This article aims to present an overview of how knowledge of and attitudes to phosphate fertiliser efficiency have developed within New Zealand, to clarify areas of confusion that exist, to discuss possibilities for improving phosphate efficiency in the future, and for reducing losses to the aquatic environment. There is far more detailed information on a range of sub-topics to be found elsewhere, but no recent attempt has been made to present an overview that addresses major areas of disagreement and uncertainty regarding phosphate fertiliser.
Origins of recommendations for phosphorus (P) use in New Zealand
The importance of phosphorus in pastoral agriculture is totally accepted in New Zealand. This is because the development of introduced ryegrass and clover pasture, which quickly came to be the backbone of the New Zealand economy, could not have happened on the previously acidic and nutrient-poor soils without lime, phosphorus (P) and sulphur (S). Clover establishment in particular, even with inoculation with Rhizobia, was very difficult without these vital inputs.
A large number of field trials conducted the Department and then Ministry of Agriculture (MAF) throughout the country enabled P requirements for establishment and maintenance to be fairly well understood by the 1950s. Summaries were made by Peter During in his excellent book ‘Fertilisers and Soils in New Zealand Farming’ in the 1960s, and updated and refined by MAF’s Ian Cornforth and Allan Sinclair into the nutrient balance-sheet driven Computerized Fertiliser Recommendation Scheme (CFAS) of 1972. This in turn has been succeeded by the dynamic-system driven ‘Overseer’, developed by AgResearch scientists, particularly Alister Metherell and Dave Wheeler.
Natural sources of phosphorus
The presence of P is inextricably interwoven with that of calcium (Ca) in nature. Ca is continually being washed from the land into the sea in rivers and precipitating with soluble carbonate on the ocean floor. The sedimentary phosphate ‘rocks’, many of which are now above ground and mined for their P content, were formed in places where the presence of unusually high concentrations of phosphate in sea water lead to replacement of carbonate in limestone over time to make tricalcium phosphate, or ‘hard’ phosphate ‘rock’, a far more stable compound, which has to be treated with strong acids to make phosphate fertilisers. ( I have put the word rock in inverted commas as it is an unfortunately misleading misnomer – most phosphate deposits actually exist as sands, and only become more ‘rock-like’ if they have become compressed by the weight of other material deposited over them over huge periods of time). The strength and durability of tricalcium phosphate is why it is nature’s building block in skeletons.
Where the conversion of limestone to tricalcium phosphate on the sea floor was not complete before earth movements or dropping sea levels left it above sea level, the crystal lattice contains both carbonate and phosphate bonded to Ca, and we have what are called ‘reactive phosphate rocks’ or RPRs, which can be used directly as a fertiliser on acid soils. New Zealand’s own underwater Chatham Rise phosphorite deposit is an RPR in the making.
Many other types of calcium phosphate compounds are formed by igneous or volcanic acivity. Many of these incorporate iron and aluminium as well, which reduces their usability for fertiliser, directly or through manufacturing. Many of these compounds have been changed further by weathering and deposition of bird or bat excreta, which in isolated island or cave conditions can result in the natural fertiliser known as guano.
Manufactured phosphate fertilisers
The earliest recorded use of ground limestone to improve ‘sour’ soil is by the Romans. Gypsum has been used to improve infiltration and drainage for a long time as well. Both ground bones and calcium phosphate minerals were widely used as P fertilisers in the 1800s. Then it became known that treating either with sulphuric acid to make what Lawes in England called ‘superphosphate’ was far more effective in most cases. The sulphuric acid converted the tricalcium phosphate to water soluble (and hence immediately, if only temorarally) available to plants. The Ca that was liberated by this forms gypsum with the sulphate from the sulphuric acid. “Super’ made from pure tricalcium phosphate is by weight 45% monocalcium phosphate and 55% gypsum. The key point is that all the Ca from the phosphate rock remains in the superphosphate. “Super’ became by far the most widely used fertiliser in New Zealand.
With the development in the USA in the 1920s of technology to separate the intermediary phosphoric acid from the gypsum during the superphosphate manufacturing process, everything changed, in most countries. First, triple superphosphate (20.5% P but only 9% Ca and 1% S) could be made, and phosphoric acid, containing no Ca, could be reacted with ammonia gas to make MAP and DAP. The latter has become increasingly widely used in NZ in the last 10-15 years. We will come back to this.
It is important to note that all phosphate rocks have some liming effect. This is due to the fact that soil acidity slowly – albeit much more slowly than in a factory – dissolves the tricalcium phosphate, and acidity is consumed in the process. Liming effects of phosphate rocks range from about 25% for apatite and other hard phosphate rocks, to over 50% for RPRs, which contain destabilising carbonate in the lattice.
Natural P in New Zealand soils
European settlers discovered we had a very good climate for growing pasture, but we needed to reduce the acidity, and put on lots of P. Surprisingly, our soil Ca levels themselves were reasonably good, given the climate. This is because the parent sedimentary material was high in the tricalcium phosphate (often called apatite) we talked about. However, the rate of dissolution of this phosphate into plant-available form, at about 1-2 kg P/ha annually, was far too slow to meet the demands of high-producing introduced pastures.
Most New Zealand soils have quite a high cation exchange capacity (CEC), so the ground lime applied to increase the soil pH meant that its Ca content was not leached too quickly. The use of ‘single’ superphosphate or ‘super’ as the main fertiliser to supply P and S also helped maintain soil Ca levels, although the gypsum component is present as a very soluble form from which the sulphate is easily leached from most soils, taking cations with it.
What factors drive the need to continue to apply P?
No-one disputes the vital importance P plays in all living organisms. All farm produce taken from the farm contains the P used to produce it. This is generally the largest or second-largest type of P removal or loss that must be replenished, the other being fixation in the soil. Other smaller losses include direct deposition of excreta into waterways, surface run-off into waterways, leaching through the soil, deposition as excreta on unproductive areas of the farm, and uneven redistribution of excreta, for example heavy deposition on camp sites. The latter mode of loss in particular can be exacerbated by the use of soluble P fertilisers, if they result in short-term excessive plant uptake and animal intake.
The original native vegetation evolved to survive on very low P inputs from slow weathering of P minerals present in the parent material, plus inputs in rainfall, minus losses in leaching and run-off. This system was necessarily very ‘tight’, with very little net loss.
With the introduction of high-producing ryegrass-clover pastures and various crops, removals increased dramatically. After a capital application phase to get concentrations of P and other nutrients in the soil up to levels where the growth of these plants are not unduly restricted, ongoing losses from the system have to be replaced with ‘maintenance’ applications, typically annually (sheep and beef farms) or twice a year on dairy farms.
As both real the cost of fertiliser and the adverse effects of losses of nutrients to the environment increase, increasing attention is being paid to finding out where P is being lost, and how to reduce the losses, ie, tighten the cycle. Lets look at how these losses occur and what could – or already can – be done about them.
Animal-induced losses of P from the production cycle
Direct deposition of excreta into waterways
This source of P loss represents an economic loss to the farmer, and is a substantial contributor to environmental pollution in the form of algal and weed growth in waterways and lakes. The campaign by the dairy industry in recent years to fence off drains and ditches to stock has substantially reduced this mode of P loss, and much remains to be done.
Direct entry of dairy shed effluent into waterways
Likewise, industry action and policing by regional councils has lead to more efforts being made to ensure effluent is irrigated onto land efficiently, and reducing the risk of accidental splillage. In addition, the nutrient content of effluent is incorporated into ‘Overseer’ fertiliser nutrient recommendations for the treated paddocks. As a generalisation, effluent needs to be spread over a much greater percentage of the farm than the current 5-15% typical range, as this leads to accumulation of excess soil K levels on the ‘effluent block’, and increases the risk of pugging and nutrient run-off.
Uneven deposition of excreta
While not achieving the uniformity of effluent irrigation, strip-grazing on dairy farms achieves quite high evenness of distribution. This is not possible on hill country farms, and the flatter camping sites receive disproportionately large amounts of excreta. Increasing subdivision reduces this, as does using sustained-release P sources such as RPR. This maintains a more consistent level of P in the herbage, whereas luxury uptake in the first few weeks after application of soluble P fertilisers can result in very elevated herbage P levels, much of which will end up being excreted in camp sites.
Dairy cows can however also excrete significant quantities in small unproductive areas, for example while waiting at gates to walk to the milking shed, and on the walking lanes to and from milking. Ideally, these lanes should have shallow culverts to collect runoff during wet weather.
Pugging of heavily-grazed areas either in wet weather or under irrigation can greatly reduce soil infiltration and drainage rates, resulting in far higher losses of nutrients, especially P, in runoff. The use of standing pads between grazing, from which effluent can be collected and irrigated onto land, markedly reduces such losses.
‘Fixation’ losses of P in the soil
While all the above factors are important, it is in reducing losses (of N as well as P ) into and from the soil is where the potential to really clean up New Zealand’s waterways and lakes, and reduce fertiliser costs, really lies.
‘Fixation’ losses of P in the soil relate to the accumulation of P in the soil that is too tightly bound to soil clays and/or organic matter to be available for uptake by plant roots, or has been precipitated as insoluble P compounds, or is incorporated in very stable (resistant to microbial breakdown) soil organic matter.
The worst ‘fixation’ offenders in New Zealand by far are the volcanic ash soils of the North Island and Southland, which contain a very highly P fixing amorphous aluminium silicate clay called allophane. These soils require typically half as much again to double the amount of P to maintain a given level of production as non-allophanic soils – simple but undeniable proof that 30-50% of the P applied to these soils is being effectively ‘lost’ in terms of availability to the plant. As it happens, because they mostly occur in areas with reliable rainfall, have good drainage characteristics and are capable of sustaining high levels of production if topdressed with appropriate amounts of nutrients, they tend to be largely devoted to intensive dairy farming, which further increases the need for P application, and with it the potential for loss in runoff.
Traditionally, the ability of soils to fix P was determined by a laboratory test called the ‘P Retention’ test, which rated this ability on an arbitrary 0-100% scale. Allophanic soils typically lie in the range of 90-100, with sedimentary soils typically ranking at a much lower 20-30%.
Unfortunately, this test has been hijacked by both ends of the ‘fixation’ schools of thought spectrum. Those seeking to claim that P requirements can be reduced by as much as 90% by applying P as a spray for foliar uptake use it as ‘proof’ that 90% of the P applied in conventional fertiliser is permanently lost through fixation. This is a nonsense on two grounds. Firstly, the test does not claim to determine permanently ‘fixed’ P but rather just the ability of the soil to strip it out of solution and hang on to it rather tightly. This is why it was called the ‘P Retention’ rather than the ‘P Fixation’ test. Secondly, in a grazed pasture, once the herbage has been consumed in the next grazing, and converted into meat, wool or milk, or excreted, what then? The pasture is back to being reliant on soil P. The economics of pastoral farming can not cover the cost of multiple split applications of foliar P fertilisers per year.
Partly as a reaction to this misuse of the P Retention test, and probably partly because the traditional fertiliser industry wished to see its products as being efficient, the test was some years ago renamed the ‘Anion Storage Capacity’ test, or ASC. This is just as misleading, because, as any dictionary will tell you, a store is somewhere you put something until you need it, and from where it can be retrieved at any time you want to. This flies in the face of scientific knowledge on allophanic soils, which many decades after development, still require much higher maintenance P applications than most non-allophanic soils. It isn’t ‘storage’ in the soil, it is in effect ‘theft’ by it.
Like in most things in life, there are equilibria involved. If one stops applying P to these soils, P will continually become available from the surface of clays, but at far, far too slow a rate to maintain a vigorous pasture once readily available P levels in the soil have been depleted, which typically takes 1-5 years. Thereafter, production inevitably falls.
So what actually can be done to reduce this loss of applied P as through fixation?
Reducing P fixation on allophanic soils
On allophanic soils, most of the fixed P is either attached to clay particles via aluminium ions that have themselves been fixed, or precipitated as (more or less) permanently unavailable aluminium and iron-aluminium phosphate compounds. Lime application doesn’t reduce this process very much, basically because the strength of the aluminium-phosphate bond is much stronger than the calcium-phosphate bond. So the already-fixed P is as I say, pretty much stolen for good, not ‘stored’.
There is however real scope to make maintenance fertiliser P applications more efficient, and therefore reduce the amount of P that is required to maintain a given level of production.
There are several approaches to this being investigated. One is to coat granular soluble P fertilisers in a slowly-dissolving coating of some sort, which allows the plant uptake to keep up with the rate of release from the granule, so there is less opportunity for the P to be fixed. Theoretically, if a granule dissolved slowly over 6 months say, the plant might get to use far more of it than at present. There is some evidence that coating fully granular soluble P fertilisers such as DAP with fine elemental S or some types of polymers does improve efficiency, perhaps by 20-40%, but at significant extra cost per unit P.
The second approach is to simply use a slow-release natural form of P such as RPR. The problem with using RPR as the sole source of P on allophanic soils is that because the P is being released from each particle very slowly – typically over 3-5 years – the concentration of soluble P at the surface of the particle is often not high enough to ensure that the majority of P gets taken up by the plant rather than fixed. The answer to this problem, as demonstrated by Dr Sunder Rajan and others, was to combine fine RPR with water-soluble P in the same particles. This can be achieved by either granulating the two components together, or by applying them in a sufficiently damp or wet manner (such as a fluid) than ensures that each particle as it lands will contain both ingredients.
What then happens is that root growth is stimulated in the vicinity of each granule (dry application) or particle (fluid application), which in turn stimulates the utilisation of the RPR component through excretion of RPR-dissolving acidity by the plant roots. Dr Rajan demonstrated a minimum of 30% of the total P needed to be in water-soluble form, and a form of P that (unlike single superphosphate) which not contain gypsum, which inhibits the dissolution of RPR. In my estimation, such a product reduces long-term maintenance P applications on allophanic soils by about 15%. Costs per unit P are comparable with straight soluble sources, and environmental losses are greatly reduced as well.
The third option is to effectively use the maintenance fertiliser applications as a convenient way of getting soil additives delivered to the site of the fertiliser granules or particle. These additives can be anything that will help the applied P get into the plant. For example, some polymers will bind with P to make a compound that the plant can still take up, but is less susceptible to fixation in the soil, at least until the polymer itself is decomposed by soil bacteria. Other additives reduce the levels of toxic soil aluminium, iron or manganese in the soil by precipitating them. Soluble aluminium stunts plant roots, blocks the synthesis of nucleic acid if it enters the roots, and blocks P uptake. These type of additives have I believe enormous potential to improve P efficiency on both allophanic soils and acid, low-pH hill-country soils.
The pumice soils of the central North Island are also of volcanic origin, and also fix P, although not quite to the same degree as the allophanic soils. They need the soil P status raised to a high level, but once there, maintenance needs slowly reduce because of lower net fixation.
Reducing P fixation on acid hill soils
On the yellow-grey and yellow-brown earth sedimentary soils that cover much of New Zealand, much less P is fixed onto clays in plant-unavailable form than is the case on allophanic soils. A much larger source of in-soil loss on these soils is the accumulation of P adsorbed onto, and incorporated into, soil organic matter, specifically has inosital phosphate compounds.
However, unlike the P fixed by allophone, this organic-fixed P can be released by the use of adequate lime. Excessive lime releases it but it then reprecipitates it as calcium phosphate. Infrequent heavy applications of lime, as traditionally practiced in New Zealand, can exacerbate this cycle. He answer is to move to regular applications of reasonably fine maintenance lime with the maintenance fertiliser. Fluidisation represents a cost-effective way of achieving this without risking uneven distribution of differently sized dry components, or aircraft safety.
Another benefit of this approach is to eliminate aluminium toxicity in the soil, which is an insidious and, I believe, increasing problem on hill country farms. Early work by Doug Edmeades and others demonstrated that soil aluminium levels were likely to reduce both ryegrass and clover growth at soil pH levels below about 5 (measured in water). However, researchers in other countries had demonstrated that at soil pH levels right up to at least 5.8, there can be sufficient soluble aluminium in the soil to inhibit the establishment and function of rhizobia in clover root nodules. Some clover plants will still grow, especially in locations where urine has been deposited, supplying very high levels of N and K, but the overall pasture growth will be severely limited by N deficiency. Losses from the system are simply not being matched by fixation of atmospheric N, if rhizobia is inhibiting N fixation.
Although catch-up research was eventually carried out in New Zealand and published a decade later, in the 1990s, the contemporaneous disbanding of the Farm Advisory Division of the MAF meant that to this day, many farmers remain totally unaware of the threat posed by aluminium.
P losses off and through the soil
Reducing P losses in surface run-off
It took many years for traditionally-taught agricultural scientists in New Zealand to accept that any P at all could be lost in surface run-off following rainfall, such was the belief in the ability of the soil to ‘retain’ P. Likewise, when this was eventually established, it took more research by Alan Gillingham, and NIWA research by Long Nguyen funded by Summit-Quinphos, to demonstrate that (i) much of this run-off occurred in the first few weeks after application, and (ii) using sustained-release RPR as the P source reduced P run-off by over 50%, especially soluble P. Losses of P in soil particles will continue to happen through erosion events of course, but the entry of this P into waterways is far less eutrofying. Subsequent work by McDowell and others confirmed lower losses using RPR, and lead to this reduced loss being incorporated into the Overseer program, albeit in a watered-down form, if one can excuse the pun.
These runoff P losses are typically only a few percent of the applied fertiliser, and therefore do not represent a major direct economic loss to farmers, but they represent by far the biggest contributor to the eutrophication of our lakes and waterways. Making RPR the required form of P (or at least 70% of the total applied) in sensitive attachments would single-handedly reduce P run-off by over 50% over a period of 2 or 3 years, to levels less than 0.5 kg/ha annually, which receiving waters could cope with without suffering excess weed growth and algal blooms. Further reductions, to nearer 75%, could be obtained by applying fertiliser in fluidised form, because of its evenness of application, accuracy of placement and consequent ability to avoid direct entry into waterways and onto unproductive areas.
However, the reality is that the superphosphate industry in New Zealand, a duopoly of two giant farmer cooperatives, aided by some ‘friendly’ researchers, has so far represented far too powerful a lobbying and marketing force for acceptance of what must be done to take root politically, nationally or at regional council level. Instead, we have witnessed the absolute charade of many millions of shareholders funds being wasted on a meaningless fight between Ravensdown and Ballance over patents over the use of nitrification inhibitors to reduce N losses from urine patches. Like fiddling while Rome burns, we are playing games while our waterways are ruined.
Reducing P leaching through the soil
Even harder for many scientists to accept was the fact that not only could fertiliser P be washed off the surface of the soil into waterways during rainfall-induced runoff events, but fertiliser and recycled P could be washed straight through the soil into the groundwater. On extremely weathered podzols of the West coast and parts of Northland, fully 30% of the applied P can be leached through the soil beyond the root zone.
On other soils, higher P retention reduces this percentage, but it can still be very significant on non-allophanic soils. The leached P is not just recently applied fertiliser P, but inorganic and organic P ions that have become desorbed from soil clay particles. Because recently-applied fertiliser P is not the only major culprit with leached as opposed to runoff P, changing to slow-release fertiliser will certainly help, but not guarantee that waterbodies will be protected. Another part of the answer here is carefully managing the amount of P in the system, and using plants that are efficient foragers of P.
What needs to be done first?
In the past, fertiliser and lime were treated as essentially separate issues. This was partly because lime in New Zealand was regarded purely as a soil ameliorant, for reducing acidity and with it aluminium toxicity, and not as a source of Ca as a nutrient. Partly because of this attitude, and partly because it suited the spreading industry to deal with dusty lime in fewer, larger applications, the accepted wisdom of maintaining soil available P and S levels, and K on dairy farms, with annual ( and biannual in the case of dairy farms) was never applied to lime itself. The assumption, if considered at all, was that the input of Ca in super would take care of that.
However, two big changes have been occurring in New Zealand for over a decade, one each in hill country farming and dairy farming, that mean that this assumption is no longer valid, even if it used to be.
On hill country farms, several things are causing the regular application of lime, particularly sufficiently fine lime, to diminish. Until 2010, low meat and wool prices meant that often, when the 5-yearly application of lime was due, cash-flow problems meant it didn’t happen. The fertiliser industry ( and the banks ) pressured farmers to put on P as a priority, when in increasingly many cases lime should have been the priority. Now, with a bit more money to spend, farmers are often finding that (a) applications of the increasingly coarser ‘aerial’ lime aren’t giving the same benefit as before, and (b) because of increasingly common aluminium toxicity, they aren’t getting the response to P they used to. In my personal experience, 70-75% of long-term hill country farmers would be of this view. In desperation, many are being sucked into ‘muck and mystery’ wonder products that promise everything, but deliver just a massive hole in the bank account.
What must be done is to increase the fineness of lime considerably. Where aluminium toxicity is possible, then the first option is to apply a capital application of whatever quantity is required to get the soil pH to at least 5.6 (typically 1-2 tonnes/ha), and then maintain it at this level with annual applications mixed with the maintenance fertiliser. The annual maintenance lime requirement will vary with type of fertiliser used, soil type, stocking rate and rainfall, but will typically be around 100 kg/ha. Doug Edmeades and I agree that the new fineness standard for aglime should be 100% minus 0.5mm and 50% minus 0.2mm. To that I would add 25% minus 0.1mm. The actual weight of lime applied must be adjusted according to the CaCO3 content of the actual product used, as commercial lime can range from 80-98% CaCO3.
Where farm budgets simply do not permit capital application of lime, the alternative is to incorporate Quin Environmentals’ “PORTAL” lime-sparer additive with the annual maintenance fertiliser and lime application. PORTAL precipitates soluble aluminium and manganese present in the soil, reducing the need for capital lime and improving P uptake. Where Al and Mn toxicity is not a problem, ryegrass and clover will perform well at soil pH levels of 5.2 and below. In addition, it takes less lime to maintain a soil pH of 5.2-5.5 than one of 5.6-5.8.
Lime as fine as I have specified may be a problem for some contractors, who will be concerned about the risk of bridging in fixed-wing aircraft hoppers and helicopter buckets. Options include the use of flow agents, attaching vibrators to hoppers and buckets, mini-granulating the lime, or fluidising the lime/fertiliser mix. I have come to view the latter as the way forward.
As far as P is concerned, sustained-release, non-leaching reactive phosphate rock (RPR) should be used wherever soil and climate conditions permit, augmented if required by a proportion of soluble P (plus other nutrients and lime as required of course).
On dairy farms, and on the more intensive sheep and beef farms, the use of ‘super’ is giving way to DAP, due to increased fertiliser N usage and relative costs of P, especially on an applied basis. I believe that overall, this is a good thing, as efficient elemental S can be added to exactly meet S requirements. Super on the other hand contains a fixed 1.4 to 1 ratio of S (all as sulphate) to P, double the actual agronomically required S to P ratio. On many soils the sulphate is so easily leached (taking valuable cations with it), farmers have to use elemental S-fortified ‘sulphur-super’ instead. On the volcanic ash soils that do retain the sulphate in plant-available form reasonably well, the excess sulphate builds up over years to the point where the soil will not hold any more, and then the excess gets leached, taking cations with it.
Where the N content of DAP relative to P is excess to requirements, RPR should be used instead of soluble P fertiliser.
As DAP contains no Ca, it is becoming increasingly important for dairy farmers to maintain soil and pasture Ca levels by ensuring that adequate quantities, again of a sufficiently fine lime, are applied mixed with the maintenance fertiliser. If P is being applied twice a year, each application should include typically 100kg/ha of lime.
Optimising soil and herbage Ca levels is particularly important on dairy farms, where high applications of potash are practiced to maximise pasture growth, but which increase the risk of mineral imbalances in the animal if other cations are marginal.
Given the massive advantages in fertiliser N efficiency that are achieved by applying urea in fluidised form containing a urease inhibitor, and the increasing frequency with which N is being applied to dairy farms, it is to me a no-brainer that fluidised application of all fertiliser and lime is the way forward on intensive farms. With increasing improvements in fluidisation technology, total costs per unit ‘plant-effective nutrient’ or PEN are now fully competitive with conventional application, and the benefits and convenience (including being able to mix anything with anything, solid or liquid) are immense.
Sufficient knowledge already exists to reduce P losses from agriculture to levels that would reduce eutrophication of our waterways and lakes to very satisfactory levels within 5 years. The barriers are a conservative fertiliser supply and application industry in which no one wants to take the risk of being the first to act, and regional councils who lack the courage to lead. Instead a ‘bit-by-bit’ approach is being taken, by setting caps on fertiliser application and dairy farming in the most environmentally-sensitive catchments. While this action is helping water quality in these areas, it is not providing the necessary impetus for the introduction of real initiatives in farm nutrient supply and management, with support rather than resentment from farmers and farming organisations.