Can we Reduce Costs and Increase Profits?

Low cost and optimised costs

This changed economic and production background demands that we question the use of high input systems. The fixed-rate, high-input approach must be challenged to ensure that cereal production remains profitable and competitive. While cost reduction can be a valid strategy targeted at improving competitiveness – it can not be an end in itself. The aim must be to get the correct balance between costs and output, i.e. cost optimisation. This objective can be difficult to achieve in practice, largely due to differences in growing seasons.

Very low input cereal production would not allow us to benefit from our natural advantages (unless in an organic system). The use of relatively high levels of inputs, particularly fungicides, to protect crops from wet weather diseases, has allowed us to consistently produce the highest cereal yields in Europe.

System trials

To research low-cost or optimised cost strategies, a number of approaches can be taken:

1. Single factor trials: The response from the application of individual inputs can be examined in isolation. E.g. we can individually look at the response to fungicide type and rate, Nitrogen, seedrate etc. This component work is the mainstay of crop production research.
2. Multi factor trials: The response to a combination of factors can be studied E.g. the effect of fungicide on different varieties. This yields information about the interaction of different components of a production system.
3. Systems trials: Complete production systems can be evaluated. In this approach, systems of production (e.g. low input systems, high input systems etc) can be compared.

There are advantages and disadvantages to each of these approaches. The single factor approach, while often giving the clearest results, fails to capture possible interactions between different inputs. The second option can pick up these interactions, but experimentally can be very complex if combinations of different levels of inputs are to be evaluated. The systems approach does not attempt to measure the interactions between the different inputs. It relies on the sensible choice of input levels which are characteristic of the system chosen. Only the complete systems can be validly compared. While this makes the trial design manageable, the downside is that we may not know what components contribute to a particular performance result.

From a research perspective, no single approach is perfect. In practice we use a combination of the three approaches. The input levels used in the system trials are frequently determined by component trials at Oak Park.

Systems trials and cost reduction

If a research objective is to reduce or optimise production costs, then by working with more than one input for example gives us a greater opportunity to reduce costs. There may also be a synergistic benefit in tackling more than one input. In its simplest form, if reducing the rate of one input lowers the yield potential, then it may be possible to reduce other inputs without further negative impact on yield. In addition there may be further biological reasons for getting a synergistic benefit from reducing more than one input. If a low seeding rate is used with winter wheat for example coupled with a lower nitrogen rate, then the widely spaced and not-too-lush plants may be less susceptible to fungal disease spread, thereby reducing the need for fungicide.

Experimental design and treatments

Three rotations and up to four input levels are being evaluated in the systems work. The rotations include a five-course break-crop rotation which includes: Beans; Winter Wheat, Spring Barley, Oilseed Rape; and Winter Barley. A three-course cereal rotation which includes oats as a break for take-all includes: Winter Wheat; Winter Barley and Winter Oats. The third growing option: monoculture, is evaluated by growing continuous winter wheat and continuous spring barley.

While just two input levels were used originally, the experimental layout used since 2004 allows four input levels to be assessed on each crop grown in each rotation. Four input options are used in Winter Wheat and Spring Barley crops. Two levels of inputs – ‘High’ and ‘Low’ are used on the other crops. The input levels used include:

• High: Commercial rates of all inputs including Teagasc recommended rates of fertilisers and typically about 80% of plant protection product label rates.
• Low: Approximately 80% of the recommended N rates and 50% of the plant protection product rates used in ‘High’.
• Decision-Based High (DB High): A strategy where knowledge about the crop and growing conditions is used to optimise input application to achieve high yields and optimum returns, using the ‘High’ treatment as a base. This strategy was only used on winter wheat where the DESSAC decision support system was used to determine fungicide application levels.
• Decision-Based Low (DB Low): A strategy where knowledge about the crop and growing conditions is used to optimise inputs in a low input approach based on the ‘Low’ treatment. This treatment is also only used on winter wheat where the ‘Septoria timer’ is used to influence the fungicide programme.
• High-Nitrogen Low-Fungicide (HN LF): A treatment coupling the Nitrogen levels of the ‘High’ strategy with the fungicide rates of the ‘Low’. It is only used on spring barley to determine which factor influences the response in a two-factor design.
• Low-Nitrogen High-Fungicide (LN HF): A treatment utilising the Nitrogen levels of the ‘Low’ strategy with the fungicide rates of the ‘High’. Only used on spring barley.

The basic ‘High’ and ‘Low’ input levels have been applied to crops for 11 seasons. Varieties and the plant protection products used have changed over this period to reflect new developments in these areas. In 2004 the decision-based and additional nitrogen/fungicide combinations were added for winter wheat and spring barley respectively. Phosphorus and potassium application levels are based on soil analysis where the low input strategies are used, while a maintenance dressing of P and K is always used with the high input strategies. In practice, no P was necessary on the ‘Low’ plots until 2005 when maintenance dressings recommenced. For most years of the trial, the variety and sowing date used for winter wheat was the same for all input treatments. For the 2004/2005 and subsequent seasons it was decided to use a later sowing date and more disease resistant variety in the ‘Low’ strategies to reduce disease and weed pressure. A full description of the input levels used in 2006 on winter wheat and spring barley is given in Appendix 1.

The trial has been carried out for eleven seasons at Knockbeg. To assess and validate the input levels at a more disease prone location, a site incorporating the continuous wheat and continuous barley elements was established in Kildalton. This site has generated two years data to date.

The performance of the treatment strategies is assessed in many ways. Crop and soil nutrient levels, disease assessments, all yield components and grain quality measurements are recorded in all cases. Financial performance is assessed using the cost of the various inputs. For the purposes of this paper, only the two main performance criteria: grain yield and production margin are examined.

Costing analysis

Production costs are estimated by using product prices typical of those available to a 100 ha producer of cereals. While these are calculated each year, to deal with product inflation, the prices used in this paper, when analysing performance over eleven seasons, are those applicable for 2006. Similarly grain prices used in these calculations are the mean of the 2005 and 2006 harvest prices corrected to 15% moisture content. The exception to this approach is when the decision-based and alternative nitrogen and fungicide combinations of the last 3 years are assessed – in this case the actual prices appropriate for the three years are used. In all cases machinery costs are included at contractor charge level. Cost samples are given in Appendix 2. While the costs used will differ from those achieved on some farms, the approach allows a valid and accurate comparisons of the treatments.

High vs Low inputs

The trends and levels of grain yield for the cereal crops grown for 11 seasons at Knockbeg are illustrated in Figures 1, 3, 5 and 7. Each graph includes data points and trend lines for yields of the ‘High’ and ‘Low’ treatments with each crop. In this series of graphs the yield figures are the combined figures for all rotations where the crop type is grown in rotation. E.g. the wheat yields in Figure 1 are the average values for wheats grown in monoculture and the two rotation types at each input level. The associated crop production margin trends are illustrated in Figs 2, 4, 6 and 8. These margins are calculated using today’s costs and returns and exclude any area aid or single farm payment.

Fig. 1: Winter wheat yield trends with ‘High’ and ‘Low’ input levels: Knockbeg.

Winter wheat: Wheat yields at both input levels varied considerably over time with yields varying from 9.2 to 13.2 t/ha for ‘High’ input levels and from 9.0 to 12.3 t/ha for ‘Low’ inputs. Variations in growing seasons influence disease development and yield potential. The average difference in yield between the two input levels was 0.87 t/ha in favour of the high input system (8.4%) but this varied from no yield difference (1998, 2005) to a 2.5t/ha (23%) difference in 2004.

Fig. 2: Winter wheat margin trends with ‘High’ and ‘Low’ input levels: Knockbeg.

The true measure of the performance of these systems though is their impact on production margins (Fig 2). On average the ‘Low’ input system has a production cost of €161/ha less than the ‘High’ system which amounts to a saving of the equivalent of 1.4t/ha in grain yield. Over the 11 years of the trial, the use of ‘Low’ input levels resulted in an average increased margin of €62 / ha. With low inputs the cost saving made was worth more than the yield foregone. The dramatic variation in margins across years is a feature of low margin production systems. In 9 out of the 11 years, the low input system generated more income. In only one year when winter wheat yield potential was at it’s highest (2004), did the high input system give a better return. Seasons which had a relatively low yield potential greatly favoured the low input approach with margin differences approaching €150/ha. It is ironic that the high input system which is considered an insurance based approach, results in a lower but also more variable production margin. Low cost approaches to winter wheat production offer considerable potential.

Winter Barley: The contribution of the ‘High’ input system to yield in winter barley was more consistent and averaged 1.1 t/ha over all years representing a 14% increase in yield (Fig 3). The difference in yield between the two input levels varied from 0.8t/ha (12%) to 1.6t/ha (20%).

Fig. 3: Winter barley yield trends with ‘High’ and ‘Low’ input levels: Knockbeg

When production costs are included, and all 11 years are considered, input level had no effect on margin (Fig 4). With low inputs the cost savings made are cancelled by the reduction in yield. While there are some differences in margin in individual years, they were small. While the low input system evaluated showed little scope for improving margin in winter barley, there is equally no yield penalty from using a low input approach.

Fig. 4: Winter barley margin trends with ‘High’ and ‘Low’ input levels: Knockbeg

Winter Oats: The yield of the winter oat crop was also increased where high inputs were used (Fig 5). Overall a yield increase of 1t/ha (11%) was recorded with a relatively consistent increase in most years, except for 1998 when the crop lodged.

Fig. 5: Winter oats yield trends with ‘High’ and ‘Low’ input levels: Knockbeg

While this yield difference is not that large, it does result in an average margin increase of €46/ha for the high input strategy (Fig 6). There is little difference in costs between the two systems as the ‘High’ input approach chosen for oats in this trial uses a quite modest level of inputs. Note the negative margin recorded for 1998 using today’s prices.

Fig. 6: Winter oats margin trends with ‘High’ and ‘Low’ input levels: Knockbeg

Spring Barley: Spring barley had the greatest difference in yield between the two input systems when expressed in percentage terms. The average yield difference over the 11 years was 1.2 t/ha which represents 18% of the low input yield level (Fig 7). While there was some variation from year to year, the trend was similar in all years.

Fig. 7: Spring barley yield trends with ‘High’ and ‘Low’ input levels: Knockbeg

Production cost differences between the low and high input strategies used on spring barley was €81/ha using today’s costs. A yield difference of just 0.7t/ha would pay for the higher input levels. Overall the high input strategy results in a margin improvement of €55/ha (Fig 8). This difference cannot be ignored in the context of the low margins achieved with spring barley. In 2002 when yields were low, at today’s prices, the low input approach would have resulted in a negative margin, while higher input levels would have prevented a negative margin being generated. These results are important for cereal growers that participate in the REP scheme where allowed nitrogen levels are similar to the low-input treatment here. The cause of this difference is discussed in a later section.

Fig. 8: Spring Barley margin trends with ‘High’ and ‘Low’ input levels: Knockbeg

Input levels and rotations

The Knockbeg trial allows the interaction between rotation type and input level to be assessed. Winter wheat and winter barley are both grown in potentially beneficial parts of the two rotations. Cereals grown after break crops may benefit less from higher levels of inputs than cereals grown in monoculture. The average (11 year) yields for winter wheat and winter barley grown in rotations or in monoculture (or a non-beneficial part of the rotation) are given in Figure 9 with the associated production margins in Figure 10. While the trends in this data would indicate that crops grown in beneficial positions of rotations benefit more from a low-input approach, the differences may not be significant. With wheat for example a low-input approach boosts margins by €72/ha compared to €42 for continuous wheat. Similarly low-input winter barley grown after oilseed rape had an increased margin of €16 /ha, whereas the low-input strategy applied to winter barley grown after winter wheat in a cereal rotation showed a reduced margin of €15/ha (difference of €31/ha).

Fig. 9: The effect of rotation on wheat and barley yields with different input levels

Fig. 10: The effect of rotation on wheat and barley margins with different input levels

Decision-based strategies

Additional decision-based input strategies for winter wheat were added at the Knockbeg site for the 2004 season. These were also incorporated in the Kildalton site for which we can report two years data (2005 and 2006). The yields and margins associated with these input strategies are illustrated in Figs 11-14. Note the actual costs used in the margins reported here differ from the average costs used in the analysis of the full 11 years data and consequently are not directly comparable. In 2004, the decision based strategies resulted in yield reduction and margin reduction compared to the fixed approaches at Knockbeg (Fig. 11 and 13). The DESSAC system was allowed select an old fungicide technology product which restricted yield potential in the DB-High treatment, while grass weed competition and subsequent expensive late-control measures reduced yields and margins in the DB-Low treatment. In 2005, the decision-based treated plots yielded similarly to the fixed-rate plots resulting in similar margins. There was a tendency for margins to be improved by the decision based strategies in 2006 as yields were maintained and costs reduced.

While the overall benefits from decision-based systems in this trial are negligible to date, the results of the last two years and, particularly the performance of decision support systems in other trials at Oak Park (Burke and Dunne, In-Press), would indicate that they have potential in system-based approaches to crop production. Our long-term wheat results (Fig 1 and 2) which showed that matching input strategy to yield potential could result in substantial margin boosts, also indicates the scope for decision-support type systems.

Fig. 11: Winter wheat yield with standard and decision-based strategies at Knockbeg.

Fig. 12: Winter wheat yield with standard and decision-based strategies at Kildalton.

Fig. 13: Winter wheat margin with standard and decision-based strategies at Knockbeg

Fig. 14: Winter wheat margin with standard and decision-based strategies at Kildalton

Spring barley nitrogen and fungicide rates

The relatively poor performance of the ‘Low’ input system with spring barley stimulated further studies to determine the contribution of the fungicide and nitrogen elements of the strategies to these results. Two further treatments HNLF (High Nitrogen, Low Fungicide) and LNHF (Low Nitrogen, Low Fungicide) as described earlier were added in 2004 at Knockbeg (3 years data) and Kildalton (2 years data). The effect of these treatments on yield and margin for all years is summarized in Figs 15 and 16. At Knockbeg, the major factor influencing yield was nitrogen level. Both low (105kg/ha) N level treatments (Low and LNHF) gave reduced yields regardless of fungicide level (Fig 15). Where a high nitrogen level was used, there was some small level of response from using full-rate (High) rather than half-rate (HNLF) fungicide. This response to fungicide just about paid for itself leaving no difference in margins (Fig 16). Where lower Nitrogen levels were used at Knockbeg, there was no benefit in using full-rate fungicide, resulting in a much lower margin where full-rate was used.

Two years data from the more disease prone Kildalton site shows a slightly different picture. Overall there is a significant response to fungicide at the higher nitrogen level resulting in the ‘High’ treatment giving the best return. At a lower applied Nitrogen level, there was no response to changes in fungicide level resulting in the ‘Low’ treatment giving a much higher margin than the LNHF treatment. At both sites, at lower N (equivalent to REPs level), the best option was to use half rate fungicide. Also the difference in margin between the basic high and low treatments was less at Kildalton than at Knockbeg.

Fig. 15: Spring barley yield with four input levels at Knockbeg (3 yrs) and Kildalton (2 yrs)

Fig. 16: Spring barley margin with four input levels at Knockbeg (3 yrs) and Kildalton (2 yrs)