Landscape change and biodiversity decline is a global problem and has sparked world-wide initiatives promoting biological conservation techniques such as reintroductions, protected area networks, and both preservation and restoration of landscape connectivity. Despite the increasing abundance of such working landscapes (i.e. “human-modified” landscapes), we know relatively little about their ecological mechanics; these landscapes can be vast, encompassing areas too large to obtain high resolution ecological data to test ecological process. To investigate the ecological mechanics of working landscapes, I use a small, tractable, landscape mesocosm situated in east-central Alberta, Canada, The Cooking Lake Moraine (a.k.a. the Beaver Hills Biosphere). The chapters within this dissertation quantify biodiversity across a hierarchy of measurements (from genes to communities) and investigate consistencies in ecological processes generating patterns in these biodiversity measurements across spatial scales. As a result, I investigate both a depth, and breadth, of spatial ecological processes underlying the efficacy of biodiversity conservation techniques in heterogeneous working landscapes. In Chapter I, I explore between-landscape functional connectivity by investigating the genetic contribution of reintroduced individuals to an ostensibly successfully reintroduced population within the mesocosm. I find that contemporary animals are the result of recolonization from adjacent sources rather than putative reintroduction founding individuals, indicating greater mesocosm functional connectivity to adjacent landscapes than previously thought. In Chapter II, I probe within-landscape functional connectivity by quantifying the contribution of protected areas, natural, and anthropogenic landscape features to animal movement across the mesocosm. I find that natural features had the largest effect on animal movements, despite the presence of protected areas. Chapter III investigates protected area network efficacy on biodiversity conservation by quantifying the contribution of protected areas, natural, and anthropogenic landscape features to mammalian functional diversity across multiple spatial scales within the mesocosm. I find that protected areas rarely predict functional diversity across spatial scales; instead natural features positively predict functional diversity at small spatial scales while anthropogenic features are negatively associated with biodiversity at large spatial scales. Finally, Chapter IV ties the previous three chapters together by testing implicit assumptions of the species occurrence data collected in each. I compare GPS collar data (Chapter II) to species occurrence data collected on wildlife cameras (Chapter III) to demonstrate that the magnitude of animal movements better predict species occurrence than the commonly assumed proximity of animal space use. Across chapters, two central themes emerge from this dissertation. First, the importance of natural features at small spatial scales, and anthropogenic features at large spatial scales, within the landscape matrix is predominant in predicting multiple measures of biodiversity. And second, we cannot assume predictable efficacy of conservation strategies or even the ecological process inferred from the data collected to test these strategies. / Graduate

During the last 25 years, the Eastern Cape Province of South Africa has seen the establishment of many small to medium sized (≤ 440km2) game reserves. These reserves have reintroduced many of the larger indigenous wildlife that had been extirpated by the early 20th century. As such, these reserves and wildlife introductions have created many research opportunities, including investigations on the ecology of reintroduced carnivores in the Thicket biome. Brown hyaenas (Hyaena brunnea) are one of the large carnivore species that have been reintroduced into the area. As these animals have predominantly been studied in more open, arid systems, their reintroduction has provided an excellent opportunity to study the species in an alternate natural habitat. Information gathered from such investigations adds to our knowledge of the species and also provides information for the management of brown hyaenas within small, enclosed reserves. Data were collected over the period of one calendaryear, from February 2013 to February 2014 at Kwandwe Private Game Reserve in the Eastern Cape, South Africa. Brown hyaena population estimates were calculated using capture-recapture methods from individually identifiable images captured during a three month camera trapping survey. Images of brown hyaenas were separated into left- and right-side profiles. Twenty-eight individuals were positively identified from left-side images and 27 from right-side images. Non-spatial and spatially explicit capture-recapture analyses were both run in the program DENSITY 5.0. Density estimates ranged from 14 to 20 individuals/100km2 (equivalent to a total abundance of 26-37 individuals) depending on the method used. Despite the range of estimates, all are considerably higher than in other areaswhere densities have been calculated. Satellite/GPS collars were fitted to three individual brown hyaenas (two males and one female) to measure their home range size and use of space. Home range size was calculated using two different methods, Kernel utilisation distributions and Brownian bridges. Home range estimates were similar using both methods; however Brownian bridge methods appeared to exaggerate the use of space by individuals. Kernel home range sizes for the three individuals ranged between 42.62km2 and 79.88km2. These estimates are considerably smaller than previous findings from other parts of Africa and suggest that sufficient resources may be available within this enclosed system. The results from this study suggest that brown hyaenas are successful generalists in this enclosed system and are able to persist at high densities and occupy relatively small home ranges. This information is important for the managers of small reserves who wish to reintroduce brown hyaenas. Should brown hyaenas be introduced into reserves in the Thicket biome with sufficient resources, the numbers are likely to increase rapidly. If high numbers are not desired then preventative measures (e.g. contraception) should be investigated before release.

The purpose of this thesis is to document river otter (Lutra canadensis) distribution and reintroduction potential in northern Utah. Distribution was studied using data from 3 sources: 1) otter sighting records from Utah Division of Wildlife Resources; 2) surveys of Utah furbearer trappers and natural resources personnel; and 3) searches of streams for otter sign. Potential for river otter habitat/reintroduction was evaluated by assessing food, cover, and reintroduction attributes. Streams were ranked using an evaluation system based on data from the otter literature. Forty-six positive otter sightings were made in Utah by trappers, natural resources personnel, and the public, 1964-1988. Only 1.3% of 844.4 km of northern Utah streams had otter sign during winter and summer searches. General characteristics of northern Utah streams such as habitat type and stream gradient are suitable for river otters. However, stream alterations and livestock grazing have negatively impacted potential otter habitat. Ninety-four percent of the studied streams are presently unacceptable for reintroductions. Escape cover is the most limited habitat attribute, but food appears to be available in adequate quantities. We recommend no otter reintroductions be made until riparian zones are rehabilitated and protected. Reestablishment of stream bank vegetation is essential to provide escape cover for reintroduced otters. We also recommend control of pollution inputs and no further construction of reservoirs. Surveys of otter distribution and evaluation of potential reintroduction should be done on the Colorado River drainage in Utah.

京都大学 / 0048 / 新制・論文博士 / 博士(農学) / 乙第13020号 / 論農博第2830号 / 新制||農||1042(附属図書館) / 学位論文||H28||N4966(農学部図書室) / 32948 / (主査)教授 祝前 博明, 教授 今井 裕, 教授 廣岡 博之 / 学位規則第4条第2項該当 / Doctor of Agricultural Science / Kyoto University / DFAM

1871 var bäver utrotad i Sverige, 1922 gjordes den första återintroduceringen i Jämtland. Det dröjde ända till 1960-talet innan bäver återintroducerades i Norrbotten. Återintroduceringarna i Sverige är ett relativt outforskat område och syftet med denna uppsats är att kartlägga återintroduceringarna i Norrbotten som skedde mellan 1960 - 1976. Resultaten som sådana är inte generaliserbara då de första utsättningarna i Norrbotten gjordes avsevärt senare än i övriga län, vilket innebär att aktörer och processer kan ha varit annorlunda. Dock visar denna undersökning på de bakomliggande motiv och restriktioner som fanns, vilket potentiellt kan tillämpas även vid andra återintroduceringar. Av bäverinventeringsmaterialet används endast information gällande Norrbotten, trots att inventeringar över hela Sverige redovisas, vilket gör det möjligt att potentiellt undersöka om samma mönster även gäller generellt. / By 1871 the beaver was extinct in Sweden. The beaver was first reintroduced in the county of Jämtland in 1922 and during the 1960s it was also reintroduced in the county of Norrbotten. This is a relatively unexplored field and the intention is to chart the reintroduction of beavers in Norrbotten between 1960-1976. This essay will not study the reintroductions in general. It will only study Norrbotten during the period mentioned above. The results as such is not generalizable because the first reintroduction in Norrbotten was considerably later than in the other counties, which means that the agents and processes involved may have been different in other parts of the country. However, this survey shows the underlying motivations and constraints that existed and this might be applicable on other reintroductions. In the materials of the beaver inventories only information regarding Norrbotten has been used, even though the inventories contains a chart of the beavers in all of Sweden. It is possible to investigate if these patterns might be generally applicable

The objectives of this study were to: (1) quantify levels of genetic intermixing in remnant Illinois eastern woodrat populations augmented with Ozark woodrats, and (2) assess genetic health and population viability in remnant and reintroduced woodrat populations. This dissertation is arranged into 3 chapters. DNA extractions, microsatellite genotyping, and mtDNA haplotype sequencing were conducted for individuals from the western Shawnee National Forest (WSNF) (i.e., LaRue Pine Hills [LPH], Fountain Bluff, Horseshoe Bluff), eastern Arkansas (AR), southern Missouri (MO), and 5 locations in the eastern Shawnee National Forest (ESNF), Illinois. Genetic data provided a strong signal in STRUCTURE which confirmed MO, AR, and LPH were genetically distinct populations prior to the genetic augmentation at LPH. Following augmentation, eastern woodrats in ESNF and WSNF have experienced admixture with introduced populations, increased allelic richness, decreased inbreeding, and increased effective population size, while genetic swamping was not observed. Haplotypes from both introduced populations were observed after augmentation at LPH but neither were detected in adjacent remnant populations. Reintroduced populations in the ESNF are acting as a single population and descendants consist of primarily AR and MO individuals. Admixture was apparent in all ESNF populations and haplotypes from all source populations persisted. Effective population size was largest at Garden of the Gods but all 5 populations had similar allelic richness and heterozygosity. In the ESNF, migration is not supported between study populations and additional habitat management may be needed. Future population monitoring of the ESNF population should focus efforts in the 3 central populations which are likely source populations. Future eastern woodrat management efforts should focus on (1) connectivity between ESNF populations and (2) connectivity in the WSNF amongst remnant populations and between remnant and reintroduced populations in the central Shawnee National Forest. I recommend the establishment of migration corridors by decreasing canopy cover in the forested matrix and the establishment of artificial rock outcrops in the form of boulder piles along migration corridors. My study indicates that genetic augmentations can be a beneficial management action in isolated eastern woodrat populations having experienced loss in gene flow, while reintroductions can benefit woodrats in areas believed to have experienced local extirpation. I also recommend identifying genetic structure of source populations for genetic augmentations and reintroductions beforehand to confirm multiple source populations, genetic variation, and introduction of multiple alleles. As rodents often have short lifespans and individual reintroductions may fail to establish, genetic augmentations and reintroductions should include the relocation of multiple individuals over time via soft release.

The extent and quality of freshwater systems is declining globally. Combined with past drainage, straightening and flow regulation, current systems are often functional but not pristine. Conservation, creation and restoration of freshwater systems is common but requires significant planning, resources and active monitoring and may only be a short-term solution to the long-term problem of destruction and loss of riparian zones. Beavers (Castor spp.) have the ability to create physical and biological habitat heterogeneity through the construction of woody debris dams, thereby restoring lost natural discontinuities in freshwater systems. Beavers may thus offer a natural, more passive solution to the need for wetland restoration or creation and the problem of homogenisation of watercourses. As such, numerous beaver reintroductions and introductions have been undertaken based in part on restoring this lost natural heritage. However, it is crucial to be able to predict the potential effects on existing biota of physical modifications by beavers to ecosystems, especially in the light of further population expansion, whilst also disentangling these effects from other influences, namely herbivory. The impact of beavers on aquatic systems was studied using a combination of field-based surveys and experiments, using aquatic plants and macro-invertebrates as indicators of hydromorphological changes and to quantify the effects of direct foraging. The research presented in this thesis demonstrates beaver adaptive foraging behaviours between terrestrial and aquatic habitats, whilst feeding highly selectively, optimally and opportunistically, using the white water lily (Nymphaea alba) as a model species. The effects of beaver foraging on the aquatic plant resource and diversity was low over short time spans (e.g. 1 year), but when selective foraging was assessed over greater time scales (e.g. 10 years) the effects of foraging were distinct. Significant changes in aquatic plant height, biomass, richness, diversity and composition were observed over this time period due to selective grazing on large rhizomatous species (e.g. Menyanthes trifoliata). These direct effects occurred even though changes in water levels, which are commonly believed to be the main driver of beaver influence on aquatic vegetation, were negligible. In a separate study in Sweden where beavers commonly constructed dams, with ponds then forming upstream, the aquatic plant and coleoptera species richness and composition differed in comparison to adjacent non-beaver created wetlands. Therefore, having a range of wetland types in the environment increases physical and biological heterogeneity creating unique niches that are exploited by disparate taxa. The construction of a series of dams within a single reach of stream flowing through a Scottish agricultural landscape also increased physical habitat diversity. Distinctive macroinvertebrate assemblages and modified functional diversity were associated with each dominant habitat type in the stream, resulting in increased landscape scale richness. The findings of this thesis confirm that beaver engineering and foraging has the potential to create unique and highly heterogeneous wetland and stream habitats within landscapes that enhances richness and diversity for multiple species groups. This thesis also supports part of the rationale for the trial reintroduction of beaver to Scotland that beavers can restore degraded habitats.

Beavers have been described as a “keystone species” and “ecosystem engineers”, and in this dual role have great potential to physically modify their environment through tree-felling, foraging and construction activities. The resultant change in habitat heterogeneity can affect the flora and fauna that share the habitat with them. There has been recent interest in reintroducing the Eurasian beaver to the United Kingdom after an absence of over 400 years. To date, no research (aside from this thesis) has focussed on beaver ecology and behaviour in Scotland. This study has investigated the ecological effects of a small number of beavers in two enclosed but semi-natural Scottish sites at Bamff in Perthshire. The research conducted over a three-year period, with particular emphasis on the effects of tree-felling, foraging and construction activities. Trees were felled for both dietary and construction purposes, with felling rates being influenced by habitat availability, quality and the degree of habitat modification required. Highest rates were evident during the initial colonisation period of marginal sites (c. >300 trees / beaver / calendar year), and lowest rates in later years of occupation of more optimal sites (c. 55 – 70 trees / beaver / calendar year). Preferences were generally for willow and aspen trees, with conifers almost entirely avoided, and smaller trees preferred over larger ones. Proximity of trees to waterbodies was also an important factor, with nearer trees favoured, and generally most felling occurred within 50 m of water. Such behaviour followed the principles of optimal central place foraging. These preferences were less predictable however when intense construction activity was undertaken, with larger trees preferred and generic preferences for deciduous trees apparently invalid. In such cases, close proximity to the construction site was of prime importance. Increased cover of herbaceous plant species was observed in beaver-created canopy gaps in riparian woodland, whilst macrophyte diversity within waterbodies increased slightly in areas of herbaceous grazing. The diversity of terrestrial ground invertebrates was highest in areas of heavy tree-felling, and invertebrate richness and abundance was greatest in areas of herbaceous grazing under an intact tree-canopy. Furthermore, the abundance, diversity and richness of macroinvertebrate communities were increased by beaver-generated woody debris in ponds and streams. Overall, 30% of all macroinvertebrate species collected were found only in beaver-affected areas, due to the refugia and food supply provided by beaver dams, caches and lodges, as well as hydrological effects of these structures. These results are discussed with reference to future plans to return the beaver to Scotland. The habitat usage and modification of riparian ecosystems in northern Britain is likely to be similar to that found in this study, and the results are believed to be relevant, applicable and transferable to many areas of Scotland.

The capercaillie is the largest member of the grouse family extant in Scotland. This species is reported to have become extinct during the 1700s. It is also reported to have been reintroduced to Scotland from Sweden during the 1800s. There have been many assertions made about the underlying causes of the decline of the species; however the specific causal factors remain unknown. The reintroduction of the capercaillie to Scotland in the 1800s is the only successful reintroduction of a grouse species ever to have occurred in the world. The specific factors behind the success of the reintroduction also remain unknown. This thesis examines the extent to which a selection of historical documentary evidence can help to establish both the causes of the 18th century decline of the capercaillie in Scotland and the successes of the 19th century reintroduction. The methodology of this thesis incorporates facets from the fields of both environmental science and history. The methodology includes three steps. The first step involves the selection of a series of potential critical factors that may have been responsible for the decline of the species in the 1700s; these critical factors were selected from the present day understanding of the ecology and the behaviour of the capercaillie. The second step of the methodology includes the surveying of a series of historical documentary sources. From these surveys historical observations of the species were gathered. The historical documentary sources selected for examination in this thesis include what are referred to here as ‘primary historical source material’ and ‘secondary historical source material’. The majority of the primary historical source material was gathered from the hand written manuscripts of the Breadalbane estate, held at the National Archives of Scotland (NAS) and the Atholl estate held at the Blair Castle Charter Room in Blair Atholl. Other select primary historical source material consulted to a lesser extent, due to time constraints, includes the Forfeited Estates (1745) Inventory and the Baron Court Records for Strathspey and Urquhart (1617–1683) from the Grant estate muniments; both held at the NAS. The secondary historical source material was gathered from published and edited literary collections that include historical accounts of the species. The third and final step of the methodology involves the synthesis of both the historical and environmental information in order to establish to what extent the causes of both the decline of the species in the 1700s and success of its reintroduction in the 1800s can be realised. The findings from this thesis assert that the capercaillie was resident in Scotland from, at least, the end of the Medieval. Moving forward from the Middle Ages this thesis presents observations of the capercaillie throughout the historical period. These observations of the capercaillie appear in many different historical accounts. In some instances these observations are fleeting and do not form the main subject of the particular document in question. In other instances accounts of the species are much more detailed and include references to the ecology and behaviour of the bird. The level of detail included in an observation aside, the frequency with which the species is referred to in the sample of historical documents suggests that sections of the Scottish human population were familiar with the species, in various locations and at various times throughout history. By the 17th century the capercaillie is reported as beginning to become rare in some locations while still remaining comparatively abundant in others. The number of instances where the species is referred to as becoming rare in the historical documents increases between the 17th and the 19th centuries. Despite the reported scarceness of the species in Scotland from around the 17th century onwards, the capercaillie is recorded as persisting in Scotland until around the end of the 1700s. By the early 1800s the number of observations of the species in the secondary historical source material increases. All of the observations in the secondary sources from the early 1800s record the absence of the species from localities and regions of Scotland. No new evidence was found in either the primary or secondary historical source material to challenge the supposition that the capercaillie did become extinct in Scotland after 1785. No detailed quantitative data was available for analysis of the decline of the species. Thus, to investigate the extent to which the historical accounts can help explain the specific causal factors of the reported decline, a synthesis of the environmental and historical data was necessary. The findings of this synthesis suggest that the naturally occurring Scottish population of capercaillie probably persisted in the form of a metapopulation. The two hundred years between the 17th and 19th century most likely saw the extinction of capercaillie sub-populations, before the loss of the overall population of capercaillie around 1785. The sample historical documentary evidence alludes to this pattern of local and/or regional extinction of sub-populations. The cause or causes of the extinction of these sub-populations has focussed on five limiting or critical factors known to affect the species today. These five factors are climate change, particularly weather effects associated with the Little Ice Age, habitat loss and deterioration, disturbance, human hunting and predation by species other than humans also contributed to the species’ extinction. The extent to which these critical factors affected each sub-population would have varied between regions of Scotland occupied by the capercaillie in history. This thesis proposes that there was no single or combination of specific critical factors that were ultimately responsible for the decline of the capercaillie in Scotland during the 1700s. In some areas the capercaillie sub-populations would have most likely died out as a result of habitat loss and deterioration and climate change. Whereas in others predation and inbreeding may have been the critical factors responsible for the species’ demise. More detailed information referring to the capercaillie was found in the historical documentary source material for the period post-extinction (i.e. 1800 onwards). Contrary to popular understanding numerous attempts to reintroduce the capercaillie to Scotland were carried out before the Marquis of Breadalbane’s successful programme in 1837. The historical documentary evidence reports early attempts to reintroduce the species to locations such as the Isle of Arran in 1807, on the Duke of Atholl’s estate in 1822 and on the Earl of Mar’s estate in 1824. None of these reintroduction programmes are reported to have been successful in establishing a ‘wild’ population. However, in some instances the captive rearing programmes initiated did bear some fruit and captive reared birds were sent from Dunkeld by the Duke of Atholl to Kenmore and were used in Breadalbane’s successful reintroduction in 1837. The historical documents report two causes for the failure of these early reintroduction attempts. The first is the sudden death of captive birds, most likely as a result of choking due to stress as observed in recent rearing programmes (i.e. Moss 1986). The reintroduced Arran population is reported to have become extinct in this fashion. The second reported cause of failure is predation by species other than man. For example the entire population of birds brought to Scotland by the Earl of Mar were predated when released on his estate. This thesis offers two critical factors as explanations for the remarkable success of the capercaillie’s reintroduction to Scotland in the 19th century. The first is the method by which the reintroduction was carried out; specifically, the re-establishment of a series of capercaillie sub-populations in different regions of Scotland.

Preservation of rare plant species often requires establishment of new populations. Survivorship surveys are the most common method of post-introduction monitoring. However, they provide an incomplete picture of establishment success. This study is an attempt to develop a model for determining establishment success by determining the factors affecting recruitment in introduced populations of a rare species. Erigeron decumbens is an endangered forb endemic to the Willamette Valley of western Oregon. Several populations of E. decumbens have been introduced by governmental and non-profit agencies. While there has been some monitoring of the survival of introduced plants, no systematic surveys have measured recruitment in the new populations. We monitored recruitment in five introduced populations, and compared abiotic and biotic characteristics in these and five stable natural populations. Seventy percent of introduced populations produced fewer than one recruit for every three survivors. Thirty percent produced at least one and one half recruits for every one survivor. The factors that affected recruitment were site specific. Low recruitment (less than one recruit per three survivors) was associated with dominance by exotic species (Dactylis glomerata, Rosa eglanteria, Vicia tetrasperma, and Leucanthemum vulgare), high litter cover, high soil electrical conductivity, and low silt levels. Recruitment was highest at sites with higher native plant species richness and soil characteristics falling within the variation of large natural populations. Viable seed number per individual had the strongest linear relationship with recruitment, demonstrating that seed viability could be a strong limitation for this species (r² = 0.83). The results of this study suggest numerous guidelines for future reintroductions of E. decumbens. This research also demonstrates the utility of recruitment surveys to determine factors important in the success of introduced populations of rare plant species. / Graduation date: 2013