Monohan, Carrie E. 2004. Riparian buffer function with respect to nitrogen transformation and temperature along lowland agricultural streams in Skagit County, Washington. Ph.D.
The impact of riparian buffers on salmonid habitat in lowland agricultural streams is not well understood. This study consists of three parts: the first describes the water quality of agricultural streams in row-crop and pasture areas; the second examines instream temperature among streams with different buffer widths; and the third investigates nitrogen transformation in riparian buffers using push-pull nitrate injections.
Grab samples were taken on a monthly basis during May-October of 2001, 2002 and 2003 from streams in row-crop and pasture areas. Row-crop streams had significantly higher concentrations of total nitrogen, ammonium, organic nitrogen, and total phosphorus than pasture streams without buffers. Large buffers composed of red alder (Alnus rubra) in pasture areas had significantly higher concentrations of nitrate than small (12m) buffer and conifer-dominated streams (p<0.1).
To investigate instream temperatures the average water temperature at 4:00 pm on Mondays was used to compare different groups of streams. The most significant environmental drivers affecting instream temperatures were buffer width and density (p=0.022, df=10). The number of days that the daily maximum instream temperature exceeded state water quality standards, the average length of time that instream temperatures were above standard, and the range of temperature fluctuation were summarized. Streams would stay above 16oC from 8-22h on average and average diurnal temperature ranges were from 0.8-6oC.
Push-pull injections of nitrate into shallow subsurface wells were used to determine the rate of nitrogen loss in a riparian area dominated by red alder to an area without a buffer. The non-buffer site demonstrated similar nitrate loss rates in both July and December (average 0.068 mg·L-1·h-1). The buffered site had high dispersion rates and showed dramatic differences in July (dispersion rate 0.002 mg/L/hr, nitrate loss rate 0.055 mg·L-1·h-1) and December (dispersion rate 0.01mg/L/hr, nitrate loss rate 0 mg·L-1·h-1). Production of nitrous oxide in the presence of co-injected acetylene suggests that the observed nitrate loss was partially due to microbial denitrification.
Riparian buffers along northwest agricultural streams should be studied independently prior to prescribing buffer widths for water quality parameters because of the unique biochemical cycling of nitrogen and subsurface environment.
Read the full dissertation (644 KB pdf)