“…Seminatural land cover was estimated to be have a stronger positive influence on EQR TDI than the negative influence of either TP or SRP (Supplemental Tables S4.1, S5.1, S5.2, S6), and combined evidence from different modeling approaches shows that land cover had an overriding influence on the ecological response, which is in line with other studies that also found a hierarchy of stressors from land cover, followed by physicochemical (e.g., sediment and nutrient concentrations) and hydromorphological (e.g., river bed and riparian corridor characteristics) variables (Villeneuve et al, 2018). Here, seminatural land cover is likely acting as a proxy for other factors (Segurado et al, 2018), such as varying bioavailability of P fractions (Ellison and Brett, 2006; Prestigiacomo et al, 2016; Stutter et al, 2018), river morphology, riparian and/or aquatic habitat structure, and absence of toxic contaminants (e.g., herbicides). In addition, the bioavailability of P forms has been shown to vary between 12 and 73% for TP and between 6 and 81% for particulate P in catchments dominated by different land cover types (Ellison and Brett, 2006; Egemose and Jensen, 2009; Poirier et al, 2012; Baker et al, 2014).…”