Effects of filtration timing and pore size on measured nutrient concentrations in environmental water samples

Nutrient monitoring is important for informing management decisions to mitigate eutrophication in aquatic systems. Many nutrient monitoring programs use filter pore sizes that allow microorganisms to pass into samples and/or wait extended times between sample collection and filtration/preservation, allowing microbial processes to alter nutrient concentrations. Here, 34 sites were sampled to determine how filter pore size and filtration timing affected measured ammonium (NH4+) and orthophosphate (ortho-P) concentrations. Three filter pore sizes (0.22, 0.45, and 0.70 mu m) were used to filter water immediately upon collection and after 5 and 22 h in a bottle. NH4+ and ortho-P concentrations varied relative to "baseline" measurements (i.e., 0.22 mu m, field-filtered samples), both over time and with different filter pore sizes, and showed no predictable direction of change based on ambient nutrient concentration or trophic status. As expected, larger relative changes occurred with lower ambient concentrations; however, for the entire dataset, samples with > 1 mu mol L-1 ortho-P and > 3 mu mol L-1 NH4+ were lower by 11 and 33%, respectively, which would result in reported nutrient concentrations that were not representative of in situ conditions. Whole-water samples filtered after 22 h varied up to 3070% for NH4+ and 480% for ortho-P from baseline concentrations. Filtering water samples with a 0.22 filter (or 0.45 mu m, at worst), immediately upon collection, should be adopted as standard practice to ensure that reported nutrient concentrations represent the most accurate measurement possible. Inconsistent and/or insufficient sampling and sample handling procedures can lead to poorly calibrated models and misinformed management and legislative decisions.