Patterns of Nutrient Limitation of Phytoplankton in Mountain Lakes Using a Hierarchical Modelling Approach

Nutrient limitation of phytoplankton growth in a lake has various effects on lake function and primary producer composition. There is a rich history of using short-term enrichment experiments to infer the type and degree of limitation in lakes. However, most experiments are analysed independently of one another, lack a common framework for expressing uncertainty, and only represent a single point in time. We conducted a series of phytoplankton nutrient limitation experiments in mountain lakes across northwestern Montana, USA to (a) assess regional patterns of nutrient limitation and (b) examine limitation through time in one large lake. We use a multilevel Bayesian model to better characterise parameter uncertainty across all experiments. The fertilisation effect of phosphorus was consistently larger than that of nitrogen, reflecting low phosphorus concentrations across lakes in this region. Most experiments showed evidence for P limitation or serial P limitation, with some evidence for independent co-limitation as well. We also show that nutrient limitation in one lake changed over both a growing season and across years from strong P limitation to serial P limitation. Future work should continue to explore methods for analysing collections of nutrient limitation experiments.