It is increasingly recognized that a sustainable future for agriculture must build on ecosystem services. Pollination is an important ecosystem service in all agroecosystems. In much of Africa the main challenge is conserving pollinator biodiversity in traditionally “ecologically-intensive” agroecosystems that are changing to meet different demands for food security and poverty alleviation, rather than safeguarding pollination in transition from conventional agricultural systems, with a high reliance on purchased inputs, to “ecologically-intensive” agroecosystems using natural inputs provided by biodiversity. Priority issues for research and development in pollination services in Africa include, inter alia: quantification and documentation of pollination deficits and finding measures to address these; socio-economic valuation of pollinator-friendly practices; assessment of lethal and sub-lethal effects of farming methods, such as pesticide use, on crop pollinators; identification of habitat management practices that enhance synergies between pollinator lifecycles and crop growing patterns; and policy analysis in relation to drivers and trends in pollination services and management.

Unprecedented growth in human populations has required the intensification of agriculture to enhance crop productivity, but this was achieved at a major cost to biodiversity. There is abundant local‐scale evidence that both pollinator diversity and pollination services decrease with increasing agricultural intensification. This raises concerns regarding food security, as two‐thirds of the world's major food crops are pollinator‐dependent. Whether such local findings scale up and affect crop production over larger scales is still being debated. Here, we analyzed a country‐wide dataset of the 54 major crops in France produced over the past two decades and found that benefits of agricultural intensification decrease with increasing pollinator dependence, to the extent that intensification failed to increase the yield of pollinator‐dependent crops and decreased the stability of their yield over time. This indicates that benefits from agricultural intensification may be offset by reductions in pollination services, and supports the need for an ecological intensification of agriculture through optimization of ecosystem services.

Agriculture's reliance on ecosystem services creates economic and ecological interdependencies between crop production and biodiversity. Interactions with mobile organisms are particularly complex because they depend on the spatial configuration of habitat at large scales. As such conserving habitat is likely to benefit multiple farmers whereas conservation costs are born individually, creating potential interdependencies among farmers. We explore under what conditions landscape-scale management of ecosystem services is likely to benefit farmers compared to managing them at the farm-scale. To do this we develop an agent-based model (ABM) to predict the landscape configuration emerging from farm-scale management under different conditions: initial landscape, crop and pollinator characteristics. As a benchmark, the landscape configuration from landscape-scale management is derived through a global optimization procedure. Not only do we find that efficiency improves with landscape-scale management, but also that all farmers would benefit from it (given dependence of crop yields on ecosystem services). However, we also find that the individual incentives to avoid maintaining habitat on one's own land are relatively high; therefore creating conditions for a Prisoner's Dilemma-type problem. On the other hand we also demonstrate that an incentive-compatible contract exists that can promote efficient landscape management (by combining side-payments with fines for defection).

Understanding the inter-play plant–pollinator is essential to develop ecosystem services of faba bean and improved populations for low-input systems. We analyzed the phenotypic selection exerted by pollinators on floral traits involved in the plant pollinator inter-play. We test whether variation in pollinator-related floral traits is associated with differences among plants in seed production patterns. We used open pollination and pollinator-exclusion environments to examine pollinator mediated selection and selection for autonomous selfing in six gene-pools of Vicia faba over three consecutive years. We recorded, by using Digital Image Analysis, functional floral traits related to attraction, sexual dimension, and vector matching/pollen transfer efficiency. Nine production components were measured to categorize seed production patterns. Our approach used a series of Multivariate Regression Analyses (MRA) to explore which floral traits provided the best models to explain seed production patterns in each gene-pool and pollination environment. MRA showed that variation in the incidence of pollinator-mediated selection on floral traits or in the selection for autonomous selfing can substantially contribute to differences in seed patterns. The underlying floral mechanisms are specific to the gene-pool and largely unrelated among gene-pools. Consistent results among gene-pools involved no evidence of pollinator-mediated selection as result of the floral traits under study on the main predictors of crop yield, pods and seeds per plant. However, relevant predictors of crop yield such as pod length and seed dimensions and weight were pollinator-dependent because of pollinator-mediated selection on sexual dimension, floral display and vector matching traits. We caution against dismissing pollinator-mediated selection as driver of seed production patterns variation which may be influenced by the gene-pool and by the gene-pool × pollination environment interaction. | We gratefully acknowledge the EU-FP7- KBBE-2009-3 project SOLIBAM (Strategies for Organic and Low-input Integrated Breeding and Management) for financial support. | Peer Reviewed

Biotic pollination is an important ecosystem service for the production of many food crops. The supply of pollination is mostly studied at the landscape scale while recent studies on the demand for pollination services provide a global-scale picture based on aggregate national-level data. This paper quantifies both demand and supply of pollination in the European Union (EU) at a relatively high spatial resolution, allowing an analysis of the match between demand and supply. Finally, we evaluate how policies interact with the spatial differences between demand and supply of this ecosystem service.We mapped the crop area requiring pollination for optimal production (demand) and both bee habitat and related visitation probability (supply) using detailed agricultural and landscape data. We compared the maps of demand and supply by visual comparison, descriptive statics and a trend surface generalized additive model to analyze the relation between visitation probability and the presence or absence of pollinator dependent crops. A sensitivity analysis was done to test the robustness of the pollination supply model. Finally, the impact of EU Biodiversity Strategy and Biofuel Directive were evaluated by identifying areas where these policies would influence the demand or supply of pollination.In the EU, 12% of the total cropland area was dependent on pollinators for optimal agricultural production. Pollinator habitat is especially abundant in mosaic landscapes as found in hilly and mountainous areas. Although covering less than 0.5% of the agricultural area, the presence of green linear elements increased the visitation probability by 5–20% while being the sole providers of pollinators in 12% of the croplands. In half of the area with a high pollination demand, the supply of pollination is also high. Irrespective of the different parameterizations, total habitat areas and visitation probability were highest in croplands without pollinator dependent crops and lowest in hotspots of pollination demand.The analysis of the match between pollination supply and demand for this service indicates that for improving or ensuring pollination one must consider both the demand and supply of the process of crop pollination for optimal results.

Insect pollination is important for food production globally and apples are one of the major fruit crops which are reliant on this ecosystem service. It is fundamentally important that the full range of benefits of insect pollination to crop production are understood, if the costs of interventions aiming to enhance pollination are to be compared against the costs of the interventions themselves. Most previous studies have simply assessed the benefits of pollination to crop yield and ignored quality benefits and how these translate through to economic values. In the present study we examine the influence of insect pollination services on farmgate output of two important UK apple varieties; Gala and Cox. Using field experiments, we quantify the influence of insect pollination on yield and importantly quality and whether either may be limited by sub-optimal insect pollination. Using an expanded bioeconomic model we value insect pollination to UK apple production and establish the potential for improvement through pollination service management. We show that insects are essential in the production of both varieties of apple in the UK and contribute a total of £36.7 million per annum, over £6 million more than the value calculated using more conventional dependence ratio methods. Insect pollination not only affects the quantity of production but can also have marked impacts on the quality of apples, influencing size, shape and effecting their classification for market. These effects are variety specific however. Due to the influence of pollination on both yield and quality in Gala, there is potential for insect pollination services to improve UK output by up to £5.7 million per annum. Our research shows that continued pollinator decline could have serious financial implications for the apple industry but there is considerable scope through management of wild pollinators or using managed pollinator augmentation, to improve the quality of production. Furthermore, we show that it is critically important to consider all production parameters including quality, varietal differences and management costs when valuing the pollination service of any crop so investment in pollinator management can be proportional to its contribution.

Previous studies focused mainly on the provision of ecosystem services by species movements between semi-natural and managed habitats, whereas data on spillover effects between two managed habitats or between habitats that provide target resources in non-overlapping time periods are lacking. We studied densities of three pollinator groups on sunflower fields as a late mass-flowering crop in 16 landscapes that differed in the relative cover of oil-seed rape as an early mass-flowering crop, in the relative cover of sunflowers and in the relative cover of semi-natural habitats. Our aim was to evaluate dynamics between two crops with non-overlapping flowering periods. Densities of bumble bees in late-flowering sunflower fields were enhanced by early-flowering oil-seed rape. Highest bumble bee densities in the late-flowering crop were reached in landscapes that combined high relative covers of oil-seed rape and semi-natural habitats. Further, low relative covers of oil-seed rape in spring led to decreased bumble bee densities in late-flowering sunflower fields in landscapes with high relative covers of sunflower fields (dilution effect), whereas in landscapes with high relative covers of oil-seed rape, no dilution of bumble bees was found. Thus, our results indicate that early mass-flowering crops can mitigate pollinator dilution in crops flowering later in the season. We conclude that the management of landscape-scale patterns of early and late mass-flowering crops together with semi-natural habitats could be used to ensure crop pollination services. Similar processes could also apply for other species groups and may be an important, but so far disregarded, determinant of population densities in agroecosystems.

Tropical countries contribute substantially to global agricultural production, but the majority of farmers are small-scale subsistence farmers. Their comparably low agricultural productivity is coupled with major yield gaps. Agricultural expansion and the intensive use of agro-chemicals are the major cause for the destruction of tropical habitats and biodiversity loss and pose a threat to ecosystem services. However, many tropical small-holder farmers rely on ecosystem services like insect pollination or natural pest control, which may play an indispensable role in closing yield gaps. In addition, losses from pollination deficits or pest pressure are usually mitigated by the use of managed pollinators or pesticides that in turn can be harmful to ecosystem services provided by wild insects. We aim to shed light on the relative importance of pollination services, common management practices and their interaction. We evaluated the benefits from pollination services for cucumber (Cucumis sativus) production and how wild bees (the main cucumber pollinators in the region) are influenced by management on four different spatial scales. We furthermore assessed income loss due to dis-services from seed predating ants in order to suggest management measures that may reduce yield gaps. Our work comprises experimental field studies on small-scale vegetable production in homegardens in tropical central Sulawesi, Indonesia and a review on pollination services to highlight existing information and to close information gaps on pollinations services for crops. In the first study (chapter 2) we tested how different management practices (insect pollination, weed control, fertilization and herbivore control) affect cucumber fruit set and yield and how these variables influence each other. We found that insect pollination, fertilization and weed control increased crop fruit set and yield in an additive way. However, fertilization and weed control alone could not compensate for pollination loss, which was the most important driver and accounted for 75 % of the yield. We found an interaction between the weed control and pollination treatment in which weed control strongly influenced insect-pollinated plants but not wind and self-pollinated plants. This indicates that weed control contributes to closing yield gaps, but only in addition to insect pollination. In contrast, insecticides to control herbivores did not influence yield. We recommend shifting the focus from common management practices towards more sustainable management to enhance pollination services and stress the importance for policy driven regulations of reduced and better targeted pesticide application in tropical agroecosystems. In the second study (chapter 3), we evaluated variables from three spatial scales to better understand bee communities indispensable for cucumber production. We further assessed if the response to these scale predictors depends on pollinator body size and sociality. Yield increased significantly with increasing number of flower-visiting bee individuals (mainly composed of wild solitary bees which translates into a net income decline of 47% if half of the bees would be lost. For optimized bee management, farmers need to consider four spatial scales. On the garden scale (1), the homegarden-wide percentage of flower cover predicted pollinator attraction best, if, on the adjacent-habitat scale (2), a higher percentage of homegardens (at least 20%, best 50% in a 200m radius) surrounded the study garden. In addition, the landscape scale (3), distance to the rainforest (up to 2.2 km) had also a significant effect on total bee flower visitor. This effect was dominated by small bees which increased closer to the forest. We conclude that farmers need to adjust bee management accordingly to reduce major yield gaps. High percentages of crops and non-crop plants flowering inside the homegarden can attract pollinators from adjacent source habitats that are mainly homegardens as well. In the third study (chapter 4), we aimed to assess the effect of seed predation by ants on sown seeds of four crop species (C. sativus, Daucus carota, Capsicum frutescens and Solanum melongena) and the resulting impact on the net income of small-scale farmers. Furthermore, we tested if ant seed predation differs with or without insecticide and herbicide applications. We found that ant seed predation was high for all crops (42.0 %, 49.4 %, 48.0 % and 50.6 % for C. sativus, D. carota, C. frutescens and S. melongena, respectively), potentially reducing farmers’ net income by half. Application of insecticides and herbicides did not influence ant seed predation or total ant abundance, but influenced ant species-specific abundances positively or negatively. Despite ant species-specific responses to insecticide and herbicide applications, we found consistently high seed predation rates across all gardens. We conclude that high seed predation is caused by high overall ant abundance mediated through functional redundancy of ant species. We recommend more environmentally friendly and sustainable practices such as overseeding or seedling production in nurseries to reduce chemical pest control. In the fourth study (chapter 5), we reviewed pollination services and their importance to crops focusing on another understudied geographical region - the Neotropics - to get an overview of the main pollinator taxa and the dependence on pollination services of crops. In addition, we summarized pollination research methodologies and discuss pollination relevant farm and landscape management, as well as socio-economic drivers affecting pollination services. We have shown that pollination services by wild pollinators are important for crop production. However, knowledge gaps exist in terms of the quantity, quality and stability of crop production provided by animal pollinators. It is also critical to understand how multiple socioeconomic drivers influence the selection of particular management systems and, thus, the environmental services delivered. In conclusion, pollination services are of major importance for closing yield gaps in tropical small-scale agriculture, such as homegardens. Although common management practices such as the use of agro-chemicals influence yield, they do not compensate yield gaps due to pollinator loss or ant seed predation. The enhancement of pollination services should be considered as well as the reduction of pesticide use in the majority of tropical agroecosystems. Farmers should adapt environmentally friendly and more sustainable practices adjusted to bee management considering four spatial scales to reduce major yield gaps.

Pollinating insects provide crucial and economically important ecosystem services to crops and wild plants, but pollinators, particularly bees, are globally declining as a result of various driving factors, including the prevalent use of pesticides for crop protection. Sublethal pesticide exposure negatively impacts numerous pollinator life‐history traits, but its influence on reproductive success remains largely unknown. Such information is pivotal, however, to our understanding of the long‐term effects on population dynamics. We investigated the influence of field‐realistic trace residues of the routinely used neonicotinoid insecticides thiamethoxam and clothianidin in nectar substitutes on the entire life‐time fitness performance of the red mason bee Osmia bicornis. We show that chronic, dietary neonicotinoid exposure has severe detrimental effects on solitary bee reproductive output. Neonicotinoids did not affect adult bee mortality; however, monitoring of fully controlled experimental populations revealed that sublethal exposure resulted in almost 50% reduced total offspring production and a significantly male‐biased offspring sex ratio. Our data add to the accumulating evidence indicating that sublethal neonicotinoid effects on non‐Apis pollinators are expressed most strongly in a rather complex, fitness‐related context. Consequently, to fully mitigate long‐term impacts on pollinator population dynamics, present pesticide risk assessments need to be expanded to include whole life‐cycle fitness estimates, as demonstrated in the present study using O. bicornis as a model.