No description available.

Recent years have seen a rapid growth in movement research owing to new technologies contributing to the miniaturization and reduced costs of tracking devices. Similar trends have occurred in how environmental data are being collected (e.g., through satellites, unmanned aerial vehicles, and sensor networks). However, the development of analytical techniques for movement research has failed to keep pace with the data collection advances. There is a need for new methods capable of integrating increasingly detailed movement data with a myriad of contextual data - termed context aware movement analysis (CAMA). CAMA investigates more than movement geometry, by including biological and environmental conditions that may influence movement. However, there is a shortage of methods relating movement patterns to contextual factors, which is still limiting our ability to extract meaningful information from movement data. This thesis contributes to this methodological research gap by assessing the state-of-the art for CAMA within movement ecology and human mobility research, developing innovative methods to consider the spatio-temporal differences between movement data and contextual data and exploring computational methods that allow identification of patterns in contextualized movement data. We developed new methods and demonstrated how they facilitated and improved the integration between high frequency tracking data and temporally dynamic environmental variables. One of the methods, multi-channel sequence analysis, is then used to discover varying human behaviour relative to weather conditions in a large human GPS tracking dataset from Scotland. The second method is developed for combing multi-sensor satellite imagery (i.e., image fusion) of differing spatial and temporal resolutions. This method is applied to a GPS tracking data on maned wolves in Brazil to understand fine-scale movement behaviours related to vegetation changes across seasons. In summary, this thesis provides a significant development in terms of new ideas and techniques for performing CAMA for human and wildlife movement studies.

This thesis presents the findings of an investigation of the spatial and temporal distribution of soil macrofauna at multiple scales within smallholder agriforestry fields in a remote, mountainous area of western Honduras. Since 1990, smallholder farmers in the study area have switched from traditional slash-and-burn agriculture to a form of slash-and-mulch agriforestry based on cultivating maize, beans and sorghum amongst dispersed trees. The principal objective was to examine the influence of the slash-and-mulch agricultural system on soil macrofauna abundance, biomass and community composition, and relate soil macrofauna distribution patterns to environmental variables. The initial stage of the research comprised transect-based sampling of soil macrofauna and biophysical variables in four common land uses of the study area. All four land uses (secondary forest, young milpa (agriforestry), mature milpa, and pasture) supported abundant, diverse and heterogeneous soil macrofauna communities, with few notable differences in soil macrofauna distribution among land uses. The most abundant soil macrofauna taxa were termites, ants, earthworms and beetles. Of the 'explanatory' environmental variables that were measured (including land use and selected soil properties, vegetation characteristics and topographic variables), those that had the strongest relationships with soil macrofauna abundance were land use, tree density and soil organic matter content. The second stage of the research was spatially-orientated and used stratified sampling based on within-field differences in farmer-defined soil type, as well as grid-based sampling of soil macrofauna surface activity. There was substantial within-field variation in soil type and topography, which was related to distribution patterns of at least one agriculturally-important soil macrofauna taxon. Earthworm activity was higher in areas of fertile soil and lower slope positions. At a finer scale, there was a positive spatial correlation between tree distribution and earthworm casting activity. The final phase situated the biophysical research in the local socio-economic context through participant observation and interviews with farmers. The results of the three phases of the study were incorporated into an original conceptual model of the relationships among soil macrofauna and environmental variables in the study area across multiple spatial scales and along a chronosequence of land use changes. Specific pointers are provided for further research on the role of soil fauna in influencing soil structure, nutrient cycling and pest species abundance, and for further investigating local knowledge and the socio-economic and cultural drivers of land use change.

Thesis (doctoral)--Swedish University of Agricultural Sciences, 2003. / Thesis documentation sheet inserted. Includes appendix of four papers and manuscripts, two co-authored with Barbara Ekbom. Includes bibliographical references. Also partially issued electronically via World Wide Web in PDF format; online version lacks appendix of papers.

Recent research has emphasized the importance of understanding the consequences of species loss, not just for biodiversity per se, but also for ecosystem resilience and functioning. Firstly, a baseline analysis of the effects of a naturally patchy landscape on avian community composition and resilience in a high altitude Afrotemperate forest landscape in South Africa is presented. Bird data from a point count survey of 706 counts at 31 forest sites was used to test whether taxonomic species diversity, functional species richness and density of birds provide insight into community resilience in this historically patchy ecosystem. Bird species richness of forest patches ranged from 17 to 38, with a total species richness of 50. Density was slightly but negatively related to both area and species richness. That density compensation is occurring in these Afrotemperate forests suggests a level of resilience in this system. Following on from this, cumulative species-area and function-area graphs were derived to quantify the loss of forest area or taxonomic bird species richness that this landscape may potentially sustain before ecosystem functioning is negatively affected. The concept that species’ patterns of distribution, abundance and coexistence are the result of local ecological processes has recently been challenged by evidence that regionalscale processes are equally instrumental in shaping local community composition. The historically and naturally fragmented Afrotemperate forests of the uKhahlamba– Drakensberg Park, South Africa, offer an interesting test case. In this study the relative effects of local and regional-scale processes on species assemblages in a naturally patchy forest system were investigated. By employing species-area and species-isolation relationships, and nested subset analyses, we showed that isolation (regional-scale process) had a greater effect on bird species richness and composition than area (local-scale process), though the species-area relationship was significant. Using generalized linear models and an information-theoretic approach to model selection, patch area, the size of the regional species pool as well as the distance to the nearest Eastern Mistbelt forest were all influential in determining local bird species richness in these montane forests. Thus, localities are regionally enriched within the constraints on species occupancy provided by the available habitat. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2009.

A large proportion of the world's land surface is covered by semi-arid grasslands, and they provide an important source of income as a grazing resource. A more comprehensive understanding of these complex ecosystems is vital for the effective management of rangelands, as it will lead to an increased and more sustainable economic output. Herbivores modify the spatial pattern of vegetation distribution and their response to spatially heterogeneous forage resources affects their performance. The spatial aspect of herbivory is often ignored although it is a necessary component of understanding grazing dynamics and the factors affecting herbivore condition. A spatial model is developed which incorporates vegetation and animal dynamics and the interactions between these two components. The effect of different spatial foraging strategies on animal performance and vegetation was investigated. Model results were compared with the output of a non-spatial model to assess the importance of spatially explicit modelling in the context of monitoring animal performance. The relative significance of a number of aspects relating to spatial grazing and animal condition was explored. The results from this research show that significant differences in output are obtained from spatial versus non-spatial models. While the purpose of a model will determine its nature, the results imply that in certain contexts, a spatial model is essential for accurate results and insight. The results also indicated that foraging strategies have a large affect on herbivore condition and that spatially explicit models are necessary in the context of investigating the effect of foraging strategies on animal performance. Various aspects that significantly affected animal condition were highlighted and are useful in directing future investigations into grazing dynamics. It is difficult to conduct field studies under spatially and temporally variable conditions where the interactions between vegetation and herbivores are so complex. In the light of this, modelling was found to be an effective tool that can be used in investigating and revealing important dynamics of semi-arid grazing systems. / Thesis (M.Sc.)-University of Natal, Pietermaritzburg, 1999.

Thesis (M.S.)--University of Tennessee, Knoxville, 2008. / Title from title page screen (viewed on Sept. 23, 2009). Thesis advisor: Frank T. van Manen. Vita. Includes bibliographical references.

Thesis (M.Sc.(Zoology))--University of Pretoria, 2000. / Abstract in English. Includes bibliographical references.

Tous les hôtes ne contribuent pas également à la transmission de parasites. Certains individus ou espèces peuvent par exemple être davantage infectés que d'autres, une observation qui a mené à la proposition de la règle des `20/80', selon laquelle 20 % des individus seraient responsables de 80 % de la transmission. Cependant, les études qui se sont intéressées à l'hétérogénéité de la transmission se sont principalement focalisées sur les sources d'hétérogénéité intrinsèques à l'espèce ou à l'individu, telles que la susceptibilité ou l’infectivité, tandis que les facteurs extrinsèques, comme la connectivité entre espèces au sein de la communauté d'hôtes et le rôle de différents types de mouvements des hôtes ont été relativement négligés. Dans ce contexte, cette thèse aborde le rôle des causes extrinsèques de l'hétérogénéité de transmission sur la propagation d'infections dans les systèmes multi-hôtes, en utilisant notamment les systèmes tiques-oiseaux marins-microparasites comme support empirique à des approches de modélisation théorique. Quatre principales sources d'hétérogénéité dans les systèmes à transmission vectorielles ont ainsi été considérées : (i) l'hétérogénéité de l'abondance des vecteurs, de leur distribution, et l'estimation des paramètres de la dynamique de leurs populations, (ii) l'hétérogénéité de contact entre espèces de communautés multi-hôtes et multi-vecteurs, (iii) l'hétérogénéité de la propagation d'infections en raison de différents types de comportements des hôtes (avec en particulier, l'importance de considérer les mouvements de prospection entre groupes d'hôtes chez les espèces sociales) et (iv) l'hétérogénéité dans les capacités de dispersion et de transmission d'infections entre vecteurs à traits d'histoire de vie contrastés (dispersion en fonction du stade de vie). Nous soulignons d'abord l'importance potentielle d'une estimation fiable des abondances d'ectoparasites, à l'aide d'approches hiérarchiques susceptibles de prendre en compte à la fois l'hétérogénéité de leur probabilité de détection et leur distribution agrégée. Ensuite, nous utilisons une approche permettant d'étudier l'impact des caractéristiques du réseau d'interactions au sein de la communauté d'hôtes sur la transmission et le maintien d'infections. Nos résultats indiquent que la structure de la communauté mais aussi les propriétés locales des espèces modèlent l'émergence d'espèces qui contribuent disproportionnellement à la transmission de l'infection (`superspreader') et d'espèces qui contribuent disproportionnellement au maintien de l'infection (`keystone') dans les communautés d'infections multi-hôtes, multi-vecteurs. Nous avons également exploré le rôle de la contribution de différents comportement de déplacement des hôtes et des traits d'histoire de vie des vecteurs sur la propagation d'agents infectieux. Une revue de la littérature nous a permis de souligner l'importance potentielle, relativement aux autres comportements de déplacement plus communément considérés, des mouvements de prospection entre groupes d'hôtes sur le rôle dans la transmission d'infections. Les résultats d'un travail théorique nous on également permis de montrer l'importance des caractéristiques des traits d'histoire de vie des vecteurs (notamment la durée de repas sanguins) et des contraintes démographiques (effet Allee) sur le potentiel de colonisation des tiques. Cette différence de dispersion en fonction du stade est ainsi susceptible d'avoir une incidence sur la propagation d'infections à transmission vectorielle et la structure génétique des populations de tiques. Dans l'ensemble, les travaux menés ont permis de mettre en évidence l'importance de l'étudie des déterminants des hétérogénéités de transmission et leurs conséquences dans les systèmes à transmission vectorielles, pour une meilleure compréhension de l’écologie et l’évolution des interactions entre hôtes et parasites, avec des implications potentielles pour le contrôle des maladies. / Different hosts may not contribute equally to parasite transmission. For instance, some individuals or species may be more heavily infected than others, an observation that lead to the `20/80' rule, stating that in many cases 20% of individuals are responsible for 80% of the transmission. However, studies on heterogeneity in transmission have primarily focused on intrinsic factors of transmission, such as susceptibility and infectivity, while the impact of extrinsic factors, such as connectivity network among individuals or species of the host community and the role of various host movements has been relatively neglected. This thesis investigates the role of extrinsic transmission heterogeneities on the spread of infectious disease in multi-host systems, using tick-seabird-microparasite system as empirical models for theoretical investigations. Four main causes of heterogeneity in transmission of vector-borne diseases were considered : (i) heterogeneity in vector abundance, distribution, and estimation thereof (ii) heterogeneity in contact among species in a multi-host, multi-vector community, (iii) heterogeneity in infection spread caused by different host mouvement behaviors (notably the potential role of ‘prospecting’ by host individual among host groups), and (iv) heterogeneity in dispersal ability and transmission competence among vectors with different life-history traits (stage-dependent dispersal). First, we highlight the need to accurately estimate ectoparasite abundances with hierarchical modeling approaches that can take into account both heterogeneity in their detection probability and their aggregated distribution among hosts. Next, using network theory to examine the impact of community context on disease transmission and maintenance, we found that network structure (modularity, nestedness) and node-based measures (e.g., centrality) both shape the emergence of ‘super-spreader’ species (i.e., species that contribute disproportionally to disease transmission) and keystone species (i.e., species that contribute disproportionally to disease maintenance) in multi-host, multi-vector pathogens communities. Finally, we explored the contribution of host behavior and vector life-history traits to the spread of infectious agents. By reviewing the recent literature, we highlight the fact that prospecting, relative to various other types of host movement, may be of key importance to disease transmission among host groups, notably in social species. We also show how vector life history characteristics (e.g. length of bloodmeals) and demographic constraints (Allee effects) affect their colonization potential. Soft ticks, which take a single, long bloodmeal at only the larval stage, should have much lower colonization rates than hard ticks, which take a single, long bloodmeal at every life stage. These stage-dependent dispersal discrepancies may have direct consequences for the genetic structure of their populations and the spread of vector-borne infectious agents. Overall, these findings highlight the importance of studying the causes and consequences of transmission heterogeneity in multi-host, multi-vector systems. A series of potentially important sources of heterogeneity in parasite transmission are outlined, together with perspectives of empirical and theoretical studies to further explore their implications for understanding ecology and evolution of host-parasite interactions and for disease management purposes.

The Cape Vulture (Gyps coprotheres), a southern African endemic, is a species currently listed on the IUCN Red List as “Threatened” due to its 66-81% decline over the past 50 years. South Africa and Lesotho hold 90% of the global population, which are focused at two core areas, namely northern central South Africa and eastern South Africa. This species provides important ecological services yet faces numerous anthropogenic threats. An emerging threat in the south eastern part of South Africa is that of wind energy development. Understanding the movement and roosting behaviour of the Cape Vulture may mitigate potential collisions if areas of high use are avoided. Juvenile Cape Vultures, who naturally suffer high mortality rates, are known to forage extensively over a wider landscape and as a consequence may face a greater assortment of threats. The overall aim of this study therefore was to determine the ranging and roosting behaviour of juvenile Cape Vultures in the Eastern Cape Province, South Africa. In order to achieve this aim, i) the ranging behaviour and habitat use were determined and ii) roost sites and factors that influence roost site selection were determined. Using five juvenile Cape Vultures tagged with Global Positioning System (GPS/GSM) transmitters, home range sizes, distance travelled from the nest and habitat use were determined. Home range was determined through kernel density estimates and distance travelled from the nest was determined through the Euclidean distance. Habitat use was determined by overlaying the home range size onto a merged layer of all protected areas and the National Land Cover Database of South Africa. To determine roost sites, data from the tagged vultures were split into encamped and exploratory movements using a mixture model in a cluster analysis setting. Encamped movements were associated with roost sites. Roost density was determined around the natal colony using predetermined buffer sizes and a generalised linear mixed model (GLMM) fitted to the data. Conditions considered favourable for roost sites were based on previous research conducted on cliff-nesting species and a GLMM conducted. Juveniles increased their home range progressively for the first two months, then exhibited a rapid increase in size associated with dispersal from the colony. Distance from the nest increased rapidly following the dispersal period. Protected areas and woody vegetation were areas of preferred habitat. The highest density of roosts for juveniles was located within 20 km’s from the breeding colony and decreased further away. Roost sites that were favoured for juveniles and adult birds were those that were located close to colonies, had low accessibility to terrestrial predators and were in areas of high wind speed. The orientation of the cliff into the prevailing wind direction was also a determining factor. Roosting sites and foraging areas are important spatial determinants of Cape Vulture behaviour and the identification of such areas can help with conservation management. With the additional threat of wind development in areas highly utilised, wind farms located too close to colonies could have a devastating impact on the Cape Vulture population. Therefore, this study highlights the importance of conservation buffer zones as no-go areas for wind energy development around vulture colonies.