Nutrient dynamics in carbonate vs. non-carbonate streams: Seasonality, sources, and relative loading

Background/Question/Methods

Nutrient impairment is a primary degrading anthropogenic influence on surface water.  Less recognized is the role that groundwater plays in the storage and conveyance of nutrients, particularly in karst watersheds where rapid infiltration and lateral transmission through conduit-flow allow agricultural and residential nutrients to bypass biogeochemical processing in the soil system.  This study compared the nutrient dynamics of carbonate (limestone) and non-carbonate streams within the Great Valley in Central Pennsylvania, a historically forested agricultural valley within the Chesapeake Bay Watershed. Nitrate-N and orthophosphate were surveyed for ten months in six carbonate spring creeks and five non-carbonate streams.  Additionally, samples of in-stream flora were collected in April, July, and November and analyzed for percent C and N, and natural abundance ¹⁵N and ¹³C.  

Results/Conclusions

Nitrate-N concentrations in limestone streams were significantly greater than those of non-carbonate streams during nine months of the study (P = 0.002 - 0.004). Nitrate-N concentration in the limestone streams peaked in August at 7.1 mg/L and fell to a minimum of 2.6 mg/L in April. Nitrate-N concentrations exhibited two peaks in the freestone streams, 2.2 mg/L in May and 1.9 mg/L in December, and fell to a minimum of 0.5 mg/L in September.  Letort Spring Run, a USGS gauged carbonate stream in this study, carried an annual mean of 4.3 mg/L of nitrate-N. Taken with mean annual discharge, estimated NO₃-N flux from this spring creek exceeds 170,000 kg/year, much of which resurges directly with groundwater at spring sources.  Spring creek orthophosphate concentrations were not significantly different from those of freestone streams except during the month of August, when the average concentration in freestone streams was comparatively elevated (P = 0.05).  Our results suggest that 1) limestone watersheds more efficiently transport nitrate-N to their surface water discharges than do non-carbonate watersheds and 2) the two stream-types exhibit virtually opposite seasonal patterns in nitrate concentration.  Preliminary δ¹⁵N ranged from -0.4 to 26.5 ^o/oo in spring creek autotrophs, with higher values indicating anthropogenic N sources occurring during summer months, and in watercress, Elodea, and filamentous algae.