Performance of Saturated Riparian Buffers in Iowa, USA.

Nitrate from artificial drainage pipes (tiles) underlying agricultural fields is a major source of reactive N, especially NO, in surface waters. A novel approach for reducing NO loss is to intercept a field tile where it crosses a riparian buffer and divert a fraction of the flow as shallow groundwater within the buffer. This practice is called a saturated riparian buffer (SRB), and although it is promising, little data on the performance of the practice is available. This research investigated the effectiveness of SRBs in removing NO at six sites installed across Iowa, resulting in a total of 17 site-years. Water flow and NO in the tile outlets, diverted into the buffers, and NO concentration changes within the buffers were monitored throughout the year at each site. Results showed that all the SRBs were effective in removing NO from the tile outlet, with the average annual NO load removal ranging from 13 to 179 kg N for drainage areas ranging from 3.4 to 40.5 ha. This is NO that would have otherwise discharged directly into the adjoining streams. The annual removal effectiveness, which is the total NO removed in the SRB divided by the total NO draining from the field, ranged from 8 to 84%. This corresponds to an average removal rate of 0.040 g N m d with a range of 0.004 to 0.164 g N m d. Assuming a 40-yr life expectancy for the structure and a 4% discount rate, we computed a mean equal annual cost for SRBs of US$213.83. Given the average annual removal of 73 kg for all site-years, this cost equates to $2.94 kg N removed, which is very competitive with other field-edge practices such as denitrification bioreactors and constructed wetlands. Thus, SRBs continue to be a promising practice for NO removal in tile-drained landscapes.

Impact of hydraulic residence time on nitrate removal in pilot-scale woodchip bioreactors.

Nitrate (NO3-N) export from row crop agricultural systems with subsurface tile drainage continues to be a major water quality concern. Woodchip bioreactors are an effective edge-of-field practice designed to remove NO3-N from tile drainage. The NO3-N removal rate of woodchip bioreactors can be impacted by several factors, including hydraulic residence time (HRT). This study examined the impact of three HRTs, 2 h, 8 h, and 16 h, on NO3-N removal in a set of nine pilot-scale woodchip bioreactors in Central Iowa. NO3-N concentration reduction from the inlet to the outlet was significantly different for all HRTs (p < 0.05). The 16 h HRT removed the most NO3-N by concentration (7.5 mg L-1) and had the highest removal efficiency at 53.8%. The 8 h HRT removed an average of 5.5 mg L-1 NO3-N with a removal efficiency of 32.1%. The 2 h HRT removed an average of 1.3 mg L-1 NO3-N with a removal efficiency of 9.0%. The 2 h HRT had the highest NO3-N mass removal rate (MRR) at 9.0 g m-3 day-1, followed by the 8 h HRT at 8.5 g m-3 day-1, and the 16 h HRT at 7.4 g m-3 day-1, all of which were statistically different (p < 0.05). Significant explanatory variables for removal efficiency were HRT (p < 0.001) and influent NO3-N concentration (p < 0.001), (R2 = 0.80), with HRT accounting for 93% contribution. When paired with results from a companion study, the ideal HRT for the bioreactors was 8 h to achieve maximum NO3-N removal while reducing the impact from greenhouse gas emissions.

Agricultural conservation planning framework: 1. Developing multipractice watershed planning scenarios and assessing nutrient reduction potential.

Spatial data on soils, land use, and topography, combined with knowledge of conservation effectiveness, can be used to identify alternatives to reduce nutrient discharge from small (hydrologic unit code [HUC]12) watersheds. Databases comprising soil attributes, agricultural land use, and light detection and ranging-derived elevation models were developed for two glaciated midwestern HUC12 watersheds: Iowa's Beaver Creek watershed has an older dissected landscape, and Lime Creek in Illinois is young and less dissected. Subsurface drainage is common in both watersheds. We identified locations for conservation practices, including in-field practices (grassed waterways), edge-of-field practices (nutrient-removal wetlands, saturated buffers), and drainage-water management, by applying terrain analyses, geographic criteria, and cross-classifications to field- and watershed-scale geographic data. Cover crops were randomly distributed to fields without geographic prioritization. A set of alternative planning scenarios was developed to represent a variety of extents of implementation among these practices. The scenarios were assessed for nutrient reduction potential using a spreadsheet approach to calculate the average nutrient-removal efficiency required among the practices included in each scenario to achieve a 40% NO-N reduction. Results were evaluated in the context of the Iowa Nutrient Reduction Strategy, which reviewed nutrient-removal efficiencies of practices and established the 40% NO-N reduction as Iowa's target for Gulf of Mexico hypoxia mitigation by agriculture. In both test watersheds, planning scenarios that could potentially achieve the targeted NO-N reduction but remove <5% of cropland from production were identified. Cover crops and nutrient removal wetlands were common to these scenarios. This approach provides an interim technology to assist local watershed planning and could provide planning scenarios to evaluate using watershed simulation models. A set of ArcGIS tools is being released to enable transfer of this mapping technology.

Pasture size effects on the ability of off-stream water or restricted stream access to alter the spatial/temporal distribution of grazing beef cows.

For 2 grazing seasons, effects of pasture size, stream access, and off-stream water on cow distribution relative to a stream were evaluated in six 12.1-ha cool-season grass pastures. Two pasture sizes (small [4.0 ha] and large [12.1 ha]) with 3 management treatments (unrestricted stream access without off-stream water [U], unrestricted stream access with off-stream water [UW], and stream access restricted to a stabilized stream crossing [R]) were alternated between pasture sizes every 2 wk for 5 consecutive 4-wk intervals in each grazing season. Small and large pastures were stocked with 5 and 15 August-calving cows from mid May through mid October. At 10-min intervals, cow location was determined with Global Positioning System collars fitted on 2 to 3 cows in each pasture and identified when observed in the stream (0-10 m from the stream) or riparian (0-33 m from the stream) zones and ambient temperature was recorded with on-site weather stations. Over all intervals, cows were observed more (P ≤ 0.01) frequently in the stream and riparian zones of small than large pastures regardless of management treatment. Cows in R pastures had 24 and 8% less (P < 0.01) observations in the stream and riparian zones than U or UW pastures regardless of pasture size. Off-stream water had little effect on the presence of cows in or near pasture streams regardless of pasture size. In 2011, the probability of cow presence in the stream and riparian zones increased at greater (P < 0.04) rates as ambient temperature increased in U and UW pastures than in 2010. As ambient temperature increased, the probability of cow presence in the stream and riparian zones increased at greater (P < 0.01) rates in small than large pastures. Across pasture sizes, the probability of cow presence in the stream and riparian zone increased less (P < 0.01) with increasing ambient temperatures in R than U and UW pastures. Rates of increase in the probability of cow presence in shade (within 10 m of tree drip lines) in the total pasture with increasing temperatures did not differ between treatments. However, probability of cow presence in riparian shade increased at greater (P < 0.01) rates in small than large pastures. Pasture size was a major factor affecting congregation of cows in or near pasture streams with unrestricted access.

Reconnecting tile drainage to riparian buffer hydrology for enhanced nitrate removal.

Riparian buffers are a proven practice for removing NO from overland flow and shallow groundwater. However, in landscapes with artificial subsurface (tile) drainage, most of the subsurface flow leaving fields is passed through the buffers in drainage pipes, leaving little opportunity for NO removal. We investigated the feasibility of re-routing a fraction of field tile drainage as subsurface flow through a riparian buffer for increasing NO removal. We intercepted an existing field tile outlet draining a 10.1-ha area of a row-cropped field in central Iowa and re-routed a fraction of the discharge as subsurface flow along 335 m of an existing riparian buffer. Tile drainage from the field was infiltrated through a perforated pipe installed 75 cm below the surface by maintaining a constant head in the pipe at a control box installed in-line with the existing field outlet. During 2 yr, >18,000 m (55%) of the total flow from the tile outlet was redirected as infiltration within the riparian buffer. The redirected water seeped through the 60-m-wide buffer, raising the water table approximately 35 cm. The redirected tile flow contained 228 kg of NO. On the basis of the strong decrease in NO concentrations within the shallow groundwater across the buffer, we hypothesize that the NO did not enter the stream but was removed within the buffer by plant uptake, microbial immobilization, or denitrification. Redirecting tile drainage as subsurface flow through a riparian buffer increased its NO removal benefit and is a promising management practice to improve surface water quality within tile-drained landscapes.

Source-pathway separation of multiple contaminants during a rainfall-runoff event in an artificially drained agricultural watershed.

A watershed's water quality is influenced by contaminant-transport pathways unique to each landscape. Accurate information on contaminant-pathways could provide a basis for mitigation through well-targeted approaches. This study determined dynamics of nitrate-N, total P, Escherichia coli, and sediment during a runoff event in Tipton Creek, Iowa. The watershed, under crop and livestock production, has extensive tile drainage discharging through an alluvial valley. A September 2006 storm yielded 5.9 mm of discharge during the ensuing 7 d, which was monitored at the outlet (19,850 ha), two tile-drainage outfalls (total 1856 ha), and a runoff flume (11 ha) within the sloped valley. Hydrograph separations indicated 13% of tile discharge was from surface intakes. Tile and outlet nitrate-N loads were similar, verifying subsurface tiles dominate nitrate delivery. On a unit-area basis, tile total P and E. coli loads, respectively, were about half and 30% of the outlet's; their rapid, synchronous timing showed surface intakes are an important pathway for both contaminants. Flume results indicated field runoff was a significant source of total P and E. coli loads, but not the dominant one. At the outlet, sediment, P, and E. coli were reasonably synchronous. Radionuclide activities of (7)Be and (210)Pb in suspended sediments showed sheet-and-rill erosion sourced only 22% of sediment contributions; therefore, channel sources dominated and were an important source of P and E. coli. The contaminants followed unique pathways, necessitating separate mitigation strategies. To comprehensively address water quality, erosion-control and nitrogen-management practices currently encouraged could be complemented by buffering surface intakes and stabilizing stream banks.

Hydrogeological constraints on riparian buffers for reduction of diffuse pollution: examples from the Bear Creek watershed in Iowa, USA.

Riparian Management Systems (RiMS) have been proposed to minimize the impacts of agricultural production and improve water quality in Iowa in the Midwestern USA. As part of RiMS, multispecies riparian buffers have been shown to decrease nutrient, pesticide, and sediment concentrations in runoff from adjacent crop fields. However, their effect on nutrients and pesticides moving in groundwater beneath buffers has been discussed only in limited and idealized hydrogeologic settings. Studies in the Bear Creek watershed of central Iowa show the variability inherent in hydrogeologic systems at the watershed scale, some of which may be favorable or unfavorable to future implementation of buffers. Buffers may be optimized by choosing hydrogeologic systems where a shallow groundwater flow system channels water directly through the riparian buffer at velocities that allow for processes such as denitrification to occur.