Evidence for rapid evolutionary change in plants in response to changing environmental conditions is widespread in the literature. The first objective of this research was to estimate rates of evolution for previously published studies to begin a foundation for comparison and to examine trends and rates of microevolution in plants and to demonstrate how estimated rates could be used to address specific microevolutionary questions. The remainder of the research in this paper focused on invasion as a stimulus for microevolution using a major invasive grass, Bromus tectorum . Identification of pairs of European, source and North American, introduced population pairs of B. tectorum provided an ideal opportunity to test for post-introduction increase in invasiveness because parallel change involving multiple introduction events suggests a non-random process. This research tested to see if B. tectorum evolved to be more invasive by becoming larger and more fecund, more competitive, and more phenotypically plastic in its invasive range than its source range. All B. tectorum experiments were done in common garden or greenhouse settings to control environmental differences, and comparisons were made between morphological characteristics of introduced and source B. tectorum populations. The experiments showed that introduced B. tectorum was taller and had higher reproductive output than source B. tectorum, but the populations did not differ in aboveground biomass. Contrary to predictions, North American B. tectorum had less competitive effect than European B. tectorum on one of the native bunchgrasses, with no other differences in competitive ability. All of the B. tectorum populations show plasticity for the traits measured in this study, however there was no strong trend to suggest that there has been a change in the level of plasticity in introduced populations. Overall, these results suggest that B. tectorum may have evolved to be more ruderal to capitalize on characteristics that increase its ability to spread in its new environment while not focusing as many resources on competitive ability. As for plasticity, it is likely that the same strategy of phenotypic plasticity from its native range could be a key to the success of B. tectorum in its invaded range with similar levels of selection for plasticity in both ranges.

This dissertation presents several new approaches to analyzing species-habitat relationships in multi-scaled, hierarchical systems, and demonstrates their utility in analyzing a forest bird data set from the Oregon Coast Range. In the first chapter I present a review of past studies of species-habitat relationships at the landscape level. I then suggest a number of attributes that studies should include in order to rigorously analyze landscape-level relationships between species and their environments. In chapter two I introduce a new method of hierarchical, multi-scale decomposition of species-environment relationships. The approach translates a hierarchical, multi-scale conceptual model into a statistical decomposition of variance. It uses a series of partial canonical ordinations to divide the explained variance in species-environment relationships into its independent and confounded components, facilitating tests of the relative importance of factors at different organizational levels in driving system behavior. In chapter three I discuss the implications of how the species response variables are recorded for analyses of multi-scale species-environment relationships. In chapter four I demonstrate the use of this method in analyzing forest bird species-environment relationships. Chapter four demonstrates the hierarchical partitioning method for a large forest bird data set from the Oregon Coast Range. I found that plot-level factors were better predictors of community structure than either patch- or landscape-level factors. There were major differences among life-history groups in terms of the relative importance of factors at the three spatial scales. In chapter five I describe the landscape-level patterns of avian diversity in the Oregon Coast Range. I used a combination of factorial analysis of variance and partial discriminant analysis to quantify the relative importance of differences in mature forest area, fragmentation and basin in influencing each response variable and community diversity overall. Bird community diversity was influenced by both the extent and fragmentation of mature forest at the landscape-level. Species richness and density responded more strongly to mature forest area than to fragmentation, and were significantly lower in landscapes that were completely dominated by mature forest than in landscapes with a mixture of seral stages. Species evenness was more strongly related to fragmentation than to area of mature forest at the landscape-level.

The cichlid fishes of the East African Lakes are noted for their explosive radiation and vast diversification over short evolutionary time. Lake Malawi, in particular, houses a diverse cichlid assemblage. There are numerous theories to explain this diversity that are often linked to sexual selection through female choice. Given that males are brightly colored, female choice cues have been thought to be primarily visual. Little, however, has been done to investigate the visual factors used by Lake Malawi cichlids. The primary objective of my dissertation is to ask four main questions using Lake Malawi mbuna (rock-dwellers): (I) what spectral abilities exist across several taxa? (II) what is the role of visual cues in interspecific mate choice? (III) what is the role of ultraviolet radiation in mate choice and feeding, and (IV) what variation in physical characteristics occur with attenuating light? A critical understanding of the visual factors affecting the Lake Malawi mbuna is a necessary step in understanding the mating systems of these fish. This understanding can aid in explaining the extreme diversity seen in these mbuna, and comes at pivotal time because of the anthropogenic stressors that currently threaten the integrity of the East African cichlid fish.

With recent estimates in excess of 151,000 elephants in northern Botswana, there is growing concern about the management of this large elephant population. Despite several telemetry studies of elephant movements, there is limited information on the seasonal home ranges and transboundary movements of elephants in northern Botswana. Our home range measures of 19 elephants with satellite collars were highly variable by age, sex, season and availability of water. One of our collared bulls had the largest home range (24,828 km2) recorded for African elephants. Their extensive transboundary movements indicate that the elephants of northern Botswana are part of a larger contiguous elephant population encompassing western Zimbabwe, the Caprivi Strip in Namibia, southeast Angola and southwest Zambia. Elephant populations in the Caprivi Strip, southeast Angola and southwest Zambia have been affected by civil war, poaching, human settlements, landmines, and veterinary fences. Based upon our dry season aerial survey of the Caprivi Strip in November 2005, elephant numbers (5,242) increased by only 15% since 1998. The recent dispersal of elephants into Angola from the Caprivi Strip may account for this small increase. In contrast, our three aerial surveys of Luiana Partial Reserve, Angola indicate that elephant numbers are increasing rapidly and expanding their range in the Reserve, from 329 in January 2004 to 1,827 in November 2005. The end of 25 years of civil conflict in Angola probably provided the requisite security for elephants to return to southeast Angola and southwest Zambia. Our wet season (January 2004) and one of our dry season (August 2004) aerial surveys of elephants in Sioma Ngwezi National Park, Zambia indicate that elephant numbers (1,099 and 899, respectively) were similar to those reported by an aerial survey in 1991. However, our second dry season aerial survey (October 2005) indicated that limited access to water and extensive fires can dramatically decrease elephant numbers in the park (385). Conservation of elephants in the Lower Kwando River Basin can be enhanced by: realigning veterinary fences, de-mining, improved land use management, and development of anti-poaching and community conservation programs.

Animals move through landscapes where their resources are unevenly and often patchily distributed. When animals move and choose among their scattered resources in predictable ways, ecologists may be able to anticipate the spatial distribution of their populations and the relative strength of their trophic interactions (e.g. predation, or facilitation). Likewise, an understanding of animal movements can inspire the design and preservation of habitat for conservation or the promotion of ecosystem services. However, movement-based predictions of animal populations—and the human interventions these predictions inspire—are only as reliable as our understanding of the determinants of animal movement. The research presented here addresses three basic aspects of animals' environments and experiences that have the potential to influence our understanding of animal movement, population distributions, and ecological interactions: (1) animals' conspecific density (2) prior experience with resources, and (3) the composition and geometry of habitat patches. These topics are well-studied, but rarely in the context of spatially heterogeneous landscapes, and many prior studies have confounded important explanatory variables. For the first study, I performed a lab experiment using the bean beetle, Callosobruchus maculatus, and two of its host beans, Vigna unguiculata and V. radiata, to explore how consumer density influences resource choice and the relative distribution of damage among resources in a patch. My results demonstrated that the damage a focal resource type receives can depend on the frequency of neighboring resource types, and that this frequency dependence decreases with regional consumer density. These findings illustrate the importance of consumer density in mediating indirect effects among resources, and suggest that accounting for consumer density may improve our use of mixed‐crop pest management strategies. For the second study, I used a field experiment to test whether prior hostplant experience influences the distribution of offspring on and damage to hostplants among and within plant patches that varied in hostplant density and composition. Specifically, I reared diamondback moths, Plutella xylostella, on either collard or mustard plants (Brassica oleracea or B. juncea, respectively) and recorded the number of offspring on and damage to plants in three patch types within large field cages: two collards, four collards, and mixed patches of two collards and two mustards. I found that in cages with collard-reared moths, there were more offspring and damage per plant in four-collard patches than in two-collard patches, while mustard-reared moths did not respond to collard density. In contrast, I found no effect of natal hostplant experience on hostplant choice within mixed patches, and no influence of mustard plants on attacks on collards in mixed patches versus two-collard patches (i.e. there were no associational effects). These findings suggest that accounting for prior hostplant experience may improve our understanding of how some herbivores and their damage are distributed in patchy environments through time. For my final dissertation study, I used a correlated random walk model to make predictions for animal density in patches that vary in area and perimeter-to-area ratio. I then tested predictions from this model by manipulating the area and perimeter-to-area ratios of plant patches and observing the densities of two predaceous beetles: a relative habitat specialist, Calosoma sayi, and a relative habitat transient, Tetracha carolina carolina. The model predicted that as habitat specialists spend more time in patches relative to non-habitat, patches with lower perimeter-to-area ratios should have higher animal density. However, for relatively transient species, defined as spending more time within non-habitat than in the habitat patches, the model predicted slightly higher densities in higher perimeter-to-area patches or no difference between patch types. Area per se, in contrast, did not influence mean density. Contrary to the model, I found that in the field experiment patch area and perimeter-to-area ratio interacted to influence the movement and density of the relative habitat specialist, C. sayi. Their density increased significantly with patch size in high perimeter-to-area patches, but patch size had no influence on C. sayi's density within low edge-to-area patches. By contrast, densities of T. carolina were slightly higher in high perimeter-to-area ratio patches once the influence of plant height was accounted for, which was consistent with the model. These results underscore the importance of considering both patch area and perimeter-to-area ratio as well as species-specific movement behaviors for the management of habitat for ecosystem services. Together these projects highlight exciting new areas of consideration for the study of plant-insect interactions in heterogeneous habitats, which could improve our ability to predict insect distributions and interactions in natural and managed populations. / A Dissertation submitted to the Department of Biological Science in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester 2018. / August 27, 2018. / associational effect, natal habitat preference induction, plant-insect interactions, resource concentration / Includes bibliographical references. / Nora Underwood, Professor Co-Directing Dissertation; Brian Inouye, Professor Co-Directing Dissertation; Nick Cogan, University Representative; Alice Winn, Committee Member; Scott Burgess, Committee Member; Joseph Travis, Committee Member.

Abstract not available

Abstract not available

The development of the discipline of landscape ecological planning is traced through 1960-1980 which is considered a period of reform associated with the rapid growth and wide global acceptance of the environmental movement; a movement which has subsequently lost its earlier momentum. Directions for further development of the discipline are suggested and center on the need to combine holistic planning concepts with the reductionisms of longer established scientific disciplines. More penetrative assessments and syntheses derived from interpreting available knowledge, drawn from the findings of many disciplines are proposed. Emphasis is, however, placed on the role that plant-derived bioenergies have as a renewable resource of global implication. Bioenergy resources have greater universal and lower cost availability than other renewables. Plant biomass also represents the one known, energy efficient mechanism for withholding the likely increased effects of environmental degradation under entropy. Projected world populations and the likely global stresses associated with the dimunition of fossil fuels in the century ahead make consideration of bioenergies and land use relationships an imperative for today. A descriptive methodology which uses computer-assisted mapping techniques is described and validated for assessing both existing and potential resource supplies as governed by the abiotic parameters of the environment. Biotic and anthropic variables act as conflicts and constraints for graphic matrices and reduce the likelihood of maximizing resource supplies in real-world situations. A provisional predictive methodology for determining the impact of five bioenergy scenarios on existing regional ecosystem health and landscape patterns utilizes a concept of indices applied to each land use of a compartmental classification system. Indices account for the percentage of land uses in protective phases, the effect of transactions between compartments and regional boundary effects. The two methodologies are interconnected at detailed land management, rather than planning levels and are framed for the formulation of positive rural planning policies.

Isozyme variation and genotypic diversity in three different-aged populations of Ammophila breviligulata (Fern.) (American beachgrass) were investigated in order to determine the nature and causes of intra-species changes during sand dune succession. Ramets of Ammophila breviligulata were collected along three randomly located transects in each of three populations on Castle Neck, a barrier island in Ipswich, MA. Populations ages (I $<$ 7 yrs., II = 10-20 yrs., III $>$ 35 yrs.) were estimated by examining a series of aerial photographs taken over the last 35 years. The ramets were randomly transplanted to a common field plot in Amherst, MA. Ten soluble isozyme systems extracted from young leaf tissure were surveyed in a total of 300 transplanted ramets using starch gel electrophoresis and the zymogram technique. Microgeographic differentiation of populations has occurred with respect to particular allozyme phenotypes. Significant differences were found among study populations in the frequencies of allozyme phenotypes at three of seven polymorphic isozyme loci (Duncan's multiple range test, p $<$ 0.05). Factors such as random drift, founder effects, and natural selection may have contributed to the differences in phenotype frequencies among populations. Changes in population structure appear to have occurred with time. Significant differences were found among populations in the frequencies of 19 multilocus genotypes (G-test, p $<$ 0.05), suggesting that populations may consist of ecotypes. Genotypic diversity increased significantly with greater population age (Duncan's multiple range test, p $<$ 0.05). The mean number of populations in which a genotype was found was significantly higher in the younger population than in the older population (Duncan's multiple range test, p $<$ 0.05). These results suggest that changes toward increased genotypic diversity and increased niche specialization of genotypes occurred within the populations during sand dune succession. The inter-species changes found in this study parallel the inter-species changes found in other studies of ecological succession. The high degree of genotypic differentiation among populations has implications for those who use Ammophila breviligulata in beach stabilization efforts. Current dune management techniques involve planting dunes with a single genotype. This study indicates that monoculture of man-made dunes should be abandoned in favor of planting various genotypes better adapted to the particular environmental conditions.

abstract: Global biodiversity is threatened by anthropogenic impacts, as the global population becomes increasingly urbanized. Conservation researchers and practitioners increasingly recognize the potential of cities to support biodiversity and foster human-nature interactions. However, further understanding of social and ecological mechanisms driving change in urban biodiversity over time is needed. In this dissertation, I first synthesized evidence for the urban homogenization hypothesis, which proposes that cities are more similar across space and time than are the natural communities they replace. I found that approaches to testing urban homogenization varied widely, but there is evidence for convergence at regional spatial scales and for some taxa. This work revealed a lack of long-term urban studies, as well as support for social and ecological mechanisms driving homogenization. Building from this systematic literature review, I tested the effects of a long-term nutrient enrichment experiment in urban and near-urban desert preserves to evaluate indirect urban impacts on natural plant communities over time. Urban preserves and nitrogen-fertilized plots supported fewer annual wildflower species, limiting their effectiveness for biodiversity conservation and nature provisioning for urban residents. Finally, I conducted research on residential yards in Phoenix, Arizona, to explore the effects of individual management behavior on urban plant community dynamics. Using a front yard vegetation survey repeated at three time points and a paired social survey, I asked, to what extent are yard plant communities dynamic over time, and how do attitudes and parcel characteristics affect native plant landscaping? Front yard woody plant communities experienced high turnover on a decadal scale, indicating that these managed communities are dynamic and capable of change for conservation benefit. Residents held positive attitudes toward native plants, but cultivated few in their yards. Priorities such as desired functional traits, attitudes toward native plants, and household income predicted native plant abundance, while knowledge of native plants did not. This body of work contributes to the growing understanding of how urban ecosystems change over time in response to local- and city-scale impacts, demonstrating opportunities to engage urban residents and land managers in local conservation action to improve the value of cities for people and biodiversity. / Dissertation/Thesis / Doctoral Dissertation Biology 2020