Integrating citizen science data to environmental impact assessments of pollinators in Sweden to improve the understanding of risk

Citizen science data on biodiversity may supplement the shortage of systematically generated biodiversity data, and improve our understanding of risk factors lying behind environmental impacts in landscapes more or less... [ view full abstract ]

Citizen science data on biodiversity may supplement the shortage of systematically generated biodiversity data, and improve our understanding of risk factors lying behind environmental impacts in landscapes more or less affected by anthropogenic stressors. Whereas, large and spectacular species for long have been popular in voluntary observation records, the emergence of citizen science data on the “smaller majority” of species, such as insects, provide an opportunity to use this new data to fill knowledge gaps. However, careful consideration of any data-specific differences in observation errors is required when integrating data of different quality and origin. Unlike data on birds and vascular plant, biodiversity monitoring data of wild pollinators, such as Bumblebees, have, until recently, been lacking. This might be due to the former lack of appropriate identifications guides and insects elusive behaviour. Wild pollinators are important for conserving biodiversity per se and provide pivotal pollination services.
In this project, we looked at the possibility to integrate bumblebee data from a Swedish national monitoring scheme and data from the Species gateway (www.artportalen.se), the latter a database where citizens contribute with their observations in Sweden on a voluntary basis. The two data sources were used to build a predictive model of trends and impacts on wild pollinators in Sweden, under influence of climate and land-use changes. Data from both sources were integrated by treating pollinator biodiversity as a partially observable process. This process was set up as a Joint Species Mixture Distribution model to allow information on spatial co-occurrence, land-use and climate dependency being shared between multiple species of wild pollinators. Using the bee model as a case-study, we discuss the limits, benefits and disadvantages by integrating citizen science data when building predictive biodiversity models for impact assessments.