Improved understanding of catchment-scale hydrology and nutrient transport in agricultural catchments is needed. Here, the annual young water fraction is determined for three southern Ontario headwater catchments using the stable water isotope δ 18 O; the dominant event contributions for three streams were also determined using a two-component isotopic hydrograph separation. On an annual average, Nissouri Creek (loamy soils, high tile drainage) sees the lowest amount of event water, followed by Big Creek (clay soils, high tile drainage) and then North Creek (clay soils, low tile drainage). Event hydrograph separations show Nissouri Creek has significantly higher median amounts of pre-event water than Big and North Creek during events. These results in indicate soil type may be more influential than tile drainage presence with respect to the contribution of event contributions to total stream flow; specifically, higher contributions of event water occurred in the two clay soil catchments, despite differing tile presence. Antecedent moisture and event characteristics did not uniformly explain the observed preevent water contributions for events across all sites, with pre-event contributions only weakly and positively correlated with discharge at event commencement and negatively with isotopic variability in stream water. The total contribution of preevent water during events was significantly correlated with mean event turbidity and the flow-weighted mean concentrations of dissolved organic carbon, total phosphorus and total dissolved phosphorus, but not total nitrogen or nitrate. The correlations we observed between pre-event water and water chemistry were stronger than between event size and water chemistry, suggesting isotopic information adds insight that simple hydrometrics are unable to provide. Our results show that event water is responsible for phosphorus loss at the watershed scale, indicating actions targeting fast pathways or targeting pools of phosphorus available to those fast pathways can aid in reducing phosphorus transported by event water.

<p>Annual and event scale water age contributions were analyzed using δ¹⁸O for three agricultural sub-catchments in southern Ontario: Nissouri Creek catchment (loamy soils, high tile drainage), North Creek catchment (heavy clay soils, low tile drainage) and Big Creek catchment (loamy clay soils, high tile drainage). The local meteoric water line was δ²H = 6.94 δ¹⁸O + 2.51 (r²=0.97, n=158, p-value<0.0001). The annual young water contribution was determined for Nissouri (0.3108), North (0.9895) and Big (0.4907). Under a two-component mixing model, hydrograph separations determined the mean total event contribution of old water for Nissouri (57%), North (41%) and Big (55%). This indicates that the catchment with the lowest presence of tile drainage had the highest contribution of young water, but also that soil type may have a stronger control on dominant flow pathways than tile drainage presence alone. Antecedent moisture had no significant correlation to event water age contributions.</p>

<p>Annual and event scale water age contributions were analyzed using δ¹⁸O for three agricultural sub-catchments in southern Ontario: Nissouri Creek catchment (loamy soils, high tile drainage), North Creek catchment (heavy clay soils, low tile drainage) and Big Creek catchment (loamy clay soils, high tile drainage). The local meteoric water line was δ²H = 6.94 δ¹⁸O + 2.51 (r²=0.97, n=158, p-value<0.0001). The annual young water contribution was determined for Nissouri (0.3108), North (0.9895) and Big (0.4907). Under a two-component mixing model, hydrograph separations determined the mean total event contribution of old water for Nissouri (57%), North (41%) and Big (55%). This indicates that the catchment with the lowest presence of tile drainage had the highest contribution of young water, but also that soil type may have a stronger control on dominant flow pathways than tile drainage presence alone. Antecedent moisture had no significant correlation to event water age contributions.</p>