Variation in Hydrologic Metrics
PCA analysis indicated the five flow metrics were highly variable across the observed intermittency gradients (Figure 3). Because we were focused on describing the broad patterns of variation between hydrologic metrics and sites, we opted to primarily interpret the first two PCA axes (Table 3). The first PCA axis explained 47.97% of variation, and was predominately driven by skew and pulse count, while the second PCA axis explained 39.13% of variation, and was driven by reversals, number of no-flow days, with some influence of reversals and the date of the first no-flow day. Sites that fully dried were negatively associated with reversals, pulse count and skew, and sites that dried to isolated pools were more positively correlated with these metrics. Each site had a relatively unique hydrologic signature, as evidenced by the variable impact of each flow metric on the locations of sites within the PCA. The relationships between drying regime (fully dry, drying to isolated pools, or remaining flowing) and the variability of our selected hydrologic metrics, underscores the variability of intermittent streams at small spatial scales.