Identifying fertiliser management strategies to maximise nitrogen and phosphorus acquisition by wheat in two contrasting soils from Victoria, Australia.
Nitrogen (N) and phosphorus (P) fertilisers are a key component of dryland grain production systems in southern Australia, where many soils are inherently low in fertility (Perry 1992). Fertiliser is the single largest variable cost for dryland grain production in this region (ABARE 1999), and decisions about the quantity, timing, and placement of fertiliser (especially N) can significantly affect crop yield and profitability.
Ideally, nutrient supply to crops should aim to achieve both temporal and spatial synchronisation between nutrient supply and crop demand (Myers 1987; Crews and Peoples 2005). Decisions about whether to apply fertiliser, and how much to apply, broadly depend on the expected crop demand (strongly linked to yield expectation and therefore seasonal conditions) and the ability of the soil to meet this demand. A range of edaphic (physicochemical and biological), seasonal, and crop physiological factors can affect how efficiently the crop will utilise these applied nutrients. McDonald (1989) suggested that new fertiliser technologies, including formulation (e.g. controlled release N, Shaviv 2001; and fluid sources, Holloway et al. 2001) and new application strategies, will improve nutrient use from fertilisers by dryland crops. Application strategies can vary the timing, placement, and quantity of fertiliser to be applied. The strategy chosen will be influenced by a range of factors including availability of appropriate machinery, soil and crop type, rotation sequence, the expected supply of residual and mineralised nutrient (Lory et al. 1995), and anticipated return on fertiliser investment. In semi-arid zones such as the Victorian Wimmera and Mallee, rainfall, which can be highly variable within and between seasons (Connor 2004), strongly influences the choice of strategy. An example of a change in fertiliser strategy in the Wimmera is the trend away from the traditional practice of incorporating urea prior to sowing to tactical topdressing applications after emergence. This has been driven by a series of poor (dry) seasonal conditions since the late 1990s, and aims to reduce financial risk by delaying application to a time when more information is available about the current season.
The 2 largest influences on nutrient use by dryland grain crops in the Wimmera--Mallee region are soil type and seasonal (rainfall) conditions. Grain production in Victoria is dominated by alkaline soil groups such as the Calcarosol and Sodosol soils that predominate in the Mallee, and Vertosols in the Wimmera region. Calcarosols and Sodosols often contain a range of physicochemical constraints to root growth in the subsoil (Nuttall et al. 2005). Crop growth and grain yield are affected by nutrient deficiencies and toxicities, particularly boron, carbonate, and aluminate, as well as high levels of salinity, sodicity, and alkalinity (Holloway and Alston 1992; Rengasamy 2002; Rodriguez et al. 2006). The benefit of boron-tolerant crop cultivars can be negated by the strong effect of salinity and sodicity on subsoil root growth and hence subsoil water uptake (Nuttall et al. 2003a, 2005). In contrast, the Vertosols of the Wimmera have high physical fertility, rarely containing the high level of physico-chemical constraints that typify the Calcarosols and Sodosols. They are, however, naturally deficient in both N and P. Rainfall is highly variable (within and between seasons) in both regions, although there is a pronounced trend of increasing rainfall southward from the Mallee (annual rainfall 325-375 mm) to the Wimmera (375-450 mm). The differences in the chemico-physical properties of these 2 soil groups, and the broad differences in rainfall, suggest that different fertiliser management strategies may be required for each soil type/region.
Spatially and temporally synchronising nutrient supply (dependent on soil type and fertiliser placement) with plant demand (determined principally by season) is important for improving the nutrient use efficiency of dryland crops. The major elements for plant growth, particularly soil P and N (Tinker and Nye 2000) do, however, contrast in availability and mobility, providing a challenge. Phosphorus is an immobile nutrient, particularly in highly P-buffering soils, with its availability typically restricted to topsoil organic layers and zones of fertiliser placement. Nitrate, on the other hand, is relatively mobile and can displace from mineralisation sources, or zones of fertiliser placement, relative to rainfall patterns over a season. This disparity in mobility can lead to differences between the spatial location of soil N, P, and water over time. The challenge for improving nutrient use efficiency is to more closely match these potentially spatially disparate soil resources with the active uptake zones of the crop root system over the growing season.
Decisions about fertiliser management need to account for the complex interactions between crop nutrient demand, and supply from soil and fertiliser sources. Mechanistic simulation models such as ROOTMAP (Diggle 1988; Dunbabin et al. 2002a) can rapidly assess the likely outcomes of different management strategies for a complex range of both abiotic (soil chemo-physical properties and water) and biotic factors that affect the fertiliser response of crops. Using the ROOTMAP model, the aim of this study was to identify strategies that may improve the utilisation of N and P by wheat, across a range of seasonal conditions and for 2 major cropping soils in Victoria. The simulations were based on the assumption that once a decision has been made to apply fertiliser (and how much), grain growers have control over 2 variables: (i) the timing of nutrient supply (for N); and (ii) the spatial position (for N and P). Knowledge gained from this analysis is being used to guide subsequent controlled and field experimentation to validate the findings.
Materials and methods
The ROOTMAP model
ROOTMAP is a model of 3-dimensional (3D) root growth, and is described in detail in Diggle (1988) and Dunbabin et al. (2002a). Briefly, the model simulates soil water and nutrient dynamics, and root growth responses to those dynamics. Solutes are transported to the root surface by mass flow and diffusion, as described by the advection-dispersion equation. Solute concentrations at the root surface are derived from local bulksoil values using the Baldwin et al. (1973) model (see also Dunbabin et al. 2002a). Plant capacity to take up ions at the root surface is described using Michaelis-Menten kinetics. The 3D spatial resolution of local soil properties, including local soil values of soil water and nutrient contents, can be varied by changing the grid resolution for the simulated rooting zone. In addition to mass flow and diffusion, ions in solution phase can be transported (leached) after rainfall events by bulk water flow. Water loss through evaporation from the soil surface is a function of pan evaporation, effective crop cover (see below), and local soil properties. The water uptake from each local soil volume is a function of pan evaporation, effective crop cover, local soil properties, and local root surface area, and is based on the Feddes sink term (Feddes et al. 1976). The redistribution of water in 3D space is described by Darcy's law.
To date, simulation studies using ROOTMAP have focused on root growth, root architecture and the theory of root/soil interactions (Dunbabin et al. 2002a, 2002b, 2003, 2004; Dunbabin 2007). Below-ground resources were assumed to be the greatest limitation to plant growth, and leaf area index and root : shoot ratios over time were created for the model from associated field or experimental studies. In the current study, ROOTMAP was coupled with the Woodruff-Hammer wheat model (Hammer et al. 1987). This was done to enable the generation of a shoot growth and shoot demand function that would interact with the root system, responding to seasonal conditions and the capture of soil resources. The Woodruff Hammer model calculated shoot dry matter accumulation, leaf area index, and effective crop cover over the season. The wheat model was parameterised with phenological data from previous field and simulation studies (O'Leary et al. 1985; O'Leary and Connor 1996) of wheat growing at Dooen in the Wimmera (one of the field sites used here; see Table 1). The field trial of wheat grown at Dooen over 2 seasons (O'Leary et al. 1985) was modelled as a test of whether the Woodruff-Hammer wheat model had been coded and linked to ROOTMAP properly and was providing realistic predictions. The model reproduced well the biomass, leaf area index, and water uptake observed in the field experiment ([R.sup.2]= 0.84, data not shown; O'Leary et al. 1985). This indicated that the model was functioning properly and could be used in this study as a shoot biomass and leaf area index generator that interacted dynamically with the root model and responded to soil type and season.
The phosphate routine in the model has not been previously described. The approach of Mendham et al. (1997) was adapted for modelling the reactivity of the labile phosphate solid-liquid phases. The phosphate adsorption isotherm for each soil type is described using the Freundlich equation:
where y is the labile solid phase P ([micro]g/g) in equilibrium with the liquid phase P, x ([micro]g/mL) and a and n are coefficients. The transfer of soil P between the labile solid and liquid phase pools is described by the solid-liquid partition coefficient ([K.sub.d]), which is calculated from the slope (first derivative) of the Freundlich isotherm for a given solution P concentration:
[K.sub.d] = dy/dx = [anx.sup.n-1] (2)
The coefficient a is equal to the phosphorus buffering index given in Table 1 (Burkitt et al. 2002). In ROOTMAP, local solution phase P concentrations constantly change due to plant uptake of P, mineralisation of organic matter, and the transport of P via diffusion and saturated and unsaturated mass flow. As a consequence, the partition of labile P between the solid and liquid phase pools is recalculated at every time step for every local soil volume in 3D space, maintaining the P mass-balance. The value of the phosphorus buffering capacity (b), used in the Baldwin et al. (1973) model, is calculated at each time step and for every local soil volume in 3D space using the following expression (Van Rees et al. 1990):
b = [theta] + [rho][K.sub.d] (3)
where [rho] is the local soil bulk density, and [theta] is the local volumetric soil water content.
ROOTMAP uses a feedback approach to simulate responsive root growth. Plant demand for soil resources (nitrate, phosphate, and water in this study) is balanced against the capacity for each individual root segment to supply those resources. This drives the allocation of assimilates to root segments for branching, maintenance, and nutrient uptake, and to root tips for root growth. Each additional unit of soil resource acquired by the simulated plant, particularly the most growth-limiting resource, enables the plant to generate further assimilate for future growth. This feedback-based approach enables the model to simulate both whole root system growth responses to water and nutrient status, and also local root proliferation and nutrient uptake responses to nutrient patches (Dunbabin et al. 2002a).
For this study, the existing code was modified to enable the model to track the proportion of N and P absorbed by the crop from a fertiliser source v. the background soil source. This was done using a simple nutrient accounting approach. For these simulations, the soil profiles had a predefined N and P content (Table 1) at Day 1, representing background soil N and P. Nutrients could be lost from this pool by plant uptake (after sowing) and leaching. Nutrients could be added to the pool through mineralisation and transported through 3D space by the processes of mass flow and diffusion. On the day of sowing, the fertiliser N and P pools were initialised and placed spatially in the 3D grid of soil properties. The total size of the fertiliser pools could not be increased except through additional fertiliser applications later in the season. Again, the fertiliser pools could be reduced in size by plant uptake or leaching. Fertiliser P could also transfer to the recalcitrant non-labile pool, which was one common pool for both background and fertiliser P. Fertiliser N and P could be transported through 3D space by the processes of mass flow, diffusion, and leaching, so that local parts of the soil volume that previously only contained background N and P could gain N and P from a fertiliser source. The nutrient mass balance is continuously calculated as a check on the functionality of the nutrient accounting code. In any one local soil volume, plant roots could not distinguish between labile P from the fertiliser v. the background soil pool. All labile P was equally available for plant uptake.
Mineralisation of soil N was calculated using the potentially mineralisable N approach (Campbell et al. 1981):
[N.sub.t] = [N.sub.o][1 - [e.sup.-kt]] (4)
where [N.sub.t] is the cumulative N mineralisation at time t. [N.sub.o] is the potentially mineralisable N (AMN in Table 1), and k is the mineralisation rate constant. The rate of N mineralisation was adjusted for soil temperature assuming a temperature coefficient [Q.sub.10] of 1.8 (Campbell et al. 1981, 1984), giving a temperature-adjusted rate constant of [k.sub.1], and adjusted for soil moisture content using the following soil moisture factor (Campbell et al. 1988):
x = (M - [M.sub.o])/([M.sub.max] - [M.sub.o]) (5)
where M, [M.sub.o], and [M.sub.max] are the actual soil moisture content, the moisture content at the lower limit of plant extraction, and at the drained upper limit, respectively. Giving:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (6)
Nitrogen mineralisation was calculated for the 0-0.1 m depth layers only, and assumed to be negligible below 0.1 m.
ROOTMAP was parameterised for soils from 2 agroecological zones: the Victorian southern Mallee (Birchip, Sodosol, 35[degrees]57'S, 142[degrees]47'E) and the Wimmera region (Dooen, Grey Vertosol, 36[degrees]40'S, 142[degrees]17'E) (Isbell 1996). Birchip is a semiarid dryland cropping environment with an average annual rainfall of 376 mm, whereas the long-term average for Dooen is 420 mm. Soil parameters (Table 1) were collated from measurements on soil samples from the 2 field sites (Officer et al. 2009). Soils were characterised using standard methods for soil moisture content, bulk density (BD), available N (mineral N), Colwell P, electrical conductivity (EC) in 1:5 extract, and total P and Na by nitric/perchloric acid digestion (Rayment and Higginson 1992). Crop lower limit (LL) was determined by pressure plate method at -1500 kPa (Klute 1986) using a <2-mm sieved sample, which avoids the shrink-swell problems of intact samples and consequent difficulty of defining volume in these soils. Field capacity was determined by sampling from large intact soil cores that had been saturated and then drained. Anaerobically mineralisable nitrogen (AMN) was determined as the difference in KCl-extractable N, between post-incubated and non-incubated extracts (Sparling and Searle 1993). Phosphate buffering index (PBI) was determined by mid infrared analysis calibrated to the PBI test (Burkitt et al. 2002).
Due to the detailed root information calculated and stored in ROOTMAP, it can take up to 20min on a standard PC to simulate crop root growth over 1 season. For this reason, the stochastic/probability method of simulating crop growth and nutrient uptake over 100 seasons was not adopted. Instead, 6 'analogue' seasons were selected from SILO records (www.nrw. qld.gov.au/silo/), to represent a range of seasonal rainfall patterns. Scenarios comprised combinations of high/low rainfall around sowing (115-166 DOY, 25 April-15 June), high/low growing season rainfall (60-344 DOY, 1 March-10 December), and high/low post-anthesis rainfall (253-273 DOY, 10-30 September). A year was identified as having good rainfall for crop establishment if, during the planting window (25 April-15 June), there was [greater than or equal to] 30mm of rainfall over a 10-day period, giving good soil moisture for sowing. Seasons of low sowing rainfall had an event with [less than or equal to] 10 mm, with no subsequent rainfall during the planting window. Post-anthesis rainfall was considered to be high if there was an event of [greater than or equal to]20 mm over 48 h with a follow-up of [greater than or equal to]20 mm over the next 7 days. This approach was taken in order to investigate the effect on fertiliser use efficiency of dry sowing (practised in the Mallee/ Wimmera regions in recent years) v. the more traditional sowing into good soil moisture, and also the effects of moisture stress during grain fill.
Rooting depth is restricted in the Mallee Sodosol soil due to physicochemical subsoil constraints. Previous studies have found the sodium profile to be an effective model for predicting the effect of multiple physicochemical constraints on subsoil root growth and function (Nuttall et al. 2003b; Sadras et al. 2003; Rodriguez et al. 2006). The profile of total sodium of the Sodosol soil was used as a proxy indicator of trends in exchangeable sodium, to generate a 'hostility factor' which was used in ROOTMAP to regulate the rate of root growth. This was applied as a simple linear multiplier on the root growth rate which reduced from 1 (no effect) for sodium levels of 1.1 mg Na/g down to 0.05 for sodium levels of [greater than or equal to]5.5 mg Na/g.
All simulations were run from January 1 (Day 1) to plant maturity. Modelled crops were sown on 21 May at the Mallee Sodosol site and 7 June at the Wimmera Vertosol site. Crop row spacing was 0.2 m, and planting density 150 plants/[m.sup.2]. Seed was sown at a depth of 0.04 m and rate of fertiliser application was 26 kg P/ha and 50 kg N/ha (as urea). Placement and timing of the fertiliser application varied with the treatment (see fertiliser treatments listed below); P was always sown directly below the seed and N was sown offset to the seed by 0.02 m for the shallow placements (0.025, 0.05 m) and directly below the seed for the deep placement (0.1 m).
In total, 14 different fertiliser management options were assessed. These were based on a range of potential commercial strategies. Cultivation continues to form an integral part of cropping systems in Victoria. For this reason, treatments incorporating N or P fertiliser into the topsoil ('plough layer') have been included. The treatments are as follows:
1. Reference fertiliser treatment (N5, P5): N and P banded at 0.05 m depth with the seed at sowing.
2. Shallow banding of N (N2.5, P5): N banded at 0.025 m, P banded at 0.05 m at sowing.
3. Shallow banding of P (N5, P2.5): N banded at 0.05 m, P banded at 0.025 m at sowing.
4. Deep banding of P (N5, P10): N banded at 0.05m, P banded at 0.1 m at sowing.
5. Deep banding of N (N10, P5): N banded at 0.1 m, P banded at 0.05 m at sowing.
6. N topsoil, P banded (NTS, P5): N incorporated into the top 0.1 m of soil at sowing, P banded at 0.05 m at sowing.
7. N top-dressed mid till, P banded (NTD MT, P5): N top-dressed onto the soil surface at mid tillering, P banded at 0.05 m at sowing, no N applied at sowing.
8. Split N, P banded (N1TS, N2TD, P5): 1/2 N incorporated into the top 0.1 m of soil at sowing with the remainder top-dressed at mid tillering, P banded at 0.05 m at sowing.
9. Split N, P banded (N1 5, N2TD, P5): 1/2 N banded at 0.05 m with the seed at sowing, 1/2 N top-dressed onto soil surface at mid tillering, P banded at 0.05 m at sowing.
10. Split N, P banded (NITS, N2MRB, P5): 1/2N incorporated into the top 0.1 m of soil at sowing, 1/2N mid-row banded at mid tillering, P banded at 0.05 m at sowing.
11. Split N, P banded (N1 5, N2 MRB, P5): 1/2 N banded at 0.05 m at sowing, 1/2 N mid-row banded at mid tillering, P banded at 0.05 m at sowing.
12. N banded, P split (N5, P5&10): N banded at 0.05m at sowing, 1/2 P banded at 0.05m at sowing, and 1/2 P banded at 0.1 m at sowing.
13. N banded, P top-dressed (N5, PTD): N banded at 0.05 m at sowing, P top-dressed on soil surface at sowing.
14. N banded, P topsoil (N5, P TS): N banded at 0.05 m at sowing, P incorporated into the top 0.1 m at sowing.
The first fertiliser management option (N5, P5) is referred to as the reference treatment. All results are presented relative to this treatment for each season. This removes the absolute effect of season on nutrient uptake and highlights the interaction between season and fertiliser management strategy.
Results and discussion
This study was designed to identify agronomic strategies (or hypotheses) that involve altering the timing and position of fertiliser. These hypotheses would then focus the design of subsequent glasshouse and field experiments that aim to identify management strategies that grain growers could use to improve nutrient use efficiency. While some of the strategies modelled are already widely used by grain growers, the study revealed particular practices that, once validated experimentally, may be of significant value. Importantly, this study suggests that region-specific (and soil-group specific) recommendations for fertiliser management would be superior to the 'one size fits all' approach currently adopted across the Wimmera/Mallee.
Soil type and season affect root system development and nutrient use
In this study, the fundamental difference between the Sodosol and Vertosol was the presence of physicochemical subsoil constraints limiting subsoil root growth in the Sodosol (Holloway and Alston 1992; Nuttall et al. 2003a, 2005; Rodriguez et al. 2006; Officer et al. 2009; Table 1). The ROOTMAP model represented the restricted rooting depth typically observed in Mallee Sodosol and Calcarosol soils. This was achieved using a sodium profile (Table 1) of the 2 soils to generate a 'hostility factor' that regulated the rate of root growth (Nuttall et al. 2003b; Sadras et al. 2003; Rodriguez et al. 2006). Using this approach, rooting depth was restricted to around 0.8m in the Sodosol (Fig. 1). This agrees with the intact soil core study of Nuttall et al. (2005), who showed that the rooting depth of wheat and barley in a Mallee Calcarosol was restricted to 0.8m and shallower, due to the effect of subsoil constraints. Holloway and Alston (1992) and Nuttall et al. (2003a) also demonstrated the degree to which physicochemical constraints in Calcareous subsoils restrict root growth in the 0.6-1.0 m depth layer. Modelled rooting depth extended down to 1.3 m in the Vertosol (Fig. 1) in 1996, a season with average rainfall (Appendix 1). This agrees with the field study of Norton and Wachsmann (2006), which showed that the rooting depth of wheat and other crops can extend below 1.1 m in a Wimmera Vertosol. The lower total root length and rooting depth (Figs 1-3) in the Sodosol resulted in lower crop N, P, and water uptake than the Vertosol (data not shown).
[FIGURE 1 OMITTED]
Physicochemical constraint to subsoil root growth was the major driver of the difference in fertiliser use efficiency between the soil types, rather than the effect of any fundamental differences in plant-available nutrient dynamics (such as P buffering capacity). These subsoil constraints had the effect of reducing both rooting depth and total root system size in the Sodosol (Figs 2, 3), placing a greater proportion of the root system in the topsoil layers. In a pot study, Officer et al. (2009) also found that wheat had a greater proportion of the root system in topsoil layers of the same Sodosol soil, compared with the Vertosol soil. As a result of this, the impact of season on subsoil moisture was not a dominant regulator of subsoil root growth in the Sodosol, making rooting depth and overall root system size less responsive to seasonal rainfall (Figs 1-3). Nuttall et al. (2005) also found this to be the case for wheat and barley root growth in a Mallee Calcarosol. Other studies have certainly shown the importance of the interaction between seasonal rainfall, subsoil constraints, and crop productivity in the southern region (Sadras et al. 2002, 2003; Rodriguez et al. 2006; Sadras and Rodriguez 2007).
[FIGURE 2 OMITTED]
In the Vertosol, rainfall over the growing season, and hence levels of subsoil moisture, impacted upon rooting depth (Fig. 4). Maximum rooting depth varied from around 1.0 m in a dry year (Fig. 4a), to greater than 1.7 m in a wet year (Fig. 4c). These rooting depths were related to the depth of moisture penetration through the soil in the dry and average rainfall years, but not in the wet year. These results agree with the field study of Norton and Wachsmann (2006), who found that the rooting depth of wheat in a Wimmera Vertosol varied by a factor of 1.8 between sites, depending on the level of plant-available water in the subsoil. Seasonal rainfall also altered the modelled distribution of root length density with depth, and the total size of the root system. In the dry year (Fig. 4a), roots were concentrated in the surface-soil layers (0-0.2 m) in response to shallow soil moisture arising from small rainfall events (Appendix 1), partially compensating for restricted rooting depth. Rooting density in the topsoil layers was lowest in the wet year (Fig. 4c). Seasonal conditions had a relatively greater effect on root growth in the Vertosol than the Sodosol (Figs 2, 3), particularly in the dry year. This reflected the greater volume of plant-available water in the fine-textured Vertosol Wimmera clay than the coarser textured Mallee Sodosol, and the influence of subsoil constraints on root growth in the Sodosol. These whole-scale changes in root length density distribution with depth in response to seasonal rainfall highlight the complexity of the interaction between season and fertiliser strategy, particularly in the Vertosol.
[FIGURE 3 OMITTED]
Because of the effect of season on root growth, there was a larger and clearer interaction between season and fertiliser strategy in the modelled Vertosol than the Sodosol (Figs 5-8). This suggests a potential for fertiliser management recommendations to be primarily based on plant-available nutrients for cropping soils in which subsoil constraints are dominant, with seasonal factors playing a more important role for fertiliser recommendations in soil types with fewer constraints to subsoil root growth.
It is important, however, to keep in mind that the primary reason for the large root growth response to season in the Vertosol was that the profile was reset to zero plant-available water at the start of each year (1 January). This approach assumed no stored subsoil moisture remaining from the previous season. Subsoil moisture in this study was therefore a direct function of seasonal rainfall, hence the large affect of season on modelled subsoil root growth. While this assumption may be realistic for many paddocks in the Wimmmera/Mallee over much of the last decade, it raises the question of what the effect of additional stored water at sowing might be, as can occur during a long fallow or with a large rainfall event late in the previous season.
A number of factors suggest that if we had considered additional stored water carried over from the previous season, we may have seen different subsoil root growth and nutrient uptake responses in the Wimmera Vertosol. First, any additional water carried over from the previous season is most likely to be located in the subsoil (e.g. O'Leary and Connor 1997). Also, water plays an important role in nutrient use efficiency, particularly N use efficiency (Aulakh and Malhi 2005), with water located in the subsoil generally more beneficial to crop yield than that located in the topsoil (Kirkegaard et al. 2007), particularly in relation to late-season N nutrition (during grainfill). Additional simulations that include different levels of stored subsoil moisture remaining from the previous year would provide an insight into how the root growth and season interactions might change if subsoil root growth was not as reliant upon current season rainfall to provide subsoil moisture. The ROOTMAP model could also be modified to account for carryover of nutrients from previous years. For example, the background soil nutrient pool could be divided into a residual fertiliser N and P pool and a background N and P pool. This could then be used to evaluate the residual benefit of fertiliser application from the previous year, in relation to the timing and placement of the fertiliser application in the current year.
[FIGURE 4 OMITTED]
Fertiliser placement and timing affect nutrient acquisition and root growth
While there were important differences, uptake of nutrients from the background soil pool showed a similar trend to that for uptake from the fertiliser pool. This was particularly true for P uptake in the Vertosol (Fig. 5), for which there was a strong interaction between fertiliser treatment and season. Generally, there was a greater spread of N uptake and P uptake responses to fertiliser strategy in the Vertosol than the Sodosol. This is in agreement with the intact soil core study of Officer et al. (2009), who found that wheat root growth responses to N and P placement were greater in the Vertosol than the Sodosol.
For both soils, fertiliser management strategy had a greater effect on P acquisition (Figs 5, 6) than N acquisition by the crop (Figs 7, 8). We hypothesise that this result is a direct consequence of the relatively low mobility of P (Tinker and Nye 2000). Because P is relatively immobile, particularly in highly P-buffering soils, it is primarily spatially restricted to those zones of the soil where it has been placed as fertiliser, or where it is produced through the mineralisation of organic matter (McLaughlin et al. 1988). In addition to this, relatively high root length densities are required for the successful interception and foraging of soil P (Fitter 1985). It is not surprising, therefore, that P acquisition was sensitive to the interplay between depth and timing of fertiliser placement (both N and P) in the soil profile, seasonal rainfall, and root length density distribution with depth.
The effect of fertiliser management strategy on P uptake, and its interaction with season, dominated modelled root growth in the Vertosol, particularly for the single fertiliser placement strategies (Figs 2a, 3a, 5a, b, 7a, b). There was a relationship between total root system size (Fig. 2a) and root exploration of the subsoil (Fig. 3a), and P uptake (Figs 5a, b, 9). This in turn directly affected foraging for the background soil N resource, even though fertiliser N uptake was relatively unaffected (Fig. 7). Total rooting depth varied by up to 86% and total root length by up to 230% between fertiliser treatments in the Vertosol soil in the dry year (1976, Figs 2, 3). This highlights the potential value of using fertiliser timing and placement as a tool for managing crop and root system size, which could be valuable for manipulating both fertiliser use efficiency and water use efficiency (Norton and Wachsmann 2006; Sadras and Angus 2006), particularly in resource-poor soils. The substantial reduction in root system size and rooting depth in the dry year directly affected nutrient uptake from both the fertiliser and background pools. This is seen clearly in Fig. 9, which shows that the relationship between root system size and total P uptake is closest in the dry year.
[FIGURE 5 OMITTED]
The placement and timing of the fertiliser application had the greatest affect on modelled P uptake by wheat in a dry (most resource-limiting) year across both soil types (Figs 5, 6). In both soils, more fertiliser P was taken up in the dry year (1976) when placed at a shallow depth (0.025 m, N5 P2.5), compared to placement at 0.05 m (N5 P5) (Figs 5c, 6c). Fertiliser P uptake from both soils was optimised when N and P were banded together at a depth of 0.05 m or less (N5, P5; N2.5, P5; N5, P2.5, Figs 5c, 6c). This supports the work of Valizadeh et al. (2003), who showed that wheat growth and P-utilisation efficiency was improved with shallow banding of P (0.05 m), compared with deeper banding (0.15 m). They suggested that this was due to the delayed access of roots to the deep band, which in turn reduced tillering. In this modelling study, deep placement of P (0.1 m; N5, P10) markedly reduced the uptake of fertiliser P in the Sodosol (Fig. 6c), but N uptake was not affected (N placed at 0.05 m, Fig. 8c). The modelled plant partially compensated for the reduced availability of fertiliser P by accessing more soil (background) P, resulting in a small net reduction in total P uptake and root growth (Figs 2c, 6a). The effects were less marked in the Vertosol.
That fertiliser management strategy had the biggest effect on nutrient uptake and root growth in the dry (1976) season appears to contrast with the observation that crops are most responsive to fertiliser in wet seasons (e.g. McDonald 1989; Asseng et al. 2001). However, the absolute uptake of N and P was still low in the dry season, despite the larger relative effect of fertiliser treatment. The multiple small rainfall events that characterise dry seasons in the Wimmera/Mallee provide topsoil moisture only. They affect biological processes and nutrient availability in the topsoil, in turn affecting soil resource uptake and root growth in topsoil layers. The pattern of rainfall events also influences subsoil water flows and processes such as nitrate leaching, affecting the depth and distribution of roots through the soil profile. Through the study of climate data for eastern Australia, Sadras and Rodriguez (2007) showed that small rainfall events make an important contribution to annual rainfall in western Victoria. They suggest that these small rainfall events may have an important impact on dryland cropping systems in southern cropping zones, over a range of seasons, but particularly in drier years (Sadras and Rodriguez 2007).
[FIGURE 6 OMITTED]
In addition to this, several previous studies in Australia have suggested that any P located in the topsoil would become positionally unavailable to plant roots as the topsoil dried (Cornish et al. 1984; Pinkerton and Simpson 1986). The dynamics of nutrient uptake from topsoil layers are, however, complex. A more recent study suggested that wheat may have the capacity to utilise hydraulic lift to redistribute subsoil moisture into topsoil layers, for accessing P from dry surface soil (Valizadeh et al. 2003). This highlights the importance of testing the hypotheses generated from this desktop study in the field. While the shallow placement (50 mm and less) of P may have shown a relative benefit in the dry year examined here, this result is likely to be sensitive to both the rainfall distribution and soil type. While frequent small rainfall events are typical of dry seasons in the agricultural zones considered here (Sadras and Rodriguez 2007), other types of dry-season rainfall distributions may not be compatible with shallow fertiliser placement.
Topdressing of P fertiliser on the soil surface (N5, PTD), or incorporation of P into the top 0.1 m of the soil profile (N5, PTS) at sowing, resulted in a reduction in fertiliser P uptake in both soils and all seasons (Figs 5d, 6d). Distributing fertiliser P through the top 0.10m of soil (or topdressing), rather than placing a concentrated band, had 2 effects (de Wit 1953). Firstly, it reduced the concentration of P seen at the root surface, thereby reducing the rate at which P is taken up at the root surface (by Michaelis-Menten kinetics). Secondly, it reduced the capacity for roots to physically intercept the fertiliser P in soil. These P fertiliser strategies did not generally lead to a reduction in the uptake of N fertiliser by the crop, the exception being the dry year of 1976 (Figs 7d, 8d). Instead, the topdressing of P led to a small increase in the uptake of fertiliser N (Figs 7d, 8d). Analysis (not shown) of the root distribution with depth showed less root proliferation in the topsoil layers and an increase in root growth in deeper layers when P was topdressed. This suggests there may have been greater interception of any mobile fertiliser N displaced to subsoil layers.
Similarly, incorporating fertiliser N into the top 0.1 m at sowing (NTS, P5), or topdressing N at mid-tillering (NTD, MT, P5), reduced fertiliser N and P uptake by the crop in both soil types, under all seasonal conditions assessed (Figs 5-8). Again, the dispersal of N through the soil reduced both N concentration in soil, thereby reducing the rate of N uptake at the root surface (calculated using Michaelis-Menten kinetics), and nutrient interception by roots (de Wit 1953). This may have a particular impact early in the growing season. When banded below the seed, there is a concentrated band of nutrient in the growth direction of the root system, improving nutrient interception and uptake rates and, hence, crop establishment. In this modelling study, the dispersal of N through the top 0.10 m resulted in lower root length density at all depths, but particularly in the top 0.10m of the soil profile (data not shown).
[FIGURE 7 OMITTED]
Soil N is a relatively mobile resource, and can displace some distance from the original source of fertiliser placement or organic matter mineralisation (Tinker and Nye 2000). It is also absorbed over a greater proportion of the growing season. Nitrogen acquisition was therefore less sensitive to the spatial placement of N fertiliser, being more sensitive to the supply of N fertiliser over time (Figs 8, 9). Splitting N fertiliser application over time [25 kgN/ha at sowing (N1), 25kgN/ha at mid-filleting (N2)] generally improved the uptake of the applied fertiliser N (Figs 7d, 8d). This effect was most apparent in the Wimmera Vertosol, occurring to a lesser extent in the Mallee Sodosol. This result suggests that crop access to, and utilisation of, a mobile nutrient such as N may be improved when it is applied on a continual basis through time (at least up to mid tillering), such as occurs when using slow-release formulations (Shaviv 2001). Again, however, there was an interaction between the acquisition of the soil N and P resources. When N application was split over time, root systems were smaller over all seasons (Fig. 2) than those in the reference treatment (N5, P5), and rooting depth was reduced in the Vertosol soil (Fig. 3b). While the crop was able to compensate at the second N application, the total root system size was still smaller than if all N was applied at sowing. There was also a reduction in the uptake of background P and fertiliser P in both soils, regardless of season (Figs 5, 6). Since P is a relatively immobile resource, plant uptake of P is sensitive to the density and number of roots present (Fitter 1985); hence, the reduction in root system size led to a reduction in P uptake (and vice versa). This effect was less pronounced under wet conditions and greatest during dry seasons (Fig. 9). The effect of N fertiliser application on early root growth may have important implications not only for nutrient use efficiency, but also for water use efficiency. Early vigour has been suggested as a method for improving water use efficiency, by reducing evaporative water losses (Turner 1997; Richards et al. 2002; Sadras et al. 2003; Norton and Wachsmann 2006). However, the potential yield advantage of early vigour also needs to be weighed against the 'haying off' phenomenon experienced during dry finishes (van Herwaarden et al. 1998).
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
Starting soil moisture, seasonal conditions and nutrient use efficiency
A range of edaphic (physicochemical and biological), seasonal and crop physiological factors can affect the nutrient use efficiency of crops. Both the definition of nutrient use efficiency and the methods used to measure it, vary widely (Batten 1992; Gourley et al. 1994; Fageria and Baligar 2005; Ladha et al. 2005); however, grain growers are solely interested in the economic return (i.e. amount and quality of grain produced) arising from the supply of the nutrient (fertiliser), termed 'agronomic efficiency'. Only a small proportion of the nutrients contained in fertiliser are directly used by dryland crops, ranging from 5 to 30% for P (Hedley and McLanghlin 2005), and from 10 to 50% for N (McDonald 1989; Crews and Peoples 2005). However, the overall response of the crop to this added nutrient can be much larger, ranging from 0 (or even negative in the case of the 'haying off' phenomenon (van Herwaarden et al. 1998) to >50%, due to the stimulatory effect of added nutrients on the uptake of other nutrients in the soil (the 'priming effect' Rao et al. 1992).
The contribution that N and P fertiliser made to total N and P uptake by the wheat crop further highlighted the interaction between rainfall, soil moisture, and nutrient availability. In both soils, the background soil N or P taken up by the crop was not directly proportional to the fertiliser N or P uptake (Figs 5-8). This reflected both stimulatory and inhibitory effects of the fertiliser on root growth and subsequent uptake of nutrients from the soil pool. It suggests a capacity for the crop to partially compensate for a poor fertiliser management option by altering the foraging strategy for background soil N and P.
The fertiliser N source became relatively more important as the overall N availability declined (such as in a dry start, or dry year, for example), particularly for these infertile soils (Table 1). In a dry year, soil N becomes less mobile (more transport-limited) due to the low soil moisture, with the plant taking up less background N and relatively more of the fertiliser N (particularly for the infertile soils in this study). While N fertiliser may make a greater contribution to N uptake in a dry year, the resulting effect on grain yield and economic return must be considered. In practice, grain growers generally reduce N fertiliser application in a dry year due to the low probability of obtaining an economic yield response.
Phosphorus fertiliser also typically made a bigger contribution (relative to background soil P) to plant P uptake in a dry year. As the low-concentration, background soil P became more transport-limited in a dry year, P uptake in the concentrated P band was up-regulated, increasing uptake from the fertiliser source. Numerous studies have highlighted reductions in total P uptake under dry soil conditions (including this simulation study). Soil moisture affects the dissolution of P fertilisers (E. Lombi and M. McLaughlin, pers. comm.), the diffusion of P in the soil to the root surface (Tinker and Nye 2000; Kovar and Claassen 2005), and the ability of the crop to absorb P through reductions in both root length and root : soil contact by water-stressed roots (Strong and Barry 1980; MacKay and Barber 1985).
Sowing time and soil water availability (throughout the growing season) strongly influence grain yield potential and response to fertiliser application in semi-arid environments (McDonald 1989; Asseng et al. 2001) such as the Wimmera and Mallee. Batten et al. (1999) found that early-sown wheat in southern New South Wales had higher grain yields, accumulated more P, and required less fertiliser P (i.e. used more soil P) to achieve maximum yields than crops sown 2 months later. Grain growers will often 'dry sow' even if the 'seasonal break' has not occurred, because of the effect of sowing time on yield potential. In recent years, farmers have generally shifted their attention from soil moisture at the time of sowing, focusing more on seasonal forecasts such as the Southern Oscillation Index (SOI) when making fertiliser decisions. This is because the SOI is believed to have a greater influence on the overall potential demand by the crop for nutrients, especially for N fertiliser where there is a risk of 'haying off' if insufficient moisture is available later in the season (Passioura 2006).
Our study indicated, however, that it may be important to consider both the SOI and soil moisture at time of sowing. The results suggested that soil moisture at the start of the season can strongly influence the effectiveness of different fertiliser management strategies, regardless of subsequent seasonal conditions. There were several different treatments for which there were substantial differences in both relative and absolute fertiliser uptake in 1975 (a dry start) compared with 1988 (a wet start). This was particularly true for relative uptake from the Sodosol soil (Fig. 6c, d). There were, however, fertiliser management strategies that maintained both the absolute and relative utilisation of fertiliser P in a dry start. Currently, many growers tend to reduce fertiliser use if dry seasonal conditions are predicted. An economic analysis is needed, but there may be a case for rates of P fertiliser to be maintained in a dry season, particularly if the depth and timing of placement are matched to the dry conditions and grain prices are favourable.
One limitation of ROOTMAP is that while it simulates root growth and nutrient uptake, it does not predict the variable most important to grain growers: yield. For this study we assumed that when nutrients limited the growth of wheat, there would be a highly significant correlation between nutrient uptake and plant tissue concentration, and grain yield. Importantly, the simulation approach permitted a rapid preliminary assessment of a range of strategies for different soil type/seasonal conditions that would have been prohibitively expensive to undertake using an empirical field (or controlled environment) study.
The complex interaction between season, N and P fertiliser treatments, fertiliser uptake, and total N and P uptake highlights the need to discriminate between crop N and P uptake from the background (native) pool and the fertiliser pool. The current study was able to differentiate between the 2 nutrient pools following changes to ROOTMAP. The ability to discriminate between fertiliser and soil sources is important as it allows a direct estimate of fertiliser use efficiency, and it also provides a mechanistic insight into soil-plant processes. An alternative approach is the use of stable isotopes such as [sup.15]N or alternatively ionising radiation sources such as [sup.32p]. [sup.15N] data must be treated with caution because mineralisation--immobilisation in the soil can result in isotope dilution in the mineral N pool, leading to an underestimate of the crop recovery of the fertiliser (Fillery and McInnes 1992), while safety regulations limit the use of [sup.32]p in the field. A study of fertiliser use efficiency using simulation modelling can be important for investigating treatment effects, before time-consuming field and glasshouse studies are undertaken.
Simulation models such as ROOTMAP can represent a summation of our current understanding of the processes regulating root growth, nutrient and water dynamics, and their interaction with soil chemico-physical properties and seasonal conditions. Model output is, however, a function of both the quality of data used for parameterisation and the algorithms used to describe the various soil-plant processes represented in the simulations. There is a level of uncertainty surrounding the data used to parameterise the Wimmera and Mallee soils, and we are aware of no published data on how P-buffering capacity changes with depth, for example. The use of a much wider range of (soil) datasets for parameterisation offers the potential to significantly improve the utility of model predictions. Keeping this in mind, the aim of this exercise was to formulate and investigate hypotheses that could be subsequently tested in the glasshouse and validated in the field, rather than provide immediate estimates of grain yield responses to different management options as required by farmers. However, the behaviour and trends observed in the model appear to agree with a range of glasshouse observations, providing a useful method for investigating the complex interactions between soil type, nutrient placement, and season.
Root system size is important for increased resource acquisition when resources are limiting (Ho et al. 2005). In these simulations, both the size and the distribution of the root system in the profile were strongly affected by soil type, season (rainfall), and fertiliser management. Measurement of root systems is notoriously difficult under field conditions, especially in sodic soils with high clay contents similar to those that dominate the Wimmera/Mallee regions. Even when great attention is paid to methods, a significant proportion of the root system can be lost during recovery (Kucke et al. 1995). Consequently, this study provided insights into a system that would have been difficult to study empirically, although subsequent validation is needed. Our study suggested that the co-location of root growth and resource availability (both water and nutrients) provides significant increases in nutrient uptake and therefore crop growth.
[FIGURE A1 OMITTED]
This work was co-funded by the Grains Research and Development Corporation through the Nutrient Management Initiative (Project UM00023). ROOTMAP has been developed as a collaborative partnership between The Grains Research and Development Corporation, The Department of Agriculture and Food Western Australia, The Centre for Legumes in Mediterranean Agriculture at the University of Western Australia, The University of Tasmania, and Dr V. Dunbabin.
Manuscript received 1 May 2008, accepted 1 December 2008
ABARE (1999) 'Australian Farm Survey Report: Australian Grains Industry Performance by Agroecological Zone.' ABARE Report prepared for the Grains Research and Development Corporation, Project 1608. (ABARE: Canberra)
Asseng S, Turner NC, Keating BA (2001) Analysis of water- and nitrogenuse efficiency of wheat in a Mediterranean climate. Plant and Soil 233, 127-143. doi: 10.1023/A:1010381602223
Aulakh MS, Malhi SS (2005) Interactions of nitrogen with other nutrients and water: Effect on crop yield and quality, nutrient use efficiency, carbon sequestration, and environmental pollution. Advances in Agronomy 86, 341-409. doi: 10.1016/S0065-2113(05)86007-9
Baldwin JP, Nye PH, Tinker PB (1973) Uptake of solutes by multiple root systems from soil. III. A model for calculating the solute uptake by a randomly dispersed root system developing in a finite volume of soil. Plant and Soil 38, 621-535. doi: 10.1007/BF00010701
Batten GD (1992) A review of phosphorus efficiency in wheat. Plant and Soil 146, 163-168. doi: 10.1007/BF00012009
Batten GD, Fettell NA, Mead JA, Khan MA (1999) Effect of sowing date on the uptake and utilisation of phosphorus by wheat (cv. Osprey) grown in central New South Wales. Australian Journal of Experimental Agriculture 39, 161-170. doi: 10.1071/EA97104
Burkitt LL, Moody PW, Gourley CJP, Hannah MC (2002) A simple phosphorus buffering index for Australian soils. Australian Journal of Soil Research 40, 497-513. doi: 10.1071/SR01050
Campbell CA, Myers RJK, Weier KL (1981) Potentially mineralizable nitrogen, decomposition rates and their relationship to temperature for five Queensland soils. Australian Journal of Soil Research 19, 323-332. doi: 10.1071/SR9810323
Campbell CA, Jame YW, De Jong R (1988) Predicting net nitrogen mineralization over a growing season: model verification. Canadian Journal of Soil Science 68, 537-552.
Campbell CA, Jame YW, Winkleman GE (1984) Mineralization rate constants and their use for estimating nitrogen mineralization in some Canadian prairie soils. Canadian Journal of Soil Science 64, 333-343.
Connor DJ (2004) Designing cropping systems for efficient use of limited water in southern Australia. European Journal of Agronomy 21, 419-431. doi: 10.1016/j.eja.2004.07.004
Cornish PS, So HB, McWilliam JR (1984) Effects of soil bulk density and water regimen on root growth and uptake of phosphorus by ryegrass. Australian Journal of Agricultural Research 35, 631-544. doi: 10.1071/ AR9840631
Crews TE, Peoples MB (2005) Can the synchrony of nitrogen supply and crop demand be improved in legume and fertilizer-based agroecosystems? Nutrient Cycling in Agroecosystems 72, 101-120. doi: 10.1007/s 10705-004-6480-1
de Wit CT (1953) A physical theory on placement of fertilizers. Verslagen van Landbouwkundige Onderzoekingen 59, 1-71.
Diggle AJ (1988) ROOTMAP--a model in three-dimensional coordinates of the growth and structure of fibrous root systems. Plant and Soil 105, 169-178. doi: 10.1007/BF02376780
Dunbabin V (2007) Simulating the role of rooting traits in crop-weed competition. Field Crops Research 104, 44-51. doi: 10.1016/j.fcr. 2007.03.014
Dunbabin VM, Diggle A J, Rengel Z (2002b) Simulation of field data by a basic three-dimensional model of interactive root growth. Plant and Soil 239, 39-54. doi: 10.1023/A:1014952728942
Dunbabin VM, Diggle A J, Rengel Z (2003) Is there an optimal root architecture for nitrate capture in leaching environments. Plant, Cell & Environment 26, 835-844. doi: 10.1046/j.1365-3040.2003.01015.x
Dunbabin VM, Diggle AJ, Rengel Z, van Hugten R (2002a) Modelling the interactions between water and nutrient uptake and root growth. Plant and Soil 239, 19-38. doi: 10.1023/A:1014939512104
Dunbabin VM, Rengel Z, Diggle AJ (2004) Simulating form and function of root systems: efficiency of nitrate capture is dependent on root system architecture and the spatial and temporal variability of nitrate supply. Functional Ecology 18, 204-211. doi: 10.1111/j.0269-8463.2004. 00827.x
Fageria NK, Baligar VC (2005) Enhancing nitrogen use efficiency in crop plants. Advances in Agronomy 88, 97-185. doi: 10.1016/S0065-2113 (05)88004-6
Feddes RA, Kowalik P, Kolinska-Mayinka K, Zarakny H (1976) Simulation of field water uptake by plants using a soil water dependent root extraction function. Hydrology 31, 13-26. doi: 10.1016/0022-1694 (76)90017-2
Fillery IR, McInnes KM (1992) Components of the fertiliser nitrogen balance for wheat production on duplex soils. Australian Journal of Experimental Agriculture 32, 887-899. doi: 10.1071/EA9920887
Fitter AH (1985) Functional significance of root morphology and root system architecture. In 'Ecological interactions in soil'. (Eds AH Fitter, D Atkinson, DJ Read, MB Usher) pp. 87-106. (Blackwell Scientific Publications: Oxford, UK)
Gourley CJP, Allan DL, Russelle MP (1994) Plant nutrient efficiency: a comparison of definitions and suggested improvement. Plant and Soil 158, 29-37. doi: 10.1007/BF00007914
Hammer GL, Woodruff DR, Robinson JB (1987) Effects of climatic variability and possible climatic change on reliability of wheat cropping A modelling approach. Agricultural and Forest Meteorology 41, 123-142. doi: 10.1016/0168-1923(87)90074-8
Hedley M, McLaughlin M (2005) Reaction of phosphate fertilizers and byproducts in soils. In 'Phosphorus, agriculture and the environment'. Agronomy Monograph No. 46. pp. 181-252. (American Society of Agronomy: Madison WI)
Ho MD, Rosas JC, Brown KM, Lynch JP (2005) Root architectural tradeoffs for water and phosphorus acquisition. Functional Plant Biology 32, 737-748. doi: 10.1071/FP05043
Holloway RE, Alston AM (1992) The effects of salt and boron on the growth of wheat. Australian Journal of Agricultural Research 43, 987-1001. doi: 10.1071/AR9920987
Holloway RE, Bertrand I, Frischke A J, Brace DM, McLaughlin M J, Sheppard W (2001) Improving fertiliser efficiency on calcareous and alkaline soils with fluid sources of P, N and Zn. Plant and Soil 236, 209-219. doi: 10.1023/A:1012720909293
Isbell RF (1996) 'The Australian Soil Classification.' (CSIRO Publishing: Collingwood, Vic.)
Kirkegaard JA, Lilley JM, Howe GN, Graham JM (2007) Impact of subsoil water use on wheat yield. Australian Journal of Agricultural Research 58, 303-315. doi: 10.1071/AR06285
Klute A (1986) Water retention: Laboratory methods. In 'Methods of soil analysis. Part I: Physical and mineralogical methods'. 2nd edn (Ed. A Klute) pp. 635-662. (American Society of Agronomy: Madison, WI)
Kovar JL, Claassen N (2005) Soil-root interactions and phosphorus nutrition of plants. In 'Phosphorus, agriculture and the environment'. pp. 379-414. Agronomy Monograph No. 46. (American Society of Agronomy: Madison WI)
Kucke M, Schmid H, Spiess A (1995) A comparison of four methods for measuring roots of field crops in three contrasting soils. Plant and Soil 172, 63-71. doi: 10.1007/BF00020860
Ladha JK, Pathak H, Krupnik TJ, Six J, van Kessel C (2005) Efficiency of fertilizer nitrogen in cereal production: Retrospects and prospects. Advances in Agronomy 87, 85-156. doi: 10.1016/S0065-2113(05) 87003-8
Lory JA, Russelle MP, Randall GW (1995) A classification system for factors affecting crop response to nitrogen fertilization. Agronomy Journal 87, 869-876.
MacKay AD, Barber SA (1985) Effect of soil moisture and phosphate level on root hair growth of corn roots. Plant and Soil 86, 321-331. doi: 10.1007/BF02145453
McDonald GK (1989) The contribution of nitrogen fertiliser to the nitrogen nutrition of rainfed wheat crops in Australia: a review. Australian Journal of Experimental Agriculture 29, 455-481. doi: 10.1071/ EA9890455
McLaughlin M J, Alston AM, Martin JK (1988) Phosphorus cycling in wheat pasture rotations. III. Organic phosphorus turnover and phosphorus cycling. Australian Journal of Soil Research 26, 343-353. doi: 10.1071/SR9880343
Mendham DS, Smethurst PJ, Moody PW, Aitken RL (1997) Modelling nutrient uptake: a possible indicator of phosphorus deficiency. Australian Journal of Soil Research 35, 313-325. doi: 10.1071/ S96046
Myers RJK (1987) Nitrogen management of upland crops: From cereals to food legumes to sugarcane. In 'Advances in nitrogen cycling in agricultural ecosystems'. (Ed. JR Wilson) pp. 257-273. (CAB International: Oxford, UK)
Norton RM, Wachsmann NG (2006) Nitrogen use and crop type affect the water use of annual crops in south-eastern Australia. Australian Journal of Agricultural Research 57, 257-267. doi: 10.1071/AR05056
Nuttall JG, Armstrong RD, Connor DJ (2003a) Evaluating physicochemical constraints of Calcarosols on wheat yield in the Victorian southern Malice. Australian Journal of Agricultural Research 54, 487-497. doi: 10.1071/AR02168
Nuttall JG, Armstrong RD, Connor DJ (2003b) Interrelationships between edaphic factors potentially limiting cereal growth on alkaline soils in north-western Victoria. Australian Journal of Soil Research 41, 277-292. doi: 10.1071/SR02022
Nuttall JG, Armstrong RD, Connor DJ (2005) The effect of boron tolerance, deep ripping with gypsum, and water supply on subsoil water extraction of cereals on an alkaline soil. Australian Journal of Agricultural Research 56, 113-122. doi: 10.1071/AR04183
O'Leary G J, Connor DJ (1996) A simulation model of the wheat crop in response to water and nitrogen supply: II Model validation. Agricultural Systems 52, 31-55. doi: 10.1016/0308-521X(96)00002-9
O'Leary G J, Connor DJ (1997) Stubble retention and tillage in a semi-arid environment: 1. Soil water accumulation during fallow. Field Crops Research 52, 209-219. doi: 10.1016/S0378-4290(97)00034-8
O'Leary G J, Connor D J, White DH (1985) Effect of sowing time on growth, yield and water-use of rain-fed wheat in the Wimmera, Vic. Australian Journal of Agricultural Research 36, 187-196. doi: 10.1071/ AR9850187
Officer S J, Dunbabin VM, Armstrong RD, Norton RM, Kearney GA (2009) Wheat roots proliferate in response to nitrogen and phosphorus fertilisers in Sodosol and Vertosol soils of south-eastern Australia. Australian Journal of Soil Research 47, 91-102.
Passioura J (2006) Increasing crop productivity when water is scarce from breeding to field management. Agricultural Water Management 80, 176-196. doi: 10.1016/j.agwat.2005.07.012
Perry MW (1992) Cereal and fallow/pasture systems in Australia. In 'Field crop ecosystems'. (Ed. CJ Pearson) pp. 451-483. (Elsevier: Amsterdam, The Netherlands)
Pinkerton A, Simpson JR (1986) Interactions of surface drying and subsurface nutrients affecting plant growth on acidic soil profiles from an old pasture. Australian Journal of Experimental Agriculture 26, 681-689. doi: 10.1071/EA9860681
Rao ACS, Smith JL, Parr JF, Papendick RI (1992) Considerations in estimating nitrogen recovery efficiency by the difference and isotopic dilution methods. Fertilizer Research 33, 209-217. doi: 10.1007/ BF01050876
Rayment GE, Higginson FR (1992) 'Australian laboratory handbook of soil and water chemical methods.' (Inkata Press: Melbourne, Vic.)
Rengasamy P (2002) Transient salinity and subsoil constraints to dryland farming in Australian sodic soils: an overview. Australian Journal of Experimental Agriculture 42, 351-361. doi: 10.1071/EA01111
Richards RA, Rebetzke G J, Condon AG, van Herwaarden AF (2002) Breeding opportunities for increasing efficiency of water use and crop yield in temperate cereals. Crop Science 42, 111-121.
Rodriguez D, Nuttall J, Sadras VO, van Rees H, Armstrong R (2006) Impact of subsoil constraints on wheat yield and gross margin on fine-textured soils of the southern Victorian Malice. Australian Journal of Agricultural Research 57, 355-365. doi: 10.1071/ AR04133
Sadras VO, Angus JF (2006) Benchmarking water-use efficiency of rainfed wheat in dry environments. Australian Journal of Agricultural Research 57, 847-856. doi: 10.1071/AR05359
Sadras VO, Baldock J, Roget DK, Rodriguez D (2003) Measuring and modelling yield and water budget components of wheat crops in coarsetextured soils with chemical constraints. Field Crops Research 84, 241-260. doi: 10.1016/S0378-4290(03)00093-5
Sadras VO, Rodriguez D (2007) The limit to wheat water-use efficiency in eastern Australia. II. Influence of rainfall patterns. Australian Journal of Agricultural Research 58, 657-669. doi: 10.1071/AR06376
Sadras VO, Roget DK, O'Leary GJ (2002) On-farm assessment of environmental and management constraints to wheat yield and rainfall use efficiency in the Mallee. Australian Journal of Agricultural Research 53, 587-598. doi: 10.1071/AR01150
Shaviv A (2001) Advances in controlled release fertilisers. Advances in Agronomy 71, 1-49. doi: 10.1016/S0065-2113(01)71011-5
Sparling GP, Searle PL (1993) Dimethyl sulphoxide reduction as a sensitive indicator of microbial activity in soil: the relationship with microbial biomass and mineralization of nitrogen and sulphur. Soil Biology & Biochemistry 25, 251-256. doi: 10.1016/0038-0717(93)90035-A
Strong WM, Barry G (1980) The availability of soil and fertilizer phosphorus to wheat and rape at different water regimes. Australian Journal of Soil Research 18, 353-362. doi: 10.1071/SR9800353
Tinker PB, Nye PH (2000) 'Solute movement in the rhizosphere.' (Oxford University Press: Oxford, UK)
Turner NC (1997) Further progress in crop water relations. Advances in Agronomy 58, 293-338. doi: 10.1016/S0065-2113(08)60258-8
Valizadeh GR, Rengel Z, Rate AW (2003) Response of wheat genotypes efficient in P utilisation and genotypes responsive to P fertilisation to different P banding and watering regimes. Australian Journal of Agricultural Research 54, 59-65. doi: 10.1071/AR02040
van Herwaarden AF, Farquhar GD, Angus JF, Richards RA, Howe GN (1998) 'Haying-off, the negative grain yield response of dryland wheat to nitrogen fertiliser. 1. Biomass, grain yield, and water use. Australian Journal of Agricultural Research 49, 1067-1081. doi: 10.1071/A97039
Van Rees KCJ, Comerford NB, Rao PS (1990) Defining soil buffer power: implications for ion diffusion and nutrient uptake modelling. Soil Science Society of America Journal 54, 1505-1507.
V. M. Dunbabin (A,D), R. D. Armstrong (B), S. J. Officer (B), and R. M. Norton (C)
(A) Tasmanian Institute of Agricultural Research (TIAR), University of Tasmania, Private Bag 54, Hobart, Tas. 7001, Australia.
(B) Primary Industries Research Victoria--Horsham, 110 Natimuk Rd, Horsham, Vic. 3400, Australia.
(C) University of Melbourne, PMB 260, Horsham, Vic. 3400, Australia.
(D) Corresponding author. Email: Vanessa.Dunbabin@utas.edu.au
Table 1. Input soil parameters for initialising the ROOTMAP model Input parameters for initialising the modelling scenarios were obtained from soil tests on intact soil cores from field sites (Officer et al. 2009). Two soil types were characterised: a Grey Vertosol (Dooen, Wimmera), and a Sodosol (Birchip, Mallee)Soil Bulk pH depth density Ca[Cl.sub.2] [H.sub.2]O (m) (Mg/ [m.sup.3]) 0-0.1 1.32 7.91 8.4 0.1-0.2 1.41 7.91 8.6 0.2-0.3 1.37 7.97 8.8 0.3-0.4 1.37 8.03 8.9 0.4-0.5 1.40 8.11 9.0 0.5-0.6 1.40 8.21 9.2 0.6+ 1.45 0.0-0.1 1.10 8.04 8.8 0.1-0.2 1.32 8.13 9.3 0.2-0.3 1.30 8.32 9.2 0.3-0.4 1.30 8.42 9.2 0.4-0.5 1.30 8.46 9.2 0.5-0.6 1.30 8.50 9.1 0.6+ 1.26 LL (A) DUL (B) Water depth (v/v) content (m) (v/v) 0-0.1 0.23 0.46 0.207 0.1-0.2 0.29 0.46 0.261 0.2-0.3 0.28 0.52 0.252 0.3-0.4 0.29 0.52 0.257 0.4-0.5 0.27 0.51 0.243 0.5-0.6 0.27 0.51 0.243 0.6+ 0.27 0.49 0.243 0.0-0.1 0.14 0.46 0.122 0.1-0.2 0.18 0.46 0.158 0.2-0.3 0.21 0.52 0.189 0.3-0.4 0.22 0.52 0.194 0.4-0.5 0.22 0.51 0.198 0.5-0.6 0.24 0.51 0.216 0.6+ 0.24 0.49 0.216 Min. N AMN (C) Tot. Colwell PBI (D) depth (KCI) P P (m) ([micro]g/g) Vertosol 0-0.1 4.55 14.47 186 20.2 99 0.1-0.2 4.47 6.49 139 5.0 90 0.2-0.3 2.83 6.12 126 4.3 96 0.3-0.4 1.65 7.04 124 3.3 93 0.4-0.5 2.87 3.70 107 3.0 97 0.5-0.6 2.27 3.03 96 3.0 100 0.6+ 2.27 3.03 96 3.0 100 Sodosol 0.0-0.1 4.21 6.49 151 7.1 60 0.1-0.2 2.44 7.34 100 2.0 97 0.2-0.3 1.74 7.90 98 2.0 87 0.3-0.4 0.99 5.92 88 2.0 89 0.4-0.5 0.95 5.47 89 2.0 117 0.5-0.6 0.87 5.25 76 2.0 80 0.6+ 0.87 5.25 76 2.0 80 EC Tot. depth paste Na (m) (dS/m) (mg/g) 0-0.1 0.15 0.30 0.1-0.2 0.16 0.40 0.2-0.3 0.19 0.54 0.3-0.4 0.24 0.71 0.4-0.5 0.27 0.91 0.5-0.6 0.30 1.10 0.6+ 0.0-0.1 0.23 0.55 0.1-0.2 0.49 1.32 0.2-0.3 0.83 2.08 0.3-0.4 1.11 2.65 0.4-0.5 1.32 3.07 0.5-0.6 1.47 3.21 0.6+ (A) Lower limit of plant water extraction. (B) Drained upper limit. (C) Aerobically mineralisable N. (D) P buffering index.
|Printer friendly Cite/link Email Feedback|
|Author:||Dunbabin, V.M.; Armstrong, R.D.; Officer, S.J.; Norton, R.M.|
|Publication:||Australian Journal of Soil Research|
|Date:||Feb 1, 2009|
|Previous Article:||Fate of potassium fertilisers applied to clay soils under rainfed grain cropping in south-east Queensland, Australia.|
|Next Article:||Wheat roots proliferate in response to nitrogen and phosphorus fertilisers in Sodosol and Vertosol soils of south-eastern Australia.|