4.1 Connectivity Strength as a Continuous Measure
Treating connectivity as a continuous value measured with geochemical and microbial metrics successfully enabled the development of target site specific connectivity functions that described site specific behavior in response to fluctuations in stage at Inflow. Our evaluation of the geochemical and microbial connectivity strength metrics confirmed that both performed well in distinguishing between target sites with high and low connectivity as identified by where tracer breakthrough curves were observed during experimental tracer injections (Figure 4). The site specific connectivity functions allowed us to predict connectivity at each target site daily throughout the main study period (May 01, 2018 through September 01, 2018) and subsequently predict connectivity behavior from previous years (2016 through 2020) using only information from the Inflow stage measurements.
Using daily connectivity predictions for the 2018 study period, we identified three major surface water connectivity regimes operating within the river-floodplain system during the study period (Figure 6). Sites were observed to have: high σm values and low spread for the majority of the study period, low σmvalues and spread, or intermittently connected with a wide range of σm values that varied with Inflow stage (Figures 5 & 6). The sites that had high σm for a large proportion of the study period were those located along the main stem of the river. Interestingly, these sites also exhibited evidence of a shift toward decreased σm at very low streamflows (Figures 2 & 5), perhaps reflecting increased transit time or alternatively, changes in source water composition within the reach. The only site to demonstrate low σm values throughout the study was Pond-Iso where we observed no surface connection to the main channel and which had a unimodal distribution of low σm values despite being geographically near Inflow (Figure 1b), demonstrating that geographic proximity does not dictate high hydrologic connectivity. At sites with intermittent connectivity regimes, interactions between river flow dynamics and floodplain geomorphic structures generated target site-specific variability in connectivity regimes. Distributions of connectivity strength at all intermittent sites, except at Main-Braid, were characterized by a dominant mode of low connectivity strength and a secondary mode of higher connectivity strength (Figure 6). This reflects that at the majority of floodplain target sites, high connectivity values are only maintained at Inflow stages well above the median stage of 424 mm during the 2018 study period.
While this bimodality at intermittently connected target sites makes a case for treating connectivity as binary, evaluation of site specific data also indicated that some sites experience prolonged periods of intermediate connectivity strength. If we consider σ values of 0.4 < σ < 0.6 to be intermediate connectivity, then during the 2018 study period, sites spent as little as 7% and as much as 33% in a state of intermediate connectivity (ranges: σm: 7%-20%; σg: 10-33%). Identifying places with longer durations of intermediate connectivity is thought to be particularly important for identifying control points in a landscape that have disproportionate influence on hydrologic and biogeochemical properties of the ecosystem (Bernhardt et al., 2017). In river-floodplain systems, sites with high durations of intermediate connectivity may act as such because they can have a substantial flux of river water moving through them but also long enough residence times for biogeochemical processes that are distinct from those in the main stem river to occur (Covino, 2017; Lynch et al., 2019).