La compréhension des processus écologiques qui sous-tendent l’assemblage des communautés végétales et la coexistence des espèces est un objectif central en écologie. Ces processus sont potentiellement nombreux et de natures contrastées. Ainsi, la composition d’une communauté de plantes dépend de processus déterministes liés aux conditions environnementales abiotiques (climat, conditions physiques et chimiques du sol, lumière) et d’interactions biotiques complexes, positives (facilitation, symbioses) comme négatives (compétition, prédation, pathogènes). En outre, les communautés sont influencées par des processus stochastiques (capacité de dispersion limitée, dérive écologique). Si les mécanismes à l’origine de ces processus sont très différents, ils ont néanmoins en commun la génération de motifs (patterns) spatiaux de distribution d’espèces dans les communautés. L’analyse de la structure spatiale des communautés permet ainsi une étude indirecte des processus régissant les communautés. La nature complexe de ces patterns spatiaux a mené au développement de nombreuses méthodes statistiques de détection et de description de patterns. Les méthodes basées sur des vecteurs propres spatiaux sont parmi les plus puissantes et précises pour détecter des patterns complexes et multi-échelles. Ces vecteurs propres, utilisés comme prédicteurs spatiaux, peuvent être combinés à un ensemble de variables environnementales dans un cadre de partition de variation. Celui-ci permet, en théorie, de démêler les effets uniques et l’effet conjoint des variables environnementales et spatiales sur la variation de composition d’une communauté. Il mène ainsi à une quantification de l’action des processus déterministes et des processus stochastiques sur l’assemblage de la communauté. Néanmoins, je montre dans cette thèse qu’un certain flou méthodologique concernant deux étapes déterminantes des analyses basées sur les vecteurs propres spatiaux a mené une proportion élevée d’études à utiliser ces méthodes de manière sous-optimale, voire fortement biaisée. Ceci compromet la fiabilité des patterns spatiaux détectés et des processus écologiques inférés. Une autre limitation de ce cadre d’analyse concerne la fraction de la partition de variation décrivant l’effet environnemental spatialement structurés qu’aucune méthode ne permet de tester.Cette thèse présente des solutions non biaisées, puissantes et précises à ces différentes limitations méthodologiques et permet d’élargir le cadre de l’inférence de processus écologique à partir de patterns spatiaux de communautés. Les différentes étapes d’amélioration de ces méthodes ont également été illustrées dans la thèse au travers de trois cas d’études fournis par deux communautés d’arbres tropicale et tempérée et une communauté de champignons symbiotiques des arbres. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

The savanna woodlands of Southern Africa, colloquially termed the miombo, are poorly described in terms of biodiversity compared to other biomes. They have therefore been underrepresented in the wider understanding of how land use intensification is shaping global biodiversity. Land use change is known to reduce biodiversity and disrupt intactness of ecological communities with consequences for ecosystem functioning, resilience, and services. Miombo woodlands are described as biodiversity hotspots due to a high endemism of species and the presence of megafauna. At the same time, they are also considered dynamic socio-ecological systems shaped by disturbances and the land use activities of people. The patterns of biodiversity change in these tropical ecosystems may, therefore, have their own unique contexts, understanding of which will be essential for biodiversity and land use management in these ecosystems. In this thesis, I identified the patterns of biodiversity change in response to the two major land use practices in the two dominant woodland types in southern African woodlands: the selective logging due to charcoal production in the mopane woodlands, and agricultural expansion in the miombo. I also examined the impact of two main disturbance agents, humans and elephants, on habitat structure and biodiversity in mopane woodlands. Across all chapters in this thesis, I investigated the effects of land use change and habitat modification on biodiversity empirically using chronosequences. To understand biodiversity change, I employed a hierarchical multilevel modelling approach making inferences at the three levels of ecological communities: species, community, and meta-community (set of ecological communities at different sites). I selected six villages in the charcoal production hotspot of southern Mozambique and carried out field surveys for three taxonomic groups: trees, mammals and ground beetles. I modelled the counts of trees and beetles and incidence of mammals using meta-community occurrence models in a Bayesian framework with the intensity class of the villages, above-ground biomass and land cover type as predictors. The results suggested that the species richness of trees and mammals declined by 12 and 8.5 % respectively while that of beetles increased by 3.5%, albeit non-significantly. In addition, the beta diversity of trees decreased while that of mammals increased. The results show that while both trees and mammals reduced in richness, they responded differently to charcoal production in terms of community organisation. The trees underwent subtractive homogenisation (decrease in alpha and beta diversities) primarily because of deterministic processes induced by selective harvesting of tree stems for charcoal. Mammal communities, on the other hand, showed subtractive heterogenization (decrease in alpha, but increase in beta diversity) mainly due to random extinctions. In the agriculture frontier of miombo-dominated northern Mozambique, I investigated the effects of fragmentation and habitat loss caused by agricultural expansion on diversity and composition of trees and mammals. I modelled the occurrences of trees and mammals using occupancy models with the fragmentation and quantity of woodland cover as predictors. The model showed that most tree species (n=10), mainly the timber and firewood species, linearly declined in population size as fragmentation increased. Mammals, on the other hand, showed a nonlinear response. Seven mammal species increased at the lower levels of fragmentation. However, at the higher levels, none of the mammal species increased while two declined. Similarly, the species richness of trees linearly declined, while that of mammals increased up to a fragmentation level of 55-65% and declined above this limit. The beta diversity of trees increased with fragmentation while that of mammals decreased. The results suggest that, although fragmentation reduces species richness of both trees and mammals, it affects their species compositions in different ways. Trees undergo subtractive heterogenization due to random species losses while mammals experience subtractive homogenisation mainly due to the combined effects of fragmentation-led habitat loss and intensified hunting. Finally, this study concludes that, above 75% fragmentation or below 26% habitat quantity, both taxonomic groups endure biodiversity loss. The threshold results here corroborate similar habitat quantity thresholds (20-30%) observed elsewhere in different ecosystems. However, they differ with the widespread notion that above 30% habitat quantity, the effect of fragmentation is non-existent. The results here emphasize that taxonomic groups respond differently, the diversity and population size of mammals reduced only after the habitat threshold, whereas, those of trees showed linear decrease with fragmentation most likely due to fragmentation-led habitat loss. Lastly, I examined the effects of disturbance by humans and elephants on habitat structure and bird diversity by conducting a space for time substitution comparison in the mopane woodlands of Zambia. To examine the woodland structure, I modelled the structural attributes of habitat (stem diameter, stand density, and basal area) using mixed models with the proportion of affected stems by humans and elephants as explanatory variables. I found that elephant disturbance was associated with higher stem diameters, low stand densities, but no change in basal area. Human disturbance, on the other hand, was related to reductions in stand density and basal area, but no change in the stem diameter. Further, I tested species and functional diversity of birds against the covariates of habitat structure and disturbance. I found that bird communities reduced in species richness in both, human as well as elephant disturbed areas. However, the functional diversity did not change with elephant disturbance. I concluded that human disturbance reduces woody biomass (basal area is correlated with woody biomass) of mopane woodlands and functional diversity of birds whilst elephants do not. In this thesis, I conclude that human driven land use change in the miombo woodlands erodes alpha diversity of all taxonomic groups. However, increases in beta diversity of mammals with charcoal land use and trees in agricultural land use may maintain their diversities at the meta-community level.

Information on deforestation and forest management is limited due to lack of understanding as to how socioeconomic factors affect deforestation. Without adequate data, it is difficult to manage deforestation; particularly in Chipuriro where deforestation is occurring at a rapid rate. This research was carried out to address the problem. Surveys and landsat images were used to collect data. To quantify deforestation, images were processed using Geographical Information System. Pearson’s Correlation Coefficient was used to deduce the relationship between socioeconomic factors and deforestation. Socioeconomic factors were obtained through household surveys and focus group discussions. Results indicated high rates of deforestation in Ward 18 (4.7% per annum) while in Ward 16 woodland cover increased throughout the study period (7.7% per annum). Population growth with its demand on fuel wood, settlement and agricultural land caused a significant decrease in woodland cover in Ward 18. Plantations increased the area under woodlands in Ward 16. / Environmental Sciences / M. Sc. (Environmental Management)

Information on deforestation and forest management is limited due to lack of understanding as to how socioeconomic factors affect deforestation. Without adequate data, it is difficult to manage deforestation; particularly in Chipuriro where deforestation is occurring at a rapid rate. This research was carried out to address the problem. Surveys and landsat images were used to collect data. To quantify deforestation, images were processed using Geographical Information System. Pearson’s Correlation Coefficient was used to deduce the relationship between socioeconomic factors and deforestation. Socioeconomic factors were obtained through household surveys and focus group discussions. Results indicated high rates of deforestation in Ward 18 (4.7% per annum) while in Ward 16 woodland cover increased throughout the study period (7.7% per annum). Population growth with its demand on fuel wood, settlement and agricultural land caused a significant decrease in woodland cover in Ward 18. Plantations increased the area under woodlands in Ward 16. / Environmental Sciences / M. Sc. (Environmental Management)

Dans la forêt boréale, l’augmentation de la fréquence et de la superficie d’îlots de pessières à lichens sur le territoire québécois a déjà été observée et pourrait résulter en une migration vers le sud de la limite nordique des pessières à mousses. Ce phénomène survient après des échecs de régénération, qui ont lieu lorsque le milieu est préalablement fragilisé lorsqu’une nouvelle perturbation affecte le peuplement. Avec la possibilité de détecter ce stress en analysant les propriétés spectrales de la végétation, les zones perturbées pourraient alors être identifiées. L’objectif principal de la présente étude est d’établir des liens entre d’une part, les informations extraites des signatures spectrales et d’autre part, les indices de végétation et les différents types de stress affectant les écosystèmes boréaux. Cela permettra de savoir s’il est possible d’identifier les pessières à mousses à risque de subir un accident de régénération en étudiant les propriétés spectrales de la végétation comme indicateur de stress. Pour répondre à cet objectif, des sites d’échantillonnage ont été positionnées aux 51e, 52e et 53e parallèles le long de la route de la baie James. Les placettes ont été regroupées en paires afin de faire des tests appariés et ainsi comparer les deux types de peuplements. Sur le terrain, les signatures spectrales ont été prises sur les feuilles aléatoirement prélevées sur cinq épinettes noires. Ces mesures ont été prises tout au long de la saison de croissance (3 campagnes d’échantillonnage). Quatre indices de végétation (NDVI, NDWI, PRI et SIPI) ont été extraits des signatures spectrales, et la pente moyenne du red-edge a été calculée. Les résultats obtenus ont permis de déterminer que certaines des pessières à mousses ont des valeurs très proches de celles des pessières à lichens, qui sont considérées comme des écosystèmes stressés. À partir de ces résultats, il est possible de supposer que le stress peut également être identifié à l’échelle du paysage (sur les images satellitaires) et ainsi permettre un suivi et une gestion après les feux et les épidémies afin de limiter les pertes de ce précieux écosystème. / In the boreal forest of northern Québec, the size and quantity of lichen woodlands patches is increasing, and taking over the spruce-moss forest territory. The phenomenon has been observed, and scientists now believe that the northern limit of the spruce-moss forest will slowly move south. This shift of ecosystem happens when the forest stand is already fragilized, and a perturbation occurs. Vegetation’s spectral properties can be used as a tool to assess and identify disturbed forest areas The main objective of this study is to establish relations between data extracted from spectral signatures, vegetation indexes and different types of stress that could affect boreal ecosystems. The identification spruce-moss woodlands prone to regeneration failure could be achieved with the study of spectral properties as stress indicators. In order to achieve this objective, sites from 3 latitudes (51, 52 and 53) have been sampled on James Bay Road. Plots have been regrouped in pairs for subsequent pairwise statistical tests to compare results from both forest stand types. Spectral signatures have been measured on 5 randomly chosen black spruces. These measurements were taken throughout the growing season (3 sampling campaigns). Four vegetation indexes have been extracted from spectral signatures (NDVI, NDWI, PRI and SIPI), and the mean slope of the red-edge area have been calculated. Results have shown that some of the spruce-moss stands have had very similar values to those from the lichen woodlands, that are considered as stressed ecosystems. From these results, it is possible to assume that stressed ecosystems can be detected at landscape level (on satellite images). Monitoring vegetation stress can help improve forest management after forest fires and insect’s epidemics to prevent the loss of this beautiful ecosystem.

<p>Here, we seek to address the importance of biodiversity in plant ecosystems. We examined the productivity-diversity relationship through the lens of the modern coexistence theory, using a combination of both experimentation and mathematical simulation. We did this by tracking and comparing the productivity of mixed and monoculture plots, analyzing the growth responses of individual trees at forest plots (Chapter 1), confirming the productivity-diversity relationship in a greenhouse experiment using local herbaceous plants (Chapter 2), and finally simulating the productivity response of monoculture vs polyculture plantations to specialist enemy attack (Chapter 3).</p><p>It is no surprise that biodiversity has been decreasing at an exponential rate on the global scale because of effects such as habitat fragmentation, invasive species, spreading pathogens, and anthropogenic influences. Ecologists often found that plants in more species rich locations often exhibited higher productivity and stability in the face of stress. One such phenomenon is known as the productivity diversity relationship that implies biodiversity is key to sustaining ecosystems. Notably, while efforts are being put forth to address ecosystem destruction, much of the current tree planting strategy in the USA is based on timber profit rather than forest productivity and species coexistence with tree biology often being a secondary consideration. These thought processes are in opposition with historical experiments that indicate polyculture communities create more biomass making them significantly more productive than monocultures. However, we also acknowledge that it is not simply biodiversity that must be taken into consideration for a productive ecosystem but also species interaction through coexistence indicate whether or not a community will persevere. These interactions can be addressed using the modern coexistence theory which depends on these complementarity and fitness similarities for species to coexist through time. Here, we seek to address the importance of biodiversity in plant ecosystems. We examined the productivity-diversity relationship through the lens of the modern coexistence theory, using a combination of both experimentation and mathematical simulation. We did this by tracking and comparing the productivity of mixed and monoculture plots, analyzing the growth responses of individual trees at forest plots (Chapter 1), confirming the productivity-diversity relationship in a greenhouse experiment using local herbaceous plants (Chapter 2), and finally simulating the productivity response of monoculture vs polyculture plantations to specialist enemy attack (Chapter 3). Our research across the combination of approaches used found that species with overlapping niches and very different finesses will exclude one another due to high competition. Further, the productivity diversity correlation is necessary for ecosystem growth, but it is not sufficient for species coexistence. However, species can maintain this positive relationship despite a lack of coexistence if they maintain niche complementarity. Lastly, using a theoretic game model we were able to identify the impacts of a specialist pest on polyculture and monoculture forest. These results showed that a polyculture forest was more productive than that of a monoculture forest regardless of the presence of a specialist enemy. The results of the multiple threads of evidence found from these combined experiments indicate that while the productivity diversity correlation is important to ecosystems it is likely due to the impacts of niche complementarity that determine whether or not species will be productive within an ecosystem.</p>

Graduation date: 2006 / Habitat loss causes a reduction in available resources for wildlife, alters the configuration of remaining habitat, and may isolate wildlife populations. White-breasted nuthatches (Sitta carolinensis) are experiencing long-term population declines in the Willamette Valley of Oregon, where they are historically associated with oak woodlands. As secondary cavity-nesters, white-breasted nuthatches may be limited by the availability of existing cavities for nesting and roosting. Oak vegetation in the Willamette Valley has changed since European-American settlement times from vast areas of open oak savanna to isolated closed-canopy stands separated by agricultural fields. We examined nuthatch density, nest cavity selection, and nest success in relation to oak woodland structure and landscape context. We conducted point transect surveys in 3 strata: woodland interiors, large woodland edges, and small woodlands. We located and monitored nuthatch nests and sampled vegetation at nest locations and matching random locations around each nest. Woodland structure and edge density were measured at a 178-m radius (home range) scale, and landscape context was measured using vegetation cover within a 1-km radius around point transects and nests. We used program DISTANCE to fit detection functions and calculate nuthatch densities. We used conditional logistic regression to compare nest locations with random locations, and analyzed nest success with Mayfield logistic regression. White-breasted nuthatch density was significantly higher in small woodlands than in edges of large woodlands, which had higher nuthatch density than woodland interiors. Density of nuthatches increased with a combination of oak cover within a 1-km radius of the point, edge density within a 178-m radius, and number of oak trees >50 cm diameter at breast height (dbh) within a 100-m radius. Nest cavities were situated in oak trees containing more cavities than random oak trees that had cavities, and oak trees used as nest trees had a larger dbh than oak trees within random plots. Local woodland structure at nest locations was characterized by larger trees, measured by greater mean dbh, canopy cover, and basal area of oaks than random locations within the home range. Nest success in natural cavities was 71% and was not predicted by attributes of nest cavities, nest trees, local woodland structure at nests, woodland structure at the home range scale, or landscape context. These results suggest that the most suitable habitat for white-breasted nuthatches in the Willamette Valley includes oak woodlands in close proximity to one another with a high proportion of edge and mature oak trees. Managers should preserve trees containing cavities and large oak trees whenever possible. Thinning of small oaks and removal of conifers in oak woodlands to create more open, savanna-like conditions may also promote the development of larger oaks with more spreading branches, providing more opportunities for cavities to form and more foraging surface area for nuthatches.

This research aimed to investigate how life coaching influenced five female high school learners (aged 16 and 17 years old) at an independent school in Gauteng province. The learners attended four life-coaching sessions with an experienced and accredited life coach over a period of eight weeks. Data was collected from a literature review of previous research on life coaching and coaching in general, interviews with the learners both before and after the life coaching experience and from journal entries kept by the learners during the process. All learners learnt through the process and experienced positive changes in their lives. This included improved balance in their lives; improved time management, reduced stress and more positive mind-sets. They were able to set goals and achieve some success during the process. They expressed increased confidence in themselves and their abilities to overcome challenges in their lives and recommended life coaching for other learners. / Educational Management and Leadership / M. Ed. (Educational Management)