Multivariate analysis of the primary production in a tropical high mountain lake in Mexico

The relationships between meteorological and physicochemical features and phytoplanktonic chlorophyll a, b and c concentrations in the water from the tropical high mountain Lake El Sol were explored with principal components analysis and canonical correlation. The first three principal components (PC) explained 73.5% of the variation. Minimum and average air temperature, wind speed, sulfates and hardness had greatest weights on the first PC, which explained 35.9% of the data variation. Rainfall, nitrates, TN/P-PO4 ratio and conductivity contrasted with chlorophyll b, maximum air temperature, and nitrite and explained 19.8% of the data variation. Cloudiness, water temperature, CO2 and pH contrasted with light and chlorophyll a and had greatest weights on the third PC, which explained 17.8% of the data variation. A biplot of PC1 vs PC2 showed a meteorological gradient that, together with a chemical gradient in the water, grouped the samples into cold-dry and warm-rainy seasons. Also, a secondary gradient of nitrate and ions, contrasting with maximum air temperature and nitrite, was defined. Finally, a third gradient of cloudiness, water temperature, CO2 and pH, contrasted with percent light attenuation and chlorophyll a, suggested a significant time lag between physical inputs and biological outputs, Canonical correlation between variables pointed out the preponderant effect of climate on the physicochemical features.