An improved analytical framework and typology of fishers are provided to improve understanding of the diverse livelihood functions of inland fishing in development policy making. Inland fisheries make an important but often neglected contribution to rural livelihoods in developing countries. A holistic and widely applicable analysis of the possible livelihood functions of such fisheries is presented, focusing on fishing as one activity within diverse livelihoods. Four different livelihood strategies are identified, involving fishing as: (i) a primary livelihood of last resort, (ii) part of a diversified semi-subsistence livelihood, (iii) a specialist occupation and (iv) part of a diversified accumulation strategy. The policy implications of these strategies are found to be differentiated and poorly represented in practice by socio-economic analysis that either undfervalues fisheries or treats them solely as livelihoods of last resort and by traditional approaches to fisheries management centred on stock conservation. The need for a more diverse and flexible range of measures, tailored to local priorities and conditions and ensuring that poor people can access the benefits of inland fisheries whilst achieving conservation objectives, is identified.

Many tropical fisheries are inherently of a multispecies nature, with any given type of fishing gear harvesting a wide range of species. Species-aggregated relationships between fishing effort and yield or catch per unit of effort (CPUE) provide important information for the management of such fisheries, as well as insights into ecosystemlevel responses to fisheries exploitation. We used a model selection approach to study species-aggregated, yield-effort relationships in spatially replicated, multispecies inland fisheries. Of three alternative models considered, the data strongly supported a sigmoid functional form (Akaike weight 0.95) over the alternative, asymptotic exponential or quadratic (Schaefer) models (Akaike weights 0.03 and 0.02, respectively). The sigmoid form implies the occurrence of an inflexion point in the rising part of the yield-effort curve and no decline in aggregated yield even at the highest effort levels observed. Aggregated CPUE declines steeply near the origin of the effort scale before stabilizing or rising to a local maximum and then declining inversely with further increases in effort. This highly nonlinear response suggests that extreme caution is required when interpreting aggregated CPUE as an indicator of fishing impacts on exploited communities. Further research is required to identify the mechanisms generating the observed relationship.Résumé : Plusieurs pêches tropicales sont multispécifiques de nature et tout type d'engin de pêche y récolte une gamme étendue d'espèces. Les relations pour l'ensemble des espèces entre l'effort de pêche et le rendement ou les captures par unité d'effort (CPUE) fournissent des informations importantes pour la gestion de telles pêches, ainsi que des perspectives sur les réactions écosystémiques à l'exploitation par la pêche. Nous utilisons une méthodologie de sélection de modèles afin d'étudier les relations rendement-effort pour l'ensemble des espèces dans des pêches inté-rieures multispécifiques qui se répètent à l'échelle spatiale. Des trois modèles considérés, les données appuient fortement le choix d'une forme fonctionnelle sigmoïde (pondération d'Akaike de 0,95) plutôt que des modèles de rechange asymptotique exponentiel ou quadratique de Schaefer (poids respectifs d'Akaike de 0,03 et de 0,02). La forme sigmoïde implique l'existence d'un point d'inflexion dans la partie croissante de la courbe rendement-effort et l'absence de déclin du rendement de l'ensemble des espèces même aux niveaux les plus élevés d'effort observés. La CPUE de l'ensemble des espèces décline abruptement près de l'origine de l'échelle des efforts avant de se stabiliser ou de remonter pour atteindre un maximum local, pour ensuite diminuer en fonction inverse des augmentations subséquentes de l'effort. Cette réaction qui est loin d'être linéaire indique qu'il faut être très prudent lorsqu'on interprète la CPUE de l'ensemble des espèces comme indicateur des impacts de la pêche sur les communautés exploitées. Des recherches supplémentaires pourraient identifier le...

Summary 1.Expanding irrigation development threatens the productive and diverse fisheries of rain-fed rice-farming landscapes. Environmental management of irrigation can minimize negative impacts on fisheries, but its effectiveness is constrained by a lack of reliable information on the nature and magnitude of impacts. 2. To quantify the impacts of small-to medium-scale irrigation schemes on aquatic habitat availability, fish catches, species richness and ecological composition of fish assemblages in these landscapes, we conducted a field study in Laos. The observational study was replicated at irrigation scheme level, covering 10 weir and 10 dam irrigation schemes paired with non-impacted control sites. 3. Weir schemes had no significant impact on aquatic habitat, but caused a significant decline ( − 36%) in fish catches that was only partly explained by a reduction in fishing effort. Weirs had no effect on species richness, but were associated with a significant increase (+17%) in the relative abundance of omnivores. 4. Dam irrigation schemes significantly reduced riverine habitat area, and increased lacustrine and dry-season rice-field areas. Dams led to a marked redistribution of catch and fishing effort from non-reservoir habitats into reservoirs, but no overall change in catch or effort occurred. The redistribution reflected a response to fishing opportunities in the reservoir, rather than a loss of productivity in non-reservoir habitats. No significant impacts were detected on fish species richness or the relative abundance of functional feeding groups. 5. Overall impacts of irrigation on fisheries were related mostly to changes in fishing effort, rather than ecological effects on the resources. The unexpectedly moderate level of ecological impacts may reflect compensatory effects at the landscape level and the fact that rice fields, which dominated the wet-season habitat, continued to be managed as rain-fed deep-water systems even where dry-season irrigation had been developed. 6. Synthesis and applications. Small-to medium-scale irrigation schemes in rain-fed ricefarming landscapes have only moderate impacts on fisheries, which remain productive and diverse. Changes in agricultural practices in the wet season are likely to have greater effects on fisheries than dry-season irrigation.