water quality and crop production improvement using a wetland-reservoir and draining/subsurface...

Canadian Water Resources Journal Vol. 32(2): 129–136 (2007) © 2007 Canadian Water Resources AssociationRevue canadienne des ressources hydriques

C.S. Tan1, T.Q. Zhang1, C.F. Drury1, W.D. Reynolds1, T. Oloya1 and J.D. Gaynor1

1 Agriculture & Agri-Food Canada, GPCRC, Harrow, ON N0R 1G0

Submitted December 2006; accepted February 2007. Written comments on this paper will be accepted until December 2007.

Water Quality and Crop Production Improvement Using a

Wetland-Reservoir and Drainage/Subsurface Irrigation System

C.S. Tan, T.Q. Zhang, C.F. Drury, W.D. Reynolds, T. Oloya and J.D. Gaynor

Abstract: In a wetland-reservoir system, tile drainage water and surface runoff water from agricultural fields are routed into a wetland reservoir, rather than into open-ended streams and drainage ditches. The collected water is then recycled back through a controlled tile drainage-subsurface irrigation system to provide subsurface irrigation during times of crop water deficit. The wetland reservoir provides wildlife habitat and serves as a sink to prevent off-site movement (loss) of water and sediments, and also provides a means for intercepting and recycling agricultural nutrients and chemicals via return irrigation. As a result, precipitation water is used more efficiently and the discharge of agricultural sediments and chemicals into off-site surface and ground water resources is substantially reduced. The controlled drainage/subirrigation system (CDS) reduced total nitrate loss by 41% compared to traditional tile drainage (DR). The CDS system also reduced losses of dissolved inorganic phosphorus, dissolved organic phosphorus and total dissolved phosphorus in tile drainage water by 18%, 47% and 36%, respectively, relative to the non-irrigated DR system. During the low rainfall growing seasons of 2001 and 2002, the CDS system increased corn grain yield by 91% (2001) and soybean yield by 49% (2002), relative to the DR system. Thus, the CDS system combined with a wetland-reservoir can be highly effective for improving crop yield and reducing non-point source pollution from agricultural fields.

Résumé : Dans un système de réservoir en milieu humide, l’eau de drainage souterrain et l’eau d’écoulement de surface provenant des champs agricoles sont dirigées vers un réservoir en milieu humide, plutôt que vers des cours d’eau ouverts et des fossés de drainage. L’eau recueillie est ensuite recyclée et retournée à travers un système de drainage contrôlé et d’irrigation souterraine pour assurer une irrigation souterraine en période de déficit d’eau des cultures. Le milieu humide « réservoir » offre un habitat faunique et sert de puits pour prévenir le mouvement vers l’extérieur (pertes) de l’eau et des sédiments, et sert également de véhicule permettant d’intercepter et de recycler les produits chimiques et les éléments nutritifs agricoles au moyen de l’eau d’irrigation qui retourne à son point d’origine. Par conséquent, l’eau des précipitations est utilisée plus efficacement et le rejet des produits chimiques et des sédiments d’origine agricole dans les eaux de surface et souterraines hors site est réduit de manière considérable. Le système de drainage contrôlé et d’irrigation souterraine (DCIS) a réduit la perte d’azote totale de 41 % par rapport au drainage souterrain traditionnel. Le système de DCIS a également permis de réduire les

130 Canadian Water Resources Journal/Revue canadienne des ressources hydriques

© 2007 Canadian Water Resources Association

pertes de phosphore inorganique dissous, de phosphore organique dissous et de phosphore total dissous dans l’eau de drainage souterrain de 18 %, de 47 % et de 36 %, respectivement, comparativement au système de drainage souterrain (sans irrigation). Au cours des saisons de croissance de faibles précipitations de 2001 et de 2002, le système de DCIS a accru le rendement des cultures de grains de maïs de 91 % (2001) et des cultures de soja de 49 % (2002), comparativement au système de drainage souterrain. Le système de DCIS, combiné au milieu humide-réservoir, peut donc s’avérer hautement efficace pour ce qui est d’améliorer le rendement de culture et de réduire la pollution de source non ponctuelle provenant des terres agricoles.

Introduction

Areas of intense agricultural production, such as Essex County, Ontario, face many challenges associated with profitability, sustainability, and off-farm environmental impacts. Of particular concern are severe growing season droughts associated with global warming (Tan and Reynolds, 2003), and non-point source agricultural pollution.

The existing network of drainage tiles and ditches in Essex County is designed to remove excess surface and soil water as quickly as possible. Unfortunately, this approach can increase non-point source agricultural pollution by enhancing the movement of agricultural sediment, nutrients and pesticides into surface and ground water resources (Rudolph and Goss, 1993; Tan et al., 1993). In addition, frequent droughts during the growing season can seriously reduce crop yield. Hence, new agricultural practices must be developed to reduce agricultural non-point source pollution (Tan et al.,1993; Drury et al., 1996) and reclaim drained water for future irrigation (Tan et al., 2002). To address these issues, an integrated reservoir-controlled drainage–subsurface irrigation study was conducted.

Controlled drainage involves installation of risers on tile outflows after planting in order to prevent excessive drainage of the crop root zone. Subsurface irrigation involves pumping water back into the tile drains during water deficit periods to provide irrigation water directly to the crop root zone. The wetland-reservoir component

of the system captures and stores surface runoff and tile drainage water for future use as crop irrigation water. The wetland-reservoir also serves as a sink to prevent off-farm movement of sediments and a means for intercepting and recycling leached agricultural nutrients and chemicals back into the crop root zone. During the growing season, water and dissolved nutrients are pumped out of the reservoir and into a specially constructed tile system that provides a highly efficient form of subsurface irrigation and fertilization for field and vegetable crops.

The objectives of this study were to determine the effectiveness of the integrated wetland-reservoir and controlled drainage-subsurface irrigation system for: i) mitigating nitrate and phosphorus losses from agricultural fields; and ii) improving corn and soybean yields with supplemental subsurface irrigation.

Materials and Methods

Experimental Design and Field Lay Out

The experiment was initiated in the spring of 2000 and consisted of two treatments: controlled tile drainage/subirrigation (CDS) and regular/traditional tile drainage (DR). The experimental site is located on the Essex Region Conservation Authority demonstration farm at Holiday Beach, Ontario.

The field site (Figure 1) includes one CDS plot and one DR plot, each 25 m by 131 m, and enclosed by a raised surface berm. Surface runoff and tile drainage are recorded and routed into a constructed wetland-reservoir in both plots. Water stored in the wetland-reservoir is then used for surface and/or subsurface irrigation of the CDS plot during the growing season.

Both plots contain six 104-mm diameter subsurface drains (4.6 m spacing, 0.6 m average depth) which are connected at their outlets to a control structure. The control structure is used for both controlled drainage and subsurface irrigation. For controlled drainage, a “riser” within the structure increases the elevation of the tile outflow, thereby effectively raising the elevation of the tile and encouraging water retention within the soil profile. For subsurface irrigation (or “subirrigation”), water is pumped into the control structure, thus creating a pressure head that forces water back up the tile lines and into the crop root zone. The pumping rate and pressure head are controlled by

Tan, Zhang, Drury, Reynolds, Oloya and Gaynor 131


a float valve and an overflow pipe permits drainage to proceed when the water table in the centre of the plot rises to the desired level. The control structure is used primarily for subirrigation during the growing season and primarily for controlled drainage during the fall, winter and spring. Each plot is also equipped with a 0.6 m diameter catch basin at its lower boundary to collect surface runoff water (data not shown).

Drainage and surface runoff are directed to a central instrumentation building via 100 mm underground solid PVC pipe. The 3.7 m by 3.7 m instrumentation building is equipped with an electrical circuit breaker panel, fan, heater, a telephone line and data acquisition facilities. Four stainless steel custom fabricated tipping buckets are used at the instrumentation building to measure drainage and surface runoff on a continuous basis. The tipping buckets were calibrated individually to determine the relationship between flow rate and

tip rate. A magnetic reed switch (normally open) is mounted on each bucket so that every tip produces a switch closure detected by a multi-channel data logger. The data logger counts and converts the signals to flow volume on a continuous basis. Samples of surface runoff and tile drainage water from each plot are collected using two separate ISCO auto-samplers. The auto-samplers contain 24 one-litre sample bottles, and are activated by a signal from the data logger and a preset volume interval. Sample collection was based on flow volume with collection volumes varying with the time of year and expected runoff volumes. The water samples were stored in glass bottles at 4oC before analysis for nitrate and phosphorus concentrations.

The nitrate samples were pre-filtered at 0.45 µm (Gelman GN-6, Gelman Science, MI) and analyzed using a TRAACS 800 auto-analyzer (Bran + Leubbe, Buffalo Grove, IL) via the cadmium reduction method

Figure 1. Experimental layout of the controlled tile drainage/

subirrigation (CDS) and regular/traditional tile drainage (DR) area.



(Tel and Heseltine, 1990). The phosphorus samples were vacuum-filtered through a 0.45 µm Millipore membrane (mixed cellulose ester) and analyzed for dissolved inorganic phosphorus (DIP) and total dissolved phosphorus (TDP), using a QuikChem Flow Injection Auto-analyzer (Lachat Instruments, Milwaukee, WI) and the ammonium molybdate ascorbic acid reduction method (USEPA, 1983). The TDP samples were digested using acidified ammonium persulphate ((NH4)2S2O8) oxidation in an autoclave (USEPA, 1983) before analysis. Dissolved organic phosphorus (DOP) was determined as DOP = TDP – DIP. Flow weighted mean nitrate and phosphorus concentrations for each plot were calculated as cumulative loss (on mass basis) from June 1, 2000 to December 31, 2004, divided by the corresponding cumulative water outflow from the plot (Baker and Johnson, 1981).

The wetland-reservoir (Figure 1) is approximately 45.5 m long by 30.5 m wide, with a shallow area around the perimeter to facilitate vegetative growth and a deep section at the centre for greater water storage. With an average depth of 2.75 m, the wetland-reservoir has a storage capacity of 2,300 m3. The volume of tile drainage and surface runoff are measured separately using two Myers MW50 half HP solid handling pumps. A QP10 Quick Prime centrifugal pump installed in the instrumentation building is used to pump water from the wetland-reservoir to the CDS control structure for subirrigation during the growing season.

Agronomy

Corn (Pioneer 34G81) was planted at a rate of 74,000 seeds ha-1 in 76 cm wide rows on May 26, 2000, May 20, 2001 and May 15, 2003. Fertilizer was applied pre-plant at 17.7 kg N ha-1, 76.8 kg P ha-1 and 17.7 kg K ha-1. Urea-Ammonium Nitrate (UAL 28% liquid) was added as a side-dress application in June at 150 kg N ha-1. Weed control consisted of post-emergence applications of Marksman (dicamba/atrazine, 1:2) at the rate of 1.5 kg a.i.ha-1 and Frontier (dimethenamid) at the rate of 0.4 kg a.i.ha-1. Yields were measured for the entire experimental field on November 15, 2000, November 7, 2001 and December 16, 2003. Grain corn yield was reported at 15.5% moisture content. Soybeans were planted at a rate of 580,000 seeds ha-1 in 38 cm wide rows on June 1, 2002 and June 15,

2004. Weeds were controlled using post-emergence applications of glyphosate (0.6 to 1.0 kg a.i.ha-1). Soybean yields were measured for the entire field on October 16, 2002 and November 15, 2004.

Results and Discussion

Precipitation and Subirrigation

The 2000, 2003 and 2004 growing seasons (May to October) were wet, with total growing season precipitation above the 43-yr average by 12% (53.3 mm), 27% (122.7 mm) and 22% (100.1 mm), respectively. The 2001 and 2002 growing seasons were dry, with total growing season precipitation being, respectively, 12% (52.7 mm) and 36% (165.8 mm) below the 43-yr average (Table 1). The controlled drainage/subirrigation plot was irrigated by maintaining a water table level at 40 cm below soil surface using subirrigation through regular tile drainage during the growing season. Subirrigation was initiated when the crop height was appproximately 60 cm, usually around the first week of July. Subirrigation was not required in 2000 and 2004. Subirrigation for CDS treatment was initiated on July 3 and terminated on August 17, 2001 with a total of 287 mm of subirrigation water added. In 2002, subirrigation for the CDS treatment was initiated on July 15 and terminated on August 14 with a total of 203 mm subirrigation water added. In 2003, subirrigation for the CDS treatment was only required during July 6 to July 30 with 153 mm subirrigation water added.

Water and Nutrient Losses

The total tile drainage volume for the period from June 2000 to December 2004 was 10,919 m3 ha-1 (10,919 KL ha-1) for the DR treatment and 7,637 m3 ha-1 (7,637 KL ha-1) for the CDS treatment (Figure 2). The CDS treatment had 43% less water loss through tile drainage than the DR treatment. The flow weighted mean (FWM) nitrate concentrations in tile drainage were similar between the DR and CDS treatments (Figure 3); however, the CDS treatment reduced total nitrate loss by 41% compared to the DR treatment (Figure 4). This was due primarily to the reduced tile drainage volume (Figure 2) and increased crop yields under the CDS treatment (Table 2). Other researchers



have also found that CDS reduced nitrate loadings through subsurface drainage relative to DR (Fogiel and Belcher, 1991; Drury et al., 1996; Tan et al., 1993). The flow weighted mean (FWM) DIP concentrations in tile drainage were similar between the DR and CDS treatments (Figure 5). However, the CDS treatment reduced the flow weighed mean DOP and TDP concentrations by 33% and 11%, respectively, compared to the DR treatment (Figure 5). The CDS treatment reduced DIP, DOP and TDP losses in tile drainage water by 18%, 47% and 36%, respectively, relative to the DR treatment (Figure 6). More importantly, tile

drainage accounted for 91% of the total phosphorus loss from the DR treatment and 62% of the total phosphorus loss from the CDS treatment (data not shown). In other words, phosphorus loss occurred primarily through tile drainage rather than by surface runoff, which is contrary to what is normally assumed. The likely reason for this is extensive preferential flow through the fine-textured soil as a result of an abundance of worm holes, root channels and cracks. Large phosphorus losses through tile drainage from fine-textured soil were also found by Zhang et al. (2002).

Table 1. Monthly precipitation (mm) for the period 2000 to 2004 compared to the long-term average (1960 to

2004) at the experimental site.

Month 2000 2001 2002 2003 2004 Long-Term Avg.

Precipitation (mm)

January 16.6 24.4 44.1 14.8 49.2 53.1February 22.4 68.6 58.5 37.4 24.8 52.7March 33.2 15.9 29.4 52.2 92.0 69.8April 75.8 48.8 80.4 68.2 42.0 80.6

May 16.3 77.9 95.4 133.8 222.8 75.1June 95.6 62.8 41.6 84.8 79.4 77.7July 82.2 28.9 64.7 58.8 57.9 81.4August 133.8 16.9 21.4 142.8 122.9 82.9September 110.5 77.0 37.8 94.4 22.6 80.8October 73.4 142.3 31.8 66.6 53.0 60.6

November 43.8 60.4 79.6 79.4 94.2 71.4December 39.4 50.6 59.6 76.6 73.8 72.4

Growing Season (May-October) 511.8 405.8 292.7 581.2 558.6 458.5

Yearly Total( January-December) 742.8 674.5 644.3 909.8 934.6 858.5

Table 2. Effect of controlled drainage/subirrigation (CDS) and regular/traditional tile drainage (DR) on corn

yields in 2000, 2001 and 2003, and on soybean yields in 2002 and 2004.

Treatments Yields (kg/ha)

Corn Corn Soybean Corn Soybean

2000 2001 2002 2003 2004

Regular/Traditional Tile Drainage/No irrigation (DR) 6663 3706 2216 7405 1269Controlled Drainage/Subirrigation (CDS) 7155 7064 3308 9029 1508



Figure 2. Total tile drainage volume for regular/

traditional tile drainage (DR) and controlled drainage/

subirrigation (CDS) for the period, June 1, 2000 to

December 31, 2004.

Figure 3. Flow weighted mean (FWM) nitrate

concentration for regular/traditional tile drainage

(DR), controlled drainage/subirrigation (CDS), and

wetland reservoir for the period June 1, 2000 to

December 31, 2004.

Figure 4. Total tile nitrate loss for regular/traditional

tile drainage (DR) and controlled drainage/

subirrigation (CDS) for the period June 1, 2000 to

December 31, 2004.

Figure 5. Flow weighted mean (FWM) phosphorus

concentrations for regular/traditional tile drainage

(DR), controlled drainage/subirrigation (CDS), and

wetland-reservoir for the period June 1, 2000 to

December 31, 2004.

The wetland-reservoir is designed to be a water and nutrient “recycling” system where most of the drained water and leached agrochemicals are intercepted and recycled back to the field. Hence, the overall loss of water, nutrients and pesticides to the off-field environment (e.g., streams and lakes) is minimal, regardless of runoff volume or type of tile drainage system (i.e., CDS versus DR). It is interesting to note, however, that the wetland-reservoir water had substantially lower nitrate and phosphorus concentrations relative

to the tile drainage water (Figures 3 and 5), which is probably the result of nitrate and phosphorus uptake by aquatic plants and algae.

Crop Yields

The CDS treatment consistently increased corn and soybean yields relative to the DR treatment (Table 2). There was a trend for soybean to show a negative



relationship between rainfall and yield for the DR treatment, however, the relationship was positive for corn possibly due to differences in water use by the two crops. In the low rainfall years of 2001 and 2002, corn yield was increased by 91% (2001) and soybean yield was increased by 49% (2002) where supplemental subsurface irrigation was applied with the CDS treatment. In the relatively wet years of 2000, 2003 and 2004, supplemental irrigation with the CDS treatment increased corn yield by 7% (2000) and 22% (2003) and soybean yield by 19% (2004).

Conclusions

The CDS system reduced total tile nitrate loss by 41% compared to the DR system (from 93.8 kg N ha-1 to 66.5 kg N ha-1), over a 4.5 year period with two different crops (corn, soybean). The CDS system also reduced phosphorus losses by 18% (DOP), 47% (DIP) and 36% (TDP) relative to the DR treatment. In dry years, irrigation from the wetland-reservoir using the CDS system increased corn grain yield by 91% (2001) and soybean yield by 49% (2002) relative to the non-irrigated DR system. Hence, the combined wetland-reservoir-controlled drainage-subirrigation system

can be highly effective for improving crop yields in water deficit situations and reducing non-point source pollution from agricultural fields.

Acknowledgements

This research was supported by grants from the Agriculture and Agri-Food Canada Matching Investment Initiative program, the Canada Trust Friends of the Environment Foundation, and the Essex Region Conservation Authority (ERCA). Appreciation is also expressed to Mr. M. Soultani, Mr. K. Rinas, Mr. B. Hohner, Ms. M. Reeb, and Ms. J. Gignac for expert technical assistance. We also acknowledge the cooperation of Mr. T. Impens (ERCA) and Mr. J. Deslippe (contractor).

References

Baker, J.L. and H.P. Johnson. 1981. “Nitrate-Nitrogen in Tile Drainage as Affected by Fertilization.” Journal of Environmental Quality, 10: 519-522.

Drury, C.F., C.S. Tan, J.D. Gaynor, T.O. Oloya and T.W. Welacky. 1996. “Influence of Controlled Drainage-Subirrigation on Surface and Tile Drainage Nitrate Loss.” Journal of Environmental Quality, 25: 317-324.

Fogiel, A.C. and H.W. Belcher. 1991. “Water Quality Impacts of Water Table Management Systems.” ASAE paper 91-2596. ASAE, St. Joseph, MI.

Rudolph, D. and M.J. Goss. 1993. “The Farm Groundwater Quality Survey.” Report to Agriculture Canada, Guelph, ON.

Tan, C.S., C.F. Drury, J.D. Gaynor and T.W. Welacky. 1993. “Integrated Soil, Crop and Water Management System to Abate Herbicide and Nitrate Contamination of the Great Lakes.” Water Science and Technology, 28(3-5): 497-507.

Figure 6. Cumulative loss of dissolved inorganic

phosphorus (DIP), dissolved organic phosphorus

(DOP) and total dissolved phosphorus (TDP) in tile

drainage water for regular/traditional tile drainage

(DR) and controlled drainage/subirrigation (CDS) for

the period June 1, 2000 to December 31, 2004.



Tan, C.S., C.F. Drury, W.D. Reynolds, T.Q. Zhang and J.D. Gaynor. 2002. “Reservoir-Irrigation System Improves Water Quality and Increases Crop Production.” Paper presented at the National Conference on Agricultural Nutrients and their Impact on Rural Water Quality, Waterloo, Ontario, April 28-30, 2002.

Tan, C.S. and W.D. Reynolds. 2003. “Impacts of Recent Climate Trends on Agriculture in Southwestern Ontario.” Canadian Water Resources Journal, 28: 87-97.

Tel, D.A. and A. Heseltine. 1990. “The Analyses of KCL Soil Extracts for Nitrate. Nitric and Ammonium using TRAACS 800 Analyzer.” Communication of Soil Science and Plant Analysis, 21: 1681-1688.

U.S. Environmental Protection Agency. 1983. “Methods for Chemical Analysis of Water and Wastes.” EPA-600/4-79-020. Method 365.3.

Zhang, T.Q., C.S. Tan, C.F. Drury, J.D. Gaynor, W.D. Reynolds, T.W. Welacky, B. Ma and R. Fleming. 2002. “Soil P Losses through Tile Drainage in a Clay Loam Soil Amended with Various Composts.” SERA-IEG 17 Meeting on Reducing Phosphorous Losses from Agriculture. June 25-29, 2002, Fort Collins, CO.

water quality and crop production improvement using a wetland-reservoir and draining/subsurface...

Documents