It has been well documented that periphyton communities play a key role in primary productivity, nutrient cycling, and food web interactions. However, a worldwide overview of research on the key themes, current situation, and major trends within the field is lacking. In this study, we applied the machine learning technique (latent Dirichlet allocation, LDA) to analyze the abstracts of 6690 publications related to periphyton from 1991 to 2020 based on the Web of Science database. The relative frequency of classical and basic research on periphyton related to colonization, biomass, growth rate, and habitats has been clearly decreasing. The increasing trends of research on periphyton are embodied in the periphyton function in freshwater ecosystems (e.g., application as ecological indicators, function in the removal of nutrients, and application in paleolimnology), the research at macroscales (e.g., spatial-temporal variation, and functional and taxonomic diversity), and the anthropogenic themes (e.g., climate warming, response to multiple stressors, and land use type). The keyword and title analysis showed that the periphyton studies are concentrated mainly on diatom aspects, especially with respect to streams relative to lakes. The thematic space based on non-metric multidimensional scaling (NMDS) showed that the classical themes such as growth rate, colonization, and environmental factors (e.g., multiple stressors and climate warming) were most linked to other research themes. We proposed that future trends in the periphyton should focus on the function of periphyton in lakes and their response to multiple environmental pressures with the increasingly extensive eutrophication in lakes and the increasingly significant change in the climate.
Read full abstract