Stream bacterial diversity peaks at intermediate freshwater salinity and varies by salt type.

Anthropogenic freshwater salinization is an emerging and widespread water quality stressor that increases salt concentrations of freshwater, where specific upland land-uses produce distinct ionic profiles. In-situ studies find salinization in disturbed landscapes is correlated with declines in stream bacterial diversity, but cannot isolate the effects of salinization from multiple co-occurring stressors. By manipulating salt concentration and type in controlled microcosm studies, we identified direct and complex effects of freshwater salinization on bacterial diversity in the absence of other stressors common in field studies using chloride salts. Changes in both salt concentration and cation produced distinct bacterial communities. Bacterial richness, or the total number of amplicon sequence variants (ASVs) detected, increased at conductivities as low as 350 µS cm-1, which is opposite the observations from field studies. Richness remained elevated at conductivities as high as 1500 µS cm-1 in communities exposed to a mixture of Ca, Mg, and K chloride salts, but decreased in communities exposed to NaCl, revealing a classic subsidy-stress response. Exposure to different chloride salts at the same conductivity resulted in distinct bacterial community structure, further supporting that salt type modulates responses of bacterial communities to freshwater salinization. Community variability peaked at 125-350 µS cm-1 and was more similar at lower and upper conductivities suggesting possible shifts in deterministic vs. stochastic assembly mechanisms across freshwater salinity gradients. Based on these results, we hypothesize that modest freshwater salinization (125-350 µS cm-1) lessens hypo-osmotic stress, reducing the importance of salinity as an environmental filter at intermediate freshwater ranges but effects of higher salinities at the upper freshwater range differ based on salt type. Our results also support previous findings that ~300 µS cm-1 is a biological effect concentration and effective salt management strategies may need to consider variable effects of different salt types associated with land-use.

Freshwater salinization increases survival of Escherichia coli and risk of bacterial impairment.

Elevated levels of Escherichia coli (E. coli) are responsible for more designated freshwater stream impairments than any other contaminant in the United States. E. coli are intentionally used as a sentinel of fecal contamination for freshwaters because previous research indicates that salt concentrations in brackish or marine waters reduce E. coli survival, rendering it a less effective indicator of public health risks. Given increasing evidence of freshwater salinization associated with upland anthropogenic land-use, understanding the effects on fecal indicators is critical; however, changes in E. coli survival along the freshwater salinity range (≤ 1500 µS cm-1) have not been previously examined. Through a series of controlled mesocosm experiments, we provide direct evidence that salinization causes E. coli survival rates in freshwater to increase at conductivities as low as 350 µS cm-1 and peak at 1500 µS cm-1, revealing a subsidy-stress response across the freshwater-marine continuum. Furthermore, specific base cations affect E. coli survival differently, with Mg2+ increasing E. coli survival rates relative to other chloride salts. Further investigation of the mechanisms by which freshwater salinization increases susceptibility to or exacerbates bacterial water quality impairments is recommended. Addressing salinization with nuanced approaches that consider salt sources and chemistry could assist in prioritizing and addressing bacterial water quality management.

Fecal indicator dynamics at the watershed scale: Variable relationships with land use, season, and water chemistry.

Tracking fecal contamination in surface waters is critical to remediating water quality; however, general and source-specific fecal indicators often provide conflicting results. To understand the spatial and temporal dynamics of multiple fecal indicators and the sources they represent, we measured weekly concentrations of two general fecal indicator bacteria (FIB), a genetic indicator of human-associated Bacteroides (HF183), and surface water chemistry in nine mixed land-use watersheds in southwest Virginia, USA. At the watershed scale, general and source-specific indicators were decoupled, with distinct spatial, temporal, and chemical patterns. Random Forest analysis of individual sample variability identified temperature, watershed, nutrients, and cations as top predictors of indicator concentrations. However, these patterns - and the specific nutrients and cations identified - varied by indicator type. Among watersheds, FIB increased with developed land cover and during the summer months, while HF183 increased during the winter and only in urban watersheds. Nutrients generally related poorly to FIB and HF183, except E. coli, which correlated with total nitrogen. In contrast, all fecal indicators showed strong correlations with cations. FIB were more strongly related to calcium, magnesium, and potassium concentrations, while HF183 was related to sodium. These results suggest that, even at the watershed scale, 1) HF183 detects mainly human fecal contamination, while FIB detect broader ecosystem fecal inputs, and 2) poor correlation between specific and generalist fecal indicators is caused by unique spatial, temporal, and transport dynamics of different fecal sources in watersheds.

Constraining estimates of global soil respiration by quantifying sources of variability.

Quantifying global soil respiration (RSG ) and its response to temperature change are critical for predicting the turnover of terrestrial carbon stocks and their feedbacks to climate change. Currently, estimates of RSG range from 68 to 98 Pg C year-1 , causing considerable uncertainty in the global carbon budget. We argue the source of this variability lies in the upscaling assumptions regarding the model format, data timescales, and precipitation component. To quantify the variability and constrain RSG , we developed RSG models using Random Forest and exponential models, and used different timescales (daily, monthly, and annual) of soil respiration (RS ) and climate data to predict RSG . From the resulting RSG estimates (range = 66.62-100.72 Pg), we calculated variability associated with each assumption. Among model formats, using monthly RS data rather than annual data decreased RSG by 7.43-9.46 Pg; however, RSG calculated from daily RS data was only 1.83 Pg lower than the RSG from monthly data. Using mean annual precipitation and temperature data instead of monthly data caused +4.84 and -4.36 Pg C differences, respectively. If the timescale of RS data is constant, RSG estimated by the first-order exponential (93.2 Pg) was greater than the Random Forest (78.76 Pg) or second-order exponential (76.18 Pg) estimates. These results highlight the importance of variation at subannual timescales for upscaling to RSG. The results indicated RSG is lower than in recent papers and the current benchmark for land models (98 Pg C year-1 ), and thus may change the predicted rates of terrestrial carbon turnover and the carbon to climate feedback as global temperatures rise.

Assessing the homogenization of urban land management with an application to US residential lawn care.

Changes in land use, land cover, and land management present some of the greatest potential global environmental challenges of the 21st century. Urbanization, one of the principal drivers of these transformations, is commonly thought to be generating land changes that are increasingly similar. An implication of this multiscale homogenization hypothesis is that the ecosystem structure and function and human behaviors associated with urbanization should be more similar in certain kinds of urbanized locations across biogeophysical gradients than across urbanization gradients in places with similar biogeophysical characteristics. This paper introduces an analytical framework for testing this hypothesis, and applies the framework to the case of residential lawn care. This set of land management behaviors are often assumed--not demonstrated--to exhibit homogeneity. Multivariate analyses are conducted on telephone survey responses from a geographically stratified random sample of homeowners (n = 9,480), equally distributed across six US metropolitan areas. Two behaviors are examined: lawn fertilizing and irrigating. Limited support for strong homogenization is found at two scales (i.e., multi- and single-city; 2 of 36 cases), but significant support is found for homogenization at only one scale (22 cases) or at neither scale (12 cases). These results suggest that US lawn care behaviors are more differentiated in practice than in theory. Thus, even if the biophysical outcomes of urbanization are homogenizing, managing the associated sustainability implications may require a multiscale, differentiated approach because the underlying social practices appear relatively varied. The analytical approach introduced here should also be productive for other facets of urban-ecological homogenization.