Biodiversity loss and conservation are increasingly coming into focus in global policy fora, requiring information and assessments at wider spatial and temporal scales than previously considered. However, the monitoring framework required to support such data collection and assessment is lacking in many countries and is not harmonized across countries, hampering these efforts. Aggregation of existing freshwater data offers a solution to the problem of assessing status and trends of ecosystems and biodiversity at large spatial scales in the absence of nationally coordinated monitoring efforts. Analysis of aggregated data from different sources, collected using different protocols and with varying levels of metadata and supporting data, can be challenging and requires decisions regarding data comparability. In this paper, we identify the challenges inherent in harmonizing aggregated freshwater data for analysis, including general concerns related to research goals, spatial and temporal scale, sample selection, sampling effort, and site integrity. We also discuss the challenges related to measured parameters, sampled habitats, sample collection and processing methods, and data integrity for phytoplankton, benthic algae, macrophytes, zooplankton, benthic macroinvertebrates, fish, and supporting variables such as water and sediment chemistry. We provide a workflow to evaluate each of these challenges and make decisions about how best to work with the data. Finally, we review a case study from a large-scale analysis of freshwater data from the circumpolar Arctic region that exemplifies the encountered challenges and the chosen solutions. Through the description of the case study, we provide practical solutions to support aggregation and analysis of existing freshwater data. As global conversations about biodiversity status and trends continue, the demand for large-scale analyses of data from different sources will only grow. In the absence of globally harmonized monitoring, we are faced with the need to ensure comparability of data, making expert judgements where needed to support sound conclusions.
Goedkoop et al. (Thu,) studied this question.