Landscape modification, especially those human-dominated changes, has been widely considered as one of the key drivers inducing the degradation of environment and ecosystem. The consequent habitat loss and fragmentation, which are closely related to ecological process, have attracted much scholarly attention. However, very few studies have systematically examined their causes and consequences at the landscape level. Integrating human activities into consideration in a holistic way remains poorly understood in particular. This thesis reports a systematic study of habitat loss and fragmentation under urbanization. This research distinguishes itself from previous works in its investigation of the spatio-temporal dynamics of causes and consequences of habitat loss and fragmentation, and its endeavor to disentangle habitat fragmentation from simple habitat loss in terms of their processes and effects. The Shenzhen River cross-border watershed, shared by Hong Kong Special Administrative Region and Shenzhen, is used as a case to illustrate how these ideas can work in practice. The study period spans from 1988 to 2008.The main contents of the study include analysis of the processes and causes of habitat loss and fragmentation, and dynamic consequences on water quality and ecosystem services for biodiversity conservation at spatial and temporal scales. Various methods are employed, mainly including conventional and newly developed landscape metrics (such as the Urbanization Isolation Effect (UIE) and Habitat Isolation Degree (HID)), a spatial statistical approach based on the Geographically Weighted Regression (GWR) model, spatio-temporal analyses and cross-border comparison. Major findings of the research are highlighted as follows. First, habitat loss and fragmentation show great spatial heterogeneity as well as temporal variation due to extensive and rapid urban sprawl in the study area. Second, by adopting the GWR model, it is found that human-related factors have played a significant role in determining not only the “amount” but also the “location” of habitat loss. Third, by using two new landscape metrics, namely the UIE and HID, both the extent (urban sprawl) and intensity (population density) of urbanization are suggested to have significant influences in habitat isolation. Fourth, water quality in the watershed is recognized to be closely correlated with habitat loss and fragmentation, and the correlations show major discrepancies between the Hong Kong and the Shenzhen sides. Moreover, by proposing a compound indicator (NP/MPS) for measuring habitat fragmentation (independent of simple habitat loss), it is found that habitat loss has exerted a bigger influence on water quality degradation than habitat fragmentation. Fifth, by integrating habitat connectivity into the evaluation of ecosystem services for biodiversity conservation, both the habitat size and connectivity are recognized to be important in explaining the degradation of ecosystem services value. Habitat size has played a dominant role in the overall change of ecosystem services value, whereas the role of connectivity (fragmentation) is more critical when the “stepping stones” sites are lost. The study provides a new framework to examine the habitat loss and fragmentation under urbanization at the landscape level. The efforts show great effectiveness in providing sound and useful information for future landscape planning, biodiversity conservation and watershed management. / published_or_final_version / Geography / Doctoral / Doctor of Philosophy

Forest fragmentation poses a serious danger to population diversity in plants and animals by increasing species isolation, thus reducing the population size and genetic diversity. However, little information is available concerning how fragmentation impacts plant diversity. AFLP fingerprinting was used to assess genetic diversity within and between populations of Asarum canadense L. (Canadian Wild Ginger) across 11 different populations in East-Central Indiana. AFLP fingerprints using two primer pairs generated 51 distinct bands with an average of 25.5 bands per primer. Forty-eight low molecular weight distinct polymorphic bands were observed (50-200 bp range). The percentage of polymorphism was low (0-25%) indicating low levels of genetic diversity within each population studied. NTSYSpc Numerical Taxonomy Analysis Software generated aphenogram that revealed high levels of homologies within populations (75-100%), with individuals from the same population typically clustered. The genetic diversity between populations ranged from 10-50%. The populations from Jay, Randolph and Henry Counties clustered together exhibiting -54% homology, while populations from Mien, Madison, and Huntington counties shared approximately 64% homology. The populations from Adams, Blackford, Delaware, and Grant counties shared approximately 66% homology. However, within this last group Blackford and Delaware counties shared 90% homology. There were no apparent effects of the size of the forest fragments on the observed diversity measures. A possible relationship between genetic diversity and spatial distance was observed between populations moving from east to west. Possible reasons for this observation may be due to forest types, age of forests, climatic factors, soil types, and/or anthropogenic activities. Overall, the low level of average diversity within the populations strongly suggests that the individuals inhabiting isolated forests primarily propagate by asexual means.Ball State UniversityMuncie, IN 47306 / Department of Biology

Thesis (Ph. D.)--Washington State University, August 2008. / Includes bibliographical references.

Thesis (Ph. D.)--Oregon State University, 2006. / Printout. Includes bibliographical references (leaves 98-104). Also available on the World Wide Web.

Thesis (Ph. D.)--University of Canterbury, 2008. / Typescript (photocopy). Includes bibliographical references (leaves 148-172). Also available via the World Wide Web.

Assemblages of algae are altered by both bottom - up ( e.g. nutrient availability ) and top - down ( e.g. herbivory ) processes. As a result of the increasing human population in coastal areas, massive changes are forecast to benthic habitats in response to increasing coastal nutrient concentrations and a reduction in consumers. To identify the scales over which nutrients may have an effect, abundance of turf - forming algae growing as epiphytes on kelp ( Ecklonia radiata ) were related to water nutrient concentration across temperate Australia. In general, the percentage cover of epiphytes was greatest at sites with the greatest nutrient concentrations. By experimentally elevating mean nitrate concentration from the low 0.064 ± 0.01 µmol L [superscript - 1 ] to 0.121 ± 0.04 µmol L [superscript - 1 ], which was still only ~ 5 % of that measured on a more eutrophic coast, I was able to increase the percentage cover of epiphytes to match those seen on nutrient rich coasts, despite not matching the nutrient concentrations on those coasts. Hence, it appears that the effects of elevated nutrients will be disproportionately large on relatively oligotrophic coasts. Nutrient concentrations were also experimentally elevated to test whether the presence of an algal canopy or molluscan grazers were able to counter the effects of nutrient enrichment on algal assemblages. The loss of canopy - forming algae is likely to be a key precursor to nutrient driven changes of benthic habitats, because nutrients had no direct effect on algal assemblages in the presence of canopy - forming algae. In the absence of canopy - forming algae, space was quickly monopolised by turf - forming algae, but in the presence of elevated nutrients grazers were able to reduce the monopoly of turf - forming algae in favour of foliose algae. This switch in relative abundance of habitat may reflect greater consumption of nutrient rich turf - forming algae by grazers, possibly creating more space for other algae to colonise. Importantly, greater consumption of turf - forming algae in the presence of elevated nutrients may act as a mechanism to absorb the disproportionate effect of nutrients on oligotrophic coasts. In southern Australia, canopy - forming algae have a negative impact on the abundance of turf - forming algae. To assess the mechanisms by which an algal canopy may suppress turf - forming algae, abrasion by the canopy and water flow were experimentally reduced. Abrasion by the canopy reduced the percentage cover and biomass of turf - forming algae. In contrast to predictions, biomass and percentage cover of turf - forming algae were also reduced when water flow was reduced. Light intensity was substantially reduced when there was less water flow ( because of reduced movement in algal canopy ). However, the reduction in available light ( shading ) did not account for all of the observed reduction in biomass and percentage cover of turf - forming algae, suggesting that other factors are modified by water flow and may contribute to the loss of turf - forming algae. Habitat loss and fragmentation are well known to affect the diversity and abundance of fauna in habitat patches. I used experimental habitats to assess how fragmentation of turf habitats affects the diversity and abundance of two taxa of macroinvertebrates with different dispersal abilities. I established that increased isolation of habitats reduced the species richness and abundance of invertebrates with slow rates of dispersal, while the species richness and abundance of invertebrates with fast rates of dispersal were greatest in habitats that were far apart. In summary, this thesis provides an insight into some of the impacts associated with human populations in coastal areas, namely increased nutrient inputs, loss of grazers ( e.g. harvesting ), and loss of canopy algae and fragmentation of habitats. I show that increased nutrient concentrations in coastal waters can alter the relative abundance of algal species, and that some effects of elevated nutrients can be absorbed by the presence of grazers. I also show that elevated nutrients have no effect on algal assemblage in the presence of canopy - forming algae, and that canopies can suppress the colonisation of turf - forming algae. Finally, I show that the fragmentation of turf habitats affects taxa of invertebrates with different dispersal abilities in different ways. Whilst the contemporary ecology of much of the temperate Australian subtidal coast is considered to be relatively unaffected by human activity, this thesis shows that changes to top - down and bottom - up processes could have large consequences for habitats and their inhabitants. / Thesis (Ph.D.)--School of Earth and Environmental Sciences, 2005.

Thesis (Ph. D.)--University of Missouri-Columbia, 2000. / Typescript. Vita. Includes bibliographical references. Also available on the Internet.

Thesis (Ph. D.)--University of Missouri-Columbia, 2000. / Typescript. Vita. Includes bibliographical references. Also available on the Internet.

Despite increasing publicity, habitat loss and fragmentation remain a serious threat to biodiversity. The main objectives of this research were (i) to study the effects of forest fragmentation on the distribution and abundance of resident birds in the fragmented littoral forests of southeastern Madagascar and temperate woodlands of southeastern Australia at three spatial scales (patch, landscape and regional) and (ii) to place the results of these case studies within the realms of the equilibrium theory of island biogeography and its descendent theories, to ascertain whether it is appropriate to use these theories to derive conservation scenarios within these threatened regions. Deforestation of Madagascar's remaining forests is considered a global concern due to both its current high intensity and the small amount of forest claimed to be remaining on the island. Surprisingly, very few studies have considered the impacts of forest fragmentation on bird diversity in Madagascar. A multi-scale study on the effects of littoral forest fragmentation and degradation on birds is therefore a major step forward for bird conservation on the island. Furthermore, prior to this study no known work has been conducted on the avifauna within the threatened littoral forests of eastern Madagascar. My results indicated that (i) the littoral forests contained bird species assemblages that were unique when compared to neighbouring forest habitats, (ii) many forest-dependent bird species were significantly affected by habitat structure and especially proximity to forest edge and (iii) many forest-dependent species were affected by landscape factors such as remnant shape and remnant size. No relationship was found between measures of landscape composition, remnant 'isolation' and bird distribution within littoral forest remnants.

Despite intense publicity, habitat loss still remains a serious threat to biodiversity. Forest destruction is its frontrunner, both in terms of physical habitat under threat and potential for biodiversity loss. In the fragmented landscape of the Taita Hills, SE Kenya, several bird species are facing the threat of extinction from forest loss. They are absent from many of the remnant forest patches and/or are showing negative effects with increasing disturbance. Using a relatively common forest-dependent bird species - the whitestarred robin Pogonocichla stellata - as a model, the current status of this ecosystem was examined, and future patterns predicted in view of the unrelenting destruction. As expected, the robin population in the largest and most intact fragment (c35 ha) was the healthiest, suggesting that this was indeed the best quality habitat patch: it had the highest population density, highest productivity (low nest predation and high juvenile to adult ratio) and lowest turnover rates. Effects of forest deterioration were evident from the fact that the medium-sized patch (c95 ha), which is undergoing severe degradation, was a worse habitat for the robin than the tiny patches (c2-8 ha): it had the lowest population density, lowest productivity (highest nest predation rates and lowest juvenile to adult ratio), and highest turnover rates. The explanation for this is twofold. Besides the smallest patches facing lower levels of habitat loss recently, they also had high levels of dispersal between them. They occasionally operated as a finegrained system with individuals moving between them in the space of a few days. In general, the robin metapopulation is demographically (rate of change, λ = 0.996) and genetically (at migration- and mutation-drift equilibrium) stable at present. The populations in the largest and smallest patches were potential sources providing emigrants that were possibly crucial in sustaining the population in the medium-sized patch (given its low productivity and high turnover rates). Overall, these findings underscore the importance of within-patch processes, both for ensuring persistence of subpopulations and providing dispersers, as well as between-patch processes (chiefly dispersal) for ensuring metapopulation persistence. Thus, by furnishing ample sample sizes that enabled work to be carried out in all fragments throughout this landscape, the model species approach was useful for identifying the need for a two-pronged conservation strategy. First, a need to focus within fragments to reduce habitat loss and degradation, and second, to address among fragment issues relating to land-use and maintaining a forested landscape, in order to enhance connectivity between patches. Finally, based on the mechanisms by which disturbance and fragmentation are affecting bird populations e.g. predator influxes from the surrounding matrix, conservation recommendations for the Taita Hills are offered.