Drainage water management combined with cover crop enhances reduction of soil phosphorus loss

STOTEN-21962; No of Pages 10 Science of the Total Environment xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Drainage water management combined with cover crop enhances reduction of soil phosphorus loss T.Q. Zhang ⁎, C.S. Tan, Z.M. Zheng, T. Welacky, Y.T. Wang Harrow Research and Development Center, Agriculture and Agri-Food Canada, Harrow, ON N0R 1G0, Canada

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Surface runoff and tile drainage flow volumes were reversely affected by CC and DWM. • Total volumes of field water discharge were similar, regardless of DWM-CC treatments. • CDS reduced DRP loss in drainage water, which however became less effective with CC. • CDS combined with CC enhanced reduction in soil PP and TP losses.

a r t i c l e

i n f o

Article history: Received 6 November 2016 Received in revised form 1 February 2017 Accepted 3 February 2017 Available online xxxx Editor: Jay Gan Keywords: Phosphorus Surface runoff Tile drainage Cover crop Drainage water management Surface water quality

a b s t r a c t Integrating multiple practices for mitigation of phosphorus (P) loss from soils may enhance the reduction efficiency, but this has not been studied as much as individual ones. A four-year study was conducted to determine the effects of cover crop (CC) (CC vs. no CC, NCC) and drainage water management (DWM) (controlled drainage with sub-irrigation, CDS, vs. regular free tile drainage, RFD) and their interaction on P loss through both surface runoff (SR) and tile drainage (TD) water in a clay loam soil of the Lake Erie region. Cover crop reduced SR flow volume by 32% relative to NCC, regardless of DWM treatment. In contrast, CC increased TD flow volume by 57 and 9.4% with CDS and RFD, respectively, compared to the corresponding DWM treatment with NCC. The total (SR + TD) field water discharge volumes were comparable amongst all the treatments. Cover crop reduced flow-weighted mean (FWM) concentrations of particulate P (PP) by 26% and total P (TP) by 12% in SR, while it didn't affect the FWM dissolved reactive P (DRP) concentration, regardless of DWM treatments. Compared with RFD, CDS reduced FWM DRP concentration in TD water by 19%, while CC reduced FWM PP and TP concentrations in TD by 21 and 17%, respectively. Total (SR + TD) soil TP loss was the least with CDS-CC followed by RFD-CC, CDS-NCC, and RFD-NCC. Compared with RFD-NCC, currently popular practice in the region, total TP loss was reduced by 23% with CDS-CC. The CDS-CC system can be an effective practice to ultimately mitigate soil P loading to water resource. © 2017 Elsevier B.V. All rights reserved.

1. Introduction

⁎ Corresponding author. E-mail address: [email protected] (T.Q. Zhang).

Increased phosphorus (P) use to meet ever-increasing demands for food, feed, fibre, and fuel has concomitantly increased P export from agricultural fields, which constitutes a critical non-point source of P

http://dx.doi.org/10.1016/j.scitotenv.2017.02.025 0048-9697/© 2017 Elsevier B.V. All rights reserved.

Please cite this article as: Zhang, T.Q., et al., Drainage water management combined with cover crop enhances reduction of soil phosphorus loss, Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.02.025

2

T.Q. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

eutrophication in streams and lakes (Sims et al., 1998). To address this environmental concern, development of soil-test and crop-demand based nutrient management plans are necessary to ensure fertilizer P being utilized in a way that is both economically and environmentally responsible (Sharpley et al., 2003). Some other complementary strategies, such as water management systems that restrict nutrient loss (Tan et al., 2007) and cover crops that biologically remove soil legacy nutrients susceptible to loss (Dabney et al., 2001), have also received increasing attention. Water management systems, such as tile drainage, allow removal of excessive soil water, resulting in improved soil physical quality and crop yields (Tan et al., 2007). However, traditional tile drainage may lead to increased nutrient loss (Tan and Zhang, 2011). To mitigate nutrient loss while increasing crop yields, the regular free tile drainage has been advanced with an option of installing a “riser” on the tile outlet to implement water table control and sub-irrigate crops, namely “controlled drainage with sub-irrigation (CDS)” (Tan et al., 1993). The outlet riser serves a dual function of controlling drainage outflow during the period of excess water and returning the water back into the tile lines to sub-irrigate crops under drought conditions. Recent field studies have shown that incorporation of CDS system with field crop production can relieve the negative impacts of drought stress on crops during the growing season and the environmental degradation associated with off-field movement of nutrients (King et al., 2014). Overwinter cover crops have been demonstrated to be effective in reducing surface runoff flow, enhancing evapotranspiration of soil water, and extracting residual nutrients (Dabney et al., 2001). Studies have reported that cover crops dramatically reduced NO3-N loss up to 84% in surface runoff (Sharpley and Smith, 1991) and up to 62% in tile drainage (Constantin et al., 2010). However, the benefits of cover crops to P transport are generally less clear than erosion and NO3-N leaching. The effects of cover crops on dissolved P loss have been inconsistent, with some studies even indicating increased dissolved P losses in runoff (Riddle and Bergstrom, 2013), while they are effective in reducing particulate P (PP) and total P (TP) losses in surface runoff (Sharpley and Smith, 1991). Having summarized the effects of cover crops on surface runoff P loss from two watersheds under clean-tilled peanuts, Sharpley and Smith (1991) found that cover crops dramatically reduced PP and TP losses, compared to bare soil. However, soluble P loss in runoff was greater in the presence of cover crops than fallow. This indicates the successful control of one form of P loss by cover crops may unintentionally exacerbate the loss of the other, and thus complicates their effects on P loss. The effects of cover crops on P loss in both surface runoff and tile drainage water, as well as the synergistic effects of drainage water management systems and cover crop, must be investigated, if truly beneficial management practices are to be developed. The objective of this study, therefore, was to determine both the individual and the combined effects of drainage water management and a winter wheat cover crop on field water discharge flow volume and P losses in both surface runoff and tile drainage on a fine-textured soil of the Lake Erie basin. 2. Materials and methods 2.1. Experimental site and design and agronomic management The four-year study with a corn (Zea mays L.)-soybean (Glycine max. (L) Merr.) rotation was established in late 1999 and continued until 2003 in a Brookston clay loam soil (fine-loamy, mixed, mesic Typic Argiaquoll) on the Eugene F. Whalen Experimental Farm, Agriculture and Agri-Food Canada, in Southwestern Ontario, Canada (42°13′N, 82°44′W). The 43-year means of annual air temperature and precipitation at the field site were 8.9 °C and 830 mm, respectively. Weather data were collected using a weather station located within 0.5 km of the site (Table S1). Annual precipitation data were presented based on the cropping year (i.e. from Nov 1st when the soil was started for preparation of cropping the next year with fall tillage until Oct 31st the following year when crop was harvested). The soil contained 280 g kg−1 sand, 350 g kg−1

silt, 370 g kg−1 clay, 21.4 g kg−1 organic C, 2.5 g kg−1 total N, 14.1 mg kg−1 NO3-N, 4.91 mg kg−1 NH4-N, 11.3 mg kg−1 water extractable P, 7.49% degree of P saturation determined using the Melich-3 procedure (Wang et al., 2010) and 191 mg kg−1 Mehlich-3 extractable K, with a pH of 7.5, prior to the initiation of the study. The experimental design of the field site consisted of eight treatment combinations of two drainage water managements (DWM) (i.e., regular free drainage, RFD, vs. controlled drainage with sub-irrigation, CDS) with two compost amendments (i.e., yard waste leaf compost and liquid pig manure compost) or two soil crop covers (i.e., no cover crop, NCC, vs. winter wheat as cover crop, CC). The treatments were arranged in a completely randomized block design with two replicates. The plot size was 15 m wide by 67 m long. However, only the four combinations of DWM with soil crop covers were selected for this study to determine the effects of cover crop along with DWM on soil P loss. Two tiles spaced out 7.5 m were located in each plot at 0.6 to 0.7 m depth. The tiles ran parallel to the length of each plot. In the CDS plots, a riser was installed for each plot to effectively “raise” the average water table from the drain depth to 0.3 m below soil surface. The risers were continuously in place except for brief periods during planting and harvesting when they were released to prevent soil structure damage from field operation or wheel traffic compaction associated with excessively wet soil. Sub-irrigation in the CDS plots was initiated during the growing season once the water level behind the risers dropped below the one set for 0.3 m from soil surface. This was accomplished by pumping water from an irrigation pond filled with municipal water to the risers via an underground 50mm diameter polyethylene pipe. Water meters (Neptune T-10; Neptune Equipment Co., Cincinnati, OH) located at the control structures recorded the total volume of irrigation water delivered to each plot. Both N and P loadings in sub-irrigation water were insignificant due to their low concentrations (Table S2). In addition, while the plots of CDS treatments were on water table controlled mode in both 2000 and 2003, no sub-irrigation was conducted due to sufficient natural rainfall during the growing season. Winter wheat (cultivar AC Ron in 1999; AC Essex in other years) was seeded as a cover crop in the designated plots shortly after corn or soybean harvest in late Oct. or early Nov. using a no-till drill at 105 kg ha−1 in the first 3 years (1999–2001) and at 112 kg ha−1 in 2002. The cover crop was killed using glyphosphate [either Roundup (Monsanto) at 0.84 kg ha−1 or Vantage (Dow Agrosciences) at 1.4 kg ha−1] in May or early June of each spring. Corn (variety N58D1) was planted on 17 May 2000 and 22 May 2002 at 79,700 seeds ha−1 in 76 cm wide rows using a Kinze four row planter. Soybean was planted on 8 June 2001 (variety A2553) and 17 June 2003 (variety S20Z5) at 112 kg seeds ha−1 in 20.3 cm wide rows using a John Deer 1590 no-till drill. Starter fertilizer (18-46-0 at 142 kg ha−1) was applied at planting and a side-dress of 28% UAN (urea ammonium nitrate) solution (150 kg N ha− 1) injected into the soil ~ 15 cm from each corn row at ~ 10 cm depth in June when corn was at six-leaf stage. No potassium (K) was applied to corn production due to the high native soil K at the site. No fertilizers were applied for soybean production. Weed control was achieved by pre-emergent broadcasting atrazine (1-chloro-3-ethylamino-5-isopropylamino-2,4,6-triazine at 1.1 kg a.i. ha− 1), metribuzin (0.56 kg a.i. ha−1) and dual magnum (1.05 kg a.i. ha− 1) for corn and dual magnum (1.2 kg a.i. ha− 1), metribuzin (0.42 kg a.i. ha−1) and pursuit (0.08 kg a.i. ha−1) for soybean. Fall cultivation included annual fall disking operation (~ 7.5 cm deep) to incorporate corn and soybean stubble before planting winter wheat. Spring tillage with two passes of an S-tine cultivator and packer was conducted prior to planting either corn or soybean. 2.2. Surface runoff and tile drainage water discharge monitoring, sampling and phosphorus determination There was a 30 cm high berm, as well as a 1.2 m deep plastic barrier installed, on three sides (excluding the lower end with the catch basin)

Please cite this article as: Zhang, T.Q., et al., Drainage water management combined with cover crop enhances reduction of soil phosphorus loss, Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.02.025

T.Q. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

of each plot to prevent surface and subsurface water movement between plots. Surface runoff and tile drainage water were routed from field plots using non-perforated tiles to a monitoring building which contained separate catch basins for both surface runoff and tile drainage water from each plot, as well as sump pumps, water meters, data loggers, and auto-samplers. In addition, a backup generator was present and automatically engaged in the event of electric power interruptions to ensure year-round data and water sample collection without any missing. Additional details on methods and procedures of data collection were described previously in Tan et al. (1993). Surface runoff and tile drainage water samples were collected using auto-samplers (CALPSO 2000S, Buhler Gmbh & Co., Uzwil, Switzerland) after every 500 to 3000 L of flow, depending on the time of year, on a year-round continuous basis. The samples were returned to laboratory after each collection. A portion of the samples was filtered through 45 μm millipore filters and analysed for dissolved reactive P (DRP) and total dissolved P (TDP). Another portion of unfiltered water samples was analysed for TP with concentrated sulphuric acid and H2O2 digestion. Particular P was then calculated as the difference between TP and TDP. All determination of P in water samples or digestions was performed using a Flow Injection Auto-Analyzer (QuikChem FIA + 8000 series, Lachat Instruments, Loveland, CO) with the ammonium molybdate ascorbic acid reduction method of Murphy and Riley (1962). Data on field water discharge and P loss are reported on a cropping year basis (i.e. from Nov. 1st to Oct. 31st the following year). 2.3. Statistical analysis The normality of data distribution was firstly examined, with logtransformations conducted when necessary to obtain homogeneous variance. The effects of DWM and CC on flow volume and P concentration and loss in surface runoff and tile drainage water were determined using a 2 × 2 factorial completely randomized design analysis of variance (ANOVA) with the GLM procedure of SAS (SAS, 2006). The ANOVA was performed separately for each cropping year. When a significant interaction between DWM and CC was found, a least square mean procedure was used to evaluate the significance of difference between treatment means. All statistical analyses were performed at the significance level of P ≤ 0.05. 3. Results and discussion 3.1. Surface runoff and tile drainage flow volumes There was considerable variation in annual surface runoff flow volume amongst cropping years over the four-year study period (Table 1). Generally, the annual surface runoff flow volumes were the lowest for all treatments in 1999–2000 and the largest for the NCC treatments in 2000–2001. Given the similarity of precipitation between 1999–2000 and 2000–2001 (Table S1), it appeared that the production of surface runoff was also related to event intensity and temporal distribution of precipitation, in addition to its total anural amount. The two large rainfall events, accounting for 30.4% of the annual total precipitation, occurred in June and July, when the crop had great demand for soil water supply during the late vegetative and early reproductive stages for corn in 1999–2000 (OMAFRA, 2002). In 2000–2001, the three large rainfall events, accounting for 45.5% of the annual total precipitation, happened in August, September, and October, when the crop demand for soil water supply declined significantly after the physiological maturity stage of soybean. Formation of surface runoff was affected by CC in each of the four years of study, except for 2001–2002 in which CC effect was not observed (Table 1). There were also interactions between DWM and CC on the formation of surface runoff in two (i.e. 1999–2000 and 2000– 2001) of the four years of study. However, annual surface runoff flow volume averaged over the four-year period was affected only by CC.

3

Table 1 Annual (Nov. 01–Oct 31) surface runoff and tile drainage flow volumes as affected by drainage water management (controlled drainage with sub-irrigation, CDS, vs. regular free tile drainage, RFD) and cover crop (no cover crop, NCC, vs. cover crop, CC) in a clay loam soil, ON, Canada. Year

CDS-NCC

CDS-CC

RFD-NCC

RFD-CC

Surface runoff flow volume (mm yr ) 1999–2000 97 (8.3) 53 (32) 2000–2001 203 (10) 80 (54) 2001–2002 111 (1.4) 81 (45) 2002–2003 103 (6.2) 85 (15) Average 129 (1.4) 75 (36)

71 (6.8) 116 (1.0) 76 (2.7) 97 (5.1) 90 (0.8)

63 (12) 82 (31) 91 (32) 62 (6.9) 75 (20)

Tile drainage flow volume (mm yr−1) 1999–2000 81 (22) 93 (15) 2000–2001 127 (3.4) 246 (51) 2001–2002 123 (0.9) 214 (16) 2002–2003 128 (9.4) 170 (3.7) Average 115 (8.9) 181 (21)

97 (17) 212 (17) 213 (15) 165 (20) 171 (17)

79 (2.6) 245 (6.6) 237 (16) 185 (27) 187 (13)

−1 a

Statistical significance (P N F)b 1999–2000

2000–2001

2001–2002

2002–2003

Average

Surface runoff flow volume DWM n.s. n.s. CC ** ** DWM CC* * *

n.s. n.s. n.s.

n.s. * n.s.

n.s. * n.s.

Tile drainage flow volume DWM n.s. CC n.s. DWM CC* n.s.

* * n.s.

* * n.s.

* * *

n.s. * n.s.

a

Numbers in parentheses are standard errors (n = 2). *, **, significant at P ≤ 0.05 and 0.01 levels, respectively; n.s., not significant at P ≤ 0.05 level. b

Cumulative surface runoff flow volume over the four-year period and the resultant annual surface runoff flow volume average for the CC treatments were 32% lower than those for the NCC treatments (Table 1 and Fig. S1A), suggesting winter wheat CC was effective in reducing surface runoff. It is well known that cover cropping reduces soil erosion by lowering the velocity and the quantity of surface runoff (Dabney et al., 2001). Also, the increased evapotranspiration and infiltration of soil water by cover crops (Sharpley and Smith, 1991), especially under drought conditions (i.e. 1999–2000 cropping year), might have further mitigated the surface runoff flow under CC, relative to NCC. The annual tile drainage flow volume varied considerably from one cropping year to the other (Table 1). Similar to surface runoff flow volume, the annual tile drainage flow volume was generally the lowest in 1999–2000, and the largest in 2000–2001. The annual tile drainage flow volume was affected by either DWM and/or CC, except for 1999– 2000 when it was influenced by neither DWM nor CC. Consequently, tile drainage flow volume averaged over the four-year period was affected interactively by DWM and CC. The CDS produced significantly smaller annual as well as four-year cumulative flow volume of tile drainage than did the RFD under NCC (Table 1, Fig. S1B). Both the annual and the four-year cumulative tile drainage flow volumes under CC were similar between CDS and RFD, but under NCC they were 33% greater with RFD than CDS. This was expected, as the effective water table depth was much shallower for CDS than RFD (i.e. 0.3 m for CDS vs. 0.6 m for RFD) and the saturated hydraulic conductivity was higher with RFD than CDS under CC than NCC (Drury et al., 2014). The cover crop canopy acts as a physical barrier between rainfall and soil surface, and its root growth results in the formation of soil pores, which, in addition to enhancing soil macro-fauna habitat, provides pathways for precipitates to infiltrate into soil profile, rather than draining off the field as surface flow (Kaspar et al., 2001; Joyce et al., 2002), This is particularly the case in soils where preferential flow predominates the pathway for water downward movement (Sims et al., 1998; Tan and Zhang, 2011; King et al., 2014). Reynolds et al. (2003) noted that better soil

Please cite this article as: Zhang, T.Q., et al., Drainage water management combined with cover crop enhances reduction of soil phosphorus loss, Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.02.025

4

T.Q. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

structure, lower bulk density, and higher wet aggregate stability induced by cover crops also enhanced the surface hydraulic conductivity and promoted greater infiltration of water. Previous studies indicated CDS was effective in reducing tile flow (Fisher et al., 1999; Elmi et al., 2005; Tan and Zhang, 2011). However, it appeared that the roles of CDS in reducing subsurface flow were partially abated, when combined with overwinter CC. This explains the comparable tile drainage flow volumes found with CDS-CC, RFD-NCC and RFD-CC (Table 1). The total field water discharge, the sum of annual flow volumes of surface runoff and tile drainage, were largely comparable amongst treatments, ranging from 244 mm yr−1 for CDS-NCC to 261 mm yr−1 for RFD-CC, even though CDS-NCC produced an extremely higher surface runoff flow volume and a lower tile drainage flow volume than the others (Fig. S1). 3.2. Flow-weighted mean (FWM) P concentrations in surface runoff and tile drainage water While both flow volume of field water discharge and P concentration are the key and direct factors contributing to soil P loss, it is the latter that plays a greater role (Zhang et al., 2015b). The annual FWM DRP, PP and TP concentrations in surface runoff water varied highly amongst treatments and cropping years, ranging from 0.15 to 0.58, from 0.17 to 0.89, and from 0.47 to 1.25 mg P L−1 for DRP, PP and TP, respectively (Table 2 and Fig. 1A). These values were markedly higher than the critical thresholds of 0.01 mg DRP L−1 and 0.03 mg TP L−1 for triggering eutrophication effects in lakes (Environment Canada, 2004). The annual FWM PP concentration in surface runoff was generally higher than FWM DRP. In over 69% of samples analysed, PP accounted for N50% of TP contained (Table 2), suggesting that PP dominated the soil TP loss in surface runoff water. This result is consistent with findings from other studies (Sharpley and Smith, 1991; Tan and Zhang, 2011; Ball-Coelho et al., 2012). The FWM DRP concentrations in surface runoff averaged over the four-year period were neither affected by the individual treatment of DWT and CC, nor the interaction between the two factors, although they were influenced by CC in one of the four years, 2001–2002 (Table 2). In contrast, the annual surface runoff FWM values of PP concentration in both 1999–2000 and 2001–2002 and TP concentration in 2001–2002, as well as the resultant averages of both PP and TP over the four-year period, were significantly affected by CC, regardless of DWM treatments. The FWM PP and TP concentrations averaged across both DWM treatments over the four-year period under CC reduced by 27 and 14% for PP (0.45 mg P L− 1 under CC vs. 0.33 mg P L− 1 under NCC) and TP (0.45 mg P L−1 under CC vs. 0.33 mg P L−1 under NCC), respectively, relative to NCC. The non-significant DWM effect on P concentrations of various forms (i.e. DRP, PP and TP) in surface runoff water was aligned with its effect on surface runoff flow volume (Table 1). While the effects of cover crop on surface runoff water quality is a function of climatic, soil and cover crop plant factors (Dabney et al., 2001), it is generally recognized that it can be effective in reducing PP and TP concentrations in surface runoff in association with sediment transport, even though its effect on soluble P was inconsistent (Sharpley and Smith, 1991; Bechmann et al., 2005). The FWM DRP concentrations in tile drainage water varied widely, with annual averages ranging from 0.05 to 0.33 mg P L−1 yr− 1 (Table 3 and Fig. 1B). The FWM DRP concentrations over the four-year period for all of the treatments were generally the highest in 1999– 2000 and the lowest in 2000–2001. This might have been attributable to the concentration- or dilution-effect of the corresponding tile drainage flow in 1999–2000 and 2000–2001, respectively (Table S1). As a result, the effects of DWM and CC on FWM DRP concentrations in tile drainage water varied depending on the cropping year. The FWM DRP concentrations were significantly affected by DWM in both cropping years of 2000–2001 and 2001–2002, but by CC in 2001–2002 and 2002–2003 cropping years (Table 3). The four-year average of the DRP

Table 2 Annual flow-weighted mean (FWM) dissolved reactive P (DRP), particulate P (PP) and total P (TP) concentrations (mg P L−1 yr−1) in surface runoff water as affected by drainage water management (controlled drainage with sub-irrigation, CDS, vs. regular free tile drainage, RFD) and cover crop (no cover crop, NCC, vs. cover crop, CC) on a clay loam soil, ON, Canada. Year

CDS-NCC

CDS-CC

RFD-NCC

FWM DRP concentration in surface runoff water (mg P L−1)a 1999–2000 0.31 (0.14) 0.31 (0.02) 0.30 (0.08) 2000–2001 0.15 (0.06) 0.16 (0.02) 0.16 (0.04) 2001–2002 0.31 (0.10) 0.20 (0.06) 0.32 (0.11) 2002–2003 0.25 (0.09) 0.26 (0.06) 0.19 (0.05) Average 0.23 (0.09) 0.24 (0.01) 0.23 (0.06)

RFD-CC 0.58 (0.16) 0.22 (0.02) 0.19 (0.03) 0.23 (0.01) 0.29 (0.04)

FWM PP concentration in surface runoff water (mg 1999–2000 0.48 (0.05) 0.28 (0.02) 2000–2001 0.39 (0.08) 0.35 (0.03) 2001–2002 0.89 (0.07) 0.30 (0.03) 2002–2003 0.17 (0.08) 0.25 (0.04) Average 0.47 (0.07) 0.30 (0.01)

P L−1) 0.36 (b0.01) 0.39 (0.01) 0.77 (0.07) 0.23 (0.02) 0.42 (b0.01)

0.25 (0.11) 0.38 (b0.01) 0.46 (0.09) 0.29 (0.11) 0.36 (0.04)

FWM TP concentration in surface runoff water (mg 1999–2000 0.94 (0.07) 0.76 (0.03) 2000–2001 0.62 (b0.01) 0.61 (0.01) 2001–2002 1.25 (0.06) 0.56 (0.11) 2002–2003 0.49 (0.04) 0.55 (0.11) Average 0.79 (0.03) 0.62 (b0.01)

P L−1) 0.90 (0.02) 0.66 (0.03) 1.13 (0.19) 0.47 (0.04) 0.75 (0.06)

1.00 (0.28) 0.69 (0.02) 0.67 (0.12) 0.60 (0.10) 0.74 (0.08)

Statistical significance (P N F)b 2002–2003

Average

FWM DRP concentration in surface runoff water DWM n.s. n.s. n.s. CC n.s. n.s. * DWM CC* n.s. n.s. n.s.

1999–2000

n.s. n.s. n.s.

n.s. n.s. n.s.

FWM PP concentration in surface runoff water DWM n.s. n.s. n.s. CC * n.s. * DWM CC* n.s. n.s. n.s.

n.s. n.s. n.s.

n.s. * n.s.

FWM TP concentration in surface runoff water DWM n.s. n.s. n.s. CC n.s. n.s. * DWM CC* n.s. n.s. n.s.

n.s. n.s. n.s.

n.s. * n.s.

a b

2000–2001

2001–2002

Numbers in parentheses are standard errors (n = 2). *, significant at P ≤ 0.05 level; n.s., not significant at P ≤ 0.05 level.

concentrations in tile drainage water, however, was only influenced by DWM. Neither CC nor its interaction with DWM was found to affect the four-year average of DRP concentration in tile drainage water. Averaged across both CC and NCC treatments over the four-year period, CDS reduced the FWM DRP concentration in tile drainage water by 19% (0.190 mg P L−1 for CDS vs. 0.154 mg P L−1 for RFD), compared with RFD. The effect of CDS on reducing DRP concentration in tile drainage water in this study was consistent with Nash et al. (2015) who reported the flow-weighted ortho-P concentration, equivalent to DRP in the current study, in tile water was significantly lower with the controlled tile drainage (0.09 mg L− 1) than with regular free tile drainage (0.15 mg L−1) in a claypan soil, Missouri, USA. As it was in surface runoff water, the annual FWM PP concentrations in tile drainage water were greater than DRP concentrations. In 75% of samples analysed, PP accounted for N 50% of TP contained in tile drainage water (Table 3, Fig. 1B), suggesting that PP also dominated total P loss in tile drainage water. This might have stemmed from the preferential flow paths that are well developed in this type of soil (Zhang et al., 2015a, 2015b). There was neither DWM treatment effect nor its interaction with CC on annual FWM PP and TP in tile drainage water in any of the four years. However, CC significantly affected the annual FWM PP and TP concentrations in tile drainage water in 2001–2002, as well as the four-year averages. The four-year averages of FWM PP and TP concentrations in tile drainage water with CC across both DWM treatments

Please cite this article as: Zhang, T.Q., et al., Drainage water management combined with cover crop enhances reduction of soil phosphorus loss, Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.02.025

T.Q. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

5

Fig. 1. Flow-weighted mean (FWM) dissolved reactive P (DRP), particulate P (PP) and total P (TP) concentrations (mg P L−1) in surface runoff and tile drainage water as affected by drainage water management (controlled drainage with sub-irrigation, CDS, vs. regular free tile drainage, RFD) and cover crop (no cover crop, NCC, vs. cover crop, CC), respectively, in a clay loam soil, Ontario, Canada.

were reduced by 21% (0.315 mg PP L−1 for CC vs. 0.40 mg PP L−1 for NCC) and 17% (0.54 mg TP L−1 for CC vs. 0.65 mg TP L−1 for NCC), respectively, relative to NCC. Some studies suggested that cover crops may even increase dissolved P and TP concentrations in leachate, due to plant destruction and release of P that is susceptible for leaching during freezing-thawing cycles (e.g. Torstensson et al., 2006), a typical winter scenario in Canada. Uusi-Kämppä (2007) commented that plants exposed to frost damage the membranes that surround the protoplasm of plant cells and released large amounts of P. Torstensson et al. (2006) observed higher P concentration in leachate from a soil with catch crops relative to the

bare soil, which was explained by the P from frozen and thawed cover crop plant materials. However, our observations indicated that it was the DWM, rather than CC, that significantly affected the overall FWM DRP concentration in tile drainage water over the four-year period. The results suggest that P contained in winter wheat CC had limited contribution to DRP concentration in tile drainage water, while the CC inversely resulted in reductions in the overall averages of PP and TP concentrations in tile drainage water during the four-year period. The beneficial roles of CC playing on PP and TP reduction can be explained as the reduced downward movement of soil particles and the subsurface soil erosion (Langdale et al., 1991).

Please cite this article as: Zhang, T.Q., et al., Drainage water management combined with cover crop enhances reduction of soil phosphorus loss, Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.02.025

6

T.Q. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

Table 3 Annually flow-weighted means (FWM) of dissolved reactive P (DRP), particulate P (PP) and total P (TP) concentrations (mg P L−1 yr−1) in tile drainage water as affected by drainage water management (controlled drainage with sub-irrigation, CDS, vs. regular free drainage, RFD) and cover crop (no cover crop, NCC, vs. cover crop, CC) on a clay loam soil, ON, Canada. Year

CDS-NCC

CDS-CC

RFD-NCC

RFD-CC

P L−1)a 0.41 (0.01) 0.09 (0.03) 0.25 (0.05) 0.14 (0.02) 0.20 (0.03)

0.35 (0.13) 0.10 (0.06) 0.21 (0.05) 0.17 (0.01) 0.18 (0.05)

FWM PP concentration in tile drainage water (mg P L−1) 1999–2000 0.45 (0.14) 0.24 (0.03) 0.25 (0.02) 2000–2001 0.40 (0.03) 0.47 (0.03) 0.42 (0.02) 2001–2002 0.52 (0.09) 0.30 (0.01) 0.58 (0.19) 2002–2003 0.28 (0.03) 0.15 (0.04) 0.23 (0.05) Average 0.40 (0.02) 0.32 (0.01) 0.40 (0.03)

0.22 (0.06) 0.40 (0.05) 0.32 (0.02) 0.24 (0.09) 0.31 (0.05)

FWM TP concentration in tile drainage water (mg P L−1) 1999–2000 0.88 (0.07) 0.73 (0.11) 0.80 (0.04) 2000–2001 0.55 (0.05) 0.60 (0.01) 0.59 (0.01) 2001–2002 0.75 (0.12) 0.46 (b0.01) 0.87 (0.24) 2002–2003 0.47 (0.02) 0.38 (0.06) 0.40 (0.02) Average 0.64 (0.05) 0.52 (b0.01) 0.66 (0.06)

0.73 (0.04) 0.60 (b0.01) 0.56 (0.03) 0.44 (0.08) 0.56 (0.01)

FWM DRP concentration in tile drainage water (mg 1999–2000 0.33 (0.06)a 0.33 (0.07) 2000–2001 0.05 (0.01) 0.06 (0.01) 2001–2002 0.21 (0.03) 0.13 (0.01) 2002–2003 0.13 (0.01) 0.19 (0.01) Average 0.16 (0.03) 0.15 (0.01)

Table 4 Annual (Nov. 01–Oct 31) losses (kg ha−1 yr−1) of dissolved reactive P (DRP), particulate P (PP) and total P (TP) in surface runoff water as affected by drainage water management (controlled drainage with sub-irrigation, CDS, vs. regular free tile drainage, RFD) and cover crop (no cover crop, NCC, vs. cover crop, CC) in a clay loam soil, ON, Canada. Year

CDS-NCC

CDS-CC −1

DRP loss in surface runoff water (kg P ha yr 1999–2000 0.29 (0.11) 0.17 (0.11) 2000–2001 0.32 (0.13) 0.11 (0.07) 2001–2002 0.34 (0.11) 0.19 (0.14) 2002–2003 0.26 (0.08) 0.21 (0.01) Average 0.30 (0.11) 0.17 (0.08)

RFD-NCC

RFD-CC

0.21 (0.03) 0.18 (0.05) 0.24 (0.07) 0.19 (0.06) 0.20 (0.06)

0.38 (0.17) 0.19 (0.08) 0.18 (0.08) 0.15 (0.02) 0.22 (0.09)

−1 a

)

PP loss in surface runoff water (kg 1999–2000 0.46 (0.08) 2000–2001 0.79 (0.13) 2001–2002 0.99 (0.09) 2002–2003 0.18 (0.09) Average 0.61 (0.10)

P ha−1 yr−1) 0.14 (0.02) 0.29 (0.01) 0.26 (0.03) 0.21 (0.01) 0.23 (b0.01)

0.26 (0.08) 0.45 (0,21) 0.58 (0.16) 0.23 (0.01) 0.38 (0.11)

0.17 (0.10) 0.31 (0.11) 0.45 (0.23) 0.17 (0.05) 0.27 (0.10)

TP loss in surface runoff water (kg 1999–2000 0.90 (0.01) 2000–2001 1.27 (0.07) 2001–2002 1.38 (0.05) 2002–2003 0.51 (0.01) Average 1.02 (0.03)

P ha−1 yr−1) 0.41 (0.05) 0.48 (0.05) 0.50 (0.12) 0.45 (0.07) 0.46 (0.05)

0.64 (0.25) 0.76 (0.32) 0.85 (0.34) 0.46 (0.01) 0.68 (0.23)

0.67 (0.29) 0.57 (0.23) 0.65 (0.33) 0.37 (0.02) 0.56 (0.21)

Statistical significance (P N F)b Statistical significance (P N F)b 1999–2000

Average

FWM DRP concentration in tile drainage water DWM n.s. * * CC n.s. n.s. * DWM CC* n.s. n.s. n.s.

n.s. * n.s.

* n.s. n.s.

FWM PP concentration in tile drainage water DWM n.s. n.s. n.s. CC n.s. n.s. * DWM CC* n.s. n.s. n.s.

n.s. n.s. n.s.

n.s. * n.s.

FWM TP concentration in tile drainage water DWM n.s. n.s. n.s. CC n.s. n.s. * DWM CC* n.s. n.s. n.s.

n.s. n.s. n.s.

n.s. * n.s.

b

2001–2002

1999–2000 2002–2003

a

2000–2001

Numbers in parentheses are standard errors (n = 2). *, significant at P ≤ 0.05 level; n.s., not significant at P ≤ 0.05 level.

3.3. Phosphorus losses in surface runoff and tile drainage water 3.3.1. Dissolved reactive P, PP and TP losses in surface runoff water The annual DRP loss in surface runoff water was neither significantly affected by DWM nor by overwinter CC in three of the four years (Table 4). Nevertheless, the annual DRP loss in surface runoff water was significantly influenced by both DWM and CC in 2001–2002, and interactively affected by DWM and CC in both 2000–2001 and 2001– 2002. The resultant four-year average of DRP loss in surface runoff water was affected by CC, as well as its interaction with DWM. The treatment combination of CDS-CC had the lowest annual DRP and cumulative DRP losses (0.68 kg P ha−1) in surface runoff water over the fouryear period amongst all of the four treatment combinations (Table 4, Fig. 2A). This would have primarily been attributed to the reduced runoff flow volume induced by cover crop (Table 1), as the corresponding DRP concentrations in surface runoff water were largely comparable in all of the treatments (Table 2). In spite of the lower surface DRP loss observed in CC than in NCC in this study (0.20 kg P ha−1 yr−1 for CC vs. 0.25 kg P ha−1 yr− 1 for NCC, averaged across the two DWM systems), other studies reported that overwinter CC increase DRP loss in runoff under repeated freezing-thawing events. Sharpley and Smith (1991) found that soluble P concentrations in surface runoff from two watersheds under clean-tilled peanuts were greater in the presence of winter cover crops than fallow. Bechmann et al. (2005) reported that

2001–2002

2002–2003

Average

DRP loss in surface runoff water DWM n.s. n.s. CC n.s. n.s. DWM CC* n.s. *

* ** *

n.s. n.s. n.s.

n.s. * *

PP loss in surface runoff water DWM * n.s. CC * * DWM CC* * *

n.s. * *

n.s. n.s. n.s.

n.s. * *

TP in surface runoff water DWM n.s. CC n.s. DWM CC* *

n.s. * *

* n.s. n.s.

n.s. * *

a b

2000–2001

* * **

Numbers in parentheses are standard errors (n = 2). *, significant at P ≤ 0.05 level; n.s., not significant at P ≤ 0.05 level.

repeated freezing and thawing significantly increased water-extractable P from catch crop biomass and resulted in significantly elevated concentration of dissolved P in runoff (9.7 mg P L−1) compared with bare soils (0.14 mg P L−l), under identical simulation rain and hydrological regime with rainfall intensity of 3.0 cm h−1. The annual PP and TP losses in surface runoff water were also interactively influenced by DWM and CC over the four-year period, except for 2002–2003 during which neither individual effect of DWM (PP only) and CC nor the interaction between the two factors was observed (Table 4). The four-year average of annual PP loss in surface runoff water was 29% less under CC than NCC with RFD, while with CDS it was 62% less under CC than NCC. Consequently, the four-year average of annual TP loss in surface runoff water under CC was reduced by 18% relative to NCC with RFD, while with CDS it was reduced by 55% under CC compared to NCC. Clearly, the combination of CC with CDS was particularly effective in reducing PP and TP losses in surface runoff water. This was plausibly explainable, as cover crops that provide canopy soil coverage beyond the normal growing season and protect soil aggregates from the impacts of raindrops, reducing soil detachment and aggregate breakdown that would otherwise pose P loss, PP in particular, along with surface runoff and erosion (Dabney et al., 2001). This effect would have been further enhanced with CDS that increased and maintained soil moisture content. The results were in agreement with Sharpley and Smith (1991) who indicated that winter wheat cover crop dramatically reduced PP and TP losses in surface runoff water

Please cite this article as: Zhang, T.Q., et al., Drainage water management combined with cover crop enhances reduction of soil phosphorus loss, Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.02.025

T.Q. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

7

Fig. 2. Cumulative dissolved reactive P (DRP), particulate P (PP) and total P (TP) losses (kg P ha−1) in surface runoff and tile drainage water as affected by drainage water management (controlled drainage with sub-irrigation, CDS, vs. regular free tile drainage, RFD) and cover crop (no cover crop, NCC, vs. cover crop, CC), respectively, in a clay loam soil, Ontario, Canada.

from two watersheds by up to 85 and 84%, respectively, compared to no cover crop. On the other hand, the four-year average and cumulative PP and TP losses in surface runoff water under CDS-NCC were markedly higher than those with RFD-CC, RFD-NCC, and CDS-CC (Table 4 and Fig. 2A), resulted from the extremely great surface runoff flow volumes produced (Table 1). This suggests that use of CDS system without overwinter cover crop would pose high environmental risk of P loss in surface runoff water.

3.3.2. Dissolved reactive P, PP and TP losses in tile drainage water There were large variations in subsurface tile drainage DRP loss, ranging from 0.06 kg P ha− 1 yr−1 for CDS-NCC in 1999–2000 to 0.55 kg P ha−1 yr−1 for RFD-NCC in 2002–2003 (Fig. 2, Table 5). This was further evidenced by the inconsistent effects of DWM and CC and their interaction amongst the years during the four-year study period (Table 5). There was neither significant individual effect of DWM and CC nor their interaction on annual DRP loss in tile drainage water in

Please cite this article as: Zhang, T.Q., et al., Drainage water management combined with cover crop enhances reduction of soil phosphorus loss, Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.02.025

8

T.Q. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

Table 5 Annual (Nov. 01–Oct 31) losses (kg P ha−1 yr−1) of dissolved reactive P (DRP), particulate P (PP) and total P (TP) in tile drainage water as affected by drainage water management (controlled drainage with sub-irrigation, CDS, vs. regular free tile drainage, RFD) and cover crop (no cover crop, NCC, vs. cover crop, CC) in a clay loam soil, ON, Canada. Year

CDS-NCC

CDS-CC

RFD-NCC

RFD-CC

0.40 (0.08) 0.19 (0.08) 0.55 (0.15) 0.24 (0.05) 0.34 (0.09)

0.28 (0.12) 0.24 (0.15) 0.51 (0.15) 0.33 (0.07) 0.34 (0.12)

PP loss in tile drainage water (kg P ha−1 yr−1) 1999–2000 0.33 (0.02) 0.23 (0.06) 2000–2001 0.50 (0.05) 1.15 (0.19) 2001–2002 0.64 (0.12) 0.63 (0.02) 2002–2003 0.36 (0.07) 0.26 (0.07) Average 0.46 (0.05) 0.57 (0.05)

0.24 (0.03) 0.88 (0.03) 1.27 (0.49) 0.37 (0.03) 0.69 (0.13)

0.17 (0.04) 0.98 (0.09) 0.75 (b0.01) 0.42 (0.09) 0.58 (0.05)

TP loss in tile drainage water (kg P ha−1 yr−1) 1999–2000 0.70 (0.14) 0.70 (0.22) 2000–2001 0.70 (0.08) 1.47 (0.27) 2001–2002 0.93 (0.16) 0.99 (0.07) 2002–2003 0.60 (0.07) 0.64 (0.09) Average 0.73 (0.11) 0.95 (0.12)

0.77 (0.10) 1.25 (0.11) 1.88 (0.64) 0.66 (0.04) 1.14 (0.22)

0.58 (0.05) 1.47 (0.05) 1.34 (0.16) 0.79 (0.02) 1.04 (0.06)

DRP loss in tile drainage water (kg 1999–2000 0.28 (0.12)a 2000–2001 0.06 (0.01) 2001–2002 0.25 (0.04) 2002–2003 0.17 (0.02) Average 0.19 (0.05)

−1

−1 a

P ha yr ) 0.32 (0.11) 0.14 (0.05) 0.29 (0.05) 0.32 (0.02) 0.27 (0.05)

Statistical significance (P N F)b 2001–2002

2002–2003

Average

DRP loss in tile drainage water DWM n.s. n.s. CC n.s. * DWM CC* n.s. *

1999–2000

** * *

n.s. * n.s.

* n.s. *

PP loss in tile drainage water DWM n.s. n.s. CC n.s. * DWM CC* n.s. n.s.

n.s. n.s. n.s.

n.s. n.s. *

n.s. n.s. n.s.

TP in tile drainage water DWM n.s. CC n.s. DWM CC* n.s.

** n.s. n.s.

n.s. n.s. n.s.

* n.s. n.s.

a b

2000–2001

* n.s. *

Numbers in parentheses are standard errors (n = 2). *, significant at P ≤ 0.05 level; n.s., not significant at P ≤ 0.05 level.

1999–2000. The insignificant treatment effects in 1999–2000 could have been caused by the significantly low but largely similar flow volumes (Table 1) as well as the DRP (Table 3) contained in tile drainage water at the early stage of establishment in soil porosity and hydraulic conductivity systems. However, there were significant effects of DWM in 2001–2002 and of CC in three of the four years, as well as particularly the interaction between DWM and CC in two of the four years (i.e. 2000–2001 and 2001–2002) on DRP loss in tile drainage water (Table 5). The resultant four-year average and the cumulative DRP losses in tile drainage water were affected by DWM as well as the interaction between DWM and CC (Table 5, Fig. 2B). Dissolved reactive P loss in tile drainage water with RFD averaged across the four years remained unchanged, regardless of the winter wheat CC treatments. With CDS, DRP loss in tile drainage water decreased by 44% under NCC and by 21% under CC, compared with those values in the corresponding cover crop treatment with RFD (Table 5). The effect of CDS on reducing subsurface DRP loss was reported previously in our earlier study (Tan and Zhang, 2011). Nash et al. (2015) also reported that the controlled tile drainage system greatly reduced ortho-P loss in tile water (36 g ha−1) by 80%, as compared with regular free tile drainage (180 g ha−1) on a Claypan soil in Missouri. The greater reduction in tile drainage DRP loss under NCC than CC with CDS over RFD indicates that the role of CDS in mitigation of DRP loss had largely been compromised by CC. The compensation effect of CC over NCC on tile drainage DRP loss with CDS was primarily attributed to the greater tile drainage flow discharge (Table 1), as the DRP concentrations were largely similar (Table 3). It

was reported on the same site that CC with CDS improved markedly soil hydraulic conductivity that would explain the increased tile drainage flow volume relative to NCC (Drury et al., 2014). The annual tile drainage PP losses were affected by CC in 2000–2001 and by the interaction between CC and DWM in 2002–2003, while there were no treatment effects, including CC, DWM, and the interaction between CC and DWM, in any of the other two years (Table 5). Winter wheat CC averaged across the two DWM treatments increased PP loss in tile drainage water by 126% relative to NCC in 2000–2001, predominately caused by the increases in flow discharge (Table 1). The annual PP loss in tile drainage water was less under CC than NCC with CDS, but with RFD it remained similar between CC and NCC in 2002–2003. Overall, however, the four-year average and cumulative PP losses in tile drainage water were not significantly different amongst all treatments (Table 5). In addition, there were marked variations in tile drainage PP losses amongst years during the four-year study period. For instance, the PP loss in tile drainage water averaged across all treatments ranged from 0.24 kg P ha−1 yr−1 in 1999–2000 to 0.88 kg P ha−1 yr−1 in 2000–2001. The regression analysis indicates that annual precipitation was the driving force for PP loss in tile drainage water (PP loss = − 0.497 + 0.0014 ∗ annual precipitation, R2 = 0.96; P = 0.021). The annual tile drainage TP loss varied depending on DWM and its interaction with CC in 2000–2001 and on DWM in 2001–2002 (Table 5). Compared with NCC, CC increased annual TP loss in tile drainage water by 110% with CDS and by 17.6% with RFD in 2000–2001. This confirms the results reported by Riddle and Bergstrom (2013) who, using an indoor lysimeter rainfall simulation study, showed that PO4-P and TP leaching losses with cover crop were 5.67 and 6.55 times of those with NCC in a clay soil. With the significant effects on annual tile drainage TP loss in 2000–2001 and 2001–2002, the treatment effects on overall four-year average and the cumulative TP losses in tile drainage water were determined by DWM (Table 5, Fig. 2B). Compared with the RFD, a 23% reduction in tile drainage TP loss was observed with CDS systems over the four-year period. This is in agreement with the results reported by Evans et al. (1995) and Tan and Zhang (2011). By summing up P losses in both surface runoff and tile drainage water, the total soil DRP loss was largely comparable amongst the four treatment combinations of the study (Fig. 3). However, the soil total PP and total TP losses were considerably lower under CC than NCC, regardless of DWM. Averaged across the two DWM systems (i.e. CDS and RFD), soil total PP loss decreased by 23% under CC, relative to NCC, over the four-year study period. In combination with the CDS systems, the total TP loss with CC was the least amongst all of the four treatment combinations, followed by RFD-CC, CDS-NCC, and RFD-NCC. Compared with RFD-NCC, the currently most popular practice in the Lake Erie watershed, the total TP loss was reduced by 23% with CDSCC, a new system investigated in this study. The results suggest that the CDS provides additional benefit for reducing TP loss over CC, and the CDS-CC system can be a highly effective practice that comprehensively reduces soil TP loss in pathways of both surface runoff and tile drainage. Historically, most researches have focused on P losses in surface processes including runoff and erosion, because subsurface P loss was often deemed to be negligible (King et al., 2014). Our observations indicated P losses in tile drainage water contribute greatly to P loadings from agricultural soils to water resource by accounting for 39–63, 43–71 and 42–67% of total DRP, PP and TP, respectively (Fig. 3). This supports the conclusion generated previously that P is transportable through downward moving processes and can represent a significant portion of the total amount of P loading to water resource (Zhang and Mackenzie, 1997; Sims et al., 1998; Smith et al., 2014). The combination of CDS with CC appears a reasonably explainable beneficial management practice, as CDS was effective in reduction of tile drainage P loss (e.g. CDS-NCC vs. RFD-NCC TP loss in tile drainage water, Fig. 3), and the CC was effective in reduction of surface runoff P loss (e.g. CDS-CC vs. CDS-NCC TP loss in surface runoff water, Fig. 3), as well as a highly

Please cite this article as: Zhang, T.Q., et al., Drainage water management combined with cover crop enhances reduction of soil phosphorus loss, Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.02.025

T.Q. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

9

Fig. 3. Total (the sum in both surface runoff and tile drainage water) dissolved reactive P (DRP), particulate P (PP) and total P (TP) losses (kg P ha−1) as affected by drainage water management (controlled drainage with sub-irrigation, CDS, vs. regular free tile drainage, RFD) and cover crop (no cover crop, NCC, vs. cover crop, CC), respectively, in a clay loam soil, Ontario, Canada. Error bars are standard errors of total (surface runoff + tile drainage) P; Stacked bars labelled with the same lowercase letter are not significantly different of the total (surface runoff + tile drainage) DRP, PP or TP loss, according to the Tukey's HSD test at P ≤ 0.05.

feasible practice, as the costs for modification (i.e. adding up a “riser”) of the existing drainage system for water table control and sub-irrigation is very minimal. As such, adoption of the CC-DWM technique can be practically meaningful to help meet the 40% P loading reduction goal to the Lake Erie by 2025, as committed by the bilateral governments of Canada and USA.

Acknowledgements Gratitude is expressed for technical assistance provided by B. Hohner, M. Soutani, M.R. Reeb, and K. Rinas from the Harrow Research and Development Center, Agriculture and Agri-Food Canada. Appendix A. Supplementary data

4. Conclusions The study demonstrated clearly the trade-off of tile drainage over surface runoff as to the effects of DWM and CC on field water discharge. Controlled drainage with sub-irrigation decreased tile drainage water flow volume, while it increased surface runoff flow volume. Reversely, winter wheat CC reduced surface runoff water flow volume, while it increased tile drainage flow volume. The total field water discharges were largely identical amongst various treatments. Both pathways of surface runoff and tile drainage must be monitored simultaneously to determine the effectiveness of a management practice on variables that are related to the quantity of field water discharge. Similarly, no significant differences were observed on the fouryear average FWM DRP, PP and TP concentrations in surface runoff water amongst various treatments. In tile drainage water, however, the four-year average FWM DRP concentration decreased with CDS from FRD, while the values for FWM PP and TP decreased with CC from NCC. Winter wheat CC reduced the annual soil total P loss in both surface runoff and tile drainage water due to the reduction in PP loss, although it had no effects on DRP loss. The reduction effect of CC on soil total P loss was further enhanced by CDS. Winter wheat CC in combination with CDS can be recommended as a beneficial management practice to mitigate soil P loss, particularly in the Lake Erie basin where non-point source P loss is currently a major concern to the lake's water quality, while free tile drainage is a highly popular agricultural management practice for improvement of crop production.

Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2017.02.025.

References Ball-Coelho, B., Murry, R., Lapen, D., Topp, E., Bruin, A., 2012. Phosphorus and sediment loading to surface waters from liquid swine manure application under different drainage and tillage practices. Agric. Water Manag. 104, 51–61. Bechmann, M.E., Kleinman, P.J.A., Sharpley, A.N., Saporito, L.S., 2005. Freeze–thaw effects on phosphorus loss in runoff from manures and catch-cropped soils. J. Environ. Qual. 34, 2301–2309. Constantin, J., Mary, B., Laurnet, F., Aubrion, G., Fontaine, A., Kerveillant, P., Beaudoin, N., 2010. Effects of catch crops, no-till and reduced nitrogen fertilization on nitrate leaching and balance in three long-term experiments. Agric. Ecosyst. Environ. 135, 268–278. Dabney, S.M., Delgado, J.A., Reeves, D.W., 2001. Using winter cover crops to improve soil and water quality. Commun. Soil Sci. Plant Anal. 32 (7&8), 1221–1250. Drury, C.F., Tan, C.S., Welacky, T.W., Reynolds, W.D., Zhang, T.Q., Oloya, T.O., McLaughlin, N.B., Gaynor, J.D., 2014. Reducing nitrate loss in tile drainage water with cover crops and water-table management systems. J. Environ. Qual. 43 (2), 587–598. Elmi, A., Burton, D., Gordon, R., Madramootoo, C., 2005. Impacts of water table management on N2O and N2 from a sandy loam soil in southwestern Quebec, Canada. Nutr. Cycl. Agroecosyst. 72, 229–240. Environment Canada, 2004. Canadian Guidance Framework for the Management of Phosphorus in Freshwater Systems. Ecosystem Health: Science-based Solution Report No. 1–8. Cat. No. En1-34/8-2004E. National Guideline and Standards Office. Water Policy and Coordination Directorate. Environment Canada, Ottawa, ON. Evans, R.O., Skaggs, R.W., Gilliam, J.W., 1995. Controlled versus conventional drainage effects on water quality. J. Irrig. Drain. Eng. 121:271–276. http://dx.doi.org/10.1061/ (ASCE)0733-9437(1995)121:4(271). Fisher, M.J., Fausey, N.R., Subler, S.E., Brown, L.C., Bierman, P.M., 1999. Water table management, nitrogen dynamics and yields of corn and soybean. Soil Sci. Soc. Am. J. 63, 1786–1795.

Please cite this article as: Zhang, T.Q., et al., Drainage water management combined with cover crop enhances reduction of soil phosphorus loss, Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.02.025

10

T.Q. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

Joyce, B.A., Wallender, W.W., Mitchell, J.P., Huyck, L.M., Temple, S.R., Brostrom, P.N., Hsiao, T.C., 2002. Infiltration and soil water storage under winter cover cropping in California's Sacramento Valley. Trans. ASAE 45, 315–326. Kaspar, Z.C., Radke, J.K., Laflen, J.M., 2001. Small grain cover crops and wheel traffic effects on infiltration, runoff, and erosion. J. Soil Water Conserv. 56, 160–164. King, K.W., Williams, M.R., Macrae, M.L., Fausey, N.R., Frankenberger, J., Smith, D.R., Kleinman, P.J.A., Brown, L.C., 2014. Phosphorus transport in agricultural subsurface drainage: a review. J. Environ. Qual. http://dx.doi.org/10.2134/jeq2014.04.0163. Langdale, G.W., Blevins, R.L., Karlen, D.L., McCool, D.K., Nearing, M.A., Skidmore, E.L., Thomas, A.W., Tyler, D.D., Williams, J.R., 1991. Cover crop effects on soil erosion by wind and water. In: Hargrove, W.L. (Ed.), Cover Crops for Clean Water. Proceedings of International Conference, Jackson, TN. 9–11 Apr. 1991. Soil and Water Conservation Society of America, Ankeny, IA, pp. 15–22. Murphy, J., Riley, J.P., 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta 27, 31–36. Nash, P.R., Nelson, K.A., Motavalli, P.P., Nathan, M., Dudenhoeffer, C., 2015. Reducing phosphorus loss in tile water with managed drainage in a claypan soil. J. Environ. Qual. 44 (2):585–593. http://dx.doi.org/10.2134/jeq2014.04.0146. OMAFRA, 2002. Agronomy Guide for Field Crops - Publication 811. Queen's Printer for Ontario, Toronto. Reynolds, W.D., Yang, X.M., Drury, C.F., Zhang, T.Q., Tan, C.S., 2003. Effects of selected conditioners and tillage on the physical quality of a clay loam soil. Can. J. Soil Sci. 83, 381–393. Riddle, M.U., Bergstrom, L., 2013. Phosphorus leaching from two soils with catch crops exposed to freeze-thaw cycles. Agron. J. 105, 803–811. SAS, 2006. Statistical Analysis Aystem, Ver. 9.1. SAS Institute, Cary, NC. Sharpley, A.N., Smith, S.J., 1991. Effect of cover crop on surface water quality. In: Hargrov, W.L. (Ed.), Cover Crop for Clean Water. Soil and Water Conservation Society (SWCS), Iowa, USA, pp. 41–49. Sharpley, A.N., Weld, J.L., Beegl, D.B., Kleinman, P.J.A., Gburek, W.J., Moore, P.A., Mullins, G., 2003. Development of phosphorus indices for nutrient management planning strategies in the United States. J. Soil Water Conserv. 58 (3), 137–151. Sims, J.T., Simard, R.R., Joern, B.C., 1998. Phosphorus loss in agricultural drainage: historical perspective and current research. J. Environ. Qual. 27:277–293. http://dx.doi. org/10.2134/jeq1998.00472425002700020006x.

Smith, D.R., King, K.W., Johnson, L., Francesconi, W., Richards, P., Baker, D., Sharpley, A.N., 2014. Surface runoff and tile drainage transport of phosphorus in the Midwestern United States. J. Environ. Qual. http://dx.doi.org/10.2134/jeq2014.04.0176. Tan, C.S., Zhang, T.Q., 2011. Surface runoff and subsurface drainage phosphorus losses under regular free drainage and controlled drainage sub-irrigation systems in southern Ontario. Can. J. Soil Sci. 91, 349–359. Tan, C.S., Drury, C.F., Gaynor, J.D., Welacky, T.W., 1993. Integrated soil, crop and water management system to abate herbicide and nitrate contamination of the Great Lakes. Water Sci. Technol. 28, 497–507. Tan, C.S., Zhang, T.Q., Drury, C.F., Reynolds, W.D., Oloya, T.O., Gaynor, J.D., 2007. Water quality and crop production improvement using a wetland reservoir and drainage/ subsurface irrigation system. Can. Water Res. J. 32 (2), 129–136. Torstensson, G., Aronsson, H., Bergström, L., 2006. Nutrient use efficiencies and leaching of organic and conventional cropping systems in Sweden. Agron. J. 98, 603–615. Uusi-Kämppä, J., 2007. Effects of freezing and thawing on DRP losses from buffer zones. In: Heckrath, G., et al. (Eds.), Diffuse Phosphorus Loss: Risk Assessment, Mitigation Options and Ecological Effects in River Basins. The 5th International Phosphorus Workshop (IPW5), Silkeborg, Denmark. 3–7 Sept. 2007. DJF Plant Sci. 130. Faculty Agric. Sci., Aarhus Univ., Tjele, Denmark, pp. 169–172. Wang, Y.T., Zhang, T.Q., Hu, Q.C., Tan, C.S., O'Halloran, I.P., Drury, C.F., Reid, K., Ma, B.L., Ball-Coelho, B., Lauzon, J., Reynolds, W.D., Welacky, T., 2010. Estimating dissolved reactive phosphorus concentration in surface runoff water from major Ontario soils. J. Environ. Qual. 39, 1771–1781. Zhang, T.Q., Mackenzie, A.F., 1997. Phosphorus in zero tension soil solution as influenced by long-term fertilization. Can. J. Soil Sci. 77 (4), 685–691. Zhang, T.Q., Tan, C.S., Zheng, Z.M., Welacky, T.W., Reynolds, W.D., 2015a. Impacts of soil conditioners and water management on phosphorus loss in tile drainage from a clay loam soil. J. Environ. Qual. 44, 572–584. Zhang, T.Q., Tan, C.S., Zheng, Z.M., Drury, C., 2015b. Tile drainage phosphorus loss with long-term consisting cropping systems and fertilization. J. Environ. Qual. 44, 503–511.

Please cite this article as: Zhang, T.Q., et al., Drainage water management combined with cover crop enhances reduction of soil phosphorus loss, Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.02.025