|Main authors:||Oene Oenema, Meindert Commelin, Piet Groenendijk, John Williams, Susanne Klages, Isobel Wright, Morten Graversgaard, Irina Calciu, António Ferreira, Tommy Dalgaard, Nicolas Surdyk, Marina Pintar, Christophoros Christophoridis, Peter Schipper, Donnacha Doody|
|FAIRWAYiS Editor:||Jane Brandt|
|Source document:||»Oenema, O. et al. 2018. Review of measures to decrease nitrate pollution of drinking water sources. FAIRWAY Project Deliverable 4.1, 125 pp|
Measures to prevent and reduce the risk of leaching and surface runoff are usually categorized according to the source-pathway-receptor concept, i.e., there are
- source-based measures or input-based measures,
- pathway-based measures, and
- receptor or effects-based measures (e.g. Burt et al., 1993; Van Boekel 2015).
Examples of source-based measures are N application limits, balanced fertilization, appropriated storage of animal manures and fertilizers, and prohibition periods for the application of manures and fertilizers. Source-based measures are often a combination of N application amounts, methods and timing, commonly referred to as the 4R strategy (IPNI, 2018). Examples of pathway-based measures are irrigation measures, drainage, buffer strips, green covers, terracing. Examples of receptor or effects-based measures are dredging, creation of riparian zones, water purification. These three categories can be understood also from the ‘hole-in-the-pipe-model’ in Figure 5 and discussed in »The nitrogen cycle and nitrogen transformation processes.
Source-based measures are often seen as effective measures, because of the restriction on N input. However, it has to be realized that the response of crop production to N input is nonlinear, which is known as the law of diminishing returns (or the law of diminishing marginal returns). When the availability of N in soil is low, the crop response is high, and the risk of N losses will be relatively low. Conversely, when the availability of N in soil is high, the response of the crop to N input is low, and the risk of N losses will be relatively high, because the crop is unable to recover the applied N. Hence, the risk of leaching is the reverse of the law of diminishing crop yield returns; the risk increase more than proportional with N input, until a certain N input level. Thereafter, the risk of leaching is more or less linearly related to N input.
Here we discuss the rational and mechanism of key measures in more depth. It starts with nitrogen management in general, as this measure is increasingly seen as the overall integrative measure, also to minimize pollution swapping.
Nitrogen management can be defined as “a coherent set of activities related to N use of farms to achieve agronomic and environmental/ecological objectives” (Oenema and Pietrzak, 2002). The agronomic objectives relate to crop yield and quality and animal performance. The environmental/ecological objectives relate to N losses from agriculture. Taking account of the whole N cycle emphasizes the need to consider all aspects of N cycling, to circumvent pollution swapping. Nitrogen management planning at farm level is increasingly seen as the starting point also of all measures aimed at reducing nitrate losses. The importance of a broader look is also emphasized by the term integrated nitrogen (nutrient) management (Sutton et al., 2011; 2013; FAO, 2018).
Depending on the type of farming systems, N management at farm level involves a series of management activities in an integrated way, including:
- Fertilization of crops;
- Crop growth, harvest and residue management;
- Growth of catch or cover crops;
- Grassland management;
- Soil cultivation, drainage and irrigation;
- Animal feeding;
- Herd management (including welfare considerations), including animal housing;
- Manure management, including manure storage and application;
- Ammonia emission abatement measures;
- Nitrate leaching and run-off abatement measures;
- N2O emission abatement measures.
Nitrogen management at farm level involves the reiterative cycle of analysing, making decisions, planning, acting, evaluation & control, and adjustment (Bittman et al., 2014). It depends strongly on the availability of easy accessible information of individual fields (i.e. soil analysis), the available nutrients in manures as well as the nutrients from additional sources, and the exports of nutrients in crops and animal products, as foreseen in view of experiences in preceding years. In addition to data of inputs and outputs, information is needed on the available time windows suitable for applying nutrients, based on pedo-climatic conditions. The success of any planning procedure also depends on the timely availability of information. A true planning must therefore not be restricted to a listing of the required items of information, but also define the recurring temporal flows of information (Anonymous, 2011). Extension services play an essential role in providing these conditions. The nutrient management planning has to be linked to all the other measures to reduce the risk of nitrate pollution of groundwater and surface waters, especially also to the application limits. It requires also regular soil fertility analyses and analyses of the compositions of the animal manures and harvested crops; these provide a solid basis for the nutrient management planning. Effects of nitrogen management do show up in nitrate concentrations of groundwater aquifers (e.g. Kirchman et al., 2002; Hansen et al., 2018).
Land use management can have a significant effect on surface run-off and leaching of nutrients (e.g. Goulding, 2000). Crop rotations systems and the proportion of the land area devoted to permanent crops relative to annual tillage crops may be adjusted when the surface run-off potential and downward leaching potential are high, because crop species differ greatly in their ability to intercept and absorb applied and mineralised N (e.g. Dalgaard et al., 2014; Hashemi et al., 2018; Schröder et al., 2010). A sequence of crops differing in ability to intercept and absorb applied and mineralised N can transfer N between individual crops, and thereby maximize N utilization. In commercial farms, however, there is a certain specialisation and consequently individual farmers may have less opportunity to optimize nutrient management through balanced rotations i.e. mitigate the adverse environmental effect of one crop with the beneficial effect of another crop. Crop rotations systems and the proportion of the land area devoted to permanent crops relative to annual tillage crops should be adjusted when the surface run-off potential and downward leaching potential are high (Hashemi et al., 2016).
Growing ‘leaky’ crop types can be compensated by the growth of crops that are much less leaky or by the nearby presence of unfertilised natural vegetation (e.g. Schröder et al., 2007; Wendland et al., 2009). Several vegetables (e.g., spinach, lettuce, strawberries, leek) and some arable crops (e.g., potatoes, peas) with shallow rooting systems and relatively short growing seasons are ‘leaky’ crops; hence, these crops should be rotated with cover crops and cereals that can mop up residual mineral N from the soil. Also, tree-lines, border strips, riparian zones, and mixed cropping systems may contribute to decreasing the risk of surface runoff and increasing biodiversity and buffering against diseases. Ploughing-up grass-leys should be done in early spring, to allow a subsequent crop to mop up the N released from the mineralized sod (e.g. Schröder et al., 1999).
Nutrient inputs must be ‘balanced’ with nutrient outputs to minimize the risk of N losses. Balanced fertilization often has two meanings, i.e., (i) all 14 required nutrient elements should be made available in the proper ratios that reflect the requirement of the crop for these nutrient elements, and (ii) the total input of N should balance the total crop N requirements. Here, we discuss the importance of the second interpretation of balanced fertilization.
Balancing N inputs to the N demand of the crop involves the assessment of the availability of the various possible N inputs, also termed the N fertiliser replacement value of the inputs. In addition, it is important to assess the recovery of the available N in the soil by the crop, the extent to which the N taken up by crops is invested in harvested plant parts, and the fate of the resultant surplus N input (Kirchman et al., 2002; Wendland et al., 2009). As far as the various types of inputs are concerned, crops can derive N from the soil mineral N present at the start of the growing season, N mineralising from earlier inputs (manures, crop residues, peat), mineral N in fertilisers and manure, atmospherically deposited N, biologically fixed N, and N in irrigation water. The N fertiliser equivalency of these sources depends partly on intrinsic characteristics, such as ratios of (readily available) carbon and (readily available) N, but also strongly on the time and method of their application, the soil type and the manuring history (Schröder et al., 2005a, 2007a).
The soil N surplus can be defined as the difference between the total N input (fertilisers, manures, biologically fixed N, mineralised N, atmospherically deposited N, N in irrigation water) and the N output (harvested N, volatilized ammonia N, N temporarily immobilised to sustain the mineralisation, N lost via denitrification). Underlying factors for the discrepancy between inputs and outputs are (i) the fertiliser equivalency (N fertiliser replacement value: NFRV) of the various input sources determining the amount of N available to crops, (ii) the uptake efficiency (apparent recovery: ANR) of the available N, determining the ultimate crop uptake, (iii) the harvest efficiency (harvest index: HI) indicating to which extent the N taken up is exported from the field (Schröder et al., 2005a, 2007a). The soil N surplus is vulnerable to N losses via leaching and denitrification; a large N surplus is a proxy indicator for the nitrate leaching losses (Osterbrug et al., 2008; Klages et al., 2018).
The utilization of inputs is determined by the product of NFRV x ANR x HI. This product is determined by the intrinsic properties of N input sources and crops, the ability of crops to assimilate N and the combined effects of climate and weather, soil characteristics, and management. As far as N demand of a farm as a whole is concerned, the relative share of crops represented in the rotation as well as their attainable yield must be considered.
The nitrate concentration in water bodies is the result of a specific load being dissolved in a specific volume of water. The N load is determined by the extent to which the soil N surplus effectively leaches or runs-off. The N load and the soil N surplus are not necessarily the same, as the initial soil N surplus can be exposed to various conversions, including denitrification (i.e. the conversion of dissolved N form into gaseous N forms) or retentions in a broad sense (e.g. N captured in vegetated buffer strips). The factor linking the soil N surplus to the N load can be called ‘leaching fraction’ (Schröder et al., 2005a, 2007b; Osterburg et al., 2007). Balanced N fertilization is considered as a most effective measure to reduce nitrate leaching losses (Velthof et al., 2009; Oenema et al., 2009).
Fertilizers and manures must be applied to land in such a way that the nutrients can be utilized by the growing crop in an effective way. Basically, this means that the fertilizers and manure must be applied at the right time, right amount, right place and the right depth. This requires that appropriated techniques are used. Split application, band application, injection, variable rate application are common techniques with a proven high effectiveness, but application in practice will depend on the local site conditions. The spatial positioning of fertilisers and manures in the field is one of the factors determining to what extent nutrients will be available to crops or exposed to loss processes (Anonymous, 2011; Chen et al., 2014).
Precision fertilization can be seen as a further clarification of balanced fertilization, i.e., precision fertilization involves in the first place balanced fertilization. Positioning has a horizontal and a vertical component. Horizontal aspects pertain to the evenness of application, variable-rate application as a function of the soil nutrient status, and the ability of machines to position nutrients to just the rooted parts of the soil profile in case of row crops. Vertical aspects pertain to the ability of (combinations of) machines to incorporate fertilisers and manures in such a way that the risks of volatilization and run-off losses are minimized, whilst assuring that the nutrients can still be timely intercepted by plant roots. In some cases, top dressings via spraying of dissolved nutrients are practices to ease the uptake by the crop.
Precision fertilization can increase the use efficiency of applied N and thereby decrease the risk of leaching loss. The use efficiency of N is highest when high-yielding crop varieties are used, all other essential nutrients and water are in adequate supply, and pest and diseases and weeds are controlled. Increasing crop yield and N withdrawal with the harvested crop, through for example genetic improvement or improved pest and disease control and/or weed control, at constant N input, reduces the risk of N leaching. Hence, N use efficiency enhancing measures reduce the risk of N leaching, especially when the N input is adjusted. The term ‘efficiency enhancing measures’ is often preferred over ‘source-based measures’, because efficiency enhancing measures address the output : input ratio, and thus consider both increases in N output and decreases in N input. However, a high efficiency may be achieved at low N input and at relatively high N input; the N surplus will be higher in the latter case than in the former. This is why the EU-Nitrogen Expert Panel recently suggested to report both NUE, N surplus and N output (EU-Nitrogen Expert Panel, 2015).
Precision fertilization may decrease potential N leaching especially in grazed pastures with a huge spatial variability in N input through the uneven spreading of urine and dung from grazing animals (Di and Cameron, 2002). However, it is notoriously difficult to identify urine and dung patches in the field and to adjust N applications via N fertilizer and/or manure (Buckthought et al., 2015; Roten et al., 2017)
Cover crops are crops grown after main crops and intended to intercept the mineral N left by or liberated from the residues of these main crops, or to minimize erosion risks (Aronsson et al., 2016; Ostenburg et al., 2007; Van Boekel., 2015. Instead of becoming lost, residual N can in this way be transferred to a next growing season where it can contribute to the N supply of subsequent main crops. Cover crops should not be fertilised, unlike the so-called green manures sensu strictu of which the production is often maximized by the application of plant nutrients to increase the input of organic matter into the soil. Crop species largely differ in their ability to scavenge soil mineral N and transpose this N effectively. The ideal species should be able to germinate in a relatively dry seedbed, should be frost or even cold resistant and should be deep rooting where residual soil mineral N happens to find itself at greater depths (Dalgaard et al., 2014).
Leguminous species may be very suitable to act as a green manure in low input cropping systems in need of additional N sources, but are less apt as scavenger of N residues. The amounts of N fixed by this type of crops may increase instead of mitigate the risks of N emissions. A successful establishment and growth of cover crops is often more difficult after crops typically associated with a late harvest (potatoes, maize) than after early crops (cereals), whereas the amounts of residual mineral N are greatest in these late crops. The potential yield of a green canopy is strongly related to the length of the period during which weather conditions (temperature, light) favour biological processes. The larger the fraction of this period being used for the production of the main crop, the smaller the remaining fraction available for a subsequent cover crop. The available heat sum (‘degree days’, i.e. the summed daily average temperatures above a threshold value allowing biological processes) determines to which extent residual N can indeed be taken up by cover crops. As a result, it is difficult to grow cover crops in cool regions (Aronsson et al., 2016).
Results indicate that cover crops are effective to reduce nitrate leaching losses, especially when the main crop is harvested early. This allows the cover crop to establish well and to mop up residual mineral nitrogen from the soil. It is also a rather cost-effective method (Ostenbrug et al., 2007; Dalgaard et al., 2014; Aronsson et al., 2016).
The storage capacity of containers for livestock manure must be large enough to store the manures produced during the period when the application of manures are inappropriate. It is one of the main measures of the Annex II of the Nitrates Directive, but not included in the list of measures of Osterbrug et al., (2007) and Van Boekel (2015). The investment costs are rather high (3-10 euro per m3) but depending on the type and volume of the storage (Bittman et al., 2014).
The construction of the storage container (or vessel, lagoon or pit) must be robust and leak-tight, and should be covered preferably to minimize the loss of gaseous ammonia and the influx of rain water. The required storage capacity may range from 3 to 9 months per year, depending the pedo-climatic zones, land use and the vulnerability of the nearby water resources. The size of the containers depends on the number of housed animals on the farm and the volume of manure produced per animal, corrected for the possible influx of spilled drinking water, cleaning water and the efflux of evaporation losses. The manure production per animal depends on animal category, production level, live weight, and the digestibility of the offered feed stuffs.
The governing factors for defining the manure storage capacity are length of the period when the land application of manure is inappropriate, number and type of animal species, manure production per animal species, manure type (solid, liquids and slurries), addition of bedding material and litter, addition of cleaning, spilling and rain water, presence of storage cover, manure processing and transport, evaporative losses and decomposition losses during storage.
The effectiveness of this measure strongly depends on the reference (or control treatment). When the reference method is daily spread (including the non-growing period), the effectiveness may be high. If the reference method is an unsealed lagoon, the effectiveness may be high too. However, if the reference is a proper storage for 4 months and the treatment measure is a storage capacity of 6 month, the effectiveness may be limited. However, there are no studies that have examined the effects of storage capacity in experimental studies.
The maximum application rate of animal manure is 170 kg N per ha per year according to the EU Nitrates Directive, irrespective of land use and climate zone. However, there is opportunity to derogate from this limit, when justified on scientifically sound arguments. The regulation of the application rate requires an accurate assessment of the amounts of N and P applied in the form of animal manure (Schröder et al., 2007b). As far as N is concerned, the effects on water quality are not only determined by the applied rate of total N, but also by the ratio of mineral-N (Nm) and organically bound N (Norg) in the manure. One of the major factors determining the Nm:Norg ratio is the housing type i.e. the decision to keep animals on slatted floors resulting in slurries or provide ample bedding material (potentially) resulting in solid manures. To be able to respect the limit of 170 kg N per ha per year, farmers have to account for the total amount of N excreted by all farm animals, and correct this amount for gaseous N losses from housing and manure storages. Farmers have usually access to tables to find out how much N is in the manure per animal. Alternatively, farmer may estimate the amounts of N per animal on the basis of the mass balance:
Nexcretion = Nintake by the animal – Nretention by the animals
The amount of N excreted must be corrected also for the gaseous N losses during storage (which may range from 10-40%, depending on the manure type and storage condition and duration. This assessment (book keeping) of amount of N (and P) in manure is likely the most accurate way of estimating N production, provided accurate information is available about total feed intake, weighted mean protein content of the feed and the amount of animal protein exported from the farm.
The manure N application limit of the EU Nitrates Directive is effective in the sense that it limits the manure N application. There is a considerable amount of literature that has investigated the differences between manure N and synthetic fertilizer N efficiency, also in terms of reducing N leaching. In short-terms experiments, N leaching losses from treatments with manure N are commonly lower than the N leaching loss from treatments with synthetic N fertilizer, because part of the N in manure is organically bound and hence not available to the crop nor vulnerable to leaching. However, long-term experiments often indicate that leaching losses from treatments with manure N are often higher than the N leaching loss from treatments with synthetic N fertilizer, because the mineralization of organic N from manure is not well synchronized to the N uptake by the crop (Basso and Ritchie, 2005; Schröder et al., 1993).
Risks of nutrient leaching are most imminent when
- the natural precipitation exceeds the evapo-transpiration and the water holding capacity of the soil,
- soils tend to crack which may lead to preferential flow, or tend to seal which may lead to overland flow,
- the land is sloping, and 4) soils contain considerable amounts of water-soluble N and P, while there is no growing crop.
Hence, application of fertilizers and manures is inappropriate when the demand by the crop of nutrients is low and the risk for surface runoff and leaching of nutrients are high. The risk of leaching depends also on the ratio of mineral N to organically-bound N; solid manures with litter commonly have a Nm/Ntotal ratio of <0.3, and are less vulnerable to leaching (but not to surface run off).
Application of fertilizers and manures just before and during the growing season has been shown to be an effective method to reduce nitrate leaching losses (Schröder et al., 1993; Thomsen et al., 1993; Beckwith et al. 1998).
The application of fertilizers and manures to steeply sloping land is associated with high risk for surface run-off of N and P, which may result in the pollution and eutrophication of surface waters. Hence, the application of fertilizers and manures to steeply sloping land must be limited and done in such a way that the risk of surface run-off of N and P is strongly minimized. Risks of surface run-off are greatest where there are nutrients (sources), and where the infiltration capacity of soils and potential residence time of water are low. This implies that risks are positively related to the soil fertility status of a soil, application rates of fertilizers and manures and the extent to which they are left on the surface, the surface roughness as related to the infiltration capacity, and the extent to which water is hold in situ instead of allowed to flow run-off as quickly as possible. Infiltration capacity can be increased by minimum tillage without the removal of crops residues (mulching), sub-soiling, ridge tillage, cover cropping, the conversion of arable land into grassland, agro-forestry or complete reforestation. Incorporation of fertilizers and manure may help to reduce the risks. This is technically feasible, but it may be more difficult in the case of the presence of stones. As nutrients in solid manures are generally less mobile than those in slurries, solid manures are somewhat less risky than slurries
The length and steepness of the slope define the slope classes:
- Flat: 0 to 2%; negligible risk of surface run-off (green)
- Rolling: 2 to 8%; moderate risk of surface runoff (yellow)
- Sloping: 8 to 15%; high risk surface runoff (pink)
- Moderately steep: >15%; very high risk of surface runoff (red)
Various studies have indeed verified that application of fertilizers and manures on sloping lands is conducive to surface run off and losses of N and P (Smith et al., 2001a, 2001b; Gilley and Risse (2000). Management practices used to control runoff include contouring, permanent green covers (grassland), strip cropping, conservation tillage, terraces, buffer strips, and appropriate timing and subsurface application of manure and fertilizer. More than one runoff-control practice may be necessary in areas with high runoff potential. Alternatively, the application of fertilizers and manure is prohibited.
The application of fertilizers and manures to water-saturated, flooded, frozen or snow-covered ground is associated with very high risk for surface run-off of N and P, which may result in the pollution and eutrophication of surface waters. Moreover, applications of fertilizers and manures are not effective as there will be no growing crop and a demand for nutrients. Hence, the application of fertilizers and manures to water-saturated, flooded, frozen or snow-covered ground should be prohibited.
However, the application of fertilizers and manures on frozen but dry soils without snow cover may be advantageous in pedo-climatic zones with a short growing season and on soils with high risk of soil compaction by traffic. Farmers may appreciate frozen soils for their carrying capacity allowing land spreading of manures without negatively affecting the soil structure. Even without a snow cover this practice can still be conducive to serious nutrient losses because of the very low permeability of the soils. The low permeability can trigger superficial run-off, also when the frozen soil layer is below the soil surface (i.e., the soil surface is thawed already) and thin. Hence, manures and fertilisers applications should be avoided when the soil is frozen and snow covered, irrespective of the thickness of the snow cover. On dry (without any snow cover) frozen soils, overgrown by a winter cereal, starter fertilizer application could be beneficial, when the soil would become compacted otherwise. However, special pre-cautionary measures should be taken in this case, such as unfertilized buffer strips, to minimize the risk of pollution of surface waters.
There is indeed quite some empirical evidence that manure application on frozen, snow covered and/or flooded soils may contributes to increased leaching and surface run-off losses. This evidence mainly comes from studies in northern America (Converse et al., 1976; Srinivasan et al., 2006; Williams et al., 2011), and less from northern Europe. The increased potential for losses of N and P originates from the facts that
- nutrients may not easily infiltrate frozen, snow-covered and/or flooded soils,
- there is no growing crop that can take up the applied nutrients,
- rainfall may not infiltrate frozen, snow-covered and/or flooded soils, and
- snow and ice melt may contribute to increased surface runoff.
Application of fertilizers and manures near water courses is accompanied with the risk of direct application of fertilizer and manures into surface waters. One of the reasons for that is the inevitable lack of preciseness of spreading equipment and the ones in charge of operating that equipment. Moreover, the indirect discharge of fertilizer and manure nutrients into surface waters through surface runoff and leaching may be also significant, especially on sloping grounds, and soils with very low infiltration capacity or just permeable soils. Unfertilized buffer strips where fertilizer and manure applications are withheld can be highly effective in this case. Unfertilised buffer strips further contribute through an increased residence time of nutrients in the field as a whole, thus enlarging the probability of denitrification (N) and retention in soil (P). If vegetated and left unfertilised, strips can also act as effective interceptors of the nutrients passing by.
The effectiveness of buffer strips is variable. Differences in width, slope, vegetative cover, and soil composition and hydrology represent some of the reasons for this. On sloping fields with relatively impermeable subsoil, water is mainly discharged via run-off and superficial flow. The effectiveness of strips acting as a filter is greater on sloping fields than when strips are established in flat landscapes on deeply drained soils. Besides, if strips are intended to remove N via denitrification, the environment needs to be conducive to that process by providing sufficient carbon substrate and by having a low oxygen concentration. When groundwater level is high and the land is drained via subsurface or surface drains, the effectiveness of buffer strips is low. In summary, buffer strips along water courses seem most appropriate whenever there is a risk of surface run-off i.e. on both sloping land and on flat land whenever the upper soil is periodically water-saturated, in particular when the discharge of water is not evenly distributed in time (e.g. summer storms, thawing snow cover. The need for buffer strips is greater if the land is tilled, managed intensively and receiving considerable inputs of nutrients.
The effectiveness of buffer strips and riparian zones has been extensively studied in COST869 (2011), and results have been summarized in among others Van Boekel (2015). Buffer strips seem to be effective measures for reducing P-loads to surface water in sloping land.
Note: For full references to papers quoted in this article see