No abstract

No abstract

Widespread observations of ecohydrological separation are interpreted by suggesting that water flowing through highly conductive soil pores resists mixing with matrix storage over periods of days to months (i.e., two ‘water worlds’ exist). These interpretations imply that heterogeneous flow can produce ecohydrological separation in soils, yet little mechanistic evidence exists to explain this phenomenon. We quantified the separation between mobile water moving through preferential flow paths versus less mobile water remaining in the soil matrix after free‐drainage to identify the amount of preferential flow necessary to maintain a two water world's scenario. Soil columns of varying macropore structure were subjected to simulated rainfall of increasing rainfall intensity (26 mm h−1, 60 mm h−1, and 110 mm h−1) whose stable isotope signatures oscillated around known baseline values. Prior to rainfall, soil matrix water δ2H nearly matched the known value used to initially wet the pore space whereas soil δ18O deviated from this value by up to 3.4‰, suggesting that soils may strongly fractionate 18O. All treatments had up to 100% mixing between rain and matrix water under the lowest (26 mm h−1) and medium (60 mm h−1) rainfall intensities. The highest rainfall intensity (110 mm h−1), however, reduced mixing of rain and matrix water for all treatments and produced significantly different preferential flow estimates between columns with intact soil structure compared to columns with reduced soil structure. Further, artificially limiting exchange between preferential flow paths and matrix water reduced bypass flow under the most intense rainfall. We show that (1) precipitation offset metrics such as lc‐excess and d‐excess may yield questionable interpretations when used to identify ecohydrological separation, (2) distinct domain separation may require extreme rainfall intensities and (3) domain exchange is an important component of macropore flow.

Preferential flow reduces water residence times and allows rapid transport of pollutants such as organic contaminants. Thus, preferential flow is considered to reduce the influence of soil matrix-solute interactions during solute transport. While this claim may be true when rainfall directly follows solute application, forcing rapid chemical and physical disequilibrium, it has been perpetuated as a general feature of solute transport—regardless of the magnitude preferential flow. A small number of studies have alternatively shown that preferential transport of strongly sorbing solutes is reduced when solutes have time to diffuse and equilibrate within the soil matrix. Here we expand this inference by allowing solute sorption equilibrium to occur and exploring how physiochemical properties affect solute transport across a vast range of preferential flow. We applied deuterium-labeled rainfall to field plots containing manure spiked with eight common antibiotics with a range of affinity for the soil after 7 days of equilibration with the soil matrix and quantified preferential flow and solute transport using 48 soil pore water samplers spread along a hillslope. Based on > 700 measurements, our data showed that solute transport to lysimeters was similar—regardless of antibiotic affinity for soil—when preferential flow represented less than 15% of the total water flow. When preferential flow exceeded 15%, however, concentrations were higher for compounds with relatively low affinity for soil. We provide evidence that (1) bypassing water flow can select for compounds that are more easily released from the soil matrix, and (2) this phenomenon becomes more evident as the magnitude of preferential flow increases. We argue that considering the natural spectrum preferential flow as an explanatory variable to gauge the influence of soil matrix-solute interactions may improve parsimonious transport models.