Effects of a beaver pond on runoff processes: comparison of two headwater catchments

Natural variations in concentrations of O-18, D, and H4SiO4 in two tributary catchments of Woods Lake in the west-central Adirondack Mountains of New York were measured during 1989-91 to examine runoff processes and their implications for the neutralization of acidic precipitation by calcium carbonate treatment. The two catchments are similar except that one contained a 1.3 ha beaver pond. Evaporation from the beaver pond caused a seasonal decrease in the slope of the meteoric water line in stream water from the catchment with a beaver pond (WO2). No corresponding change in slope of the meteoric water line was evident in stream water from the other catchment (WO4), nor in ground water nor soil water from either catchment, indicating that evaporative fractionation was not significant. Application of a best-fit sine curve to delta(18)O data indicated that base how in both catchments had a residence time of about 100 days, Ground water from a well finished in thick till had the longest residence time (160 days); soil water from the O-horizon and B-horizon had residence times of 63 and 80 days, respectively. Water previously stored within each catchment (pre-event water) was the predominant component of streamflow during spring snowmelt and during spring and autumn rainfall events, but the proportion of streamflow that consisted of pre-event water differed significantly in the two catchments. The proportion of event water (rain and snowmelt) in WO2 was smaller than at WO4 early in the spring snowmelt of March 13-17, 1990, but the proportions of source water components for the two catchments were almost indistinguishable by the peak flow on the third day of the melt. The event water was further separated into surface-water and subsurface-water components by utilizing measured changes in H4SiO4 concentrations in stream water during the snowmelt. Results indicated that subsurface flow was the dominant pathway by which event water reached the stream except during the peak flow of a rain-on-snow event on the last day of the melt. Streamflow from a spring rain storm with dry antecedent conditions two months later (May 16-18, 1990), was less than 5% event water at peak flow in WO2 and 26% in WO4. This change from the runoff pattern in March is attributed to retention of event water in the beaver pond favored by relatively low pre-event storage and isothermal (nonstratified) conditions in the pond that allowed mixing. Streamflow during several autumn storms was about 15-25% event water at peak flow in WO4; the highest values for event water were associated with wet antecedent moisture conditions. These results indicate that a beaver pond can significantly affect the downstream delivery of event water through evaporation and mixing, but provides minimal retention during large runoff events such as snowmelt. Beaver ponds are expected to provide greater opportunity for neutralization of acidic waters during most of the year in catchments treated with calcium carbonate, but little neutralization effect during snowmelt. (C) 1998 Elsevier Science B.V. All rights reserved.