Os padrões de associação entre epífitos e forófitos não podem ser considerados espécie-específicos, mas as árvores que os epífitos colonizam também não são um conjunto aleatório das espécies forofíticas de um determinado local. Ao invés disso, parece haver uma preferência de certos epífitos por diferentes forófitos. No entanto, se conhece pouco sobre os fatores que determinam essa preferência. Nosso objetivo nesse trabalho é avaliar como os atributos funcionais e a filogenia dos epífitos vasculares influenciam na associação dos epífitos com os forófitos em uma Floresta Ombrófila Mista no sul do Brasil. Para isso nós (1) investigamos padrões de associação positiva e negativa entre grupos funcionais de epífitos e grupos de forófitos e como a diversidade e os atributos funcionais dos epífitos variavam em função do tamanho do forófito e (2) inferimos sobre a existência de sinal filogenético no uso de árvores hospedeiras pelos epífitos, procuramos por estrutura filogenética nas comunidades epifíticas e investigamos diferenças de composição filogenética de epífitos vasculares em diferentes clados de forófitos. Foram amostrados 70 forófitos compreendendo 15 espécies pertencentes a diversos clados e com arquiteturas e características variadas. A amostragem compreendeu 31 espécies epifíticas com os principais clados sendo Polypodiaceae, Bromeliaceae e Orchidaceae. A associação de grupos de epífitos vasculares com diferentes grupos de forófitos sugere que as características dos forófitos proporcionam ambientes contrastantes e que diferentes valores de atributos são necessários para colonizar esses ambientes. Mais especificamente espécies epifíticas com menor área específica foliar (SLA) parecem predominar em árvores maiores e maior SLA em árvores menores. Encontramos sinal filogenético na utilização dos forófitos, sugerindo que a conservação das interações com os forófitos deve ter sido importante ao longo da evolução dos epífitos. A tendência a agrupamento filogenético nas comunidades epifíticas sugere a influência de filtros ambientais representados pelas diferentes características dos forófitos estruturando as assembleias de epífitos. Clados mais basais de forófitos apresentaram composição filogenética distinta devido, sobretudo, a presença de diferentes clados de monilófitos epifíticos nessas árvores. Angiospermas epifíticas ocorreram principalmente em forófitos pertencente as eurosídeas. A preferência de epífitos por forófitos parece ser influenciada pelo surgimento de novidades morfológicas e ecofisiológicas em alguns clados, enquanto outros mantiveram o seu nicho ancestral. A composição florística das florestas quando da origem dos clados epifíticos também parece influenciar a associação entre epífitos e forófitos. Ao utilizar informações sobre os atributos e filogenia das espécies de epífitos vasculares nós podemos melhor compreender os mecanismos ecológicos e históricos que influenciam os padrões de associação entre epífitos e forófitos. / It has been shown that patterns of association between epiphytes and phorophytes can not be considered species-specific, although the trees that epiphytes colonize are not a random subset of phorophyte species in a particular location. Instead, there seems to be a preference of some epiphytes for different phorophytic species. However, little is known about the factors determining this choice. Our objective in this study is to assess how functional attributes and phylogeny of vascular epiphytes influence the association of epiphytes with the phorophytes in an Araucaria Forest in Southern Brazil. For that we (1) investigated the positive and negative association patterns between functional groups of epiphytes and groups of phorophytes and how functional diversity and functional traits of epiphytes varied with host tree size and (2) inferred about the existence of phylogenetic signal on the host trees use by epiphytes, looked for phylogenetic structure in the epiphytic communities and investigated differences in the phylogenetic composition of vascular epiphytes in different phorophyte clades. We used a sample of 70 phorophytes comprising 15 species and belonging to different clades and with different architectures and traits. The sample comprised 31 epiphytic species, the major clades being Polypodiaceae, Bromeliaceae and Orchidaceae. The combination of vascular epiphyte groups with different groups of 15 phorophytes suggests that phorophyte traits provide contrasting environments and that different trait values are needed to colonize these environments. More specifically, epiphytic species with lower specific leaf area (SLA) seem to predominate on larger trees and with higher SLA values on smaller trees. We found phylogenetic signal on the host tree use, suggesting that conservatism of the interactions with phorophytes must have been important throughout the evolution of epiphytes. The tendency to phylogenetic clustering in the epiphytic communities suggests the influence of environmental filters represented by phorophyte traits structuring epiphyte assemblages. More basal clades of phorophytes showed different phylogenetic composition mainly due to the presence of different epiphytic monilophyte clades on these trees. Epiphytic angiosperms occurred mainly on those trees belonging to eurosids. The preference of epiphytes for phorophytes seems to be influenced by morphological and ecophysiological novelties in some lineages, while other clades kept their ancestral niche. The floristic composition of forests at the origins of epiphytic lineages also appears to influence the association between epiphytes and phorophytes. By using information about the traits and phylogeny of species of vascular epiphytes we can better understand the ecological and historical mechanisms that influence the patterns of association between epiphytes and phorophytes.

L’accélération du processus d’urbanisation, constatée à l’échelle mondiale depuis les dernières décennies, conduit à une artificialisation progressive des milieux naturels. La construction d’infrastructures de transport ou de nouveaux bâtiments fragmente les paysages de manière irréversible et cause une réduction des habitats écologiques et de leur connectivité. Le maintien de la fonctionnalité des réseaux écologiques, s’intègre désormais dans les politiques d’aménagement du territoire ou d’urbanisme soucieuses de la préservation de la biodiversité.En se focalisant plus particulièrement sur les évolutions urbaines à l’horizon 2030 dans l’Aire Urbaine de Besançon (développement résidentiel et variations de trafic routier), cette thèse cherche à évaluer l’impact potentiel des formes d’urbanisation sur la connectivité des réseaux écologiques des espèces animales. Ce travail de recherche privilégie l’approche par la modélisation en s’inscrivant à la fois dans le champ de la géographie théorique et quantitative et de l’écologie du paysage.L’application de cette démarche se fait en trois étapes : (1) simuler le développement résidentiel et ses évolutions de trafic associées à l’horizon 2030, à l’aide de cinq scénarios prospectifs présentant des formes urbaines différenciées ; (2) modéliser les réseaux écologiques de plusieurs espèces animales avec des graphes paysagers construits à partir de cartes d’occupation du sol et de données écologiques ; et (3) évaluer les impacts potentiels de chaque scénario sur les réseaux écologiques à partir de ces graphes à l’aide de métriques de connectivité, par mesure de la perte de connectivité imputable à chaque scénario de développement résidentiel.Les résultats obtenus montrent que les formes de villes denses et compactes, contrairement aux villes étalées, sont celles qui favorisent le mieux le maintien des connectivités écologiques pour la plupart des groupes d’espèces analysés. Des analyses plus approfondies mettent en avant la contribution importante des variations de trafic aux impacts écologiques de chaque scénario.D’après les analyses de sensibilité effectuées, le modèle utilisé est robuste, ce qui montre l’intérêt de la modélisation dans le processus d’aide à la décision pour la protection environnementale et la planification urbaine afin de penser la ville de demain de manière durable. / The global increase of urbanization during the past decades have induced a progressive artificialization of natural environments. The building of transport infrastructures and new housings causes a landscape fragmentation in an irreversible way and a strong decrease of the connectivity of ecological habitats. Maintaining the functionality of ecological networks is becoming a major goal of sustainable urban planning policies. With a special focus on urban evolutions in the horizon 2030 in the urban area of Besançon in eastern France (residential development and road traffic evolutions), this thesis aims to assess the potential impact of urban forms on landscape connectivity of animal species’ ecological networks. This research work promotes a modelling approach both on the field of theoretical and quantitative geography and landscape ecology.This approach follows three main steps: (1) simulating residential development and its associated road traffic changes using five prospective scenarios of differentiated urban forms; (2) modelling landscape graphs of various animal species using land-cover maps and ecological data; (3) assessing the potential impacts of each scenario on ecological networks from these graphs using connectivity metrics, with measures of the connectivity decrease attributable to each residential development scenario. Contrary to sprawled cities, the results show that compact and dense urban forms best promote the maintenance of ecological connectivity for the majority of species groups. Further analysis highlights the great contribution of road traffic evolutions regarding the ecological impacts of each scenario.According to some sensitivity analysis, the model used is quite robust. It demonstrates the interest of modelling in the decision-making process for environmental conservation and urban planning to think out the city of tomorrow in a sustainable way.

Le paysage constitue à la fois un cadre de vie pour les populations humaines et un support du cycle de vie des espèces animales. Les modifications du paysage induites par les dynamiques d’occupation du sol se répercutent sur ces deux dimensions, l’une esthétique et l’autre écologique. Ces logiques étant généralement étudiées dans des champs disciplinaires différents, peu de recherches ont porté sur la manière dont elles s’articulent selon les modifications des structures paysagères.Ce travail cherche donc à modéliser de manière rétrospective la coévolution spatiale des fonctions esthétique et écologique du paysage à partir de métriques spatiales basées sur des données d’occupation du sol. Il se focalise sur les changements intervenus dans les franges urbaines de deux agglomérations françaises (Besançon et Paris) durant les 30 dernières années.La démarche adoptée a d’abord visé à modéliser, à partir des données d’occupation du sol, (1)les préférences paysagères d’un ensemble d’individus et (2) la connectivité écologique pour un ensemble d’espèces animales. En mobilisant de manière complémentaire des analyses statistiques multivariées et des analyses spatiales, le cœur du travail a ensuite consisté à étudier comment ces deux fonctions ont évolué de manière convergente ou divergente au cours du temps. Les résultats donnent de nouveaux éléments de compréhension des relations entre esthétique et écologie du paysage et amènent à s’interroger sur l’intérêt de la modélisation spatiale pour une gestion du paysage conciliant la préservation du cadre de vie des habitants et la conservation de la biodiversité. / Landscape is both a backdrop to the lives of human populations and a medium for the life cycle of animal species. Landscape changes induced by land-use and land-cover dynamics affect both these dimensions, the one aesthetic, and the other ecological. Because these rationales areusually studied within different disciplines, little research has been done into how the two clashor combine as and when landscape structures change. This work seeks therefore to model the spatial co-evolution of the aesthetic and ecological functions of landscape retrospectively usingspatial metrics based on land-cover data. It focuses on changes in the urban fringes of two French cities (Paris and Besançon) over the last 30 years. The approach attempts first to use land-cover data to model (1) the landscape preferences of a set of individuals and (2) the ecological connectivity of a set of animal species. Drawing on both multivariate statistical analysis and spatial analysis, the core of this work consists in investigating how the two functions have evolved in convergent or divergent ways over time. The results provide fresh insight into the relationship between landscape aesthetics and landscape ecology and raise questions about the value of spatial modelling for a landscape management approach that endeavours to reconcile the preservation of residents’ living environments and the conservationof biodiversity.

Thesis (PhD)--University of Stellenbosch, 2011. / ENGLISH ABSTRACT: Land use change is leading to rapid biodiversity loss in terrestrial ecosystems worldwide. Ecological networks (ENs) are systems of remnant natural habitat which remain intact in a transformed matrix, and have been suggested as a means to mitigate the effects of habitat loss and transformation and the resulting loss of biodiversity. Conceptually, ENs are similar to habitat corridors but are larger in scale and more heterogeneous in their design and management. Like corridors, the efficacy of ENs for biodiversity conservation has been called into question and requires empirical investigation. South Africa boasts a unique system of extensive ENs associated with exotic timber plantations, mostly within the highly endangered grassland biome and Indian Ocean Coastal Belt in KwaZulu-Natal Province, South Africa. In these regions, grasshoppers respond sensitively to management practices such as grazing, mowing and burning, justifying their a priori selection as an ecological indicator. In this study, I utilize grasshopper assemblages to determine the efficacy of ENs for biodiversity conservation and suggest guidelines for EN optimization. Specifically, I assess grasshopper assemblage sensitivity and robustness to habitat quality within ENs, congruence of grasshoppers with butterflies, and I identify indicator species which can be utilized by managers for EN assessment in future. I also perform a case study of isolated fragment utilization by a highly mobile generalist species, the bird locust, Ornithacris cyanea (Stoll, 1813) (Acrididae: Cyrtacanthacridinae). Grasshopper assemblages were sampled within ENs in two geographic regions (Zululand and the KwaZulu-Natal Midlands) and in two years (2007 and 2008) during the peak season of grasshopper abundance, late summer (February-April). Local-scale environmental variables relating to management practices and landscape-scale environmental variables relating to design of the ENs were quantified. Management practices explained, on average, two-thirds of the variability in grasshopper assemblages that could be explained, while design variables explained one-third. Grass height and the time since the last fire event were most consistently influential, while area, context, isolation, proportion of bare ground and proportion of forbs at a site, proved influential in some analyses but not others. This response was robust over time and among geographic regions. Grasshopper species richness and abundance were highly congruent with that of butterflies and did not differ among isolated fragments, connected corridors and reference sites, although they did differ among geographic regions. Not all grasshopper species responded similarly to ENs. There was a strong phylogenetic signal in species response to environmental variables, with the distribution of highly mobile families varying more among years. Linear, heavily disturbed power line servitudes had higher grasshopper abundance, and generalist species consistent with early successional assemblages. Reference sites had more graminivorous species and those with intermediate mobility. Three species of grasshopper were identified with IndVal and validated on an independently collected dataset as indicators of high habitat quality. A case study of the bird locust showed that populations of this highly mobile species were effectively isolated and undergoing anthrovicariance even at short distances. These results indicated that heterogeneous ENs supported diverse grasshopper assemblages, although movement among isolated fragments may have been limited. A grasshopper bioindication method for South Africa’s ENs is suggested and has great potential for assessment of a crucial and sensitive trophic layer within the ENs. This method should be field-tested and revised over time as grasshopper relative abundances and species compositions may change. Increased heterogeneity, simulation of multiple successional stages, and increased connectivity are expected to positively impact biodiversity, particularly of insect primary herbivores. Globally, ENs, if managed and designed appropriately, have potential to enhance biodiversity, particularly of smaller resident organisms which can utilize the ENs for movement and live within them. / AFRIKAANSE OPSOMMING: Veranderinge in grondgebruik lei tot ‘n vinnige verlies van biodiversiteit in terrestriële ekosysteme wêreldwyd. Ekologiese netwerke (ENs) is sisteme van oorblywende natuurlike habitat wat ongeskonde bly in getransformeerde matrikse, en is voorgestel as ‘n manier om die verlies aan biodiversiteit as gevolg van habitatverlies en transformasie te verminder. Konseptueel, is ENs soortgelyk aan habitat korridors, maar is groter in omvang en meer heterogeen in hulle ontwerp en bestuur. Soos korridors, is die effektiwiteit van ENs om biodiversiteit te bewaar bevraagteken, en vereis dit empiriese ondersoek. Suid-Afrika spog met ‘n unieke stelsel van uitgebreide ENs wat geassossieer is met uitheemse plantasies, meestal in die hoogs bedreigde grasveld bioom en Indiese Oseaan kusstrook in KwaZulu- Natal Provinsie, Suid Afrika. In hierdie streek reageer sprinkane sensitief op bestuurspraktyke soos beweiding, sny en brand. In hierdie studie het ek sprinkaan gemeenskappe gebruik om die effektiwiteit van ENs vir bewaring van biodiversiteit te bepaal en om riglyne te stel vir EN optimering. Spesifiek het ek sprinkaan gemeenskap sensitiwiteit en robustness tot habitat kwaliteit bepaal, oreenstemming tussen sprinkane en skoenlappers getoets, en het ek indikator species geïdentifiseer wat in die toekoms deur bestuurders gebruik kan word vir EN assessering. Ek het ook ‘n gevallestudie gedoen van geïsoleerde fragment benutting deur ‘n hoogs mobiele, algemene spesie, Ornithacris cyanea (Stoll, 1813) (Acrididae: Cyrtacanthacridinae). Sprinkaan gemeenskappe was versamel in twee geografiese gebiede (Zululand en die KwaZulu-Natal Midlands) en in twee jare (2007 en 2008) gedurende die seisoen van hoogste sprinkaan aktiwiteit, laat somer (Februarie – April). Plaaslike-skaal omgewingsveranderlikes met betrekking tot bestuurspraktyke en landskap-skaal omgewingsveranderlikes met betrekking tot die ontwerp van ENs is gekwantifiseer. Bestuurspraktyke verduidelik, gemiddeld, twee derdes van die variasie in die sprinkaan gemeenskappe, terwyl ontwerpveranderlikes een derde verduidelik. Grashoogte en tyd sedert die laaste brand was konsekwent invloedryk, terwyl area, konteks, isolasie, hoeveelheid kaal grond en hoeveelheid kruidagtige plantegroei invloedryk was in sekere analyses, maar in ander nie. Hierdie reaksie was konstant oor tyd en tussen geografiese gebiede. Sprinkaan spesies rykheid en hoeveelheid was hoogs kongruent met die van skoenlappers, en het nie verskil tussen geïsoleerde fragmente, verbinde korridors en verwysingsareas nie, maar hulle het verskil tussen geografiese streke. Nie alle sprinkaan spesies het soortgelyk gereageer tot ENs nie. Daar was ‘n sterk filogenetiese sein in spesies reaksies op die omgewingsveranderlikes, met meer variasie tussen jare in die verspeiding van hoogs mobiele families. Liniêre, hoogs versteurde kraglyn dienspaaie het hoer sprinkaan hoeveelhede gehad en algemene spesies wat geassosieer is met gemeenskappe in vroeë stadiums van suksessie. Verwysingsareas het meer graminivorous spesies gehad en spesies met intermediêre mobiliteit. Drie sprinkaan spesies is geïdentifiseer met IndVal en bevestig met ‘n onafhanklik versamelde dataset as indikators van hoë kwaliteit habitat. ‘n Gevallestudie op O. cyanea het aangedui dat populasies van hierdie hoogs mobiele spesie geïsoleerd is en dat hulle anthrovicariance ondergaan, selfs op kort afstande. Hierdie resultate dui aan dat heterogene ENs diverse sprinkaan gemeenskappe ondersteun, hoewel beweging tussen geïsoleerde fragmente dalk beperk is. Verhoogde heterogeneïteit, simulasie van verskeie suksessiewe fases en meer konnektiwiteit sal moontlik ‘n positiewe invloed op biodiversiteit hê, aangesien sprinkane verteenwoordigend was van insek primêre herbivore. Wêreldwyd, as ENs op ‘n gepaste manier bestuur en ontwerp word, het dit die potensiaal om biodiversiteit te bevorder. Dit is veral belangrik vir kleiner, plaaslike organisms wat ENs kan benut vir beweging en as leefhabitat. ‘n Sprinkaan bioindikasie metode vir Suid-Afrika se ENs is voorgestel en het groot potensiaal vir die assessering van ‘n belangrike en sensitiewe trofiese laag binne die ENs. Hierdie metode moet in die praktyk getoets word en hersien word oor tyd om veranderinge in sprinkaan relatiewe hoeveelhede en spesies komposisies in ag te neem.

There is concern that climate change may cause mismatches between timing of flowering and activity of pollinators (phenology). However, concluding that mismatches will occur, and have serious consequences for pollination services, requires assumptions that have not yet been tested. I begin by discussing a set of these assumptions, bringing past research into the context of mismatch. Briefly, the assumptions are that 1) dates of first-flowering or emergence (DFFE) correctly describe phenology (and therefore mismatch); 2) differences in DFFE represent the magnitude of mismatch; 3) advancement of DFFE will be the primary phenological change; 4) shifts will be random and independent for each species; 5) populations of plants and pollinators are “bottom-up” regulated by their mutualistic interactions; 6) all interactions are of similar strength and importance; 7) dispersal, and the spatial context of phenological mismatches can be ignored; and ecological processes including 8) phenotypic plasticity and adaptive evolution of phenology, 9) competition and facilitation, and 10) emergence of novel interactions, will not affect mismatches. I then describe novel experiments, which could help to account for some of these assumptions, clarifying the existence and impacts of mismatches. Next, I present an original field experiment on factors affecting seed set in an alpine meadow in the Coast Range of British Columbia, Canada. I found evidence contradicting the assumption that seed set is primarily limited by pollination. My data highlight the roles of phenology, temperature (degree-days above 15°C, and frost hours), and interactions with pollinators (mutualists) and seed-predators (floral antagonists) in driving patterns of seed set. Seed set of early and late-flowering species responded differently to a 400m elevation gradient, which might be explained by phenology of bumble bees. My data suggest that the consequences of mismatch may be smallest for plants that are fly-pollinated and self-fertile. Non-selfing, bee-pollinated species might be more prone to reproductive limitation through mismatch (affected by snowmelt and cumulative degree-days). Plants that are limited by seed-predators might be negatively affected by warming temperatures with fewer frost hours, and extreme events such as late-season frosts and hail storms can prevent plants from setting seed entirely. Overall, my work emphasizes the importance of complementing theory, data-driven simulations, and meta-analyses with experiments carried out in the field. / Graduate