Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Effects of modelling studies on controlled drainage in agricultural land on reduction of outflow and nitrate losses–a meta-analysis

  • Barbara Kęsicka ,

    Roles Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Visualization, Writing – original draft, Writing – review & editing

    barbara.kesicka@up.poznan.pl

    Affiliation Department of Land Improvement, Environmental Development and Spatial Management, Faculty of Environmental and Mechanical Engineering, Poznań University of Life Sciences, Poznań, Poland

    ⨯
  • Rafał Stasik,

    Roles Conceptualization, Methodology, Supervision, Visualization, Writing – review & editing

    Affiliation Department of Land Improvement, Environmental Development and Spatial Management, Faculty of Environmental and Mechanical Engineering, Poznań University of Life Sciences, Poznań, Poland

    ⨯
  • Michał Kozłowski

    Roles Conceptualization, Methodology, Supervision, Writing – review & editing

    Affiliation Department of Soil Science and Land Reclamation, Faculty of Environmental and Mechanical Engineering, Poznań University of Life Sciences, Poznań, Poland

    ⨯

Abstract

A review with meta-analysis of outflow and nitrate loss reduction in controlled drainage (CD) vs conventional, free drainage (FD) was carried out in the study. Since the results of experimental field studies usually cover short periods of data collection, hence in this paper, meta-analyses were based on model studies that usually cover a longer time range. The databases Web of Science and Scopus were searched for eligible English articles, published until December 2020, that describe the quantity and quality of drainage water. The meta-analysis of outflow and nitrate loss reduction in CD vs FD using the mean difference (MD) with a confidence interval (CI) of 95%. The influence of each study was measured through heterogeneity, sensitivity analyses and publication bias using STATISTICA (version 13.3) for all analyses. Of the 107 works identified, 18 were finally included in the analysis based on established criteria required for an appropriate meta-analysis. In general the results indicate a reduction in average drainage outflow of 30.5% (MD = -71.26 mm; 95% CI, -103.49 –-39.04; p = 0.000) in arable land with CD in comparison to FD practice. In the case of nitrate load the reduction was 33.61% and in the drainage water there was lower content in CD practice by an average of 8.36 kg NO3 ha-1year-1 (95% CI, -9.93 –-6.79; p = 0.000). Subgroup analysis of two meta-analyses indicates that the results concerning these associations may vary with the calculated weight for each article, in which the number of years of study had the most significant impact.

Citation: Kęsicka B, Stasik R, Kozłowski M (2022) Effects of modelling studies on controlled drainage in agricultural land on reduction of outflow and nitrate losses–a meta-analysis. PLoS ONE 17(4): e0267736. https://doi.org/10.1371/journal.pone.0267736

Editor: Pramod K. Pandey, University of California Davis, UNITED STATES

Received: September 9, 2021; Accepted: April 13, 2022; Published: April 28, 2022

Copyright: © 2022 Kęsicka et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: The publication was co-financed/financed within the framework of Ministry of Science and Higher Education programme as „Regional Initiative Excellence” in years 2019-2022, Project No. 005/RID/2018/19. This research was funded by the Polish National Centre for Research and Development grant number BIOSTRATEG3/347837/11/NCBR/2017. The funders were not involved in the design of the study, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Drainage systems, comprising ditches or subsurface drains, are important for agricultural production of the world’s cropland. Subsurface tile drainage, i.e. free drainage (FD), is widely practiced in productive agricultural land with poorly drained soils in different parts of the world [1–3]. Benefits of this management practice are increased crop productivity and improved economic returns for crop producers. The consequences of drainage water from drainage network systems have been identified as an important factor contributing to increasing nutrient loads, mainly nitrate-nitrogen (), , total nitrogen (TN) and total phosphorus (TP) into surface receiving waters [4–6]. Globally, it is estimated that the losses of from agricultural systems account for 19% of the total amount of N used worldwide [7].

Tile drainage plays a significant role in soil nitrogen losses from agricultural lands, thus affecting surface and groundwater and leading to deterioration of water quality, especially as pollutants originating from agricultural drainage outflow include sediment, nitrogen and also phosphorus, pesticides, pathogens, salts, trace elements, and dissolved organic carbon [8–13]. This threatens the hydrological environment due to the accumulation of nitrates in water sources and oxygen deficiency [14–16]. This is confirmed by long-term studies carried out in various catchment areas, which indicate that agricultural systems affect nitrate levels in river waters. The research concerns facilities located in other climatic zones, soil properties and different cultivation systems, which allows a greater understanding of nitrate losses to subsurface drainage [17, 18]. This is a particularly important issue from the point of view of the environmental risk related to nitrate losses from agricultural sources. Several management practices have been developed to reduce the nitrate loading from artificially drained agricultural areas, including controlled drainage (CD), controlled tile drainage (CDT), drainage water management (DWM) [19–22], denitrifying bioreactors (DBR) [23–25], free water surface flow constructed wetlands (FWS) [26–28], saturated buffer zones (SBZ) and integrated buffer zones (IBZ) [29–31], and drainage water recycling [32].

CD is a conservation practice, to artificially raise and adjust the level to which the water table in a tile-drained field is allowed to rise, using a water control structure near the outlet of a drain to adjust the effective outlet elevation. Also, it can reduce nutrient loss during wet periods by storing more water in the field. This drainage water management affects the hydrological changes in the field cycle, which, depending on the location and weather conditions, increases some or all of the following flow components: root zone water storage, seepage, surface drainage, plant uptake and evaporation [33, 34]. Another benefit is reducing losses of nitrogen (N) from agricultural subsurface drained fields to surface waters. The consequence is an increase in groundwater levels, and longer anaerobic conditions are created in the soil, ideal for denitrification [35–37]. This process involves microbial respiration under anaerobic conditions in which nitrates from the ecosystem return reactive nitrogen (Nr) to the atmosphere as N2 and N2O emissions [38]. This practice is one of the mitigation measures targeting nutrient losses from agricultural drainage systems’ water before it enters streams [39].

Most of the research related to CD practice has been tested and widely used in US and Canadian studies for more than two decades [35, 36]. DWM is gaining popularity in other countries of the world due to improved quality of water flowing from drained agricultural fields to surface water. The practice has been tested on research facilities in several countries including Lithuania [40], Sweden [41], Denmark [39], Italy [42], China [43, 44], Iran [45], India [46], Japan [47], and Egypt [48]. Unfortunately, the results of experimental field studies usually include a short measurement period of CD effectiveness. Therefore, on the basis of these measurements, model studies covering a longer time range are carried out.

Among the many studies on the influence of CD on the drainage water quality and quantity, there are disparities in the obtained results. Some of them show a positive effect while others show a negative or no effect of CD on the quantity and quality of drainage water. Hence, in order to comprehensively analyze the effect of CD on the quantity and quality of drainage water, which will be statistically confirmed at the same time, a meta-analysis can be used, especially since there are no such studies. Increasing the implementation of the practice provides more data measured from field studies, and modelled data assessing the impact of CD on the quantity and quality of water in the discharge of tiles from fields or small catchment areas. One of the soil-specific models used to support tile drainage research is DRAINMOD created by Skaggs et al. [34, 49, 50, 51]. DRAINMOD is one of the most widely used hydrologic models to simulate subsurface drainage systems. It simulates surface runoff, infiltration, evapotranspiration, subsurface drainage and seepage from the soil profile in response to given climatological conditions, crop rotation, soil type, and drainage system parameters and management [34]. The program is used to simulate the water table depth, drained outflow or nitrate-nitrogen in the drainage water of drained soils in different parts of the world, for present, future and past data. In addition, there are several models to assess the long-term impact of agricultural management and climate change on crop production and water quality [52]. In recent decades, the ability of water quality models in the root zone to study the impact of agricultural management practices on water quality and crop growth in places significantly different in terms of climatic and pedological conditions has been widely created and improved.

Due to the shortage of previous studies comprehensively analyzing model data on the effect of CD on the quantity and quality of drainage water, including statistical validation, we applied a meta-analysis for this purpose. In this study, we focused on comparing the results obtained for DRAINMOD model studies under CD vs FD conditions and its effect on reduction of outflow and nitrate losses of drained agricultural land. Therefore, we used meta-analyses to synthetically and also statistically indicate the effectiveness of CD use in quantitative and qualitative aspects of drainage outflow.

Materials and methods

Search strategy

This meta-analysis was conducted (18 January 2021) by an advanced search in the Web of Science (WoS) by the ISI Web of Knowledge published by Thomson Reuters and Scopus; the owner of the database is Elsevier. The search scope was developed using the symbol "*" and the advanced search was performed for words or instructions created by logical operators. Keywords searched included (“controlled drainage” AND “drainmod”) until December 31, 2020. The scientific documents are included in the database Web of Science Core Collection: Science Citation Index Expanded (SCI-EXPANDED). The meta-analysis was limited to articles published in English.

Inclusion and exclusion criteria

For further analysis, several articles were rejected after an initial review of the summaries and titles. A model study was selected considering CD and FD and their impact on the amount of outflow and nitrogen loss from fields. Articles from field studies, as well as those which provided insufficient data to conduct reliable meta-analysis or were reproduced publications, were also excluded.

Data extraction

A standard protocol for obtaining data and information on the basis of which eligible articles were established was prepared after consultation. The data collected included the name of the article, the name of the first author, the year of publication, the amount of outflow, the loss of N, the number of years of modelling, and additionally the texture of the soil, the crop, the amount of annual rainfall, the average temperature, and depth and spacing of the drainage. Disputes regarding the eligibility of the article for meta-analysis were resolved through a group discussion. The information and data were entered in the standard Microsoft Excel data extraction form (S1 Data).

Weights for individual articles in meta-analysis

In comparison to field research modelling of CD and FD practices gives the possibility of analyzing many combinations of different tile drainage and soil parameters, weather conditions and future climate predictions as well as CD variants and so on. Hence the derived conclusions of model simulations after validations can result from greater amounts of data than those obtained from field research. Pooled mean difference weighting by a function of sample size was used in meta-analysis weights [53, 54]. The following calculation is the sample weight for an individual study: (1) where ntx,i = total number of modelled variants of CD, nctr,i = total number of modelled variants of FD. The number of variants ntx,i and nctr,i, represents the result of multiplying the number of: years, locations, experimental plots, drainage spacing, drainage depth, variants of CD application as well as soil types (S1 Data).

Data synthesis and analysis

On the basis of the data from each article, the average value of outflows and nitrogen losses for CD and FD were calculated according to the number of variants. A mean difference (MD) with a confidence interval (CI) of 95% was used as a standard for measuring the relationship between CD and FD. Statistical heterogeneity was assessed in studies using Q and I2 and L’Abbé plots. Q is Cochran’s heterogeneity statistic usually computed by summing the squared deviations of each study’s estimate from the overall meta-analytic estimate, weighting each study’s contribution in the meta-analysis. I2 explains the percentage of total variation across studies that is due to heterogeneity rather than chance, and it is a helpful estimate to investigate the causes and type of heterogeneity. Based on the Q value, T2 is determined, defined as the estimator of the variance of the actual effects [55, 56]. A cumulative meta-analysis was performed taking into account the chronological effect of individual studies on the calculated overall effect and the error in its estimation varying over time after accounting for subsequent publications. A sensitivity analysis was performed for the change in cumulative effect due to the exclusion from the meta-analysis of individual studies as this will change the standard error. Wherever the results were heterogeneous, a random-effects model was used in the meta-analysis. In this selected statistical model, we assumed that the true effect may vary depending on the study. We chose this variant because the studies differed significantly in depth and drainage spacing parameters, duration simulation, modelling accuracy, soil and climate context, and cultivated plant species used. Subgroup analyses were carried out for possible causes of heterogeneity. Integrated estimates and associated CI of 95% were evaluated using forest plots as visualizations. Publication bias was evaluated qualitatively using funnel plots, Egger regressions, and the Begg–Mazumdar correlation test [57, 58].

Values of p < 0.05 were considered as valid for heterogeneity tests. For statistical analyses, STATISTICA software (version 13.3), the Set Plus module with meta-analysis and meta-regression, was used.

Results

Selected articles

In an electronic search of the literature, 107 potential articles were identified; 43 were found in the WoS database and 64 articles in the Scopus database. Following the completion of the preliminary screening of abstracts and titles, 57 articles were excluded on the basis of the inclusion criteria (8 duplicate articles and 49 on the basis of unrelated titles and summaries) and 50 articles remained for a full review of the text. In the secondary study and after the full-text review, a further 32 articles were excluded; they were rejected due to missing information to create a meta-analysis data statement for outflows and nitrate losses. Ultimately, eighteen studies after these exclusions were selected for the final analysis (Fig 1).

thumbnail
Download:
Fig 1. Flowchart of the literature search.

https://doi.org/10.1371/journal.pone.0267736.g001

Description of selected articles

The publications used in this meta-analysis were published between 1995 and 2020. Most of the articles were published in Agricultural Water Management (9 papers), Transactions of the ASABE (2 papers). One article each was published in 7 journals on the effects of CD on drainage water quantity and quality using DRAINMOD modeling results: European Journal of Agronomy, Geoderma, Irrigation and Drainage, Journal of Environmental Quality, Journal of Soil and Water Conservation, Sustainability, and Transactions of the ASAE. Of the selected articles, there are 10 (Skaggs et al. [59], Breve et al. [60], El-Sadek et al. [61], Ma et al. [52], Salazar et al. [62], Luo et al. [63], Skaggs et al. [64], Negm et al. [65], Negm et al. [66], Youssef et al. [67]) that are included in two meta-analyses of subsurface drainage outflow reduction and nitrate in drainage outflow loss. Subsequently, 4 publications (Skaggs et al. [68], Pease et al. [69], Sojka et al. [70], Singh et al. [71]) concerning outflows and 4 publications (Breve et al. [72], Singh et al. [73], Ale et al. [74], Ale et al. [75]) related to nitrogen losses were included in the relevant analyses. Most studies concern the modelling of outflows from research facilities located in the United States (83%), while the remaining 3 studies concern European countries: Belgium, Sweden and Poland (Fig 2). Table 1 presents the characteristics for selected articles regarding authors and year of publication, country, soil texture, years for which CD use was modelled, and depth and spacing of drains.

thumbnail
Download:
Fig 2. Countries to which the articles used in the record screening of articles relate.

https://doi.org/10.1371/journal.pone.0267736.g002

thumbnail
Download:
Table 1. Characteristics and information of studies evaluated in the meta-analysis.

https://doi.org/10.1371/journal.pone.0267736.t001

Weights for individual articles

Due to differences in the basic characteristics of meteorological, soil and drainage networks introduced into models in different selected articles, a weight was established for each article. There is a significant difference in the years of modelling drainage in articles, from 1 to 90 years; it affected in some cases the final value of the weight. In addition, in some cases, the final assessment was also influenced by significant values in other parameters such as locations (Youssef et al. [67]), plots (Ma et al. [52]), spacing (Breve et al. [72]), and soils (Ale et al. [75]). Luo et al. [63] modelled for different spacing, others to determine outflows and others to determine nitrate loss. The range of weights for accepted articles is from 1 to 630. Three articles have the smallest weight values (Breve et al. [60], Salazar et al. [62], Negm et al. [66]) whose value reflects the number of years used for modelling. The largest weight values are published (Luo et al. [63], Youssef et al. [67]) with the largest number of modelling years and more variants of other parameters. The values of NCD and NFD for selected articles were used to determine the number of meta-analyses performed for the reduction of outflow drains and nitrate losses with outflow.

Meta-analysis of drainage outflow and nitrate reduction

Two meta-analyses for subsurface drainage outflow reduction and nitrate in drainage water by CD versus FD were initially performed (Fig 3A and 3B). The first meta-analysis therefore used a random effect model to demonstrate how controlled drainage affects the reduction of drainage compared to conventional drainage. The overall effect from all articles analyzed indicates that the use of CD statistically significantly reduces drainage outflow by 71.26 mm per year. In absolute terms, this represents 30.5% of FD drainage outflow. All publications used in the meta-analysis indicated that the use of CD reduces outflow compared to FD. In this list, 5 publications were statistically insignificant (Fig 3A). Considering the year the study was published (Fig 4A), in the first period 1995–2007, when the first 5 publications were published, there was no statistically significant effect of CD on reducing drainage outflow. It is only with the inclusion of the 2009 study by Salazar et al. [62] that the effect was statistically significant and maintained after the inclusion of subsequent studies. Since this study, the 95% confidence intervals have narrowed significantly for the combined effect. The last 3 studies had high variability of results, which influenced the expansion of the 95% confidence interval for the combined effect.

thumbnail
Download:
Fig 3.

Meta-analysis for all articles: (a) subsurface drainage outflow reduction by CD versus FD; (b) reduction of nitrate load in drainage outflow by CD versus FD.

https://doi.org/10.1371/journal.pone.0267736.g003

thumbnail
Download:
Fig 4.

Cumulative meta-analysis: (a) subsurface drainage outflow reduction by CD versus FD; (b) reduction of nitrate load in drainage outflow by CD versus FD.

https://doi.org/10.1371/journal.pone.0267736.g004

The second meta-analysis also used a random-effects model to demonstrate how CD affects nitrate losses at low tide compared to FD. The findings of this analysis indicate a higher annual nitrate load in drainage waters in FD than in CD. The overall effect was statistically significant (Fig 3B) and showed an average reduction in nitrate losses of 8.36 kg NO3·ha-1year-1 (MD = -8.36; 95% CI: -9.93 –-6.79, p < 0.05) and this represents 33.61% of nitrate FD outflow. It is also apparent that in 4 out of 14 studies the results were statistically insignificant. The results of the cumulative meta-analysis indicate that from 1995 [59], each addition of subsequent studies (except study [72]) had a statistically significant impact on the summary effect for the cumulative analysis (Fig 4B). In general, each further addition of a literature item resulted in a significant narrowing of the 95% confidence interval for the summary effect.

Based on the heterogeneity analysis (Fig 5A), there was significant variation between studies analyzing the effect of CD on drainage outflow reduction (I2 = 99.89%; p = 0.000). The variance of T2 effects was 3239.08 and represented 99.89% of the observed variability. The relationship between the drainage outflow effect in CD and FD practice modelling was shown in the L’Abbé plot (Fig 5A). The solid line corresponds to the equilibrium level (MD = 0.00) and the dashed line to the overall effect determined in the meta-analysis (MD = -71.26 mm). The size of the markers is proportional to the contribution of a given study to the meta-analysis. A significant dispersion of points around the dotted line indicates high variability in the results of the research. The results of the meta-analysis indicate a positive correlation between drainage outflow in FD versus CD. The most far removed from the cumulative effect of the meta-analysis is the publication by Peace et al. [69], indicating a significantly higher average outflow value for FD and at the same time the greatest reduction in drainage outflow using CD. In three literature items [60, 61, 70, 71], the results indicated a small effect of CD on outflow reduction, of which two results of items [60 and 61] were not statistically significant. The outcome of study [68] is not much different from the combined effect of meta-analysis but has the largest drainage outflow values for both FD and CD. Analyzing the points (literature items except [69]) located within the solid and dashed lines, one can point out a certain regularity: the points are closer to the solid line the smaller the FD values are, while they move towards the dashed line as the FD increases.

thumbnail
Download:
Fig 5.

Heterogeneity analysis for all articles based on L’Abbé plot: (a) subsurface drainage outflow reduction by CD versus FD; (b) nitrate load in drainage outflow by CD versus FD.

https://doi.org/10.1371/journal.pone.0267736.g005

Similarly, the results of the heterogeneity analysis were obtained for a meta-analysis of nitrogen losses, because Q is 1683.26 (df = 13) and the p value is less than 0.05. The variance of actual T2 effects is equal to 6.58 and represents 99.23% of the observed variability (I2 = 99.23%). In this case, the MD values of the individual tests are higher than the equilibrium level (MD = 0.00) (Fig 5B). In general, for nitrate in drainage outflow, the points (literature items) have less distribution around the dashed line than for water outflow. In addition, results from publications showing the lowest nitrate losses with both FD and CD were closer to the solid line (item [60, 62]), while those showing the highest nitrate losses with FD and CD (item [64, 65, 73]) were near the dashed line (except item [61]).

Sensitivity and subgroup meta-analysis

A meta-analysis of subsurface drainage outflow reduction in current studies and sensitivity analysis were also conducted. After the sequential removal of each study from the analysis, the average overall effect changed. Thus, it was found that the most significant impact on the results was from the research article by Pease et al. [69]. After its exclusion, we would get a slightly larger average effect (MD = -47.48 mm), and the standard error for the overall effect would decrease by about 60%. The article had a strong impact on the results of the meta-analysis. This was the study [69] with the largest absolute reduction in drainage outflow among the articles analyzed (MD = -314.17 mm).

Analyzed subgroups in the current meta-analysis take into account the number of years of model studies to minimize heterogeneity among different studies. Three groups are distinguished on the basis of the product of parameters used to calculate the weights of individual articles. The first includes studies up to 10 values of NCD and NFD, the second from 10 to 100, and the third above 100. Fig 6 shows the results of the analysis in such defined groups. The first group’s overall effect score is statistically insignificant, MD = -75.90 mm (95% CI: -195.70–43.90; p = 0.331) and represents 33.73% of FD outflow. This group includes publications that analyzed the effect of CD on reducing drainage outflow for the fewest number of years. The final results of the overall effect for the second group, MD = -84.89 mm (95% CI: -123.73 –-46.05, p = 0.001) and the third, MD = -70.15 mm (95% CI: -75.23 –-65.08, p = 0.002) were statistically significant. These reductions represent 22.63% and 31.10% respectively. In addition, first and second groups contain publications (3 and 2, respectively) with an insignificant effect. When analyzing the forest plot for the distinguished groups the presented 95% confidence intervals for the group effect decrease from the first cluster to the third group. The closest to this effect is the D value of the third group, and from the selected articles the average difference between CD and FD for Youssef et al. [67] is 82.50 mm. The resulting total meta-analysis effect was most influenced by the three publications of Youssef et al. [67], Luo et al. [63] and Ma et al. [52], respectively 33.70%, 25.74% and 20.22%, in total explaining 79.66% of the results obtained. The smallest impact on the resulting total meta-analysis effect was found for the publications in the first group, two publications each of 0.11% and one of 0.06%. The most important in the first subgroup are two publications of 40% (Salazar et al. [62], Negm et al. [66]). The largest weights in the second subgroup are the two publications of Skaggs et al. [64] (37.88%) and El-Sadek et al. [61] (26.52%). In the third group, the three publications with the highest weights are those of Youssef et al. [67], Luo et al. [63] and Ma et al. [52], respectively 36.51%, 27.88% and 21.91%.

thumbnail
Download:
Fig 6. Meta-analysis in groups of subsurface drainage outflow reduction by CD versus FD.

https://doi.org/10.1371/journal.pone.0267736.g006

Likewise, three article groups were distinguished when analyzing the effect of CD on nitrate load reductions (Fig 7). For the first group, the mean effect score is statistically insignificant, MD = -4.95 kg NO3 ha-1year-1 (95% CI: -12.58–2.67; p = 0.096) representing a 28.18% reduction compared to FD. There was a statistically significant mean effect of CD practice on reduction of nitrate load for the second group, MD = -9.41 kg NO3 ha-1year-1 (95% CI: -16.05 –-2.76, p = 0.000), and the third, MD = -8.27 kg NO3 ha-1year-1 (95% CI: -8.98 –-7.56; p = 0.000), representing 23.19% and 34.59% respectively. This meta-analysis indicates that the overall effect of nitrate load in drainage is lower for CD by 8.36 kg NO3 ha-1year-1 (33.61%) compared to FD. In the second group is the article by Negm et al. [65], with a similar average value of 9.20 kg NO3 ha-1year-1. In the third group there is also a article by Breve et al. [72] with a mean nitrate load reduction effect (8.42 kg NO3 ha-1year-1) similar to that obtained in the meta-analysis. The total effects for the second and third groups are 1.05 kg NO3 ha-1year-1 higher and 0.09 kg NO3 ha-1year-1 lower, respectively, compared to the overall meta-analysis effect. In addition, a test based on Q statistics showed significant differences between the three test groups (p = 0.000). Also, in 4 (three of the first group and one of the second group) of the studies considered, the results were statistically insignificant for group effects. The resulting total meta-analysis effect was influenced the most by four publications: Youssef et al. [67], Ma et al. [52], Ale et al. [74] and Luo et al. [63], respectively 31.65%, 18.99%, 13.19% and 12.66%, together accounting for 76.49% of the results obtained. Here, the publications of the first group had the least impact on the total meta-analysis, two publications each of 0.11% and one of 0.05%, together affecting 0.27% of the final result. Two publications, by Salazar et al. [62] and Negm et al. [66], in the first subgroup are the most important, with 40%. In the second subgroup, the largest weight in the subgroup are three publications: Skaggs et al. [64] (31.15%), Singh et al. [71] (23.36%). In the third group, the two publications with the highest weights are those of Youssef et al. [67] and Ma et al. [52], respectively 34.68% and 20.81%.

thumbnail
Download:
Fig 7. Meta-analysis in groups of nitrate load in drainage outflow by CD versus FD.

https://doi.org/10.1371/journal.pone.0267736.g007

Publication bias

No signs of publication bias in the studies under consideration for two meta-analyses were shown by the funnel graphs presented in the Begg-Mazumdar test (Fig 8A and 8B). The Egger and Begg tests showed no evidence of publication bias for subsurface drainage outflow reduction and nitrate in drainage outflow by CD versus FD, respectively (p = 0.336 for Egger; p = 1.00 for Begg’s and p = 0.143 for Egger; p = 0.293 for Begg’s). In both graphs, a significant number of studies are at the top of the funnel outside the triangle, but they are close to the edge of the funnel. These are the studies with the smaller standard error and therefore have the strongest influence on the outcome of the analysis. In the funnel plot (Fig 8A), one study is outside the perimeter of the triangle (Pease et al. [69]); the only concern is the selection of this study.

thumbnail
Download:
Fig 8. Funnel plot (from Begg-Mazumdar test) for publication bias: (a) subsurface drainage outflow reduction by CD versus FD; (b) nitrate load in drainage outflow by CD versus FD.

https://doi.org/10.1371/journal.pone.0267736.g008

Limitations

The meta-analysis presented above has several limitations. Firstly, there were significant differences in the number of years of outflow and the nitrate reduction modeling for CD and FD between studies, which may affect average values. Hence, meta-analyses were conducted in groups. Secondly, due to insufficient data and statistical information in some studies, statistical values were calculated on the basis of the values given in the publications for each CD and FD variant. Publications were excluded if they did not contain data on the results of model studies. In the case of missing information, selected meta-analysis publications were not verified and consulted with their authors. Thirdly, this meta-analysis used mean outflow values and nitrate load in outflows in modelled studies. However, the input for each model differed among different studies. In addition, in meta-analysis, sample size and SD are particularly important in combining test results; this issue influenced our study. Furthermore, the effects of publication bias related to meta-analyses cannot be excluded.

Discussion

In their project report from 2002 Abbott et al. [76] stated that CD has the potential to improve water use efficiency, maintain crop yields in periods of water stress, and ensure that land drainage systems work to the maximum benefit of farmers. Many articles supporting this thesis have appeared since this report was prepared.

The current systematic study and quantitative meta-analyses found an average 30.5% reduction of drainage outflow in CD in comparison to FD practice, which averaged 71.26 mm in absolute values. Nevertheless, saving an average of about 71 mm of water by applying CD represents a significant amount of water considering global climate warming that can be used to mitigate adverse climate changes, including droughts. The present results are consistent with those of previous meta-analyses involving field and/or model studies. A meta-analysis presented by Wang et al. [77], including measured rather than simulated data, showed that the combined effect of CD on outflow volume reduction was 19.23%. However, it is worth noting that in this work [77] the combined result was based on both controlled surface drainage (ditches) and subsurface drainage (tile). Considering CD for subsurface drainage only, this effect was significantly higher at about 27.5%. Also according to the review conducted by Ross et al. [78] CD is an effective conservation practice for reducing drainage outflow. The results obtained [78] were significantly higher and CD reduced tile drainage volume by 46% on average. The authors also concluded that the greatest potential for CD to reduce discharge is during the non-growing season and confirmed that the efficiency of CD is especially influenced by drain spacing and management. Also Ale et al. [79] stated that there is reduction of drainage outflow using CD due to its principle of operation, but its effectiveness can vary significantly according to the different course of meteorological conditions, the strategy used, or the different parameters of the drainage network. Based on data from 1915–2006, Ale et al. [79] reported that CD has the potential to reduce annual drain outflow by 114 mm (52%), 94 mm (55%) and 75 mm 55%, respectively for 10 m, 20 m and 30 m drain spacings, relative to FD. Similar conclusions are found in some of the papers used in our meta-analysis [59]. For example, the highest outflow reduction using CD was found in the work of Pease et al. [69] and Salazar et al. [63], where outflow reductions of 95% and 61% were achieved at spacings of 10 and 9 m, respectively. On the other hand, the reductions reported by El-Sadek et al. [61] showed little effectiveness of CD in reducing outflow, averaging about 2.13%. In this case, it may have been due to the wide range of spacing adopted, from 10 to as much as 300 m (Table 1), and the fact that these simulations were carried out for soils with sand texture. The efficiency of drainage outflow reduction depends, among other factors, on climatic conditions, as pointed out by Wang et al. [77]. Based on the data and results in the articles selected for our meta-analysis, it can be concluded that there is a statistically significant negative relationship between total precipitation and reduction efficiency by CD expressed as a % (Fig 9). This suggests that the effectiveness of the CD system is higher for areas with lower precipitation. Such a relationship may be due to the fact that high precipitation amounts force the user to apply the FD practice more frequently to maintain optimal soil moisture.

thumbnail
Download:
Fig 9. Reduction of drainage water with controlled drainage versus precipitation.

https://doi.org/10.1371/journal.pone.0267736.g009

In meta-analyses conducted in groups the results vary between 22.63% and 33.73% reduction of drainage outflow for CD practice compared to FD. In general, the meta-analysis of the subgroups indicates that as the weights became higher, the heterogeneity of drainage outflow reduction efficiency within the subgroup decreased and statistical significance increased. These weights were mainly determined by the number of study years.

We also found using meta-analyses in the present paper that CD practice is an effective way of reducing nitrogen losses. According to our results, CD significantly reduced the nitrate loading of the drainage outflow. The average reduction in nitrate losses using CD was 8.36 kg ha-1year-1 compared to FD, and this was slightly lower than that obtained by Carstensen et al. [80] - 12.00 kg NO3 ha-1year-1 (Fig 3B). However, the heterogeneity of our meta-analysis results was relatively high, and the effectiveness showed a large dispersion around the mean (from to 0.88 kg NO3 ha-1year-1 to 12.32 kg NO3 ha-1year-1), which was not observed in previous studies [80]. This was expected, however, as the effect of CD on the efficiency of nitrate loss reduction depends on many factors, such as cropping system, drainage methods and control drainage management method, climatic and other conditions [43, 64, 77, 78], which varied among the literature items we analyzed. The absolute reduction of nitrate losses was mainly regulated by limiting the amount of water outflow into the drains, which was also emphasized in previous review studies [78]. This is confirmed by our results of the relationship between absolute reduction of drainage outflow and absolute reduction of nitrate expressed by the high value of the determination coefficient (R = 0.92 (Pearson), p = 0.018) (Fig 10A). The reduction in nitrate loading loss using CD solutions expressed in relative values (%) is on average 33.6%, with dispersion ranging from 3% to 53%. This average effect is comparable to that reported by Wang et al. [77] (36%) and lower than that reported by Carstensen et al. [80] (50%). In the case of the relationship between relative reduction of drainage outflow and relative reduction of nitrate losses, the correlation is not statistically significant (R = 0.41 (Pearson), p = 0.274) (Fig 10B), which does not confirm previous findings [80]. As with drainage outflows, in general, a negative correlation between increasing publication weights and heterogeneity within a subgroup and a positive correlation with statistical significance can be found for the meta-analysis of nitrate reduction efficacy subgroups. Since these weights were mainly determined by the number of study years (Table 1), it can be concluded that studies covering a longer period were more statistically significant than those conducted for shorter time intervals. Hence, the overall effect of CD on the reduction of nitrate loading loss in subgroup 1 was statistically insignificant. This first subgroup includes only 3 items of articles with a study period of only 2–3 years. In the other two separated subgroups, the articles were for a much larger number of years; therefore, in both subcluster 2 and subcluster 3, the subgroup effects of CD on reducing nitrate loading loss were statistically significant. For subgroup 2, one literature item [61] had statistically insignificant results for the subgroup effect, which was related to the high variation around the average. The high variable effect of CD on reducing nitrate loading losses in that article was related to drain spacing, which varied widely from 10 to 300 m. This large variation in drain spacing, according to previous findings [78], significantly affects the amount of nitrate leached with drainage water and thus the effectiveness of CD versus FD.

thumbnail
Download:
Fig 10.

Absolute (a) and relative (b) reduction of nitrate versus absolute and relative reduction of drainage water with controlled drainage.

https://doi.org/10.1371/journal.pone.0267736.g010

Although the meta-analysis results obtained indicate that CD statistically significantly reduces drainage outflow as well as nitrate loading in drainage water, there is high heterogeneity of study results within the articles used in our meta-analysis. This is due in part to the varying number of years covered by the studies, as well as factors highlighted in other papers [77, 78, 80]. Hence, we suggest that future studies evaluating the effectiveness of CD compared to FD should be based on measured or model data covering long time periods. While the latter are obtainable from calibrated and validated models, the former are expensive and time-consuming. In addition, modeling studies can be conducted for a wide range of factors affecting CD effectiveness, as well as at different scales [78].

Conclusions

This review study, based on meta-analyses of the scientific literature available until December 2020, indicates that CD is an effective practice for reducing drainage outflows and nitrogen load losses on agricultural land. Overall, CD application reduced the amount of drainage outflow by 30.5%, averaging 71.26 mm in absolute terms. It should be noted, however, that not every application of CD will yield similar quantitative results of water saved. In general, studies covering a longer period of analysis have higher statistical significance for the meta-analysis result than studies conducted for shorter time intervals. The results of this study showed that in agricultural areas, the effect of CD on reducing drainage outflow in relation to FD is more effective where there is less precipitation. For nitrate, the application of CD solutions reduces nitrate load losses with drainage water by an average of 33.61%, which is 8.36 kg NO3 ha-1year-1 in absolute terms. However, the findings of the meta-analysis indicate that the heterogeneity of the results of the effect of CD on reducing nitrate losses with drainage water is very high (from to 0.88 kg NO3 ha-1year-1 to 12.32 kg NO3 ha-1year-1). It decreases with the increasing number of years included in the presented results in individual articles. The reduction of nitrate losses in absolute values was mainly regulated by the reduction of drainage outflows, while a weaker relationship is found between the relative reduction of nitrate losses and the relative reduction of drainage outflows.

References

  1. 1. Ritzema HP, Nijland HJ, Croon FW. Subsurface drainage practices: From manual installation to large-scale implementation. Agric. Water Manag. 2006; 86.1–2: 60–71.
  2. 2. Ritzema HP. Drain for Gain: Managing salinity in irrigated lands—A review. Agric. Water Manag. 2016; 176: 18–28.
  3. 3. Valipour M, Krasilnikof J, Yannopoulos S, Kumar R, Deng J, Roccaro P, et al. The Evolution of Agricultural Drainage from the Earliest Times to the Present. Sustainability. 2020; 12(1): 416.
  4. 4. Mitchell JK, McIsaac GF, Walker SE, Hirsch MC. Nitrate in river and subsurface drainage flows from an east central Illinois watershed. Trans. ASAE. 2000; 43: 337–342.
  5. 5. Grigg BC, Southwick LM, Fouss JL, Kornecki TS. Drainage system impacts on surface runoff, nitrate loss, and crop yield on a southern alluvial soil. Trans. ASAE. 2003; 46.6: 1531.
  6. 6. Zhou M, Butterbach-Bahl K. Assessment of nitrate leaching loss on a yield-scaled basis from maize and wheat cropping systems. Plant Soil. 2014; 374.1: 977–991.
  7. 7. Lin BL, Sakoda A, Shibasaki R, Suzuki M. A modelling approach to global nitrate leaching caused by anthropogenic fertilisation. Water Res. 2001; 35.8: 1961–1968. pmid:11337842
  8. 8. Carpenter SR, Caraco NF, Correll DL, Howarth RW, Sharpley AN, Smith VH. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecol. Appl. 1998; 8: 559–568.
  9. 9. Camargo JA, Alonso A. Ecological and toxicological effects of inorganic nitrogen pollution in aquatic ecosystems: A global assessment. Environ. Int. 2006; 32: 831–849. pmid:16781774
  10. 10. O’Geen AT, Budd R, Gan J, Maynard JJ, Parikh SJ, Dahlgren RA. Mitigating nonpoint source pollution in agriculture with constructed and restored wetlands. Adv. Agron. 2010; 108: 1–76.
  11. 11. Gramlich A, Stoll S, Stamm C, Walter T, Prasuhn V. Effects of artificial land drainage on hydrology, nutrient and pesticide fluxes from agricultural fields–a review. Agric. Ecosyst. Environ. 2018; 266: 84–99.
  12. 12. Hansen B, Thorling L, Kim H, Blicher-Mathiesen G. Long-term nitrate response in shallow groundwater to agricultural N regulations in Denmark. J. Environ. Manage. 2019; 240. pmid:30928796
  13. 13. Liu W, Youssef MA, Birgand FP, Chescheir GM, Tian S, Maxwell BM. Processes and mechanisms controlling nitrate dynamics in an artificially drained field: insights from high-frequency water quality measurements. Agric. Water Manag. 2020; 232: 106032.
  14. 14. Kladivko EJ, Frankenberger JR, Jaynes DB, Meek DW, Jenkinson BJ, Fausey NR. Nitrate leaching to subsurface drains as affected by drain spacing and changes in crop production system. J. Environ. Qual. 2004; 33(5): 1803–1813. pmid:15356241
  15. 15. Sutton MA, Oenema O, Erisman JW., Leip A, van Grinsven H, Winiwarter W. Too much of a good thing. Nature. 2011; 472(7342): 159–161. pmid:21478874
  16. 16. Zhao X, Christianson LE, Harmel D, Pittelkow CM. Assessment of drainage nitrogen losses on a yield-scaled basis. Field Crops Res. 2016; 199: 156–166.
  17. 17. Blackwell MSA, Pilgrim ES. Ecosystem services delivered by small-scale wetlands. Hydrol. Sci. J. 2011; 56: 1467–1484.
  18. 18. Radcliffe DE, Reid DK, Blomback K, Bolster CH, Collick AS, Easton ZM, et al. Applicability of models to predict phosphorus losses in drained fields: A review. J. Environ. Qual. 2015; 44(2): 614–628. pmid:26023980
  19. 19. Evans RO, Skaggs RW. Development of controlled drainage as a BMP in North Carolina. In: Drainage VIII, 21–24 March 2004. American Society of Agricultural and Biological Engineers. 2004; p. 1.
  20. 20. Woli KP, David MB, Cooke RA, McIsaac GF, Mitchell CA. Nitrogen balance in and export from agricultural fields associated with controlled drainage systems and denitrifying bioreactors. Ecol. Eng. 2010; 36: 1558–1566.
  21. 21. Drury CF, Tan CS, Welacky TW, Reynolds WD, Zhang TQ, Oloya TO, et al. Reducing nitrate loss in tile drainage water with cover crops and water-table management systems. J. Environ. Qual. 2014; 43: 587. pmid:25602660
  22. 22. Wesström I, Joel A, Messing I. Controlled drainage and subirrigation: A water management option to reduce non-point source pollution from agricultural land. Agric. Ecosyst. Environ. 2014; 198: 74–82.
  23. 23. Christianson L, Bhandari A, Helmers M, Kult K, Sutphin T, Wolf R. Performance evaluation of four field-scale agricultural drainage denitrification bioreactors in Iowa. Trans. ASABE. 2012; 55: 2163–2174.
  24. 24. Christianson LE, Bhandari A, Helmers MJ. A practice-oriented review of woodchip bioreactors for subsurface agricultural drainage. Appl. Eng. Agric. 2012; 28.6: 861–874.
  25. 25. David MB, Gentry LE, Cooke RA, Herbstritt SM. Temperature and substrate control woodchip bioreactor performance in reducing tile nitrate loads in East-Central Illinois. J. Environ. Qual. 2016; 45: 822–829. pmid:27136147
  26. 26. Tanner CC, Sukias JPS. Multiyear nutrient removal performance of three constructed wetlands intercepting tile drain flows from grazed pastures. J. Environ. Qual. 2011; 40: 620–633. pmid:21520769
  27. 27. Tournebize J, Chaumont C, Fesneau C, Guenne A, Vincent B, Garnier J, et al. Long-term nitrate removal in a buffering pond-reservoir system receiving water from an agricultural drained catchment. Ecol. Eng. 2015; 80: 32–45.
  28. 28. Groh TA, Gentry LE, David MB. Nitrogen removal and greenhouse gas emissions from constructed wetlands receiving tile drainage water. J. Environ. Qual. 2015; 44: 1001–1010. pmid:26024280
  29. 29. Zak D, Kronvang B, Carstensen MV, Hoffmann CC, Kjeldgaard A, Larsen SE, et al. Nitrogen and phosphorus removal from agricultural runoff in integrated buffer zones. Environ. Sci. Technol. 2018; 52: 6508–6517. pmid:29733209
  30. 30. Jaynes DB, Isenhart TM. Performance of saturated riparian buffers in Iowa, USA. J. Environ. Qual. 2019; 48: 289–296. pmid:30951131
  31. 31. van’t Veen SGW, Kronvang B, Zak DH, Ovesen NB, Jensen H. Intelligente bufferzoner. 2019. Notat fra DCE—Nationalt Center for Miljø og Energi. Available from: http://dce.au.dk/fileadmin/dce.au.dk/Udgivelser/Notater_2019/Intelligente_bufferzoner.pdf
  32. 32. Reinhart B, Frankenberger J, Abendroth L, Ahiablame L, Bowling L, Brown L, et al. Drainage water storage for improved resiliency and environmental performance of agricultural landscapes. In: 2016 10th International Drainage Symposium Conference, 6–9 September 2016, Minneapolis, Minnesota. American Society of Agricultural and Biological Engineers. 2016; p. 1–8.
  33. 33. Skaggs RW, Fausey NR, Evans RO. Drainage water management. J. Soil Water Conserv. 2012; 67.6: 167A–172A.
  34. 34. Skaggs RW, Youssef MA, Chescheir GM. DRAINMOD: Model use, calibration, and validation. Trans. ASABE. 2012; 55(4): 1509–1522.
  35. 35. Gilliam JW, Skaggs RW, Weed SB. Drainage control to diminish nitrate loss from agricultural fields. J. Environ. Qual. 1979; 8 (1): 137–142.
  36. 36. Lalonde V, Madramootoo CA, Trenholm L, Broughton RS. Effects of controlled drainage on nitrate concentrations in subsurface drain discharge. Agric. Water Manag. 1996; 29.2: 187–199.
  37. 37. Wesström I, Messing I, Linner H, Lindstrom J. Controlled drainage—effects on drain outflow and water quality. Agric. Water Manag. 2001; 47.2: 85–100.
  38. 38. Bakken LR, Bergaust L, Liu B, Frostegard A. Regulation of denitrification at the cellular level: a clue to the understanding of N2O emissions from soils. Philos. Trans. R. Soc. Lond., B, Biol. Sci. 2012; 367.1593: 1226–1234. pmid:22451108
  39. 39. Carstensen MV, Børgesen CD, Ovesen NB, Poulsen JR, Hvid SK, Kronvang B. Controlled drainage as a targeted mitigation measure for nitrogen and phosphorus. J. Environ. Qual. 2019; 48.3: 677–685. pmid:31180423
  40. 40. Ramoska E, Bastiene N, Saulys V. Evaluation of controlled drainage efficiency in Lithuania. Irrig. Drain. 2011; 60.2: 196–206.
  41. 41. Wesström I, Messing I. Effects of controlled drainage on N and P losses and N dynamics in a loamy sand with spring crops. Agric. Water Manag. 2007; 87.3: 229–240.
  42. 42. Bonaiti G, Borin M. Efficiency of controlled drainage and subirrigation in reducing nitrogen losses from agricultural fields. Agric. Water Manag. 2010; 98.2: 343–352.
  43. 43. Yu Y, Xu J, Zhang P, Meng Y, Xiong Y. Controlled Irrigation and Drainage Reduce Rainfall Runoff and Nitrogen Loss in Paddy Fields. Int. J. Environ. Res. Public Health. 2021; 18: 3348. pmid:33805028
  44. 44. Gao S, Yu S, Wang M, Meng J, Tang S, Ding J, et al. Improving Water Productivity and Reducing Nutrient Losses by Controlled Irrigation and Drainage in Paddy Fields. Pol. J. Environ. Stud. 2018; 27.3.
  45. 45. Jouni HJ, Liaghat A, Hassanoghli A, Henk R. Managing controlled drainage in irrigated farmers’ fields: A case study in the Moghan plain, Iran. Agric. Water Manag. 2018; 208: 393–405.
  46. 46. Karegoudar AV, Vishwanath J, Anand SR, Rajkumar RH, Ambast SK, Kaledhonkar MJ. Feasibility of Controlled Drainage in Saline Vertisols of TBP Command Area of Karnataka, India. Irrig. Drain. 2019; 68.5: 969–978.
  47. 47. Matsuo N, Takahashi M, Yamada T, Takahashi M, Hajika M, Fukami K, et al. Effects of water table management and row width on the growth and yield of three soybean cultivars in southwestern Japan. Agric. Water Manag. 2017; 192: 85–97.
  48. 48. Wahba MAS, El‐Ganainy M, Abdel‐Dayem MS, Gobran ATEF, Kandil H. Controlled drainage effects on water quality under semi‐arid conditions in the western delta of Egypt. Irrig. Drain. 2001; 50.4: 295–308.
  49. 49. Skaggs RW. DRAINMOD Reference Report. Methods for Design and Evaluation of Drainage-water Management Systems for Soils with High Water Tables USDA-SCS. South National Technical Center, Fort Worth, Texas; 1980.
    • 50. Luo W, Skaggs RW, Chescheir GM. DRAINMOD modifications for cold conditions. Trans. ASAE. 2000; 43.6: 1569.
    • 51. Brown RA, Skaggs RW, Hunt WF III. Calibration and validation of DRAINMOD to model bioretention hydrology. J. Hydrol. 2013; 486: 430–442.
    • 52. Ma L, Malone RW, Heilman P, Jaynes DB, Ahuja LR, Saseendran SA, et al. RZWQM simulated effects of crop rotation, tillage, and controlled drainage on crop yield and nitrate-N loss in drain flow. Geoderma. 2007; 140.3: 260–271.
    • 53. Sánchez-Meca J, Marin-Martinez F. Weighting by inverse variance or by sample size in meta-analysis: A simulation study. Educ. Psychol. Meas. 1998; 58.2: 211–220.
    • 54. Manolov R, Guilera G, Sierra V. Weighting strategies in the meta-analysis of single-case studies. Behav. Res. Methods. 2014; 46.4: 1152–1166. pmid:24488814
    • 55. Cochran WG. The combination of estimates from different experiments. Biometrics. 1954; 10.1: 101–129.
    • 56. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003; 327.7414: 557–560. pmid:12958120
    • 57. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994; 1088–1101. pmid:7786990
    • 58. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997; 315.7109: 629–634. pmid:9310563
    • 59. Skaggs RW, Breve MA, Gilliam JW. Predicting effects of water table management on loss of nitrogen from poorly drained soils. Eur. J. Agron. 1995; 4.4: 441–451.
    • 60. Breve MA, Skaggs RW, Gilliam JW, Parsons JE, Mohammad AT, Chescheir GM, et al. Field testing of DRAINMOD-N. Trans. ASAE. 1997; 40(4): 1077–1085.
    • 61. El‐Sadek A, Feyen J, Ragab R. Simulation of nitrogen balance of maize field under different drainage strategies using the DRAINMOD‐N model. Irrig. Drain. 2002; 51.1: 61–75.
    • 62. Salazar O, Wesström I, Youssef MA, Skaggs RW, Joel A. Evaluation of the DRAINMOD–N II model for predicting nitrogen losses in a loamy sand under cultivation in south-east Sweden. Agric. Water Manag. 2009; 96.2: 267–281.
    • 63. Luo W, Sands GR, Youssef M, Strock JS, Song I, Canelon D. Modeling the impact of alternative drainage practices in the northern Corn-belt with DRAINMOD-NII. Agric. Water Manag. 2010; 97.3: 389–398.
    • 64. Skaggs RW, Youssef MA, Chescheir GM. A DRAINMOD-based method to estimate effects of drainage water management on annual nitrogen loss to surface water. Trans. ASABE. 2012; 55.3: 799–808.
    • 65. Negm LM, Youssef MA, Chescheir GM, Skaggs RW. DRAINMOD-based tools for quantifying reductions in annual drainage flow and nitrate losses resulting from drainage water management on croplands in eastern North Carolina. Agric. Water Manag. 2016; 166: 86–100.
    • 66. Negm LM, Youssef MA, Jaynes DB. Evaluation of DRAINMOD-DSSAT simulated effects of controlled drainage on crop yield, water balance, and water quality for a corn-soybean cropping system in central Iowa. Agric. Water Manag. 2017; 187: 57–68.
    • 67. Youssef MA, Abdelbaki AM, Negm LM, Skaggs RW, Thorp KR, Jaynes DB. DRAINMOD-simulated performance of controlled drainage across the US Midwest. Agric. Water Manag. 2018; 197: 54–66.
    • 68. Skaggs RW, Youssef MA, Gilliam JW, Evans RO. Effect of controlled drainage on water and nitrogen balances in drained lands. Trans. ASABE. 2010; 53.6: 1843–1850.
    • 69. Pease LA, Fausey NR, Martin JF, Brown LC. Projected climate change effects on subsurface drainage and the performance of controlled drainage in the Western Lake Erie Basin. J. Soil Water Conserv. 2017; 72.3: 240–250.
    • 70. Sojka M, Kozłowski M, Stasik R, Napierała M, Kęsicka B, Wróżyński R, et al. Sustainable water management in agriculture -The impact of drainage water management on groundwater table dynamics and subsurface outflow. Sustainability. 2019; 11.15: 4201.
    • 71. Singh S, Bhattarai R, Negm LM, Youssef MA, Pittelkow CM. Evaluation of nitrogen loss reduction strategies using DRAINMOD-DSSAT in east-central Illinois. Agric. Water Manag. 2020; 240: 106322.
    • 72. Breve MA, Skaggs RW, Parsons JE, Gilliam JW. Using the DRAINMOD-N model to study effects of drainage system design and management on crop productivity, profitability and NO3–N losses in drainage water. Agric. Water Manag. 1998; 35.3: 227–243.
    • 73. Singh R, Helmers MJ, Crumpton WG, Lemke DW. Predicting effects of drainage water management in Iowa’s subsurface drained landscapes. Agric. Water Manag. 2007; 92.3: 162–170.
    • 74. Ale S, Bowling LC, Owens PR, Brouder SM, Frankenberger JR. Development and application of a distributed modelling approach to assess the watershed-scale impact of drainage water management. Agric. Water Manag. 2012; 107: 23–33.
    • 75. Ale S, Bowling LC, Youssef MA, Brouder SM. Evaluation of simulated strategies for reducing nitrate–nitrogen losses through subsurface drainage systems. J. Environ. Qual. 2012; 41.1: 217–228. pmid:22218190
    • 76. Abbott C, Lawrence P, Pearce G, Abdel-Gawad S. Review of the potential for controlled drainage around the world. Project Report. HR Wallingford Ltd. 2002. Available from: https://eprints.hrwallingford.com/471/.
      • 77. Wang Z, Shao G, Lu J, Zhang K, Gao Y, Ding J. Effects of controlled drainage on crop yield, drainage water quantity and quality: A meta-analysis. Agric. Water Manag. 2020; 239: 106253.
      • 78. Ross JA, Herbert ME, Sowa SP, Frankenberger JR, King KW, Christopher SF et al. A synthesis and comparative evaluation of factors influencing the effectiveness of drainage water management. Agric. Water Manag. 2016; 178: 366–376.
      • 79. Ale S, Bowling LC, Frankenberger JR, Brouder SM, Kladivko EJ. Climate variability and drain spacing influence on drainage water management system operation. Vadose Zone J. 2010; 9.1: 43–52.
      • 80. Carstensen MV, Hashemi F, Hoffmann CC, Zak D, Audet J, Kronvang B. Efficiency of mitigation measures targeting nutrient losses from agricultural drainage systems: A review. Ambio. 2020; 49: 1820–1837. pmid:32494964