Classifying lakes to quantify relationships between epilimnetic chlorophyll a and hypoxia.

Excess nutrient loading increases algal abundance which can cause hypoxia in many lakes and reservoirs. We used a divisive partitioning approach to analyze dissolved oxygen profile data collected across the continental United States to increase the precision of estimated relationships between chlorophyll a (chl a) concentrations and the extent of hypoxia in the water column. Chl a concentrations predicted the extent of hypoxia most accurately in lakes that were stratified at the time of sampling with a maximum temperature gradient of at least 1.2 °C/m. Lake elevation, Secchi depth, and lake geometry ratio further refined the specification of groups of lakes with different relationships between chl a and the extent of hypoxia. The statistical relationships between chl a and the extent of hypoxia that were estimated can be used directly for setting management thresholds for chl a in particular types of lakes.