Multivariate Condition Assessment of Watersheds with Linked Micromaps

Abstract

A challenge for statewide stream monitoring is visualizing the spatial and statistical characteristics of such data to compare the biotic condition of watersheds and relate that condition to watershed‐level estimates of instream variables. We used linked micromaps on stream survey data of 25 subbasins (766‐5,982 km2) for biotic condition, nine water quality, and two habitat variables. Subbasin biotic condition was negatively correlated with conductivity, magnesium and sulfate concentrations, and weakly positively correlated with habitat scores of sedimentation and embeddedness, with higher scores indicating better habitat. Positive spatial autocorrelation was detected among the subbasins in both habitat variables, iron concentration, pH, and exceedances of fecal coliform criteria as shown in linked micromaps. A spatial principal components analysis reduced the 11 environmental variables to two principal axes. The first axis synthesized a gradient of water quality and habitat scores among the subbasins. Subbasin biotic condition regressed on first axis subbasin scores had a significant, negative slope and accounted for 55% of the variation. Subbasins in degraded biotic condition had elevated conductivities and ion concentrations in northern and southern subbasins, and low habitat scores in western subbasins. Through linked micromaps, we compared the biotic condition among subbasins and identified stressors prevalent among subbasins that affected biotic condition.

Spatial Convergence in Major Dissolved Ion Concentrations and Implications of Headwater Mining for Downstream Water Quality

“…In addition, spatial stream network modeling approaches (e.g., Peterson et al 2013;McGuire et al 2014;Ver Hoef et al 2014;McManus et al 2016;Scown et al 2017) are now available to further evaluate effects of headwater disturbance and cumulative downstream impacts. .…”

“…A similar approach of using assumed values of proposed mine impacts was used to develop a predictive model for downstream conductivity based on a weighted average approach at stream confluences ). In addition, spatial stream network modeling approaches (e.g., Peterson et al 2013;McGuire et al 2014;Ver Hoef et al 2014;McManus et al 2016;Scown et al 2017) are now available to further evaluate effects of headwater disturbance and cumulative downstream impacts. Such geospatial techniques are necessary to advance watershed management and next-generation watershed assessment.…”

“…Several studies have documented scale-dependent patterns and variability in water chemistry within drainage networks (Likens and Buso 2006;Zimmer et al 2013;Tiwari et al 2017;Abbott et al 2018). Evaluating spatial patterns (e.g., longitudinal confluences) is further useful for identifying areas of disturbance and previous studies have also demonstrated how spatial patterns can be important in watershed assessment (Temnerud and Bishop 2005;Peterson et al 2013;McManus et al 2016). Nevertheless, watershed assessment programs in the United States typically rely on fixed or random site selection study designs (USEPA 2003;Olsen and Peck 2008) that rarely allow investigation of spatial patterns or network analysis.…”