There has already been considerable recent change in fertiliser recommendations for potatoes - for instance the new SNS Index system for nitrogen and the revised levels of phosphate adopted in the 7th edition (2000) of Fertiliser Recommendations Book RB209. Suggestions are being made by other sources that there is still considerable scope for economies of nutrient use on this crop. This technical note in conjunction with PDA leaflet No 15 reviews the role of potash and examines optimum rates of use for potatoes.

Principles

It is important to identify the different reasons for applying potash to the potato crop which are threefold :-

• to achieve economic optimum yield response
• to maintain soil K reserves
• to achieve quality requirements for marketing

Recommendations for potash in PDA publications and RB209 are based on the amount removed in the harvested crop with adjustments according to soil K reserves. The resulting recommendations (applicable to potatoes whatever their target market) are normally adequate to achieve full potential yield. Additional K nutrient beyond these recommended rates will normally neither increase yield nor further improve quality aspects such as dry matter/bruising resistance/fry colour etc. If "optimum response" rates which are less than crop removal are used, soil K reserves will decline. The effect that this may have on future yields of the whole rotation for different soils is not adequately known to run the risk of such a policy. So the principle of K manuring for arable crops is to maintain soil K for most soils in the range 120-180 mg/l (K Index 2-). Farm practice which decreases this level of fertility is not sustainable in the long term. Specific soils are now recognised as exceptions - true sands and loamy-sands have lower target ranges for soil K because their low clay content means they are not able to retain the same level of K reserves as other soils, whereas some clay soils have the capacity to release K reserves to crops over a number of years and thus require lower levels or nil potash additions. Whilst increasing areas of potatoes are grown on contracted land where the tenant may have no long term financial interest in maintaining nutrient reserves, it is not felt appropriate to provide general recommendations that are incorrect for the long term well-being of the land.

Difficulties of experimentation

The BPC Research Review - Potato Agronomy by EJ Allen and RK Scott emphasises the difficulties of experimentation with the potato crop. Frequently the size of variation between plots results in "standard errors" that are so large that yield differences between treatments are not "statistically different". This means that yield effects cannot with certainty be attributed to the input being examined and may only be the result of general yield variation across the field. However this does not mean that the yield effects are definitely not related to the input being examined. Additional small consistent yield increases which cover the cost of inputs are more important to farmers than statistical significant differences in experiments.

This effect is highlighted in a recent series of 33 experiments which have been reported in Responses of potato to potassium fertilisers by Allison, Fowler and Allen (Cambridge University Farms) in the Journal of Agricultural Science (2001) 136, 407-426. The authors analysis of results in this paper indicates a very poor relationship between soil K Index and yield response to potash fertiliser and the conclusion drawn is that the optimal fertiliser rate is rarely more than 200-250 kg/ha K₂O. This is in conflict with recommendations given elsewhere (RB209 and PDA literature).

Review of response data

The data from the Cambridge University Farm trials published in the J. Agric Sc. paper referred to above, has been considered by PDA and its independent advisors. This review has identified two points of concern in relation to the author's conclusions.

First - From the information given, many of the trials were factorial in design with varying rates of N, P, K or Mg. There is no information as to whether or not the response to K was affected by different amounts of the other nutrients. Second - the fertiliser was applied over the ridges after planting and raked into the soil. This is not normal commercial practice. K is not readily mobile in soil, except for very sandy soils as mentioned above, and so the method of applying it in these experiments would minimise rather than maximise the opportunity of roots growing into soil enriched with the newly applied K. Much of the variability in the response to K was probably due to the method of application.

Potash Response Curve

By combining the data from a group of trials it is possible to derive information on the average response. Not all experiments had a plot without added K and not all tested the same rates of K so we have used only data from experiments with a nil K plot and calculated the increase in yield for each amount of K tested in those experiments. There were data from 23 experiments and the increase in yield at each level of K is shown in figure 1 (blue points). The responses are very variable, a few trials had large increases in yield, most had a small increase and in some there were negative effects. In figure 1 the red line shows the calculated average response curve. This shows that the increase in yield gradually diminishes and reaches a maximum as applied K increased.

Yield increase v Potassium applied

Figure 1. All soils	Figure 2. K Index 1 soils
K Fertiliser kg/ha K2O	K Fertiliser kg/ha K2O
Yield Increase Average response curve

The optimum rate of potash can be calculated as the point at which the cost of applying an extra kg/ha potash is balanced by the value of the yield increase. On this basis the optimum potash rate from all these trials was 444 kg/ha K₂O. The same approach has been repeated for those trials where soil K was Index 1 as shown in figure 2. For these trials the average increase in yield was given by 448 kg/ha K₂O.

NB Figure B on page 8 of PDA leaflet 15 plots the yield increase for Index 1 sites with the green line giving the best fit response curve and the red line indicating the cost of applied potash in terms of tuber yield.

Economic implications

These trials showed that possible financial benefits of potash application can be very large, on one site 250 kg/ha potash increased yields by 21 tonnes/hectare. That could be worth £1500 per ha. The cost of potash is minor, 250 kg/ha costs about £55.

Because of the variability of the data a common method of assessing practical fertiliser rates is the amount which will give 90% of the maximum yield. Using this approach the practical economic implications are clear. For all 23 sites the value of the response with potash to achieve 90% of maximum yield assuming potatoes at £80/t was £381/ha from an input of potash of 344kg/ha costing £75/ha assuming potash at 22p/kg. For Index 1 sites the value of the yield response was £585/ha from an optimum potash rate of 330 kg/ha costing £73/ha.

These estimated response optima are very close to the current rates recommended for 50 t/ha crops on soils at K Indices 0 & 1 in RB209 and PDA literature.

Comparison with sugar beet

Recent trials with sugar beet have shown that soil Index is not a good predictor of yield response for that crop (Potash News - September 1999, British Sugar Beet Review - Autumn 1999 and Summer 2001) and it has been suggested that potatoes are directly comparable. However we do not believe that this comparison is valid. The main explanation for sugar beet being able to produce full yield regardless of topsoil K Index, is that a considerable amount of potash is taken-up from deeper soil zones and thus K is not limited. Whilst some potato varieties can be encouraged to root more extensively than traditionally believed, it is not accepted that they can recover 50% or more of total K requirement from depth as can beet. Thus maintenance of top soil K fertility remains the key principle for potato manuring.

Soil K demand

Whilst factors such as improved cultivation techniques, de-stoning, changing varieties, increased use of irrigation, etc. have all contributed to an improvement in the crop's ability to take-up K from the soil and thus reduce its dependency on and response to potash fertilisers, there has also been a continuing increase in average yields. Higher yields require greater total uptake and result in larger removals of potash. National average maincrop yields have increased from around 30 t/ha in 1980 to around 50 t/ha now which represents an increase in potash offtake of over 110kg/ha. This trend of increased yield and offtake is also true for other crops in the rotation. In view of this higher demand it would seem prudent to ensure that adequate supplies of this important nutrient are maintained rather than run-down and that application rates are based on the replacement of potassium.

Conclusion

The general guidelines and recommendations for potash for potatoes in PDA leaflet 15 and Fertiliser Recommendations Book RB209 (2000) 7th edition provide the best guide to the correct use of this nutrient based on a review of current knowledge. These general recommendations should be fine-tuned by the use of site specific knowledge where this is available and relevant.

For further information please contact:
info@pda.org.uk
Potash Development Association

update: Mar 2002