| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Primary Industries Research Victoria (PIRVic), Dep. of Primary Industries, 1301 Hazeldean Rd., Ellinbank 3821, Victoria, Australia
b present address: Wollongong City Council, 41 Burelli St., Locked Bag 8821, Wollongong 2521, NSW, Australia
* Corresponding author (Nicole.Mathers{at}dpi.vic.gov.au).
Received for publication October 25, 2006.
ABSTRACT
Cropping is one of the many industries contributing to the excessive loading of nitrogen (N) and phosphorus (P) to rivers and lakes in Australia. Nitrogen and P exports from cropping systems have not been systematically investigated to the same extent as those from other agricultural sectors, such as dairy pastures. Therefore, this review relies heavily on information derived from agronomy and other fundamental studies on soil–nutrient interactions to determine the potential for nutrient export from high rainfall zone (HRZ) cropping. There is a great deal of variation in environmental and management strategies across cropping in the HRZ, which suggests that nutrient exports could occur under a range of scenarios. The potential for exports is therefore discussed within a conceptual framework of nutrient sources, mechanisms for mobilization, and transport pathways in HRZ cropping. Transport refers to nutrient movement by flowing water after it has been mobilized, and export refers to the transfer of nutrients from one landscape compartment (e.g., a soil) to another (e.g., a stream or lake). The transport of nutrients from HRZ cropping can occur through surface and/or subsurface pathways depending on factors such as landform and infiltration and nutrient sorption characteristics of the soil profile. Surface pathways are likely to be more significant for phosphorus. For N, subsurface movement is likely to be as significant as surface movement because nitrates are generally not bound by most soils. Information about mechanisms of nutrient mobilization is essential for developing management strategies to control nutrient exports from HRZ cropping.
Abbreviations: BMP, best management practice • HRZ, high rainfall zone • SOM, soil organic matter • WA, Western Australia
NUTRIENT EXPORTS from agriculture contribute to water quality problems in many parts of the world (Heathwaite, 1993; Granlund et al., 2005). Nutrients, particularly nitrogen (N) and phosphorus (P), exported from agricultural industries into adjacent water bodies can accelerate eutrophication, which is the natural ageing of water bodies (Sharpley et al., 2003). This "anthropogenic" eutrophication promotes algal blooms and other changes that degrade water quality and restrict its use for fisheries, recreation, industry, and drinking.
With increasing emphasis on the need to protect water quality, considerable research over the past two decades has focused on developing agricultural systems that are environmentally and economically sustainable (Cannell and Hawes, 1994; Di Pietro, 2001). In Australia these studies have often focused on dairy systems (Nash and Halliwell, 1999; Nash et al., 2004). Nutrient exports from broad-acre cropping (large-scale farming of a single crop on parcels of land >10 ha) have received far less attention.
Modern cropping systems with minimal or no tillage and retention of plant residues (i.e., conservation cropping) are often perceived to have less adverse environmental impacts than conventional cropping systems (Holland, 2004). Conservation cropping can reduce the export of sediment-bound nutrients (>0.45 µm) in runoff from farms into waterways. However, management strategies aimed at preventing the export of soil-bound nutrients may not alleviate and may even increase dissolved nutrient exports.
Dissolved nutrients (<0.45 µm) are generally more environmentally damaging than particulate nutrients because they are likely to be transported further into the water system and tend to be more bioavailable (Walton and Lee, 1972; Peters, 1981). In conservation cropping systems, dissolved nutrients can form a larger proportion of total nutrient exports than in conventional systems with deep tillage (>25 cm depth) and little residue retention (Sharpley et al., 1992; Kimmell et al., 2001). Therefore, the long-term environmental impact of all cropping systems should be investigated in more detail.
Annual cropping has been expanding in the high rainfall zone (HRZ) of southern Australia (Zhang et al., 2006), where wool and meat industries previously dominated (Poole et al., 2002). Wheat (Triticum aestivum) and canola (Brassica napus) are the current crops of choice, although there is the potential for other crops to be grown, such as barley (Hordeum vulgare), oats (Avena sativa), lucerne (Medicago sativa), and chick-pea (Cicer arietinum), with the breeding of cultivars specifically adapted to the HRZ (Zhang et al., 2006). The major climatic differences between the HRZ (Fig. 1 ) and the traditional wheatbelt are higher rainfall (annual rainfall 450–800 mm), longer and cooler growing seasons, and a longer period of high frost risk, which if managed properly could lead to higher yields of crops and the potential for increased cropping by growing dual-purpose crops for grazing and grain (Poole et al., 2002; Zhang et al., 2006).
|
|
In this paper we review the environmental and management factors affecting P and N exports in runoff from cropping with an emphasis on dissolved nutrient exports from Australian cropping systems in the HRZ. We also examine the likely sources and processes responsible for those exports and possible remedial options.
Effects of Cropping Systems on Nutrient Export
Cropping systems involve a range of interrelated biological, chemical, and physical processes that are modified by management in growing and nongrowing seasons. Some of the possible combinations of tillage and residue management are presented in Table 2 . Tillage, for example, can vary from cultivation to a depth of 20 to 30 cm (conventional tillage), to various degrees of reduced tillage, to no tillage (zero tillage). Residue management practices can be similarly variable, involving incorporation, mechanical removal, burning, or retention on the soil surface.
|
Nutrient Sources
Soil Phosphorus and Nitrogen
The total P content of topsoils in temperate regions can vary from 300 to 2000 mg kg-1, depending on the parent material, soil type, land use, and fertilizer/manure application rate and method (Leinweber et al., 2002). Soil P is often in the form of orthophosphate but can also be associated with the organic and inorganic soil components. Soil organic P can range from 25 to 50% of total P in mineral soils and can range from 60 to 90% in organic soils (Harrison, 1987). For many Australian soils, it is common for about 50% of total P to be in an organic form (Richardson, 1994). Inorganic P has been studied far more widely than organic soil P and is well reviewed elsewhere (Wild, 1949; Sanyal and DeDatta, 1991; Holford, 1997).
Soil P pools vary in their degree of extractability and availability to plants (Holford, 1997). The three basic soil P pools are (i) soil solution P, (ii) labile (readily available) P, and (iii) nonlabile (recalcitrant) P.
Orthophosphate, a common form of P in soil solution, is readily available to and used by plants (Holford, 1997). Soil solution P is in quasi-equilibrium with the labile pool, meaning that the soil solution P is nearly in equilibrium with the labile P pool and reaches a new state of equilibrium after an infinitesimal change in its properties. The labile pool of P replenishes the soil solution when the P concentration declines through plant or microbial uptake or dilution. The labile pool usually comprises P that is weakly adsorbed to inorganic soil constituents and P that is readily mineralized from soil organic matter (SOM) (Barrow, 1980; Holford, 1997). Labile P is often measured by extracting soil with various salt solutions, such as sodium bicarbonate for Colwell P or Olsen P (Colwell, 1963; Olsen, 1967; Rayment and Higginson, 1992). Nonlabile P includes the more insoluble precipitates and P that has been occluded by iron and aluminum minerals (Barrow and Shaw, 1975). In terms of determining the potential for dissolved P exports from cropped soils, common soil P tests, such as Olsen P, Colwell P, Bray P, Mehlich P, or Morgan P, are less relevant because they estimate the P available to a plant over the cropping season, which is a combination of the soil solution P and a variable proportion of the labile P (Moody and Bolland, 1999). These soil tests usually extract P from a larger portion of the soil profile than is involved in P mobilization (Nash and Halliwell, 1999). The value of extraction procedures for predicting P exports has been improved by using less aggressive extractants (e.g., water or calcium chloride) and by using ratios of P to other analytes in the extracts (i.e., P sorption saturation) (Kleinman et al., 2002).
Phosphorus is an essential component of SOM, and its degradation by microorganisms and other soil fauna releases P into the soil solution (Dalal, 1977; Stewart and Tiessen, 1987). In conservation cropping systems, organic matter accumulation in the surface layers of the soil where residues may be retained rather than burnt can be accompanied by a buildup of organic P (Kingery et al., 1996). In many agricultural systems, P has accumulated at the surface after years of P fertilizer application; it is here that some plant residues decompose. For example, more than 40 years of P fertilization in the wheatbelt of WA has resulted in areas that have nutrient concentrations that could pose a risk for water contamination (Wong et al., 2000). Moreover, because conservation cropping practices and stubble retention are likely to result in greater P accumulation at the soil surface (stratification), the risk of dissolved P exports from these systems increases. This is especially true because SOM, which may also increase in this zone, can interfere with P sorption (Scherer and Sharma, 2002). Soil organic matter can affect P sorption by increasing (Saunders, 1965), decreasing (Toreu et al., 1988), or remaining unchanged (Moody and Standley, 1979). The net effect of SOM on P sorption may depend on the nature of the organic matter and how much iron and aluminum is associated with it (Moody and Bolland, 1999).
The major forms of organic P in soils include the inositol phosphate ester, inositol hexaphosphate, and to a lesser extent the di-, tri-, and tetraphosphates of inositol in various combinations with fulvic and humic acids. The high charge density of inositol phosphates causes them to be strongly adsorbed to reactive surfaces (e.g., on clays), thereby protecting them from microbial attack (Harrison, 1987; Turner et al., 2002) and lessening their contribution to plant available P (Richardson et al., 2000). Phosphatase enzymes catalyze the hydrolysis of labile forms of organic P to release P as inorganic orthophosphate. Microbial phosphodiesterase catalyzes the hydrolysis of phosphodiesters to phosphomonoesters, which are then hydrolyzed to orthophosphate by phosphomonoesterase enzymes (Turner and Haygarth, 2005). The degradation of phosphodiesters to orthophosphate is therefore dependent on the presence and relative activities of the two microbial enzymes (Turner and Haygarth, 2005).
Like soil P, N is a major component of SOM, which is the major store of N in most uncultivated soils. During decomposition of SOM, ammonium is released into the soil solution and is converted by microbes to nitrite and then to nitrate. This process has a profound impact on nutrient exports because cationic ammonium is held on exchange complexes (van Olphen, 1977), whereas anionic nitrate poorly adsorbs to soil, making it more susceptible to leaching (Strong and Mason, 1999). Factors that can exacerbate nitrate movement include failure of the plant root system to penetrate subsoil and preservation of macropores, particularly in zero-tilled systems, which increases preferential displacement of nitrate (Dalal, 1992; Strong and Mason, 1999). Soil organic matter commonly has a C to N ratio of around 10:1 and contains about 5% N, approximately 95 to 99% of the total soil N (Price, 2006). For example, a soil with an organic matter content of 1.2% may have 0.1% N, equivalent to 1.6 tonnes of N per hectare (assuming a bulk density of 1.6 g cm-3 and a depth of 10 cm). When this amount is compared with the amount of N added as fertilizer (e.g., 50–100 kg ha-1), the N present in SOM seems to be more than adequate to meet crop requirements. However, in practice, the release of N from SOM may be too slow to meet the entire need of the crop, especially at vulnerable growth stages (Strong and Mason, 1999). This is because N release from native SOM is considerably slower than that from fertilizers, such as urea (also known as carbamide), or from more easily decomposed plant material, such as leguminous plant residues (Wang et al., 2004).
Phosphorus and Nitrogen Fertilizers
Natural reserves of P in agricultural soils are often inadequate to support commercial agriculture; fertilizer amendments are therefore necessary. Various types of P fertilizers are available and can be classified as (i) water soluble, (ii) partially water soluble, and (iii) relatively water insoluble. The relative solubilities and reactions of different P fertilizers in each of these categories are well described elsewhere (Nash and Halliwell, 1999). Although P from slow-release fertilizers can be mobilized, greater P mobilization is likely to occur if rainfall occurs soon after the application of water-soluble fertilizers such as monocalcium phosphate. In pastures, large increases in the concentration of P in runoff have been recorded close to the time of fertilizer application, and in one study, up to half the annual export occurred in the first irrigation after fertilizer application (Nash et al., 2000; McDowell and Catto, 2005). Studies of this nature are not common for the Australian cropping sector, but one can surmise that the pathways for P exports are likely to depend on fertilizer type and placement (i.e., surface, subsurface, incorporated), tillage practices (Lorimer and Douglas, 2001; Bünemann et al., 2006), and cropping systems, such as on-the-flat versus raised-bed cropping (Torbert et al., 1996).
In most commercial cropping systems, the natural supply of N is also insufficient to meet crop demand, and N fertilizers are therefore applied. In Australia, common N fertilizers include urea, di-ammonium phosphate, and mono-ammonium phosphate. One third or more of N applied to a crop can be lost if urea is leached too deeply into the soil by rainfall. Urea may be applied at times when the potential for export in surface runoff is high (i.e., before 10 mm rainfall) (Palta et al., 2003). Di-ammonium phosphate and mono-ammonium phosphate are more commonly applied subsurface at sowing and are therefore less susceptible to export through surface pathways. Exports of N from urea under optimum management, however, can be much lower than those from other N fertilizers, such as ammonium sulfate and calcium nitrate (Craswell, 1979).
Isotope-labeled (15N) fertilizers are often used to distinguish fertilizer N from other N sources (Craswell, 1979; Fillery, 2001), especially in soils with contrasting texture (Fillery and McInnes, 1992). In red-brown earths from southeastern Australia, between 10 and 40% of the applied N could not be accounted for in the soil–plant system 85 d after sowing and was presumed to be lost through denitrification. The form of fertilizer, as ammonium sulfate or urea, and the placement or timing of application did not seem to affect the deficit (Smith et al., 1989; Fillery and McInnes, 1992). In the sandy soils of WA, between 17 and 70% of fertilizer N applied as urea was lost, and the role of denitrification was less clear (Fillery and McInnes, 1992). However, transport of N in interflow can occur (Nash et al., 2002); interflow is the lateral motion of water through the upper layers until it reaches a stream channel or returns to the surface at some point downslope from its point of infiltration. It was hypothesized that most N was lost as a result of fluctuation of the watertable and the subsequent lateral movement of water in the soil profile (Fillery and McInnes, 1992). Exports of N in surface runoff are rarely considered in such studies.
Because nitrate is more mobile than ammonium in most soils (Strong and Mason, 1999), nitrification inhibitors have been used to lessen N availability and therefore N losses (Di and Cameron, 2005). Their effects are most likely to be greater on soils that are N rich and where the N losses due to leaching and nitrification are large (Edmeades, 2004). Fertilizer urea can be less efficient than nitrate-type fertilizers; the major reason for this is that soil pH in the vicinity of urea granules increases as a result of hydrolysis, facilitating the volatilization of ammonia to the atmosphere (Watson et al., 1990; Edmeades, 2004). The benefits of treating urea with urease inhibitors are variable and dependent on the same variables that control ammonia volatilization (Edmeades, 2004). It cannot be assumed that a reduction in ammonia volatilization translates into an increase in crop yield. These inhibitors have not been widely adopted by the Australian cropping sector, presumably because of cost considerations.
Plant Residues
Cropping produces residues (above and below ground) that are major sources of N and P (Baldock and Ballard, 2004) and that at low concentrations can immobilize native soil N and P (Fuller et al., 1956; McLaughlin and Alston, 1986). Wheat (Triticum aestivum) is one such crop that can leave large amounts of residue after harvesting, and stubble loads of up to 8 t dry matter ha-1 are common (Butler, 2004). In some regions, residues are mechanically removed or burnt in situ, an option that is commonly used in conventional tillage systems. In addition to altering the pool of available nutrients, burning can make nutrients more available for mobilization and transport by changing their particle size, surface area, and chemical form. Residues can also be left standing in the field to slowly decompose during establishment of the next crop, or they can be slashed and left on the soil surface or incorporated into the soil (Butler, 2004). Residue retention can assist in maintaining or improving structural stability, soil fertility, and water availability (Heenan et al., 2004) by reducing erosion. Despite containing low nutrient concentrations, stubble can be a source of nutrient exports with surface runoff in conservation cropping systems where stubble is not incorporated in the soil but left on the surface. In these situations, nutrients released or leached from stubble may enter runoff directly without any interaction with the soil matrix (Bünemann et al., 2006). In the HRZ, however, stubble is often burnt, which can result in significant losses of N though volatilization. Burning can release the P in stubble into the soil environment, and this can affect the P export potential. The transformations of this source of P and its significance have not been investigated but might be significant if rain falls soon after stubble burning.
Inputs of P nutrients to plant crops are essential to maintain profitable crop production and meet global food requirements (Sharpley et al., 2000). In cropping systems, a proportion of these nutrient inputs can be retained in the system and contribute to subsequent nutrient exports. This usually occurs with biological decomposition of plant residues but also with leaching from live plant tissues (Sharpley, 1981). Physical damage to plants that may occur during the early stages of crop growth would be expected to increase nutrient exports and this can be significant (Nexhip et al., 1997; Mundy et al., 2003). Where stubble is not incorporated into the soil but retained on the soil surface, nutrients can be released or leached directly without any interaction with the soil matrix. Retention of stubble can be effective in reducing particulate nutrient exports by decreasing erosion, but when they are subjected to rainfall leaching, the residues are a significant source of dissolved nutrients (dissolved reactive P and nitrate N) to runoff (Schreiber, 1999). In a North American study using wheat straw containing 0.35% N and 0.04% P, nutrient mobilization was dependent on rainfall intensity, straw source, and stubble loads (Schreiber, 1985; Schreiber and McDowell, 1985). Although the straw N content was greater than P, P concentrations in the leachate were always greater than N, indicating that P in wheat straw is more mobile than N. Between 10 and 40% of P in the wheat straw was removed after 25 mm of rainfall compared with less than 1% of N (Schreiber and McDowell, 1985). Differences in nutrient exports can also occur between crops. For example, although the percentage of N leached from corn (Zea mays L.) residues was the same as that for wheat, the amount of P leached from corn decreased to between 3 and 6% (Schreiber, 1999). Variation in the quantity of nutrients leached from different straws suggests that results from these studies cannot be directly applied to the Australian situation, where different cultivars have been adapted to the various growing areas. Therefore, further investigation is necessary, particularly for P exports.
Leguminous Nitrogen.
Legume rotations are an important component of many cropping systems. The amount of N fixed by leguminous plants can vary considerably, but the upper limits of the ranges in Table 3
suggest that these quantities can be substantial, especially in systems with high potential productivity, such as the HRZ of Australia.
|
Nitrogen and Phosphorus Export Pathways
Nitrogen and P are exported from cropping systems through surface and/or subsurface pathways depending on factors such as landform, soil infiltration rates, and the sorption characteristics of soil layers within the soil profile. The nutrient export pathway affects nutrient concentrations and loads by exposing different sources of the nutrient, enabling demobilization processes (i.e., adsorption) to remove nutrients from drainage.
Surface Pathways
Water and nutrients can move over or through the soil surface as (i) infiltration-excess overland flow, (ii) saturation-excess overland flow, or (iii) interflow (Fig. 2
). Infiltration-excess overland flow occurs when water reaching the soil surface as rainfall or run-on exceeds the infiltration capacity of the underlying soil. The soil surface may be naturally impermeable, or its infiltration capacity may be impaired through cultivation and cropping. Saturation-excess overland flow occurs when water is able to infiltrate the surface layer, but subsurface layers restrict further infiltration. With saturation of the surface layers, water usually re-emerges in "wet areas" and flows onto the soil surface. This situation particularly applies to texture contrast soils or where subsurface compaction has occurred. Interflow occurs where water moves laterally through the surface soil. For example, some soils in WA are shallow sands overlying slowly permeable, lateritic ironstone gravel or clay. In these soils, where subsurface drainage is restricted, P is known to move laterally (Bolland et al., 1999a). Prefixes are used to define the zone in which interflow is moving. For example, A/B interflow occurs at the junction of the A and B horizons. Interflow often reappears on the surface with changes in the landform.
|
Subsurface Pathways
Water and nutrients may move through subsurface pathways by matrix flow (saturation excess overland flow) or macropore flow (Fig. 2). Matrix flow is an important nutrient transport pathway, particularly in highly permeable soils. Nutrient transport in matrix flow has been amply demonstrated for some of the sandy soils found in the HRZ of WA (Weaver et al., 1988). In general, matrix flow results in the transport of N, but, with the low P retention capacity of the sandy soils in WA, there is also potential for P exports (Gerritse, 1995). Despite the potential for nutrient exports in matrix flow, their catchment scale impact is uncertain.
Macropores are internal channels that facilitate the rapid movement of water and dissolved materials through the soil profile with little interaction with the soil matrix. Phosphorus mobilization studies using intact soil cores have demonstrated the importance of preferential flow within the soil profile (Cox et al., 2000). Even heavy (high clay content) soils that were not P-saturated showed some vertical movement of P due to the presence of macropores. The importance of macropore flow in nutrient transport over landscape scales has also been demonstrated (Stevens et al., 1999; Cox et al., 2000).
Factors Affecting Nutrient Export Pathways
The extent to which the different flow paths are responsible for mobilizing and transporting nutrients depends on the infiltration depth and the distribution of nutrients within that depth. For example, if fertilizer is applied on the surface of a soil that is prone to crusting and it is not incorporated into the soil, then it will most likely be retained on the soil surface. Under these circumstances, most nutrients are likely to be mobilized by and transported in infiltration excess overland flow. However, if fertilizer is placed or washed into the root zone before overland flow occurs, then saturation excess overland flow could dominate nutrient movement. Fertilizer and cropping practices can vary markedly, such that the exact transport processes that operate in any particular situation are dependent as much on soil physicochemical properties as on management factors.
Surface crusting can affect flow patterns in cultivated soils. Crusting, or surface sealing, can develop with cultivation in most soils as a result of aggregate breakdown with successive cycles of wetting and drying and frequent mechanical disturbances (Bresson, 1995). This effect is more pronounced in sodic soils (i.e., those soils with high exchangeable sodium cations) (Sumner, 1993) with inherent structural stability problems. Surface crusting restricts water infiltration (McKenzie et al., 1993) and therefore favors surface rather than subsurface nutrient exports. Water repellent layers that form on the soil surface due to the release of waxes and other hydrophobic substances from plants and fungi (McKissock et al., 1998) have a similar effect.
Conservation cropping can lessen surface sealing by protecting the soil surface from mechanical disruption by rain drop impact (Li et al., 2004). Presumably this is partly why some studies suggest that runoff volumes are lower under reduced tillage and residue retention systems (McDowell and McGregor, 1984). Other studies have shown similar or more runoff from zero tillage than reduced tillage systems or conventional systems (Richardson and King, 1995; Kimmell et al., 2001). The expectation is that conservation tillage and residue retention practices would improve soil porosity (Vogeler et al., 2006) and increase water infiltration into the soil, leading to decreased runoff volumes. However, increased porosity under zero tillage does not necessarily translate to increased infiltration (Lipiec et al., 2006). It may be that although hydraulic conductivity (i.e., the rate of water movement through soil) can be higher under zero tillage systems, a surface seal develops that is not physically disrupted by tillage, thereby lowering infiltration in subsequent events.
Subsoil compaction from the use of heavy machinery in farm operations is another factor affecting soil hydrology and varies according to soil texture, soil structure, soil mechanical strength at different depths, pore structure, and soil moisture when farming operations occur (Lipiec and Hatano, 2003). Compaction lowers subsoil hydraulic conductivities and encourages exports by surface pathways. Compaction can occur at depths as shallow as 5 to 15 cm (Silburn et al., 2003) and may account for little improvement in infiltration under conservation cropping (i.e., zero tillage) (Kimmell et al., 2001). Although subsurface compaction is natural in many parts of Australia, mechanically induced compaction has significantly reduced time to ponding, steady infiltration rate, and total infiltration compared with nonwheeled soil, with or without residue cover (Li et al., 2001).
Surface pathways are likely to be more significant for P transport because matrix flow generally lowers P concentrations. Exceptions include sands, some organic soils, and soils with significant macropores. The lack of subsurface movement of P is evident in several studies that have shown that P is concentrated in the surface layers of the soil even after many years of fertilizer application. For example, in the acid soils of Queensland with high Fe and Al contents, P was not transported beyond a depth of 500 mm even after several decades of fertilization (LWRRDC, 1998). Similar results have been achieved in pasture soils in southwest Victoria (McCaskill and Cayley, 2000).
Surface and subsurface pathways are likely to be important for N transport. Situations where nitrates are located away from the immediate soil surface favor subsurface pathways (Craswell, 1979; Stevens et al., 1999). However, especially in texture contrast soils, N that is initially mobilized and transported subsurface may re-emerge as surface flow lower in the topographic sequence.
A technology that has become increasingly popular in Australia's HRZ is the use of formed, raised beds for planting broad-acre crops (Zhang et al., 2006). The beds are primarily used in situations where soil drainage is poor. In addition to improving soil drainage, vehicular traffic in these systems is restricted to the drainage lines (Fig. 3 ). In conventional "on-the-flat" cropping systems nutrients move to the surface via diffusion against a mass flow gradient into the soil or via mass flow where interflow or subsurface water is being discharged. In raised bed systems, the nutrients are transported by mass flow to the drainage lines from where they are efficiently transported off-site because infiltration in the channels is decreased by compaction of the soil from vehicular traffic. Consequently, there may be increased potential for N and P exports from raised bed systems compared with on-the-flat systems. However, this is dependent on rainfall, whereas increased biomass production in the raised beds may decrease available moisture and water exports overall because of increased plant uptake.
|
The bioavailability of nutrients exported in the particulate form can be highly variable (Sharpley et al., 1992). Dissolved P can attach to suspended material after mobilization even during storage (Sharpley et al., 1981). Consequently, P can enter a stream in dissolved form, attach to sediment, and become particulate P shortly afterward (Sharpley et al., 1981). The dissolved P fraction of P exports comprises mostly inorganic orthophosphate ions, which are immediately available for uptake in aquatic environments (McDowell and Sharpley, 2001). However, the presence of colloidal P, especially in drainage from dispersive soils, although not immediately bioavailable, may influence the environmental behavior of nutrients (Haygarth et al., 1997; Heathwaite et al., 2005). Dissolved P exports from overland flow can often exceed stream target values of 0.1 mg P L-1 (Cox and Pitman, 2001; Nash et al., 2003).
Nitrogen Exports
Nitrogen is exported in many forms, including nitrate, nitrite, ammonium, dissolved organic components, and colloidal material (Hagedorn et al., 2001). Nitrate is often the most important species in the dissolved fraction. As for P, conventional tillage seems to have higher particulate N exports than reduced tillage systems. For example, on a heavy clay soil of 1 to 3% slope under a rotation with wheat, sorghum (Sorghum bicolor), and corn, mean annual particulate N exports were 0.3 and 2.5 kg ha for the zero tillage and conventional tillage treatments, respectively (Richardson and King, 1995). Similarly, on a silt loam soil of 5% slope under corn, the exports of particulate N were higher from conventional tillage than from zero tillage systems (McDowell and McGregor, 1984). In that study, particulate N exports were greater than dissolved N exports in both tillage systems, whereas in the heavy clay soil (Richardson and King, 1995), the reverse was true. The reason for this difference could be the erodibility (Wischmeier and Smith, 1978) of the silt loam compared with that from the heavy clay soil.
As was the case with P, it seems that measures to lessen particulate N exports may lower nutrient loads but increase the relative potency (i.e., impact per kilogram of nutrient accession) of the remaining exports. However, reduced tillage does not always increase dissolved N exports (McDowell and McGregor, 1984; Richardson and King, 1995). In many situations, however, dissolved N exports from cropping systems are greater than the lowland river trigger value of 500 µg L-1 (ANZECC, 2000).
Mitigation Measures
Measures to lessen nutrient exports can generally be categorized as (i) source control and (ii) transport interruption. In the past, the second category has been pre-eminent (Dillaha et al., 1989), with extensive literature available on reduced tillage systems and vegetative buffer strips that trap nutrients and thereby interrupt transport (McDowell and McGregor, 1984; Crozier et al., 1999). However, more recently, efforts are being directed toward source management by optimizing fertilizer inputs and timing of application to control dissolved exports, particularly in surface runoff (Sharpley et al., 2001).
Improved fertilizer management practices, such as the use of more appropriate compounds and formulations, better timing of fertilizer application to avoid high-risk periods, and the placement of fertilizer to limit mobilization, are important goals. The physical and chemical properties of P fertilizers vary, and, as has been demonstrated for pastures, these properties can affect P exports (Nash and Halliwell, 1999; Nash et al., 2004). In the cropping sector, fertilizer choice and placement can have implications for crop yield. Placement of fertilizer below the seed rather than with the seed was more effective for seed germination of wheat and barley (Radford et al., 1989) and establishment of canola (Brassica napus L.) (Hocking et al., 2003). For other crops, such as faba beans (Vicia faba L.), drilling or banding of P fertilizers has no effect on yield (Bolland et al., 2001). Fluid fertilizers enhanced crop yield and P uptake when compared with granular fertilizers applied at the same rate (Lombi et al., 2004), and soil chemical analysis indicated that liquid forms of P were not correlated with total soil P concentration or P availability as measured by Colwell-P (McBeath et al., 2005). However, dissolved P was increased threefold in surface runoff after granular application of fertilizer compared with that of liquid application at the same rate (Sharpley and Syers, 1983). Moreover, the addition of any effective source of P increases the P in soil water that is available for plant uptake but also raises it above the threshold for exports into waterways.
In pasture systems, some slow-release P fertilizers have been shown to be as effective as more soluble fertilizers. For example, reactive phosphate rock was shown to be as effective as triple superphosphate for some pastures in very acid soils in the HRZ, although at other sites this was not the case (Sale et al., 1997). In other studies, partially acidulated rock phosphate was as effective as single or coastal superphosphate (a granulated mix of superphosphate, rock phosphate, and elemental sulfur) for clover production on sandy, humic podosols (Isbell, 2002) or arenic alaquods (Soil Survey Staff, 2006) in the HRZ of southern WA (Bolland, 1995). However, monocalcium phosphate has been found to be a more effective source of P than rock phosphate and partially acidulated rock phosphate for wheat (Kumar et al., 1993). The need to increase fertilizer application rates to achieve similar yields may negate some beneficial effects of using less-soluble P fertilizers.
The interval between fertilizer application and water application also affects nutrient export potential (Tan et al., 2002; Nash et al., 2004). For pasture systems, best management practice (BMP) guidelines on the timing of fertilizer application in relation to the likelihood of rain have been developed for farmers (Government of Victoria, 1995; Waters, 1996) and fertilizer distributors, such as the Australian Fertilizer Services Association and Fertcare. The flexibility with timing of fertilizer application in pasture systems may not be available in the cropping sector. Fertilizer is crucial in the initial growth stages of the crop, and leaving the fertilizer in contact with the soil for too long before sowing can lessen its effectiveness through immobilization reactions in the soil (Bolland and Barrow, 1996). Also, fertilizer application rates can vary for different crops because they have different nutrient requirements, making a singular fertilizer recommendation for all crop types more difficult to encourage or enforce (Bolland et al., 1999b).
Given that there is limited information on which to develop a BMP for fertilizers in cropping systems, particularly given the episodic nature of rainfall and the associated variation in plant nutrient requirements, mobilization, and transport potential, BMPs are likely to vary within and between years, necessitating a probability distribution approach (e.g., Bayesian Networks) (Robertson and Wang, 2004) to their development. The flexibility of fertilizer source reduction could be investigated using such systems that define the causes and effects of decisions in probabilistic terms.
The use of soil amendments to increase the nutrient retention capacity of soils can decrease net nutrient mobilization and, hence, nutrient exports. For example, bauxite has been added to sandy soils to improve their capacity to retain P fertilizers and thereby decrease exports in leachate (Summers et al., 2001). Polyacrylamide can be used to decrease particulate and colloidal nutrient exports, although its economic viability is uncertain (Soupir et al., 2004). Other authors advocate the control of soil acidity to restrict nitrification and retain N fertilizers in the less easily transported ammonium form (Kemmitt et al., 2005). This strategy has worked better in grasslands than in arable systems.
Amending soil through physical changes can also lessen nutrient exports. Using artificial drainage and deep ripping to ensure a closer interaction between nutrients and the soil matrix has been used for the control of P exports, especially in surface pathways (Haygarth et al., 1998). However, such strategies have also been shown to enhance the export of N as nitrates through subsurface pathways (Zucker and Brown, 1998).
Another proposal for restricting nutrient transport has been the development of cropping systems that are more efficient at using water so that the systems are more "leak proof" (Connor, 2004). Restricting the water exiting the cropping paddock by the use of dams could be similarly effective. Given that the wheat zone of southeastern Australia has been shown to have runoff that varies from 1 to 6%, drainage that varies from 0 to 18%, soil evaporation that varies from 37 to 81%, and transpiration of approximately 53% (Sadras, 2003), such proposals need to consider the episodic nature of nutrient exports and the effects of these proposals on downstream water users.
A number of other mechanisms aimed at intercepting nutrient transport are used in the cropping industry. Buffer strips have been shown to lessen the exports of particulate nutrients (Hairsine, 1996) and to a lesser extent dissolved nutrients where total water volumes are decreased through infiltration in the buffer strip (Dillaha et al., 1989). However, the effectiveness of these measures for episodic events is questionable (Nash et al., 2000; Nash et al., 2004).
Nutrients in drainage water can also be attenuated by wetlands and other systems through which the drainage water is channeled (Wen and Recknagel, 2002). Wetlands can be a sink for nutrients if their size, location, and overall design are appropriate (Raisin et al., 1997). However, wetlands may provide only a short-term solution because nutrients could be released at a later stage (Richardson, 1985). These issues may not be apparent in short-term studies, and longer-term investigations are required to evaluate the real benefits of wetlands and other measures as remedial strategies to address nutrient export from cropping systems.
Conclusions
The HRZ receives more rainfall than other cropping areas in Australia; therefore, the likelihood of water and associated nutrient exports is a concern within this zone. Because water availability is generally not a constraint for crop growth in the HRZ, cropping in these areas could intensify as a land use activity in the future. Nutrient exports and their off-site impacts could then become a more serious issue in terms of catchment-scale water quality, including some catchments of the HRZ such as the Glenelg-Hopkins and Corangamite catchments. The lack of adequate information on nutrient exports from cropping in the HRZ of Australia makes it difficult to determine the significance of cropping on water quality impacts in catchments in the HRZ. At this stage, what can be determined is only the potential for nutrient exports on the basis of information derived from nutrient export studies conducted for agronomic purposes or from more fundamental research on the effects of cropping on the physical, chemical, and biological properties of soils. In this review we have shown that nutrient exports can occur in the HRZ where environmental and/or management factors can create situations that predispose nutrients to export.
Management initiatives to address nutrient exports and impacts from the HRZ cannot be based on the potential for nutrient exports because the information is inadequate but should be based on actual measurements. Further studies are necessary that provide unequivocal information about the various environmental and management scenarios that predispose cropping systems to nutrient export. In addition to investigating nutrient exports from cropping systems at the paddock scale, catchment-scale studies are necessary to appreciate the role of transport. New and emerging technologies, such as highly adapted cultivars, new biological products to improve soil health, and innovative machinery, can be explored that could provide better insight on nutrient mobilization and transport processes. Only then can meaningful progress be made in Australia to understand and to address the issue of nutrients and water quality in the HRZ cropping areas. Willingness to act without delay will provide the best opportunity to improve water quality in areas that are already affected or to prevent water quality degradations on the scale that has been experienced in other parts of the world.
ACKNOWLEDGMENTS
The Grains Research and Development Corporation and the Department of Primary Industries Victoria provided funds for this work. The authors thank Dr. Fiona Robertson, Dr. Roger Armstrong, and Professor Aldo Bagnara of the Victorian Department of Primary Industries and two anonymous reviewers for providing suggestions and comments on previous versions of the manuscript.
NOTES
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
