• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss in redOrbit Knowledge Network

Integrated Practices for Reducing Sediment Loss From Piedmont Tobacco Fields

Posted on: Friday, 2 May 2008, 06:00 CDT

By Hazel, D W Franklin, E C; Thomas, K T; Jennings, G D

Abstract: We evaluated the effectiveness of three best management practices for flue-cured tobacco production-reduced tillage, grassed field-side filter zones, and forested filter zones-to determine the total reduction in suspended solids from storm water runoff. Use of reduced tillage in comparison with conventional tillage decreased total suspended solids leaving tobacco fields by 82%. Grassed field- side filter zones functioned very well in retaining solids in early summer but were overloaded by late summer. Forested filter zones were able to back up the grassed filter zones when they overloaded and exported total suspended solids to the forested filter zones. In combination, grassed filter zones and forested filter zones retained 68% to 69% of total suspended solids, respectively. Dense vegetation in the cutover forested filter zone more than doubled its capacity to detain solids, compared to the same forested filter zone when it was covered by mature mixed pine-hardwood. Use of these best management practices in series can significantly reduce sediment loss from tobacco; however, use of reduced-till may reduce tobacco yield and quality.

Key words: best management practices-filter zones-nonpoint source pollution-runoff-sediments-tillage

Nonpoint source (NPS) pollution from agricultural runoff is a major concern regarding water quality throughout the state of North Carolina, the southeastern region, and the nation (Neary et al. 1989; North Carolina Department of Natural Resources and Community Development 2001; Smolen and Shanholtz 1980). Farmers are being encouraged to adopt best management practices (BMPs) to conserve soils and reduce NPS pollution. Many studies have evaluated individual BMPs such as conservation tillage, grassed filter zones (GFZs), or forested filter zones (FFZs) in reducing NPS pollution. However, on most agricultural watersheds, multiple BMPs can be used in series to provide maximum NPS pollution reduction and to maintain . soil productivity. As an example, conservation tillage and vegetated filter zones (VFZs) are frequently recommended, but little research has been done documenting effectiveness of a complete system consisting of reduced tillage (RT), GFZs, and FFZs.

If the use of RT and other BMPs can be demonstrated to reduce soil sediment and nutrient loss and to improve water quality, adoption of these BMPs may be accelerated, especially if crop yields and quality are not significantly compromised. Tobacco was considered to be a good crop to use for this study because heavy applications of fertilizers are used in predominantly conventionally tilled fields. If a system of BMPs can handle runoff from tobacco fields in the Piedmont, it can probably handle most other field crops.

Conservation tillage has been used to reduce soil erosion, and many studies have documented its effectiveness (Johnson et al. 1979; McDowell and McGregor 1980; Smith et al. 1995). Generally, conservation tillage includes those practices where more than 30% of the soil surface is covered with crop residue (Unger 1990). Water as surface runoff and subsurface drainage serves as a carrier for dissolved chemicals. Soil suspended in surface water also serves as a carrier of sediment-adsorbed chemicals. Use of conservation tillage affects chemical losses by trapping soil and by increasing infiltration and chemical exchange of dissolved chemicals (Baker and Laflen 1983). In many studies, crop residue on the soil surface has been shown to reduce surface runoff from all but the heaviest storms (Baker and Laflen 1983;Barisas et al. 1978; Hamlett et al. 1984; Johnson et al. 1979; Laflen and Tabatabai 1984; Mostaghimi et al. 1991; Schreiber and Cullum 1992) and eliminate runoff from most small storms (Baker and Laflen 1983). However, several other researchers have found that RT, a form of conservation tillage, may sometimes produce more runoff than conventional tillage (CT) (Gaynor and Findlay 1995; Lindstrom and Onstad 1984; Muelleretal. 1984).

Peedin and Smith (2001) have expressed the view that "growing flue-cured tobacco by the reduced-till method is the best option ... to meet conservation compliance requirements" thus avoiding loss of government support programs for the whole farm. They reviewed results from comparison tests of conservation tillage versus CT tobacco in the Piedmont and Coastal Plain from 1993 to 1999. In all cases, tobacco was successfully grown with conservation tillage, but in most cases yields were lower than for CT.

Researchers have demonstrated that FFZs significantly reduced sediments and nutrients in agricultural runoff (Franklin et al. 1992, 2000; Verchot et al. 1996, 1997a, 1997b, 1998; Peterjohn and Correll 1984;Lowrance et al. 1984a, 1984b;Jacobs and Gilliam 1983; Parsons et al. 1990; Phillips 1989; Doyle et al. 1975). However, results regarding GFZs are mixed (Daniels and Gilliam 1989; Dillaha et al. 1989); results suggest that GFZs may be effective for particulate and sediment-bound nitrogen including organic and adsorbed NH4-N, and may be relatively ineffective for dissolved N (NH4-N and NO3-N). In addition, grass -can be quickly submerged by runoff rendering these zones ineffective (Hayes and Hairston 1983). Others have felt that GFZs may offer specific characteristics that may complement other types of vegetation when used together in filter zones with mixed vegetation. Schultz et al. (1995) noted deficiencies in GFZs compared with FFZs and proposed a VFZ consisting of trees, shrubs, and grasses to take advantage of the differing above- and below-ground structures of different plant forms. Welsch (1991) proposed a three-zone VFZ system where channelized runoff would be distributed within a field-edge GFZ approximately 6 m (20 ft) in length (up-slope to downslope). Nutrient "trapping" would primarily be accomplished by a two-zoned FFZ. The up-slope FFZ would be approximately 18m (59 ft) in width and would have periodic timber harvesting to remove sequestered nutrients. The second and streamside FFZ would be about 5 m (16.4 ft) in width and would not be harvested. We believe that vegetation management to increase woody ground vegetation will improve the ability of filter zones to detain pollutants by physically slowing storm flow so that sediment and nutrients are dropped in the zone even if infiltration does not take place. This is especially important during major events when infiltration capacity of most zones will be inadequate for large volumes of storm flow.

Several recent studies have indicated that filter zones sometimes do not function because of channelized flow through the riparian areas (Dillaha et al. 1986a, 1986b, 1989; Nutter and Gaskin 1989). We conducted a preliminary study designed to test the feasibility of water quality improvement through dispersed runoff using level spreaders and documented reductions in sediment and nutrients (Franklin et al. 1992, 2000; Hazel 2000).

Objectives. Our main objective was to compare the effectiveness of three BMPs for tobacco production including RT versus CT, GFZs, and FFZs. Specific objectives were to compare RT to CT practices for tobacco in terms of reduction of runoff and yields of solids (as measured by total suspended solids) to field edges, crop yields, and leaf quality; document the effectiveness of GFZs and FFZs in terms of reduction of runoff and removal of solids along field edges; and compare performance of a FFZ with dense woody and herbaceous vegetation after a clear cut to its prior performance with a mature mixed pinehardwood forest.

Materials and Methods

Experimental Watersheds. Two 1.6 ha (4 ac) previously instrumented watersheds (Franklin et al. 1992, 2000) on the Oxford Tobacco Research Station in Oxford, North Carolina, included fields and forested areas and are typical of farm watersheds in die Piedmont throughout the southeast. Each watershed contained four fields from which surface runoff drained into a grassed waterway and then into a GFZ. Runoff flowed out of the GFZ of each watershed as channelized flow that was then dispersed in the receiving portion of the FFZ with level spreaders. The effective portion of the FFZs made up 5.0% and 5.3% of'watersheds II and I, respectively (table 1). Effective filter zone area was determined by mapping actual surface flow patterns during repeated irrigation events. Soils in watershed I were Helena loamy sand (Aquic Hapludult),Vance sandy loam (Typic Hapludult), and Durham sandy loam (Typic Hapludult). Soils on watershed II were Vance and Durham sandy loams. Topographic maps for each watershed were developed using land surveying.

Tobacco Tillage Comparison. The field portion of each watershed consisted of two fields, one designated for CT and the other for RT tobacco (figure 1). The four fields were roughly comparable in size, soils, and slope (table 1). Cover crops were established (table 2) during the fall of 1998, 1999, and 2000 on all four fields by disking, fumigating, shaping tobacco beds, and then sowing abruzzi rye (secale cereals L. [Poaceae, cv Wrens Abruzzi]). In spring, each field on which CT was assigned was plowed and cross-disked in late April and re-bedded just before transplanting. In fields where RT tobacco was to be grown, the rye cover crop was sprayed with herbicide (glyphosate) twice in late April or early May prior to transplanting (table 2). Cultivation practices followed recommendations of Collins and Hawks (1963), North Carolina Cooperative Extension Service (2001), and Peedin and Smith (2001) with regard to uses of pesticides and fertilizers for tobacco. Both CT and RT received the same treatments except tillage.Tobacco was planted for both tillage treatments the same day each year. In 1999, nursery-grown bare-root transplants were used for both treatments. In 2000, greenhouse-grown containerized transplants were used for both treatments. The tobacco variety used each year was K-346. Irrigation was used when needed. A tillage treatment within a watershed consisted of two series of half-rows on either side of the improved waterway (figure 1). Thus, fields were divided in half at the field ridge which bisected the field such that for each complete tobacco row, half the row was in one tillage treatment and the other half row in the other treatment with runoff moving laterally to the waterway. Rows were spaced 1.2 m (4 ft) apart between row centers, and each half row was 12.2 m (40 ft) in length with 20 tobacco plants per half row. Watershed I CT had 256 half rows, and RT had 248. Watershed II CT had 248 half rows, and RT had 258. RT and CT beds were approximately the same height and width. To compare tobacco yields and quality between tillage treatments, tobacco was bundled by half rows. A price index based on regional averages for each year was used to determine value of the crop rather than actual receipts. Each field had between 124 and 134 20-plant rows depending on field and year. For each year, there were three harvests or "primings" during the harvest season. For each priming, tobacco was bundled and tagged by replication (watershed), treatment, and row. Then, each bundle was later separated by grade and weighed by grade.

Grassed Filter Zones. Grassed filter zones with a mixture of fescue and coastal Bermuda grass were located between the field and the FFZs. In crop year 2000, both zones met the minimum standard of at least a 6 m (20 ft) flow length through the zone and grassed level spreaders to ensure sheet flow entering the zone (USDA Natural Resources Conservation Service CPS Code 393). However, runoff from each of the four flumes remained mostly channelized in a roughly 2 to 3 m width (6.5 to 9.8 ft) through the zone. To try to enhance GFZ performance, a low, grassed level spreader was built below each field flume in the fall of 1999 by constructing a low (15 to 20 cm [6 to 8 in]) berm on the contour at the raised waterway/GFZ junction. Spreaders were 12 to 18 m (40 to 60 ft) in length and allowed channelized flow to impound above the spreader and then flow as sheet flow over the level top of the spreader. A 2-in (5-cm) diameter PVC drain was installed under the spreader berm to allow all runoff to drain from above the spreader at the end of a runoff event. Level spreaders were well-grassed before initiation of sampling during the 2000 crop season (figure 1). Spreaders were constructed to permit trafficking over them by trucks, tractors, and other farm vehicles.

Forested Filter Zones. During this study, runoff for all events was dispersed on the contour across the forested area using level spreaders. Level spreaders consisted of two treated wooden troughs extending laterally from each side of a distribution box. The wooden spreaders were placed along the contour to allow water to flow over the whole FFZ. During our previous studies on these watersheds, both FFZs were fully stocked with mature trees (Franklin et al. 1992, 2000; Verchot et al. 1996, 1997a, 1997b, 1998). In September 1996, most trees within the watershed I FFZ were uprooted by Hurricane Fran. In the spring of 1997, a few standing trees and all downed trees within the FFZ were removed with a large crane positioned outside the FFZ. The crane was also used to right overturned stumps. Thus, during this study, the FFZ on watershed I was considered to be functioning as a recently clear cut FFZ. The overstory on the watershed II FFZ was essentially untouched by the same storm and remained a fully stocked, mature forest stand of mixed pine- hardwoods during this study.

Flumes, Samplers, and Data Recorders. Four 46-cm (18-in) H- flumes were installed in 1998 in the lower end of each of the four grassed waterways that drain the four tobacco fields (figure 1). Two 61-cm (24-in) H-flumes between the GFZs and FFZs and two 61-cm H- flumes at the lower end of each FFZ were re-activated for this study. Plywood wing walls extended out from each of the eight flumes to direct runoff into the flumes. Wing walls extended approximately 30 cm (12 in) below ground and were set into a concrete base. Stage was measured using float-driven potentiometers monitored by Campbell Scientific (Logan, Utah) CR-10 data loggers. One logger was equipped with a 12-volt modem for remote monitoring of general conditions via telephone. When a total of 50 mm (2 in) absolute change in stage occurred within a two-hour period for any flume, a 24-bottle automatic sampler was activated to take a sample at that flume. Rainfall was measured by three tipping-bucket rain gauges and was stored as five-minute totals.

Revised Rating Tables for Field Flumes. Flumes were fabricated open-channel flowmeasuring devices that consisted of three sections that converge, restrict, and expand flow in a deliberate fashion (Gwinn and Parsons 1976).The H-flume was designed by the USDA Natural Resources Conservation Service for estimation of low-volume flows from experimental watersheds. When published, empirically derived rating tables or models are to be used, rigid design and installation specifications and methods must be followed (Brakensiek et al. 1979). Any deviation from the design requires field calibration that is not generally performed due to the normally impractical requirement that large volumes of water be delivered in a wellmonitored mariner.

Calibration of Grassed and Forested Filter Zone Flumes. During previous studies on the Oxford site (Franklin et al. 1992), it was observed that mid- to high-flow rates appeared turbulent, possibly due to shorterthan-specified approach sections. All other specifications were measured and found to be within published tolerances (Brakensiek et al. 1979). This led us to question the wisdom of using standard discharge equations for determining flow rates and ultimately sediment and nutrient flow through the watersheds. The permanency of the site required calibration rather than alteration of the approach section to conform to specifications. Prior to calibration, wooden baffles were built outside the approach section of all flumes to lower velocities and provide more uniform flow throughout the cross-section of the approach sections.

For field calibration, two irrigation pumps were used to supply water at constant flow rates via two 12.7-cm (5-in) irrigation pipes. Flow rates were controlled using the pump engine throttles and valves at the pipe outflow just above the flume. Flume stage was measured using a data logger and potentiometers to monitor stilling- well floats. Medium and high flow (0.24 to 6.86 m^sup 3^ min^sup - 1^ [63.4 to 242.3 ft^sup 3^ min^sup -1^]) calibrations were determined by measuring stage height for a constant flow rate using two paddle-wheel flow sensors installed in the irrigation pipes. The flow sensors were calibrated volumetrically once installed in the irrigation system by repeatedly timing the filling of a 1.89 m^sup 3^ (500 gal) tank. Because flow sensors were erratic below a flow velocity of 0.30 m s^sup -1^ (1 ft s^sup -1^), low flow rates (0.0060 to 0.240 m^sup 3^ min^sup -1^ [0.21 to 8.48 ft^sup 3^ min^sup -1^]) calibrations were done by timing the filling of a 21.30-L (5.63-gal) bucket at constant flume stage. Regression equations were developed from the data acquired and used to estimate flow rates based on stage for the GFZ and FFZ (Franklin et al. 1992, 2000).

Observed flow rates varied from 20% to 35% greater than the standard rating curves. In the range of stage in which most runoff volumes occurred (15 to 45 cm [6.0 to 17.7 in]) the average difference was about 30%.

BMP Evaluation. To evaluate BMP effectiveness, runoff volumes and concentrations (nutrients and sediment) were measured at three locations: (1) field edge above the GFZ, (2) woodland edge above the FFZ, and (3) lower end of the FFZ (figure 1).

To evaluate BMPs, runoff volumes, flow rate, time to peak flow, and total suspended solids (TSS) concentrations were measured between treatment plots for tillage practices, over the two replications; above and below GFZs; above and below FFZs; and between current (clear cut) and prior (mature woodland) performance of the FFZ in watershed I.

Sample Collection and Analysis. Discrete 500 ml (16.9 02) samples were retrieved within 12 to 24 h after a rainfall event (shorter intervals during hot weather), placed on ice in coolers in the field, and then refrigerated in the laboratory until analyzed. For most events, average concentrations and total loadings for the event at each flume were computed by mixing a single 500 ml flowproportional aliquot for each flume. This was done by compositing the samples in the laboratory from hydrographs at each flume. For a subset of events, each discrete sample was analyzed for concentrations of solids as TSS.

Standardized analytical procedures were used to estimate yields of TSS Method 160.2 Gravimetric, dried at 103[degrees]C to 105[degrees]C (217[degrees]F to 221[degrees]F) for nonfilterable residues (Greenberg et al. 1992).

All results for TSS are expressed in kilograms for all watershed components (fields, GFZs, and FFZs). However, statistical analyses for comparisons of TSS and sediment leaving fields under different tillage treatments results are expressed on a kilograms per hectare basis because field area is different for each field. Runoff in each component is expressed as millimeters equivalent depth over the component. For example, field runoff is computed by measuring runoff at the H-flume below the field (m^sup 3^) and then expressing that volume as equivalent depth (mm) over the field area. Storm flow into the GFZ below the fields is computed by measuring field runoff with the field flume (m^sup 3^) and expressing that volume as equivalent depth over the area of the GFZ. Statistical Analyses. Comparison of sample means, using the student's t-test (Steel and Torrie 1980), were pooled over tobacco crop seasons and watersheds for total runoff, peak runoff, and TSS. Tobacco yields and quality (price per pound used as surrogate for quality) were analyzed by analysis of variance separately for 1999 and 2000 because the two crop years differed significantly in natural rainfall and irrigation schedules applied to all fields. Analyses of variance were computed using the General Linear Models procedure with SAS software (SAS Institute 1980). Because grade was not a continuous variable, the published price index for each grade was used to analyze quality. Each observation was the weighted average price index for each row within each treatment and watershed for each year. The price index was weighted by the total weight of each grade for that row within treatment and watershed.

Performance of Water Sampling System. This report summarizes results of two separately funded one-year projects. Although the two projects spanned two tobacco growing seasons, neither project covered 12 months of monitoring in any one year. There was additionally a brief funding gap between projects. The initial data collection period began December 15, 1998, and ended September 16, 1999. Three of the last four events prior to shutdown were tropical storms (Dennis twice and Floyd) as they passed through Granville County. No runoff events were collected for the 1999 to 2000 winter rye cover crop. The collection system was reactivated April 1, 2000, with new funding and then shut down on December 19, 2000. The resulting periods for which runoff data were obtained included the periods December 16, 1998, to September 15, 1999, and May 27, 2000, to December 16, 2000, for a total of 40 rainfall-producing events. Many of these events did not generate enough rainfall to submerge water sampler intake ports. Instrument failures further reduced the number of events usable for analyses to 30 (table 2). Causes of failures included transient voltage due to electrical storms and animal damage. These data included 8 events during the 1999 winter cover crop, 4 during the 1999 tobacco cultivation season, 17 during the 2000 tobacco cultivation season, and a single event or! December 16, 2000, while fescue was being established as the winter cover crop.

Results and Discussion

Runoff. Average runoff rates for CT were 50% in 1999 and 44% in 2000. For RT they were 54% in 1999 and 55% in 2000. The higher observed rate of runoff for RT was statistically significant (p > I = 0.0400) and was consistent with observations in other studies (Gaynor and Findlay 1995; Lindstrom and Onstad 1984; Mueller et al. 1984). Runoff during the 1998 to 1999 cover crop season was relatively low (38%) but was very high (82%) during the single large event in December 2000 (table 3).

Runoff rates from GFZs were generally very high, averaging 85% and ranging from 76% during the 1998 to 1999 cover crop season to 97% during the 1999 tobacco crop season (table 3). This is consistent with other observations in this study indicating that GFZs did little to slow or detain runoff. However, it must be recognized that these GFZs are small relative to fields (functionally, they usually are for turning equipment), and they not only receive precipitation directly on them, but they are the receiving zones and potentially the filtration zones for field runoff. Thus, equivalent depths for these were much larger than for fields (table 3).

Runoff from FFZs averaged 42%, ranging from a low of 34% during the 2000 tobacco crop season to a high of 54% during the 1999 tobacco crop season. The runoff rate during the large single event in December 2000 was surprisingly low (36%), probably due to the fact that very little moisture was in the soil just prior to that event (table 3) and infiltration capacity of forest soils is relatively high unless compacted during harvesting.

Peak Flow Rates. No clear differences in peak flow (m^sup 3^ h^sup -1^) emerged between tillage treatments for the four largest rainfall events during each crop year. During the single cover crop event in December 2000 when all four fields were in poorly established cover, peak flow rates from the fields averaged 581 m^sup 3^ h^sup -1^ (20,518 ft^sup 3^ hr^sup -1^), over five times the highest rate previously observed during the two-year study (table 4).

Peak flow rates in GFZs were generally equal to the sum of rates for respective contributing fields, again indicating little or no retention of funoff by the GFZs (table 4).

Peak flow rates in FFZs ranged from three fourths to less than half of the flow rates in GFZs, indicating a substantial reduction in energy of flow in the FFZs (table 4).

Total Suspended Solids. Total suspended solids transported from fields in surface runoff during the tobacco crop season was reduced by 84% in 1999 and 81% in 2000 by the use of RT (table 5). These results were statistically significant (p > |t| = 0.0062). Other researchers reported similar reductions in sediment with various forms of conservation tillage (Angle et al. 1984; Johnson et al. 1979; McDowell and McGregor 1980; Mostaghimi et al. 1991; Smith et al. 1995).

GFZs detained 36% of TSS during the cover crop season of 1998 to 1999, but exported -3% in crop year 1999, retained 0% in crop year 2000, and had a net export rate of 59% during the one large event of the cover crop season of 2000. The mean export rate over the entire study period was -9% (table 5).

FFZs detained 50% of TSS during the cover crop season of 1998 to 1999, 70% and 60% respectively during the tobacco crop seasons, and 44% during the one major event in the cover crop season of 2000. Although this "detention" rate (44%) was somewhat lower than the average for the other crop seasons (60%), it was still impressive when one considered that the December 16, 2000, event delivered substantially more sediment than the total from all other events monitored over the entire two-year study period. The mean retention rate over the entire study period was 56% (table 5).

The 82% of TSS reduction resulting from reduced-till compared to conventional tillage, corroborated by many other studies, points conclusively to varying forms of conservation tillage as the most effective solution to tillage sediment discharge problems in agriculture. Issues related to potentially lower yields must be addressed.

Sediment was by far the largest pollutant measured in this study and therefore deserved closer study. So we bar-graphed the retention of TSS by GFZ and FFZ for each event monitored to look for seasonal patterns in retention. The general pattern observed was that through the spring and summer the grassed zones tended to load up with sediment. Then when a large rainfall event occurred, such as on August 25, 1999 (Hurricane Dennis I), the GFZs showed a net export of sediment to the FFZs. Having been scoured of sediment on JD 237, the GFZs showed net retention of TSS during Hurricanes Dennis II and Floyd (figure 2). The spring and summer of 2000 showed similar patterns. The GFZs detained TSS during the spring, showed net exports during large events on August 3 and 4, 2000, detained TSS during the next two events and eventually released TSS again on September 25, 2000 (figure 3). Even though there was very little rain during the fall of 2000, when the big event of December 16, 2000, arrived, the GFZs again showed net releases of substantial amounts of TSS (figure 4).

Vegetation Management in a Forested Filter Zones. During earlier studies of the same watersheds and FFZs (1989 to 1994), watershed I FFZ (figure 1) performed rather poorly in removing nutrients, removing as little as 15% of NO^sub 3^ up to about 48% of ammoniacal nitrogen (Franklin et al. 2000). With vegetation resulting from clear cutting of the overstory in this FFZ as described above, effectiveness of the zone improved somewhat in detention of TSS. Mean detention of TSS for 47 events from October 1989 to April 1995 was 46% with a mature forest canopy on the filter zone. In 30 events following overstory removal, mean TSS detention improved to 51%. Improved performance, though modest, probably resulted from decreased velocity of runoff and increased retention time due to high structural resistance of the dense, woody ground vegetation. This effect would also be expected to provide for higher infiltration and more contact with adsorption surfaces. Results of this research indicate that vegetation management in FFZs can improve the ability of these zones to detain nutrients. A heavy thinning of trees would probably be as effective as cutting all the trees as long as enough light reached the forest floor to stimulate plant growth. Low-intensity, controlled fires could be used to maintain dense, low vegetation. Mowing would not be preferred because of the resulting compaction of soil. The transition area between GFZs and FFZs can be improved by sowing crops such as lespedeza or other crops with closely spaced, stiff, woody stems.

Tobacco Yield and Quality. Treatment as an independent ANOVA variable was significant for both years for weight (p > F < 0.0001). Yield for RT tobacco was about 32% less than that for CT tobacco in 1999 and about 19% less in 2000. Even so, weight for both RT and CT exceeded the regional average in 2000 by almost 200 kg ha^sup -1^ (178 lb ac^sup -1^) for RT and by over 800 kg ha^sup -1^ (714 lb ac^sup -1^) for CT (table 6). The 1999 crop was especially poor at least in part because the bed-grown transplants were smaller than greenhouse-grown transplants for which the modified no-till transplanter used was designed. Thus, water from the transplanter was distributed below the root zone resulting in poor initial survival. After spot replanting, fields were intensively irrigated to improve survival. However, irrigation may have excessively leached nitrogen from the soil. The combination of late replanting and possible N leaching probably accounts for the low yields for both RT and CT treatments in 1999. However, in both years, it may be assumed that these factors impacted both treatments equally, thus, yields from RT tobacco as done in this study may be expected to be lower than those for CT. Tobacco quality, based on price/quality index published for tobacco research (North Carolina State University 2000), was lower in both years for RT compared to CT. Similarly, the average grade for CT and RT was lower than the regional average grade (table 6). Treatment as an independent ANOVA variable was significant for both years for quality expressed as weighted price index (p > F < 0.0001).

Designing Systems for Large Events. During previous research on the FFZs, it was observed that 80% or more of TSS and nutrients were delivered to filter zones in only three to eight events per year (Franklin et al. 2000). These observations indicated that filter zones must be designed to handle the largest events. If not, most pollutants will reach receiving waters even though filter zones are in place. In the previous research, it was also observed that the proportion of pollutants detained during large events by the FFZs was as great or only slightly less than that detained for all events combined. Mass quantities of pollutants detained were almost always larger for the large events, and that was accomplished without excessive damage to the filtration capacity of the zone (Franklin et al. 2000).

For events in this study, detention of TSS was calculated for each GFZ and FFZ based on the percent of the total gravimetric loading detained (total input minus total output as a percentage of input). Then, percentage detained was calculated for only the largest events that together contributed about 80% of the total of all events for the analyte (referred to as "80% events"). In all cases, if the GFZ had a net export to the FFZ, then the amount exported to the FFZ was used as the denominator (base) for the percentage reduction for the combined detention percentage (figure 5), to properly represent the detention percentage in both zones combined. FFZs never had a net export situation, even during the largest events.

For the 30 runoff events monitored in this study during major portions of 1999 and 2000, five events contributed 80% or more of the total gravimetric loading of TSS (figure 5). The December 16, 2000, event delivered more sediment than the entire sum for all other events monitored. Four additional, much smaller events were required to cover 80% of TSS over the two-year period.

Frequently used design criteria for VFZs include bypass channels to divert the heaviest runoff around the zone, ostensibly to protect the integrity of the zone during large runoff events. However, these data and those of earlier studies demonstrate that these experimental FFZs handle large events very well. Thus, it is illogical to design FFZs so that runoff from major events by-passes the filter zone. These large events are the ones most in need of treatment. If the system fails to effectively handle the large events (including hurricanes and other major events), the system has failed to effectively reduce the amount of sediment and nutrients that reach receiving waters.

Recommendations. Tobacco and other tilled crops should be grown with an effective form of conservation tillage to effectively minimize delivery of sediment to receiving waters. Reductions of TSS of 70% can be made and would go a long way toward restoring sediment- impacted waters. Indications were that yields from reduced-till are lower than from conventional tillage, thus cost-share payments may be appropriate to supplement the practice. On the other hand, retention of so much soil in farm fields, which under conventional tillage was routinely lost, should eventually help to increase yields and/or lower costs of crop production.

Grassed filter zones, adjacent to farmed fields, should be used for the first phase of reduction of sediments and nutrients in agricultural runoff. Grassed zones are typically low in infiltration capacity because of the tightness of the sod and compaction caused by equipment. We recommend use of level spreaders in grassed zones when they would effectively improve the utilization of the available area. We recommend establishing woody-stemmed cover crops such as lespedeza in the lower portion of the grassed zones, next to the woodland filter zone. We do not recommend that GFZs be relied upon as the only filter zone for a farm field because they are typically inadequate to handle the large events which are the ones delivering most of the sediment and nutrients to receiving waters.

We recommend the use of FFZ where they are adjacent to agricultural fields, and in some cases the planting of forested, brushy and GFZs. If topography or hydrologic engineering concentrate runoff in the zone, level spreaders should be installed in the upper end of the zone. Vegetation in all zones should be managed to encourage growth of dense, woody ground vegetation, except in the upper portions of the grassed zone. If available forested zones are too small in area or otherwise only partially functional, removing the overstory would be expected to greatly improve performance of the filter zone. This practice may also address shading and/or root competition problems for farm field edges. If any portion of a field- side filter zone is riparian, restrictions may apply to manipulation of vegetation and/or other activities within all or portions of the zone.

Acknowledgements

The research on which this report is based was financed in pan by the United States Department of the Interior, Geological Survey, through the North Carolina Water Resources Research Institute. Contents of the publication do not necessarily reflect the views and policies of the US Department of the interior, nor does mention of trade names or commercial products constitute their endorsement by the US government.

References

Angle, J.S., G. McClung, M.S. McIntosh, P.M. Thomas, and D.C. Wolf. 1984. Nutrient losses in runoff from conventional and no-till watersheds. Journal of Environmental Quality 13(3):431-438.

Baker, J.L., and J-M. Laflen. 1983. Water quality consequences of conservation tillage. Journal of Soil and Water Conservation 38(3): 186-193.

Barisas, S.G., J.L. Baker, H.P. Johnson, and J.M. Laflen. 1978. Effect of tillage systems on runoff losses of nutrients, a rainfall simulation study. Transactions of the American Society of Agricultural Engineers 21(5):893-897.

Brakensiek, D.L., H.B. Osborn, and W.J. Rawls. 1979. Field Manual for Research in Agricultural Hydrology, 75-214. USDA Handbook No. 224. Washington, DC: US Government Printing Office.

Collins, W.K., and S.N. Hawks Jr. 1963. Principles of Flue-Cured Tobacco Production, ed. W.K. Collins and S.N. Hawks Jr., 97-139. Raleigh, NC.

Daniels, R.B., and J.W Gilliam. 1989. The effect of grass and riparian filters on runoff quality. In Proceedings of the 32nd Meeting of the Soil Science Society of North Carolina, ed. H. Byrd and S. Buol, 37-52.

Dillaha, T.A., R.B. Reneau, S. Mostaghimi, and D. Lee. 1989. Vegetative filter strips for agricultural nonpoint source pollution control. Transactions of the American Society of Agricultural Engineers 32(2):513-519.

Dillaha, T.A., J.H. Sherrard; and D. Lee. 1986a. Long-Term Effectiveness and Maintenance of Vegetative Filter Strips. Virginia Water Resources Research Center Bulletin 153. Blacksburg, VA: Virginia Tech.

Dillaha, T.A., J.H. Sherrard, D. Lee, V.O. Shanholtz, S. Mostaghimi, and W.L. Magette. 1986b. Use of Vegetative Filter Strips to Minimize Sediment and Phosphorus Losses from Feedlots; Phase I. Experimental Plot Studies. Virginia Water Resources Research Center Bulletin 151. Blacksburg, VA: Virginia Tech.

Doyle, R.C., D.C. Wolf, and DE Bezdicek. 1975. Effectiveness of forest buffer strips in improving the water quality of manure polluted runoff. In Managing Livestock Wastes, Proceedings. 3rd International Symposium on Livestock Wastes, University of Illinois.

Franklin, E.C., J.D. Gregory, and M.D. Smolen. 1992. Enhancement of the Effectiveness of Forested Filter Strips by Dispersion of Agricultural Runoff. Water Resources Research Institute Report Number 270. Raleigh, NC: University of North Carolina.

Franklin, E.C., J.D. Gregory, J.E. Parsons, and D.W. Hazel. 2000. Management of Forested Filter Zones for Dispersion and Treatment of Agricultural Runoff. Water Resources Research Institute Report Number 312. Raleigh, NC: University of North Carolina.

Gaynor, J.D., and W.I. Findlay. 1995. Soil and phosphorus loss from conservation and conventional tillage in corn production. Journal of Environmental Quality 24(4):734-741.

Greenberg, A.E., L.S. Clesceri, and A.D. Denton, eds. 1992. Standard Methods for the Examination of Water and Wastewater, 18th ed., Washington, DC: American Public Health Association.

Gwinn, W.R., and D.A. Parsons. 1976. Discharge equations for HS, H, and HL flumes. Proceedings of the American Society of Civil Engineers Journal of the Hydraulics Division, 102(HY1):73-88.

Hamlett, J.M., J.L. Baker, S.C. Kimes, and H.P. Johnson. 1984. Runoff and sediment transport within and from small agricultural watersheds. Transactions of the American Society of Agricultural Engineers 27 (5): 1355-63, 1369. Hayes, J.C., and J.E. Hairston. 1983. Modeling the long-term effectiveness of vegetative filters as on-site sediment controls. St. Joseph, MI: American Society of Agricultural Engineers, Paper 83-2081.

Hazel, D.W. 2000. Effectiveness and Cost of Improving Vegetated Filter Zones by Installing Level Spreaders to Disperse Agricultural Runoff. PhD dissertation. North Carolina State University, Raleigh, NC.

Jacobs, T.C., and J.W. Gilliam. 1983. Nitrate Loss from Agricultural Drainage Waters: Implications for Nonpoint Source Control. Water Resources Research Institute Report Number 209. Raleigh, NC; University of North Carolina.

Johnson, H.P., J.L. Baker, WD Shrader, and J.M. Laflen. 1979. Tillage system effects on sediment and nutrients in runoff from small watersheds. Transactions of the American Society of Agricultural Engineers 22(1979):1110-1114.

Laflen, J.M., and M.A.Tabatabai. 1984. Nitrogen and phosphorus losses from corn-soybean rotations as affected by tillage practices. Transactions of the American Society of Agricultural Engineers 27(1):58-63.

Lindstrom, M.J., and C.A. Onstad. 1984. Influence of tillage systems on soil physical parameters and infiltration after planting. Journal of Soil and Water Conservation 39(2): 149-152.

Lowrance, R.R., R.L. Todd, and L.E. Asmussen. 1984a. Nutrient cycling in an agricultural watershed: I. Phreatic Movement. Journal of Environmental Quality 13(1):22-27.

Lowrance, R.R., R.L., Todd, and L.E. Asmussen. 1984b. Nutrient cycling in an agricultural watershed: II. Streamflow and artificial drainage. Journal of Environmental Quality 13(1):27-32.

McDowell, L.L., and K.C. McGregor. 1980. Nitrogen and phosphorus losses in runoff from no-till soybeans. Transactions of the American Society of Agricultural Engineers 23:643-648.

Mostaghimi, S.,T.M., Younos, and U.S. Tim. 1991. The impact of fertilizer application techniques on nitrogen yield from two tillage systems. Agriculture, Ecosystems, and Environment 36(1-2): 13-22.

Mueller, D.H., R.C. Wendt, and T.C. Daniel. 1984. Soil and water losses as affected by tillage and manure application. Soil Science Society of America Journal 48:896-900.

Neary, D.G., W.T. Swank, and H. Riekerk. 1989. An Overview of Nonpoint Source Pollution in the Southern United States. USDA Forest Service Southeastern Forest Experiment Station, General Technical Report. Asheville, NC: USDA Forest Service.

North Carolina Department of Agriculture and Consumer Services. 2000. Field Crops, Annual Summary: Crop Estimates, NC, 1998-2000. Raleigh, NC: Agricultural Statistics Division, North Carolina Department of Agriculture and Consumer Services.

North Carolina Cooperative Extension Service. 2001. Flue-cured Tobacco Information-2001. Raleigh, NC: North Carolina Cooperative Extension Service.

North Carolina Department of Natural Resources and Community Development. 1989. North Carolina Nonpoint Management Program. Raleigh, NC; North Carolina Department of Natural Resources and Community Development.

North Carolina State University. 2000. 1999 and 2000 Flue-Cured Tobacco Grade Prices. Raleigh, NC: Official Variety Testing Program, Crop Science Department, North Carolina State University.

Nutter, W.F., and J.W. Gaskin. 1989. Role of Streamside Management Zones in Controlling Discharges to Wetlands. USDA Forest Service Southeastern Forest Experiment Station General Technical Report. Asheville, NC: USDA Forest Service.

Parsons, J.E., R.D. Daniels, J.W. Gilliam, and T.A. Dillaha. 1990. Water Quality Impacts of Vegetative Filter Strips and Riparian Areas. American Society of Agricultural Engineers Paper 90-2501.

Peedin, G.F., and W.D. Smith. 2001. No-till flue-cured tobacco production. In 2001 North Carolina Flue-Cured Tobacco Production Guide. Raleigh, NC: North Carolina State University.

Peterjohn, W.T, and D.L. Cortell. 1984. Nutrient dynamics in an agricultural watershed: Observations on the role of a riparian forest. Journal of Ecology 65(5):1466-1475.

Phillips, J.D. 1989. Nonpoint source pollution control effectiveness of riparian forests along a Coastal Plain River. Journal of Hydrology 110:221-237.

SAS Institute. 1980. SAS User's Guide, 1982 edition. Cary, NC: SAS Institute Inc.

Schreiber, J.D., and R.F. Cullum. 1992. Nutrient Transport in Loessial Uplands of Mississippi. American Society of Agricultural Engineers Paper No. 92-2612. St. Joseph, MI: American Society of Agricultural Engineers.

Schultz, R.C., J.P. Colletti, T.M. Isenhart, W.W. Simpkins, C.W. Mize, and M.L. Thompson. 1995. Design and placement of a multi- species riparian buffer strip system. Agroforestry Systems 29(3):221- 226.

Smith Jr., S., J.D. Schreiber, and R.R. Cullum. 1995. Upland soybean production: Surface and shallow groundwater quality as affected by tillage and herbicide use. Transactions of the American Society of Agricultural Engineers 38(4):1061-1068.

Smolen, M.D., and V.O. Shanholtz. 1980. Agricultural Land Uses: Effects on the Chemical Quality of Runoff. Virginia Water Resources Research Center Bulletin 125. Blacksburg, VA: Virginia Tech.

Steel, R.C.D., and Torrie, J.H. 1980. Principles and Procedures of Statistics, 2nd ed. New York: McGraw-Hill.

Unger, P.W. 1990. Conservation tillage systems. Advances in Soil Science 13:27-28.

Verchot, L. V., E.C. Franklin, and J.W. Gilliam. 1996. Nitrogen Biogeochemical Cycling in Piedmont Vegetated Filter Zones. Water Resources Research Institute Report Number 310. Raleigh, NC: University of North Carolina.

Verchot, L.V., B.C. Franklin, and J.W Gilliam. 1997a. Nitrogen cycling in Piedmont vegetated filter zones. Surface soil processes. Journal of Environmental Quality 26:327-336.

Verchot, L.V., B.C. Franklin, and J.W Gilliam. 1997b. Nitrogen cycling in Piedmont vegetated filter zones. Subsurface nitrate removal. Journal of Environmental Quality 26:337-347.

Verchot, L.V., B.C. Franklin, and J.W. Gilliam. 1998. Effects of agricultural runoff dispersion on nitrate reduction in forested filter zones. Soil Science Society of America Journal 62:1719-1724.

Welsch, D.J. 1991. Riparian forest buffers: Function and design for protection and enhancement of water resources. USDA Forest Service Northeastern Area State and Private Forestry Paper NA-PR-07- 91. Newtown Square, PA: USDA Forest Service.

Dennis W. Hazel is an assistant professor and E. Carlyle Franklin is a professor in the Department of Forestry and Environmental Resources at North Carolina State University, Raleigh, North Carolina. Kathleen T. Thomas is an accounting associate with the SAS Institute in Gary, North Carolina. Gregory D. jennlngs Is a professor in the Department of Biological and Agricultural Engineering at North Carolina State University, Raleigh, North Carolina.

Copyright Soil and Water Conservation Society May/Jun 2008

(c) 2008 Journal of Soil and Water Conservation. Provided by ProQuest Information and Learning. All rights Reserved.

Source: Journal of Soil and Water Conservation

More News in this Category