ASAE Conference Proceeding
This is not a peer-reviewed article.
Subsurface Drain Flow Characteristics During a 15-Year Period in Minnesota
Pp. 17-24 Drainage VIII the Proceedings of the Eighth International Drainage Symposium, 21-24 March 2004 (Sacramento, California, USA), ed. Richard Cooke. ,21 March 2004 . ASAE Pub #701P0304
Two tile drainage research facilities were installed on a Webster clay loam at the University of Minnesota's Southern Research and Outreach Center in southern Minnesota in the mid-1970s. One facility consists of 8 individual tile-drained plots (facility A) with a simulated tile spacing of 27.4 m and the other (facility B) contains 36 individual tile-drained plots with a simulated 15.2 m spacing. Drain flow and nitrate concentration measurements were taken on these plots planted to continuous corn (Zea mays L.) (facility A) and corn and soybean (Glycine max) annually (facility B) from 1987 through 2001. Sixty eight to 71% of the annual drain flow and 71 to 73% of the annual nitrate loss occurred during the 3-month (April-June) period. Drain flow was greatly affected by major precipitation events. Greater than 50% of the annual nitrate loss occurred in 10 to 18% of the days drainage occurred. The drainage hydrograph showed flashier characteristics for the 15.2 m spacing compared to the 27.4 m spacing. Results from these two facilities during this 15-yr period have significant implications on: (1) design and approach to drainage research, (2) design and performance of landscape storage structures, including controlled drainage and constructed wetlands, to mitigate nitrate levels in subsurface drainage water (3) fertilizer and manure N management, and (4) the role of subsurface drainage on loss of nitrate to surface waters including the Mississippi River and Gulf of Mexico.KEYWORDS. Drainage, Precipitation, Water quality, Nitrate
Subsurface drainage is a common water management practice in highly productive agricultural areas with poorly drained soils that have seasonally perched water tables or shallow groundwater. This management practice increases crop productivity, reduces risk, and improves economic return to crop producers. In the North Central Region of the USA, eight states (Illinois, Indiana, Iowa, Michigan, Minnesota, Missouri, Ohio, and Wisconsin) are some of the most highly drained states in the nation (Zucker and Brown, 1998). Together these eight states account for over 20.6 million hectares or 37% of the total cropland drained by surface and subsurface drainage (Fausey et al., 1995).
However, subsurface drainage water often contains significant amounts of nitrate-nitrogen (N), which are then transported to surface water bodies including lakes and streams. The zone of hypoxia in the Gulf of Mexico has been attributed to increased loadings of nitrate in the Mississippi River (Rabalais et al., 1996). Nitrate concentrations in the Mississippi River are generally highest in tributaries emanating from Illinois, Iowa, and Minnesota and vary seasonably, usually being greater in winter, spring, and early summer and lower in late summer and early autumn (Antweiler et al., 1995). The elevated nitrate-N concentrations in rivers in these states generally coincide with the geographical location of extensive drainage systems and the time of greatest subsurface tile drainage.
Tile drainage has been monitored in many research studies to assess the impact of agricultural management practices on surface and groundwater quality (Hallberg et al., 1986; Goss et al., 1988; Kanwar et al., 1988; Randall and Iragavarapu, 1995; Bjorneberg et al., 1996; Randall et al., 1997; Randall et al., 2000; and Randall and Mulla, 2001) and can be used to identify best management practices (BMPs) for N management. Nitrate-N concentrations and losses in subsurface drainage in these studies have been shown to vary greatly among years due primarily to differences in annual and growing season precipitation (Randall and Goss, 2001) and by dry and wet climatic cycles (Logan et al., 1994; Randall, 1998). Moreover, a few days of high-flow events can lead to most of the annual nitrate loss in some years (David et al., 1997).
Because the influence of precipitation on drain flow and nitrate loss can be huge, long-term drainage research is vital to our understanding of subsurface drainage and nitrate losses from agricultural production systems. The purpose of this investigation was to identify drainage flow characteristics and patterns across a 15-year period from two drainage research facilities in Minnesota to assist in the approach to and design of future drainage and water management research.
Materials and Methods
Two drainage research facilities were installed at the University of Minnesota's Southern Research and Outreach Center in 1975 and 1976. The facility installed in 1975 consisted of eight 13.7 m by 15.2 m plots separated from each other by 4.6m and further isolated to a 1.8 m depth by the installation of 12-mil plastic around each plot's perimeter. Further details on these plots (facility A) having a simulated tile spacing of 27.4 m were reported by Buhler et al. (1993) and Randall and Iragavarapu (1995). In 1976, 36 additional drainage plots were installed using a somewhat similar design except that plots measured 6.1 by 9.1 m with a 1.5 m isolation distance between plots and a simulated tile spacing of 15.2 m. Further details on this site (facility B) are reported in Randall et al. (2003a).
The data presented herein were all collected under ambient conditions. Daily flow rates were determined at 0800 hrs on a Monday through Friday schedule and on Saturday and Sunday when precipitation events occurred that increased flow rates during the weekend. In addition, flow rates were taken more than one time per day when large precipitation events occurred within that day. The amount of water flowing from each tile line during a 1-min interval was measured and converted to millimeters per plot per day. Water samples were collected manually in 250-mL plastic bottles for nitrate-N analysis three times a week (Monday, Wednesday, and Friday) when tile flow exceeded 0.30 mm d -1 (10 mL min -1 plot -1 ). In addition to the Monday, Wednesday, and Friday collection schedule, water samples were also collected on days when peak flow was occurring due to a large precipitation event, on the first three days of flow in the spring, and on the first three days of flow after a summer period when no flow occurred. Water samples were stored frozen until subsequent laboratory analysis. Nitrate was analyzed by the colorimetric Cd-reduction method; therefore, concentration data include nitrite-N, which was assumed to be extremely small. Total nitrate-N lost (flux) was calculated by multiplying the nitrate-N concentration for each sample by the total calculated flow for the same period.
Continuous corn was grown on facility A throughout the 15-year period. Experiments compared two tillage systems from 1982-1992 (Randall and Iragavarapu, 1995), dairy manure vs. urea fertilizer from 1994-1997 (Randall et al., 2000), and hog manure vs. urea from 1998-present. A corn-soybean rotation with corn grown on 16 plots and soybean grown on 16 plots each year was planted on facility B in all years. Time and rate of fertilizer N studies were conducted to determine BMPs for N (Randall et al., 2003b).
Results and Discussion
Annual precipitation at Waseca from 1987-2001 ranged from 557 mm (32% below normal) in 1989 to 1282 mm (56% above normal) in 1991 (Table 1). Annual precipitation averaged 912 mm for the 15-yr period and was 11% wetter than normal. Twelve of 15 years were wetter than normal. Growing season precipitation during the 5-month period ranged from 317 mm (37% below normal) in 1989 to 807 mm (61% above normal) in 1993. For the 15-year period, growing season precipitation averaged 548 mm or 9% above normal. Ten of 15 years had wetter-than-normal growing seasons. Growing season precipitation as a percent of annual precipitation ranged from a low of 49% in 1998 to a high of 70% in 1987 and 1993 and averaged 61% for the 15-year period, which is the same as the 30-year normal. In summary, the 1987-2001 period was marked by wetter-than-normal conditions and a 2.5-fold difference among years in growing season precipitation. These conditions afforded a wide range of plentiful drainage events for this analysis.
Table 1. Annual and growing season precipitation and annual subsurface drainage from two drainage research facilities at Waseca, MN from 1987-2001.
- - - - - - - - - - - - mm - - - - - - - - - - - -
- - - - - - - drainage (mm) - - - - - - -
† = May - September
‡ = 1961-1990.
Annual drainage from the plots in facility A averaged 307 mm, but ranged from < 60 mm in the three driest years (1987-1989) to > 500 mm in four very wet years (Table 1). Drainage facility B, which has less un-tiled border between the plots and a subsurface tile 2.1 m deep located 6.8 m from the end of each plot, produced lower amounts of drain flow. Drainage averaged 178 mm across the 15-year period with < 35 mm in the three driest years and > 300 mm in three wet years. The data from both facilities show that high amounts of annual precipitation do not necessarily lead to high subsurface drainage. This was particularly true in 1992, 1996, and 1998 when growing season precipitation totaled only about 50% of annual precipitation compared to the normal of 61%. Similar comparisons can be made for growing season precipitation, i.e., high precipitation in 1993, 1997, and 2000 did not result in as much drainage as in 1990 and 2001. The correlation coefficients between annual precipitation and subsurface drainage for facilities A and B are 0.87 and 0.81, respectively. The correlation coefficients between growing season precipitation and subsurface drainage for facilities A and B are 0.82 and 0.77, respectively. All correlation coefficients were significant at the P=99% level, however.
Monthly distribution of annual subsurface drainage for the two facilities is found in Table 2. Sixty eight and 71% of the annual drainage occurred in the 3-month period (April-June) for facilities A and B, respectively. Virtually no drainage occurred in December, January and February when the soils are usually continuously frozen and in September when the water content of the soil profile is lowest. Slight drainage occurred in late March as the soils were thawing and in October and November in years when fall precipitation recharged the soil water content to field moist capacity.
Table 2. Monthly distribution of annual subsurface tile drainage and nitrate losses in drainage from two drainage research facilities with different drain tile spacings and plot design for the 15-yr (1987-2001) period at Waseca, MN.
- - - - - - - - - - % - - - - - - - - - -
- - - - - - - - - - - - - - - % - - - - - - - - - - - - - - -
The monthly distribution of subsurface drainage in relation to 30-year normal monthly precipitation and average seasonal water use for corn in southern Minnesota is shown in Fig. 1. This graphical presentation of the water budget clearly shows: (1) the large effect of water usage by corn during the 3-month (June, July, and August) period on subsurface drainage and (2) the effect of spring rainfall on subsurface drainage in April and May prior to significant water use by corn. Corn and soybean are usually planted in late April or early May in southern Minnesota. Thus, water usage by these crops is minimal during these months and into June. Because soil moisture in the rooting profile is normally recharged in the fall by rainfall and/or early spring by rainfall and snow melt, substantial drainage occurs in April, May, and June when precipitation exceeds evapotranspiration and before water usage by the crop accelerates. Water usage by corn exceeds precipitation in late June, July and much of August, resulting in depleted soil water reserves in the rooting profile. As a result very little drainage occurs from August until fall freezeup. When drainage does occur in August and October, it is usually associated with a very heavy precipitation event. Additionally, the drainage period tends to be short after large rainfall events in August due to plant demand for water.
Figure 1. Relationship between monthly subsurface tile drain flow from facility B in 1987-2001 and 30-yr normal monthly precipitation and water use (ET) by corn at Waseca, MN.
Precipitation received in December through March almost always occurs as snow. However, much of the water in the snow does not recharge the available soil water profile because of sublimation and evaporative losses during the winter months when the soils are frozen. Moreover, the snow usually blows resulting in zones of accumulation (drifts) in protected areas, thus causing great spatial variability with respect to soil water recharge.
Nitrate-N losses in subsurface drainage were closely aligned with the monthly distribution of drainage water (Table 2). Seventy one percent of the nitrate from the continuous corn plots (facility A) was lost during the April-June period while 73% of the nitrate from the corn-soybean rotation plots (facility B) was lost during that 3-month period. Virtually no nitrate was lost in January, February, September and December.
Major drainage events and other drainage characteristics during the 15-year period are shown in Table 3. Subsurface tile drainage occurred an average of 80 and 66 days per year in facility A and B, respectively. The difference between the two facilities may have been partially due to the wider simulated tile spacing (27.4 m) in facility A compared to 15.2 m in facility B.
Table 3. Subsurface tile drainage characteristics, including major drainage events, for the two drainage research facilities for the 15-yr (1987-2001) period at Waseca, MN.
Avg. no. days drainage occurred, annually
Avg. no. days of drain flow to obtain
= 50% of annual drainage
Percent of days drainage occurred to obtain
= 50% of annual drainage
Avg. greatest 1-day drainage event (mm)
Avg. percent of annual drainage occurring in
maximum 1-day event
Daily drain flow amounts for each drainage facility were arranged in order from greatest to least to develop cumulative drainage for each year and to determine the number of days to achieve = 50% of annual drainage. The data in Table 3 show that = 50% of the annual drainage was achieved in 14 days or 18% of the days the tiles flowed in facility A. For facility B with narrower tile spacing, = 50% of the annual flow occurred in 7 days or 10% of the days the tile flowed. Greatest 1-day events averaged 23 and 24 mm of drainage for facilities A and B, respectfully. On average, from 8 (facility A) to 13% (facility B) of the total annual discharge occurred in this one day. Thirty percent of the time this 1-day event occurred in April with an equal distribution among May, June, July, August, and October. These data show the flashy nature of subsurface drainage in a corn and soybean-dominated landscape in southern Minnesota. In addition, the data suggest that decreasing tile spacing from 27.4 to 15.2 m may exacerbate the flashy drainage response to convective, thunderstorm precipitation events in these soils and under these climatic conditions and cropping system.
Implication of Findings
Two primary characteristics of subsurface drainage from corn and soybean cropping systems were identified in this 15-year study in southern Minnesota. First, 68 to 71% of the annual drain flow and 71 to 73% of the nitrate was lost in drainage occurring in April, May, and June. Second, 50% or more of the annual drain flow total occurs in 7 to 14 days (10 to 18% of the days the tile lines flow) and 8 to 13% of the annual drainage can occur in one-day events. These primary findings can have significant implications on: (1) design and approach to drainage research, (2) design and performance of landscape storage structures to mitigate nitrate levels in subsurface drainage water, (3) fertilizer and manure N management, and (4) role of subsurface drainage on loss of nitrate to surface waters including the Mississippi River and the Gulf of Mexico.
Design and Approach to Drainage Research
Because almost 75% of the drain flow occurs in the 3-month period beginning when the soil profile thaws or shortly after, the sampling protocol needs to be designed and implemented in such a manner that all early-season drainage is measured and sufficient samples collected. This means that instrumentation needs to be in order and protected from the cold, and perhaps freezing conditions, during these times. On the other hand, if sample collection is done manually, labor must be sufficient close to the study area to allow convenient checking for drainage initiation and subsequent frequent flow measurement and sample collection. In addition, the sampling protocol and instrumentation needs to be designed to capture the major 1-day drainage events when as much as 13% of the annual drainage and nitrate can be lost. In other words, sample collection systems must be designed to capture the rapidly ascending and descending portion of the hydrograph in addition to accurately capturing the descending tail of the hydrograph.
Major storm events causing large 1- to 2-day drainage events can also contribute to drainage outflow problems from research plots if the outflow mains receiving drainage from the plots are completely filled with drainage water from other portions of the field. Thus, site selection for drainage plots with respect to outflow capability to handle drainage from the plots and the rest of the drainage watershed is critical.
Design and Performance of Landscape Storage Structures
Controlled drainage has been shown to minimize subsurface drain flow and nitrate losses from agricultural cropping systems by limiting drainage during those times of the year when field operations are not being conducted. Given the drainage characteristics described in this study, controlled drainage would have limited utility and impact on losses of water or nitrate from a southern Minnesota landscape. Field operations, including tillage, fertilizer and pesticide application, and planting, are conducted from shortly after the soils thaw in late March to early April until early June. During this period, controlled drainage would not be acceptable because farmers desire to drain soils to about a meter to reduce compaction and facilitate these field operations. However, this is the period when most of the drain flow and nitrate loss in drainage occurs. Controlled drainage would be acceptable to most corn and soybean producers in southern Minnesota from mid-June through September and again in November, but less than 30% of the annual drainage occurs during this time. Fear of creating shallow rooting systems with controlled drainage in June would also be a potential barrier to many producers.
Re-constructing wetlands in strategies places in the drained landscape has been proposed as method to mitigate nitrate losses and reduce peaks in the hydrograph of smaller rivers accepting subsurface drainage after significant rainfall events. Given the nature of subsurface drainage being concentrated in April - June plus the magnitude of major storm events and their distribution (60% occurring in April - June), the size and location of the constructed wetland needs very careful consideration. If not, the drainage characteristics found in this study suggest these constructed wetlands could easily be non-functional and ineffective due to very short residence time during much of the year.
Fertilizer and Manure N Management
Fertilizer N and manure are often applied in the fall for a number of reasons, including more available time and labor, suitable field conditions, lack of compaction, manure availability, storage limitations and lower price. However, the potential for nitrate losses in subsurface drainage is increased because nitrification converts the fertilizer and manure to nitrate before and during the period of maximum drainage and before plant uptake. In other words, fall application of these N sources, especially when nitrification inhibitors are not used, leads to greater losses of nitrate in the drainage water and poorer N use efficiency.
Spring and in-season application of fertilizer N and manure reduces the potential for nitrate loss and generally increases N use efficiency, but other challenges facing the farmer and supplier like those mentioned above may outweigh the agronomic, economic, and environmental advantages associated with spring application. Thus, the drainage characteristics described in this study do present additional N management challenges for Minnesota farmers.
Role of Drainage Losses of Nitrate on Hypoxia
Hypoxia in the Gulf of Mexico has been linked to nitrate loading from the Mississippi River into the Gulf, primarily from April through June. Because >70% of the annual nitrate loss in drainage from corn-soybean agriculture in southern Minnesota occurs in this same time period, it is likely this drainage contributes significantly to the hypoxia condition. This would be especially true if: (1) fertilizer N and manure were applied primarily in the fall; (2) N rates applied were greater than needed for optimum corn yield thereby creating high levels of residual soil N susceptible to leaching into tile drainage, and (3) a year or two of dry weather preceded a wet spring where ample to excessive subsurface drainage occurred.
Antweiler, R. C., D. A. Goolsby, and H. E. Taylor. 1995. Nutrients in the Mississippi River. pp. 73-86. In R. H. Meade (ed.) Contaminants in the Mississippi River, 1987-92. U.S. Geological Survey, Circ. 1133.
Bjorneberg, D. L., R. S. Kanwar, and S. W. Melvin. 1996. Seasonal changes in flow and nitrate N loss from subsurface drains. Trans. ASAE 39:961-976.
Buhler, D. D., G. W. Randall, W. C. Koskinen, and D. L. Wyse. 1993. Atrazine and alachlor losses from subsurface tile drainage of a clay loam soil. J. Environ. Qual. 22:583-588.
David, M. B., L. E. Gentry, D. A. Kovacic, and K. M. Smith. 1997. Nitrogen balance in and export from an agricultural watershed. J. Environ. Qual. 26:1038-1048.
Fausey, N. R., L. C. Brown, H. W. Belcher, and R. S. Kanwar. 1995. Drainage and water quality in Great Lakes and Cornbelt states. J. Irrig. and Drain. Engrg. 121(4):283-288.
Goss, M. J., P. Colburn, G. L. Harris, and K. R. Howse. 1988. Leaching of nitrogen under autumn sown crops and the effects of tillage. pp. 269-282. In D. S. Jenkinson and K. A. Smith (ed.) Nitrogen efficiency in agricultural soils. Elsevier Applied Science. England.
Hallberg, G. R., J. L. Baker, and G. W. Randall. 1986. Utility of tile-line effluent studies to evaluate the impact of agricultural practices on groundwater. pp. 298-326. In Proc. Conf. Agricultural Impacts on Groundwater, Omaha, NE. 11-13 Aug. 1986. Nat'l Well Water Assoc., Dublin, OH.
Kanwar, R. S., J. L. Baker, and D. G. Baker. 1988. Tillage and split N-fertilization effects on subsurface drainage water quality and crop yields. Trans. ASAE 31:453-460.
Logan, T. J., D. J. Eckert, and D. G. Beak. 1994. Tillage, crop and climatic effects on runoff and tile drainage losses of nitrate and four herbicides. Soil & Tillage Res. 30:75-103.
Rabalais, N. N., R. E. Turner, D. Justic, Q. Dortch, W. J. Wiseman, Jr., and B. K. Sen Gupta. 1996. Nutrient changes in the Mississippi River and system responses on the adjacent continental shelf. Estuaries 19:386-407.
Randall, G. W. 1998. Implications of dry and wet cycles on nitrate loss to subsurface tile drainage. pp. 53-60. In Proc. 7th Annual Drainage Symposium. Drainage in the 21st Century. 8-10 March, 1998. Orlando, FL.
Randall, G. W., and M. J. Goss. 2001. Nitrate losses to surface water through subsurface, tile drainage. pp. 95-122. In R. F. Follett and J. L. Hatfield (eds.) Nitrogen in the Environment: Sources, Problems, and Management. Elsevier Science B. V. England.
Randall, G. W., D. R. Huggins, M. P. Russelle, D. J. Fuchs, W. W. Nelson, and J. L. Anderson. 1997. Nitrate losses through subsurface tile drainage in CRP, alfalfa, and row crop systems. J. Environ. Qual. 26: 1240-1247.
Randall, G. W., and T. K. Iragavarapu. 1995. Impact of long-term tillage systems for continuous corn on nitrate leaching to tile drainage. J. Environ. Qual. 24:360-366.
Randall, G. W., T. K. Iragavarapu, and M. A. Schmitt. 2000 Nutrient losses in subsurface drainage water from dairy manure and urea applied for corn. J. Environ. Qual. 29:1244-1252.
Randall, G. W., and D. J. Mulla. 2001. Nitrate-N in surface waters as influenced by climatic conditions and agricultural practices. J. Environ. Qual. 30:337-344.
Randall, G. W., J. A. Vetsch, and J. R. Huffman. 2003a. Nitrates in subsurface drainage from a corn-soybean rotation as affected by time of nitrogen application and use of nitrapyrin. J. Environ. Qual. 32:000-000.
Randall, G. W., J. A. Vetsch, and J. R. Huffman. 2003b. Corn production on a subsurface-drained mollisol as affected by time of nitrogen application and nitrapyrin. Agron. J. 95:1213-1219.
Zucker, L. A., and L. C. Brown (Eds.). 1998. Agricultural drainage: Water quality impacts and subsurface drainage studies in the Midwest. Ohio State Univ. Ext. Bul. 871.