To provide a complete portrayal of the multiple factors negatively impacting insects in agricultural landscapes it is necessary to assess the concurrent incidence, magnitude, and interactions among multiple stressors over substantial biogeographical scales. Trans-national ecological field investigations with wide-ranging stakeholders typically encounter numerous challenges during the design planning stages, not least that the scientific soundness of a spatially replicated study design must account for the substantial geographic and climatic variation among distant sites. ‘PoshBee’ (Pan-European assessment, monitoring, and mitigation of Stressors on the Health of Bees) is a multi-partner transdisciplinary agroecological project established to investigate the suite of stressors typically encountered by pollinating insects in European agricultural landscapes. To do this, PoshBee established a network of 128 study sites across eight European countries and collected over 50 measurements and samples relating to the nutritional, toxicological, pathogenic, and landscape components of the bees’ environment. This paper describes the development process, rationale, and end-result of each aspect of the of the PoshBee field investigation. We describe the main issues and challenges encountered during the design stages and highlight a number of actions or processes that may benefit other multi-partner research consortia planning similar large-scale studies. It was soon identified that in a multi-component study design process, the development of interaction and communication networks involving all collaborators and stakeholders requires considerable time and resources. It was also necessary at each planning stage to be mindful of the needs and objectives of all stakeholders and partners, and further challenges inevitably arose when practical limitations, such as time restrictions and labour constraints, were superimposed upon prototype study designs. To promote clarity for all stakeholders, for each sub-component of the study, there should be a clear record of the rationale and reasoning that outlines how the final design transpired, what compromises were made, and how the requirements of different stakeholders were accomplished. Ultimately, multi-national agroecological field studies such as PoshBee benefit greatly from the involvement of diverse stakeholders and partners, ranging from field ecologists, project managers, policy legislators, mathematical modelers, and farmer organisations. While the execution of the study highlighted the advantages and benefits of large-scale transdisciplinary projects, the long planning period emphasized the need to formally describe a design framework that could facilitate the design process of future multi-partner collaborations.

The unicellular spore-forming parasites Nosema apis and Nosema ceranae are considered to be one of the causes of increased honey bee mortality in recent years. These pathogens attack their honey bee hosts through their gut, causing changes in behavioral stress responses and possibly resulting in decreased honey yield and increased honey bee mortality. The present study aimed to determine the prevalence of Nosema spp. (nosemosis) in Estonia and Latvia, as well as the persistence of the disease in previously infected hives. Currently, N. ceranae is considered the most virulent species and is predominant worldwide. However, in some regions, usually with colder climates, N. apis is still prevalent. To achieve better disease control, it is important to determine the species distribution. For this purpose, we selected 30 apiaries in Estonia and 60 in Latvia that were positive for Nosema spp. in the EPILOBEE (2012–2014) study, which was 5 years prior to the present study. The results show that, while both species are present in Estonia and Latvia, N. apis is dominant in Estonia (43%), and N. ceranae is dominant in Latvia (47%). We also found that the pathogens are very persistent, since 5 years later, only 33% of infected apiaries in Estonia and 20% of infected apiaries in Latvia, we could not detect any pathogens at the time of sampling.