Abstract:Understanding the influence of storm events on nitrate (NO 3 ) dynamics is important for efficiently managing NO 3 pollution. In this study, five sites representing a downstream progression of forested uplands underlain by resistant sandstone to karst lowlands with agricultural, urban and mixed land-use were established in Spring Creek, a 201 km 2 mixed land-use watershed in central Pennsylvania, USA. At each site, stream water was monitored during six storm events in 2005 to assess changes in stable isotopes of NO 3 (υ 15 N-NO 3 and υ 18 O-NO 3 ) and water (υ 18 O-H 2 O) from baseflow to peakflow. Peakflow fractions of event NO 3 and event water were then computed using two-component mixing models to elucidate NO 3 flow pathway differences among the five sites. For the forested upland site, storm size appeared to affect NO 3 sources and flow pathways. During small storms (<35 mm rainfall), greater event NO 3 fractions than event water fractions indicated the prevalence of atmospheric NO 3 source contributions at peakflow. During larger storms (>35 mm rainfall), event NO 3 fractions were less than event water fractions at peakflow suggesting that NO 3 was flushed from stored sources via shallow subsurface flow pathways. For the urbanized site, wash-off of atmospheric NO 3 was an important NO 3 source at peakflow, especially during short-duration storms where event water contributions indicated the prevalence of overland flow. In the karst lowlands, very low fractions of event water and even lower fractions of event NO 3 at peakflow suggested the dominance of ground water flow pathways during storms. These ground water flow pathways likely flushed stored NO 3 sources into the stream, while deep soils in the karst lowlands also may have promoted NO 3 assimilation. The results of this study illustrated how NO 3 isotopes and υ 18 O-H 2 O could be combined to show key differences in water and NO 3 delivery between forested uplands, karst valleys and fully urbanized watersheds.