Urbanization drives geographically heterogeneous freshwater salinization in the northeastern United States
Data files
Dec 29, 2022 version files 26.51 MB
-
README.txt
-
Utz_et_al._northeast_salinization_data.xlsx
Abstract
Rising trends in freshwater salinity, collectively termed the Freshwater Salinization Syndrome (FSS), constitute a global environmental concern. Given that the FSS has been observed in diverse settings, key questions regarding the causes, trend magnitudes, and consequences remain. Prior work hypothesized that FSS is driven by state factors, such as human-centered land use change, geology, and climate. Here, we identify the fundamental overriding factors driving FSS within the northeastern United States and quantify the diversity of FSS severity within the region. Specifically, we analyzed decadal-scale trends in specific conductance (a salinity proxy) for 333 lotic sites over four decades. Next, we quantified potential variables driving the rising or falling trends, including impervious surface cover (ISC), winter temperature and precipitation, watershed size, and ambient conductance. Temperature and ISC were considered the most likely candidates for predicting FSS severity because road salts have previously emerged as the fundamental regional driver.Most (62.5%) sites exhibited patterns of significantly increasing conductance; thus, the overall regional state reflects advancing FSS. However, others exhibited an absence of change (28.8%) or decreasing values (8.7%), and slope magnitude did change with latitude. Linear modeling demonstrated that two variables—ISC and watershed size—constitute the best predictors of long-termconductance trends and that an intercept not significantly different than zero suggests that the FSS does not reign in the absence of urbanization. We also detected areas with consistently decreasing trends despite moderate ISC. Therefore, within the region, advancing urbanization causes the typical condition of advancing FSS, but heterogeneity also exists.
Methods
All water quality data were collected from the Water Quality Portal, an online database maintained by the National Water Quality Monitoring Council. Data were collected using the following filters (selected options as listed on the portal provided in italics): states inclusive of the 10 listed above, site type = Stream (NWIS, STEWARDS, STORET), sample media = Water (NWIS, STEWARDS, STORET), characteristic = Specific conductance (NWIS, STORET), a date range between 1 Jan.1980 and 12 Jan. 2021, and minimum results per site = 100. Sites with records that spanned <10 yr were omitted from further analyses. Included here are also metrics depicting land use and climate attributes for each site.