Phosphorus nutrition in southern Mallee and northern Wimmera farming systems

Ben Jones, Birchip Cropping Group and Mallee Focus

Introduction

A series of low rainfall years, and high fertiliser prices, have focused attention on phosphorus nutrition in southern Mallee and northern Wimmera farming systems. There is a need to reduce costs as much as possible, without compromising yields in the event that another average to good rainfall year is experienced! At the same time, there have been developments in the use of Colwell P soil tests and their use in conjunction with the Phosphorus Buffer Index that should be tested against crop performance in the area. This report presents soil and plant measurements made on paired no-till and conventional (till) farmer paddocks between 2006 and 2008, with the aim of highlighting phosphorus-related issues for crop management in the area.

Methods

Soil samples were oven-dried at 40ºC, then analysed for Colwell P and PBI by CSBP, Bibra Lake, Western Australia. Plant samples were oven-dried at 60ºC, then analysed for total P using ICP-AES after extraction with nitric acid and hydrogen peroxide (also CSBP).

Results

Profile of phosphorus in soil

The profile of available (Colwell) phosphorus with depth was measured twice, pre-sowing in 2006 and pre-sowing in 2007 at some sites (thanks to over-enthusiastic analysis by CSBP). In the 2007 measurements, Phosphorus Buffer Index (PBI) was also measured on some soil profiles. There was little difference between the 2006 and 2007 measurements of Colwell P at depth, and so 2006 measurements only have been presented. Where available, critical P calculated from 2007 PBI according to Moody (2007) has been included as a reference (though strictly it has no meaning outside a 0-10cm sample).

The majority of phosphorus was concentrated in the 0-10 cm horizon (Figure 1a, b). In some paddocks small amounts of phosphorus was present in the 10-40 cm horizon (Sea Lake, Yaapeet no till, Patchewollock till) but in most paddocks and at deeper depths the amounts were negligible. Higher levels of Colwell P at depth may reflect leaching, or a history of deep fertiliser placement. These paddocks had a combination of sandy texture and acidity in topsoil (5.6-6.7 pH in CaCl₂), and leaching seems more likely. The other acid topsoils in the study came from clay textured Carron paddocks (pH 6.7-6.8).

According to critical phosphorus calculated from PBI, phosphorus measured by Colwell P was also readily used by the plant in surface soils, less so in subsoils. Colwell P in topsoil was sufficient to achieve 90% of maximum yield in cereal crops at all paddocks apart from Minyip, where both Till and No Till paddocks had low Colwell P for their PBI in 2006. There were no consistent patterns with Till or No Till paddocks in 2006, but the measurements highlight the importance of plant access to phosphorus in topsoil in both systems and at most locations.

Figure 1. Available (Colwell) phosphorus between 0 and 100 cm, in soils from no till (a) and till (b) focus paddocks pre-sowing in 2006. Critical Colwell P for 90% maximum yield for wheat (in a 0-10cm soil sample) is included as a reference, where available.

Topsoil phosphorus 2006-2008

Colwell P was measured on topsoil (0-10cm) each year pre-sowing, and also after harvest in 2007. PBI was also measured at each sampling except 2006. All paddocks except the Minyip paddocks had sufficient phosphorus at sowing (Figure 2a, b) to achieve 90% of maximum wheat yield as defined by Moody (2007). In many paddocks there was a clear drop in Colwell P from sowing to harvest in 2007, interestingly accompanied by a slight increase in PBI in some of the No Till paddocks. The decrease in Colwell P also occurred in the Sea Lake Till paddock, which was in fallow in 2007. This possibly reflects normal fixation and immobilisation processes, but a similar drop was not measured in the Donald Till paddock (also in fallow). Apart from Harvest 2007, Colwell P was quite steady or increased between 2006 and 2008. The exception was Sea Lake No Till in 2007, which was almost double the pre-sowing measurements in 2006 and 2008.

Figure 2. Available (Colwell) phosphorus in topsoil over measurements between 2006 and 2008 (PS = pre-sowing, H = post-harvest), in soils from no till (a) and till (b) focus paddocks. The post-harvest measurement from Yaapeet No Till in 2007 was lost at analysis.

Plant phosphorus uptake

Mineral phosphorus in the soil pre-sowing is added to by fertiliser, and further changed by processes of fixation, mineralisation/immobilisation (microbial activity), and to a lesser extent, leaching and runoff. Phosphorus is typically highly fixed to soil particles and also moves with them – in erosion, biological and mechanical mixing processes, and with fine soil particles in water (eluviation). Different systems have different sowing configurations (up to 61cm spacing, inter-row sowing), which may enhance or limit their ability to extract phosphorus from topsoil and different areas within topsoil (eg. previous crop inter-row).

Plant uptake (tissue test) is a more certain test of phosphorus supply than measurements of soil mineral phosphorus (Colwell P and PBI). Ultimately plants need to put a certain amount of phosphorus into grain, from literature at least 0.17% and possibly over 0.4% (Table 1). This translates to between 1.7 and 4 kg P/tonne of yield. Studies on a range of Western Australian soils indicate that canola requires less phosphorus for maximum yield than wheat and field pea (Bolland et al. 2006).

Table 1. Grain phosphorus measurements in cereals from literature.

Plant uptake of phosphorus, particularly at GS30, was small compared to the amount of phosphorus applied as fertiliser, and measured by the Colwell P test (Table 2), but 5-30 times more than was applied in seed. Higher GS30 phosphorus usually reflected higher dry matter growth, rather than phosphorus percentage (there was a much greater range in dry matter than phosphorus percentage). Tissue measurements also ignore any phosphorus present in the roots.

For some sites there was little change in plant P uptake between GS30 and GS65 (eg. Patchewollock 2007), but at other sites P uptake increased up to five times during the same period. At sites where Colwell P dropped appreciably between pre-sowing and harvest measurements in 2007 (Figure 2a, b), the drop was far greater than would be accounted for by plant P uptake to GS65, especially if some plant P uptake from fertiliser was also allowed for. There could be some impact of plant P uptake after anthesis but in other studies in a similar rainfall environment the amounts have been negligible (Zubaidi et al. 1999).

Table 2. Plant phosphorus uptake at GS30 and GS65 for crops in focus paddocks, compared with soil, fertiliser and seed sources of P.

·         All soil measurements from topsoil (0-10cm).

·         Critical = estimated from PBI and relationship of Moody.

·         Seed = estimated from seed rate and P concentrations in Table 6 of Bolland et al. (1999).

·         Fert = not all fertiliser details have been received from farmers; this is likely to be incomplete particularly in 2007.

Phosphorus uptake with respect to nitrogen

A confounding factor in interpreting plant phosphorus concentration is nitrogen nutrition, which affects the amount of phosphorus taken up by plants during growth (in addition to supply, Ziadi et al. 2008). Ziadi et al. defined two relationships between tissue nitrogen and phosphorus, for nitrogen unlimited and nitrogen limited spring wheat crops in Canada, yielding at most between 1.6 and 4.1 t/ha on 250-380mm growing season rainfall. These may have some relevance to Mallee and Wimmera crops in better years but are included for reference. Sadras (2006) considered the issue in terms of relationships between plant N and P uptake at maturity, across a wide range of field experiments (and also oilseeds and legumes), defining a N:P ratio often achieved at maximum yield (Table 1).

Compared to the Canadian study, most Mallee and Wimmera cereal crops were phosphorus limited in either system (high plant N concentration for the plant P concentration) both at GS30 (Figure 3a) and GS65 (Figure 3b). When total uptake was considered, Mallee and Wimmera crops at anthesis were closer to the N:P uptake relationship of Sadras (both cereals, Figure 4a and legumes, Figure 4b). Measured this way, crops still tended to have high N uptake for the level of P uptake (again indicating relatively P limited), particularly 2008 Till wheat crops. Some other crops further from the Sadras line were barley crops at Yaapeet (No Till in 2008, Till in 2007), the Till wheat crops at Sea Lake in 2008 and Culgoa in 2007. The Patchewollock Till oat crop in 2007 stood out as the only nitrogen limited crop.

Given P and N harvest indices of 0.6 – 0.75 (Zubaidi et al. 1999), and possible grain P concentrations of 0.17 to over 0.4, P in these seasons is probably more limiting than N for absolute grain yield, to the extent that uptake of both nutrients may exceed water-limited yield potential anyway. At the lower limit of approximately 2.5 kg P uptake/t grain yield, all crops in Figure 4a have taken up sufficient P for 1 t/ha yield, and most have sufficient for 2 t/ha yield. It should be noted that P uptake particularly in 2008 would be hampered by the lack of pre-anthesis rain, given that most P comes from surface soil and in no-till crops, probably from mineralisation of previous crop residue.

Figure 3. Plant nitrogen and phosphorus concentration at GS30 (a) and GS65 (b) for cereal crops on focus paddocks. The relationships for luxurious and N limited crops from Ziadi et al. (2008) are included for reference.

Figure 4. Plant nitrogen and phosphorus uptake at GS65 for cereal crops on focus paddocks (legumes measured at cereal GS65). The curve is the N:P uptake relationship of high  yielding crops in field experiments, at maturity in Sadras (2006). Note that the 2008 No Till Vetch (Carron paddock) has had dry matter estimated – the sample was taken from a windrow.

Discussion

Location of phosphorus in the soil

The profile of phosphorus with depth indicates the importance of the surface soil layer for phosphorus nutrition. There were no consistent differences between till and no-till paddocks, although less soil mixing is often cited as a reason for requiring more fertiliser initially in no-till crops. In no-till crops there is less deliberate soil mixing but it occurs nonetheless. Sowing with a knife point continually buries surface soil with soil from a mix of depths around the knife. Ridges of thrown soil slump into surface depressions through the process of rain, and possibly in a few years the soil is near-completely mixed in no-till crops too.

It is interesting to consider the impact of row spacing and dry years. At fixed plant density, wider row spacing decreases the distance between each plant and a fertiliser granule, but increases the distance roots must traverse to fully exploit the topsoil. In soil where most of the phosphorus is in the soil, this may disadvantage wide row crops, but where most of the phosphorus is in fertiliser this may advantage wide row crops.

In crops that are largely dependent on rainfall, the topsoil is wetter for longer in wet (high phosphorus demand) years and access to topsoil phosphorus should be greater. There may be a mis-match between root access to topsoil and crop growth potential (phosphorus demand) when crops are growing on fallow moisture in dry years, and this is potentially indicated by the high N:P ratios in 2008 Till wheat crops at Carron and Minyip.

There is also some research that indicates that plants can hydraulically ‘lift’ subsoil moisture into dry topsoil to access fertiliser phosphorus if it becomes available (Valizadeh et al. 2003). This would obviously only happen if there was subsoil moisture, and may help to minimise the problem of crops growing on fallow moisture with dry surface soil. Plants also increase root growth in particular parts of the soil if phosphorus supply is low and phosphorus is supplied heterogeneously (Ma and Rengel 2008). This may be an advantage for uptake of other nutrients, for example zinc.

Soil tests to predict phosphorus requirement

In these paddocks, pre-sowing topsoil Colwell P and PBI often indicated that plants had sufficient P for 90% of maximum yield (ie. criteria of Moody 2007), yet analysis of nitrogen and phosphorus uptake suggested that plants were often phosphorus limited. This is possibly because much nitrogen is located below the topsoil (wetter for longer) can can be taken up in excess, but phosphorus in these dry years is more often supply limited. The difference between Colwell P and critical P (90%) was related to the excess of nitrogen uptake by cereal crops in these experiments particularly at GS30 (Figure 5a). {possibly better in relative terms}. Fertiliser P (not included in Colwell P or PBI measurements) may be important for explaining some of the variation in this relationship, but has not been collated completely, yet (Table 2). No till and till crops were on a similar relationship at GS30 (Figure 5a), but at GS65 till crops were more likely to have accumulated a surplus of nitrogen than no-till crops (Figure 5b).

Figure 5. Surplus nitrogen uptake (compared to N:P ratio at maximum yield of Sadras), related to surplus of Colwell P over critical P (calculated from PBI according to Moody), at GS30 (a) and GS65 (b). Lines of best fit relate to both Till and No Till paddocks in 2007 and 2008.

Timing of phosphorus demand

The results also highlight the relatively small amounts of phosphorus taken up by the crop, especially early-on, in relation to the amounts of mineral P implied by the Colwell P test, and amounts of P applied in fertiliser (Table 2). Given that phosphorus concentrations in seed are similar to or greater than phosphorus concentrations in plant tissue, the plant already has sufficient phosphorus to reach a biomass similar to the sowing rate until it becomes dependent on soil and/or fertiliser phosphorus for further growth.

At least one of the reasons for poor efficiency of phosphorus fertiliser use may be that at the time when phosphorus from fertiliser granules is most available (shortly after sowing), the plant has little capacity for uptake. In a wet year with early May sowing, cereal crops reach 100 kg/ha biomass (approximately when seed phosphorus is insufficient for further growth) about 25 days after sowing. In dry years this may happen 10 days later, and later still with later sowing. To take up 1 kg P, with concentration 0.4%, requires 250 kg/ha biomass, achieved about 35 days after sowing in a wet year and up to 60 days or more after sowing in a dry year.

This poses the question: “why provide all phosphorus fertiliser at sowing?”. It seems necessary only to provide sufficient to allow each plant access to fertiliser phosphorus while roots are growing throughout the topsoil sufficient to access bulk soil mineral phosphorus. After that, Colwell P/PBI measurements suggest that in these environments, this source of supply will be enough. The driver for higher phosphorus fertilisation at sowing appears to be distribution (making sure each plant has ready access to a granule) rather than quantity. The success of fluid fertilisers seems to be in providing an appropriate amount of phosphorus in the right place (under the seed row) that is sufficient to meet plant requirements until roots have grown enough to access bulk soil P. This leads to the early dry matter, but not final yield responses observed in fluid fertiliser experiments in Victoria.

Apart from a small, evenly distributed source of ‘starter’ phosphorus fertiliser, applying phosphorus post-emergent in wetter years (where deficiency may be a problem) seems to be a real possibility. Cereal plants are able to take up foliar phosphorus, at least until ear emergence (Alston 1979). The problem with post-emergent, soil applied phosphorus is that roots need to grow into it (rather than having it leach onto them, as is the case with nitrogen). In pot experiments wheat has been shown to have the ability to grow into and take up sources of phosphorus a number of weeks after sowing (Ma and Rengel 2008). This may be a slower process in dry soils.

Phosphorus dynamics

Measurements of Colwell P at sowing and harvest in these paddocks have shown changes in Colwell P far in excess of anything related to plant uptake (Figure 2a), indeed in some fallow paddocks (ie. without plants!). In some No Till paddocks, changes up to 20 mg P/kg were observed – equivalent to the entire soil Colwell P at Minyip (about 25 kg P/ha) – as a decrease between pre-sowing (late March) and post-harvest (December), and then again as an increase by pre-sowing the following year. It seems unlikely that these changes are related to chemical ‘fixation’ of soil phosphorus, which is usually regarded as a one-way process.

A plausible alternative is microbial immobilisation and then mineralisation of mineral phosphorus. There were reasonable rainfalls in November (before sampling), and again in December and January at most sites. There was also a reasonable correlation with changes in topsoil nitrate measured in the same samples (Figure 6a), but not when changes in nitrate and Colwell P were tracked from year to year (Figure 6b). As with Colwell P, topsoil nitrate changes between pre-sowing and harvest will include the net effects of fertiliser and plant uptake as well as mineralisation/immobilisation. Regression on changes in Colwell P estimates they were on average 0.72 of the changes in nitrate N (Figure 6a).

Differences in phosphorus mineralisation between till and no-till systems, including immobilisation of phosphorus by wheat residue in no-till systems, have been observed over 52 week periods in Alberta, Canada, but the amounts are small for grain crop residues (Lupwayi et al. 2007). Estimates of mineralisation rates for 4 t/ha of legume residue in warm, saturated conditions were in the order of 1 mg/kg Olsen P over about a fortnight (Kabba and Aulakh 2004).

Another possibility is that some of immobilisation/mineralisation could be an artefact of the oven-drying process for soil samples. Harvest samples in 2007 were as wet as pre-sowing samples, both considerably wetter than 2008 pre-sowing samples. The reduction in nitrogen and phosphorus between pre-sowing and harvest measurements in 2007 were both related to the moisture difference between 2007 harvest and 2008 pre-sowing samples (as an estimate of ‘available’ water at 2007 harvest, not shown), but subsequent increases were not well related.

Whether ‘real’, or an artefact of soil sample water content, the potential for such large changes in soil phosphorus and nitrate, possibly related to the time of sampling, need to be considered when managing crops in the area and interpreting soil sample data.

Figure 6. Change in topsoil Colwell P, compared to change in topsoil nitrate, within the season 2007 (pre-sowing to harvest, then harvest to pre-sowing 2008), and between pre-sowing measurements in subsequent seasons (2006-7, and 2007-8).

Conclusions

Soil measurements showed that most available phosphorus in southern Mallee – northern Wimmera paddocks was present in topsoil, and in most paddocks was sufficient that crop yield would not be limited by phosphorus. There were no differences related to whether the crop was in a no-till, or a till paddock.

In the relatively dry years when plant phosphorus uptake was measured, phosphorus was more limiting to crops than nitrogen, and till crops tended to take up more nitrogen than required to match phosphorus uptake. At GS30 the excess uptake of nitrogen across both till and no-till crops was related to deficiency indicated by Colwell P and the critical P level calculated from the PBI.

Plant phosphorus uptake to GS30 was relatively small compared to fertiliser and soil phosphorus supply.

Large changes between pre-sowing and harvest Colwell P measurements were related to nitrate nitrogen measurements, and may have been caused by mineralisation/immobilisation, or be an artefact of the soil sampling process. Whatever the cause, the possibility of large within-season changes in soil measurements that may not be reflected in soil nutrient status at sowing needs to be considered in crop management.

Acknowledgements

Thank you to the Grains Research and Development Corporation for funding the project (BWD00008), to the farmers who have hosted focus paddocks and assisted with data collection, and to BCG and Vic No Till staff involved in collecting, processing and collating data.

References

Alston AM (1979) Effects of soil water content and foliar fertilization with nitrogen and phosphorus in late season on the yield and composition of wheat. Australian Journal of Agricultural Research 30, 577-585.

Bolland MDA, Brennan RF, White PF (2006) Comparing responses to phosphorus of field pea (Pisum sativum), canola (rape, Brassica napus) and spring wheat (Triticum aestivum). Australian Journal of Experimental Agriculture 46, 645-657.

Bolland MDA, Siddique KHM, Loss SP, Baker MJ (1999) Comparing responses of grain legumes, wheat and canola to applications of superphosphate. Nutrient Cycling in Agroecosystems 53, 157-175.

Calderini DF, Torres-Leon S, Slafer GA (1995) Consequences of Wheat Breeding on Nitrogen and Phosphorus Yield, Grain Nitrogen and Phosphorus Concentration and Associated Traits. Ann Bot 76, 315-322.

Kabba BS, Aulakh MS (2004) Climatic conditions and crop-residue quality differentially affect N, P, and S mineralization in soils with contrasting P status. Journal of Plant Nutrition and Soil Science-Zeitschrift Fur Pflanzenernahrung Und Bodenkunde 167, 596-601.

Kim JC, Mullan BP, Selle PH, Pluske JR (2002) Levels of total phosphorus, phytate-phosphorus, and phytase activity in three varieties of Western Australian wheats in response to growing region, growing season, and storage. Australian Journal of Agricultural Research 53, 1361-1366.

Lupwayi NZ, Clayton GW, O'Donovan JT, Harker KN, Turkington TK, Soon YK (2007) Phosphorus release during decomposition of crop residues under conventional and zero tillage. Soil & Tillage Research 95, 231-239.

Ma Q, Rengel Z (2008) Phosphorus acquisition and wheat growth are influenced by shoot phosphorus status and soil phosphorus distribution in a split-root system. Journal of Plant Nutrition and Soil Science-Zeitschrift Fur Pflanzenernahrung Und Bodenkunde 171, 266-271.

McDonald GK (2006) Effects of soil properties on variation in growth, grain yield and nutrient concentration of wheat and barley. In 'Australian Journal of Experimental Agriculture' pp. 93-105

Moody PW (2007) Interpretation of a single-point P buffering index for adjusting critical levels of the Colwell soil P test. Australian Journal of Soil Research 45, 55-62.

Sadras VO (2006) The N:P stoichiometry of cereal, grain legume and oilseed crops. Field Crops Research 95, 13-29.

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 depths and watering regimes. Australian Journal of Agricultural Research 54, 59-65.

Ziadi N, Belanger G, Cambouris AN, Tremblay N, Nolin MC, Claessens A (2008) Relationship between Phosphorus and Nitrogen Concentrations in Spring Wheat. Agron J 100, 80-86.

Zubaidi A, McDonald GK, Hollamby GJ (1999) Nutrient uptake and distribution by bread and durum wheat under drought conditions in South Australia. Australian Journal of Experimental Agriculture 39, 721-732.