Understanding nitrogen fixation in two shallow eutrophic lakes in Central Florida

Our objective is to improve our understanding of the physicochemical and biological influences on cyanobacterial N2 -fixation in Lakes Jesup and Monroe, Florida. The overall goal of this project was to determine rates of nitrogen added to the water column due to nitrogen fixation and thus contributing to a more accurate assessment of nitrogen loading from the watershed. The isotopically enriched 15 N2 method was used to mearsure N2 fixation in a sub-set of samples to calibrate the acetylene reduction approach used at all stations. One station from each lake was pre-selected for calibration purposes. To calibrate, we compared linear regression of N2 fixation nmol L-1 h-1 as calculated by the 15 N-method compared with ethylene produced (nmol L--1 h-1) via the acetylene reduction method in the same water collected at both stations for each sampling. A statistically significant relationship was found between the two methods in both lakes however, the regression was a better fit for Lake Jesup than for Lake Monroe. We therefore built our calibration data set by using all the data from Lake Jesup and several points that were < 5 nmol L-1 h-1 as determined by the 15N-method. In general, Lake Monroe had much lower rates of fixation than Lake Jesup. It was therefore difficult to calibrate the Lake Monroe methods independently. Thus we found the best calibration resulted from a combination of the data from the two lakes. Lake Monroe provided the N-fixation values between 0 to <5 nmol L-1 h-1, and Lake Jesup data included values >5 nmol L-1 h-1. The calibration slope of 6.37, is within the range of previously published measurements/calibrations. The approach used to measure the enriched samples also provides natural abundance δ15 N values from the "untreated" POM. Isotopic values less than 1 permil and near 0 permil , typically reflect an atmospheric N source via fixation of N2, versus nitrate utilization. All data in the time series from both Lakes were <5 permil , except for one value from Lake Monroe in September of 2007, which had an extremely enriched value of 9.3 permil, indicating a different N source for this sampling. δ15N values near this level of relative enrichment are typically associated with denitrification, and or DIN coming into the watershed from pasture lands (manure). However, our data do not fall within the typical range of δ15N values for denitrification when Lake Monroe's DIN concentrations are considered.