Nutrient Atmospheric Deposition on Utah Lake: A Comparison of Sampling and Analytical Methods

We describe modified sampling and analysis methods to quantify nutrient atmospheric deposition (AD) and estimate Utah Lake nutrient loading. We address criticisms of previous published collection methods, specifically collection table height, screened buckets, and assumptions of AD spatial patterns. We generally follow National Atmospheric Deposition Program (NADP) recommendations but deviate to measure lake AD, which includes deposition from both local and long-range sources. The NADP guidelines are designed to eliminate local contributions to the extent possible, while lake AD loads should include local contributions. We collected side-by-side data with tables at 1 m (previous results) and 2 m (NADP guidelines) above the ground at two separate locations. We found no statistically significant difference between data collected at the different heights. Previous published work assumed AD rates would decrease rapidly from the shore. We collected data from the lake interior and show that AD rates do not significantly decline away from the shore. This demonstrates that AD loads should be estimated by using the available data and geostatistical methods even if all data are from shoreline stations. We evaluated screening collection buckets. Standard unscreened AD samples had up to 3-fold higher nutrient concentrations than screened AD collections. It is not clear which samples best represent lake AD rates, but we recommend the use of screens and placed screens on all sample buckets for the majority of the 2020 data to exclude insects and other larger objects such as leaves. We updated AD load estimates for Utah Lake. Previous published estimates computed total AD loads of 350 and 153 tons of total phosphorous (TP) and 460 and 505 tons of dissolve inorganic nitrogen (DIN) for 2017 and 2018, respectively. Using updated collection methods, we estimated 262 and 133 tons of TP and 1052 and 482 tons of DIN for 2019 and 2020, respectively. The 2020 results used screened samplers with lower AD rates, which resulted in significantly lower totals than 2019. We present these modified methods and use data and analysis to support the updated methods and assumptions to help guide other studies of nutrient AD on lakes and reservoirs. We show that AD nutrient loads can be a significant amount of the total load and should be included in load studies.