Summary

In this paper attention is drawn to the need for sulfur in some very important processes in crop production and the increasing development and/or discovery of S deficiencies.The minimum requirement for a sustainable system is a balance between the inputs and losses of S and within this boundary crop and animal production is controlled by the storage and recycling of S, and other nutrients. As well as nutrient balances for the system there is a need to have a balanced supply of all nutrients if plants are to use them efficiently. A simple diagram of the inputs to and losses from a cropping system, and the S pools and recycling pathways within it, are used to highlight key processes that control crop production. Examples of some crop residues managements show how important good management is in reducing S losses A diagram of a carbon cycle is used to show the interactions between energy, recycling of nutrients and balanced supplies of nutrients. Methods for the detection of S deficiencies are considered, and the problems with the development of S soil tests are discussed. Radiotracer studies using 35S have shown turnover rates ranging from hours for the small available S components to years for the large resistant organic S pool and to show recycling and/or losses of S from residues, the efficiency of fertiliser uptake by plants in relation to fertiliser placement and nutrient interactions. Tracers were also important in evaluating a range of soil S tests and the development of the 0.25M KCl at 40° C soil test by researchers at the University of New England. The agronomic effectiveness of S containing fertilisers is related to their water solubility and granular or particle sizes. There is a large range of pasture and crop fertilisers available with varying proportions of sulfate and elemental S (ES). The fineness of the ES is the major determinant of its oxidation rate and this can be easily manipulated in the formulation of ES containing fertilisers. Techniques for coating commercial fertilisers with ES, and computer simulation studies of S release and uptake show the potential for the development of special fertilisers for specific crops. Such fertilisers have the potential to reduce leaching losses and enhance the ability of plants to utilise S by matching release rates to plant demand and providing a balanced supply of nutrients.

Introduction

Sulfur is one of the essential elements required for the normal growth of the plants and a deficiency of S will cause basic metabolic impairment which will not only reduce the crop yield but also the quality of produce. Plant concentrations of S are about the same as P and much lower than that of N but S is incorporated directly into a variety of plant products and is essential for many of the plant production processes such as the synthesis of:-

• proteins and essential S containing amino acids,

• coenzyme A as well as biotin, thiamine and glutathione;

• chlorophyll, volatile oils in Brassicas, and

• the fixation of nitrogen by leguminous plants.

Figure 1. Within an overall S balance the production is determined by recycling processes.

  

Spencer (1975) classified crops into three groups according to their sulfur requirement:

• crops with high S requirement (rapeseed, lucerne and cruciferous forages);

• moderate S requirement (coconut, sugarcane, clover and grasses, coffee and cotton);

• low S requirement (sugar beet, cereal forages, cereal grains and peanut).

For many years, little attention was paid to sulfur as a plant nutrient mainly because it has been applied to soil in incidental inputs in rainfall and volcanic emissions and as a component of some fertilisers. In some early studies, to correct what was perceived as specific nutrient deficiencies, using fertilisers such as (NH₄)₂SO₄, K₂SO₄, ZnSO₄ and Superphosphate, the responses attributed to the N, K, Zn and P may have been partially due to S or its interactions with other nutrients. Any such unobserved interactions emphasise the need to accurately detect deficiencies and provide plants with a balanced supply of nutrients. S deficiency differs from N deficiency symptoms in that the yellowing of the leaves occurs on the new growth.

Intensification of cropping systems using high-yielding varieties has accelerated S removal from the soil, which will ultimately result in more soils becoming S-deficient. Increased use of high-analysis S-free fertilisers has aggravated the S deficiency problem in lowland rice and in other cropping systems. As a consequence of these changes, the efficient use of S and other nutrients in cropping systems must be given greater consideration.

If the overall production system is to be sustainable there must at least be a balance between the total S input and output/loss. The losses can be reduced by better management of the products and their residues and any deficit made up by inputs of appropriate fertiliser materials.

Given a situation where the removal of S in products is approximately in balance with S supply the most critical parts of the production system are the storage pools and recycling processes in the soil, and the management of the residues which plays a significant part in reducing the loss of S from the cycle.

If fertilisers are to be used to maintain a balance in the supply of S to the crop there is a need to know the efficiency with which the fertiliser S reaches the plant, especially as this also depends on the competition between the recycling processes in the soil. However, there is a large range of sizes of the S pools in the cycle relative to the amounts of S added in fertilisers and we need to know where the S goes and how fast it gets there. This is not possible with the usual techniques and an example of the problem is that as soon as S from the fertiliser enters the cycle it is indistinguishable from that already there. The use of stable and radioactive isotopes of S has greatly assisted in understanding the S cycle.

Radiotracer studies of sulfur cycling using ³⁵S fertiliser applied to temperate pastures (May et al. 1968, Till and May 1970 a,b) showed that S from the fertiliser could still be detected two years after the initial application. This showed a long residual effectiveness and emphasized the role of organic matter as the major temporary storage pool for added fertiliser. It also provided a basis for a simple model, based on a cycle similar to that in Figure 2, and simulation studies which showed the importance of process rates within the cycle ( May et al. 1973)

In this paper we describe what we consider as some of the key S containing components and processes in crop production and how they relate to S requirements, especially fertilisers and crop residue management in a No-Till system.

The sulfur cycle; inputs, losses, pools and processes

The sulfur cycling literature has been extensively reviewed and a simplified diagram of a S cycle is shown in Figure 2. The figure shows the major inputs and losses that control the overall nutrient balance.

Figure 2. Simple representation of the inputs, recycling and losses of S for a cropping system.

If the system is in balance the productivity of the system is controlled by the pool sizes and the rates of transfer of S between them. The cycle shown is essentially that used in the simulation studies of May et al. (1973) but slightly modified to suit a cropping system. The boxes represent pools of sulfur and their definitions and brief listing of important properties are listed below, the arrows are the flows between the various fractions. For a better understanding of S recycling it is necessary to recognize the wide range of pool sizes that we are dealing with and an example, from ³⁵S studies of a grazed pasture, is shown in Figure 3. A more detailed description of the various pools and flows follows.

Figure 3. Relative sizes of S pools in a grazed pasture in NSW Australia

  

The removal of the domestic animal pool from Figure 3. makes very little difference to the relative sizes of the remaining pools but does remove some significant alternative recycling pathways. The sizes of the S pools show that the available pool is too small to satisfy the crops needs without replenishment. If there is not a sufficient S input to the available pool from recycling within the system and the natural inputs (Figure 2.) then the deficit must be made up by some other process such as fertiliser application. Inputs of readily soluble S fertiliser will significantly boost the size of the available pool for a short time so it is particularly important to give plants the best opportunity to compete for fertiliser nutrients.

The large OM pools dominate the system and changes in management and nutrient inputs may take a long time to reach a stable state.

Consideration of Figure 2 and Figure 3 emphasises two very important points

• In almost all circumstances any S input to the system that can provide S to the plant does so via the available pool.

• The plant must compete with other processes for the available S.

In some cases reasonable estimates can be made of the S balance of a system but to be able to make reliable estimates of the amounts of fertiliser needed requires an intimate local knowledge and/or an understanding of the recycling processes. Because of the range of sizes of the S pools in the cycle relative to the amount of S added in fertilisers, rainfall, and crop residues there is a need to know where the S in any particular pool goes and how fast it gets there. The use of stable and radioactive isotopes of S has greatly assisted in understanding the S cycle.

The S Pools

Available Sulfur

• A pool of sulfate that can be drawn upon by the plant.

• Fertiliser S must be liberated as sulfate and enter via this pool to be available to plants

• Sulfate is also supplied by mineralisation from other pools

• Note that the plant has to compete with other processes for the available sulfate

Labile Organic and Biota Sulfur

• A pool of organic S that turns over fairly rapidly (eg time for half to disappear 35 days) utilising and releasing sulfate. Because of its large size and rapid turnover the management of the labile OM pool is critical in the flow of sulfate into the available pool.

• In the cropping system a significant portion of the crop residues and soil biota are included in this pool.

Resistant Organic Sulfur

• The resistant organic pool is even larger than the labile pool and together they will usually contain about 60-90% of the total soil sulfur.

• The input to this relatively slowly cycling pool will come from the more resistant carbon-bonded-S fractions in plant residues and the soil biota.

Inorganic Sulfur

• Small amounts of adsorbed S,

• Leached sulfate that may accumulate as gypsum at depth,

• Reduced S in anaerobic, flooded soils,

• Various other mineral deposits that release S very slowly.

Plant Sulfur

• Plants must compete with other processes for available S.

• Efficiency of using fertiliser depends on placement, release rates of nutrients, timing of application, and interactions between nutrients.

• Products (grain, livestock) represent a real loss of S from the system.

Sulfur in Plant Residues

• The type of crop and method of harvesting are important in determining how much residue is left in the field.

• Management of residues left in the field influences the soil condition and amount of nutrient recycled or lost.

• Management of residues removed with product will also influence nutrient losses.

Crop Production, System Processes and Management

Various environmental inputs and losses may make significant differences to the sulfur balance of the whole system. Irrigation water can be a major source of S since it usually includes S accessions from rainfall over a large area, which may become concentrated through evaporation, enriched by S inputs from volcanic springs and fertilisers.

Leaching losses of S can be a significant problem in non-flooded soils with coarse texture and low S sorption capacity.

Tropical acid soils, with their high quantities of 1:1 type clays and Fe and Al hydrous oxides, can adsorb considerable amounts of sulfate. Typically the amount of adsorbed sulfate increases with depth within these soil profiles.

Plants require SO₄^2-_, and take it, and other nutrients, from the available pool. The S cycle has similarities to both N and P cycles. The role of organic sources in supplying sulfate to plants is similar to both the N and P cycles and the adsorption reactions are similar to P reactions although the strength of sulfate adsorption is considerably less than for phosphate.

In terms of maintaining crop production any discrepancies in the S balance of the system are usually made up by the application of some form of fertiliser.

The major storage and supply of S in soil relies on organic S so the cycling of S in soil is closely related to organic matter turnover and the S mineralisation and immobilization processes play a key role in determining the availability of S for plant growth (Figure 4). Consequently it is now appropriate to introduce the carbon cycle and the release of other nutrients.

Figure 4. The carbon cycle and nutrient recycling

For optimum production plants need a balanced supply of nutrients. The cycle shows the release of N, P and S but because of different uptake and loss processes in the systems components they are unlikely to be available to the crop in optimum proportions. There are many factors involved in the balance and just two are given here as examples.

• Sulfate is only weakly adsorbed in most soils and may be displaced by applied P.

• Nitrate and soluble sulfate are very mobile and are susceptible to leaching.

Typical results of imbalance are

• If S is deficient N cannot be used efficiently for protein production, and vice versa.

• Legumes cannot fix N efficiently if there is not a balanced supply of S and P.

• If there is not an overall balance between input and removal/loss of any nutrient the system is not sustainable. An example of the benefits of a balanced nutrient supply can be seen in Figure 5. which shows the $profit of balanced S and P applied to corn production in the Santa Fe region of Argentina.

Figure 5. Profit from balanced S and P nutrition of corn in Santa Fe, Argentina

Estimating the S balance of a production system on a local or regional level can prove helpful in predicting S requirements.and to its long term stability. If it is negative the system cannot be sustained.

Atmospheric inputs of S in rain, dryfall and gaseous forms can contribute significantly to plant S supply and, together with other inputs such as irrigation and losses in crops and by leaching, will allow better calculation of the S balance and fertiliser needs. In many countries surveys of S accessions in rainfall have shown large differences both the amount and seasonal distribution. For example in Australia the highest input was 20.3 kg/ha/yr on the W. Coast of Tasmania and the lowest was 0.41 kg/ha/yr in central Tasmania, while in Malaysia mapping of the inputs (Figure 6) showed areas where a change in crop from rubber to oil palm would result in moving from a positive to a negative S balance. Locations near the sea and industrial centres often receive sufficient S in rainfall to obviate the need for fertiliser S

It is important to know what the environmental inputs are, especially if cropping systems are to be changed.

Figure 6. Contour map of S inputs in rainfall.

The management of crop residues can also have a significant influence on S fertiliser requirement. Studies in Thailand (Table 1) show that when no fertiliser S was applied removal of crop residues turned the system from a positive S balance into a deficit. The calculations in table 1 are probably optimistic as they assume all S in rain and fertiliser is available to the crop.

Table 1. Sulfur input/output balance sheet for rice production in Thailand.

As well as the removal or return of crop residues the way that residues are processed can also have a significant effect on the S balance. The burning of residues will result in significant losses of nutrients (Figure 7, Lefroy et al. 1994), which are not the same for all nutrients, so it may upset the balanced supply of nutrients. It will also change the availability of the remaining nutrients.

Figure 7. Summary diagram showing the fate of S applied in 35S labelled straw (0.108% S) or ash derived from that straw (0.374% S).

Even if the nutrient balance is positive the system can still be in a decline if changes in the various processes allow the relative sizes of some pools to change. For example a cropping system could be in an overall S balance but the management might cause the relative amounts of labile and resistant organic matter change. A build up in resistant OM would send the system into a decline due to a fall off in the supply to the available pool, while an increase in labile OM could allow production to rise.

There is a considerable amount of energy in crop residues and they will play an important part in supplying sufficient energy and nutrients for the biota to keep the recycling processes going.

Removal of residues will constitute a loss of nutrients and energy for the soil biota recycling system. The burning of residues and return of ash will cause significant loss of nutrients and energy but any remaining S will be in a plant available form.

Fertiliser required for a particular level of crop production depends on the efficiency with which the plant competes for the nutrients in the cycle, and the overall balance between all the other rates of input and loss.

The information presented so far shows that to know how much fertiliser might be needed for efficient production of a particular crop requires the integration of a considerable amount of information. It is obvious that a S balance must be maintained, good management of crop residues is essential and the efficiency with which plants compete for available S needs to be known

Nutrient models 

One approach to easing the problems just outlined is to develop a dynamic model of the system. The model should be as simple as possible, using whatever information is available, and additional components only added when they are shown to be critical.

Unfortunately a nutrient model will only show the likely outcome of changes in inputs and the recycling rates and/or pool sizes. There also needs to be an understanding of the management practices that bring about the changes in the system.

No-Till agriculture

Some positive benefits of no till agriculture can be the improved ground cover that gives better control of surface water flows which then carry lower amounts of soil, plant particulate matter and nutrients. This reduces erosion and lowers stream sediments and pollution with excess amounts of nutrients.

Reduced cultivation does less damage to the soil structure and, especially in row crops, less compaction and better root development. Water infiltration and retention is increased due to better soil structure.

As for conventional cropping the type of crop and method of harvesting is important in determining how much residue is left in the field. The amount and management of these residues in no-till is especially important because the intact nature of the material, and less intimate contact with the soil biota, can lead to slower physical breakdown and release of nutrients. This is not necessarily a disadvantage as it might be that the nutrient release rate is more closely related to the varying rate of demand by the plant.

The management of residues removed in harvesting and recovered during separation from the crop is just as important as in conventional tillage even though it may be more difficult to manage. Changes in the breakdown rate of residues and the timing of weed control, fertiliser application and sowing may mean that fertilisers with different nutrient release characteristics are needed for conventional and no-till operations. There may also be a need for a shift in the relative proportions of the components in the recycling system if production is to be maintained.

Sulfur deficiencies and their detection

There is an increasing awareness of the need for S as a plant nutrient, and the recognition of S deficiencies is increasing, partly due to more studies of S and also through a number of direct and indirect management practices. The main reasons for this are:-

• Increasing use of ‘high analysis’ fertilisers that contain little or no S.

• Higher yields obtained as a result of other technological advances.

• Decreasing use of S containing pesticides and fungicides.

• Environmental control of sulfur dioxide emissions from industrial processes.

• More experiments where S is being studied as a nutrient in its own right.

• Awareness of the need for a balanced supply of nutrients if plants are to use them efficiently.

The detection of S deficiency in plants is fairly straightforward and symptoms include a yellowing of the younger leaves (Figure 8) as a result of a low chlorophyll production and S non-mobility and a marked reduction in plant height and tiller number in cereals.

Visual symptoms and plant tissue testing are acceptably good indicators of an S deficiency but by the time that they become obvious it is frequently too late to apply corrective measures to the crop.

In the past soil tests have been unreliable because they only extracted the available soil S pool which is usually very small compared with total crop uptake. The major problems being that the methods used did not estimate the potential rate of recharge of the available pool and could not unequivocally relate the S extracted to the fraction that was actually used by the plants. The difficulty was that without isotopic tracers there is no way of measuring whether the extractant recovered S from the same source as that used by the plant. Using this technique the UNE group developed the KCl-40 test where 0.25M KCl is added to soil and extracted at 40oC for 3 hours (Blair et al. 1991). This extractant was found to be the best indicator of S status among 5 extractants evaluated (Table 2).

 

Figure 8. S deficiency in corn

Researchers at UNE used 35S isotope dilution techniques to study the relationship between the S taken up by pasture plants and the S extracted by various reagents (Table 3).

Table 2. Coefficient of determination (r2) between extractable S and percent of maximum yield for a range of extractants on a range of 18 pasture soils collected from Northern New South Wales, Australia.

Table 3. The degree of association between the extracted S and that taken up by the plant. (The nearer to 1 the closer the S used by the plant is to that extracted).

The variation of S distribution down the profile (Figure 9) must also be taken into account in relation to the particular crop and the likelihood of S deficiency at various stages of growth, and estimated fertiliser amount and SO42- release rate.

Figure 9. S distributions down the profile for soil from some different sites

Fertiliser materials and application

There are a wide range of S containing fertilisers available in Eastern Australia (Table 4) The major forms of S in fertilisers are SO42- and ES, and as SO42- is the form required by plants its availability is only limited by its rate of solution. However, SO42-, like NO3-, is highly mobile and is vulnerable to leaching.

Table 4. Sulfur containing fertilisers available in Eastern Australia and New Zealand.

ES is not soluble so is not leached but it has to be oxidized to SO42- before it can be used by plants. In the soil ES oxidation is performed by various microorganisms and this can be exploited as an additional advantage over SO42-. The ES needs to be colonized by S oxidizing microorganisms so the rate of oxidation is partially controlled by the surface area of the ES as well as by environmental variables which are the same those controlling plant growth. The lack of leaching, and the ability to manipulate the rate of sulfate release by using ES particles of different sizes, makes ES a very versatile fertiliser material as it gives us the potential to match release rate of SO42- to plant demand at different stages of growth.

ES is obviously a very ‘high analysis’ fertiliser so is ideal for incorporating into other fertilisers in the granulation stage or being coated onto commonly available N and P fertilisers.

Some fertilisers use a mixture of sulfate and ES to get an initial high level of available S but they run the risk of significant leaching losses if rains are heavy before the plant is sufficiently developed to take up enough of the SO42-.

Other essentially pure ES fertiliser preparations are available aimed at improving the safety of handling the fine material required for efficient oxidation while still providing a large surface area.

Methods such as foaming and/or aggregating with other materials are used which potentially allow separation of the fine particles on wetting after application. These materials are frequently poor performers in the field due to problems of wide separation of ‘granules’ at normal S application rates and lack of the beneficial association with other nutrients, especially as many S oxidizing organisms require a high supply of P. Consequently ES coated materials such as UNE-COAT are more suited to drilled and broadcast applications than products where pure ES is formed into a granule, such as sulfur bentonite (Degrasul, Agricsul, Sulchem etc.), and subsequently mixed with a P source (Table 5 and Table 6).

Table 5. Distribution of S fertiliser when applied at 10kg S/ha

Table 6. Effect of S application rate and mixed or separate band application of TSP and ES (<0.250 mm particles) on yield and fertiliser S uptake by maize in pots (Friesen, 1987).

The findings of Friesen et al. (1987) that S oxidation rate is enhanced when P and ES are mixed together has important implications for the development of fertilisers. Friesen et al. (1987) reported that, on average, maize yields were 29% greater and S fertiliser recovery 124% greater from mixed P/S bands than when applied separately.

Santoso et al. (1995) not only found enhanced fertiliser S uptake by corn when P and S was banded together but also lower leaching losses of S (Table 7). In addition, Santoso et al.(1995) found that fertiliser S uptake was increased and leaching losses decreased when the P application rate was increased from 10 to 30 m g P/g soil. This clearly demonstrates that balanced fertilisation is necessary in order to maximize fertiliser efficiency.

More recent research at UNE has shown that the enhanced oxidation rate of S in the presence of P is due to an enhancement in the number of Thiobacillus thiooxidans present and that this nutrient effect is restricted to P (Soleh et al. 1997). Other research has demonstrated the effect of temperature on S oxidation rate (Table 8).

Table 7. Percentage of applied fertiliser taken up by 30 day old corn and the loss of fertiliser S (%) from the pots (Santoso et al. 1995).

Table 8. Oxidation rate of 0.4 mm diameter ES as affected by temperature (Shedley et al. 1979).

The data presented above suggests that it should be possible to use a simulation model to adjust the supply rate of S from fertilisers to match plant demand and hence maximise fertiliser use efficiency.

Matching fertiliser S release rate to Plant demand

A simple model (McCaskill and Blair 1989) has been used to simulate the release of sulfate from various fertilizers (Figure 10) and examples are shown for matching the release rate from ES mixtures in both temperate (Figure 11) and tropical (Figure 12) environments.

Figure 10. Release pattern predicted for various S fertilisers applied to a grazing trial at Armidale, NSW. SSP: Single superphosphate; SF45: sulfur-fortified superphosphate; ES agricultural grade crushed ES(total), and size fractions within it if <0.5, 0.5-1.0 and > 1.0mm diameter.

Figure 11. Plant S demand and calculated S released from ES at Armidale, Australia.

Figure 12. Plant S demand and calculated S released from ES at Darwin, Australia

Conclusion

In this paper we have attempted to raise the awareness of the need for S and a balanced supply of nutrients to the plants. For the production system the minimum requirement is an overall nutrient balance and minimisation of losses by management of crop residues and the use of appropriate fertilisers for the particular crop. The majority of the S is stored in various organic forms and the dynamics of the mineralisation and fixation processes control the amount of nutrients available for plant uptake. However the efficiency with which plants compete for the available S also depends on the management practices and interactions with other nutrients. We have outlined an approach to soil testing and the potential for the development of specific fertilisers that can enhance plant competition for nutrients by matching their changing demand during growth to the fertilisers rate of S release.

References

• Blair, G.J., Chinoim, N. Lefroy, R.D.B., Anderson G.C. and G.J. Crocker (1991). A sulfur soil test for pastures and crops. Australian Journal of Soil Research 29:619-626.

• Friesen, D.K., Sale P.W.G. and Blair G.J. (1987). Long-term greenhouse evaluation of partially acidulated phosphate rock fertilisers. II. Effect of co-granulation with ES on availability of P from two phosphate rocks. Fertiliser Research 13:45-54.

• Lefroy, R. D. B, Chaitep, W. and Blair, G.J. (1994). Release of sulfur from rice residues under flooded and non-flooded soil conditions. Australian Journal of Agricultural Research, 45, 657-667.

• May, P.F., Till, A.R., and Cumming, M.J. (1973). Systems analysis of the effect of application methods on the entry of sulphur into pastures grazed by sheep. Journal of Applied Ecology 10, 607-627.

• May, P.F., Till, A.R., and Downes, A.M. (1968). Nutrient cycling in grazed pastures. 1. A preliminary investigation of the use of 35S gypsum. Australian Journal of Agricultural Research 19, 531-543

• McCaskill, M.R., and Blair, G.J. (1989). A model for the release of sulfur from elemental S and superphosphate. Fertilizer Research. 19, 77-84.

• Santoso, D, Lefroy, R. D. B. and Blair, G.J. (1995). Sulfur and phosphorus dynamics in an acid soil/crop system. Australian Journal of Soil Research, 33, 113-24.

• Shedley, C.D., Till, A.R. and Blair G.J. (1979). A radiotracer technique for studying the nutrient release from different fertiliser materials and its uptake by plants. Communications in Soil Science and Plant Analysis 10:737-745.

• Sholeh, R.D.B. Lefroy R.D.B. and Blair G.J. (1997). Effect of nutrients and elemental sulfur particle size on elemental sulfur oxidation and the growth of Thiobacillus thiooxidans. Australian Journal of Agricultural Research 48:497-501.

• Spencer, K. (1975). Sulphur requirements of plants. In ‘Sulphur in Australian Agriculture.’ (Ed K. D. McLachlan) pp. 98-108 ( Sydney University Press, Sydney, Australia.)

• Till, A.R., and May, P.F., (1970a). Nutrient cycling in grazed pastures. 2. Further studies with (35S) gypsum. Australian Journal of Agricultural Research 21,253-260.

• Till, A.R., and May, P.F., (1970b). Nutrient cycling in grazed pastures. 3. Studies on labeling of grazed pasture systems by solid (35S) gypsum and aqueous Mg35SO4.. Australian Journal of Agricultural Research 21, 455-463.