Nutritional ecology is a key determinant of bee health, yet species-specific preferences and how bees respond nutritionally to real-world pesticide exposure remain unclear. We collected pollen stores from three managed bee species in 128 sites across two widespread agro-ecosystems (i.e., oilseed rape fields and apple orchards) in eight European countries. We measured protein content, lipid content, protein-to-lipid ratio as well as pesticide residues in pollen stores. We provide the first evidence of species-specific macronutritional patterns in the pollen stores of three managed bee species. The buff-tailed bumble bee stored pollen with significantly lower lipid content (mean: 44.1 mu g/mg) than the European honey bee (mean: 57.2 mu g/mg) and the red mason bee (mean: 54 mu g/mg). This reduced lipid content translated into a higher protein-to-lipid (P:L) ratio in the buff-tailed bumble bee (mean: 8.53) when compared to the European honey bee (mean: 5.85) and the red mason bee (mean: 5.60). Pesticide risk, measured as toxicity-weighted concentrations in pollen stores, did not influence P:L ratios in any species. However, increasing pesticide risk was associated with reduced protein content and lipid content in buff-tailed bumble bee stores, potentially leading to suboptimal colony development, whereas it was associated with increased protein content in red mason bee stores. Our findings suggest that bees maintain consistent macronutritional profiles in their pollen stores under pesticide exposure, at least in terms of P:L ratios, suggesting limited capacity to adapt their macronutritional dietary choices to mitigate agrochemical stress. These findings highlight a critical need to balance floral enhancements with a reduction in pesticide use to safeguard pollinators in agricultural landscapes.

Trace elements are persistent contaminants that enter bee diets through pollen, nectar, soil-derived particles and atmospheric deposition, yet field-based exposure data for pollinator taxa remain limited. We examined 18 trace elements (Al, As, Ba, Cd, Co, Cr, Cu, Fe, Hg, Mg, Mn, Mo, Ni, Pb, Se, U, V, Zn) in pollen stored by managed western honey bees (Apis mellifera), buff-tailed bumble bees (Bombus terrestris), and red mason bees (Osmia bicornis) from 128 apple orchards and oilseed rape fields across eight European countries. We assessed species-, crop- and landscape-related drivers of element accumulation in stored pollen and honey bee workers, and used these data to estimate dietary exposure and risk. Stored pollen showed crop- and species-specific differences in trace element levels. Red mason bee pollen contained several-fold higher concentrations than honey bee and bumble bee pollen, suggesting that honey bees may not be a suitable model for pollen-based exposure studies. Stored pollen from apple orchards contained higher Cu than that from oilseed rape sites, likely linked to Cu-based fungicide use. Landscape characteristics, including crop and urban cover, influenced element concentrations in stored pollen and honey bee workers. Acute risks were assessed for As, Cd, Cu, and Hg, and chronic risks for As and Cd. For elements with available toxicity endpoints, estimated dietary risk was negligible or low; for elements lacking such endpoints, risk could not be quantified despite exposure estimates. Higher exposure of solitary bees, together with scarce toxicity data, indicates that future studies should focus on establishing species-specific toxicity endpoints.

Abstract Predicting species interactions remains a major challenge, as multiple species attributes operate simultaneously and their relative importance may vary seasonally and with temporal resolution. Here, we assess how the relative contributions of abundance, trait matching, and phenology to plant-pollinator interactions vary through a flowering season and across temporal resolutions using interaction data from three European botanical gardens. We show that predictive models explain a substantial proportion of variation in visitation patterns, with floral and pollinator abundances consistently explaining most variation across the flowering season. Trait matching between pollinator body size and floral size also contributes to visitation patterns, playing a secondary but persistent role in shaping interactions, while phenology plays a relatively minor role at broader temporal resolutions but becomes more important at finer temporal scales. Additionally, null models accounting for spatio-temporal variation in floral and pollinator abundance reveal consistent patterns of pollinator preference and avoidance, indicating that abundance alone cannot explain the observed interaction patterns. Our results show that seasonal variation and temporal resolution differentially shape the importance of species attributes, highlighting the need for multi-variable approaches that account for temporal dynamics to accurately explain and predict ecological interactions.

Viruses can impact individual host fitness and host population dynamics, especially following host shifts. Thus, the decline of wild solitary bee populations over the last few decades may be linked to viruses or other pathogens. However, evidence for the impact of viruses-transmitted from other genera or resident in solitary bees-on their fitness remains scarce. Here, by assessing solitary bee (Osmia cornuta) foraging, offspring sex ratio, survival and body mass across seven locations in northern Switzerland, we show that resident viruses-but not honeybee-associated viruses-can impact fitness proxies in the field. Loads of Osmia-resident viruses (Ganda bee virus-GABV; Scaldis River bee virus-SRBV) and honeybee-associated viruses (black queen cell virus-BQCV; deformed wing virus B-DWV-B) were quantified in foraging females. Prevalence and loads of GABV and SRBV were higher than BQCV and DWV-B. Females with high SRBV or GABV loads had reduced offspring survival or lower male offspring body mass, respectively. Honeybee-associated viruses had no impact on O. cornuta fitness proxies. We demonstrate that viruses can affect solitary bee fitness negatively, but the degree of impact varies with viral species and provenance. Further research is needed to unravel the dynamics of multi-host pathogens in pollinator communities.

Viral transfer from managed pollinators potentially threatens wild pollinators and may be exacerbated by land-use changes. Our causal models and plant-pollinator network data from 48 European urban and agricultural landscapes revealed the ecological mechanisms underpinning viral transmission. Host identity, network architecture and land-use modulated viral dynamics (black queen cell virus, BQCV; deformed wing virus, DWV-A and DWV-B). Viral prevalence in wild pollinators was driven by viral density in the reservoir host: honey bees, and secondarily by trophic niche overlap with these managed pollinators. Modular networks limited BQCV prevalence, which was driven by reduced honey bee niche overlap, suggesting minimal onward transmission among wild pollinators. Landscapes supporting greater wild pollinator abundance diluted DWV-B transmission; in urban landscapes managed honey bees and wild pollinators experienced higher and lower BQCV prevalence, respectively. Disease in managed bee colonies and land-use changes that concentrate pollinator foraging interactions present potential viral risks to wild pollinator health.

Global declines in insect pollinators have triggered the need for standardised, scalable methods to monitor pollinator populations. Automated photomonitoring and machine-learning-based identification are emerging as promising non-lethal approaches for large-scale insect observation, but their effectiveness requires reliable and consistent monitoring designs. Artificial flowers offer a solution to limitations associated with natural flowers, including variability in floral availability, lack of standardisation, and wind-induced noise in visual data, yet their performance under real field conditions remains largely untested.Here, we developed 3D-printed artificial flowers, using Malva multiflora (Malvaceae) as a model species, incorporating increasingly complex sensory cues and rewards. We evaluated their performance by comparing insect pollinator species richness and visitation rates in artificial and co-occurring natural flowers using visual observations across 16 field sites in Southern Spain. Although artificial flowers received four times fewer species and visits per flower than natural flowers, they nonetheless accounted for around 50% of all recorded species. Variation in sensory cues among artificial flower treatments had little influence on attractiveness. Importantly, pollinator richness and visitation rates on artificial and natural flowers were strongly positively correlated, indicating that artificial flowers can provide reliable relative indicators of pollinator diversity and activity. Smaller-bodied bee species and those with warmer climatic affinities were more likely to visit artificial flowers, suggesting trait-based differences in exploratory behaviour or floral discrimination.Our findings demonstrate the potential of 3D-printed artificial flowers, showing that, while not direct substitutes for natural flowers, they can function as effective, reproducible, and scalable tools for standardised pollinator monitoring.

Agricultural landscapes feature marked seasonal changes in the quality and quantity of habitats and floral resources supporting pollinating insects. Seasonal dynamics can affect the structure of plant-pollinator interactions, yet the relative importance of both landscape elements with spatio-temporal dynamics and those elements that are more static in space and time remains largely unknown. Such an understanding is needed to identify resource-mediated modifications of plant-pollinator network structures and their functional and management implications. To understand the spatio-temporal effects of landscape heterogeneity on the structure of plant-pollinator networks, we sampled plant-pollinator (Apiformes-except Apis mellifera; Syrphidae) communities over three seasonal periods in 12 landscapes in central Germany. The landscapes comprised spatial gradients in the proportion of semi-natural habitat cover and edge density. To assess temporal changes, we evaluated the cover of mass-flowering crops in bloom, floral diversity and honey bee density at each plant-pollinator sampling event. Spatio-temporally dynamic characteristics, particularly the cover of mass-flowering crops, were more important than static characteristics in explaining variation in plant-pollinator network structure across the three seasonal periods. The richness of plants and pollinators was generally lower when the proportion of mass-flowering crops was high. Under such conditions, networks were more connected, with greater niche overlap among pollinators, and decreased network specialization (H2 '). Richness was higher in landscapes with high edge density, with an increasing effect on network connectance up to a certain threshold. The proportion of semi-natural habitat cover and floral diversity had differential effects on the richness of plants and pollinators, with strong effects on the dietary niche overlap of the pollinators, potentially indicating a decrease in competition when semi-natural habitat cover and flower diversity are high. Synthesis and application. To better support plant-pollinator communities in agricultural-dominated landscapes, we suggest incentivizing the planting of complementary floral resources and preserving or restoring semi-natural habitat areas. Especially in intensively used agroecosystems, the negative effects of mass-flowering crops can be mitigated by maintaining flower-rich edge habitats and relatively small field sizes, which help support plant and pollinator communities, avoid potential negative effects of exploitative competition, and ensure the sustainability of pollination services via increased functional redundancy.

Background Wild pollinator populations in the European Union lack defined, quantitative Specific Protection Goals (SPGs) for pesticide risk assessment. Without an agreed threshold, lower-tier toxicity data cannot be evaluated for acceptability, nor can higher-tier study results be assessed against a protection goal, leaving the tiered ERA process at an impasse. The existing honey bee surrogate compounds the problem: this was selected for laboratory convenience rather than functional representativeness, and their managed, colonial life history cannot reliably predict outcomes for the diverse life-history strategies of wild pollinators. Results This paper presents a pragmatic, staged framework combining trait-based species selection with demographic toxicity assessment. It focuses on Stage 1 (non-spatial) of a the three-stage process, where life-stage-structured population models are implementable now (2026–2027), uses existing data and tools, and generates quantitative, population-level SPG thresholds within the current [1] tiered assessment framework. The framework combines two methodological advances: trait-based species selection identifies functionally vulnerable species by grouping taxa according to exposure- and sensitivity-related traits and population resilience; and demographic toxicity assessment translates individual-level pesticide effects into population-level outcomes through life-stage-structured projection matrices. Parameter space exploration demonstrates that Stage 1 yields defined, quantitative thresholds: a maximum acceptable population decline and a maximum acceptable exceedance risk probability. Conclusions Stage 1 resolves the immediate operational constraint by providing provisional, population-level thresholds for regulatory use. A roadmap for adaptive implementation and progressive refinement is provided: Stage 2 (operational 2028–2029) incorporates landscape-scale spatial dynamics, and Stage 3 (operational 2030+) implements full agent-based modelling using the ALMaSS system. The approach is demonstrated for bees (bumble bees and solitary bees), but the trait-based framework is designed to be flexible enough to generalise across hoverflies, butterflies, moths, and other pollinating taxa.

Temperate agricultural landscapes are experiencing unprecedented biodiversity declines. Landscape simplification is commonly identified as a driver of species loss across taxonomic groups, but the contribution of crop and non-crop habitats to farmland biodiversity conservation is surprisingly poorly known. Using 86 paired permanent grasslands and oilseed rape fields in five European countries, we assess how habitat type shaped plant, butterfly, wild bee, and carabid assemblages and whether increasing grassland amount in surrounding landscapes fosters the spillover of grassland-associated biodiversity to oilseed rape fields. We find habitat type rather than landscape-level grassland amount determines diversity and shapes species assemblages: plants and butterflies are more diverse in grasslands, while wild bees and carabids are equally or more diverse in oilseed rape fields. Increasing landscape-level grassland amount affects species assemblage composition but only reduces turnover between habitats in wild bees. Overall, both grasslands and oilseed rape fields harbour distinct sets of species, together contributing complementarily to regional diversity. Safeguarding biodiversity in agricultural landscapes therefore requires not only the conservation of permanent semi-natural habitats but also biodiversity-friendly management of disturbed habitats such as crop fields that can contribute valuable species.

Translating global commitments to protect and restore nature into context-specific action remains difficult. Earth-system functions, including carbon storage, erosion regulation and pollination, are spatially heterogeneous, legacy-rich and rarely captured by a single global indicator. We introduce Earth-positive futures, a trajectory-based framework for assessing focal Earth-system functions and guiding efforts to recover, safeguard or strengthen them. The framework focuses on the wicked intermediate scale, the landscape-toregional and decadal-to-millennial domain where no single driver generally determines trajectories. Here, interacting natural and human processes create complex dynamics, but also reveal intervention points where actions can be planned, monitored and revised. For each use case, Earth-positive futures translate the complexity of a focal function and its controls into a tenable, decision-relevant suite of essential variables and driver metrics. Past-to-future trajectories built from these variables and metrics connect reference envelopes, degradation, legacies and future scenarios, and provide the basis for co-defining an Earth-positive future state, including target ranges for essential variables and guardrails for selected drivers. These trajectories can draw on harmonised time-series from palaeoproxy-, monitoring-, and model-derived data. An adaptive steering pathway is aligned to the trajectory; it details the phase-out of harmful practices and the phase-in of restorative, protective or transformative interventions. Steering denotes co-designed, revisable attempts to redirect selected essential-variable trajectories rather than controlling complex Earth-system functions. We illustrate the framework with palaeoproxy-supported forest-cover trajectories, demonstrating how reference envelopes, biodiversity trade-offs and colonisation credits can inform site selection and pathway design. Earthpositive futures evaluate whether focal Earth-system functions move toward recovery or strengthening, or remain within safeguarded target ranges, while supporting case-specific, trajectory-based Theories of Change.

Agricultural intensification reduces the diversity and continuity of floral resources, exposing pollinators to multiple stressors. Although resource limitation is a key driver of pollinator declines, the mechanisms linking environmental conditions to pollinator performance remain unclear. We tested whether alleviating food limitation improves bumble bee colony performance and whether agricultural intensification constrains worker development and reproduction. We hypothesised that supplemental nutrition would enhance colony performance, particularly by reducing developmental instability and enhancing reproductive success, whereas agricultural intensification would negatively affect worker development and reproduction. We placed Bombus terrestris colonies across 24 locations spanning a gradient of agricultural cover and experimentally supplemented a subset with pollen and nectar. We quantified worker body size, wing asymmetry, and reproductive output in relation to supplementation, agricultural cover and floral resources. Food supplementation did not increase body size but reduced wing asymmetry and enhanced reproductive output, indicating that nutritional stress constrains developmental stability and colony fitness. Worker body size declined with increasing agricultural cover across treatments, suggesting a consistent developmental cost of landscape intensification that was not alleviated by additional food provisioning. Furthermore, we found weak evidence for a positive effect of agricultural cover on wing asymmetry and of floral diversity on reproductive success across treatments. These results demonstrate trait-specific responses to environmental stress: nutritional limitation primarily constrains developmental stability and reproduction, whereas worker body size appears to be more strongly shaped by landscape-level conditions. This highlights the importance of maintaining sufficient and continuous floral resources to support pollinator populations in intensively managed agricultural systems.

Roads are vital for human societies, yet they can also have negative impacts on the ecological communities that live in close proximity to them. Insect pollinators, which nest and forage in road verges running alongside roads, are a group of particular importance. These verges may act as an "ecological trap," drawing insect pollinators into contact with traffic, increasing the risk of pollinator-traffic collisions. Spanning six European regions, we evaluated the complex relationships between traffic, road verge floral composition, and surrounding land use to understand how these factors influence abundance and richness of bees, butterflies, and hoverflies sampled within road verges. Across the study, we observed 10,960 pollinators belonging to 293 species of bees, butterflies, and hoverflies. We observed greater pollinator abundance in verges with higher flower cover, and greater pollinator richness in verges with more species of flowering plants. Lower abundances of bees and butterflies and lower species richness of bees were observed when traffic speed in the adjacent road was higher. This study indicates that road verges with abundant and diverse floral resources support more abundant and diverse pollinator populations, especially on verges alongside lower speed roads. We recommend that lower speed roads should be prioritized for floral enhancements.

Due to the high diversity of both habitats and wild species in China, biological communities are characterized by a high functional diversity. Butterfly communities are often classified into two functional groups, specialists and generalists. But a finer trait-based classification might be needed for a better understanding of community assembly processes. Here, we used larval diet, voltinism, and dispersal ability to identify functional groups and relate their responses in alpha and beta diversity to gradients in human disturbance, altitude, and resource availability. Across all species, we found a negative impact of anthropogenic disturbance on abundance, a decrease in species richness with altitude, and a positive effect of resource availability. We further found that butterfly communities are best classified into three functional groups: univoltine specialists, bivoltine specialists, and generalists, whereby univoltine specialists showed the strongest dependence on resource availability compared to bivoltine specialists and generalists. Overall beta diversity responded strongly to the effective distance, the mechanisms structuring local communities differed among the functional groups. Univoltine specialists were mainly structured by environmental filtering. Bivoltine specialists were driven by mobility limitation and environmental filtering via resource availability. Generalist species were entirely driven by mobility. Our study highlights the importance of considering multiple functional traits to assess impacts of environmental change on community assembly. This approach enabled us to differentiate between primary ecological processes such as habitat filtering and mobility limitation. We conclude that the conservation of univoltine specialists should be a priority with a focus on targeted habitat management and conservation strategies.

Wild insect pollinators contribute significantly to agricultural productivity, biodiversity, and ecosystem functioning. Wild pollinators are increasingly affected by multiple interacting stressors. Proactively identifying emerging risks and feasible mitigation strategies will be critical to ensuring the long-term stability of wild pollinators biodiversity and pollination services. We conducted the first continental scale horizon scan focused on wild pollinators in Europe. A structured Delphi-based approach was used to identify emerging issues that may have significant implications for wild pollinators over the coming decade. Ten priority issues were identified, including both potential risks and opportunities. For the first time in a pollinator-focused horizon scan, legislation was identified as a key opportunity, with the European Union Nature Restoration Regulation recognised for its potential to influence pollinator conservation through mandatory restoration and monitoring targets. In contrast, political developments such as the rise of populist parties and post-truth discourse may impede policy implementation. Several issues relating to pesticide use were also identified, including developments in RNA interference technologies and precision application methods, which may reduce non-target impacts if risks are appropriately assessed. These findings provide a foundation for further research and policy evaluation in support of pollinator conservation under changing environmental and political conditions.

Biodiversity has been declining in the last decades, including pollinating insects. As an important ecosystem service, pollinators contribute to functioning ecosystems and sustainable crop yields. There have been growing concerns that competition for floral resources between managed and wild pollinators might contribute to the current decline of pollinators, however comprehensive field data to support these concerns are scarce, particularly from northern Europe. Here, we focus on semi-natural calcareous grasslands to assess potential competition between managed and wild pollinators within the context of habitat restoration. Based on transect walks at 30 sites across Estonian islands and coastline, most of which were restored from overgrown areas to open grasslands, we measured the abundance and species richness of pollinator species from five groups (bumblebees, butterflies and burnet moths, hoverflies, solitary bees, honey bees). We assessed the potential impact of honey bee competition by relating the number of honey bee colonies within a 2 km radius to the abundance and species richness of wild pollinators while correcting for the restoration status of the grassland (restored vs reference sites), local flower cover, and the percentage of forest within a 2 km radius. We found that honey bee colony density had no effect on the abundance or species richness of wild pollinators. Flower cover had some relevance, while the other variables revealed no effects. In conclusion, our results do not indicate detrimental effects of managed honey bees on flower-visiting wild pollinators at current colony densities in our study region.

1. As there is growing interest among non-specialists in participating in pollinator conservation initiatives, developing proxies for wild bee species richness that could be potentially used by non-experts can aid conservation, decision support systems for managers and policymakers, and create entry points for future taxonomists. 2. We used data from 63 independent studies in which wild bees were sampled, covering all main European habitats and climates. We tested two proxies for wild bee species richness: abundance of all wild bees (excluding the honeybee) and abundance of bumblebees. These proxies require basic taxonomic training and have the potential for routine implementation by non-experts. 3. Within-region, the abundance of wild bees was a strong predictor of wild bee species richness at the local scale, with an average correlation exceeding 0.80. Bumblebee abundance was a poorer proxy for total wild bee species richness (correlation coefficient of similar to 0.55) and was unsuitable for warm-temperate and Mediterranean climates. Observed abundance-richness correlations were consistent across climates, habitats and sampling methods, suggesting that, after a simple training, counting all bee individuals in transect walks provides a non-lethal, robust estimate of wild bee species richness. We observed a weak negative relationship between sample coverage and the strength of the abundance-richness correlation. 4. Synthesis and applications. The use of a simple but effective proxy could be an important starting point for the expansion of wild bee monitoring initiatives at the regional scale, particularly given the growing involvement of non-specialists. We provided here a simple implementation framework to use this proxy in rapid biodiversity assessments, such as the evaluation of payment-by-result schemes by farmers and measuring the efficacy of conservation actions in urban green areas or protected areas by citizens and site managers, respectively. It is important to stress that we are not advocating for the replacement of the urgently needed long-term monitoring of pollinator status and trends. Instead, we propose that employing an effective proxy for non-experts could enhance the evaluation of many local and regional conservation initiatives that currently lack any basic assessment schemes.

Because of its close ties to numerous ecological and life history characteristics, body size is regarded as one of an organism's most important characteristics and is frequently considered a significant indicator of fitness. According to recent studies, ectotherms in particular, may see a reduction in body size with rising temperatures. How life history and ecological traits influence, however, shifts in butterfly body size in response to environmental changes, particularly focusing on the effects of temperature and land use is poorly studied. Using Generalized Additive Models (GAM), we analyzed forewing length data alongside various life history (phenology, overwintering developmental stage, voltinism, diet breadth, gender) and ecological traits (mobility, thermal tolerance) as well as environmental parameters for a period that lasts 110 years. Smaller body sizes were linked to early-season emergence and increasing forest cover, while larger sizes were linked to longer flight durations and later seasonal appearance. Males exhibited a pronounced decline in body size while females showed an opposite trend, suggesting sex-specific vulnerabilities to climate change. This research highlights the complex interplay between climate change, habitat fragmentation, and butterfly morphology, emphasizing the need for further investigation into sexual size dimorphism as anthropogenic influences continue to reshape butterfly populations. ### Competing Interest Statement The authors have declared no competing interest.

Climate change has had strong impacts on biodiversity, including well-documented shifts in the distributions and phenology of species. Reductions in body size represent a third pervasive biological response; it has garnered significantly less attention despite great ecological relevance. Theoretical frameworks-the temperature-size rule, Bergmann's rule, and James's rule-predict that warmer temperatures are associated with smaller body sizes in ectotherms. In contrast, empirical evidence concerning this pattern is mixed across both taxa and environments. In the present study, we applied computer vision techniques to historical data from two large butterfly museum collections, totaling 593 species across 10 terrestrial biomes over more than a century, in order to investigate long-term trends in butterfly body size. We measured forewing length through both manual image analyses and automated computer vision algorithms proxying body size and analyzed trends by using generalized additive models in order to consider a temporal, biome-specific pattern assessment. We have tested two hypotheses: that butterfly body size (1) declines over time, in conjunction with increasing ambient temperatures, in agreement with the temperature-size rule, and (2) its variation is more marked in warm, dry biomes. Results indicate a significant overall reduction in the body size of butterflies during the last century and that this reduction is indeed more pronounced in those biomes facing higher rises in temperature. These findings constitute large-scale evidence in support of the temperature-size rule and indicate a potential ecological impact of climate change on butterfly populations. ### Competing Interest Statement The authors have declared no competing interest.

Pesticide use and habitat loss are major anthropogenic drivers of bee decline, raising global concerns about impaired crop pollination. However, the relative importance of these stressors and their combined impact on bee assemblages comprising species with different traits, such as body size or nesting strategy, remains unknown. Here we addressed these key knowledge gaps in a global quantitative synthesis analysing bee assemblage data from 681 crop fields across three continents. We found that both local pesticide hazards and decreasing proportions of semi-natural habitats in surrounding landscapes negatively affected wild bee abundance and species richness in crop fields, while pesticides additionally reduced functional and phylogenetic diversity. Semi-natural habitat availability did not buffer against these negative pesticide effects, nor did we identify any specific traits rending bees more vulnerable to one of the two drivers. Our findings highlight the pressing need to reduce non-target effects of pesticide use and emphasize that conservation and restoration of semi-natural habitats successfully promote wild bees, but are insufficient strategies to mitigate pesticide-driven losses of wild bee pollinators from crop fields.