Scottish Government

2 EXECUTIVE SUMMARY

2.1 Objectives

The aim of this study has been to systematically review and present the operation of the crop and herbage variety trialling system in Scotland, to present its costs and to appraise the value of Scottish Government interventions in the system. This has involved interviewing representatives of key organisations involved in the system both within and outwith Scotland. In addition details of the financing of the system as well as related activities have been researched and used to conduct an evaluation of the return on Scottish Government investment.

2.2 The combinable crops NL and RL system

The system for evaluating combinable crops through the National List ( NL) and Recommended List ( RL) approach has been the subject of regular review over the last 20 years, resulting in what some of those involved in the system report as a "slimmed down" regime. For some crops, particularly in the North (Scotland and North of England) there is a perception across the industry, reflected in interviews, that the system is fragile and that trial loss through reduced funding or trial failure due to weather conditions will result in delayed recommendations and/or reduced confidence in varietal comparisons.

When a variety is National Listed it enters the EU Common Catalogue and can be marketed anywhere within the EU. The trials requirement for National Listing in the UK is minimal with 1, 2 or 3 trial sites required per year depending on the crop. However, more sites are grown in order to build up a body of data which is then used to underpin the RL system. These extra NL (or NL equivalent) sites are funded primarily by British Society of Plant Breeders ( BSPB), but also by Home Grown Cereals Authority ( HGCA) and Scottish Government ( SG). The RL system is entirely funded and run by HGCA. In total £269,047 (2008) is required to run NL (and NL equivalent) and RL combinable crop trials in Scotland of which just over £50,000 per annum comes from Scottish Government.

2.3 The herbage NL and RL system

The grass system is similar to that for combinable crops in terms of legal requirements for National Listing. For example, for perennial ryegrass ( PRG) 3 sites are the minimum required out of the total of 6 in the UK (of which there are 2 in Scotland, 1 in Northern Ireland, 3 in England). Funds proportionate to the Scottish rotational grass area are provided by BSPB to help support the Scottish NL equivalent system which is largely funded by Scottish Government. The RL system is completely funded through a voluntary Grass Levy contributed by participating merchants, though NL trials are critical in the provision of sufficient data for the RL stage. The total cost of the NL/ RL system in Scotland is £255,898 (2008) of which the Scottish Government funds £159,008 (62%).

2.4 Physical returns from the NLRL system

The NL/ RL system is effective in that uniform protocols allow ready statistical assimilation of NL with RL data. End users are very supportive of the system and the Recommended Lists for both combinable crops and herbage are well received and well used, the latter more heavily by the trade given that grass is sold in mixtures. It is anecdotally estimated that only 5% of combinable crop seed sales are of non recommended varieties. For grass, 93% of all seed sales contribute to the levy scheme and there is no knowledge of anything other than category 1 or 2 varieties (the highest selection grades) being sold. For combinable crops approximately 1 in 12 NL1 candidate varieties will ultimately be recommended. For herbage approximately 1 in 8 enter category 1.

Recent analysis of the yield increase attributable to variety improvement suggests a 1% per annum gain for combinable crops. Improvements in non yield factors such as disease resistance, marketable quality, straw strength, etc may enhance this figure. Of greater importance may be the ability of the system to weed out varieties with defects which could prove catastrophic under certain conditions.

Similarly the data for herbage suggests a Dry Matter yield increase of at least 0.5% per annum with some data suggesting nearer 1% per year if quality factors are taken into account. Certainly there is clear data to show progressive improvement in sward persistence, winter hardiness, seasonality of yield (more early spring and late autumn growth) and nutritional quality. It is important to note that for herbage, unlike cereals, on-farm evaluation is very difficult. It is impossible to measure the contribution of one new variety when grown in a mixture of varieties/ species for different end uses, where management is a major influence on yield and where yield expression is via animal performance. This means that critical plot trials are the only accurate form of evaluation.

2.5 Economic Evaluation of Scottish Government Input to the NL/ RL system

2.5.1 Methodology

Evaluating a trials system involves estimating the value of an individual trial and the impact of stopping that trial or reducing its scale on the accuracy of selection of varieties which have improved economic traits. There is no benefit from a trials system which, through lack of data, makes selection mistakes and results in marketed varieties which do not give regular long term improvements in economic performance. However, it is very difficult to convert the probability of selecting a good or bad variety from a particular set of trials into a financial impact for Scottish agriculture and to attribute the share due to Government interventions. Major assumptions, to which individual risk ratings have been applied, have been used in the evaluation and readers should bear these in mind when drawing conclusions from the results. The main areas where assumptions have been made are as follows.

• Yield improvement due to variety selection (0.8% fresh weight per annum for cereals, 0.5% dry matter per annum for herbage)
• Cereal prices (based on 2008 averages)
• Number of years a variety will be marketed (based on life in Recommended Lists)
• Delay in industry selecting a good variety if there is no systematic trials system
• 15 point spread of performance of trialled varieties around the mean.
• The selection "grey area" around the mean within which there is low certainty of the true performance of the variety in the long term.
• The use of an increase in the size of the "grey area" as a result of receiving data from fewer trials, as a means of assessing the value of Government funding of extra trials.
• The contribution of non field trial disease testing
• Yield loss through disease due to poor variety selection
• The value of the non Scottish trials as a means of selecting varieties for Scottish conditions

2.5.2 Cereals

Scottish Government has provided financial support in 2008 for NL equivalent trials on winter wheat, winter barley, spring oats and untreated spring barley. The aim of these trials is to provide more reliable data on performance of varieties in Scottish conditions as most (except spring barley) are under-represented in Scotland. No financial support from the Scottish Government is provided at the RL stage. In addition SG also provides financial support under the section in the contract referred to as "advice to Ministers", which enables SAC to represent Scotland on the national committee structure and to provide for knowledge transfer. Indirectly this also supports Scottish research and the knowledge transfer infrastructure in the area of crop trials.

The annual benefit of higher yields due to improved cereal varieties in Scotland is estimated at around £2.6M. However, some variety improvement would continue without an NL/ RL system as companies compete for sales. The value of the NL/ RL system is largely due to the much faster identification of the truly better varieties. The assumption in this report is that it halves selection time, so the value of the NL/ RL system is around £1.3M per annum. The Scottish Government financial support improves the accuracy of selection for Scottish conditions. Of the £1.3M benefit from the NL/ RL system, the majority (around £1.1M) would still be gleaned from the UK wide NL/ RL system even if SG provided no extra funding in Scotland. The economic value to Scotland of the extra Government funded trials is estimated at around £0.2M per annum. These figures ignore quality aspects of improved varieties.

2.5.3 Herbage

Scottish Government financial support, of herbage variety selection in Scotland, is far greater than the support for combinable crops. Scottish Government funds the bulk of the Scottish NL equivalent system, which underpins the entire NL/ RL regime, even although the RL is funded by the Grass Levy on herbage seed sales in Scotland. In addition Scottish Government provides the same type of financial support for "advice to Ministers" as in the cereal trials system.

The value of improved herbage varieties is very difficult to measure. Research provides clear figures on the increase in yields and quality factors as a result of improved varieties, but the economic benefit of these is expressed as livestock performance and the production of meat, milk and wool. The output of livestock products is influenced by a very wide range of factors and teasing out the impact of herbage improvements is near impossible. Several methods are used in this evaluation, resulting in an estimated annual benefit to Scotland from improved herbage varieties of around £4.15M. It is very difficult for a farmer to identify the impact of one new variety in a mix, so the contribution of the NL/ RL system to this annual benefit is high - estimated at approximately £3.1M. However, the value created by Scottish Government support to an independent Scottish NL/ RL system is relatively small (estimated at £0.45M of the total value of £3.1M). This is because reliance on the other UK Recommended Lists would throw up approximately 40% of the same varieties and even a single trial NL/ RL system as in Northern Ireland, funded by BSPB and the Grass Levy, would have some impact.

2.6 Conclusions

2.6.1 General Conclusions

2.6.1.1 Variety improvement has a large economic impact despite generating small annual gains, as it is accumulative and has industry wide uptake resulting in a large national impact. It is not surprising that this was one of the earliest areas of Government intervention in agricultural research and development. The question for today is the role of Scottish Government funding in achieving this benefit. The understanding of the perceived benefit of improved varieties is high within the industry, so should it fund the system and if Government support continues where can it have maximum additionality?

2.6.1.2 There are big differences between selecting improved cereal and herbage varieties. Herbage varieties are sold in a large number of different mixes so identifying the impact of one new variety through on-farm observation is near impossible. This has major implications for the role of Government and for the regionalisation of selection.

2.6.1.3 Some features of Scottish Government funding are a historic carry over from a previous era. There is a need for evidence based support both to make the best use of taxpayer's funds, but also to ensure the trial system delivers for Scottish agriculture.

2.6.1.4 Advice to Ministers ( ATM) support

The current ATM contract with SAC involves direct support (attending meetings on behalf of Scottish Government), breeder requests (for example where breeders are guided around trials) and Knowledge Transfer (where information from variety trials is extended to end users).

The need for Scottish Government to employ SAC to represent them at meetings depends on the ability, experience and knowledge of staff currently employed by Government either centrally or at SASA. It is clearly desirable that Scotland is represented at the various UK committees, but may not be essential. Absence, however, may result in the system drifting toward procedures and outcomes which favour southern Britain. It is up to Scottish Government to decide if they wish to resource this function and if so in what manner they wish to fund and organise activities.

It is highly questionable whether Scottish Government should fund SAC to guide breeders through their trials, as this is a business development opportunity for SAC and has minimal wider public benefit. Breeders Open Days would allow breeders to view the trials more cost effectively than a series of individual visits.

Knowledge transfer, funded by Government, should not be a major feature of trialling at the NL level as results from NL level trials are largely confidential. RL related open days funded by HGCA or closed days for Grass Levy contributors do not seem to be appropriate for Scottish Government support. Likewise Scottish Government financial support for SAC to brief its own consultants alone on the recommended lists is inappropriate. However, any decision to reduce KT support to SAC under the ATM contract needs to be considered within Scottish Governments overall policy towards KT. KT and KE (Knowledge Exchange) is now a major part of much of Scottish Governments mainstream agricultural research. In addition Scottish Government has started providing funds, to which SAC would be eligible, via the Scottish Government Science Engagement Grants 2009-10 to improve links between science and the public.

2.6.2 Conclusions on Scottish Government inputs to the cereals NL/ RL system

• The figures suggest that the value of Scottish Government interventions in cereal trialling create around £200,000 of value at an annual investment cost of around £50,000, excluding the relevant parts of the "advice to Ministers" funding which in terms of supporting the selection process and recommended list publication might total another £25,000 or so of annual investment.
• Based on the assumptions used in the study, Scottish Government support to Spring Barley ( SB) untreated trials gives a large return (around £56,000) on the £8,000 cost (2008/09) because there are specific Scottish issues which make local trials more statistically important and because the Scottish crop is large. Other factors might also need consideration here, including the reducing range of fungicide active ingredients allowed under EU legislation, the environmental impact of increased spray use and the impact of climate change on disease patterns.
• The extra impact that Scottish SB trials have on yield selection is lower, because there are plenty of official trial sites in Scotland and throughout the UK, but the SG funded trials still give a good return (around £19,000) on funds invested, especially as the figures suggest the disease benefits alone justify the additional replicates.
• The value of the intervention in winter barley ( WB) trials is surprisingly low (£18,666 added value versus £13,540 cost), but still positive. It is a fairly small crop in Scottish terms and UK wide trials could be argued to have more relevance for Scottish variety selection than is the case for other cereals. This may be underestimating the different agronomic conditions in Scotland, including winter hardiness and disease, which can be very damaging to winter barley in a wet season. The value of influence over the choice between 2 row varieties better suited to Scotland, but with lower yield than 6 rows suited to English conditions, may also be underestimated.
• The value of the intervention in spring oat ( SO) trials is small in absolute terms (around £17,500) because the area of crop grown is small, though the benefit of the extra trial is much higher per hectare of oats grown than is the case per hectare of winter barley grown. The niche market importance of the milling oat crop may also be underestimated, though trials throughout the UK will focus on oats for this end use.
• On the basis of the above, the overall conclusion would be that the untreated spring barley work, and the extra winter wheat and spring oat sites add value. The relative value of winter barley and winter oilseed rape trials needs to be reviewed.
• Of the 5 options for change investigated the most promising are scrutinising parts of the "advice to Ministers" support which are hard to justify (as discussed in general terms earlier), reviewing the relative value of winter barley trials versus other trial opportunities, and a shift to private sector responsibility for disease control work. Total removal of Scottish Government funding from the trials system gives a net loss in Scottish industry yield benefit of almost £85,000 per annum. The option of moving to funding only economic trait work and laboratory based disease assessments is difficult to quantify.

2.6.3 Conclusions on Scottish Government inputs to the herbage NL/ RL system

• The figures suggest that the value of Scottish Government interventions in herbage variety selection creates around £450,000 of value at an annual investment cost of around £159,008 (plus BioSS support). This is based on selection for herbage yield only, though this is likely to be the largest component of improvement. Persistence, nutritional quality and farm system effects have not been included.
• The total value of variety led yield improvement is very large, given the importance of sown grass swards to Scottish agriculture. The NL/ RL system is estimated to account for 75% of the improvement, if not more. This is because, unlike single variety cereal crops which are harvested and accounted once per year, it is very difficult to measure the relative performances of different herbage varieties in a mixture, especially if this is done simply through observation on commercial farms. The NL/ RL system therefore adds more value by greatly increasing the rate of improvement.
• The Scottish Government contribution to this improvement, however, is much lower (11%) because the UK wide system augmented by a single site largely funded by the remaining BSPB and Grass Levy funds could provide at least a proportion of the correct varieties (around 40% of the perennial ryegrass varieties are common to all three UK lists). However, the return on Scottish Government investment under the current regime is still good.
• Five options for change have been assessed. Relying on the rest of the UK for NL lists and only doing RL in Scotland results in a big loss of value (over £200,000 per annum). Copying Northern Ireland and having only one combined NL/ RL site produces less of a potential loss, but still suggests a reduction in benefit to Scottish agriculture of around £250,000 (a net loss once the saving in Government funding is included of around £122,000). The option of reducing the number of harvests in the trials must be considered given the nine year selection process, but the information does not exist to allow an assessment of the financial impact in this report. There is scope to review the "advice to Ministers" funding, some of which is hard to justify, and to look at the costs of the whole system, especially as the RL system seems to be a quarter of the cost of the largely Scottish Government funded NL equivalent trials. The most promising options may be to look at the sheer length of the selection process and to review the ATM support.