General enquiries on this form should be made to:Defra, Science Directorate, Management Support and Finance Team,Telephone No. 020 7238 1612E-mail: research.competitions@defra.gsi.gov.uk

SID 5 Research Project Final Report

SID 5 (2/05) Page 1 of 22

NoteIn line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The SID 5 (Research Project Final Report) is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra website. A SID 5 must be completed for all projects.

A SID 5A form must be completed where a project is paid on a monthly basis or against quarterly invoices. No SID 5A is required where payments are made at milestone points. When a SID 5A is required, no SID 5 form will be accepted without the accompanying SID 5A.

This form is in Word format and the boxes may be expanded or reduced, as appropriate.

ACCESS TO INFORMATIONThe information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors.

Project identification

1. Defra Project code IS0210

2. Project title

YIELDS OF UK CROPS AND LIVESTOCK: physiological and technological constraints, and expectations of progress to 2050.

3. Contractororganisation(s)

ADASADAS BoxworthBattlegate RoadBoxworthCambridgeCB3 8NN

54. Total Defra project costs £ 224,021

5. Project: start date................ 01 January 2003

end date................. 30 June 2005

SID 5 (2/05) Page 2 of 22

6. It is Defra’s intention to publish this form. Please confirm your agreement to do so...................................................................................YES NO (a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They

should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow.Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer.In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

(b) If you have answered NO, please explain why the Final report should not be released into public domain

Executive Summary7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the

intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work.

BackgroundYields of crops and livestock fundamentally affect government policies, environmental impacts and the economics of farming. This study provides information on how agricultural yields may change through to 2050. The rapid increases in UK farm yields over the last few decades are well recognised, but continuing increases cannot be presumed. Increases are attributed in large part to the plant and animal breeding, to better husbandry systems, and to chemical protection and nutrition. Further increases will depend on yield potentials of each farmed species in the UK environment, and the extent to which new technology can move yields towards these potentials. There have been suggestions that rates of increase in yields have slowed in recent years. This project estimates whether yield potentials are now being approached or whether there are technological constraints on progress.

ObjectivesThe aims of this Project were:

i) To form initial estimates, based on literature and data to hand, of the yield potentials of the major species farmed in the UK according to physiological and technological considerations.

ii) To collate and publish, through an International Conference, what extant views were on likely progress in on-farm yields of livestock and crops through to 2050.

iii) To review and report to Defra the best consensus on yield potentials of the major crops and livestock farmed in the UK, along with the main factors expected to affect progress in yield through to 2050.

ApproachesCase studies were carried out for nine major farm species: wheat, oilseed rape, peas, potatoes, grass, milk, beef & sheep, pigs and poultry. For each species, experts were chosen to estimate potential yield, predict progress towards this and identify the most likely constraints. Particular attention was paid to the potential negative effects of continued yield improvement. Scenarios of changes to conditions for farming in 2010, 2020 and 2050, which were developed under parallel project IS0209, were used to assess the rate of yield improvement under different futures. These studies, along with generic issues relating to yield, were presented to 60 agricultural scientists at the International Conference ‘Yields of Farmed Species’ in June 2004 held at the University of Nottingham Sutton Bonington Campus. The authors of the case studies were able to incorporate feedback arising from the main presentation and the workshop session that focussed on the detail of the calculations used. The revised case studies, representing a consensus of an international group of scientists, were then collated and integrated in this report, with predictions expressed in common terms. The conference proceedings were published as a 600 page hard-back book. The book provides the evidence and rationale for the yield predictions set out in this report; references to specific pages of the proceedings are provided throughout.

SID 5 (2/05) Page 3 of 22

Summary of Results Potential yields

It is important that units for yield are defined. In this project yields were calculated on an annual basis and where necessary accounted for animal growth rate and fecundity. The units used for crops are t/ha/yr and the units used for animal species include: milk (l/cow/yr), beef and poultry (kg live weight/animal/yr), eggs (eggs/bird/yr), pigs and sheep (kg of offspring/dam/yr). Current farm yields are significantly below the estimated potential yields. When farm yields are expressed as a proportion of the unirrigated potential the farm species are ranked as follows; eggs (0.82), poultry meat (0.81), potatoes (0.66), pigs (0.56), beef (0.53), peas (0.45), wheat (0.44), oilseed rape (0.38), sheep (0.27), grass (0.27) and milk (0.22). These estimates assume that crops will be able to use almost all of the annual rainfall. If the irrigated yield potential is used for estimating the potential yield of potatoes then the ratio of farm yield to potential yield decreases to 0.33. If potential yields could be achieved then the crop area required to produce each unit of livestock production would decrease by between two- and four-fold. Climate change is predicted to have a relatively minor effect on potential yield under rainfed conditions because positive effects of greater CO2 and, to a certain extent, higher temperatures, will largely offset effects of less summer rainfall and greater potential evapo-transpiration. In an irrigated environment the potential yield of potatoes may rise by >20% by 2050.

Constraints for achieving potential yields

Constraints for achieving potential yield can be split into 1) the development of new technology for increasing yield and 2) uptake of existing and new technology by practitioners. On balance it appears that technology uptake will be the most important constraint. For example, the gap between research yields and farm yields is widening for oilseed rape, probably because decreasing commodity prices reduce the economic optimum level of inputs, and reduce the incentive for farmers to invest in technology.

Increasing environmental regulation to reduce pollution and increase biodiversity, social concerns about technologies such as genetic modification, and refocusing market requirements towards free range animal produce and quality (e.g. potatoes) will also restrict the uptake of technologies for maximising yields. Only some of these objectives are incompatible with high yields, e.g. higher yielding animals produce less pollution per unit of produce, waste management and pollution is easier to control for housed livestock and biodiversity can be improved by careful management of non-cropped areas.

To an extent, the development of new technologies is less constrained than its uptake because some innovation takes place outside agriculture, global companies are buffered against fluctuating regional demands for their products, and research institutes and universities are often centrally funded and less influenced by market forces (at least in the short to medium term). Development of new germplasm is however a potential constraint. Plant and animal breeders have consistently increased yields for several decades but several factors now pose a threat. In particular, as yields approach species’ potentials, breeders will increasingly encounter and need to overcome traits that have negative associations with yields (crop stability, reproductive performance, metabolic stress, immunity to disease and functional fitness), whilst political pressures for greater sustainability will demand greater pest resistance and greater resource use efficiency. Diminishing returns that are ultimately linked with low commodity prices, may also slow the rate of breeding progress (as has already occurred with peas and beans).

Estimates of future yields under different scenarios

IS0209 constructed four scenarios to span the range of possible UK futures; World Market (free trade), Global Sustainability (internationally competitive, moderated by environmental compliance), National Enterprise (protected domestic markets), Local Stewardship (emphasis on environmental objectives). A future ‘business as usual’ scenario was added to represent extrapolation of recent trends. In order to assess future yields under each scenario, a matrix of indicators considered to influence farm yields was drawn up by this study and IS0209. Factors predicted to have the greatest positive effect on farm yields included commodity ‘farm gate’ prices and the cost of farm inputs (including management). The most negative factors included environmental regulation as it constrains the levels of inputs. Under the Business As Usual scenario recent yield trends are predicted to continue with wheat yields increasing at 1% per year, the carcass weights of cattle, sheep, pigs and poultry will increase by 0.5% per year, eggs/bird/year and milk/cow/year will increase at 0.5% per year and no increase will occur for oilseed rape, peas, potatoes and grass. In the short-term, yields may appear to increase as production becomes focused on high yielding land (due to decoupling of farm subsidies from production). However, in the medium term the consensus amongst experts was that environmental regulation, weakening economic signals and increasingly diverse breeding targets will slow the rate of yield improvement.

Considering other futures, greatest yield improvements are expected under the National Enterprise and World Market scenarios. Under National Enterprise, high commodity prices, medium input costs and low environmental regulation enable a high level of innovation and inputs. Predicted yield increases under this scenario by 2050 are 20 to 80%. Under a World Market scenario yield improvement of up to 60% is expected due to lower input prices, less environmental regulation and the abandonment of less productive

SID 5 (2/05) Page 4 of 22

land. Yields are expected to remain static or rise slightly under a Global Sustainability scenario and to remain static or fall by up to 25% under a Local Stewardship scenario due to the high level of regulation and lower inputs.

DiscussionOn a global scale, improvement of farm yields is just one of several means by which an increasing population may be provided with adequate nutrition, fuel, clothing and building materials. Other means include (1) extension of farming, by claiming natural habitat, (2) reducing wastage, e.g. 15% cereal production is wasted (p13), (3) reducing over consumption and (4) using more efficient species, e.g. fish are 8 times more efficient producing protein from plant material than cattle (p14). On a regional scale, IS0209 illustrated that yield estimates are a key factor for predicting land uses and environmental impacts. UK land required for agricultural production was predicted to decrease under Business As Usual, World Market and Global Sustainability scenarios. These changes were mainly due to greater yields and competition from imports. By contrast insufficient land would be available to meet demand under the Global Sustainability and Local Stewardship scenarios as a result of lower yields and greater local demand for produce. The overall environmental load produced by agriculture is predicted to be least under World Market and Business As Usual scenarios. This is due to the large amount of land released from agricultural production, although the burden is high in the areas that are farmed intensively. Yield estimates also affect the prediction of economic factors and social factors such as the amount of land available for recreation.

Implications for Defra Priorities

There is considerable scope to further increase yield of all species within the constraints of the UK environment and thereby food and fuel security could increase, imports could be reduced, and environmental costs could be curtailed. Realising potential yields would significantly reduce the amount of agricultural land required to produce each unit of crop and livestock product (by as much as 4-fold), thus land could potentially be released for non-agricultural uses. Project IS0209 predicted this would have significant and positive effects on the environment. However, under current economic and political environments it seems unlikely that the industry will be able to exploit this advantage, because high costs, low commodity prices, environmental regulation and social issues will all tend to restrict development and exploitation of new and existing technologies for increasing yields.

Future work

There is ample justification for publicly funded research to be directed towards yield improvement, in particular by facilitating the uptake and exploitation of new technology.

It will be important to devise joint strategies, involving both land uses and productivity targets, that will optimise between environmental and economic outcomes.

It is particularly important that better methods are developed to breed simultaneously for several traits, especially where these are negatively associated. This is to enable joint delivery of commercial and public agricultural goods.

Project Report to Defra8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with

details of the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: the scientific objectives as set out in the contract; the extent to which the objectives set out in the contract have been met; details of methods used and the results obtained, including statistical analysis (if appropriate); a discussion of the results and their reliability; the main implications of the findings; possible future work; and any action resulting from the research (e.g. IP, Knowledge Transfer).

SID 5 (2/05) Page 5 of 22

1. INTRODUCTION

1.1. Rationale and objectivesThe rapid increases in UK farm yields since 1945 are well recognised. They are attributed in large part to concurrent strides in science and technology. The UK now has the capacity for self-sufficiency in food. For some species, it also competes well in world trade. Yields still appear to be increasing for some major farmed species, but rates of increase have not been stable through time and have differed considerably between species. Clearly it is not safe to conclude that increases will continue for the foreseeable future. There may be absolute limits to the yield of a species, and there may be constraints set by the environments (physical, technological, sociological, economic or political) in which a species may be farmed. In many cases, yield improvements are associated with undesirable trends, for example in environmental pollution, animal welfare or CAP expenditure. As government seeks to promote sustainable development into the future, its policy determination will require good assessments of changes in yield for crops and livestock over a significant timeframe. The Scientific Objective of this project was:

To estimate the physiological potentials for yield improvement in eight major crop and livestock species farmed in the UK, to specify physiological and technological constraints on yield improvement, and provide estimates on how agriculture yields may change in future.

Technical Objectives were:

1. Based on literature and data to hand, form initial estimates of the yield potentials of the major species farmed in the UK according to physiological and technological considerations.

2. Through an International Conference, collate and publish extant views on likely progress in on-farm yields of livestock and crops through to 2050.

3. Review and report to DEFRA the best consensus on yield potentials of the major crops and livestock farmed in the UK and the main factors expected to affect progress in yield through to 2050.

1.2. ApproachesThe International conference included plenary sessions on crops and livestock, identifying generic issues, estimating potential for and limitations to genetic advance, the relevance of physiological estimates to commercial farm outputs, trait recognition, trade-offs between beneficial traits (e.g. yield & quality), impact of production on biodiversity, opportunities for design in breeding and the impacts of climate change, and considerations of technological influences on yields: molecular genetics & proteomics, nutrition, agrochemicals and pharmaceuticals, machinery, distribution and retailing environments. Case studies were made on the physiological and technological potentials of wheat, oilseed rape, peas, potatoes, grass, milk, sheep & beef, pigs and poultry. Particular attention was paid to the potential negative effects of continued yield improvement. In animals these include impacts on animal welfare, reproductive performance, the environment and the feasibility of maintaining nutrient supply and balance. In crops these include tolerance of biotic and abiotic stresses, hence requirements for pesticides, fertilisers and other resources, and impacts on pollution and biodiversity. Scenarios of changes to conditions for farming in 2010, 2020 and 2050, which were developed under parallel project IS0209, have been used to assess the rate of yield improvement under different futures. The proceedings of the conference held under this project have been published in a 600 page hard-back book (Sylvester-Bradley and Wiseman, 2005). This is an essential companion when reading this report; throughout, there are references to specific pages of the proceedings.

1.3. Definition and units of yieldYield is defined here as the amount of saleable product, expressed in relation to a constraint which the farmer has no means of altering. For crops these are time and light, and light is determined by the area of ground occupied by the crop. It is assumed that water can be supplemented if necessary. The yield of wheat, oilseed rape, peas, potatoes and grass are therefore expressed in t/ha/yr. Unalterable inputs for animal species are time and the number of animals. Animal yields are therefore expressed in kg of live weight, or l of milk, or number of eggs /animal/yr. For animal species that complete their lifecycle in less than one year (e.g. broiler chickens) the accumulated yield that could be achieved if each bird is replaced immediately after slaughter is calculated (kg/bird space/year). The annual yield of pig and lamb is dependent on the rate at which they produce offspring in addition to the rate at which an individual animal accumulates body weight. This is accounted for by expressing yield in kg of liveweight produced by all of the offspring of a single dam per year. These units were devised to enable project ISO209 to estimate the number of cropped hectares or individual cows, laying hens, broiler chicken spaces, breeding sows and ewes required to meet the market demand for each crop and animal product. Feed conversion ratios can also be used to estimate the number of crop hectares required to produce each animal product. Defra statistics express animal weights as carcass weights, however, live weights are more useful because they are compatible with published feed conversion ratios and growth rates. To enable conversion from one form of weight to another, the ratio of carcass weight to live weight is assumed to be 0.85 for poultry, 0.75 for pigs and 0.50 for

SID 5 (2/05) Page 6 of 22

beef and sheep (p496, Table 22.1). The amount of space occupied by animals is a very important welfare and economic factor. However, it is not used to express potential yield because other factors can be supplemented to achieve a particular yield for different amounts of space. The important implications for animal welfare and economics of space and animal yield are considered in the discussion.

2. YIELD TRENDS

2.1. Crop speciesOn-farm yields of wheat averaged across England and Wales have risen steadily over the last 50 years by 0.1 t/ha/year (Figure 1; Table 1). Maincrop potato yields increased from an average of 28 t/ha between 1973 and 1978 to over 40 t/ha in 1992. From 1992 onwards potato yields have remained static at 42 t/ha. There has not been a clear trend in the yield of oilseed rape and peas since government records for these species began (in 1984), with yields averaging 3.0 t/ha for oilseed rape and 3.6 t/ha for peas. Interestingly the average yield of new oilseed rape and pea varieties tested in the recommended lists have increased by 0.05 t/ha/year for both species over the same period. On-farm grass yields have been estimated at 6 t/ha (dry weight) across the country (Robson, 1981). However, regular records of the national average yield of grass have not been made. Yields of new varieties of perennial ryegrass have increased by 0.3 to 0.6% per year in recommended list trials (p550). These improvements may not have transferred to increases on-farm as illustrated by the oilseed rape and pea case studies.

Table 1. Farm yields trends and estimated potentials

For soil types with a AWC of 0.14-0.16 cm3/cm3 (sandy loams, clays) to 0.20-0.22 cm3/cm3 (loams)

Figure 1. Average crop yields for England and Wales: Defra.

SID 5 (2/05) Page 7 of 22

2.2. Animal speciesNational average animal yields reported annually by Defra include the carcass weight of cattle, pigs, lambs and poultry together with the number of eggs per hen per year and the number of litres of milk produced per cow per year. These have been reported since 1973 for milk, 1980 for cattle, and 1984 to 1986 for the other species. During these recording periods, each yield parameter has increased apart from lamb carcass weight which has remained at an average of 17.9 kg (Figure 2; Table 1; Table 2). This is more a reflection of market demand than potential for increase. Milk yields have increased by 88 l/cow/year and egg yields have increased by 2.12 eggs/bird/year. Over 101 years the rate of increase in egg yield has been 1.49 eggs/bird/year (p522). The annual increases of the carcass weights of cattle, pigs and poultry have been 2.0 kg, 0.5 kg and 0.0053 kg respectively (Figure 2; Table 2).

Trends in growth rate must be accounted for to express the yields of cattle weight and poultry weight on an annual basis. The mean live weight gain of cattle is currently about 1.0 kg/day or 365 kg of live weight /year (D. Chapple pers. comm.). Over the last decade it is estimated that the rate of live weight gain may have increased by between 0.05 and 0.10 kg/day (D. Chapple pers. comm.). Therefore the annual yield increase of cattle may be between 1.8 and 3.6 kg of live weight per year (Table 1). In 1958, broiler chickens reached a live weight of 1.9 kg at 70 days of age. By 1996 a weight of 2.2 kg was typically reached at 42 days (p522). This equates to an annual live weight yield of 19.1 kg per broiler chicken space. Historically, the annual live weight yields have increased at a rate of 0.22 kg/year (Table 1).

The annual yield of pigs and sheep is a function of the reproductive efficiency and the weight at slaughter of the offspring. The number of piglets produced per sow per year has increased at a rate of 0.19 per year to reach 21.1 in 2003 (Figure 22.7, p505). This trend is an average over 30 years and the rate of increase has been much smaller from 1990 onwards as the reduction in weaning age slowed. Live weight of pigs at slaughter has increased by 0.5 kg/year to 95 kg in 2003 (p507, Figure 22.9). Combining these trends gives an increase in the live weight/sow/year of 28 kg/year to reach 2005 kg/sow/year in 2003 (Table 1). The number of lambs produced per ewe per year has increased by 0.0075/year to 1.2 in 2003 (R. Webb, pers. Comm.). The carcass weight has not changed significantly since 1985 at 17.9 kg (Figure 2); it is estimated to correspond to a live weight of 36 kg. The live weight/ewe/year has therefore increased by 0.27kg/year to reach 43 kg/ewe/year in 2003 (Table 1). It should be recognised that carcass weights are determined by current market requirements and extrapolating current trends is not necessarily valid. Thus a continuation of the recent annual increase in live weight of pigs at slaughter is unlikely as this would take animals beyond the weight demanded by the market.

Figure 2. Average animal yields for England and Wales: Defra. 1986 yields: Beef (274 kg/carcass), sheep (17.9 kg/carcass), pigs (61.3 kg/carcass), poultry (1.44 kg/carcass), eggs (259 eggs/bird/year), milk (4970 l/cow/year).

2.2.1. Feed conversion efficiencies

Feed conversion efficiencies (MJ metabolizable energy (ME)/kg live weight gain) are estimated at 23 (p496) for poultry, 35 for pigs (p496), 70 for cattle (Anon., 1980), 62 for lambs (Anon., 1980), 0.9 MJ/egg (P523) and 5.4 MJ/l milk (B. Cottrill, pers. comm.). Feed conversion ratios for poultry and pigs given in kg feed/kg live weight gain (p496) have been converted to MJ (ME) by assuming the feed is wheat with a dry matter content of 90% and energy content expressed for 100% dry matter of 14 MJ/kg (Anon., 1980). There is evidence that these efficiencies have improved (i.e. less feed per unit yield) for some species. Estimated annual reductions in feed

SID 5 (2/05) Page 8 of 22

conversion efficiency include 0.05 MJ/kg for poultry (p523), 0.064 MJ/kg for pigs (p508, Figure 22.10) and 0.028 MJ/kg for sheep (R. Webb, pers. Comm.). These improvements must be interpreted with caution because they may have arisen through improved feed technology rather than through an improvement in animal physiology, e.g. use of enzymes in pig and poultry diets, improvements in silage making for ruminants. Therefore lesser amounts of feed have been required to provide the same amount of energy/digestible nutrients. Evidence for this can be found for pigs (p508, Figure 22.11) for which feed intake (kg) has decreased by about 17% between 1982 and 2000. Evidence of a trend in feed conversion ratio is even more difficult to find for cattle due to the wide range of feed stuffs used.

Table 2. Parameters for animal yields

ME – metabolisable energy Extrapolation of current trend to 2050

3. POTENTIAL YIELDS

Potential yield of crop and livestock species is defined here as the yield that could be produced by combination of the best germplasm from within a species, with the best management, on deep soil (>1.5m) and in an environment with the current average radiation, temperature and rainfall for the UK. Potential yields of crops have been estimated for both rainfed and irrigated environments. Use of molecular genetic modification is assumed, but not to the extent that the fundamental characteristics of the speciesis changed, e.g. C4 metabolism cannot be introduced into C3 plants. The effect of climate change is considered in section 3.3.

3.1. Crop speciesPotential yields of crop species have been estimated for the average UK environment (with and without irrigation) from estimates of the longest possible yield forming period and the maximum photosynthetic efficiency. Estimates of these parameters were based on physiological principles wherever possible.

Following the principles of Monteith (1977) primary biomass production (Pn) is determined by equation 1, where St is the integral of incident solar radiation (MJ m-2) during the growth of the harvested organs, i the efficiency with which that radiation is intercepted by the crop; c the efficiency with which the intercepted radiation is converted into biomass (p167).

(1)

Saleable, yield (Y) may be calculated from the primary biomass production (Pn), the efficiency with which biomass is partitioned into the harvested product () (p167), plus any biomass produced before the growth of the harvested organs which contributes to yield (w).

(2)

3.1.1. Radiation

If we assume that incident radiation for the UK will not change over the next few decades then historical radiation data may be used to define the radiation environment for the production of the potential yield. A cosine function predicts the daily incident solar radiation (S) for the mean of five sites in east, south and middle England, Wales and Scotland between 1989 and 1994.

SID 5 (2/05) Page 9 of 22

(3)

Where J is the Julian day, a = 1.8 (vertical location), b =8.5 (amplitude), c = 58.0 (frequency) and d = 6.5 (horizontal location). Integrating this equation then enables the total incident solar radiation between Jx and Jy to be calculated, where Jx and Jy are the first and last Julian days of any particular time period.

(4)

It is straightforward to estimate the amount of radiation intercepted when the canopy size is relatively constant, as is the case during the majority of the seed filling periods. The rate of canopy expansion must be accounted for when estimating light interception during earlier growth periods. The mean GAI during this period can be used to provide an adequate estimate of the amount of light intercepted.

3.1.2. Crop parameters

Physiological principles have been used to estimate the extent to which the key yield influencing parameters could be improved within the UK environment (Table 3). The extent to which the yield forming period can be brought forward for wheat, oilseed rape, peas and potatoes is limited by the risk of frost damage. Greater scope for improvement lies in prolonging the period of seed filling. It is predicted that the seed filling period may be increased by 8 to 13 days for wheat (P247), oilseed rape (p322) and peas (P281). A greater increase of 35 days may be assumed for potatoes if the current practice of desiccating the crop is not carried out. Current crops are estimated to intercept between 75 and 100% of incident light during the seed filling period, so there is no major scope for increasing the proportion of light intercepted. Grass is assumed to intercept 100% of the incoming radiation throughout the year.

Maximum potential radiation use efficiency (RUE) is difficult to estimate because measurements vary widely between experiments. Differences appear to be caused by environmental factors rather than methods of measurement (Sinclair et al., 1999). Measured values are considerably less than both the theoretical maximum and the maximum measured under controlled conditions, indicating that there is scope for improvement. Several routes for improvement have been described (p91-92, p173-186). Many of these involve insertion of genes from other species and must be regarded as long-term objectives. To date, breeding improvements in RUE have only been reported for wheat (p245). Accordingly, the potential RUE for wheat over the whole growing season is predicted to be slightly greater than current values at 1.4 g/MJ. RUE for peas, potatoes and oilseed rape is taken as the highest RUE that has been recorded in NW Europe; with peas at 1.46 g/MJ (p268), potatoes at 1.80 g/MJ (p296) and oilseed rape at 1.70 g/MJ pre-anthesis (Rao et al., 1991) and 0.86 g/MJ post-anthesis (Dreccer et al., 2000). Variation in RUE between species is determined to an extent by variation in the energy content of plant mass which depends upon the content of carbohydrate, protein and lipids. According to the energy contents of these constituents (Roberts et al., 1993) and the chemical compositions of the organs of different plant species (Sinclair and de Wit, 1975) it can be estimated that wheat seed has a similar energy content to vegetative biomass at 0.0170 MJ g-1, carbohydrate rich organs such as potatoes have an energy content of 0.0145 MJ g -1, pea seeds have 0.0186 MJ g-1 and lipid rich oilseed rape seeds have 0.0247 MJ g-1. It therefore seems likely that the large RUE for potatoes is due to the high proportion of carbohydrate within its tissues. Oilseed rape probably has a low RUE post-anthesis due to the high energy content of oilseeds and the low photosynthetic rate of pods (Gammelvind et al., 1996).

Potential dry matter yield of forage grasses in the UK has been estimated at 25-30 t/ha (p547; Alberda, 1977; Cooper, 1970; Wilson, 1982). Annual production of 23 t/ha has been achieved at the field scale (Adams et al., 1983). This compares with average farm dry matter yield, excluding rough grazing, of only 6 t/ha (Robson, 1981). Yields with best agricultural practice are estimated at 15 t/ha (p547). Potential RUE is taken to be 1.36 g/MJ, assuming the main period of growth is between 1 March and 31 October, zero light is intercepted for a seven day period immediately after two of the harvests (the third harvest is assumed to occur on 31 October) and a harvest index of 0.7. The potential yield of grazed grass is lower than conserved grass due to fouling and treading by livestock.

Currently only wheat yields benefit from the relocation of biomass formed before the harvested organs. The mechanism of relocating biomass is likely to be complex and it seems unlikely that it can be introduced into other crop species without major genetic modification. Therefore, the potential yields of peas, oilseed rape and potatoes do not include contributions from earlier formed biomass. In wheat, calculations of w for potential crops has been based on the total biomass accumulated prior to the yield formation period minus the biomass required for leaves and to physically support the potential yield against lodging. These calculations indicate that the translocatable dry matter could be increased from 300 g/m2 to 550 g/m2 (p247).

SID 5 (2/05) Page 10 of 22

Table 3. Parameters for potential crop yields.

start of rapid growth denoted by beginning of stem extension in wheat, oilseed rape and peas or emergence in potatoesYFP – Yield forming period pre-anthesis / post anthesis RUEa Radiation use efficiency for irrigated crops

3.1.3. Potential yields

Irrigated

The assumptions described in the previous paragraphs give potential yields of 19.2 t/ha for wheat, 7.93 t/ha for oilseed rape, 8.36 t/ha for peas, 144 t/ha for potatoes and 30 t/ha for forage grass (Table 1). These potentials are more than double current UK farm averages for the arable species.

Record yields can exceed these estimates of potential average yields because environments, especially weather, may have been exceptional. Their value is in demonstrating lack of any insuperable limits to yield of a species. The official UK record yield for unirrigated wheat is 14.0 t/ha and yields of >16 t/ha have been observed on yield maps over smaller areas in the UK (and on a field-scale in New Zealand). Unirrigated oilseed rape and pea yields of greater than 7 t/ha have been recorded in the UK. Irrigated potatoes have yielded >80 t/ha on UK farms and 100 t/ha on UK experimental plots (130 t/ha on experimental plots in the USA). Maximum grass yields of 15t/ha have been reported using best agricultural practice (p547). Irrigated experimental plots in Northern Ireland have recorded grass yields of 23 t/ha (Adams et al., 1982). These record yields are 60 to 75% greater than the average farm yields of potatoes and wheat, and >100% greater than the average farm yields of peas, oilseed rape and grass.

Rainfed

The amount of water available to the growing crop is determined by the available water capacity (AWC) of the soil at the onset of rapid growth and the amount of rain during the growth period. The amount of available water stored in the soil is dependent on soil type with sandy loams, clays and silt clays having an AWC of 0.14-0.16 cm3/cm3 and loams having an AWC of 0.20-0.22 cm3/cm3 (Rowell, 1994). Sandy loams, clays and silt clays make up approximately 50% of the arable soils in England and Wales (Clark, 1999). Rooting depths have been measured at between 1.5 and 1.8 m for wheat and oilseed rape (Barraclough and Leigh, 1984; Barraclough, 1989), 0.6 and 1.0 m for potatoes (Burton, 1966; Hebblethwaite et al., 1985), 0.7 and 1.2 m for peas (Hebblethwaite et al., 1985; Summerfield and Roberts, 1985), and 1.0 and 1.4 m for grass (J. King pers. comm; Troughton, 1957). The proportion of available water that is accessed by the roots has been measured at 0.65-0.68 for wheat and oilseed rape crops to a depth of 1.5 m (Barraclough and Leigh, 1984; Barraclough, 1989). We assume that potential yielding crops will extract 80% of available water to their potential rooting depth, taken as 1.5 m for wheat and oilseed rape, 1.4 m for grass, and 1.2 m for peas and 1.0 m for potatoes. Average UK rainfall of 1.81 mm/day is then used to estimate the water supply from direct rainfall during the period of growth. This approach has been used to estimate the maximum amount of water available for crop uptake between the onset of rapid growth in the spring and the completion of the main growth phase in the summer or autumn. It has been assumed that the soil is at field capacity at the onset of rapid growth in the spring.

The crop’s demand for water is estimated by assuming that 1 litre of water is transpired during the production of 5 g of vegetative biomass (p238). The different energy contents of non-vegetative organs means that the water use efficiencies of potato tubers, pea seeds and oilseeds are 5.86, 4.57 and 3.45 g/l respectively. These figures are used with previous calculations of the amount of biomass accumulated from the onset of rapid growth in the spring to estimate the yield that could be accumulated before the supply of water is exhausted. This indicates that the availability of water will prevent the potential yield of all crops from being reached on both high and low AWC soils apart from the oilseed rape yield on high AWC soils (Table 1). On low AWC soils, potential average yield of wheat is predicted to be 14.0 t/ha, oilseed rape 4.0 t/ha, peas 5.5 t/ha, potatoes 62 t/ha and grass 20 t/ha. On high AWC soils, wheat is predicted to yield 18.3 t/ha, peas 8.1 t/ha, potatoes 83 t/ha and grass 23 t/ha. There appears to be sufficient time to fully recharge the soil profile between the end of growth and the start of the rapid phase of growth the following season for any combination of crop species apart from when grass or oilseed rape are grown after grass on a water retentive soil type. It should be noted that reducing the RUE in water limited scenarios may increase yield potential by reducing water use before the yield forming period. Oilseed rape may benefit from this

SID 5 (2/05) Page 11 of 22

strategy most, e.g. reducing RUE by 15% is estimated to increase potential yield on low AWC soils from 4.0 t/ha to 6.5 t/ha. However, this strategy may result in insufficient biomass to physically support the yield.

3.2. Livestock speciesYield potential for animal species can largely be regarded as genetic potential because the environment can be optimised through controlled housing, nutrition, health and management, although there is considerable variation in output depending on the extent to which production systems are controlled through some or all of these factors. An example of this variation would be intensive lowland lamb production vs hill sheep. Long-term breeding experiments have shown continued improvement of traits over many generations (p126-129). For example, continued responses in the body weight of quail over 90 generations and the body weight of mice over more than 110 generations have been reported. Yield plateaux were found in these experiments, but these were broken with continued selection. Traits which do not appear to show long-term improvements include the speed of race horses and greyhounds. It is concluded that there is still potential for improving traits associated with yield and it will be very difficult to quantify the genetic limit due to the uncertain nature of mutation. Therefore, record yields and historic yield trends are used to gauge where genetic potential may lie. Current trends have been extrapolated to 2050 to provide an estimate of potential yield for yield parameters without an estimate of potential. It has been impossible to gauge whether feed conversion efficiencies have improved and very little information exists about the potential for improvement particularly in terms of ME MJ. Potential feed conversion efficiencies are therefore assumed to be the same as current efficiencies (Table 2).

3.2.1. Cattle

Milk yields by individual cows have been recorded at 14129 l/year as early as 1967 (p439) and the best Holstein cow in the USA produces 30000 l/year with a herd average of 15000 l/cow/yr (p353). The potential herd average is estimated here at 30000 l/cow/year. This is more than four times the current UK average and demonstrates the high potential yield for milk production. The carcass weight of beef cattle currently averages 315 kg and has been increasing at a rate of 2 kg per year (Table 2). Carcass weights are about 50% of the weight at slaughter. Beef cattle seldom reach their maximum potential weight because they must be slaughtered earlier to ensure that they achieve the confirmation that the market requires. The potential for increasing cattle size would appear to be very large; historical accounts of breeding oversized black and white oxen show that live weights of over 2000 kg are possible (p457). Beef cattle take about 18 months to reach the optimum conformation at an approximate live weight gain of 1 kg/day. Thus the live weight gain of cattle is about 365 kg/year. The best recorded rates of live weight gain over the whole lifecycle are between 1.8 and 1.9 kg/day (D. Chapple pers. comm.). If the potential live weight gain is achieved then the live weight of an animal could increase by 694 kg/year.

3.2.2. Sheep

The biological ceiling for production is 8 lambs per year (over 2 litters). However, a more feasible potential is for a ewe to produce 3 litters every two years with 3 lambs in each litter (p473) to give 4.5 lambs per year (Table 2). Lamb carcass weight at slaughter has remained relatively constant since the late 1970s at 17.9 kg as this is the optimum size for the UK market. If this is assumed to remain constant then the potential productivity per ewe is 79 kg of lamb carcass weight per year (Table 2).

3.2.3. Pigs

Pig productivity is a function of the number of litters per year per sow, the litter size, the pig weight at slaughter and the daily live weight gain. In 2002, top herds achieved 25.5 pigs per sow per year (based on 2-3 litters per year) (p502). Historic increases in litter number and size appear to have reached a plateaux, however it is possible that the potential yield could be 30 pigs per sow per year. Pig live weight at slaughter is currently 95 kg and has been increasing by 0.5 kg/year (p507). If this trend continues then pig weight at slaughter will increase to 120 kg by 2050 although it is the market which will determine whether this will be achieved. As castration is not permitted in animals entering meat quality chains, it is very unlikely that entire males will be taken to anywhere near this weight. Daily live weight gain is currently 650 g increasing from 560 g since the early 1980s (p507). Individually housed pigs in University of Nottingham pig trials regularly achieve over 950 g (p513). If historic trends are extrapolated to 2050 then the daily live weight gain would be in the region of 1400 g. Thus the potential productivity of a single sow could be as much as 3600 kg/year. Feed conversion efficiency is currently 35 MJ ME/kg liveweight gain. This equates to 9000 kg feed per sow and offspring per year.

SID 5 (2/05) Page 12 of 22

3.2.4. Poultry

Individual birds have been recorded as producing 361 eggs in 364 days (p523). Experiments in the 1980s with continuous lighting or shortened (<24hr) lighting cycles have indicated that the rate of egg production could be greater than one per day (p128). Thus it may be possible to achieve >365 eggs/bird/year, however the ability to digest, absorb and utilise nutrients, and the ability to transfer them to the oviduct, may become a constraint. Therefore the ultimate yield is estimated at 365 eggs per bird/year. It is estimated that a broiler chicken could achieve 2.2 kg in 34 days (p522). This equates to 23.6 kg/bird space/year assuming that a bird is replaced immediately after slaughter (Table 2).

3.2.5. Animal yields expressed in terms of crop area

Animal production can be expressed in terms of the crop area required from the feed conversion ratio (MJ/kg) (Table 2), the energy content of the feed (MJ/kg) and the crop yield (t/ha) (Table 1). Feed is assumed to be wheat (14 MJ/kg dry weight) or grass silage (10.9 MJ/kg dry weight) (The UK Tables of Nutritive Value and Composition (1990)). The area of wheat currently required to produce 1 kg of live weight is estimated to range from 2.4 m2 for poultry to 7.4 m2 for beef (Table 4). The area of wheat required to produce one litre of milk is estimated at 0.57 m 2

and the area for one egg is 0.095 m2. The area of land required almost doubles when grass is used as the main feed stuff due to its lower yield and energy content.

The minimum crop area required to produce livestock can be estimated from the potential crop yields. Potential feed conversion efficiencies are assumed to remain the same as current values. The minimum area of wheat required to produce 1 kg of live weight is estimated to more than halve from current values to 1.1 m2 for poultry to 3.2 m2 for beef (Table 4). The area of wheat required to produce one litre of milk is estimated at 0.25 m 2 and the area for one egg is 0.04 m2. A similar requirement for land area is estimated for grass due to its higher potential yield and lower energy content compared with wheat. NB greater reductions in the area of crop required could be achieved if feed conversion efficiencies can be improved.

Table 4. Crop area required for livestock produce

3.3. Effects of climate change on potential yieldPredictions of climate change in the UK over the next 50 years have been made under UKCIP98 (Hulme and Jenkins, 1998). UKCIP produces four scenarios which reflect the range of uncertainty about the sensitivity of the atmospheric system to climate change and future emission levels of greenhouse gases. Defra project CC0333 used the UKCIP98 Medium High scenario to examine the impact of climate changes on crop yields and available work time. The Medium High scenario assumes a 1% per annum increase in carbon dioxide. This emission scenario is similar to the IS92a “business-as-usual” scenario produced by the IPCC (IPCC, 1996). This scenario provides a warming rate of 0.27°C/decade, with greater warming in winter than summer. For crop production the significant climate changes by the 2050s that were estimated include:

A reduction in summer precipitation of between 0 and 20% An increase in winter and autumn precipitation of 9-13% An increase in summer potential evapotranspiration of 6-17% A reduction in frost days of around 70% An increase of around 300% in the number of days with maximum temperatures greater than 25°C An increase of incident solar radiation in summer of 3-8% and in autumn ~4%. An increase of atmospheric CO2 of 66% on present day levels

It should be noted that the climate change estimates have been updated in Hulme and Jenkins (2002) since the publication of CC0333, however the directions of the changes have remained the same.

Climate change is expected to have a significant impact upon the potential yields of crop species and only a minor influence on the potential yield of animal species via increasing the likelihood of heat stress for animals kept outside (p150). Predicting the effects of these climate changes on the potential yields of crops is not straightforward because some climatic factors increase yield whereas other factors decrease yield, and the effects depend strongly

SID 5 (2/05) Page 13 of 22

on crop species. The complexity of the problem can be illustrated by first assessing the effects of changing individual climatic factors on potential yield independently of the other factors. Doubling CO2 increases the biomass production of C3 plants by 30% and increases water use efficiency (p146). The effect of CO 2 enrichment on C4 plant species is much less than for C3 plants. A 0.1% increase in yield for every 1 ppmv increase in CO2 derived from the estimates of Kimball (1983), was used to calculate the CO2 fertilisation factor for the crops for Defra Project CC0333. Reduced summer rainfall and potential evapotranspiration (PE) will decrease potential yield in rainfed environments. Given sufficient water, higher temperatures are predicted to increase respiratory losses and increase the growing period of indeterminate crops such as grass and certain potato varieties. The effect on the length of the growing season of determinate crops is complex. The yield forming period of crops such as wheat, oilseed rape and peas occurs for a defined thermal period. Therefore, the period of yield formation could be shorter within a warmer climate and yields would be smaller. Relationships given by Monteith (1981) showed that UK wheat yields had an inverse relationship with mean temperature during grain filling. However, this effect may be countered by bringing the seed filling period forward, usually into brighter conditions, in response to the lower risk of frost damage. The greater radiation predicted under climate would additionally increase yields.

The combined effect of these factors on the potential yield of crops depends on whether irrigation can be used. In a rainfed environment, water is predicted to be the main constraint on potential yield within the current climate. Climate change that reduces summer rainfall by 10% and increases summer PE by 10% is estimated to reduce the available water for the various crop species by between 20 and 40 mm. The negative effects of this on yield are expected to be offset by the greater water use efficiency resulting in a relatively small effect on yield. In an irrigated environment, the potential yields of grass and potatoes are likely to be increased due to the longer growing season, greater CO2 and radiation, which are expected to outweigh the extra respiratory costs, i.e. increasing CO 2 by 66% is estimated to increase yield by about 20%. Smaller increases are expected for the potential yields of wheat, oilseed rape and peas due to the complex effects of temperature. NB the yield predictions described here are potentials so differ from the farm yield predictions given in Project CC0333.

4. IDENTIFICATION OF THE MAIN CONSTRAINTS FOR ACHIEVING POTENTIAL YIELDS ON-FARM.

There is a hierarchy of exploitation of potential productivity as shown in Figure 3 (p609). The maximum technological potential is set by the bio-physical limits of the species (quantified in Section 3). The existing technological boundary of productivity is evident in the trials of ‘researchers’. There is often a small but recognisable gap between the yields of research trials and those achieved by the best farm practitioners which can be attributable to a ‘research attention’ bias. For a variety of reasons farmers may not be able, or may choose not to take up the potential benefits of well established and new technologies: there is often a considerable ‘systems gap’ between the productivity of the most efficient farm (or group of farms) and that of the mean or typical farm. Figure 3 illustrates the two major possible constraints for achieving potential yield: (1) technological innovation and (2) uptake of new and existing technology.

Figure 3. Exploitation of technological potential (p609)

4.1. Uptake of new and existing technologyThis represents the gap between the average farm yield and what can be achieved if all available technology is exploited. This includes technology which may not be available to the farmer due to legislation such as high use of

SID 5 (2/05) Page 14 of 22

nutrients. The ‘systems gap’ resulting from not taking up new technology may be due to a range of economic, social and environmental factors associated for example with farmer knowledge, motivation, resources, and perceptions of the potential advantage and suitability of the technology. It is often the case that when farmers are under pressure, e.g. when product prices are low, the systems gap widens (609).

Current constraints identified for crops include; supplying enough nitrogen for bread making wheats (p252); reducing soil compaction, optimising nitrogen and irrigation, avoiding potato cyst nematode and late blight on potatoes (p297-306); optimising sulphur nutrition and minimising disease and soil compaction for peas (p275-280) and oilseed rape (p318-320); providing sufficient N fertiliser and avoiding pests and diseases in grass (p554). Potato yields are particularly restricted by the greater focus on achieving the correct quality. It has not been possible to prioritise these constraints and it seems likely that the most important constraint will depend on the farm. Most of these constraints arise from low product prices which reduce the economically optimal level of input. In future, it seems likely that low product prices, together with restricted water availability, environmental regulation to both reduce nutrient pollution and increase biodiversity, will further reduce inputs below levels required to maximise yield. Of these, restricted water may be the greatest constraint if crop irrigation is heavily regulated. The absence of irrigation is predicted to reduce the potential yields on all of the crop species studied by between 30% and 50% (Table 3). Climate change is predicted to exacerbate this problem by reducing summer rainfall and increasing PE. Restriction on nutrient use is also predicted to significantly restrict yield improvement (p252). Methods for improving biodiversity could cause smaller effects on yield because improvements can be made through appropriate management of non-crop areas (such as the headlands, and ‘skylark scrapes’), and by including a greater mix of crops in the rotation (e.g. spring and winter sown) (p345). Stronger adaptation involving wholesale reduction in herbicides and nutrition (particularly on grass) will cause greater yield reductions. Other constraints caused by climate change may include increased spatial and temporal distribution of existing pests/diseases, as well as occurrence of new ones, and by reductions in the number of available work days in autumn, due to greater autumn rainfall (p150; Defra Project Report CC0333).

The shift towards free range stocking is an example where some technology is not being exploited for animal production. This is predicted to restrict poultry and pig yields by reducing the quality of the environment (light, temperature, ventilation) and the quality of diet through greater herbage intake (p537). Welfare concerns, together with high costs and complex anatomy, are predicted to limit the extent to which in vitro fertilisation and cloning can be exploited in sheep to improve fertility and quality of offspring (p487). Legislation to increase the weaning age of piglets to 28 days (and probably older) will constrain the number of pigs that can be produced per sow per year to below the genetic potential (p515). Disease constraints include anthelmintic resistance in sheep and the removal of antibiotic growth promoters the poultry (p534) and other livestock. Commercial stocking density reduces the yield per animal in pigs and poultry. However, legislation also means that stocking densities are below that required to optimise productivity per unit of space (p515). Pollution legislation already restricts the amount of nutrients (eg copper and zinc) that may be fed and is predicted to reduce the amount of nutrients excreted to reduce the amount of nutrients than can be fed. This may affect outdoor systems most, since nutrition can be most easily managed and waste can be stored in indoor systems (p397). It should also be noted that higher yielding animals produce less pollution per unit of yield.

4.2. Technological innovationTechnological innovation includes the development of improved genotypes, new chemistry (e.g. pesticides and fertilisers), new mechanisation / engineering (e.g. to improve recovery of yield, or to help deliver inputs more accurately (p44)) and greater understanding of the processes of yield formation. Improved genotypes are probably the most important and have previously been estimated to account for at least 50% of yield improvements in broiler chickens (p126) and wheat. Environmental regulation and negative impacts of high yields may deflect breeding priorities away from increasing yield. For example genetic resistance to pests may be sought to reduce the use of chemical pesticides, greater nutrient use efficiency may be sought to reduce pollution, and social issues may reduce the ability to exploit genetic modification (p228). Breeding for high yield may increasingly encounter adverse correlations, for example with crop stability (p243) or reproductive performance, metabolic stress, immunity to disease and functional fitness (leg weakness and gait disorders) (p383-389).

It is concluded that increasing yield is unlikely to impose animal welfare problems because methods such as multi-trait selection, changing husbandry, housing and nutrient allowances can be used to overcome them (p389). For example, mult-trait selection has already been successfully employed in broiler chickens to reduce tibial dyschondrioplasia and ascites. However, ensuring welfare is likely to slow yield improvement. Diminishing returns may reduce the number of breeding companies which is likely to reduce the rate of yield improvement (p227). This appears to be the case with peas and beans in the UK (p265).

The chemical crop protection (CP) industry has undergone significant consolidation in recent years, but despite this the number of new CP chemicals has not diminished and looks set to be maintained over the next few years (p203). Environmental concerns over chemical use may be offset by the fact that new chemicals are used at much lower rates and are often more specific (p203). New technology has also increased the efficiency of inventing and screening for new chemicals (p205). Perhaps the main threats to the development of new chemicals is GM technology which has resulted in the more widespread use of cheaper broad spectrum chemicals and lower

SID 5 (2/05) Page 15 of 22

commodity prices, both of which have been linked with a reduction in the market value of CP chemicals (p196). Mechanical innovation frequently takes place outside agriculture (e.g. satellite technology for precision farming).

Overall, technological innovations that increase yields on farm usually incur expense, hence they commonly depend on product price. Rate of innovation will be stimulated if demand, hence price, increases. In addition, technological innovation will become easier if farms enlarge, or if markets become simpler. Technological innovations will also often depend on increased understanding of yield formation, hence on continued investment in research. If markets are less regulated, short- and medium-term dynamics of stimuli affecting the innovation systems in agriculture are such that there may be periods when progress is constrained by lack of personnel with appropriate training.

5. ESTIMATES OF YIELDS IN 2010, 2020 AND 2050 UNDER DIFFERENT SCENARIOS

5.1. Agricultural ScenariosProject IS0209 constructed future agricultural and related environmental scenarios by drawing on the methodology developed for the UK Foresight programme (Berhout and Hertin, 2002; DTI, 1999, 2002) which considers long term futures and possible implications for UK industry and society. Scenarios are not intended to predict the future, but rather to help think about how it might turn out. Scenario analysis can help to map out the features and consequences of possible futures and how decisions made in the interim might help shape or better cope with possible outcomes.

Scenarios have been generated as a consequence of modelling drivers of economic and social change, new trends and innovation, and of unexpected events. They are usually made up of a qualitative story-line and a set of quantitative indicators. The Foresight Futures framework was used to construct four possible scenarios, distinguished in terms of social values and governance (Figure 4). Details about how the scenarios were constructed are given in pp577-582. Characteristics of the scenarios include:

World Markets (WM): high economic growth, an emphasis on private consumption and a highly developed and integrated world trading system. Average incomes increase, but with marked disparities in distribution. Regulation focuses on supporting trade.

Global Sustainability (GS): (also referred to as Global Responsibility): pronounced social and ecological values, evident in global institutions and trading systems. There is collective action to address social and environmental issues. Growth is slower but more equitably distributed compared to the World Markets scenario.

National Enterprise: (NE) emphasis on private consumption but with decisions made at national and regional level to reflect local priorities and interests. Although market values dominate, this is within national/regional boundaries.

Local Stewardship: (LS) strong local or regional governments which emphasise social values, encouraging self-reliance, self sufficiency and conservation of natural resources and the environment.

Figure 4. Possible futures (p618)

5.2. Factors influencing exploitation of yield potential of cropsIn order to assess future yields under each scenario, a matrix of indicators considered to influence farm yields was drawn up by ISO209 and presented to a panel of crop and livestock specialists as part of this project (Table 5). The specialists validated of the relative weights given to the indicators for each scenario. A future ‘business as usual’ scenario was added, to represent extrapolation of recent trends.

Commodity ‘farm gate’ prices, as determined by the interaction of market demand and supply and interventions by government, are perceived to be strongly positively correlated with on-farm crop yields. In some cases, however, markets may offer greater rewards for products which are differentiated in terms of quality, placing less emphasis on

SID 5 (2/05) Page 16 of 22

high yields. The application of production inputs is also strongly associated with yields. These include a varied array of physical and knowledge-based inputs and processes, such as crop and livestock genetics, crop nutrition and crop protection, mechanisation, irrigation and general levels of farm husbandry. Other positive drivers include area based payments linked to particular crop types (although single farm payment regimes attempt to break this link), farm size in that larger farms exhibit greater take-up of technological possibilities, and genetic modification which may be able to enhance yield potential or overcome constraints.

Factors perceived to be negatively correlated with yield include environmental regulation as it constrains the levels of inputs. The adoption of organic production techniques is similarly perceived to be associated with lower yields, partly as a result of reduced use of artificial inputs. Business uncertainty, particularly linked with unpredictable variation in prices of outputs and inputs and outcomes of farm business decisions is also likely to influence take up of yield benefits negatively. For the most part farmers are risk averse, and business risk tends to reduce willingness to invest in new technology.

Table 5 describes the present situation and possible future scenarios in terms of relative (along the row) values of these selected indicators, including a Business as Usual (BAU) case which extrapolates recent trends. For example, farmgate output prices are perceived to be relatively high under the protected National Enterprise (NE) and Local Stewardship (LS) scenarios, and relatively low under the Business As Usual case and World Markets as agriculture is exposed to international competition. Global Sustainability (GS) demonstrates moderate farm gate prices, partly reflecting increased supply costs associated with greater regulation.

Table 5. Relative Values of Factors Influencing Exploitation of Yield Potential by Future Scenario

Notes to Table 5 : Relationship with yields + positive correlation, - negative correlation, weighted by strength of association. Relative value of parameter amongst scenarios: 0 = not applicable or zero, L = low, M = medium, H = high. *Organics: under WM organic farming is low as a % of total crop production but there is an important market in differentiated organic products compared to business as usual, driven by concern about food quality and facilitated by high incomes. Under LS, food production using organic methods is a common feature accounting for a relatively high % of total cropping.

5.3. Estimates of realisable yield by future scenarioTable 5 was used to derive estimates of the relative magnitude of crop yields on farms under each of the scenarios for the years 2012, 2025 and 2050 drawing on the panel of crop and livestock specialists engaged in the project. Future yields were expressed as an index of the current (2000-2003) average farm yields, bearing in mind the genetic potential of each species. Panel members were asked to consider the main factors affecting yields under each scenario and the main uncertainties affecting estimation. The derived estimates are given in Table 6 for crops and Table 7 for animals. Two messages emerge: there is a considerable difference amongst scenarios in terms of perceived exploitation of yield potential for a given species, and there are differences in yield take-up amongst species for any one scenario.

5.3.1. Crop species

Under the BAU scenario current yield trends are expected to continue with wheat yields continuing to grow at 0.1 t/ha/yr to reach 13 t/ha by 2050 and the yield of other species remaining static, or increasing very slightly, to give predicted yields of 3.2 t/ha for oilseed rape, 4 t/ha for peas and 42 t/ha for potatoes. For grass, farm yield trends have not been quantified and genetic improvements have been estimated at 0.3 to 0.6% per year for perennial ryegrass. We assume that these genetic improvements have not been transferred to the farm in recent years due to the lower commodity prices and higher input prices. Failure to transfer genetic gains to farm yields over recent years have been demonstrated for oilseed rape and peas. The yield of grass is therefore predicted to remain at 6 t/ha under BAU.

The greatest yield improvements are expected under the PE scenario due to high commodity prices together with medium input costs and low environmental regulation which will enable a high level of inputs. All of the breeding

SID 5 (2/05) Page 17 of 22

improvement is expected to be transferred to farm yields under this scenario. Breeding improvements are currently estimated at 0.1 t/ha/yr for wheat, 0.05 t/ha/yr for oilseed rape, 0.05 t/ha for peas and 0.006%/yr for grass. Therefore, yields are expected to reach 13 t/ha for wheat, 5.7 t/ha for oilseed rape, 6.1 t/ha for peas and 7.8 t/ha for grass by 2050. A more modest increase in potato yields to 50 t/ha is expected due to the greater focus on quality compared with yield.

The second greatest yield improvements are expected under the WM scenario due to lower input prices, less environmental regulation and abandonment of less productive land. GM technology may also be exploited in the long term. Commodity prices and are predicted to be lower and more variable under this scenario which is predicted to reduce the level of inputs to below PE. It is estimated that the inputs and management will be high enough to realise about half of the genetic improvement quantified above. By 2050, predicted yields are 12 t/ha for wheat, 4.5 t/ha for oilseed rape, 46 t/ha for potatoes, 4.7 t/ha for peas and 6.9 t/ha for grass.

The GS scenario is predicted to have medium inputs, but a high level of environmental regulation is likely to restrict optimum management, e.g. lower yielding spring wheat and oilseed rape crops may be grown. Crop yields are unlikely to rise significantly under this scenario and may even decline in the medium to long term for crops which rely heavily on inputs such as potatoes.

The LS scenario is expected to have the least intensive inputs and farm yields will probably fall in this scenario. Yields of wheat, oilseed rape and potatoes are predicted to be intermediate of those produced by integrated farming methods (80-90% of conventional) and organic methods (50-60% of conventional). Yield reductions are predicted to be less for grass because inputs to this crop are already low relative to arable crops. Pea yields are predicted to increase slightly in this scenario due to strong relative prices (high protein crop), the potential to fix nitrogen and the important contribution to crop rotations. By 2050, yields are predicted to be 6 t/ha for wheat, 2.4 t/ha for oilseed rape, 34 t/ha for potatoes, 4.3 t/ha for peas and 5.7 t/ha for grass.

Table 6. Indices of estimated crop yields by future scenario (current yields = 100)

*current average farm yields shown in brackets

5.3.2. Animal species

The yield units investigated for the animal species included the carcass weight of cattle, pigs, sheep, poultry and the production of milk and eggs per animal per year. We expect that predictions for other units (e.g. kg live weight per dam per year) will be similar to kg carcass weight per individual animal. Under the BAU scenario, the current yield trends are expected to continue with yield increases of 20-25% by 2050. Sheep yields are expected to rise, despite being static in previous years, due to the uptake of new technologies such as sire referencing schemes. The greatest yield increases are predicted under the WM scenario due to the lack of constraints and the low input costs. Under this scenario, yield increases of 65% for milk yields and between 30 and 35% for beef, sheep, pig and poultry yields are predicted. More constraints and greater input costs meant that the GS and PE scenarios are predicted to have smaller yield increases than the WM scenario. Yield increases for the GS and PE scenarios by 2050 are predicted to be 30 to 35% for milk and 20 to 25% for the other species. Under the LS scenario yields of milk and beef are predicted to decrease by 5%, and yields of sheep, pig and poultry are expected to increase by 5-15%.

SID 5 (2/05) Page 18 of 22

Table 7. Indices of estimated animal yields by future scenario (current yields = 100)

*current average farm yields shown in brackets

KEY FINDINGS When farm yields are expressed as a proportion of the unirrigated potential the farm species are ranked as

follows; eggs (0.82), poultry meat (0.81), potatoes (0.66), pigs (0.56), beef (0.53), peas (0.45), wheat (0.44), oilseed rape (0.38), sheep (0.27), grass (0.27), milk (0.22).

Irrigation was estimated to increase the potential yield of crops grown on soil with a low AWC by between 37% (wheat) and 133% (potatoes), and the potential yield of crops on high AWC by 5% (wheat) to 100% (potatoes).

Climate change was predicted to increase the potential yields of irrigated crops by at least 20% by 2050 and to have a small effect on the potential yield of unirrigated crops.

If potential crop yields could be achieved then the crop area required to produce each unit of livestock production (kg of meat, l of milk or eggs) would be reduced by between two- and four-fold. The area of crop required could be reduced further if feed conversion efficiencies can be improved.

Constraints for achieving the potential yields are more likely to be caused by failure to take up and exploit technology than by failure the generate new technology

Low commodity prices, environmental regulation to reduce pollution and enhance biodiversity, social concerns about new technology (e.g. GM) and greater market demand for high quality are possible reasons for failure to take up technology.

Raising yields is not incompatible with reducing pollution and enhancing biodiversity, but yield improvement will be more difficult and therefore slower.

Breeding for non-yield traits such as pest resistance, resource use efficiency and enhanced welfare will slow the development of high yielding germplasm and may represent a constraint for generation of new technology

If recent yield trends continue then wheat yields will increase at 1% per year, the carcass weights of cattle, sheep, pigs and poultry will increase by 0.5% per year (with the qualification that the carcass meets the market requirements operating), eggs/bird/year and milk/cow/year will increase at 0.5% per year and no increase will occur for oilseed rape, peas, potatoes and grass.

Yields were predicted to increase fastest under ‘provincial enterprise’ (crops) or ‘world market’ (stock) scenarios due to lower environmental regulation, low input prices and abandonment of the least productive land. Static or negative yield changes are predicted for a ‘local stewardship’ scenario.

6. DISCUSSIONOn a global scale, improvement of farm yields is just one of several means by which an increasing population may be provided with adequate nutrition, fuel, clothing and building materials. Other means include (1) farm extension, by claiming natural habitat, (2) reducing wastage, e.g. 15% cereal production is wasted (p13), (3) reducing over consumption and (4) using more efficient species, e.g. fish are 8 times more efficient at producing protein from plant material than cattle (p14). Political and socio-economic forces, rather than science or technology, will largely determine the relative emphasis to be placed on each of these approaches. And farm yields cannot be considered

SID 5 (2/05) Page 19 of 22

as independent. Farm yields affect other approaches, and other approaches affect farm yields. Even if farm yields are considered to be the most technologically-driven, hence primary, means of sustaining the global population, yield improvements must ultimately be dependent on economics because, as has been shown over the last 50 years, new technologies require new investment, not just at the stage of initial innovation, but throughout the production chain. Predictions of farm yields reported here were made according to expert views of how economic forces will affect yield progress, the economic scenarios coming from Project IS0209. However, expert judgements of economic effects on yield were relatively crude.

On a regional scale, Project IS0209 did not further optimise farm yields. Rather, it considered yields predicted in this project to be largely independent of further socio-economic change, and used these as a key factor determining land use, environmental impacts, and other economic and social factors. It estimated the future demand for domestic food production under different scenarios from a range of factors including predicted economic growth, population growth, consumer preferences, and assumptions about the balance of imports and exports. The amount of UK land required for agricultural production was predicted to decrease under the Business As Usual, World Market and Global Sustainability scenarios. The greatest decrease was predicted under the World Market scenario for which 73% of lowland and <50% upland was predicted to be required for food and energy production. These changes were mainly due to greater yields and greater competition from imports. By contrast, insufficient land would be available to meet demand under the Global Sustainability and Local Stewardship scenarios as a result of lower yields and greater local demand for produce. The overall environmental load produced by agriculture was predicted to be least under the World Market and business as usual scenarios. This is due to the large amount of land released from agricultural production, although the burden is high in the areas that are farmed intensively. Local Stewardship, although demonstrating reduced environmental load in lowland farm areas, offers limited scope for land release in lowlands and uplands because of low yields. Because yields affect the amount of land released from agricultural production, they had a strong effect on social factors such as the amount of land available for recreation, levels of employment, and security of livelihoods. Yield estimates are clearly central to estimates of economic criteria. Interestingly, Project IS0209 predicted that the future scenarios with rapid yield increases did not necessarily have good aggregated economic scores (farm profit/ha, dependency on subsidy, food security).

6.1. Reliability of yield predictionsThe estimates of potential yield, predictions of progress towards them and identification of the most likely constraints were presented to 60 agricultural scientists at the International Conference ‘Yields of Farmed Species’ in June 2004 held at the University of Nottingham Sutton Bonington Campus. The authors were able to incorporate feedback arising from the main presentation and the workshop session which focussed on the detail of the calculations used. The findings reported here therefore represent a good consensus from an international group of scientists.

However, it is pertinent to recognise that a similar exercise conducted 50 years ago (Anon., 1955) was substantially inaccurate in foreseeing both the type and extent of changes that would bring national self-sufficiency in food. For example, it was envisaged that wheat production might increase from 1.4 to 3.8 Mt largely by extending the cropped area from 0.5 to 1.2Mha, yields increasing from 2.6 to only 3.1 t/ha, whereas in fact yields have increased to 8 t/ha, and the area to 2.0Mha. Forecasts of oil, fat and meat production were similarly inaccurate. We claim no greater foresight or sagacity than our learned predecessors, so must ask that the predictions made here be judged in terms of their rationale and reasoning, rather than their results. They have been devised with every effort at integrity, but without a confident view of future’s path.

6.2. Implications for the Research Priorities GroupThere is considerable scope to further increase yield of all species within the constraints of the UK environment and thereby food and fuel security could increase, imports could be reduced, and environmental costs could be curtailed. The large yield potentials in the UK confer on farmers a significant potential advantage in world markets by reducing unit costs. However, under current economic and political environments it seems unlikely that farmers will be able to exploit this advantage, because high costs, low commodity prices, environmental regulation and social issues will all tend to restrict development and exploitation of new and existing technologies for increasing yields.

For example, genetic yield improvements will be slowed by breeding for additional traits such as greater pest resistance, efficiency of resource use and avoidance of detrimental traits associated with high yields (e.g. low quality, low yield stability, reduced structural integrity, reduced reproductive performance, metabolic stress, reduced immunity to disease and functional fitness).

Realising potential yields would significantly reduce the amount of agricultural land required to produce each unit of crop and livestock product (by as much as 4-fold), thus potentially releasing land for non-agricultural use. This is predicted by Project IS0209 to have a significant and positive effect on the environment.

Possibly the most serious environmental repercussion of realising large crop yields would be on water resources. High yielding crops would probably utilise the majority of in situ rainfall, so would greatly reduce drainage into the water courses and aquifers of eastern UK. Climate change will tend to accentuate this problem. At present, it

SID 5 (2/05) Page 20 of 22

seems that hydrologists do not recognise the possible implications of changes in land management and crop improvement for water resources.

There may be a tendency for minor species and species with poor commercial performance, hence weak investment and poor prospects of yield improvement, to be displaced by more productive species, hence affecting landscape diversity.

6.3. Future work Publicly funded research should continue to be directed towards yield improvement, and in particular to

facilitating the uptake and exploitation of new technology.

Farm productivity is not closely related to use of inputs on-farm. It will be important to devise joint strategies, of both land use and husbandry, that optimise environmental and economic outcomes.

Better methods are needed to breed for improvement of several traits simultaneously, particularly where these are negatively associated traits. This particularly relates to improving yield without adversely affecting environmental or welfare factors.

References to published material9. This section should be used to record links (hypertext links where possible) or references to other

published material generated by, or relating to this project.

Adams, S.N., Easson, D.L., Gracey, H.I., Haycock, R.E., O’Neil, D.G. (1983). An attempt to maximise yields of cut grass in the field in Northern Ireland. Record of Agricultural Research 31, 11-16.

Alberda, T. (1977). Possibilities of dry matter productionfrom forage plants under different climatic conditions. Proceedings of the 13th International Grassland Congress. Pp 61-69. Leipzig, Germany.

Anon. (1955). Feeding the 50 million. A report of the Rural Reconstruction Association Research Committee on the Increase of Agricultural Production. London: Hollis & Carter. 138 pp.

Anon. (2000). Nutrient allowances and composition of feeding stuffs for ruminants. Defra (formerly MAFF) Booklet 2087.Barraclough, P.B. (1989). Root growth, macro-nutrient uptake dynamics and soil fertility requirements of a high-yielding winter

oilseed rape crop. Plant Soil 119: 59-70.Barraclough P B, Leigh R A. 1984. The growth and activity of winter wheat roots in the field: the effect of sowing date and soil

type on root growth of high yielding crops. J. Agric. Sci. 103 : 59-74.Burton, W.G. The Potato. H. Veenman & Zonen N.V. Wagenungen, Holland, pp 70.Clarke, J. (1999) Potential for the effects on environmental impact of farming (SA0111). Report for Defra by ADAS and CSL.Cooper, J.P. (1970). Potential production and energy conversion in temperate and tropical grasses. Herbage Abstracts 40,

1-15. Dreccer, M.F., Schapendonk, A.H.C.M., Slafer, G.A. and Rabbinge, R. (2000) Comparative response of wheat and oilseed

rape to nitrogen supply: absorption and utilisation efficiency of radiation and nitrogen during the reproductive stages determining yield. Plant and Soil, 220, 189-205.

Gammelvind, L.H., Schjoerring, J.K., Mogensen, C.R., Jensen, C.R. and Bock, J.G.H. (1996). Photosynthesis in leaves and siliques of winter oilseed rape (Brassica napus L.). Plant and Soil 186, 227-236.

Gregory PJ 1994. Resource capture by root networks. In: Monteith JL, Scott RK, Unsworth MH, eds. Resource capture by crops. Nottingham: Nottingham University Press, 77-97.

Heath, M.C. and Hebblethwaite, P.D. (1985). Agronomic problems associated with the pea crop. In: The Pea Crop. P.D. Hebblethwaite, M.C. Heath and T.C.K. Dawkine (Eds). Butterworths, London, pp27.

Hulme, M and Jenkins, G (1998). Climate change scenarios for the United Kingdom. Technical Report No 1, Scientific Report, UK Climate Impacts Programme, Norwich. 60.

Hulme, M, Jenkins, GJ, Lu, X, Turnpenny, JR, Mitchell, TD, Jones, RG, Lowe, J, Murphy, JM, Hassell, D, Boorman, P, McDonald, R and Hill, S (2002). Climate change scenarios for the United Kingdom. The UKCIP02 Scientific Report, School of Environmental Science, University of East Anglia, Norwich. 120

Kimball, B.A. (1983) Carbon dioxide and agricultural yield: an assemblage and analysis of 770 prior observations. WCL Report 14, November 1983

Leafe, E.L. (1988). Introduction – the history of improved grasslands. In The Grass Crop, pp 1-23. Edited by M.B. Jones and A. Lazenby. Chapman and Hall, London.

Monteith J.L. (1977) Climate and the efficiency of crop production in Britain. Philosophical Transactions of the Royal Society of London, 281, 277-294.

Monteith, J.L. (1981). Climatic variation and growth of crops. Quarterly Journal of the Royal Meteorological Society 107, 749-774.

SID 5 (2/05) Page 21 of 22

Rao, M.S.S., Mendham, N.J. and Buzza, G.C. (1991) Effect of apetalous flower character on radiation distribution in the crop canopy, yield and its components of oilseed rape (Brassica napus). Journal of Agricultural Science Cambridge, 117, 189-196.

Roberts M.J., Long S.P., Tieszen L.L. and Beadle C.L. (1993) Measurement of plant biomass and net primary production of herbaceous vegetation. In: Photosynthesis and Production in a Changing Environment: A Field and Laboratory Manual (eds D.O. Hall, J.M.O. Scurlock, H.R. Bolhàr-Nordenkampf, R.C. LeegoodandS.P. Long), pp. 1-21. Chapman & Hall, London.

Robson, M.J. (1981). Potential production what is it and can we increase it? In Wright C.E. (Ed) Plant Physiology and Gerbage Production. Occasional Symposia of the British Grassland Society 13, 5-18.

Bowell, D.L. (1994). Soil Science methods and applications. Longman Group Ltd.Sinclair, T.R. and Muchow, R.C. (1999). Radiation use efficiency. Advances in Agronomy 65, 215-265.Sinclair, T.R. and de Witt, C.T. (1975). Photosynthate and nitrogen requirements for seed production by various crops.

Science 189, 565-567.Summerfield, R.J. and Roberts, E.H. (1985). Pea. In: Grain Legume Crops. D.R. Davies, G.J. Berry, M.C. Heath and T.C.K.

Dawkins (Eds). William Collins Sons and Co. Ltd, London, pp 158.Sylvester-Bradley, R. and Wiseman, J. (2005). Yields of Farmed Species, Constraints and opportunities in the 21st Century.

Nottingham University Press, UK. 651 pp.Troughton, A. (1957). The underground organs of herbage grasses. Bulletin, Commonwealth Bureau of Pastures and Field

Crops, Hurley, No. 44, pp174.Wilson, D. (1982). Response to selection for dark respiration rate of mature leaves in Lolium Perenne and its effects on young

plants and simulated swards. Annals of Botany 55, 303-312.

SID 5 (2/05) Page 22 of 22