Wheat roots proliferate in response to nitrogen and phosphorus fertilisers in Sodosol and Vertosol soils of south-eastern Australia.
Root systems are of fundamental importance to crop productivity because they acquire the water and nutrients that support shoot growth. Unfortunately, root systems and root function are difficult to study in the natural soil environment, primarily because such studies disturb the soil/root system. Instead, many root growth studies have been conducted on single plants in artificially uniform environments, such as solution culture or pots where the soil is thoroughly mixed (Hodge 2004). This contrasts markedly with the field situation where the plants compete in a soil environment where the physiochemical properties vary spatially and temporally within a profile (Burrough 1993). Responses to fertiliser are consequently best studied in realistic agronomic environments even though the large variability can make the experimental effects difficult to isolate. Despite these difficulties, we cannot optimise the efficiency of cropping systems unless we understand the complex effects that fertiliser, soil, and the environment have on the root growth of crop plants such as wheat.
Wheat (Triticum aestivum L.) is one of the major food crops in the world and is the most important grain crop gown in Australia (ABARE 2007). Wheat originated in semi-arid, relatively low-nutrient environments (Huang et al. 2007). The drought-resistant cultivars adapted to Australian dryland conditions exhibit sensitivity to environmental conditions, which typically constitute plant populations growing under moderate water stress and often a shortage of N (Sadras and Roget 2004; Sadras and Angus 2006). In the Victorian Wimmera and Mallee, the soils are generally alkaline with a moderate P-buffering capacity, of relatively low organic matter content, and often have a saline or sodic subsoil that can restrict root growth (Nuttall et al. 2003; McBeath et al. 2005). The fertilisers are generally granular and placed in bands with separate placement of the P source and the main N source (Strong and Barry 1980; Simpson and Pinkerton 1989; Passioura 2006).
The response of wheat to phosphate (P) fertiliser has been examined in many studies, although the effects found under controlled conditions may be difficult to identify in the field (Kamper and Claassens 2005). Phosphate is relatively immobile in soil, moving slowly to the plant roots by diffusion rather than by mass flow. Therefore, roots must actively acquire the P and many studies of P response patterns in wheat roots show root proliferation specifically around a band of fertiliser (e.g. Strong and Barry 1980; Vance et al. 2003). Most P is taken up through the fine root hairs that are situated behind the growing root tip, so that P uptake is closely related to root surface area and related parameters such as root length (Tinker and Nye 2000; Lynch and Ho 2005; Solaiman et al. 2007). Increasing root length density is therefore important for efficient P foraging (Manske et al. 2000). Studies also show that the rate of P influx can change, in particular high rates of P absorption have been found during early growth, which will have the effect of maximising the rate of P diffusion through the soil to the root surface (Robinson 1994; Elliott et al. 1997). Plants can also respond to the presence or absence of plant-available P by adjusting rhizosphere conditions and interacting with the rhizosphere microbial population to further solubilise P and increase uptake (Manske et al. 2000; Marschner et al. 2005; Solaiman et al. 2007).
In contrast to P, nitrate is soluble and moves readily to the plant root through the processes of mass flow in the plant transpirational stream, and also by diffusion, since the plant acts as a nutrient sink. Roots therefore do not need to physically intercept N sources, although they do require a hydraulic connection to those sources. Plants in general can exploit the mobile soil N resources more effectively by either manipulating the rate of nitrate uptake into the roots, or by elongating the roots that determine the volume of soil which is exploited, typically axial roots (Robinson et al. 1994). Wheat has shown a 'plastic' root growth response to N, exhibiting both strong localised increases in the rate of uptake and increases in root length, in response to N fertiliser when N is deficient elsewhere in the rooting volume (Diggle and Bowden 1990; van Vuuren et al. 1996; Hodge 2004). The localised response tends to be followed by responses down the profile, as the plant 'chases' N that is leaching downwards (Diggle and Bowden 1991), although mineralisation of nitrate from the organic fraction can encourage root growth in the topsoil (Herrera et al. 2007), while oversupply of N can inhibit root growth (Zhang and Rengel 2002).
Few studies have considered the simultaneous root response of wheat to both N and P, although many studies have considered the response to one nutrient under conditions of adequate background concentrations of the other nutrients. Zhang and Rengel (2002) found that wheat roots proliferated in proximity to bands of di-ammonium phosphate (DAP), provided that the ammonium did not reach toxic concentrations. Valizadeh et al. (2002) found that banding N and P fertilisers together increased the root growth of wheat more than mixing the fertilisers throughout the profile. In contrast, Alston (1976) examined the interactions between N and P fertiliser and post-anthesis water regime on root growth at harvest, and found that total root growth was mainly affected by the watering regime and that the effect of fertiliser was limited.
Low soil moisture is expected to reduce the availability of N and P to a plant, although concentrated sources of nutrient such as fertiliser bands may be less affected (Strong and Cooper 1980; Simpson and Pinkerton 1989). Decreased soil moisture reduces nutrient availability by increasing the penetration resistance of soil, which reduces root exploration of the profile (Tsegaye et al. 1995). The tortuosity of the ion diffusion pathways also increases as soil moisture decreases, which reduces the rates of N and P diffusion in soil solution (Watt et al. 2006). Plant turgor pressures will also eventually reduce under declining water availability so that plant shoot and root growth will decrease, although plants tend to partially compensate with an increased root: shoot ratio (Tinker and Nye 2000; He et al. 2002).
Evidently, the large variability that is possible in the natural soil environment can create a wide range of responses in the root systems of crop plants. Quantitative data are needed on the fertiliser response of whole root systems of crop plants growing under realistic conditions (Robinson 2001; Hodge 2004). We examined the early growth of wheat roots in response to both N and P fertiliser, at the 3-leaf and stem extension growth stages, growing under moderate moisture stress in undisturbed cores of 2 major alkaline soil types from south-eastern Australia. Our initial hypothesis was that, in a simulated low-fertility, semi-arid soil environment, wheat root growth in the very early stages would be restricted but would still respond to N and P fertiliser with an increased length of very fine roots in the vicinity of the respective fertiliser bands. Later in the season, the N response would include an increased length of thicker roots further down the profile, while any P fertiliser response would remain more restricted to the vicinity of the P fertiliser band.
Intact soil cores (0.15m diameter, 0.6m depth) were collected from fields with a history of cropping, in the Wimmera region (36[degrees]40.296'S, 142[degrees]17.25CE) and the Mallee region (35[degrees]57.331'S, 142[degrees]47.757'E). The previous crop in the Wimmera field was wheat and in the Mallee the field where the cores were collected was fallow. The soils were classified respectively as a Wimmera Grey Vertosol (alkaline smectitic vertisol) and an alkaline Subnatric Sodosol (saline and sodic subsoil) (Isbell 2002). The cores were collected when the profile was relative dry, thus minimising any potential shrinkage of these smectitic soils within the PVC sleeve, which may have led to preferential root proliferation around the wall of the core. Soils were characterised (Table 1) using standard methods (Rayment and Higginson 1992) for bulk density (BD) in the field, soil moisture content, pH (1 : 5 water and Ca[Cl.sub.2]), electrical conductivity (EC) in 1:5 w/v solution, plant-available N (2 M KC1), plant-available P (Colwell and Olsen P tests), and total nitrogen (%N) (Leco FP2000, MI). The permanent wilting point (PWP) of the soil was determined by the pressure plate method at -1500 kPa (Klute 1986) using a sieved sample (<2 mm), which avoided the shrink-swell problems of intact samples and the consequent difficulty of defining volume. Soil moisture at field capacity (FC) was determined by sampling from cores of the same size that had been saturated and then drained. Anaerobically mineralisable nitrate (AMN), as a more readily obtainable estimate of aerobically mineralisable nitrate, 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). Total P was determined by nitric/perchloric acid digestion.
Four fertiliser treatments were applied to both soils, consisting of (1) no fertiliser, (2) P fertiliser only, (3) N fertiliser only, (4) N and P fertiliser, using the rates and forms of 26 kg/ha P (K[H.sub.2][PO.sub.4]) and 50 kg/ha N (urea). Trace elements Mn, Zn, and Cu, in sulfate form, were watered in after planting. Equivalent extra K ([K.sub.2]S[O.sub.4]) was applied to treatments 1 and 3 to balance K applications across all treatments. All fertiliser applications were made to pre-wetted cores. Topsoils were disturbed to the depth of placement of the fertiliser, simulating the light minimum tillage that is typical of the region. Four wheat seeds (Triticum aestivum cv. Yitpi) were planted 40 mm deep and thinned to 3 plants per core after emergence. 'Side banding' was simulated by placing the granules in fertiliser bands 20 mm to the side of the seed row and 60 mm deep. The bands of N and P were kept separate and applied to opposite sides of the seed row (A side, P; B side, N), approximating the regional practice in wheat crops of placing the P fertiliser in the seed row and side banding or pre-sowing the main application of N fertiliser.
Plants were grown with soil moisture maintained at approximately 50% of plant-available water content in the topsoil and less in the subsoil (Fig. 1). The natural vertical distribution of available water within the intact cores was uneven and varied among cores. The moisture status was also constantly changing because the cores were watered using pulses of surface-applied water that were equivalent to a 10- or 20-mm rainfall event. The moisture regime can therefore only be regarded as nominal (Strong and Barry 1980). Plants were grown in a naturally lit (winter) glasshouse with a temperature regime of a 5[degrees]C night minimum and cooling that commenced at 20[degrees]C during the day.
Plants were harvested at the 3-leaf growth stage (GS 13) and at the commencement of stem elongation (GS30) (Zadoks et al. 1974), and these stages occurred approximately 27 and 71 days after sowing, respectively. Three replicates cores were harvested for each treatment at GS13 and 2 replicates at GS30. At each harvest, leaf number and plant characteristics were recorded and leaf area was measured (Li-Cor Li-3000, LAMBDA Laboratory Instruments, USA). Shoots were dried at 70[degrees]C, weighed, and ground, prior to nitric acid digestion and analysis for total P, and total N by Leco FP-428 Nitrogen Analyser (CSBP Limited, Bibra Lake, WA).
The soil cores were cut horizontally into 25-mm-depth sections for the top 100mm, and 100-mm-depth sections thereafter. The segments were also cut in half vertically, into an A side corresponding to the location of the P fertiliser band and a B side where the N fertiliser band had been applied, so that the effects of each fertiliser on each side of the core could be kept separate. For each soil segment, the moisture content (dried at 105[degrees]C) was measured on a subsample prior to pressure washing or wet sieving followed by collection of the roots on a sequence of 1-mm and 0.5-mm sieves.
Root material was split into subsamples as necessary and scanned to measure root length in 16 root diameter categories using WinRhizo[R] 2005 system (Regent Instruments Inc. Quebec City, Canada) (600 dpi, back lit image). Dead root material and rubbish had been separated by hand from the material collected on the 1-mm sieve but was very difficult to separate from the small material collected on the 0.5-mm sieve. Instead, estimates of the amount of dead material in each diameter class were made by scanning unsorted subsamples, then removing all the dead material and rescanning the subsamples. The proportional change in root length in each diameter class was then applied to the scans of the whole unsorted sample.
Statistical analysis of the shoot growth, and nutrient uptake into the shoots, was carried out using the general ANOVA procedure, modelling for main effects of soil type and fertiliser treatment (GENSTAT Release 9.1, Copyright 2006, Lawes Agricultural Trust, UK). Blocking effects were the glasshouse positional blocks of replicate cores.
For the roots, values of total cm root length per [cm.sup.3] of core volume per plant (RLD) were summed from the root diameter categories. The smallest diameter category of <50 [micro]m constituted only a small part of the root length and contained erroneous values, such as interpreting scratches and shadows on the scan as root material. Much of the material of this diameter may have washed through the sieve, so the category was disregarded. The statistical method of restricted maximum likelihood (REML) was used to model the RLD measurements over depth, using a selected variance--covariance structure. The selection was via a sequence of likelihood ratio tests on several nested models. A power model was selected as the most appropriate, in which the correlation between observations from the same plot decays as the delay between the observations increases. Data were log-transformed (natural log) with 0.2 added to values before transformation for time 1. After the variance-covariance model had been selected, the appropriate interactions were investigated via Wald tests and specific pairs of means were compared using the SEDLSI procedure in GENSTAT. Given a variate or table of parameter estimates (typically treatment means) and a corresponding standard error of the difference (s.e.d.), the procedure SEDLSI computes [delta] such that [[delta].sub.i] [congruent to] s.e.d, and constructs least significant intervals (LSIs) for graphical presentation of the means. LSIs are intervals (or error bars) that are designed to overlap where there is no significant difference between estimates, and to be disjoint where there are significant differences. Data were backtransformed after analysis and presented on the original scale.
[FIGURE 1 OMITTED]
A root diameter distribution was calculated for each soil segment from the proportion of RLD in each root diameter category. The correlations between the diameter distributions were summarised by the multivariate technique of principal component (PC) analysis. This dimension reduction technique identifies orthogonal linear re-combinations (PCs) of correlated attributes that summarise the principal sources of variability in the data. The PC analysis method is established in soil science (Burrough and Webster 1976) and has been used more recently to summarise the large datasets gathered in soil quality and precision agriculture research (Maddonni et al. 1999; Officer et al. 2004). Principal components were calculated based on the correlation matrix and only those PCs that had eigenvalues [greater than or equal to] 1 were retained (Jolliffe 1986; Khattree and Naik 2000). Each PC is a linear combination of the original variables, and the eigenvector loading associated with each variable represents the contribution of the original variable to the PC. In this study, only variables with eigenvector loadings [greater than or equal to] 0.3 were considered in the interpretation (Officer et al. 2004).
At GS 13, shoot weight, leaf area, and tiller number did not differ between the soil types or fertiliser treatments (P>0.05) (Table 2). In contrast, at GS30 the application of fertiliser significantly affected shoot growth, leaf area, and tiller number (P<0.001). Fertilising either with N, or with both N and P, increased shoot growth, leaf area, and tiller number to similar extents at GS30. There was no significant difference between the 2 soils in shoot response. There was no shoot response to P fertiliser at either growth stage.
The response of the plants to N fertiliser dominated the nutrient uptake patterns at GS30. Shoot N concentrations were relatively low for this growth stage (Reuter and Robinson 1986) and similar across the treatments, although there was a trend for shoot N concentration to decrease when only P fertiliser was applied (P=0.03). Total N uptake was closely related to shoot growth and increased when N was applied (P<0.001) (Table 2). Adding fertiliser N also altered shoot P concentration (P < 0.001). Shoot P concentrations were <0.15%, indicating a deficiency (Reuter and Robinson 1986), when only N fertiliser was applied to either soil, or when both N and P were applied to the Vertosol (P=0.033). The strong growth in response to N appears to have diluted the already marginal plant P content, although total P uptake into the shoots still increased significantly when N fertiliser was applied (P=0.006).
N and P recovery
Soil tests indicated a potential shortage of plant-available N in both soil profiles at the start of the experiment (Table 1). Total plant-available N, including N released by mineralisation as estimated by AMN, was approximately 105 mg N in each core. The plant-available N was therefore approximately doubled when 90ms N was added in the fertiliser. The 3 plants in each core were evidently competing strongly and scavenged the majority of the available N.
Plant recovery of available P was not of the same magnitude as the N recovery. Plant-available P values were relatively low in the Sodosol and marginal for a response to P in the Vertosol. Total plant-available P based on the Cowell P test was approximately 35 mg P in the Sodosol cores and 71 mg P in the Vertosol, so that plant-available P was approximately doubled when 50mg P was added in the fertiliser. However, plant uptake into the shoots was only 9-14 mg P, so that even if a similar amount of P was taken up into the roots, the plants were probably not under the same pressure to recover P as they were for N.
[FIGURE 2 OMITTED]
Root length at GS13 growth stage
At GS13, there were large differences in RLD between the soil types and a significant 3-way interaction between the effects of soil type, fertiliser treatment, and depth in the soil profile (Table 3). Plant roots were concentrated between 50 and 75 mm depth, which contained approximately 40% of the RLD, regardless of fertiliser treatment. There was greater overall root length density (RLD) at this growth stage in the Vertosol, which showed a marked response to the fertiliser treatments (Figs 2 and 3). There was no corresponding fertiliser response in the Sodosol. Root length increased in the Vertosol in the order of fertiliser treatments: no N or P = N only < P only <N and P. This response pattern was most marked between 50 and 75 mm, but also persisted down the profile to 200 mm. The N and P fertiliser had been banded separately on either side of the seed row, but there was no significant difference between RLD on either side of the cores, so that RLD increased in response to fertiliser to a similar extent on either side of the core.
[FIGURE 3 OMITTED]
Root diameter at GS13
Principal component analysis of the proportion of root length in each of 12 diameter classes at GS13 formed 4 PCs with eigenvalues >1, and explained 82% of the variation in the dataset (Table 4). The first PC formed a contrast between the proportion of roots of relatively fine diameter (100-200 [micro]m) and a medium diameter category (300-500 [micro]m). The second PC also formed a contrast, grouping fine/medium diameters (200-300 [micro]m) against both medium/large diameters (500 [micro]m-1 mm) and very fine diameter roots (50-100 [micro]m diameter). The third PC best summarised the largest root diameter classes of 750 [micro]m-2 mm. The fourth PC represented a very mixed set of categories and probably had little meaning.
The first 2 PCs mainly expressed the effect of depth on root diameter, grouping categories that had similar trends in the soil profile (Fig. 4a). The proportion of fine diameter roots increased in the upper part of the profile to a maximum at 50-75 mm depth and then decreased, while the proportion of medium diameter roots had a contrasting pattern, as indicated by PC1, of decreasing to minimum values at 50-75 mm depth and then increased deeper in the profile. Similarly, the proportion of fine/ medium diameters increased down the profile to a maximum at 200-400mm depth, and then decreased again, while the proportion of medium/large diameter roots was greater in the top and bottom of the soil profiles. The proportion of very fine roots had a different pattern, decreasing steadily with depth. The large diameter class comprised only a small proportion of root length, with some large diameter roots occasionally occurring in the very top and bottom of the profile, as indicated by the scatter in the score plot (Fig. 4b).
The diameter classes indicated by the PC analysis were adjusted so that the categories did not overlap and were applied to the RLD distribution. The categories constituted very fine (50-100 [micro]m), fine (100-2001 [micro]m), fine/medium (200-300 [micro]m), medium (300-500 [micro]m), and large (>500 [micro]m) (Table 5). At GS13, the average RLD over all depths constituted 39% fine diameter roots, 29% fine/medium diameter roots, 16% medium diameter roots, 13% very fine roots, and 4% large diameter roots. The root class distributions were very similar, despite the marked RLD response to fertiliser in the Vertosol (Fig. 5). Most of this increase in root length occurred at 50-75mm in the profile. At this depth, the proportion of fine diameter roots was 63% of RLD in the P only treatment, and 59% in the N and P treatment, so that the majority of the increase in root length consisted of fine diameter roots.
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
Root length at GS30
Between GS13 and GS30, the root length increased by a factor of 37, from 0.06 cm/[cm.sup.3] soil/plant (GS13) to 2.2 cm/[cm.sup.3] soil/plant (GS30). The RLD was now dominated by a proliferation of roots in the top 0-25 mm layer (Fig. 6). The effect of the fertiliser was consistent throughout the soil profile, so there was no significant fertiliser x depth interaction at GS30, although soil type interacted significantly with depth (Table 6). There was more RLD in the topsoil of the Sodosol, but RLD was similar below 100 mm in the 2 soils. Plants growing in the Sodosol where no fertiliser, or only P, had been applied had greater RLD than the same treatments applied to the Vertosol (Fig. 7). The RLD then increased to similar values in both soils when only N, or both N and P, were applied, which reflected the similarity in shoot response to N at this growth stage. The significant difference between RLD on either side of the soil cores was consistent regardless of the fertiliser treatments and therefore was an experimental artefact.
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
Root diameter at GS30
Principal component analysis of the proportion of root length in each of 15 diameter classes was carried out for wheat at the GS30 growth stage. The diameter classes formed 4 PCs with eigenvalues >1, explaining 81% of the variation (Table 7). Unlike the PC analysis of root diameters at GS13, no contrasts were formed, although the diameter groupings were similar. The first PC summarised a group of medium diameter categories (250-750 [micro]m) that were well correlated. The second PC summarised roots of relatively large diameter (>750 [micro]m). The third PC strongly expressed the proportion of very fine roots (50-100 [micro]m diameter). The fourth accounted for fine diameter roots (100-250 [micro]m).
The effect of depth on diameter once again dominated the PC analysis, as indicated by the consistent grouping by depth along the axis of the score plots, except for PC 3 (Fig. 8). The proportion of fine roots decreased steadily with depth at GS30, as the proportion of medium diameter roots increased (Table 8). There were more large-diameter roots in the upper part of both soil profiles. The proportion of very fine roots was similar at each depth, as indicated by the plot (Fig. 8b).
At GS30, the average RLD in the profile constituted 61% fine diameter roots, 27% medium diameter roots, 10% very fine roots, and 2% large diameter roots, which was similar to the distribution at GS13. The diameter distributions were again relatively insensitive to soil type and fertiliser treatment (Fig. 9), despite the marked RLD response to N.
The expected response of a proliferation of roots in the fertilised zones of a soil was not found in this study. We had expected increased RLD in the soil where bands of N and P fertiliser had been placed, especially early in plant growth, as well as changes in the root diameter distribution that indicated changes in the structure of the root system. Instead of specific responses to the location of the fertiliser, the roots responses consisted of spatially generalised increases in RLD and the root diameter distributions were similar, regardless of fertiliser treatment, soil type or plant growth stage.
[FIGURE 8 OMITTED]
At the 3-leaf growth stage, P fertiliser caused a limited but spatially generalised increase in RLD, the scale of which increased in the presence of N fertiliser, although this response was only recorded in 1 of 2 soil types and only in the roots not the shoots. The P response during early growth was followed by a marked shoot and root response to N at the stem elongation growth stage, which occurred in both soil types and regardless of the application of P. The RLD responses to N in later growth were also generalised both horizontally and vertically within the soil profile, and root diameter distributions again did not change, so that there was also no change in the structure of the root system in response to N.
A generalised response to P fertiliser on both sides of the soil cores was particularly unexpected. Many P uptake studies have found increases in root surface area in a fertilised zone, although changes in root length and branching have also been recorded outside the fertilised zone (Strong and Barry 1980; Sun et al. 2002). In this study, it is possible that specific physical responses to the location of P fertiliser occurred but that the generalised increase in RLD was the only response that was detectable at the scale of this study. For instance, a small root proliferation that was very specific to the location of the P fertiliser granules may have occurred but not have been identified, especially if a compensatory reduction in root growth also occurred within a few cm of the proliferation around the granules (Hodge et al. 2000). However, any location-specific changes that did occur were evidently of a small scale. Root hair density was not examined in this study, as the washing method breaks up the roots, and much of the very small root material probably passed through the sieves (Amato and Pardo 1994). An increase in the N or P influx rates into the plant may also have occurred (Robinson et al. 1994; Elliott et al. 1997); however, plant concentrations of N and P were not measured at GS13 due to the small size of the plants.
A second unexpected effect in this study was the triggering at GS13 of additional RLD when both N and P were applied to the Vertosol. It is not clear why N in the presence of P would increase RLD, when N alone had no effect, despite the later marked response to N at GS30. A similar marked increase in root length in response to both N and P fertiliser was found by Weligama et al. (2008). In fact, studies of root proliferation in response to N that have been done in a rich background of P, or vice versa, may have been examining the same synergistic response. Tennant (1976) found delays in the initiation and early growth of nodal roots when wheat plants were deprived of either N or P, compared with plants supplied with both nutrients. The delay effect was more severe in plants deprived of P and the same mechanism may have operated in this study. Additional N in a P-responsive situation may therefore have sped up the initiation of roots. An early maximisation of RLD may be an important factor in the competition between plants to exploit soil resources (Robinson 1996; Hodge et al. 2000; Lynch and Ho 2005). Faster initiation of basically the same root system, rather than a specific alteration in the structure of the root system (such as more very fine roots), explains why the root diameter distributions did not change when the plants responded to fertiliser at GS13.
[FIGURE 9 OMITTED]
In contrast to GS13, shoots and roots of wheat responded strongly to N fertiliser later in the season at GS30, although there was no additional response when P fertiliser was also applied. There was also no growth response to P (without N) at GS30 and the relatively small changes in the root systems at GS13 evidently conferred no advantage to later growth. Instead, the fertiliser response at GS30 was as if only N had been applied and total P uptake was determined by the N response and ultimately RLD, as has been found in other studies (Solaiman et al. 2007).
The root response to N at GS30 also consisted of a generalised increase in root length in the topsoil. Once again, this was not the expected specific response to N, rather it appeared to be a continuation of the rapid expansion of roots in the topsoil that is the expected root growth pattern of wheat at this growth stage (Belford et al. 1987). Plant-available nitrate in the soil is very mobile, and when sufficiently supplied, it is most efficiently acquired by the rapid extension of roots that increase the root surface area within a soil volume, or by manipulation of the influx rate, rather than by an increase in the surface area contacting the fertilised zone (Robinson et al. 1994; Robinson 2001; Dunbabin et al. 2003b). Therefore, an increase in root length was expected, but lower in the profile as the plant roots responded to the leaching of nitrate. A change in the root diameter distribution was also expected because the plant would be extending existing (possibly axial) roots, rather than extending new roots. However, in this study, the increase in root length in response to N occurred in the topsoil, not the subsoil, and root diameter decreased, as the majority of roots produced were of relatively fine diameter. Approximately 60% of the root length consisted of relatively fine-diameter roots (100-250 [micro]m) at GS30, regardless of fertiliser treatment and the changes in RLD, which probably indicated an increase in the initiation of lateral roots. Qin et al. (2004) also found that the diameter class distribution of mature wheat roots was relatively insensitive to agronomic treatments.
Lateral increases in RLD in the topsoil to produce a more branched, finer diameter, root system is a typical P-foraging mechanism in plants (Vance et al. 2003), and reduced plant-available moisture in this study may have triggered a switch in the plants to a P-style acquisition of N (Hodge 2004). Soil moisture and plant-available N were both constrained in this study. Low soil moisture will curtail the mobility of the nitrate, limiting both the movement of N to the plants by mass flow and leaching down the profile. Soil moisture was approximately 50% of plant-available water-holding capacity in the top 100mm, and soil was relatively dry below this depth, especially at GS30. The topsoil was also allowed to dry out periodically, between watering events. Dry soil conditions have been shown to cause a reduction in root growth and a change in the morphology (Acuna and Wade 2005; Kirkegaard and Lilley 2007), possibly increasing branching of the root system in the topsoil (Watt et al. 2006). Transpirational efficiency is also expected to increase (Alston 1976; Norton and Wachsmann 2006; Sadras and Angus 2006). In this study, moisture stress appears to have triggered the formation of a root system that was more suitable to dry conditions. The plants appear to have used the fertiliser N to increase the scale of this response, possibly because the increased root surface area of a proliferation of fine roots was useful to acquire N that was immobilised in the relatively dry topsoil (Hodge 2004).
In this study, realistic plant growth conditions, of competition under moisture constraints, appeared to trigger a 'who dares wins' strategy of increasing the rate of root initiation and therefore increasing the size of the root system in response to fertiliser. Evidently, plant foraging strategy is more complex than predictions based only on ion-diffusion theory under ideal moisture conditions (Robinson et al. 1994; Dunbabin et al. 2003a). Exactly why plants grow more roots in response to an increase in nutrient supply seems to be an enduring puzzle (Robinson et al. 1994; Hodge 2004). Instead of increasing the influx rate, or growing a few specialised roots into a nutrient patch, plants often grow many extra roots, which must then be maintained when the supply of nutrients is exhausted (van Vuuren et al. 1996). A general root proliferation response to fertiliser is therefore risky in terms of carbon allocation and does not make physiological sense in terms of efficient nutrient acquisition in single plants. It may however confer a substantial advantage when plants are competing for a finite resource (Robinson et al. 1999; Hodge 2004; Liao et al. 2004). In this study, a generalised proliferation of roots evidently did make sense for wheat plants that were competing for nutrients in a semi-arid soil environment.
The authors would like to acknowledge the very patient technical assistance of Graham Price, Corey Mathews, and Aleem Khan. Funding for the project was provided through the Department of Primary Industries, Victoria, and the Grains Research and Development Corporation through the Nutrient Management Initiative (project UM00023). The authors would particularly like to thank an anonymous reviewer for their considerable input into the paper.
Manuscript received 30 April 2008, accepted 2 December 2008
ABARE (2007) Outlook for 2007-08 and industry productivity. Australian Grains Report 07.2, Australian Bureau of Agricultural and Resource Economics. p. 16.
Acuna TLB, Wade LJ (2005) Root penetration ability of wheat through thin wax-layers under drought and well-watered conditions. Australian Journal of Agricultural Research 56, 1235-1244. doi: 10.1071/ AR05067
Alston AM (1976) Effects of depth of fertiliser placement on wheat grown under three water regimes. Australian Journal of Agricultural Research 27, 1-10. doi: 10.1071/AR9760001
Amato M, Pardo A (1994) Root length and biomass losses during sample preparation with different screen mesh sizes. Plant and Soil 161, 299-303. doi: 10.1007/BF00046401
Belford RK, Klepper B, Rickman RW (1987) Studies of intact shoot-rnot systems of field-grown winter wheat. II. Root and shoot developmental patterns as related to nitrogen fertilizer. Agronomy Journal 79, 310-319.
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
Burrough PA (1993) Soil variability: a late 20th century view. Soils and Fertilizers 56, 529-562.
Burrough PA, Webster R (1976) Improving a reconnaissance soil classification by multivariate methods. Journal of Soil Science 27, 554-571. doi: 10.1111/j.1365-2389.1976.th02024.x
Diggle AJ, Bowden JW (1990) The effect of rate of water addition on the response of wheat roots to added nitrogen in a leaching environment. Australian Journal of Soil Research 28, 973-980. doi: 10.1071/ SR9900973
Diggle AJ, Bowden JW (1991) The response of wheat tops and roots grown in a leaching environment to rates of nitrogen added as calcium nitrate or organic residues containing 1, 2 or 6% nitrogen. Australian Journal of Agricultural Research 42, 1053-1064. doi: 10.1071/AR9911053
Dunbabin V, Diggle A, Rengel Z (2003b) 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 V, Rengel Z, Diggle A (2003a) Root architecture and nutrient capture--the complex riddle of what constitutes optimality of root form and function. In Innovative soil plant systems for sustainable agricultural practices'. (Eds JM Lynch, JS Schepers, I Unver) pp. 2-16. (OECD: Paris)
Elliott DE, Reuter DJ, Reddy GD, Abbott RJ (1997) Phosphorus nutrition of spring wheat (Triticum aestivum L.). 1. Effects of phosphorus supply on plant symptoms, yield, components of yield, and plant phosphorus uptake. Australian Journal of Agricultural Research 48, 855-868. doi: 10.1071/A96159
He YQ, Zhu YG, Smith SE, Smith FA (2002) Interactions between soil moisture content and phosphorus supply in spring wheat plants grown in pot culture. Journal of Plant Nutrition 25, 913-925. doi: 10.1081/PLN120002969
Herrera JM, Stamp P, Liedgens M (2007) Interannual variability in root growth of spring wheat (Triticum aestivum L.) at low and high nitrogen supply. European Journal of Agronomy 26, 317-326. doi: 10.1016/j. eja.2006.11.003
Hodge A (2004) The plastic plant: root responses to heterogeneous supplies of nutrients. New Phytologist 162, 9-24. doi: 10.1111/j.1469-8137.2004.01015.x
Hodge A, Stewart J, Robinson D, Griffiths SB, Fitter HA (2000) Spatial and physical heterogeneity of N supply from soil does not influence N capture by two grass species. Functional Ecology 14, 645-653. doi: 10.1046/j.1365-2435.2000.t01-1-00470.x
Huang M, Deng X, Zhao Y, Zhou S, Inanaga S, Yamada S, Tanaka K (2007) Water and nutrient use efficiency in diploid, tetraploid and hexaploid wheats. Journal of Integrative Plant Biology 49, 706-715. doi: 10.1111 / j.1744-7909.2007.00463.x
Isbell RF (2002) 'The Australian Soil Classification.' Revised edn (CSIRO Publishing: Collingwood, Vic.)
Jolliffe IT (1986) 'Principal component analysis.' (Springer-Verlag: New York)
Kamper M, Claassens AS (2005) Exploitation of soil by roots as influenced by phosphorus applications. Communications in Soil Science and Plant Analysis 36, 393-402. doi: 10.1081/CSS-200043109
Khattree R, Naik DN (2000) 'Multivariate data reduction and discrimination with SAS software.' (SAS Institute Inc.: Cary, NC)
Kirkegaard JA, Lilley JM (2007) Root penetration rate--a benchmark to identify soil and plant limitations to rooting depth in wheat. Australian Journal of Experimental Agriculture 47, 590-602. doi: 10.1071/ EA0607 I
Klute A (1986) Water retention: Laboratory methods. In 'Methods of soil analysis, Part 1: Physical and mineralogical methods'. 2nd edn (Ed. A Klute) pp. 635-662. (American Society of Agronomy: Madison, WI)
Liao M, Fillery IRP, Palta JA (2004) Early vigorous growth is a major factor influencing nitrogen uptake in wheat. Functional Plant Biology 31, 121-129. doi: 10.1071/FP03060
Lynch JP, Ho MD (2005) Rhizoeconomics: Carbon costs of phosphorus acquisition. Plant and Soi1 269, 45-56. doi: 10.1007/s11104-004-1096-4
Maddonni GA, Urricariet S, Ghersa CM, Lavado RS (1999) Assessing soil quality in the rolling pampa, using soil properties and maize characteristics. Agronomy Journal 91, 280-287.
Manske GGB, Ortiz-Monasterio JI, Van Ginkel M, Gonzalez RM, Rajaram S, Molina E, Vlek PLG (2000) Traits associated with improved P-uptake efficiency in CIMMYT's semidwarf spring bread wheat grown on an acid Andisol in Mexico. Plant and Soil 221, 189-204. doi: 10.1023/ A:1004727201568
Marschner P, Solaiman Z, Rengel Z (2005) Growth, phosphorus uptake, and rhizosphere microbial-community composition of a phosphorus-efficient wheat cultivar in soils differing in pH. Journal of Plant Nutrition and Soil Science 16g, 343-351. doi: 10.1002/jpln.200424101
McBeath TM, Armstrong RD, Lombi E, McLaughlin MJ, Holloway RE (2005) Responsiveness of wheat (Triticum aestivum) to liquid and granular phosphorus fertilisers in southern Australian soils. Australian Journal of Soil Research 43, 203-212. doi: 10.1071/SR04066
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 (2003) Evaluating physicochemical constraints of Calcarosols on wheat yield in the Victorian southern Mallee. Australian Journal of Agricultural Research 54, 487-497. doi: 10.1071/AR02168
Officer S J, Kravchenko A, Bollero GA, Sudduth KA, Kitchen NR, Wiebold WJ, Palm HL, Bullock DG (2000) Relationships between soil bulk electrical conductivity and the principal component analysis of topography and soil fertility values. Plant and Soil 258, 269-280. doi: 10.1023/B:PLSO.0000016557.94937.ed
Passioura JB (2006) The perils of pot experiments. Functional Plant Biology 33, 1075-1079. doi: 10.1071/FP06223
Qin R, Stamp P, Richner W (2004) Impact of tillage on root systems of winter wheat. Agronomy Journal 96, 1523-1530.
Rayment GE, Higginson FR (1992) 'Australian laboratory handbook of soil and water chemical methods.' (Inkata Press: Melbourne, Vic.)
Reuter DJ, Robinson JB (1986) 'Plant analysis: An interpretation manual.' (Inkata Press: Melbourne, Sydney)
Robinson D (1994) Tansley review No. 73. The responses of plants to non-uniform supplies of nutrients. New Phytologist 127, 635-674. doi: 10.1111/j.1469-8137.1994.tb02969.x
Robinson D (1996) Resource capture by localized root proliferation: Why do plants bother? Annals of Botany 77, 179-186. doi: 10.1006/ anbo. 1996.0020
Robinson D (2001) Root proliferation, nitrate inflow and their carbon costs during nitrogen capture by competing plants in patchy soil. Plant and Soil 232, 41-50. doi: 10.1023/A:1010377818094
Robinson D, Linehan D J, Gordon DC (1994) Capture of nitrate from soil by wheat in relation to root length, nitrogen inflow and availability. New Phytologist 128, 297-305. doi: 10.1111/j.1469-8137.1994. tb04013.x
Robinson D, Hodge A, Griffiths BS, Fitter AH (1999) Plant root proliferation in nitrogen-rich patches confers competitive advantage. Proceedings of the Royal Society of London. Series B. Biological Sciences 266, 431-435. doi: 10.1098/rspb.1999.0656
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, Roget DK (2004) Production and environmental aspects of cropping intensification in a semiarid environment of southeastern Australia. Agronomy Journal 96, 236-246.
Simpson JR, Pinkerton A (1989) Fluctuations in soil moisture, and plant uptake of surface applied phosphate. Fertilizer Research 20, 101-108. doi: 10.1007/BF01055434
Solaiman Z, Marschner P, Wang D, Rengel Z (2007) Growth, P uptake and rhizosphere properties of wheat and canola genotypes in an alkaline soil with low P availability. Biology and Fertility of Soils 44, 143-153. doi: 10.1007/s00374-007-0188-8
Sparling GP, Searle PL (1993) Dimethyl sulpboxide 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
Strong WM, Cooper JE (1980) Recovery of nitrogen by wheat from various depths in a cracking clay soil. Australian Journal of Experimental Agriculture andAnimal Husbandry 20, 82 87. doi: 10.1071/EA9800082
Sun H, Zhang F, Li L, Tang C (2002) The morphological changes of wheat genotypes as affected by the levels of localized phosphate. Plant and Soil 245, 233-238. doi: 10.1023/A:1020472608609
Tennant D (1976) Root growth of wheat. I. Early patterns of multiplication and extension of wheat roots including effects of levels of nitrogen, phosphorus and potassium. Australian Journal of Agricultural Research 27, 183-196. doi: 10.1071/AR9760183
Tinker PB, Nye PH (2000) 'Solute movement in the rhizosphere.' (Oxford University Press: New York)
Tsegaye T, Mullins CE, Diggle AJ (1995) Modelling pea (Pisum sativum) root growth in drying soil. A comparison between observations and model predictions. New Phytologist 131, 179-189. doi: 10.1111/j.1469- 8137.1995.tb05719.x
Valizadeh GR, Rengel Z, Rate AW (2002) Wheat genotypes differ in growth and phosphorus uptake when supplied with different sources and rates of phosphorus banded or mixed in soil in pots. Australian Journal of Experimental Agriculture 42, 1103-1111. doi: 10.1071/EA01087
van Vuuren MMI, Robinson D, Griffiths BS (1996) Nutrient inflow and root proliferation during the exploitation of a temporally and spatially discrete source of nitrogen in soil. Plant and Soil 178, 185-192. doi: 10.1007/ BF00011582
Vance CP, Uhde-Stone C, Allan DL (2003) Phosphorus acquisition and use: critical adaptations by plants for securing a nonrenewable resource. New Phytologist 157, 423-447. doi: 10.1 006/j.1469-8137.2003.00695.x
Watt M, Silk WK, Passioura JB (2006) Rates of root and organism growth, soil conditions, and temporal and spatial development of the rhizosphere. Annals of Botany 97, 839-855. doi: 10.1093/aob/mc1028
Weligama C, Tang C, Sale PWG, Conyers MK, Liu DL (2008) Localised nitrate and phosphate application enhances root proliferation by wheat and maximises rhizosphere alkalisation in acid subsoil. Plant and Soil 312, 101-115. doi: 10.1007/s11104-008-9581-9
Zadoks JC, Chang TT, Konzak CF (1974) A decimal code for the growth stages of cereals. Weed Research 14, 415-421. doi: 10.1111/j.1365- 3180.1974.tb01084.x
Zhang XK, Rengel Z (2002) Temporal dynamics of gradients of phosphorus, ammonium, pH, and electrical conductivity between a di-ammonium phosphate band and wheat roots. Australian Journal of Agricultural Research 53, 985-992. doi: 10.1071/AR01149
S. J. Officer(A,E), V. M. Dunbabin, (B) R. D. Armstrong, (A) R. M. Norton, (C) and G. A. Kearney (D)
(A) Department of Primary Industries, PMB 260, Horsham, Vic. 3401, Australia.
(B) Tasmanian Institute of Agricultural Research, University of Tasmania, Private Bag 54, Hobart, Tas. 7001, Australia.
(C) The University of Melbourne, PMB 260, Horsham, Vic. 3401, Australia.
(D) 36 Paynes Road, Hamilton, Vic. 3300, Australia.
(E) Corresponding author. Email: Sally.Officer@dpi.vic.gov.au
Table 1. Characteristics of a Sodosol and Vertosol from the dryland cropping zone of south-eastern Australia, before trial commencement BD, Bulk density; AMN, anerobically mineralisable nitrogen; PBI, phosphate buffering index Depth BD pH EC paste Total N (mm) (g/[m.sup.3]) Ca[Cl.sub.2] Water (dS/m) (%) Vertosol 0-100 1.1 7.91 8.4 0.15 0.08 100-200 1.1 7.91 8.6 0.16 0.07 200-300 1.2 7.97 8.8 0.19 0.07 300-400 1.4 8.03 8.9 0.24 0.06 400-500 1.4 8.11 9.0 0.27 0.05 500-600 1.4 8.21 9.2 0.30 0.04 Sodosol 0-100 1.2 8.04 8.8 0.23 0.09 100-200 1.2 8.13 9.3 0.49 0.05 200-300 1.3 8.32 9.2 0.83 0.04 300-400 1.3 8.42 9.2 1.11 0.04 400-500 1.3 8.46 9.2 1.32 0.04 500-600 1.3 8.50 9.1 1.47 0.03 Depth Mineral N AMN Total P PBI (mm) ([micro]g/g) (mg/g) ([micro]g/g) Vertosol 0-100 4.55 14.47 0.19 196 100-200 4.47 6.49 0.14 204 200-300 2.83 6.12 0.13 206 300-400 1.65 7.04 0.12 205 400-500 2.87 3.70 0.11 194 500-600 2.27 3.03 0.10 190 Sodosol 0-100 4.21 6.49 0.15 158 100-200 2.44 7.34 0.10 156 200-300 1.74 7.90 0.10 153 300-400 0.99 5.92 0.09 169 400-500 0.95 5.47 0.09 171 500-600 0.87 5.25 0.08 165 Depth Colwell P Olsen P (mm) (mg/kg) Vertosol 0-100 20.18 9.50 100-200 5.00 2.23 200-300 4.33 2.00 300-400 3.33 2.03 400-500 3.00 1.43 500-600 3.00 1.57 Sodosol 0-100 7.09 3.53 100-200 2.00 0.77 200-300 2.00 0.87 300-400 2.00 0.93 400-500 2.00 0.83 500-600 2.00 0.60 Table 2. Shoot growth and nutrient uptake of winter wheat at 2 growth stages after 4 fertiliser treatments were applied to intact cores of a Vertosol and a Sodosol All results are the mean of the 2 soils, as indicated by the statistical analysis, except for shoot P concentration, where both soil type and fertiliser treatment effects were significant. n.s., Not significant Fertiliser treatment: No N P N N l.s.d. or P only only and P (P=0.05) 3-leaf (GS13) growth stage Shoot growth (g/plant) 0.07 0.08 0.08 0.08 n.s. Leaf area ([cm.sup.2]/plant) 24.2 28.8 26.9 26.7 n.s. Tiller no. (no./plant) 0.1 0.0 0.2 0.2 n.s. Stem elongation (GS30) growth stage Shoot growth (g/plant) 1.7 2.0 3.1 3.2 0.7 Leaf area ([cm.sup.2]/plant) 213 206 360 329 53 Tiller no. (no./plant) 1.6 2.1 3.0 3.1 0.7 Shoot N (%) 1.8 1.4 1.7 1.6 0.2 Shoot P (%): Vertosol 0.18 0.17 0.14 0.14 0.03 Sodosol 0.16 0.17 0.11 0.17 0.03 Total shoot P (mg/plant) 2.9 3.5 4.0 4.6 1.1 Total shoot N (mg/plant) 30.3 28.2 53.7 50.1 11 Table 3. Results of a restricted maximum likelihood analysis of the root length density of wheat at the GS13 growth stage, in the profile of 2 soil types with 4 fertiliser treatments Only relevant or significant fixed effects are shown Fixed term d.f. F-statistic P Profile depth 7 107.29 <0.001 Soil type 1 26.04 <0.001 Fertiliser 3 2.96 0.033 Profile depth.soil type 7 0.54 0.806 Profile depth.fertiliser 21 1.33 0.154 Soil type.fertiliser 3 3.99 0.009 Profile depth.soil type.ferfliser 21 1.64 0.040 Table 4. Principal component (PC) analysis of the distribution of wheat root diameter at the CS13 growth stage Values were derived from the proportion of root length in each diameter class. Results of the analysis are summarised as the variance (eigenvalue) accounted for by each PC and the coefficients applied to each diameter class to create each PC (PC composition) PC: 1 2 3 4 Eigenvalue: 4.3 2.9 1.4 1.1 Variation (A) (%): 36.1 24.1 11.6 9.3 Root diameter ([micro]m) PC composition 50-75 0.05 0.37 0.27 0.46 75-100 -0.27 0.30 0.36 0.34 100-150 -0.42 0.12 0.11 -0.14 150-200 -0.38 -0.12 -0.14 -0.32 200-250 -0.04 0.00 -0.25 0.14 250-300 0.26 -0.40 -0.01 0.34 300-350 0.41 -0.17 0.18 0.13 350-400 0.41 0.06 0.26 -0.18 400-500 0.36 0.22 0.10 -0.38 500-750 0.24 0.36 -0.29 -0.24 750-1000 0.11 0.31 -0.54 0.10 1000-2000 0.02 0.21 -0.46 0.40 (A) Proportion of the total variance accounted for by each PC. Table 5. Root diameter distribution of wheat at growth stage GS13 Diameter is expressed as a percentage of the root length density in each diameter category at each depth in the soil profile, and showing the average coefficient of variation (CV) for each class Depth Very fine Fine Fine/medium (mm) 50-100 [micro]m 100-200[micro]m 200-300 [micro]m 0-25 21 40 16 25-50 17 48 23 50-75 15 55 22 75-100 12 52 25 100-200 11 40 34 200-300 9 28 41 300-400 8 24 39 400-500 8 22 29 CV (%) (51) (44) (44) Depth Medium Large (mm) 300-500[micro]m >500[micro]m 0-25 9 13 25-50 8 3 50-75 7 1 75-100 10 1 100-200 14 1 200-300 20 2 300-400 27 2 400-500 32 9 CV (%) (80) (178) Table 6. Results of a restricted maximum likelihood analysis of root length density of wheat at GS30 growth stage, in intact cores of 2 soil types with 4 fertiliser treatments Each core was segmented horizontally into 8 depths and vertically into 2 sides. Only the significant fixed effects are shown Fixed terms d.f. F-statistic P Fertiliser 3 34.89 <0.001 Profile depth 7 206.18 <0.001 Soil type 1 9.36 0.003 Fertiliser.soil type 3 6.46 <0.001 Profile depth.soil type 7 8.27 <0.001 Core side 1 7.03 0.009 Table 7. Principal component (PC) analysis of the distribution of wheat root diameter at the GS30 growth stage Values were derived from the proportion of root length in each diameter class. Results of the analysis are summarised as the variance (eigenvalue) accounted for by each PC and the coefficients applied to each diameter class to create each PC (PC composition) PC: 1 2 3 4 Eigenvalue: 4.9 3.8 2.0 1.5 Variation (A) (%): 32.6 25.4 13.0 10.0 Diameter ([micro]m) PC composition 50-75 0.11 -0.19 -0.44 -0.21 75-100 -0.15 -0.16 -0.52 -0.22 100-150 -0.28 -0.13 -0.26 -0.36 150-200 -0.06 0.00 0.19 -0.72 200-250 0.28 0.14 0.29 -0.46 250-300 0.39 0.10 0.16 -0.18 300-350 0.43 0.01 0.00 0.00 350-400 0.42 -0.07 -0.12 0.07 400-500 0.40 -0.15 -0.17 0.06 500-750 0.31 -0.29 -0.12 0.01 750-1000 0.12 -0.42 -0.06 -0.01 1000-2000 -0.04 -0.43 0.06 0.02 2000-3000 -0.08 -0.41 0.26 0.02 3000-4000 -0.08 -0.38 0.33 0.01 >4000 -0.08 -0.33 0.28 0.02 (A) Proportion of the total variance accounted for by each PC. Table 8. Root diameter distribution of wheat at growth stage GS30, in categories formed by principal component analysis Diameter is expressed as a percentage of the root length density in each diameter category, at each depth in the soil profile, and the average coefficient of variation (CV) for each category Depth Very fine Fine Medium (mm) 50-100 [micro]m 100-250 [micro]m 250-750 [micro]m 0-25 12 70 16 25-50 11 67 18 50-75 10 68 20 75-100 10 66 24 100-200 10 62 26 200-300 9 58 32 300-400 9 57 33 400-500 8 52 39 CV (%) (36) (11) (20) Depth Large (mm) >750 [micro]m 0-25 2 25-50 3 50-75 2 75-100 1 100-200 1 200-300 1 300-400 1 400-500 2 CV (%) (57)
|Printer friendly Cite/link Email Feedback|
|Author:||Officer, S.J.; Dunbabin, V.M.; Armstrong, R.D.; Norton, R.M.; Kearney, G.A.|
|Publication:||Australian Journal of Soil Research|
|Date:||Feb 1, 2009|
|Previous Article:||Identifying fertiliser management strategies to maximise nitrogen and phosphorus acquisition by wheat in two contrasting soils from Victoria,...|
|Next Article:||Plant availability of phosphorus from fluid fertiliser is maintained under soil moisture deficit in non-calcareous soils of south-eastern Australia.|