Hydrology and Groundwater Nutrient Concentrations in a Ditch-Drained

Water Quality Protection: Symposium. College Park, MD: University of Maryland. \vw\:v.sawgal.umd. edu/DrainageDitches/2006_tourandsymposium/. Skaggs, R.W, M.A. Breve, . and J.W Gilliam. 1994. Hydrologic and water- quality impacts ofagricultural drainage. Critical Reviews in Enviromnental Science and Technology 24:1-32. Skaggs, R.W, and J.V Schilfgaarde, eds. 1999. Agricultural Hydrology and groundwater nutrient drainage. Agronomy Monograph 38. Madison, WI: American Society of Agronomy-Crop Science Society ofAmerica-Soil Science Society ofAmerica. concentrations in a ditch-drained Skaggs, R.W, G.M. Chesheir, and BD. Phillips. 2005a. Methods to determine lateral effect of a drainage agroecosystem ditch on wetland hydrology. Transactions of the ASAE 48:577-584 Skaggs, R.W, A.A. Youssef, G.M. Chescheir, and J.W P.A. Vadas, M.S. Srinivasan, P.j.A. Kleinman, J.P. Schmidt, and A.L. Allen Gilliam. 2005b. Effect of drainage intensity on nitro­ gen losses from drained lands. Transactions of the ASAE 48:2169-2177. Abstract: Groundwater nitrogen (N) and phosphorus (P) transport from ditch-drained, culti­ Smith, DR., and E.A. Pappas. 2007 Effect of ditch dredg­ vated fields has not been adequately investigated in the Chesapeake Bay watershed.We moni­ ing on the fate of nutrients in deep drainage ditches of tored hydrology and groundwater Nand P concentrations in 26 shallow (~3 m [10 ft]) wells the Midwestern United States.Journal ofSoil and Water for 27 months on a heavily ditched, poultry-grain farm on Maryland's Lower Eastern Shore. Conservation 62(4):252-261. Water tables fluctuated above and below shallow ditches, but were always higher than deep Smith, DR., E.A. Warnemuende, B.E. Haggard, and e. Huang. 2006. Dredging of drainage ditches increases ditches. Thus, groundwater flow to shallow ditches was intermittent, but flow to deep ditches short-term transport of soluble phosphorus. Journal of was continuous.Water tables rose rapidly with rain, but drained back from 15 to 60 cm (6 to Environmental Quality 35:611-616. 24 in) the first day after rain. The rate of water table fall decreased rapidly thereafter. Water Stott, R., and e.e. Tanner. 2005. Influence of biofilm on tables frequently perched on top of subsoil clay horizons. Although perching persisted only removal of surrogate faecal microbes in a constructed wetland and maturation pond. Water Science and 24 to 48 hours, nutrient transport could be accelerated ifrapid, lateral movement ofwater to

Technology 51:315-322. ditches occurs. Frequent and widespread concentrations ofgroundwater N03-N greater than Strock,J.S., C.J. Dell, and J.P Schmidt. 2007. Managing natu­ 10 mg L- 1 show subsurface N loss from the farm is probable. High concentrations ofdissolved ral processes in drainage ditches for nonpoint source P existed in groundwater, but P movement in groundwater was restricted. Rain infiltrating nitrogen control.Journal ofSoil and Water Conservation through topsoils mobilized soil P into groundwater and moved considerably high concen­ 62(4):188-196 Vadas, PA., M.S. Srinivasan, PJ.A. Kleinman,J.P Schmidt, and trations of P as deep as 1.5 m (4 ft), where elevated P concentrations persisted for days or A.L. Allen. 2007. Hydrology and groundwater nutrient weeks. Groundwater P concentrations were greatest where high water table hydrology com­ concentrations in a ditch-drained agroecosystem.Journal bined with the greatest soil P concentrations. Delivery of groundwater P to shallow ditches ofSoil and Water Conservation 62(4):178-188. was apparently controlled by near-ditch soil P conditions, while P delivery to deep ditches van Schilfgaarde, J. 1971. Drainage yesterday, today, and tomorrow. II/. Proceedings of the ASAE National was controlled by how deep groundwater flowed. Therefore, limiting soil P accumulation in Drainage Symposium. St. Joseph, MI: American Society near-ditch zones may help reduce P delivery to shallow ditches, and increasing the length of ofAgricultural Engineers. groundwater flow paths through low-P subsoils may help reduce P delivery to deep ditches. van Strien,A.J.,J.Van Der Linden,T.e.P Melman,and M.A.W Noordervliet. 1989. Factors affecting the vegetation of Key words: ditch-drained agroecosystem-groundwater-hydrology-nitrogen concentra­ ditch banks in peat areas in the western Netherlands. Journal ofApplied Ecology 26:989-1004. tion-phosphorus concentrations-water tables van Strien, A.J., T. Van Der Burg, WJ. Rip, and R.e.W Strucker. 1991. Effects ofmechanical ditch management on the vegetation of ditch banks in Dutch peat areas. Journal ofApplied Ecology 28:501-513. Nitrogen (N) and phosphorus (P) are nearly 600 million birds at a wholesale value Vaughan, R.E., B.A. Needelman, PJ.A. Kleinman, and A.L. ofnearly $2 billion. Nearly all ofthe 750,000 Allen. 2007. Spatial variation of soil phosphorus within the primary nutrients that accelerate a drainage ditch network. Journal of Environmental eutrophication in fresh surface waters tons of poultry manure annually produced Quality 36:1096-1104. (Carpenter et at. 1998). Eutrophication is Vaughan, R.E., B.A. Needelman, PJ.A. Kleinman, and M.e. an aging process where influxes of nutrients Peter A. Vadas is a soil scien- Rabenhorst. Forthcoming. Morphology and character­ promote excessive plant and algae growth and ization of ditch soils at an Atlantic Coastal Plain farm. tist at the Dairy Forage Research Soil Science Society ofAmerica Journal. can limit water use for recreation, industry, Center, USDA Agricultural Research Service and drinking. On the Delmarva Peninsula in (ARS), Madison, Wisconsin. M.S. Srinivasan the Mid-Atlantic region ofthe United States is a hydrologist at the Invermay Agricultural (figure 1), the Chesapeake Bay and Delaware Centre, AgResearch Limited, Mosgiel, New Zealand. Peter J.A. Kleinman and John P. and Maryland's Coastal Bays are subject to Schmidt are soil scientists at the Pasture Sys­ eutrophication-related algal blooms (Boesch tems and Watershed Management Research et al. 2001; Boynton 2000). Unit, USDA ARS, University Park, Pennsylvania. The Delmarva Peninsula is home to one Arthur L. Allen is an associate professor at the of the most concentrated poultry industries University of Maryland Eastern Shore, Princess in the United States, and annually produces Anne, Maryland.

178 JOURNAL OF SOIL AND WATER CONSERVATION Figure 1 Location of the University of Maryland Eastern Shore research farm on the Delmarva Peninsula and the location of the 26 groundwater wells relative to ditches and the Manokin River on transport and the integral nature of drain­ the farm. age ditches, groundwater nutrient transport in ditch-drained soils is not well understood, especially for P (Sims et al. 1998). Given the accumulation ofN and P in soils and the hydrologic activity of ditch-drained agroecosystems on the Delm.arva Peninsula, we hypothesized that Nand P transport in groundwater may be greater than is cur­ ,,' rently realized. The long-term objective in the ongoing field study described in this paper is to quantifY Nand P transport in groundwater from a heavily ditch-drained farn'l on Maryland's Lower Eastern Shore. In this paper, we report the findings of the first two years of field monitoring and describe the hydrology and groundwater N and P status.We also describe how variations in soil chemical and physical properties and drainage hydrology may apparently control the fate ofgroundwater Nand P.

Materials and Methods Site Description. We conducted research at the University of Maryland Eastern Shore (UMES) research farm (38°12'22" N, 75°40'35" W) in Princess Anne, Maryland

by this industry is land-applied locally to N03-N, which is the dominant chemical (figure 1). The farm relief is flat, with an agricultural fields as a source of nutrients form. ofN in aerobic soils; N added to soils in approximate elevation of5.5 m (18 ft) above for crops, often at agronomically excessive fertilizers ormanures often leaches downward sea level. Annual rain averages 111 cm (44 rates. Manure application has been impli­ in soils and moves with groundwater (Staver in). Farm soils belong to the poorly-drained cated in the contamination of groundwater and Brinsfield 1998). This is not always the Othello series (fine-silty, mixed, active, mesic and surface water as Nand P are transferred case for P. Soils typically have a strong affinity Typic Endoaquult) and typically have 50 to from soils to the surrounding envirorunent to bind P, and most P added to soils in fertiliz­ 75 cm (20 to 30 in) thick, silt loam topsoils (Boesch et al. 2001; Butler and Coale 2005; ers or manures remains in the topsoil (Malhi that are underlain by 20 cm (8 in) thick, silty Mozaffari and Sims 1994; Sims and Wolf et al. 2003; Rechcigal et al. 1985). Downward clay subsoils, which are in turn underlain by 1993). Delaware, Maryland, and Virginia leaching of P and transport in groundwater very coarse marine sediments. Silty clay sub­ have thus implemented policies to reduce has thus been considered unimportant and soils and their abrupt texture change with agricultural Nand P transport to surface and is not well understood. However, P leach­ the deeper coarse marine sediments impede groundwaters (Sims and Gartley 1996; Coale ing and transport in groundwater has been drainage. Therefore, all fields are drained by et al. 2002;Weld et al. 2000). observed in sandy soils with limited capacity shallow field (-0.3 to 1.0 m [1 to 3 ft] deep) On the Delmarva Peninsula, poultry to retain P, in soils where accumulation ofP and deeper Public Drainage Association manure is routinely applied to cultivated in topsoils fi'om fertilization has saturated soil (PDA) ditches (-1 to 3 m [3 to 9 ft] deep) fields drained by ditches. Organized land retention capacity, and in artificially drained (figure 1). Shallow ditches drain individual drainage on the Delmarva Peninsula began soils with preferential flow pathways and fields in a parallel pattern and flow into in the late 1700s, became a large-scale relatively short groundwater flow distances deeper collection or PDA ditches running endeavor by the early 20th century, and (Sims et al. 1998). perpendicular to shallow ditches. Collection has left an extensive network of drainage In the United States, considerable research ditches drain into PDA ditches, which in ditches throughout portions ofthe Delmarva has investigated the role ofagricultural drain­ turn drain into the Manokin River, a tribu­ Peninsula. Ditches improve crop production age on nutrient transport; however, it has tary ofthe Chesapeake Bay. in poorly drained soils by rapidly removing been conducted primarily in North Carolina, The farm was a poultry operation until sUlface water fi'om fields and lowering water South Carolina, Florida, and Georgia, and 1997, with a more than 20-year history of tables. Consequently, ditches can accelerate more for N than P (Thomas et al. 1995). manure application to soils. Farm soils have groundwater nutrient transport from fields Research in the Mid-Atlantic region is P concentrations greatly in excess of crop to sUlface waters (Mozaffari and Sims 1994; limited and in many instances nonexistent requirements (MeWich-3 P > 60 mg kg-1 1 Sallade and Sims 1997a, 1997b). (Shirmohammadi et al. 1995). Despite the [0.00096 oz Ib- ]), with Mehlich-3 P con­ Soils do not have a great capacity to retain widespread study of agricultural nutrient centrations as great as 450 mg kg-1 (0.0072

I JULY IAUGUST 2 007 VOLUME 62, NUMBER 4 ~ - Figure 2 Schematic of the groundwater wells with discrete-depth sampling devices. l OZ lb-l) in smface horizons and 80 mg kg- (0.0013 OZ lb-l) in subsurface horizons as deep as 1 m (3 ft). Chickens are still raised on the farm, but less frequently and in lesser quantities than before 1997. Fields are cropped to corn, wheat, and soybean and are fertilized to meet corn N demand (~150 Samples

bushel acl yield goal) using manure from 5 cm diameter the farm (50 to 150 kg N ha-1 [133 lb acl] screened PVC pipe 1 1 and 40 to 120 kg P ha- [36 to 107 lb ac- ]) Top or liquid anU110nia (120 to 150 kg N ha- 1 view 3.75 cm diameter 1 [107 to 133 lb ac ]). These nutrient and PVC pipe sealed at / hydrology conditions are representative of bottom with cap ----A;;;.QIII many farms in the region. An ongoing, intensive research proj­ K-Packers to ect at the University of Maryland Eastern seal screened Shore research farm is being conducted sections by the University of Maryland at Eastern Shore and College Park, and the USDA 1.8cm inner Agricultural Research Service in University diameter PVC Sample Park, Pennsylvania. A wide array of run­ pipe open at tubing bottom to off, drainage, ditch, and climate monitoring monitor water equipment has been installed to investigate table the effect offield, manure, and ditch manage­ ment on Nand P export (see also Kleinman et al. 2007 and Schmidt et al. 2007). Groundwater U'ell Design and Installation. In the fall 2003, we installed 26 ground­ water wells at both field centers and ditch edges in northern and southern portions of the farm (figure 1). We positioned wells so water table data would show the direction and magnitude of groundwater flow from ing a hole of the same diameter in the soil a water-tight seal. At the bottom of specific fields to ditches or the Manokin River. The and inserting the screened PVC pipe. We pipe sections between K-packers, we drilled northern portion of the farm is drained by installed wells to ~3 m (10 ft), which was sampling ports and fit rubber grommets into two PDA ditches (~3 m [10 ft] deep) and one as deep as the lowest expected water tables. the holes. We ran plastic tubing from the shallow field ditch (~1 m [3 ft] deep) that We inserted devices into the 21 groundwater ports to the top of the pipe and sealed any run east to west and empty directly into the sampling wells so we could collect sam­ gaps between the gronU11ets and tubing. We Manokin River (figure 1). In the monitored ples from discrete depths in the soil profile then inserted a second solid PVC pipe inside area, two shallow ditches run south to north (figure 2). Discrete-depth sampling devices the first pipe so that it ran through the pipe and empty into a PDA ditch. The south­ consisted of an outermost, solid, 3.75 em cap into the screened well section beneath ern portion of the farm is drained by three, (1.5 in) diameter PVC pipe, on which we it (figure 2). This construction allowed parallel shallow field ditches (0 to 1 m [0 to glued a series of rubber K-packers (Dean groundwater entering screened wells below 3 ft] deep) running east to west and empty­ Bennet Supply, Denver, Colorado).We sealed the sampling device to rise up the innermost ing into a deeper (~1 to 2 m [3 to 6 ft] deep) the bottom of the solid pipe with a cap so PVC tube to a height equal to the water collection ditch (figure 1). The collection no groundwater could penetrate inside the table. Thus in the same well, we could both ditch drains south into a deep PDA ditch pipe. The K-packers slipped directly on the measure depth to groundwater and collect (3 to 4 m [10 to 13 ft] deep) that empties outside of the PVC pipe. They had three groundwater samples at discrete depths for into the Manokin River. rubber ribs and formed a water-tight seal Nand P analysis. With the design of these We designed wells so we could only mea­ between the smaller pipe on which they wells, we were not measuring the water table sure depth to the water table (n = 5; wells were attached and the larger well casing per se, but rather the hydraulic head at the 2,6,14,15,24) or both measure water table into which the smaller pipe was inserted. bottom ofwells. depth and collect groundwater samples for Thus, water entering a well in a section Well sampling devices were either 1.5 or Nand P analysis (n = 21; all other wells). between two K-packers could not mix with 2.4 m (5 or 8 ft) long.The 1.5 m (5 ft) wells All wells consisted of a 5 em (2 in) diam­ water entering well sections above or below had three sampling depths-from 0 to 30,45 eter, PVC pipe screened with thin slits over (figure 2). Our initial laboratory tests con­ to 75, and 90 to 140 em (0 to 12,18 to 30, and its entire length. We installed wells by drill- firmed that the K-packers did indeed create 36 to 56 in)-and were in field centers and

180 JOURNAL OF SOILAND WATER CONSERVATION Figure 3 Weekly water table data from December 2003 to December 2005 for wells on the northern half next to shallow ditches where we expected ofthe farm. higher water tables. Wells 25 and 26 in a field center had four sampling depths-from 0 to 30.0 20,35 to 60, 75 to 100, and 115 to 140 cm Shallow (0 to 8, 14 to 24, 30 to 40, and 46 to 56 ---e lateraI ~.... 29.5 +l-+-t---tl---+--'_~~-e----4'--ii-!\:---fIlh.tt----A-.p.f---- in). The 2.4 m (8 ft) wells had four sampling J::. depths-from 0 to 30,45 to 75,90 to 135, 'Q.O ~ and 150 to 235 cm (0 to 12, 18 to 30,36 to 29.0 Q) 56, and 60 to 94 in)-and were next to col­ ~ lection and PDA ditches where we expected ....III 28.5 deeper water tables. When sampling, we ...... Q) attached tubing fi'om all well depths to a III 28.0 peristaltic pump and simultaneously pumped ~ Deep Q) > PDA all tubes. This ensured groundwater flowed :;::; 27.5 only laterally into wells at the discrete depths III Qj (Kaleris et al. 1995). Some rainfall simulation c:: 27.0 experiments with the wells showed that as water tables were rising, groundwater nutri­ ent concentrations from sampling depths 26.5 -1==~~~==~======r====r======r=J--­ below the initial water table did not change, 12/31/03 5/19/04 10/6/04 2/23/05 7/13/05 11/30/05 while concentrations above the water table Date were greater and changed as water tables rose. Thus, our well design did indeed result Notes: Data from wells with similar water table elevations are averaged for clarity. The approximate depth of shallow field ditches, deep Public Drainage Association ditches, and in only lateral groundwater flow at discrete the Manokin River are indicated. depths. Specific sampling depths ensured we could sample groundwater deeper than all shallow layer. Existence of the perched water table to ditches (Reuter et al. 1998; Walter et al. ditches and slightly deeper than collection would provide initial evidence of relatively 2000). and PDA ditches.We could thus assess Nand rapid lateral groundwater flow and nutri­ Hydrology and TMlter Quality Monitoring. P transport from fields to all ditches. Discrete ent transport along the top of the clay layer We manually measured depth to water table sampling also allowed greater scrutiny of differential groundwater Nand P concentra­ tions with depth (Delin and Landon 1996; Figure It Graham and Downey 1992). Consistent Weekly water table data from December 2003 to December 2005 for wells on the southern half patterns of differential concentrations of the farm. through time could help reveal chemical or physical mechanisms controlling subsurface 30.3 Nand P transport. g Next to two field-center wells (7, 19) .... 29.8 J::. and three ditch-adjacent wells (5, 8, 17), we 'Q.O Shallow also installed shallow piezometers to investi­ "Qj J::. 29.3 lateral gate the possibility of perched water tables. Q) A piezom.eter consisted of a 5 cm (2 in) ~ ....III diameter, 75 cm (30 in) long PVC pipe that .. 28.8 Eastern Q) was screened with thin slits over its bottom .... collection III 30 cm (12 in) and sealed at the bottom.We ~ 28.3 installed piezometers by drilling a hole ofthe Q) :;::;> same diameter in the soil and inserting the III PVC pipe. Even though they would at times Qj 27.8 c:: Southernmost go dry, we installed piezometers to only deep PDA 75 cm (30 in) to ensure that their bottoms 27.3 +----r---~---..,___---_._--_____,_' were above the clay layer in soils and that 12/31/03 5/19/04 10/6/04 2/23/05 7/13/05 11/30/05 we were measuring the hydraulic head above Date the clay layer. Monitoring wells would mea­ sure hydraulic head below the clay layer. A Notes: Data from wells with similar water table elevations are averaged for clarity. The greater hydraulic head in piezometers would approximate depth of shallow field, collection, and deep Public Drainage Association ditches are indicated. indicate a perched water table above the clay

I JULY I AUGUST 2007 VOLUME 62, NUMBER 4 ~ - . Figure 5 Water level data fo'r wells 8, 17, and 19 and their adjacent piezometers. in all wells weekly from November 2003 to February 2006. In June 2004, in wells 5 to 30.1 8,15 to 19, and all piezometers, we installed automated water level recording devices that measured water table depth in wells every 15 minutes. Automated data allowed us to 29.6 determine the nature of water table fluctua­ tions in response to rain and drainage and to detect the occurrence of perched water 29.1 tables. This would help determine when and ~ Piezometer how long water and nutrients might be flow­ Well 8 5 Well ing to shallow ditches. We sampled groundwater montWy in all 28.6 21 sampling wells from November 2003 to February 2006. When collecting samples, we 30.1 let water flow for 5 minutes to purge tubes and ensure we were collecting fresh ground­ water samples. We collected approximately 29.6 250 ml (8.5 fl. oz.) for each sample and fil­ tered samples within 24 hours through 0.45 !-tm (17.7 micro-in) filters and stored them at 4'C (39'F) until analysis.We analyzed samples 29.1

for N03-N by flow injection analysis using a Lachat autoanalyzer (Quick Chem FIA+ Well 19 8000 Series, Lachat Instruments, Loveland, Colorado) and the method of Wendt 28.6 (2000) and dissolved P colorimetrically by 30.1 the method of Murphy and Riley (1962). MontWy samples represented groundwater Nand P concentrations that were in chemi­ I cal equilibrium with soil. We also sampled 29.6 I. wells 9 to 12 and 17 to 20 within 48 hours • after rainfalls greater than 1.25 cm (0.5 in); we sampled wells 25 and 26 during storms when possible, but more typically imme­ 29.1 diately after storms and then every 8 to 12 hours for the next 24 to 48 hours. More Well 17 frequent storm samples represented ground­ water Nand P concentrations that recently 28.6 leached through topsoils and may not have 2/7/05 4/8/05 6/7/05 8/6/05 10/5/05 12/4/05 reached equilibrium with soils. Soil Sampling and Analysis. In December Date 2005, we collected intact soil cores adjacent Notes: Piezometer data are shown only when piezometers were not dried out. Higher water levels to all sampling wells to depths of 2 m (6 ft). for piezometers than wells indicate perched water tables. We sectioned cores and collected all soil from the same depths as discrete-depth ground­ water sampling. We air-dried soil samples (figure 3).Thus, groundwater may flow from flow more east and south to the deep ditch and ground them to pass a 2 nun (0.08 in) this field to shallow ditches only intermit­ and Manokin River than north into the shal­ sieve. We analyzed all soils for Mehlich-3 P tently, and the ditches may often dry out and low ditch. Because water tables were always (Mehlich 1984). lose hydraulic connection with groundwater. higher than deep ditches and the Manokin During drier times, water will instead flow River, groundwater flow to deep ditches and Results and Discussion to deeper PDA ditches or Manokin River. the Manokin River was continuous through Hydrology. In the northern portion of the Water tables in the field with wells 2 and 3 time. farm, water tables in one field center (wells were most often below the shallow ditch to Hydrology was similar in the southern 6 to 11) fluctuated above and below shallow the north but always above the deep PDA portion of the farm (figure 4). Water tables field ditches, but were always higher than ditch and the Manokin River bordering the in the center (wells 15 and 19) and along the deep PDA ditches or the Manokin River field. Thus, groundwater from this field may edges (wells 14,16,17,18,20, and 21) ofthe

182 JOURNAL OF SOIL AND WATER CONSERVATION Table 1 The range and average dissolved P and NO,-N concentrations and number of samples taken for four different sampling depths as compiled across 21 sampling wells from December 2003 through March 2006. N0 -N Dissolved P 3 Sampling depth Range (mg L"l Average (mg L"l n Range (mg L"l Average (mg L"l n

oto 30 em 0.01 to 4.60 0.65 27 0.20 to 39.30 9.7 24 45 to 75 em 0.00 to 12.30 0.53 186 0.01 to 54.80 9.4 159 90 to 135 em 0.00 to 7.40 0.14 297 0.06 to 59.10 14.2 257 150 to 235 em 0.00 to 0.17 0.03 70 3.10 to 38.20 18.4 65

two fields between the shallow ditches fluc­ persisted for only 24 to 48 hours. Perched for groundwater dissolved P concentrations. tuated above and below the shallow ditches water either infiltrated through restrictive However, 0.2 mg L-I (0.000027 oz gal-I) is and the eastern collection ditch. However, layers or drained laterally to ditches. Lateral about the concentration required in the soil water tables were always higher than deep drainage on top of restrictive clay horizons solution for crop growth (Tisdale et al. 1993), PDA ditches to the north or south (wells may result in greater nutrient transport than double the USEPA recommended 0.1 mg L­ 13, 22, 23, and 24). Groundwater may thus infiltration through these horizons (Reuter I (0.000013 oz gal-I) concentration limit for flow to the collection ditch and shallow field et al. 1998;Walter et al. 2000). total P in streams, butless than the 1.0 mg L-I ditches only intermittently, and the ditches Groundwater Nutrient Concentrations: (0.00013 oz gal-I) total P USEPA standard for may dry out and lose hydraulic connec­ Nitrogen. For 540 groundwater samples, wastewater eflluent. To more clearly present results in this paper, we thus used 0.2 mg L-I tion with groundwater. During drier tinles, N03-N concentrations ranged from 0.01 groundwater will flow either north or south to 59.1 mg L-I (0.0000013 to 0.0079 oz (0.000027 OZ gal-I) to represent a lower limit to deep PDA ditches.Water levels in wells 22, gal-I) and exceeded the United States for ground-waterP concentrations that would 23, and 24 in the southern field were nearly Environmental Protection Agency (USEPA) adversely affect water quality if delivered to always below shallow ditches or the eastern 10 mg L-I (0.0013 oz gal-I) drinking water surface waters. For 620 groundwater samples, collection ditch, but always higher than the standard in 54% of all samples. These con­ dissolved P concentrations ranged from 0.0 southernnlOst PDA ditch. Thus, groundwa­ centrations are similar to those observed by to a very high 12.3 mg L-I (0.0 to 0.0016 ter from this field flows continuously to the Staver and Brinsfield (1998) and Sims et al. oz gal-I), which was collected at a depth of deep PDA ditch and not to other ditches. (1996) in Maryland and Delaware coastal 45 to 75 cm (18 to 30 in). Groundwater P Water tables rose rapidly in response to plain soils. Every well had concentrations exceeded 0.2 mg L-1 (0.000027 oz gal-I) in rain (see figure 7 as an example for two field­ of N0 -N greater than 10 mg L-1 (0.0013 only 15% ofall samples and in only 11 wells. 3 center wells, as discussed later), but drained OZ gal-I) on at least one sampling date. Groundwater P concentrations were gener­ back from 15 to 60 cm (6 to 24 in) the In individual wells, 6% to 100% of samples ally greatest in topsoils and decreased with

first day after rain. Nutrient transport may collected had N03-N concentrations greater depth (table 1), but dissolved P as great as 7.4 be greatest during rapid drainage because than 10 mg L-I (0.0013 oz gal-I). In 14 ofthe mg L- I (0.00098 oz gal-I) existed as deep as groundwater rises into nutrient-rich topsoils 21 sampling wells (3 to 5,7 to 12,18,19,22, one meter. These high P concentrations are and because groundwater travel times and and 23) that were located at both field cen­ of a magnitude similar to those observed by

distances decrease as nearby shallow ditches ters and ditch edges, N03-N concentrations Nelson et al. (2005) at depths of 45 and 90 become active (DeWalle 1994). After the were predominately greater than 10 mg L-I cm (18 and 36 in) in two P-enriched soils first day, the rate ofwater table fall decreased. (0.0013 oz gal-I). Of the 350 samples from in the North Carolina coastal plain. These data clearly show that groundwater P trans­ At 5 days after rain, water table fall ranged these 14 wells, 80% had N03-N concen­ from 1 to 10 cm (0.4 to 4.0 in) per day. At trations greater than 10 mg L-I (0.0013 oz port around the farm was not as ubiquitous gal-I). In the remaining wells, only 10 samples as for N0 -N. The following discussion ten days after rain, water fall ranged from 3

1 to 5 cm (0.4 to 2.0 in) per day. Nutrient had N03-N concentrations greater than 10 explores how soil chemistry and hydrology mg L-I (0.0013 oz gal-I). For all wells, N0 ­ transport may be less during slower drainage 3 may have affected dissolved P movement in because shallow ditches become inactive and N concentrations tended to increase with groundwater. _ groundwater must travel farther and longer depth in the soil (table 1), demonstrating the Data for wells 25 and 26, which were in to deep ditches. This presumably allows soils ability of N to leach through soil profiles. field centers and were sampled frequently to adsorb more nutrients from groundwater, There was no apparent pattern to changes after rain storms, suggest a mechanism for soil P mobilization into groundwater. The especially for P, as discussed later. in groundwater N03-N concentrations from Figure 5 shows continuous water table one sampling time to another. It is clear from soil profile around wells 25 and 26 was data from three pairs of wells and piezom­ these data that soil and hydrology condi­ very enriched in Mehlich-3 P (400 mg kg-I eters (8,19, and 17) that had long periods of tions and management practices on the farm [0.0064 oz Ib-I] from 0 to 20 cm [0 to 8 in], 200 mg kg-! [0.0032 oz Ib-I] from 20 to 40 cm record. Wells 8 and 17 were next to shallow support widespread N03-N transport in field ditches, and well 19 was at a field center. groundwater. Environmentally important N [8 to 16 in], and 40 mg kg-I [0.00064 oz Ib-I] Water levels were often higher in piezom­ loss through groundwater flow £i.-om this and from 40 to 60 cm [16 to 24 inJ).These soil P eters than in wells, showing that water tables similar farms in the region is probable. data suggest a high potential to mobilize P in did periodically perch (assumedly on top of Groundwater Nutrient Concentrations: groundwater and that the greatest source of subsoil clay horizons), but perching typically PhosphoYlts. There is no regulatory standard P is topsoil. Figure 6 shows water table and

I JULY I AUGUST 2007 VOLUME 62, NUMBER 4 I~ - Figure 6 Water table and groundwater dissolved P data for wells 25 and 26 from July 13 to August 20, 2004· concentrations were greatest in well 7 (1.9 Date mg L- I [0.00025 oz gal-I]), followed by well 7/13 7/18 7/23 7/28 8/2 8/7 8/12 11 (1.3 mg L-I [0.00017 oz gal-I]), and then 0 24 I I E well 19 (0.2 mg L- [0.000027 oz gal- ]).Well ~ 7 also had more samples with groundwater 20 21 I CI) P concentrations greater than 0.2 mg L- :c 18 (0.000027 oz gal-I) (75%) than well 11 (43%) ell 40 ...... or well 19 (31 %). Figure 8 shows water tables CI) 15 .... 60 were generally highest and rose most often ell ~ E into P-enriched topsoils in well 7, followed 12 ~ 0 80 .... c by well 11 and then well 19. Therefore, P 0.93 0.98 0.80 9 .c.... ell mobilization into groundwater may increase Co 100 c::: with the frequency or duration of water CI) 6 Q tables rise into P-enriched topsoils. 120 0.58 0.29 3 0.04 In contrast to field-center wells, only three (12, 18, and 21) out of 14 wells adja­ 140 0 cent to ditches or the Manokin River had Note: Gray horizontal bars represent depths in the soil profile from which groundwater samples groundwater P concentrations conSIS­ were taken. Values in the gray horizontal bars are dissolved P concentrations measured from tently greater than 0.2 mg L-I (0.000027 oz those depths at different dates. The black line represents depth to water table for the period. gal-I). Soil P was a major difference between Black vertical bars represent depths of rain for individual days. hydrologically similar field-center wells that had high groundwater P concentrations and groundwater P data for the two wells from gest that rain storms can mobilize soil Pinto ditch-adjacent wells that did not. Field-cen­ July 13 to August 20,2004. During this 38­ groundwater to concentrations much greater ter wells had an average MeWich-3 soil P day period, water tables started at about 110 than those in equilibrium with soil. Rain can of 450 mg kg-1 (0.0072 oz lb-I) from 0 to em (44 in) deep, where groundwater P was also leach P down through the soil profile, 20 em (0 to 8 in), 380 mg kg-I (0.0061 oz 0.04 mg L-I (0.0000053 oz gal-1).Two storms and groundwater P concentrations at depths lb-I) from 20 to 40 em (8 to 16 in), and 40 in July raised water tables to about 30 em as great as 120 to 140 ClTI (48 to 56 in) can mg kg-I (0.00064 oz lb-I) from 40 to 60 em (8 in) and increased groundwater P through­ remain elevated for days or even weeks. (16 to 24 in). At the same depths in ditch­ out the soil profile to concentrations that Figure 7 shows groundwater P concentra­ adjacent well profiles, soil MeWich-3 P was were about an order of magnitude greater tions in several wells around the farm were 290,141, and 7 mg kg-I (0.0046,0.0023, and than what long-term sampling suggests are consistently greatest shortly after storms and 0.00011 oz lb-I). Overall, data from field­ P concentrations in equilibrium with soil decreased with tin~e, and they were often as center and ditch-adjacent wells show the (0.2 mg L-I [0.000027 oz gal-I] from 30 to great as 1 to 4 mg L-1 (0.00013 to 0.00053 greatest amount ofsoil P was mobilized into 60 em [12 to 24 in], 0.1 mg L-I [0.000013 oz oz gal-I) at depths greater than 50 em (20 groundwater when high water table hydrol­ gal-I] from 70 to 100 em [28 to 40 in], and in). Once P is mobilized in groundwater, soil ogy combined with excessive soil P. When 0.04 mg L- I [0.0000053 oz gal-I] from 120 will re-adsorb P over time, but complete re­ soil P was less or water tables were deeper, to 140 em [48 to 56 in]). Because topsoils adsorption may apparently take a week or less P was mobilized in groundwater. around these wells had the most P, the source more. Phosphorus mobilized in groundwater Data from field-center wells clearly show of elevated groundwater P was most likely can potentially be delivered to ditches during considerable P was mobilized in ground­ from rain leaching P from topsoils through this re-adsorption period. water during storms, thus providing at least the soil profile. While the above data for wells 25 and 26 a potential for groundwater P delivery to Figure 6 shows that from the end of show a mechanism for P mobilization in ditches. In evaluating sUlface runoff and July onward, frequent rain kept water groundwater, other data show the degree of ditch nutrient export data, Kleinman et al. tables above about 60 em (23.6 in), where P mobilization can vary. In five (7, 11, 19, 25, (2007) also conclude groundwater was a groundwater P concentrations ranged from and 26) ofthe field-center wells, 60% of430 significant contributor ofP to ditches on the 3.2 to 6.3 mg L- I (0.00043 to 0.00084 oz groundwater samples had P concentrations farm. For our data, we looked at groundwater gal-I). In comparison, sUlface runoffcollected greater than 0.2 mg L-I [0.000027 oz gal-I]). P concentrations in ditch-adjacent wells as on August 4, 2004, had a similar dissolved Only wells 3 and 23 were exceptions, but the most concrete evidence for groundwater P concentration of 3.9 mg L-I (0.00052 oz this was more likely due to very few samples P delivery to ditches. Greater groundwater gal-I) (see Kleinman et al. 2007). Because able to be collected (only eight during more P concentrations in field-center wells than the deepest t:\vo salTlpling depths remained than !:\Vo years). Some of the five field-cen­ ditch-adjacent wells (as described above) below the water table, they were presumably ter wells had greater P concentrations than suggests more P was mobilized in groundwa­ not further influenced by P leaching from others. For example, wells 7, 11, and 19 had ter in fields during storms than was actually the topsoil. However, groundwater P from similar soil P concentrations, which suggests delivered to ditches. However, close exami­ these sampling depths remained elevated a similar potential to mobilize P in ground­ nation ofhydrology and groundwater P data for almost a m.onth. Overall, these data sug- water. However, average groundwater P from ditch-adjacent wells provides evidence

184 JOURNAL OF SOIL AND WATER CONSERVATION Figure 1 Groundwater dissolved P concentrations for several wells where samples consistently had groundwater P concentrations greater than 0.2 mg L", plotted against when groundwater the magnitude and duration of groundwa­ samples were collected relative to days after a rain event. ter flow to deep ditches was likely greater 7.0 than to shallow ditches. Therefore, water (A) 0 to 50 em flow to deep ditches or the Manokin River was from more than just near-ditch zones, 6.0 and P delivery to these ditches might be controlled by more than just near-ditch soil 5.0 P conditions. For example, soil P concen­ trations were least around well 12 next to the Manokin River, greater around well 4.0 18 next to a collection ditch, and greatest around well 10 next to a deep PDA ditch 3.0 (figure 9). However, average groundwater P concentrations were greatest in well 12 (0.79 mg L-I [0.00011 oz gal-I]), less in well 2.0 18 (0.09 mg L-I [0.000012 oz gal-I)), and f f f least in well 10 (0.02 mg L-I [0.0000027 oz 1.0 I gal-I)). Therefore, near-ditch soil P condi­ Il tions apparently did not control P delivery to these ditches. Instead, there appeared to be A 11. 0.0 a better relationship between groundwater 0 2 4 6 8 'C Q) P concentrations and the depth in the soil > '0 1.6 where groundwater typically flowed. Water f/) (B) Deeper than 50 em f/) tables fluctuated between 50 and 100 cm is 1.4 (19.7 to 39A in) below the soil surface in well 12, between 100 and 200 cm (39A and 1.2 78.7 in) in well 18, and between 200 and 300 cm (78.7 and 118.1 in) in well 10 (figure 9). 1.0 Deeper water flow suggests longer flow paths from fields to ditches and a greater potential 0.8 for subsoils to adsorb P out of groundwa­ Il ter. Nelson et al. (2005) also observed that 0.6 low-p subsoils in the North Carolina coastal .. Il plain were able to adsorb P out of the soil Il solution. Therefore, increasing the length of 0.4 groundwater flowpaths through low-P sub­ soils may reduce groundwater P delivery to 0.2 deep ditches or the Manokin River. Data from well 12 next to the Manokin Il 1 "i :l[ A A 0.0 + River provide the most compelling evidence o 3 6 9 12 15 for P delivery to ditches or the Manokin. Days after rain Soil P in the topsoil around well 12 (fig­ ure 9) was not enriched compared to field Note: Data are separated into sample depths (Al above 50 em in the soil and (B) below 50 em. centers, suggesting a limited potential for P mobilization in groundwater. However, groundwater P concentrations exceeded 0.2 I for P delivery to ditches. It also suggests Because water tables fluctuated above and mg L- (0.000027 oz gal-I) in 70% ofsamples. variations in drainage hydrology and soil below shallow ditches, water flow to these Furthermore, soil P was 50 mg kg-I (0.0008 I properties may have controlled the extent of ditches was intermittent and likely from only oz Ib- ) at the depth of 60 cm (23.6 in) and I P delivery. near-ditch zones. Thus, P delivery to shal­ 30 mg kg-I (0.00048 oz Ib- ) at the depth of For all six wells adjacent to shallow field low ditches was apparently controlled ITlOre 100 cm (39A in), even though soil P in the I ditches, there was a strong relationship by near-ditch soil P conditions.This suggests topsoils was only 150 mg kg-I (0.0024 oz Ib- ). between average soil P content (mg kg-I) that limiting soil P accumulation in near­ Comparatively, around four other wells (2,3, from 0 to 75 cm (0 to 30 in) deep and aver­ ditch zones might help reduce groundwater 5, and 17) where topsoil P was about 150 mg age groundwater P concentrations (mg L-I) P delivery to shallow ditches. kg- 1 (0.0024 oz Ib-I), soil P at 60 and 100 cm I during the two years of sampling (ground­ Because water tables were always higher was less than 10 mg kg-I (0.00016 oz Ib- ). water P = 0.0003 x soil P + 0.017, /2 = 0.93). than deep ditches or the Manokin River, This suggests that soil P enrichment deeper

I JULY I AUGUST 2007 VOLUME 62. NUMBER 4 ~ - Figure 8 Vertical distribution of Mehlich-3 P in the soil profiles ofwells 7. 11. and 19, along with weekly depth to water table data from December 2003 through February 2006. ated above and below shallow ditches, but were always higher than deep PDA ditches Soil Mehlich-3 P (mg kg-i) or the Manokin River. Thus, groundwater flow to shallow ditches is only intermit­ tent, and shallow ditches may often dry out and lose hydraulic connection with ground­ water. During drier times, water will instead 50 flow to deep ditches or Manokin River. All water tables were always higher than the bottom of deep ditches and the Manokin 100 River. Thus groundwater flow to deep ditches and the Manokin River was continu­ ous through time.Water tables rose rapidly in 150 Soil P response to rain, but drained back from 15 to 60 cm (6 to 24 in) the first day after rain.The -- Water table Well 7 rate of water table decline decreased rapidly 200 thereafter. Data show perched water tables 0 did occur fairly frequently, most likely on top ofrestrictive subsoil clay horizons. Although perching persisted for only 24 to 48 hours, groundwater nutrient transport could be ~ 50 E accelerated if rapid, lateral movement of ~ perched water to ditches occurs. J: +' Data reveal N0 -N can move readily in c. 100 3 Q) groundwater in high concentrations (>10 "C mg L-1 [>0.0013 oz gal-I]) throughout the ·0 farm. Therefore, loss of N through subsur­ VI 150 face groundwater flow from this and similar Well 11 poultl-y-grain farms in the region is prob­ 200 able. While very high concentrations of dissolved P existed in groundwater, P trans­ 0 port in groundwater was restricted compared

to N03-N. During storms, rain infiltrating though high P topsoils was able to mobilize 50 soil P into groundwater and move consider­ ably high concentrations of this mobilized P as deep as 1.5 m (4 ft).These elevated ground­ 100 water P concentrations persisted for days or even weeks. High groundwater P concentra­ tions occurred in essentially all field centers, but were greatest where high water table hydrology combined with the greatest soil Well 19 P concentrations. When water tables were 200 L- ---' deeper or soil P was not as great, ground­ water P concentrations were less. Data also suggest P mobilized in groundwater can be in the profile around well 12 was greater to 0.0002 oz gal-I) regardless of sampling delivered to ditches or the Manokin River. than what leaching ofP down from the top­ date.This suggests a consistent source ofrela­ Phosphorus delivery to shallow ditches was soil might have caused. The hydrology, soil tively high groundwater P concentrations apparently controlled by near-ditch soil P P, and groundwater P characteristics from being delivered to the Manokin River. conditions, while P delivery to deep ditches well 12 combined thus suggest that P mobi­ or the Manokin River was controlled by lized in groundwater from the field to the Summary and Conclusions how deep in subsoils groundwater flowed. east may indeed be moving past well 12 into We monitored groundwater hydrology and Therefore, limiting soil P accumulation in the Manokin River. Furthermore, in the 70% Nand P concentrations during more than near-ditch zones may help reduce P delivery of groundwater P samples that exceeded 0.2 two years on a heavily ditch-drained, poul­ to shallow ditches, and increasing the length mg L-' (0.000027 oz gal-I), P concentrations try-grain farm on Maryland's Lower Eastern of groundwater flowpaths through low-P I ranged from only 0.6 to 1.5 mg L- (0.00008 Shore. Water tables in field centers fluctu- subsoils that can adsorb P may help reduce

186 JOURNAL OF SOIL AND WATER CONSERVATION Figure 9 Vertical distribution of Mehlich-3 P in the soil profiles ofwells 12, 18, and 10, along with weekly depth to water table data from December 2003 through February 2006. Butler, J.S., and FJ. Coale. 2005. Phosphorus leaching in manure-amended Atlantic Coastal Plain soils. Journal of 1 Soil Mehlich-3 P (mg kg- ) EnvirolUllental Quality 34: 370-381. Carpenter, S.R., N.F Caraco, D.L. Correll, R.W Howarth, 0 100 200 300 400 500 A.N. Sharpley, and VH. Smith. 1998. Nonpoint pollu­ 0 tion of surface waters with phosphorus and nitrogen. Ecological Applications 8:559-568. 50 Coale, FJ., j.T. Sims, and A.B. Leytem. 2002. Accelerated deployment of an agricultural nutrient management tool: The Maryland Phosphorus Site Index. journal of 100 Environmental Quality 31: 1471-1476. Delin, G.N., and M.K. Landon. 1996. Multiport well design 150 for sampling of ground water at closely spaced vertical intervals. Ground Water 34:1098-1104. DeWalle, D.R., and H.B. Pionke. 1994. Streamflow genera­ 200 tion on a small agricultural catchment during autunm • Soil P recharge. II. StormfJow periods. journal of Hydrology 250 163:23-42. Water table Graham, W.O., and D. Downey. 1992.A comparison of\vater 300 quality information obtained from depth-integrated versus depth-specific groundwater monitoring devices. Soil and Crop Science Society of Florida Proceedings 0 51: 58-63. Kaleris, v., C. Hadjitheodorou, and A.C. Demetracopoulos. 50 1995. Numerical simulation of field methods for esti­ mating hydraulic conductivity and concentration E 100 profiles. journal of Hydrology 171:319-353. ~ Kleinman, PJ.A.,A.L.Allen. B.A. eedelman,A.N. Sharpley, .=..... 1'A. Vadas, L.S. Saporito, GJ. Folmar, and R.B. Bryant. Q. 150 2007. Dynamics of phosphorus transfers from heavily CI) "C manured Coastal Plain soils to drainage ditches. Journal ofSoil and Water Conservation 62(4):225-235. ·0 200 In Malhi, S.S., J.T. Harapiak, R. Karamanos, K.S. Gill, and N. Flore. 2003. Distribution of acid extractable P 250 and exchangeable K in a grassland soil as affected by long-term sur£1ce application of N, g and K fertilizers. 300 Nutrient Cycling in Agroecosystems 67:265-272. Mehlich, A. 1984. Mehlich 3 soil test extractant: A modifica­ 0 tion of MeWich 2 extractant. Communications in Soil Science and Plant Analysis 15:1409-1416. Well 10 Mozafarti, M., and J.T. Sims. 1994. Phosphorus availability 50 and sorption in an Atlantic Coastal Plain watershed dominated by animal based agriculture. Soil Science 100 157: 97-107. Murphy LJ., and Riley J.1' 1962. A modified single solu­ tion method for determination of phosphate in natural 150 waters. Analytica Chimica Acta 27: 31-33. Nelson, N.O.,J.E. Parsons, and R.L. Mikkelsen. 2005. Field­ 200 scale evaluation of phosphorus leaching in acid sandy soils receiving swine waste. Journal of Enviromnenral Quality 34:2024-2035. 250 Rechcigal, J.E., R.B. Reneau, and D.E. Starner. 1985. Movement of subsurface fertilizer and limestone under 300 irrigated and non-irrigated conditions. Soil Science 140:442-448. Reuter, RJ., P.A. McDan.iel, J.E. Hanmlel, and A.L. Falen. 1998. Solute transport in seasonal perched \vater tables P delivery to deep ditches or the Manokin Reiner. David Otto, joan Weaver, MaryKay Lupton, Todd in loess-derived soilscapes. Soil Science Society of River. Future field and modeling work at the Strohecket, and jim Richards). This paper is dedicated in America journal 62:977-983. site will attempt to more firmly demonstrate memory ofthe late Bill StOll[. Sallade, YE., and J.T. Sims. 1997a. Phosphorus transfor­ mations in the sedunenrs of Delaware's agricultural and quantifY groundwater nutrient transport References drainageways. I. Phosphorus forms and sorption.Journal and delivery to ditches, especially for P. Boesch, D.F, R.B. Brinsfield, and R.E. Magnien. 2001. ofEnvironmemal Quality 26:1571-1579. Chesapeake Bay eutrophication: scientific understand­ Sallade,Y.E., andJ. T. Sims. 1997b. Phosphorus transformations in the sediments ofDelaware's agricultural drainageways. Acknowledgements ing, ecosystem restoration, and challenges for agriculture. The authors thank the scaff of the Nutrient Management journal ofEnvironmental Quality 30:303-320. II. Effects ofreducing conditions on phosphorus release. Laboratory at University Maryland Eastern Shore (Don Boynton, WR. 2000. Impact of nutrient inflows on journal ofEnvironmental Quality 26:1579-1588. Mahan and janice Donohoe) and the Pasmre Systems Chesapeake Bay. 11/ Agriculmre and Phosphorus Schinidt, J.P., c.J. Dell, 1'A. Vadas, and A.L. Allen. 2007. itrogen eX"port from Coastal Plain field ditches.Journal and Watershed Management Research Unit at USDA Management: The Chesapeake Bay, ed. AN. Sharpley, Agricultural Research Service (Terry Troutman, Mike 23-40. Boca Raton, FL: Lewis Publishers. ofSoil and W1ter Conservation 62(4):235-243.

I JULY IAUGUST 2007 VOLUME 62, NUMBER 4 ~ - Shirmohammadi, A., R. D. Wenberg, W F. Ritter, and F. S. WrighL 1995. Effect of agriculrural drainage on water quality in Mid-Atlantic states. Journal of Irrigation and Drainage Engineering 121 :302-306. Sims, JT, and D.C. Wolf. 1993. Poultry waste manage­ ment: Agricultural and environmental issues. Advances in Agronomy 52:1-81. Sims, JT, A.S. Andres, JM. Denver, and WJ Gangloff. Managing natural processes in drainage 1996. Assessing the impact of agricultural drainage on ground and sur£1Ce water quality in Delaware. Final Project Report. Dover, DE: Delaware Department of ditches for nonpoint source nitrogen Natural Resources and Environmental ControL Sims, JT, and KL Gardey. 1996. The phosphorus index: control A phosphorus management strategy for Delaware's agri­ cultural soils. Soil Testing Program Factsheet ST-05. University of Delaware. J.S. Strock, c.J. Dell, and J.P. Schmidt Sims, JT, R.R. Simard, and B.C. Joern. 1998. Phosphorus loss in agricultural drainage: Historical perspective and current research. Journal of Environmental Quality Abstract: In watersheds dominated by agriculture, artificial drainage systems can efficiently 27:277-293. and quickly transport excess water from agricultural soils. The application of more nitro­ Staver, KW, and R.B. Brinsfield. 1998. Using cereal grain gen (N) than a crop uses creates a surplus in the soil and increases the risk of N loss to the winter cover crops co reduce groundwater nitrate con­ environment. We examine issues associated with agricultural N use, N transfer from arti­ tamination in the mid-Atlantic coastal plain. Journal of Soil and Water Conservation 53:230-240. ficially drained agricultural land to drainage ditches, N cycling within ditches, and options Thomas, D.L., CD. Perryy, 11...0 Evans, F.T lzuno, KC. for management. Watercourses in agricultural watersheds often have high concentrations of Stone, and JW Gilliam. 1995. Agricultural drainage N and are effectively N saturated. Numerous processes are involved in N cycling dynam­ effects on water quality in southeastern U.S. Journal of ics and transport pathways in aquatic ecosystems including N mineralization, nitrification, Irrigation and Drainage Engineering 121 (4):277-282. and denitrification. Flow control structures can lower N losses related to artificial drainage Tisdale, S.L., WL. Nelson, JD. Bearon, and JL. Havlin. 1993. Soil fertility and fertilizers. Macmillan Publishing by increasing water retention time and allowing greater N removal. An ongoing study in Company, NewYork, NY Minnesota compares the impact of £low control structures on N losses from paired ditches Walter, M. T, JS. Kim, TS. Steenhuis, JY. Parlange, A. with and without £low control. During the first year of observation, results were mixed, Heilig, RD. Braddock, JS. Selker, and J Boll. 2000. with lower N concentrations in nonstorm event samples from the ditch with the £low con­ Funneled flow mechanisms in a sloping layered soil: Laboracory investigation. Water Resources Research trol structure, but no significant difference in annual total N load between the two ditches. 36:841-849. Appropriate managem.ent of drainage ditches represents a potential opportunity to remove Weld, JL., D.B Beegle, JT Sims, JL. Lemunyon, A.N. biologically available forms of N from drainage water through a combination ofphysical and Sharpley, WJ. Gburek, F.C Coale and A.B. Leytem. biogeochemical processes. 2000. Comparison of state approaches to the devel­ opment of phosphorus indices. III 2000 Agronomy abstracts, 329. Madison, WI: American Society of Key words: ditch-nitrate-nitrogen-total nitrogen-water quality Agronomy-Crop Science Society of America-Soil Science Society ofAmerica. Wendt, K. 2000. Determination of nitrate/nitrite in surface Artificial drainage can increase crop pathogens, and pesticides exported from wastewaters by flow injection analysis. QuickChem yields, reduce risk of saturated soil at agricultural fields to waterways (Gilliam et al. Method 10-107-04-1-A F1A Methodology. Loveland, planting and harvest, and improve eco­ 1999; Randall and Goss 2001). One of the CO: Ladut Instruments. nomic returns for crop producers in many most significant water quality impairments regions of the United States (Pavelis within aquatic ecosystems is accelerated 1987). Starting in the late 1700s, drainage in eutrophication caused by nutrient over­ the United States has been improved by con­ enrichment, a problem often associated structing open channel ditches (frequently with agricultural production (USEPA 2002). by artificially deepening and straightening Although m.any £1Ctors contribute to eutro­ natural waterways) and by installing subsur­ phication, most attention has focused on face tiles that transfer excess water to ditches the supply of carbon (C), nitrogen (N), and or natural watenvays. In the regions with phosphorus (P). A comparison of N:P ratios slowly permeable soils, like much of the among freshwater and estuary ecosystems Midwest, extensive net\¥orks of subsurface indicated that freshwater ecosystems were tile drains and vegetated ditches have been established. In other regions, such as the Jeffrey S. Strock is an associate professor at Delm.arva Peninsula, which have highly the Southwest Research and Outreach Center, permeable soils but also high regional water University of Minnesota, Lamberton, Minne­ tables, improved drainage can be obtained sota. Curtis J. Dell and John P. Schmidt are soil using only open channel ditches. scientists at the Pasture Systems and Watershed Artificial drainage can considerably Management Research Unit, USDA Agricultural increase the amounts of sediment, nutrients, Research Service, University Park, Pennsylvania.

188 JOURNAL OF SOIL AND WATER CONSERVATION