disturbance levels, indicating that the tested wetlands are generally invisible. Germination was highest in open wetlands and riparian zones. Increasing seed sowing density increased invasion success, but the disturbance treatments had little effect. The fact that seeds germinated and survived for 2 to 3 years in all wetland habitats indicates that wetland species with sufficiently high dispersal capacity and propagule pressure would be able to germinate and establish here in their respective wetland type. Our results clearly demonstrate that a changed climate will result in substantial changes to functioning, structure and diversity of boreal wetland and riparian ecosystems. To preserve species rich habitats still unaffected by dams and other human stressors, additional protection and management actions may have to be considered.

Master of Science / Department of Plant Pathology / Robert L. Bowden / John F. Leslie / Fusarium graminearum Schwabe sensu lato (teleomorph: Gibberella zeae (Schwein.) Petch), a homothallic ascomycete fungus, is the causal agent of Fusarium head blight (FHB) of wheat (Triticum aestivum L.), barley (Hordeum vulgare L.), and other small grains. FHB occurs worldwide and serious outbreaks have been reported in North America, South America, Asia, and Europe. According to the phylogenetic species concept (PSC), F. graminearum is composed of at least 15 phylogenetic lineages known as the Fusarium graminearum species complex. Although F. graminearum is homothallic, some members of different phylogenetic lineages are known to intercross in the laboratory. It has been suggested that F. graminearum sensu lato fits the biological species concept (BSC). According to the BSC, “species are groups of actually or potentially interbreeding natural populations, which are reproductively isolated from other such groups”. Previous reports of intercrossing were qualitative, so the degree of reproductive isolation, if any, is not clear. Since intrinsic reproductive isolation is the key criterion to identify species by the BSC, more detailed quantitative information is needed. Chromosome rearrangements between fungal strains may reduce fertility in sexual crosses through the production of genetically inviable recombinant progeny. As such, rearrangements can be important postzygotic reproductive barriers between species. Following methods used in Neurospora crassa, ascospore tetrads were analyzed for patterns of ascospore viability. Crosses were made with three lineage 7 (F. graminearum sensu stricto according to PSC) strains as female. Each female was a MAT1-2 knockout mutant that rendered it obligately heterothallic. Males were several members of lineages 6 (F. asiaticum according to PSC) and lineage 7. Crosses with lineage 7 males formed complete asci with 8 ascospores indicating that their genomes are isosequential with the testers. Crosses with one strain from lineage 6 with two known inversions produced asci containing 8, 6, and 4 ascospores, consistent with it not being isosequential. However, three other strains of lineage 6 appeared to be isosequential with the testers. Therefore, chromosome rearrangements did not appear to be common to strains of lineage 6 and probably do not contribute significantly to reproductive isolation of lineage 6 and lineage 7. Interlineage fertility studies with the three lineage 7 tester strains were performed to quantify interlineage fertility parameters including the total number of ascospores produced, perithecial density, and perithecium internal development scores. All lineage 7 female testers successfully crossed to all 23 male strains from lineages 1 to 9. For total ascospore production, one female tester crossed equally well with all lineages and the other two testers showed statistically significant differences for a few lineages. For perithecial density, there was a significantly lower density with all three testers when crossed with lineage 6, but the other lineages were not statistically different from lineage 7. For perithecial development, there was large variation for every lineage. Therefore, in the crosses with reduced fertility, the reduction can be attributed to a postzygotic effect since mature perithecia and asci developed. All of the tested lineages of the Fusarium graminearum species complex can produce viable progeny with F. graminearum lineage 7, which was the taxonomic type of the original species before it was split into phylogenetic species. There are a few examples of reduced fertility with two lineage 7 testers, the remaining tester crossed equally well with all lineages. Therefore members of lineages 1-9 all should be considered members of Fusarium graminearum according to the BSC. The existing female testers could be used to identify members of the F. graminearum clade by performing test crosses in the laboratory. The PSC and BSC species concepts do not agree for this group of fungi. This disagreement indicates that the F. graminearum species complex is in the early stages of speciation. The lack of intrinsic reproductive barriers supports the hypothesis that these lineages have developed in geographic isolation. As the lineages have apparently been brought together through global trade, interlineage hybrids have been reported in the field. The discrepancy between PSC and BSC will eventually be resolved by whether the lineages fuse or remain separate in nature. Even if the lineages remain separate, this study demonstrates the potential for gene flow between lineage 7 and lineages 1 through 9.

This thesis analyzes the ichthyofauna in Czech Republic in historical terms, i.e. how was the data on the occurrence of native species of fishes and lampreys collected. Particular attention is given to species alien to this territory - the following species are elaborated in more detail: Carassius auratus, Ameiurus nebulosus, Pseudorasbora parva, Ctenopharyngodon idella, Hypophthalmichthys molitrix, Aristichthys nobilis, Gasterosteus aculeatus, Coregonus maraena,Coregonus peled, Acipeneser baerii, Acipeneser stellatus, Acipenser gueldenstaedtii. The history of the introductions, the current situation in the Czech Republic and the impact on native ichthyofauna are listed (a more detailed evaluation of invasiveness is included for Carassius auratus, Pseudorasbora parva). Finally, the efforts to enrich the local ichthyofauna by introducing new species are generally evaluated.

Conservation of wood-decaying fungi requires improved knowledge about the long-term effects of forest management; regarding habitat loss, fragmentation and fire suppression. To better understand such effects, I examined the influence of area, isolation, fire history and forest stand characteristics on communities of wood-decaying fungi. Species richness and composition were studied along a gradient of 22 forested islands varying in size (0.16 to 17.58 ha) and fire history (spanning 5000 years) in a boreal forest archipelago in northern Sweden. A total of 490 records of 41 polypore species were found in 33 circular plots, each 0.1 ha in size. Species richness and the number of red-listed species were analyzed using generalized linear models (GLMs), while species composition was examined using non-metric multidimensional scaling (NMDS) ordination. The species composition was clearly different between recent-fire (< 300 years since last fire) and old-fire (≥ 300 years since last fire) islands, mirroring the shift in tree species composition as pine-associated fungal species were replaced by spruce-associated fungal species. The volume of logs was the only variable influencing the species richness, although the diversity of logs showed a clear trend of also influencing species richness positively. The results demonstrate the importance of having both recent-fire and old-fire forests as landscape-level habitats and species pools, where fire naturally would constitute a key role for maintaining forest biodiversity in the boreal forest landscape. The results also stress the importance of dead wood for species richness at the individual forest stands.