L’expansion et l’intensification agricole ont d’énormes conséquences sur la biodiversité. L’étude des compromis entre production et biodiversité peut aider au ciblage de mesures en faveur de la biodiversité.L’objectif de la thèse était de révéler des stratégies d’utilisation des sols agricoles qui permettraient d’enrayer le déclin des oiseaux agricoles tout en maintenant la production dans un territoire partagé entre grandes cultures et élevage de ruminants.Notre travail s’est basé sur l’étude de corrélations entre des variables décrivant l’utilisation des sols agricoles, la production agricole et plusieurs descripteurs des communautés d’oiseaux dans les milieux agricoles français.Nous avons observé des communautés d’oiseaux très spécialisées et pauvres en espèces dans les régions à haute production végétale ; et des communautés dominées par des oiseaux généralistes dans les régions à haute production animale. La stratégie qui maximise le FBI (un indice d’abondance des oiseaux agricoles) en maintenant la production dans ses composantes animale et végétale combinait une légère extensification des grandes cultures, le renforcement des régions d’élevage extensif et l’intensification des régions d’élevage déjà intensives.Nos résultats soulignent la nécessité d’adopter des solutions différenciées à la production animale et à la production végétale, au-delà de la dichotomie land sparing/land sharing. / Agricultural expansion and intensification have disastrous consequences on biodiversity. Studying the relationship between biodiversity and food production can help devise appropriate measures.The objective of this work was to reveal land use strategies to halt the decline of farmland birds while maintaining agricultural production.Our work was based on the study of correlations between agricultural land use, food production, and various bird community descriptors across French agroecosystems.In high crop production regions, we observed low-diversity communities dominated by few arable specialist species. In high livestock production regions, we observed communities dominated by generalist bird species. The land-use strategy that maximized the Farmland Bird Index under crop and livestock production constraints involved a decrease of crop intensity in the most intensive crop regions and an increase of livestock intensity in the most intensive livestock regions. Extensive livestock production was reinforced in current extensive areas.Our results have shown that different strategies are needed in crop-dominated and in livestock dominated regions.

Agriculture is arguably the greatest threat to tropical forest species. Conservation scientists disagree over the relative importance of two opposing strategies for minimising this threat: enhancing on-farm biodiversity, through wildlife-friendly farming practices, or sparing land for nature by using high-yielding farming methods on the smallest possible area to reduce the need to convert natural habitats. Previous theoretical work shows that understanding the relationship between population density and yield for individual species is crucial for determining whether one of these strategies, or a mixed strategy, will maximise their populations for a given food production target. In this thesis, I aim to identify what land-use strategy will permit increases in food production with least impact on species in the forest zone of Ghana. Farm-fallow mosaic landscapes with shifting cultivation and native canopy trees produced only around 15% as much food energy per hectare as the highest-yielding oil palm plantations. In farm mosaics where perennial tree crops dominate, food production and profits were higher, but did not reach those of oil palm plantations. I surveyed birds and trees in forest, farm mosaic, and oil palm plantation, and combined these data with information on yields to assess the likely consequences of plausible future scenarios of land-use change. My results provide evidence of a strong trade-off between wildlife value and agricultural yield. Species richness was high in low-yielding farming systems, but there was considerable turnover between these systems and forests, with widespread generalists replacing narrowly endemic forest-dependent species. Species most dependent on forest as a natural habitat, those with smaller global ranges and those of conservation concern showed least tolerance of habitat modification. For virtually all species, including even widespread generalists, future land-use strategies based on land sparing are likely to support higher populations of most species and minimise their risk of extinction compared to land-use strategies based on wildlife-friendly farming. If food production is to increase in line with Ghana‘s population growth, a combination of efforts to improve forest protection and to increase yields on current farmed land is likely to achieve this at least cost to forest species. Efforts to better protect forests, which require further restrictions on human use, might be most effective if they can be closely linked to support for farmers to improve their yields. In the long term however, this strategy will only delay and not avert biodiversity loss, unless global society can limit its consumption.

Converting exotic forages to native warm-season grasses (NWSG) such as big bluestem (Andropogon gerardii), little bluestem (Schizachyrium scoparium), and Indian grass (Sorghastrum nutans) offers a sustainable alternative because NWSG may yield comparative livestock gains with less fertilizer, as well as offer habitat for arthropods and declining grassland bird species such as Dickcissels (Spiza americana). In the Southeastern United States, the predominant forage species are exotics such as bermudagrass (Cynodon dactylon) and tall fescue (Schedonorus arundinaceus), so NWSG conversion could substantially improve sustainability and wildlife habitat on private lands in the region. In 2011 and 2012, I studied response of Dickcissels and arthropods to forage origin, diversity, and grazing at the Mississippi State University Prairie Research Unit in Monroe Co., MS, USA. Four treatments were established among 12 pastures representing a gradient in management intensity, including a grazed bermudagrass and tall fescue mix (GMEP), grazed Indian grass monoculture (GINP), grazed mixed native pasture (GMNP), and a non-grazed mixed native pasture (NMNP). Grazed treatments were stocked with steers from May through August each year. I also evaluated the economic implications of each grazing system. In general, there was a positive response to native grasses among Dickcissels and arthropods and a negative effect from grazing. Dickcissel productivity (fledgling/ha) was highest in NMNP and lowest in GMEP, whereas productivity was intermediate and similar among grazed native pastures. This pattern was attributable to availability of suitable nest sites because nest survival and brood size did not vary by treatment. Several arthropod taxa responded positively to greater vegetation density in NMNP, but communities were largely similar among grazed pastures irrespective of forage origin or diversity, suggesting little short-term response to NWSG. In spite of unfavorable growing conditions due to drought, cattle grazing NWSG consistently outperformed conspecifics in GMEP, with 45–72% higher average daily gain. Partial budget analysis indicated that NWSG pastures yielded up to 36% marginal rates of return despite establishment costs. These results suggest NWSG conversion can benefit tall grass specialists such as Dickcissels while offering competitive alternatives to exotic forages, resulting in net benefits for both conservationists and producers.

Land-use intensification in agroecosystems has led to population declines in many taxonomic groups, especially farmland birds. Two contrasting conservation strategies have therefore been proposed: land sharing (the integration of biodiversity conservation in low-intensity agriculture) and land sparing (the spatial separation of high-yielding agriculture and areas for conservation). Despite the large academic interest in this field, only few studies have taken into account stakeholders’ perspectives of these strategies when assessing conservation implications. We modeled the effects of three land-use scenarios (a business-as-usual, a land-sharing, and a land-sparing scenario), developed together with regional stakeholders, on the habitat area of 13 regionally endangered bird species in the Middle Mulde River Basin (Saxony, Germany). We used random forest models based on environmental variables relating to land-use/cover, climate and soil characteristics, occurrence of linear landscape elements (hedges and tree rows), and distance to water and major roads. Responses to the three land-use scenarios were species-specific, but extensively managed permanent grassland and the density of forest edges were positively associated with the occurrence of most bird species. Overall, the land-sharing scenario provided the largest breeding habitat area: 76% of the species had a significant (p < 0.05) increase in breeding habitat, and none showed a significant decrease. Our findings confirm that balancing the different, often contrasting habitat requirements of multiple species is a key challenge in conservation and landscape management. Land sharing, which local stakeholders identified as the most desirable scenario, is a promising strategy for the conservation of endangered birds in agricultural landscapes like our study region.

The effectiveness of land sharing and land sparing (LS/LS) approaches to conservation in the face of rising agricultural demands has been widely debated. While numerous studies have investigated the LS/LS framework from an ecological lens (yield-biodiversity relationship) the relevance of the framework to real life depends on broader considerations. Some of the key caveats include: i) limited knowledge regarding the feasibility of interventions given diverse stakeholders’ interests, ii) the social acceptability (uptake) of these contrasting strategies to direct land users, and iii) limited knowledge regarding their impacts on individuals’ livelihoods and food security. Without considering these social science dimensions proponents of the framework risk an incomplete picture that is not grounded in local realities and can paradoxically force into opposition the very conservation and development interests they seek to reconcile. Using a Companion Modelling approach, which comprises the development of a role-playing game (RPG) and an agent-based model (ABM), this thesis addressed these caveats. The research was based in the Nilgiris of Western Ghats India, a tropical agricultural system at the forest frontier. The main findings show that through engaging local stakeholders in a participatory process, plausible land use strategies that align with their objectives could be identified. Stakeholders proposed three land use interventions. Two of them resemble a form of land sparing (‘monofunctional’ landscapes) on the farms: sparing land for Wildflower Meadows or Tree Plantations while increasing yield on the remaining land. The third intervention asks farmers to accept yield penalties for Intercropping more trees on their farms, a form of land sharing (‘integrated’ landscapes). In terms of decision-making regarding the adoption of these three interventions by direct land users, the study reveals several findings. Firstly there are three main types of motivations that influence farmers’ decision to adopt interventions, in order of importance: monetary benefits, pro-environmental motivations and social norms. Secondly, land use, the type of management preferred on the farm and whether land users accept trees on the farm or not are factors that influence what type of interventions is socially acceptable on individual farms. These factors have been detected in the in-depth household survey and also validated by the RPG. When assessing the adoption of the three interventions, ex ante their implementation, using an ABM, there are some important differences observed between the interventions. Wildflower Meadows is the intervention adopted by the largest number of households, whereas Intercropping is adopted across the largest area of land. Forest Plantations is significantly more unpopular than the other two interventions. The third line of investigation, about the outcomes of adoption, has important policy implications. Adding a socioeconomic dimension to the ecological one adds a level of complexity and creates a less straightforward choice between the LS/LS strategies. None of the three interventions can provide optimal outcomes for production, aspects of biodiversity conservation, livelihoods and food security. Each intervention has indicators that score better compared to the other two interventions. The findings demonstrate that the ecological focus of the LS/LS framework is insufficient to deal with real-world complexities and lends itself to overly simplistic policy prescriptions. More meaningful policies could be achieved when bridging natural and social sciences to better understand the merits and limitations of the LS/LS approaches.