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 47

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 LS 3601

2. Project title

Finishing Pigs - Systems Research

3. Contractororganisation(s)

Meat and Livestock CommissionPO Box 44, Winterhill houseSnowdon Drive, Milton KeynesMK6 1AX          

54. Total Defra project costs £ 1744842

5. Project: start date................ 01 September 2000

end date................. 31 August 2005

SID 5 (2/05) Page 2 of 47

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.BACKGROUND

The Finishing Pigs Systems Research Project was a major multidisciplinary initiative co-ordinated by the MLC. It brought together scientists and specialists to work on areas of strategic importance to the industry and to future policy. The project covered research objectives under cost of production, environment, food safety, pig health and welfare and meat quality.

A central objective was to explore the potential of liquid feeding technology to improve competitiveness and address food safety issues relating to salmonella. The work established that liquid feeding holds the potential to reduce cost of production by a net of around 13.5p/kg dead weight and is an effective management tool for controlling the presence of salmonella in pigs at slaughter.

The project was based around four large-scale production studies at MLC’s Stotfold Pig Development Unit. These were carried out consecutively over a three-year period, from April 2002 to February 2005, with each study based on two feeding treatments evaluated within two contrasting systems of housing (fully-slatted and straw-based). Field evaluations in the production studies were underpinned by detailed laboratory investigations of liquid feeding for the development of standard operating procedures.

Project results will benefit the competitive position of the British pig industry and will provide sound information to policy makers so that appropriate balance can be achieved between the requirements for production, welfare, food safety and the environment towards sustainable development.

PRODUCTION STUDIES

Feeding treatmentsStudy 1: Dry vs. liquid feeding. Dry pelleted diets were compared with diets fed in liquid form. All diets contained a common background of raw ingredients and were formulated to the same nutrient specification. Study 2: Single diet vs. phase feeding. Two liquid diets were formulated to meet the energy and ideal protein requirements of pigs at the extremes of the weight range (30kg to 110kg live weight). Under phase-feeding the proportionality of the diets delivered at the troughs using a computer controlled system was adjusted daily to match protein and energy requirements according to growth curves. The proportionality of the two diets was fixed for single diet feeding from entry to slaughter. Study 3: Fermented cereals. The cereal fraction of a liquid diet was pre-fermented using a registered strain of Pediococcus acidilactici (Bactocell) under controlled temperature (35oC) and time (24-hour). This was

SID 5 (2/05) Page 3 of 47

compared by feeding the same diet in liquid form but without pre-fermentation of the cereal fraction. Study 4: Low protein liquid feeding. A control liquid diet was formulated without synthetic amino acids to meet the ideal protein requirements of growing/finishing pigs. This resulted in a crude protein content of 233g/kg dry matter (DM). Crude protein was reduced in the low protein diet to 175g/kg DM by the use of synthetic amino acids (lysine, methionine, threonine and tryptophan) whilst supplying an equivalent level of standardised ileal digestible ideal protein.

HousingThe housing systems were purpose built on a single site in an effort to minimise potentially confounding farm effects, and the timing of the studies were representative of the different seasons of the year. The buildings had identical shells and each consisted of four rooms containing four pens. The straw based pens measured 5.8m x 3.7m, including the scrape-through passage, which was cleaned out daily. Solid concrete was used for both the lying and the scraped dunging areas. After cleaning out, around 0.4kg of straw per pig per day was added to the pens. Fully slatted pens measured 5.5m x 3.7m and had flooring of concrete slats with 83mm width and 18mm gap. A one-metre wide central walkway ran throughout all four rooms in each building. The ventilation and environment in both housing systems was automatically controlled to set maximum and minimum ventilation, relative humidity and temperature against occupancy day. Each room had two windows, allowing natural daylight, and artificial light was mainly used during husbandry tasks, weighing and behaviour observations.

PigsIn each study, 1024 pigs ((Large White x Landrace) x Large White) were received in 8 equal batches of 128 over a period of 14 ± 4 weeks. In Study 4, two additional batches were evaluated for growth performance. Batches were allocated alternately between the housing systems until all rooms were full. After a period of 4 to 5 days acclimatisation to the housing, each pig was ear tagged for individual identification and weighed. The batch was divided into four equal groups of 32 pigs in order of weight, and then each group was randomly allocated to one of four pens within a single room. The mean live weight of animals at entry was 34.3 ± 0.65kg. Numbers per pen were reduced at week six (mid-point) to 25 in the fully slatted system and 20 in the straw based system, in accordance with normal commercial stocking densities for these housing types. Pigs were fed ad libitum and water was available ad libitum from four nipple drinkers in each pen. Pigs were slaughtered at around 104kg live weight. Feeding treatments were applied at the room level for the determination of environmental impact and to reduce microbial contamination, with the pen as the experimental unit for statistical analysis of variance (ANOVA).

KEY RESULTS

Production and competitivenessStudy 1: Liquid feeding improved daily gain (796 vs. 754g/day; P<0.001) and the conversion of feed to gain (2.27 vs. 2.53 feed conversion ratio or FCR; P<0.001) resulting in a net financial benefit of 4.6p/kg dead weight in reduced cost of production. Study 2: There were no pig performance or cost benefits from phase feeding. Cost of production could be potentially reduced by a further 8.8p/kg dead weight by food industry liquid co-products (Study 2 and 4). Study 3: Fermenting the cereal fraction increased intake (1.88 vs. 1.80 kg/day of meal equivalent; P<0.05) but not daily gain resulting in a poorer FCR (2.39 vs. 2.27; P<0.001). Controlled fermentation proved expensive with no associated performance benefits. Study 4: Reducing protein content of the liquid diet by supplementation with synthetic amino acids increased intake (2.11 vs. 2.03 kg/day of meal equivalent; P<0.05) but reduced gain (821 vs. 851 g/day; P<0.05) resulting in a poorer FCR (2.68 vs. 2.49; P<0.001)). The low protein diet increased carcase fatness at the P2 position (12.77 vs. 12.22mm; P<0.05) and cost of production by a net of 2.8p/kg dead weight.

Environmental impactStudies 1 and 2: There were no significant feeding treatments effects on measures of environmental impact. Study 4: The ammoniacal nitrogen content of the effluent and ammonia emission were reduced (P<0.05; P<0.001) by dietary protein reduction.

MicrobiologyIn all liquid feeding treatments elevated microbial counts (e.g. total viable aerobic and anaerobic, lactic acid bacteria and yeast) plus presence of end products (e.g. lactic and acetic acid and ethanol) confirmed that natural fermentation was occurring even in the absence of controlled inoculation. Study 1: Liquid feeding reduced the percentage of pigs found positive for caecal presence of salmonella at slaughter (39% vs. 23%; P<0.05). Studies 2, 3 and 4: Salmonella levels at slaughter remained low (<1.5%; <2%; <4% of pigs tested) under all liquid feeding treatments. There were no measurable advantages to the control of salmonella from cereal fermentation but there was statistical evidence of a benefit from a reduction in dietary protein content (1.7 vs. 5.6% caecal positive; P<0.05).

SID 5 (2/05) Page 4 of 47

Meat qualityStudies 1 and 4: There were no consistent significant feeding treatments effects on the quality of fresh and cooked meat samples. Study 3: The juiciness and tenderness of cooked loin was improved (P<0.05; P<0.01) by feeding fermented cereals. The underlying mechanisms are unclear.

Pig health and welfareDetailed health and behavioural investigations covering the four production studies were carried out under a separate Defra welfare research contract (AW0130) and will be reported separately.

HOUSING SYSTEMS COMPARED OVER FOUR TRIALS

There were no significant differences in the performance and carcase quality of pigs under the two housing systems. There were no consistent significant and important effects on measures of microbial status, meat quality and environmental impact. Cost of production was on average 3p/kg dead weight higher in the straw based system due to increased labour input and the requirement for bedding.

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).

Project Report To Defra

FINISHING PIGS – SYSTEMS RESEARCH

BACKGROUND

The Finishing Pigs Systems Research Project was a major multidisciplinary initiative co-ordinated by the MLC. It brought together scientists and specialists to work on areas of strategic importance to the industry and to future policy. The project covered research objectives under cost of production, environment, food safety, pig health and welfare and meat quality.

A central objective was to explore the potential of liquid feeding technology to improve competitiveness and address food safety issues relating to salmonella. The work established that liquid feeding holds the

SID 5 (2/05) Page 5 of 47

potential to reduce cost of production by a net of around 13.5p/kg dead weight and is an effective management tool for controlling the presence of salmonella in pigs at commercial slaughter.

The project was based around four large-scale production studies at MLC’s Stotfold Pig Development Unit. These were carried out consecutively over a three-year period, from April 2002 to February 2005, with each study based on two feeding treatments evaluated within two contrasting systems of housing (fully-slatted and straw-based). Field evaluations in the production studies were underpinned by detailed laboratory investigations of liquid feeding for standard operating procedures.

Project results will benefit the competitive position of the British pig industry and will provide sound information to policy makers so that appropriate balance can be achieved between the requirements for production, welfare, food safety and the environment towards sustainable development.

SCIENTIFIC OBJECTIVES

The scientific objectives are presented below including changes to the work programme as agreed with Defra.

1. Cost of production

Objective 1.1 To establish the effect of flooring (straw vs. fully slatted) and feeding and their interaction within a finishing system on the growth performance, carcase quality and cost of producing pigs from around 35kg to slaughter weight at around 105kg (Production Studies 1, 2, 3 and 4).

2. Environmental Impact

Air quality and gaseous emissions

Objective 2.1 To establish the effect of flooring (straw vs. fully slatted) and feeding and their interaction within a finishing system on air quality and gaseous emissions (Production Studies 1, 2 and 4).

Waste output and composition

Objective 2.4 To establish the effect of flooring (straw vs. fully slatted) on waste output and composition from a finishing system (Production Studies 1, 2 and 4).

Objective 2.5 To establish the effect of feeding on waste output and composition from a fully slatted finishing system (Production Studies 1, 2 and 4).

3. Microbial ecology of the finishing system

Objective 3.1 To monitor the microbial loading of individual feed ingredients and complete diets fed to pigs (Production Studies 1, 2, 3 and 4).

Objective 3.2 To monitor dynamic changes in faecal shedding of salmonella during growth and caecal presence of salmonella at slaughter according to flooring (straw vs. fully slatted) and feeding as a risk assessment for food safety (Production Studies 1, 2, 3 and 4).

Objective 3.3 To determine the effect of flooring (straw vs. fully slatted) and feeding and their interaction within a finishing system on the microflora of the gut as an indicator of gut health (Production Studies 1 and 3).

Objective (3.4) To establish the relative risk of dust and airborne bacteria as potential contaminants according to flooring (straw vs. fully slatted) and feeding and their interaction within a finishing system (Production Studies 1, 2, 3 and 4).

Objective (3.5) To monitor straw as a potential risk factor for the introduction of salmonella into the pig production system.

Objective (3.6) To monitor slurry as a potential risk factor as an environmental contaminant of salmonella and coliform bacteria.

Objective (3.7) To monitor drinking water as a potential risk factor for the introduction and loading of pathogenic bacteria into the pig production system.

SID 5 (2/05) Page 6 of 47

4. Meat eating quality

Objective (4.1) To compare the impact of flooring system (fully slatted vs. straw based) on meat quality (Production Studies 1, 3 and 4).

Objective (4.2) To establish any adverse effects on meat quality from the application of liquid feeding (Production Studies 1, 3 and 4).

5. Pig health, welfare and behaviour

Objectives covering this element of the research programme fell within a separate CSG 7 application to Defra by the University of Newcastle.

6. Development of liquid feeding technology

Objective (6.1) To develop standard operating procedures (SOPs) for liquid feeding to optimise the growth performance of growing and finishing pigs.

Objective (6.2) To develop SOPs for processing feed ingredients under liquid feeding to optimise the growth performance of growing and finishing pigs.

Objective (6.3) To develop SOPs for controlled fermentation of feed ingredients under liquid feeding to optimise biosecurity and the growth performance of growing and finishing pigs.

All of the above objectives were met in full.

MAIN IMPLICATIONS

The research has clearly demonstrated that liquid feeding pigs during the growing and finishing stages (35kg live weight to slaughter at 105kg live weight) can significantly improve their speed of growth and conversion of feed to body weight gain offering a net economic benefit of around 5p/kg dead weight in reduced cost of pig meat production. Further cost benefits of around 9p/kg dead weight in reduced cost of production could be achieved by the inclusion of liquid co-products from the human food industry in liquid cereal based diets for growing and finishing pigs.

Under field studies, further refinements in liquid feeding technology, such as the application of phase-feeding, controlled fermentation of the cereal fraction of the diet and reduction in crude protein content by supplementation with synthetic amino acids did not deliver any additional growth and feed conversion improvements and proved financially neutral or negative.

Liquid feeding provides a mangement tool for controlling and reducing the caecal presence of salmonella in finished pigs at slaughter thereby improving the safety of pig meat for consumers. There are no undersirable effects on the fresh and sensory quality of meat from liquid fed pigs.

Under the conditions of research it proved difficult to establish statistically positive effects of liquid feeding on measurements relating to environmental impact (e.g. dust and ammonia emissions and nutrient content of waste products). However ammonia emission was significantly reduced by a dietary reduction in crude protein content but the lack of performance and cost benefits present current barriers to uptake.

There were no significant differences in the performance and carcase quality of pigs housed in fully slatted and straw based accommodation. There were no consistent significant and important differences in measures of microbial status, meat quality and environmental impact between the two housing systems. Cost of production was on average 3p/kg dead weight higher in the straw based system due to increased labour input and the requirement for bedding.

The competitive position of the British pig production industry could be significantly improved by the adoption of liquid feeding technology and utilisation of food industry liquid co-products. Whilst there are indications of increased uptake, a lack of confidence in primary production remains a barrier to investment in new technology including modern liquid feeding systems despite financial and other benefits (e.g. food safety) established under this research and development programme.

FUTURE RESEARCH

The work clearly demonstrated the potential benefits of liquid feeding to enhance pig growth performance, reduce cost of production and reduce the gut presence of salmonella in finished pigs slaughtered for meat production.

SID 5 (2/05) Page 7 of 47

The research could not establish performance, cost and environmental benefits from phase-feeding where a computer controlled liquid feeding system was used to accurately deliver diets that matched the energy and protein requirements of pigs according to their growth curve. The research showed that producers could simplify their feeding systems for growing and finishing pigs by delivering a single diet from 35kg to slaughter at 105kg without performance and cost penalties. This is at variance to current recommendations that seek to reduce the environmental impact of pig production by stepped reductions in dietary protein (see Integrated Pollution and Prevention and Control IPPC, Standard Farming Installation Rules and Guidance for Pig Rearing Version 4, June 2005, Environmental Agency). Furthermore dietary reductions in crude protein content using synthetic amino acids in liquid diets for growing and finishing pigs had an adverse affect on growth and feed conversion efficiency, although reductions in ammonia emissions pointed to environmental benefits.

Further research is therefore urgently required on the efficacy of feeding strategies that aim to deliver environmental benefits but do not compromise the efficiency and economics of pig production.

Liquid feeding was found to reduce the caecal presence of salmonella in pigs at commercial slaughter. The underlying mechanisms warrant research at the fundamental level to establish if they can be transferred to pigs fed dry diets. It is estimated that only 15 to 20% of pigs finished in the UK are fed using liquid feeding systems.

TECHNOLOGY TRANSFER

The Meat and Livestock Commission facilitated and co-ordinated the technology transfer of the research programme. Transfer mechanisms included industry, technical and scientific conferences, shows, workshops and seminars, presentations at pig discussion groups and roadshows, scientific and technical publications, leaflets, press communications, articles in trade journals and websites. A high profile was maintained during the lifetime of the research programme and key results were communicated to industry following the completion of each of the four major field studies. Technology transfer activity will continue beyond the completion of the project to maximise impact and facilitate uptake. Full details of the technology transfer activity completed to date is given at the end of this report.

RESULTS

PRODUCTION STUDIES

Detailed individual reports for each of the four production studies have been published separately. The key findings according to contracted scientific objectives are presented below.

Cost of production

Major pig performance and cost benefits were established from the application of liquid feeding in Study 1. Pigs were fed ad libitum growing and finishing diets in either liquid or dry pellet form. Diets contained a similar background of individual raw ingredients (cereals and soya bean, rapeseed and fish meals, supplemented with minerals and vitamins) and formulated to the same nutrient specification. Feeding the diets in liquid form improved gain from 754g to 796g/day (P<0.001) and the conversion of feed to gain (feed conversion ratio or FCR) from 2.53 to 2.27kg feed (as meal equivalent at 87% dry matter) per kg body weight gain (P<0.001). Good carcase quality was maintained under liquid feeding with backfat depths averaging 11.45 and 11.39 mm for liquid and dry fed pigs respectively. Detailed financial analysis showed that the performance and efficiency advantages associated with liquid feeding generated a cost benefit equivalent to a net reduction of 4.6p/kg dead weight in the cost of production.

Modern computer controlled liquid feeding systems can be used to achieve greater nutritional accuracy by mixing and delivering diets that match the nutrient requirements of pigs according to their growth potential as in phase-feeding. Phase-feeding should reduce nutrient wastage and cost and holds potential benefits for the environments in terms of nutrient loading. In Study 2 two liquid diets were formulated to meet the energy and ideal protein requirements of pigs at the extremes of the weight range (30kg to 110kg live weight). Under phase-feeding the proportionality of the diets delivered at the troughs using a computer controlled system was adjusted daily to match protein and energy requirements according to growth curves. In contrast phase-feeding was compared with a single diet treatment in which the nutritional compostion of the diet remained constant over the growth curve from entry to slaughter by fixing the proportionality of the two diets delivered at the feeding troughs. Disappointingly phase-feeding reduced intake (1.99 vs. 2.06kg meal equivalent/day; P<0.05) and daily gain (860 vs. 886 g/day; P<0.05) with no benefits in feed conversion (2.36 vs. 2.36 FCR; ns). Carcase quality was similar in phase and single diet fed pigs. There were no financial benefits from phase-feeding with cost of production for phase and single fed pigs totalling 86.4 and 85.8p/kg dead weight respectively.

SID 5 (2/05) Page 8 of 47

There has been considerable interest within European in the fermentation of liquid diets using lactic acid bacteria to improve pig performance and control salmonella. In Study 3 the cereal fraction of a liquid diet was pre-fermented using a registered strain of Pediococcus acidilactici (Bactocell) under controlled temperature (35oC) and time (24-hour). This was compared by feeding the same diet in liquid form but without pre-fermentation of the cereal fraction. Although voluntary intake was enhanced with the inclusion of fermented cereals (1.93 vs. 1.83kg meal equivalent/day; P<0.05) this did not improve daily gain (818 vs. 844g/day; ns) resulting in a poorer FCR (2.47 vs. 2.26; P<0.001). Carcase fatness was increased with the inclusion of fermented cereals in the liquid diet ( 11.97 vs. 10.87mm; P<0.001). Controlled fermentation involved increased equipment and energy expenditure giving a net increase of 20.2 p/kg dead weight in cost of production. Currently controlled fermentation offers no performance advantages and involves significant cost penalties in the production of growing and finishing pigs.

Growing and finishing pigs account for over 50% of the total nitrogen emissions from pig production. In Study 4 the scope for reducing nitrogen emission by lowering the dietary crude protein content of liquid diets and the consequences for pig performance and cost of production were evaluated. . A control liquid diet was formulated without synthetic amino acids to meet the ideal protein requirements of growing/finishing pigs. This resulted in a crude protein content of 233g/kg dry matter (DM). Crude protein was reduced in the low protein diet to 175g/kg DM by the use of synthetic amino acids (lysine, methionine, threonine and tryptophan) whilst supplying an equivalent level of standardised ileal digestible ideal protein. Reducing protein content of the liquid diet by supplementation with synthetic amino acids increased intake (2.11 vs. 2.03kg/day of meal equivalent; P<0.05) but reduced gain (821 vs. 851g/day; P<0.05) resulting in a poorer FCR (2.68 vs. 2.49; P<0.001). The low protein diet increased carcase fatness at the P2 position (12.77 vs. 12.22 mm; P<0.05) and cost of production by a net of 2.8p/kg dead weight. Feed sample analysis showed that there was no loss of synthetic amino acids in the liquid diet with complete recovery of free lysine, methionine and threonine.

Growth and cost of production (CoP) by feeding system are summarised below for all trials. Table 1 Growth rate and cost of production according to feeding treatment

Study 1 Study 2 Study 3 Study 4Dry Liquida Singleb Phaseb Controlc Fermentedc Controld Low Proteind

Growth rate (g/day) 754 796 886 860 844 818 851 821CoP (p/kg dead weight) 99.2 94.6 85.8 86.4 93.6 113.8 87.3 90.1

a No liquid co-products used, major ingredients included cereals, wheatfeed, soya bean meal, rapeseed meal and fish meal. b Liquid co-products used (Greenwich Gold and Lactose 16) with cereals, wheatfeed, soya bean meal and rapeseed meal as

other major ingredients. c No liquid co-products used, major ingredients included cereals, wheatfeed, soya bean meal and rapeseed meal. d Liquid co-product used (Greenwich Gold) with cereals, soya bean meal and rapeseed meal as other major ingredients.

A further reduction of up to 8.8p/kg dead weight in the cost of production from Study 1 to Study 2 was largely attributable to the inclusion of food industry co-products.

Over the four studies, cost of production was on average 3p/kg dead weight higher for the straw based system due to higher labour costs and the requirement for straw bedding.

Environmental impact

Air quality and gaseous emissions

Measurements focused on the concentration of dust in the aerial environment of the strawed and fully slatted housing and emissions of dust and ammonia from the two systems. The effects of feeding during studies 1, 2 and 4 could only be achieved by applying treatments at the room level therefore restricting the number of replications for statistical comparison to 4 per treatment. Results are summarised in Table 2 below.

Table 2 Ammonia and dust emission and dust concentration

In Studies 1 and 2 there were no significant effects of feeding treatment on dust concentration and the emission of ammonia and dust from the strawed and fully slatted housing. In Study 4, reducing the protein content of the

SID 5 (2/05) Page 9 of 47

Study 1 Study 2 Study 4 HousingDry Liquid Single Phase Control Low Protein Fully

slattedStraw based

Ammonia emission (g NH3-N per lu hour)

1.79 1.96 0.74 0.70 1.11 0.73 1.04 0.92

Dust concentration (mg/m3)

2.60 1.80 2.00 2.90 0.70 0.70 1.40 1.10

Dust emission (g per lu hour)

1.51 0.97 0.37 0.58 0.24 0.20 0.41 0.37

diet reduced ammonia emission(P<0.001). Within each study there were no significant interactions between feeding treatment and housing system on measurements of emission and dust concentration.

Overall (Studies 1, 2 and 4) there were no significant effects of housing system on ammonia and dust emissions and dust concentration.

The ammonia emission measured in these studies are lower than the average, but within the spread reported in the national ammonia inventory, which lists fattening pigs on straw or fully slatted as 2.84 and 2.94 g NH3-N per lu hour, respectively.

Controlled accuracy in the weighing and mixing of feed ingredients in the production of liquid diets is likely to be the main factor contributing to the low emission levels. Secondly, the high level of building management and stockman input is also a contributing factor to the lower than average emission factors.

The implication of this extensive research program is evidently that tightly specified diets and in particular low protein diets are the best strategy to lower ammonia emissions from grower/finisher accommodation, irrespective of the housing type.

The dust emissions are also in line with literature values (range 0.18-0.39 g per lu hour). Animal activity is the largest factor influencing dust emission, hence no significant effects were found for feeding treatment. Higher dust emissions were expected from the straw based system. This was not evident in this study, potentially due to the low amount of straw used in this system (up to 25cm depth in the lying area), compared to the deep straw systems referred to in the literature.

Waste output and composition

The production and composition of waste by feeding treatment within the fully slatted housing and overall for the straw based system is summarised in Table 3. In Study 1, liquid feeding increased the volume of effluent production (P<0.001) in the fully slatted system. In Study 4, reducing dietary protein content reduced the ammoniacal (P<0.05) and Kjeldahl (P<0.05) nitrogen content of the effluent and hence reduced ammonia emission.

Table 3 Production and composition of waste

Microbial ecology of the finishing system

Microbial loading of the feed

In total of 55 samples of dry feed ingredients (barley, wheat, wheatfeed and rapeseed, soya bean and fish meal), total aerobic, total anaerobic, lactic acid bacteria, enterobacteracia and yeast counts averaged 5.26, 4.62, 3.20, 2.91 and 3.18 Log10 colony forming units (CFU) per gram) respectively. Enterobacteracia could not be detected in 14 samples of minerals and vitamins supplements but total aerobic, total anaerobic, lactic acid bacteria and yeast counts averaged 3.42, 3.50, 2.68 and 4.77 Log10 CFU/g respectively.

Elevated microbial counts (total aerobic, total anaerobic, lactic acid bacteria and yeast counts averaging 8.28, 7.95, 8.34 and 6.18 Log10 CFU/g respectively) plus presence of end products (lactic and acetic acid and ethanol averaging 117, 32 and 42 mmol/litre respectively) confirmed that natural fermentation was occurring in liquid diets (106 samples) even in the absence of controlled microbial inoculation.

SID 5 (2/05) Page 10 of 47

Fully slatted Straw basedStudy 1 Study 2 Study 4

Dry Liquid Single Phase Control Low ProteinProduction (pig/day)

Effluent (litres) 5.20 7.31 7.48 7.58 6.98 7.68 -Farm yard manure (kg) - - - - - - 4.10

CompositionDry matter (%) 6.4 6.9 9.2 8.0 7.3 8.3 17.8

Ammoniacal nitrogen (mg NH4-N/kg fresh)

3033 3110 3340 3270 4150 3530 2640

Kjeldahl nitrogen (mg N/kg fresh)

4070 4350 5530 5220 5690 5210 6580

Total phosphorus (mg/kg dry matter)

19900 13090 13480 11620 23060 19710 10240

Salmonella could not be detected in samples of individual dry feed ingredients, liquid co-products and final dry and liquid feed mixtures. A lactic acid content of 100 mmol/litre has been found to be effective in controlling salmonella in liquid diets (see Defra project report LS0812).

Straw and drinking water

Salmonella could not be detected in a total of 88 samples of fresh straw analysed during the production studies. Coliform counts were absent in drinking water sampled (94 samples) from drinking points in the straw based and fully slatted housing.

Salmonella status of the pigs

All pigs were tested at entry and at slaughter for salmonella antibodies using enzyme-linked immunosorbent assay (ELISA). Sixteen pigs were slaughtered at entry in each of studies 1 and 3 and tested for the caecal presence of salmonella. All pigs at slaughter were tested for the caecal presence of salmonella.

Results suggested that pigs were the most likely vehicle for the entry of salmonella to the finishing systems. Approximately 10% of pigs at entry tested positive for salmonella antibodies with one pig testing positive for caecal presence.

Results at entry and slaughter using the ELISA and caecal methods are presented in Tables 4 and 5 according to feeding treatment and housing system for each of the four studies.

Table 4 Salmonella status of pigs at entry and slaughter according to feeding treatmentStudy 1 Study 2 Study 3 Study 4

Dry Liquid Single Phase Control Fermented Control Low ProteinEntry

Ceacal positive (%) See footnotea None tested See footnoteb None testedELISA positive (%) 9 6 14 17 3 3 14 12

SlaughterCeacal positive (%) 39 23 See footnotec 3 1 6 2ELISA positive (%) 35 16 6 6 6 3 29 20

a One pig found positive out of 16 tested. b No pigs found positive out of 16 tested.

c Only 9 pigs positive out of 695 tested (6 single, 3 phase).

Table 5 Salmonella status of pigs at entry and slaughter according to housing systemStudy 1 Study 2 Study 3 Study 4

Fully slatted

Straw based

Fully slatted

Straw based

Fully slatted

Straw based

Fully slatted

Straw based

EntryCeacal positive (%) See footnotea None tested See footnoteb None testedELISA positive (%) 10 5 10 20 3 3 19 7

SlaughterCeacal positive (%) 28 34 See footnotec 1 3 3 4ELISA positive (%) 22 29 3 9 4 5 28 20

a One pig found positive out of 16 tested. b No pigs found positive out of 16 tested. c Only 9 pigs positive out of 695 tested (5 fully slatted, 4 straw based).

In Study 1 liquid feeding reduced the percentage of pigs found positive for salmonella antibodies (P<0.05) and with caecal presence of salmonella (P<0.05). In Study 4, reducing protein content of the liquid diet reduced the percentage of pigs testing positive for caecal presence (P<0.05) and the reduction in ELISA positives approaching statistical significance (P=0.087). Following Study 1, caecal presence at slaughter remained subsequently low in Studies 2, 3 and 4 (<1.5%; <2%; <4% of pigs tested) against a background of different liquid feeding treatments.

Within each study there were no significant interactions between housing and feeding system for salmonella status at slaughter.

Overall (Studies 1 to 4) there were no significant effects of housing system on salmonella status at slaughter.

SID 5 (2/05) Page 11 of 47

These results indicate the potential of liquid feeding as an effective management tool in the control of salmonella in pigs at slaughter with scope for dietary manipulation to achieve further control such as reduction in dietary protein content.

Microbial loading of dust, faeces and effluent

A total of 279 room dust, 384 pen faecal and 98 effluent samples were taken and tested for salmonella using microbial culture methods. Within study, sample distribution was spread evenly across feeding treatment and housing system. The number of samples which tested positive according to feeding treatment and housing system under the four studies are presented in Tables 6 and 7.

Table 6 Number of room dust, pen faecal and effluent samples tested positive for salmonella presence according to feeding treatment

Study 1 Study 2 Study 3 Study 4Dry Liquid Single Phase Control Fermented Control Low Protein

Room dust 3 4 0 0 0 0 1 0Pen faeces 7 8 0 0 0 3 3 2Effluent 9 1 0 0 1 0 1 2

Table 7 Number of room dust, pen faecal and effluent samples tested positive for salmonella presence according to housing system

Study 1 Study 2 Study 3 Study 4Fully

slattedStraw based

Fully slatted

Straw based

Fully slatted

Straw based

Fully slatted

Straw based

Room dust 4 3 0 0 0 0 0 1Pen faeces 4 11 0 0 1 2 3 2Effluent 10 - 0 - 1 - 3 -

Results indicated that room dust, faeces and effluent were potential sources of contamination once salmonella had become established in the housing system as under Study 1 where spread was assisted by the movement of manure along the scraping passage.

Dust generally presented a lower level of contamination carrying around 2 to 4 Log10 CFU/g of salmonella compared with effluent containing between 1.6 and 160 2 Log10 CFU/g. Faecal samples were tested only for the presence or absence of salmonella.

Faecal, room dust and effluent samples were also tested for total aerobic,anaerobic, lactic acid bacteria and enterobacteraciae counts. Results are presented Tables 8 and 9 according to feeding treatment and housing system under the four studies.

The effects of feeding treatment on the microbial status of room dust, pen faecal and effluent samples are small and are likely to be of little practical importance. A consistent and significant finding was that liquid feeding improved (P<0.05) the ratio of lactic acid bacteria and coliform in faecal samples primarily due to a reduction in coliform counts (P<0.01). This microbial shift was also observed in digesta samples removed from the ileum, caecum and colon of finished pigs during commercial slaughter (see Table 10). These findings complement the beneficial effects of liquid feeding in reducing salmonella at slaughter and provide further evidence of the positive effects of liquid feeding on gut health.

The total aerobic viable count in room dust was consistently higher (0.001<P<0.01) in samples removed from the straw based system. The implications of a higher microbial dust burden for animal and stockworker respiratory health warrant further investigation.

SID 5 (2/05) Page 12 of 47

Table 8 Microbial status of faecal, dust and effluent samples according to feeding treatment

Table 9 Microbial status of faecal, dust and effluent samples according to housing system

Table 10 Gut lactic acid bacteria and coliform counts at slaughter (Study 1)Feeding system P

Liquid DryLactic Acid Bacteria (LAB) Count

Ileal 7.94 8.33 **Caecal 7.85 8.65 **Colon 8.39 9.16 **

SID 5 (2/05) Page 13 of 47

Study 1 Study 2 Study 3 Study 4Dry Liquid Single Phase Control Fermented Control Low

ProteinTotal Aerobic Viable Count

Faeces 8.74 8.65 8.56 8.63 8.76 8.74 9.04 8.77Dust 7.44 7.41 7.99 8.03 7.21 7.68 8.32 7.92

Total Anaerobic Viable Count

Faeces 9.06 8.87 9.75 9.93 9.31 9.44 9.84 9.60Lactic Acid Bacteria (LAB) Count

Faeces 8.56 8.25 8.18 8.52 8.41 8.24 9.21 8.86Coliform Count

Faeces 7.01 6.31 5.19 5.52 5.68 5.78 5.81 5.98Dust 2.71 2.72 1.76 1.85 1.48 1.52 2.23 2.62

LAB : Coliform RatioFaeces 1.22 1.32 1.61 1.56 1.45 1.47 1.64 1.52

Enterobacteraciae Count

Effluent 5.51 5.47 4.96 5.58 5.58 5.25 4.46 4.76

Study 1 Study 2 Study 3 Study 4Fully

slattedStraw based

Fully slatted

Straw based

Fully slatted

Straw based

Fully slatted

Straw based

Total Aerobic Viable Count

Faeces 8.73 8.66 8.43 8.76 8.72 8.78 8.86 8.94Dust 6.91 7.92 7.42 8.69 6.96 8.04 7.71 8.53

Total Anaerobic Viable Count

Faeces 8.95 8.98 9.76 9.93 9.37 9.39 9.59 9.85Lactic Acid Bacteria (LAB) Count

Faeces 8.35 8.46 8.19 8.05 8.36 8.29 9.03 9.03Coliform Count

Faeces 6.73 6.59 5.32 5.39 5.65 5.81 5.98 5.81Dust 2.11 3.30 1.51 2.13 1.47 1.53 2.52 2.32

LAB : Coliform RatioFaeces 1.25 1.29 1.57 1.61 1.51 1.40 1.55 1.61

Enterobacteraciae Count

Effluent 5.48 5.26 5.43 4.62

Coliform CountIleal 7.55 8.33 **

Caecal 5.84 7.02 **Colon 5.79 7.10 **

LAB : Coliform RatioIleal 1.07 1.01 *

Caecal 1.37 1.26 **Colon 1.48 1.31 **

Meat quality

Fresh and cooked meat samples from Studies 1, 3 and 4 were evaluated for attributes of visual, biochemical and sensory quality. Measurements included drip loss, colour saturation, oxidative rancidity, indole and skatole content, cooking loss and sensory determinations for juiciness, tenderness, flavour and odour.

There were no undesirable effects of the various liquid feeding treatments on meat quality. In Study 3, the juiciness (P<0.05) and tenderness (P<0.01) of cooked loin samples was improved in pigs fed a liquid diet containing fermented cereals. The underlying mechanisms for this improvement are unclear.

There were no consistent significant effects of housing system on sensory quality of cooked loin and fat samples and on the presence of compounds associated with boar taint (e.g. skatole).

During simulated retail display the colour saturation of fresh loin was increased in samples from pigs finished in the fully slatted system. Although these observations reached statistical significance (0.001<P<0.05), the differences were small and unlikely to be of commercial importance.

STANDARD OPERATING PROCEDURES (SOPs) FOR LIQUID FEEDING

A series of detailed laboratory based in vitro and in vivo studies were completed on parameters that could influence the efficacy of liquid feeding under field conditions. These included work on:

Dry content of liquid dietsParticle size of dry ingredients in liquid dietsIn vitro enzyme treatmentFermentation of liquid dietsLow protein liquid diets

Results

Key findings of SOP studies are as follows.

Dry content of liquid diets

Intake and growth study

Four liquid feeding treatments containing 89, 178, 267 and 356g oven dry matter (ODM) per kg liquid diet (or 100, 200, 300 and 400g dry ingredients/kg liquid diet) were evaluated in a growth study using 288 pigs fed ad libitum at target live weights of 35, 55 and 75kg.

The pigs’ ability to compensate intake was significantly affected when the dry matter content of the liquid diet was reduced to 89g/kg by increased additions of water. As a consequence gain was reduced from 802 to 461g/day and increased feed wastage resulted in a loss of FCR which avaraged 7.03:1 (kg feed dry matter per kg gain) at the lowest dietary dry matter content.

Although drinking water use decreased as the water content of the liquid feed was increased, total water use (water in liquid feed plus drinking water) increased with increased dilution of the dry matter fraction of the liquid diet averaging 6.66, 7.21, 11.35 and 18.21 litres per pig/day at 356, 267, 178 and 89g ODM/kg liquid diet respectively.

Digestibility study (1)

SID 5 (2/05) Page 14 of 47

Five feeding treatments containing 174, 217, 261, 348 and 869g ODM per kg liquid diet (or 200, 250, 300, 400 and 1000g air dry ingredients/kg diet) were subjected to a digestibility study using 2 pigs per treatment housed in metabolism crates either restrict or ad libitum fed.

Dry matter intake and the digestibility of dietary components (dry matter, gross energy, protein, oil and fibre) were not significantly influenced by the dry matter content of the diets over the range studied (174 to 869g ODM per kg diet). However urine output increased by about 0.7 litres per pig/day for every 10g reduction in the dry matter content of the liquid diets. Faecal output (fresh or dried) was not affected by the dry matter content of the liquid diets.

Implications

These studies indicate that the dry matter of liquid diets can vary over a wide range (17 to 35%) without significantly affecting growth performance and digestibility. However the volume of urinary waste will be increased as dry matter content falls and at very low dry matter levels (e.g. < 9%) the ability of pigs to adjust their intake and achieve normal levels of dry matter and nutrient intake will be seriously compromised resulting in a loss of growth performance and feed conversion.

Particle size of dry ingredients in liquid diets

Particle size distribution and effects of steeping

Liquid feed has to be pumped through pipes for delivery at feeding troughs. The dry matter content of the diet and the distribution of particle size affect a number of physico-chemical properties of the process. The dry matter content determines the initial power needed to start a pump from the stationary position and also the power needed to circulate the feed through the system. These parameters are also affected by the size of particles making up the dry matter fraction. A further consideration is that particle size affects the initial homogeneity of the diet and the extent to which separation occurs during pumping and after delivery to feeding troughs. These changes could have a significant effect on the viscosity of the feed and hence the dry matter content that could be pumped at a particular power loading.

Studies were conducted to:

Determine the extent to which particle size distribution was changed by different disk mill settings.Examine the change in particle size distribution resulting from steeping the cereal component in water for 24h.

Samples of barley and wheat were prepared using a disk mill. A disk mill consists of a stationary plate and a revolving plate that can be set at varying distances from one another. The settings range from 0 closest to 10 widest setting and produce grist with different particle size. In this study cereal samples were prepared using setting points of 0, 0.25, 0.75 and 1.0, and retained for particle size analysis.

Particle sizes were not normally distributed, but highly skewed towards the larger particle sizes. Irrespective of grind setting, the majority of particles (64 to 88%) were greater than 500m and 82 to 94% were greater than 250m.

For both barley and wheat, steeping reduced the number of particles in the 250 to 2000m range but significantly increased the proportion of particles in the 10 to 250m range. Modulus of fineness decreased (more particles of a smaller size) as a result of steeping.

Steeping produced different effects on the viscosity of wheat and barley. Steeping the <1000m milled barley fraction for 24h had little effect on viscosity. In contrast, steeping the <1000m milled wheat fraction for 24h resulted in a gradual reduction in viscosity for the 63 to 250m, 250 to 500m and 500 to 1000m fractions. The viscosity of the <63m fraction remained relatively constant.

The results obtained suggest that steeping wheat particles <1000m, for as little as 2h, could be beneficial in reducing viscosity.

Digestibility study (2)

A basal dry cereal based diet supplemented with soya bean meal and minerals and vitamins was formulated for growing/finishing pigs. The cereal fraction or the whole diet was hammer milled using different screen sizes (5mm, 4mm, 3mm, 2mm or twice through 2mm) giving 10 dietary treatments. Treatments were subjected to a digestibility study using 10 pigs housed in metabolism crates and fed twice daily diets in liquid form at a fixed water to meal ratio of 2.3:1.

SID 5 (2/05) Page 15 of 47

There were no significant treatment effects on feed intake, total water use (feed and drinking) and faecal and urine output. There were no clear indications that digestibility of dietary components (dry matter, energy, protein, oil and fibre) improved with decreasing screen size. Individual treatment effects on the digestibility of dietary components were statistically significant but the magnitude of the effect was very small and unlikely to be of practical importance. There were no significant treatment effects on nitrogen and energy retention.

Implications

In vitro studies illustrated the difficulty of describing particle size distribution in feed ingredients and feeds. Particle size was not normally distributed but highly skewed towards the larger particle sizes. Irrespective of grind setting, the majority of particles (64 to 88%) were greater than 500m and 82 to 94% were greater than 250m.

There were some unexpected findings. Despite having a more fibrous pericarp, particle size distribution in barley was more influenced by grind setting than wheat. Although hydration led to an increase in the size of small particles it also reduced the apparent proportion of large particles in the sample. The implication of this observation is that a proportion of small and medium sized particles are loosely attached to large particles and that these disaggregate during steeping, either as a result of a simple physical process, or as a result of enzymic action. Results suggest that steeping wheat particles <1000m, for as little as 2h, could be beneficial in reducing viscosity. This in turn could reduce the power required to move the material around liquid feeding systems.

The digestibility of dry dietary components in liquid diets remained relatively constant over a wide range of hammer mill screen sizes. This is in contrast to research on the effects of particle size on the digestibility of diets fed in dry form. Liquid feeding appears to overcome any losses in digestibility associated with increased particle size when dry ingredients are pre-milled using high screen sizes (e.g. 5mm).

Liquid feeding therefore offers the potential to reduce power expenditure during milling without a loss of digestibility, nitrogen and energy capture and feed conversion in growing and finishing pigs. These results point to potential environmental and cost benefits associated with reduced processing requirements of dry feed ingredients used in liquid diets for growing and finishing pigs.

In vitro enzyme treatment

In liquid diets endogenous enzymes (e.g. phytases and xylanases) in feed ingredients may be activated resulting in the hydrolysis of dietary components (e.g. phytate/phytic acid and cell was non-starch polysaccharides). This may improve diet digestibility and nutrient retention. It is important to know the optimum pH, time and temperature conditions for endogenous enzyme activity in order to assess the contribution this activity may have in liquid feeding systems and the need if any, for supplementation with commercially available feed enzymes.

Phytase

The objectives were to determine:

The optimum pH for the endogenous phytases of wheat and barley.The dose response to exogenous phytase(s), in liquid feed, at pH and temperature optima provided by the manufacturers.The effect that temperature has on the activity of commercial phytases (Natuphos and Ronozyme) in selected raw materials.The effect of temperature on the activity of the endogenous phytases of wheat and barley.The effect of particle size on the release of phosphorus from phytic acid/phytates.Mineral availability in specified raw materials steeped at optimum conditions, overtime with and without exogenous phytase.Potential synergy between raw materials.

The soluble phosphorus content of wheat and barley averaged 0.26 and 0.19g/kg. This increased (P<0.001) to 1.33 and 1.78g/kg for wheat and barley respectively after steeping at 30oC for 6h under different pH treatments (4.0, 4.5, 5.0, 5.5 and 6.0). The pH had no significant effect (P > 0.05) on the amount of soluble phosphorus released. The endogenous phytase in barley tended to perform better at pH 4.5 whilst the endogenous phytase in wheat performed best at pH 5.

The rate of release of soluble phosphorus from wheat steeped at 30oC in response to increasing doses (0 to 1200 phytase units or FU/kg air dry cereal) of exogenous phytase enzymes (Natuphos and Ronozyme) was time and dose dependent. There was a significant dose response to the addition of Natuphos (pH 5) in the first 1 to 2h of steeping. However, after 6h steeping all the phosphorus present was in the soluble form with, or without, the addition of (Natuphos). There was a significant dose response to the addition of Ronozyme (mixture corrected to

SID 5 (2/05) Page 16 of 47

pH 4) over 6h. At an inclusion rate of 1000 FU/kg, and after 6h of steeping 100% of phosphorus was in the soluble form whereas only circa 60% was soluble in the absence of exogenous enzyme.

There was a synergistic effect of steeping temperature (20, 30, 40 and 50oC) and the presence of exogenous phytases (Natuphos and Ronozyme at 500 FU/kg air dry cereal) in the rate of soluble phosphorus release from wheat and barley during the first hour of steeping, but after 6 to7h over 95% of the total phosphorus in wheat and around 60 to 80% of the total phosphorus in barley was released by endogenous wheat and barley phytases at all temperatures.

The rate of release of soluble phosphorus from soya bean meal (which has no endogenous phytase) steeped at 30oC in response to increasing doses (0 to 1000 FU/kg air dry soya meal) of exogenous phytase enzymes (Natuphos and Ronozyme) was time and dose dependent. Maximum release was achieved after 6h of steeping using 1000 FU/kg of exogenous phytase when soluble phosphorus accounted for 50% of the total phosphorus content of soya bean meal. In the absence of exogenous enzymes soluble phosphorus levels remained constant at around 10% of total phosphorus content.

Grind size had no significant effect on the rate of soluble phosphorus release in wheat. In barley, increasing the fineness of grind increased the soluble phosphorus released by either exogenous or endogenous phytase by circa 20 percentage units after 6h.

The release of minerals (iron, manganese, copper and zinc) from rapeseed and soyabean meal was not significantly different in samples steeped with, or without, exogenous enzymes after 6 to7h of incubation.

Co-steeping soyabean meal with wheat, wheatfeed or pelleted wheatfeed significantly (P<0.01) increased the rate of release of soluble phosphorus. After 7h steeping with exogenous enzyme, or with wheat, wheatfeed or a combination of wheat and wheatfeed, between 93 and 97% of the total phosphorus present in soyabean was in the soluble form.

Implications

Exogenous commercial feed grade phytases may be strategically used for the rapid release of soluble phosphorus from milled cereals and protein rich supplements such as soya bean and rapeseed meal used in liquid diets if the feed mixture is steeped for 1 to 2 hours at 30oC before feeding. However endogenous phytases found in cereal and cereal by-products, especially in wheat and wheatfeed are highly effective in releasing over 95% of the total phosphorus if the feed mixture is allowed to steep for around 6 hours at 30oC. Phosphorus release is reduced under lower temperatures due to a decrease in the activity of both endogenous and exogenous enzymes, though in wheat endogenous phytase remains highly effective when steeping temperature is reduced to 20oC.

Non-starch polysaccharide (NSP) degrading enzymes

In vitro research on the use of feed grade NSP enzymes (xylanase and ß-glucanase) to reduce viscosity and increase the concentration of NSP derived sugars in liquid cereal mixtures proved inconclusive with no consistent and major measurable benefits. With the approval of Defra, research effort was redirected to protein reduction of liquid diets using synthetic amino acids (see below).

Fermentation of liquid diets

In vitro studies

Studies were completed to cover the following objectives.

To assess the effectiveness of organisms currently registered as probiotics as inoculants for the production of fermented liquid feed.To assess the effect of the provision of a trace mineral and vitamin supplement on the production of lactic acid by selected organisms.To assess the effect of liquid or freeze dried starter cultures on lactic acid production in non- sterile liquid wheat.To assess the effect of the addition of a protein balancer on the lactic acid concentration of liquid wheat fermented with Bactocell.Following the addition of protein balancer to fermented cereal to determine the extent and rate of exclusion of introduced pathogens.To determine the effect of particle size on the time-course of the fermentation process.To assess the variation in the natural fermentation of wheat and barley.To assess the effect of additional fermentable sugars (molasses) on natural fermentation.To assess the variation in natural fermentation of wheat and barley after storage.

SID 5 (2/05) Page 17 of 47

The key findings from these studies are presented below.

Three commercial products registered as probiotics, namely Bactocell (Pediococcus acidilactici), Biacton (Lactobacillus farciminis) and Gardion (Enterococcus faecium and Lactobacillus casei), and four organisms constituting the product Stabisil (Pediococcus acidilactici, Pediococcus pentosaceus, Lactococcus lactis and Lactobacillus plantarum) were investigated as possible inoculants for fermenting milled wheat mixed with water at a ratio of 1:2.5 parts. Fermentation were conducted at 30, 35 and 40oC with or without the addition of molasses (1.6% v/v of the liquid mix).

Lactobacillus plantarum (Medipharm) produced the most lactic acid (137mmol). Pediococcus acidilactici (Bactocell), Lactobacillus farciminis (Biacton), Pediococcus acidilactici (Medipharm) and Pediococcus pentosaceus (Medipharm), all produced in excess of 100mmol lactic acid within 24h. Lactic acid production was greatest at 35º C and acetic acid production was below 20mmol. Overall addition of molasses to liquid wheat significantly increased lactic acid production in 24h. On the basis of these results, two individual organisms were selected for further study, Pediococcus acidilactici (Bactocell) and Lactobacillus farciminis (Biacton). Both organisms are registered as probiotic feed additives and both produced > 100mmol lactic acid per litre in the presence and absence of molasses. With these organisms lactic acid production was greatest at 35o C, with Pediococcus acidilactici (Bactocell) producing 124 and 136 mmol lactic acid per litre and Lactobacillus farciminis (Biacton) producing 129 and 134mmol lactic acid per litre in the absence and presence of molasses respectively. The differences in lactic acid production due to molasses treatment for these organism were not statistically significant.

The addition of a vitamin and mineral mix to liquid wheat had no significant effect on the production of lactic acid by Pediococcus acidilactici (Bactocell) or Lactobacillus farciminis (Biacton).

After 24h incubation, there was no significant difference in lactic acid concentration in feeds inoculated with freeze dried or liquid starter cultures of Bactocell, or a liquid starter culture of Biacton. However, the initial generation of lactic acid was more rapid in feeds inoculated with Bactocell (liquid or freeze-dried) than Biacton. Acetic acid concentrations in feeds inoculated with starter cultures of Bactocell and the liquid starter culture of Biacton remained below 20mmol.

The addition of a protein balancer to fermented liquid wheat reduced the lactic acid concentration of the mixture. The reduction in lactic acid concentration was minimised by the use of a cereal to water ratio of 1: 2.5 parts for the production of fermented liquid cereals and addition of water with the protein balancer in order to produce a liquid feed that can be pumped.

There was no difference in lactic acid produced by fermenting liquid wheat samples made from wheat ground at a disc mill setting of 0 compared with a disc mill setting of 1.0. This suggests that regardless of the grind size there is sufficient carbohydrate available in liquid wheat to effect an efficient fermentation.

Enteropathogens were rapidly excluded from liquid feed containing fermented wheat using Bactocell. With a contamination level of the order 101 and 103 CFU/ml feed, it is estimated that it would take between 30 and 90 minutes for the whole population of Salmonella or E.coli to die.

Natural fermentation of cereals

One hundred samples of cereal grains, comprising 56 samples of wheat and 44 samples of barley, taken at harvest, were obtained from locations across the UK. The type of grain, variety, location and weather conditions at harvest were recorded by the producer.

Lactic acid bacteria counts were 0 to 4.89 Log10 CFU/ml for wheat and 0 to 5.00 Log10 CFU/ml for barley.

Lactic acid content in liquid mixtures of milled wheat and barely averaged 22.2 and 34.4 mmol /litre respectively (target for bacteriostasis is 75 mmol/litre). Only nine of the 300 fermentations conducted produced more than 75 mmol/litre lactic acid after 24h fermentation. Even after 72h fermentation only 33% exceeded 75 mmol/litre. The wheat samples had a significantly (P < 0.001) higher pH (4.53) than barley samples (4.30) after fermentation.

Acetic (>20 mMol) and butyric acid are undesirable in fermented liquid feed. The concentration of acetic and butyric acids averaged 22.4 and 15.3 mmol/litre respectively for wheat and 27.0 and 19.5 respectively for barley.

Ten percent of the cereal samples were also fermented following the addition of 2% molasses (v/v) and/or a vitamin and mineral supplement. None of the additions had any significant effect on the mean production of lactic acid, acetic acid, butyric acid or ethanol after 72h incubation.

SID 5 (2/05) Page 18 of 47

The fermenting ability of six cereal samples was determined at harvest and after six months storage. Although differences were not statistically significant, mean lactic acid production declined from 65.4 to 21.8 mmol/litre after six months storage.

Digestibility study (3)

A basal dry cereal based diet supplemented with soya bean meal and minerals and vitamins was formulated for growing/finishing pigs. The milled cereal fraction or the whole diet was fermented separately using an inoculation of Pediococcus acidilactici (Bactocell) at the rate of 0.1g/kg dry cereals. Fermentations were held at 35oC for either 12, 24 or 48 hours. Fermented cereals were mixed with the remaining non-cereal liquid fraction (fullfat soya bean and rapeseed meal, wheatfeed, amino acids and minerals and vitamins) of the balanced diet either immediately (0h) or 3 hours (3h) before feeding. These six treatments were compared with a positive and negative control in which the basal diet was fed immediately after mixing in liquid form with and without Pediococcus acidilactici (Bactocell) supplementation. Treatments were subjected to a digestibility study using 8 pigs housed in metabolism crates and fed twice daily diets in liquid form at a fixed water to meal ratio of 2.5:1 parts.

Results showed that fermentation of the cereal fraction significantly improved the digestibility of gross energy (P<0.05), nitrogen (P<0.05), starch (P<0.01) and oil (P<0.05), with the greatest improvement from 24 hours of fermentation. Mixing the fermented cereal and non-cereal fraction of the diet either 0h or 3h before feeding had no significant effect on the digestibility of dietary components. There were no significant differences between the negative and positive control treatments.

Implications

The data indicated that natural fermentation is inconsistent and cannot be relied upon to produce a liquid pig feed that is safe and palatable.

A commercial strain of Pediococcus acidilactici (Bactocell) registered as a probiotic feed additive was effective in producing desirable levels of lactic acid (e.g. > 100 mmol lactic acid per litre) from fermented liquid cereals under 35oC for 24 hours. Digestibility studies showed that the fermentation of the cereal fraction under these conditions had the potential to improve the digestibility of energy and nitrogen in liquid diets fed to growing and finishing pigs.

These laboratory based studies indicated that the controlled fermentation of the cereal fraction of liquid diets with a registered strain of Pediococcus acidilactici (Bactocell) holds potential for controlling undesirable organisms (E. coli and salmonella) and improve the digestibility and hence efficiency of grower/finisher pig production. This potential was investigated under field conditions in Study 3 (see above).

Low protein liquid diets

The use of synthetic amino acids to idealise protein in diets with reduced protein levels has been investigated widely in pigs fed dry feeds. Such diets can offer environmental benefits in terms of reduced nitrogen emissions without adversely affecting pig performance and cost of production (depending on the relative cost of soya protein and synthetic amino acids). Low protein diets supplemented with amino acids have not been evaluated under liquid feeding, where there may be risk of amino acid degradation from microbial activity and loss of protein quality. Studies were conducted to investigate the stability of naturally present amino acids in a liquid co-product from the distillery industry (Greenwich Gold) which is commonly used in commercial pig production and to determine nutrient digestibility in low protein liquid diets fed to growing and finishing pigs.

Stability of amino acids in Greenwich Gold

The stability of amino acids in Greenwich Gold was investigated in samples held over 7 weeks at either 4 or 18oC.Weekly determinations showed that stability was greater at 4oC but there was no major loss of amino acids (e.g. lysine, methionine and threonine) at 18oC that would be detrimental to the feeding value of the co-product. The amino acid content was in line with expected values for lysine and threonine and methionine plus cystine.

Digestibility study (4)

A control liquid diet was formulated without synthetic amino acids to meet the ideal protein requirements of growing/finishing pigs. This resulted in a crude protein content of 219g/kg dry matter (DM). Crude protein was reduced in the low protein diet to 177g/kg DM by the use of synthetic amino acids (lysine, methionine, threonine and tryptophan) whilst supplying an equivalent level of standardised ileal digestible ideal protein. Diet formulations were identical to those evaluated under field conditions during Study 2 conducted at MLC’s Stotfold Pig Development Unit. The diets were evaluated for nutrient digestibility using four pigs housed in metabolism crates.

SID 5 (2/05) Page 19 of 47

Digestibility of nitrogen (crude protein) was significantly reduced in the low protein diet (0.816 vs. 0.857; P<0.01) but there were no significant differences in nitrogen retention (0.524 control vs. 0.553 low protein; P = 0.086). There was a trend towards reduced digestibility of gross energy in the low protein diet (0.840 vs. 0.861; P=0.105) resulting in a significant reduction in digestible energy (DE) content (14.52 vs. 15.63 MJ DE/kg DM; P<0.01)). This difference in DE content (1.11 MJ/kg DM) was greater than the expected formulation difference of 0.45 MJ DE/kg DM. Although there were differences in the formulated and determined DE values for the two diets, they were formulated to supply equivalent levels of net energy (NE) at 11.09 MJ NE/kg DM. The total lysine and threonine contents were around 10% and 15% below the formulated targets for the control and low protein diets respectively and the total methionine content was around 15% lower than target for both diets.

Implications

Greenwich Gold can be stored at ambient temperature for several weeks without a detrioration of amino acid content. However the product may be subjected to much higher temperatures as the product leaves the distillery plant at a high temperature and this decreases slowly during on farm storage and use. Further work is required in this area. The digestibility study points to a reduction in gross nutritional value (energy and crude protein) of the low protein diet but this does not imply that there are differences in net energy (NE) and protein value given the diets were formulated to supply equivalent levels of NE and ideal protein using standardised ileal digestible coefficients for essential amino acids.

MATERIALS AND METHODS

Production studies

Details are presented in the individual published reports on each of the four studies attached in Appendix 1. The methodology for Study 1 given below was repeated for Studies 2, 3 and 4 and any modifications were presented in the individual reports. Research objectives covering meat quality were restricted to Studies 1, 3 and 4 and those covering environmental impact were restricted to Studies 1, 2 and 4.

Standard operating procedures

Brief commentary on approach has been presented along with the key findings as above. Detailed methodology can be found in the individual reports attached in Appendix 1.

Materials and methods covering production studies

Pig production

Housing

The Finishing Systems Research Unit (FSRU) located at MLC’s Stotfold Pig Development Unit consists of a Feed Centre, which manufactures and delivers liquid feed to growing/finishing pigs in two contrasting systems of production, fully slatted v straw based housing.

SID 5 (2/05) Page 20 of 47

Liquid co-product storage tanks

Straw-based house

Fully-slatted house

Feed centre

Each house consists of four rooms and each room contains four pens. Both houses have a reception area with pig weighing facilities, ventilation controllers and storage for clothing and other equipment. The drawings for the two housing systems and pen layouts are further details can be found below.

Feed centre

The feed centre houses a dry feed storage and milling unit and a state-of-the-art liquid processing and feeding system (see below for further detail).

The dry feed storage holds six ingredients in cloth silos, whole grain wheat and barley, pelleted wheatfeed, soyabean meal and rapeseed meal. Wheat, barley and wheatfeed are hammer milled into small cloth bins (fitted with load cells) before transfer by auger to the central processing tank of the liquid feeding system. The dry feed storage houses 3 small rip and tip bins for mineral and vitamin supplements and fishmeal. These bins also deliver material to the central processing tank by independent augers.

The liquid feeding system was supplied by Meyer Lohne (Germany), and designed in association with MLC to meet the requirements of the research programme. It consists of 6 tanks fitted with load cells, with a central processing tank as the focal point for the intake, proportioning and blending of all individual ingredients (dry and up to 4 liquid co-products, plus soya oil) to any formulation specified in the computer controlled Winfeed programme. The central processing tank can heat materials to 95oC and then chill before transfer to one of 5 different tanks. Depending on protocol, the blend can be transferred into one of two fermentation tanks, as in the case of a cereal mix, or into a protein tank as in the case of a protein rich blend. These can then be transferred independently to one of two feeding tanks linked to the dual pipeline system delivering feed to the two pig finishing buildings. All tanks have the facility for micro additions of dry and liquid products, such as enzymes and organic acids and the processing and fermentation tanks are equipped with pH control systems.

The dual pipeline arrangement can phase-feed to deliver different proportions of the two mixes held in the feeding tanks according to the nutritional requirements of each pen group of pigs.

Feeding

Pigs were fed a growing and a finishing diet either in dry pelleted or in liquid form (Study 1). The liquid and dry diets were formulated to similar nutrient specification and using similar feed ingredients (see Table 1).

SID 5 (2/05) Page 21 of 47

Table 1 Meal equivalent formulations and nutrient specifications of dietsGrower diets Finisher diets

Liquid Dry Liquid DryIngredient (%)Wheat 41.30 38.25 38.62 35.15Barley 13.77 12.65 12.87 11.75Wheatfeed 10.00 10.00 20.00 20.00Soya bean meal (HP) 18.37 2.00 10.05 ---Fullfat soya --- 21.50 --- 12.50Rapeseed meal 5.00 5.00 10.00 10.00Fish meal 2.50 2.50 --- ---Fat blend --- --- --- 2.50Soya oil 5.96 2.00 5.36 2.00Molasses --- 3.00 --- 3.00Minerals and vitaminsa 3.10 3.10 3.10 3.10

Nutrient Specificationb

DE (MJ/kg) 14.75 14.72 14.13 14.22Lysine (%) 1.22 1.17 0.94 0.89Ca (%) 0.79 0.90 0.73 0.83P (%) 0.68 0.67 0.59 0.56Na (%) 0.40 0.42 0.40 0.40

a Mineral and vitamin supplement provided per kg of grower diet: Vitamins A, D and E 9000, 1500 and 75 iu respectively, Vitamin K 1000 ug, Riboflavin 5 mg, Pyridoxine 6 mg, Cyanocobalamin 45 mg, Biotin 100 ug, Pantothenic acid 24 mg, Niacin 23 mg, Copper 175 mg, Zinc 85 mg, Manganese 35 mg, Iron 40 mg, Iodine 1.75 mg and Selenium 0.4 mg. Mineral and vitamin supplement provided per kg of finisher diet: Vitamins A, D and E 6000, 1500 and 70 iu respectively, Vitamin K 1000 ug, Riboflavin 2 mg, Pyridoxine 4 mg, Cyanocobalamin 30 mg, Biotin 66 ug, Pantothenic acid 16 mg, Niacin 15 mg, Thiamine 0.5 mg, Copper 100 mg, Zinc 80 mg, Manganese 25 mg, Iron 100 mg, Iodine 0.5 mg, Cobalt 0.5 mg and Selenium 0.45 mg. The supplements also provided per kg of grower diet: 1.74 g lysine as lysine hydrochloride, 9.10 g salt, 6.04 g Calcium and 1.53 g Phosphorus; and per kg of finisher diet 1.27 g lysine as lysine hydrochloride, 9.75 g salt, 6.67 g Calcium and 0.69 g Phosphorus.b Differences in expected nutrient specifications between dry and liquid formulations are due to differences in MLC and Farm Nutrition matrix values for individual ingredients.

Dry feed was manufactured (Farm Nutrition, Provimi Ltd) in 3mm pellets and delivered in 25kg bags. The feed was manually weighed and tipped into ad libitum hoppers. Feed was offered ad libitum. To estimate intake and feed conversion ratios (feed intake/body weight gain), residual feed was weighed when pigs were weighed and deducted from feed inputs.

The liquid feed was manufactured on site in the Feed Centre (see below) according to formulations given in Table 1. The formulations were adjusted according to sample analysis of newly delivered batches of major individual feed ingredients. Water was pre-weighed into the Central Processing Tank (Tank 3) to produce liquid feed (grower or finisher) in batches of around 370kg with a 20% target dry matter content.

Whole grain wheat and barley and wheatfeed pellets were hammer milled (4.5 mm screen) and transferred with other feed ingredients to the Central Processing Tank.

The components were mixed and transferred to Tanks 5 and 6 (grower) and 4 (finisher) for temporary storage at ambient temperature before delivery in complete batches to feeding Tanks 1 (A) and 2 (B) respectively. This process was led by feed demand at the troughs using sensors, which signalled for refill on empty. Troughs were refilled with 15kg drops of either grower or finisher feed according to growth stage of each pen group (when mean weight of pigs in pen exceeded 60kg, they were transferred from the grower to the finisher diet). Liquid feed was available ad libitum except during 24:00 and 01:00 when the system was automatically paused, allowing pigs to clear troughs of any accumulated residues.

Feed sampling and laboratory analysis

Each newly delivered batch of individual feed ingredient and compounded dry pelleted feed was sampled and dispatched (Sciantec Analytical Services Ltd., North Yorkshire, England) for nutrient analysis. The results from the analysis of feed ingredients were used to adjust formulations in the production of liquid diets to meet target DE and total lysine content.

Grower and finisher liquid diets were sampled during the course of the trial from Feed Tanks A and B for on site determination of oven dry matter content (24hrs at 100oC), and pH (Hanna Instruments HI 991000). Weekly liquid diet samples were stored for subsequent dispatch and laboratory analysis for nutrient content (Sciantec Analytical Services Ltd., North Yorkshire, England).

A summary of samples taken and associated laboratory analysis are given in Table 2.

SID 5 (2/05) Page 22 of 47

Table 2 Feed and feed ingredients sampled and their laboratory analysis

Dry

mat

ter (

DM

)

Cru

de p

rote

in (C

P)

Oil

(B)

Neu

tral d

eter

gent

plu

s am

ylas

e fib

re (N

DA

F)

Ash

Tota

l lys

ine

Cal

cium

(Ca)

and

pho

spho

rus

(P)

Sod

ium

(Na)

Cop

per (

Cu)

Fatty

and

vol

atile

aci

d pr

ofile

s,

etha

nol a

nd p

H

Feed ingredientWhole grain wheat Wheatfeed pellets

Whole grain barley Rapeseed meal

Soya bean meal (HP) Fish meal

Grower mineral and vitamins premix Finisher mineral and vitamins premix

Complete liquid diet Complete dry pelleted diet

DM, Oil (B), Ash, Ca, P, Na and Cu were determined according to the methods described by MAFF (1982). CP was determined by nitrogen gas analyzer (Leco FP – 528) using induction furnace and thermal conductivity. NDAF was determined according to the method described by MAFF (1993). Total lysine content in feed samples was analysed according to Liu et al. (1995). The component fatty acid content of the lipid fraction were determined according to ISO (1995) and ISO (2001). Volatile fatty acid content, lactic acid and ethanol in liquid feed samples were analysed according to Fussel and McCalley (1987). pH was measured using a pH probe (Hanna Instruments HI 991000).

Animals

A total of 1056 (Large White x Landrace) x Large White pigs weighing between 30 to 40kg were received in 8 equal batches of 132 over 11 weeks commencing 12th April 2002. Pigs were delivered on a Friday and transferred to a room in the finishing systems housing according to the pattern given below.

Room 1 2 3 4Batch 7 Batch 5 Batch 3 Batch 1

Strawbased

Pen 2 Pen 4 Pen 6 Pen 8 Pen 10 Pen 12 Pen 14 Pen 16Pen 1 Pen 3 Pen 5 Pen 7 Pen 9 Pen 11 Pen 13 Pen 15

Room 1 2 3 4Batch 8 Batch 6 Batch 4 Batch 2

FullySlatted

Pen 31 Pen 29 Pen 27 Pen 25 Pen 23 Pen 21 Pen 19 Pen 17Pen 32 Pen 30 Pen 28 Pen 26 Pen 24 Pen 22 Pen 20 Pen 18

Pig identification and weighing

On the following Monday, two pigs were randomly identified for slaughter for the baseline assessment of gut microbial status and two surplus pigs were randomly selected and removed to alternative accommodation. The remaining 128 pigs were ear tagged for individual identification, weighed and then sorted by weight, from lightest to heaviest. The batch was divided into 4 equal groups of 32 pigs in order of weight: Light Light (LL), Light Medium (LM), Medium Heavy (MH) and Heavy Heavy (HH). Each group was randomly allocated to one of four pens in the room.

Pigs were weighed at two weekly intervals for growth performance monitoring and to establish the timing (week 6) at which pen weight within room averaged 60kg for stocking density reduction and the switch from grower to finisher diet. On week 6 pigs were weighed and stocking density in the fully slatted and straw based systems was

SID 5 (2/05) Page 23 of 47

reduced from 32 to 25 and 20 pigs per pen respectively. Pigs removed were pre-selected to represent the range (minimum and maximum) and average weight in the pen so that the overall distribution of individual weights was not potentially skewed by random selection.

Pigs were weighed 9 days prior to slaughter and those weighing more than 95 kg were selected for slaughter so that weight at slaughter could be as close as possible to a target of 105kg. Pigs were weighed again the day before slaughter for determination of end point liveweight.

Pigs selected for slaughter were slap marked for individual identification and this was recorded on the slaughter sheets with corresponding information for pen number, pig ear tag number, sex, dispatch weight and pre-selection code for focal pigs (see Health and Welfare Monitoring). Slap marking clearly distinguished focal pigs from other trial pigs.

Management

Pigs were managed according to standard protocol covering the following key responsibilities:

The day before entry the environmental control system was set up to achieve target temperature and ventilation rates (see below: Ventilation system).

Staff changed overalls and wellington boots before entering each building. Foot dips (Virudine, Antec International Ltd., Suffolk) were used on entry and exit. Separate foot dips were used for each building and these were cleaned and replenished twice per week.

Pigs in the slatted building were provided with chains and plastic mats were used in the first two weeks in the fully slatted house to reduce draughts in the lying area.

Pigs were inspected twice daily for signs of ill health and welfare. Appropriate action was taken to manage pigs with health conditions. Health conditions were recorded and appropriate action was taken to safe guard the welfare of each pig. Treatments were carried out according to veterinary recommendations and all veterinary treatments were documented. Pigs that did not respond to veterinary treatment and where welfare was at risk, were weighed and removed from the study and recorded with date.

All deaths and culls for health conditions were recorded and a post-mortem was conducted by a Veterinary Surgeon to assess the cause of death.

In each room, the ventilation system was checked and temperature readings were taken and recorded using a hand held probe during the morning inspection of animals every Monday, Wednesday and Friday. Readings were taken at pig level in the middle of each of 4 pens in each room.

Soiled bedding was removed daily from each pen. The straw based system was scraped out daily (farm scraper was cleaned and disinfected (Virkon S, Antec International Ltd., Suffolk) before use to limit microbial cross contamination) and fresh straw was made available at the rate of about a third of a bale per pen per day.

All nipple drinkers were inspected daily to ensure satisfactory water flow. Feed troughs and probe sensors were inspected twice daily, any contaminated feed, faeces and straw were removed. Central passages were cleaned daily.

Straw and water use and power consumption

Straw use was monitored by recording the number of bales used daily and a sub-sample of bales were weighed during the trial as an estimate of total weight used.

Drinking water used within each room and water used for cleaning each finishing building was measured using 12.5 mm water meters (Minomess, Apartment Water Meters, Minol Messtechnik, Germany).

Power consumption in the Feed Centre for the milling, material transfer, mixing, circulation and delivery of liquid feeds to troughs was measured using electric meters (RDL Power Rail 303). Power consumption for heating and ventilation of each finishing building was measured using electric meters (RDL 3P100ANL Power Meters).

Waste production

In the fully slatted house, the volume of effluent produced by each pen group of pigs was measured by dipstick at four locations within the pen area. Measurements were taken before and after emptying to estimate difference in

SID 5 (2/05) Page 24 of 47

surface fall (h). Slurry pit width (w), length (l) and h were used to calculate the volume of effluent produced by each pen group as w x l x h.

Manure production from the straw based system was weighed daily by scraping waste onto a trailer fitted with weight cells. The weight of manure produced was estimated by taking load cell readings before and after loading with waste. It was not practical to estimate waste production at the pen level due to the use of common scraping passages either side of the central passage (see below, buildings).

Pig health and welfare

Standard management procedures for monitoring and safeguarding the health and welfare of pigs on trial were complemented by more detailed observations on sub-samples of pigs (focal pigs).

Following entry and allocation of pigs within batch to pens, 3 pairs (balanced by weight and sex) were selected as focal pigs; one pair at approximately median weight of the group, one pair within the upper quartile weight band and another pair in the lower quartile weight band. Focal pigs remained in the pen after stocking density reduction at around 60kg for subsequent monitoring to finishing and at slaughter. Any focal pig removed or lost from the trial (health, welfare or death) was replaced with a pig of the same sex and approximate start weight.

Blood sampling

Pigs were blood sampled under veterinary supervision (Home office Licence number PPL 70/5367) for the determination of health and Salmonella exposure status at entry, at mid-point in the trial (at stocking density reduction) and during exsanguination at slaughter. The number of pigs sampled and analysis is summarised in Table 3 below.

Table 3 Blood sampling and analysis for health and Salmonella status for Trial 1

Analysis PigsSalmonella ELISA test

All pigs at entry (8 batches x 128 pigs per batch, n = 1024)Focal pigs at mid-point, batches 1 to 4 (4 batches x 24 pigs per batch, n = 96)All pigs at slaughter (n = 720)

Acute Phase Proteins

All focal pigs at start (8 batches x 24 pigs per batch, n = 192)Focal pigs at mid-point, batches 1 to 4 (4 batches x 24 pigs per batch, n = 96)All focal pigs at slaughter ( 8 batches x 24 pigs per batch, n= 192)

Generalised immunity

Focal pigs at start, mid-point and slaughter, batches 1 to 4 (n = 288)

PRRS virus Focal pigs (15 from 24 per batch), batches 1 and 8 only, start, mid-point and slaughter (n = 90)

Health monitoring

All pigs, including focals were subjected to weekly detailed health monitoring. This included records of any incidence of external clinical signs of respiratory, locomotory and enteric disorders, behavioural vices such as tail, flank or ear lesions and any other problems (e.g. hernias, abscesses and rectal prolapse).

The average faecal consistency was scored weekly for each pen on a subjective visual scale of 1 to 5 (1 very loose to 5 very solid).

Skin lesions and cleanliness

Hygiene status of focal pigs was scored weekly by visual estimation of the % whole body surface area which was clean (as opposed to soiled).

Body lesions were counted weekly on each of the focal pigs. Seven areas of the body were assessed: face and ears, neck, shoulders, flank, rump, buttocks and tail. For each area (excluding the tail), lesions on the left and right hand sides were recorded separately.

Bursitis of the hock in focal pigs was visually scored weekly on a subjective scale of 0 to 5 according to Lyons et al. (1995) as follows: 0 no bursitis, 1 small raised swelling, 2 moderate swelling, 3 fairly extensive swelling, 4 very severe swelling and 5 eroded/ulcerated bursa with infection.

SID 5 (2/05) Page 25 of 47

Behavioural recording

Behavioural time budgets for focal pigs were recorded concurrently by time sampling at 10 minute intervals for three 2-hour periods (09.00-11.00, 12.00-14.00, and 15.00-17.00hrs). This was repeated 3 times for each room: in the week of entry, the week prior to stocking density reduction and the week before slaughter. Data were entered onto a standard recording sheet using an ethogram.

A 24-hour video record of behaviour at the feed trough was taken in each of the 3 weeks detailed above. Focal pigs within each pen were uniquely spray marked so that they could be recognised individually.

The number of pigs feeding and queuing for the trough was recorded at 5 minute intervals. The feeding bouts of focal pigs were individually documented. For each bout, the start and end time and method of initiation and termination were recorded.

Slaughter assessments

Foot damage of focal pigs was subjectively evaluated on the slaughter line based on the method described by Lyons et al. (1995). Both claws of the left hind foot were inspected for the presence of white line lesions, false sandcracks, toe erosions, sole erosions, and heel flaps (torn heels). These conditions were scored on a scale of 0 to 3, with 0 no damage and 3 as severe damage.

The hearts and lungs of all pigs were removed and scored for lesions by a Veterinary Surgeon. All seven lung lobes were scored: 4 were scored out of 10, and 3 scored out of 5 to give a maximum total score of 55 for any given pig (Goodwin and Whittlestone, 1979). Lesions on the lobes were scored proportionately; if lesion covered 20% of the surface area of a lobe it was scored as 2. If other conditions were present (for example pleurisy, scarring), they were scored on an increasing severity scale (+,++ or +++). The heart was inspected for evidence of pericarditis and scored on an increasing severity scale (+,++ or +++).

The gastrointestinal tract of focal pigs was scored for gastric ulceration (including hyperkeratosis of the gastric pars oesophagus) according to Potkins and Lawrence ( 1989). Any digesta was removed from the pars oesophagus and its surface condition was subjectively scored on a scale of 0 to 5: 0 apparently normal, 1 hyperkeratosis just beginning, 2 slight hyperkeratosis, 3 moderate hyperkeratosis, 4 severe hyperkeratosis and 5 severe hyperkeratosis and ulcer (active or healed).

The dressed carcases of focal pigs were scored for skin damage using the standard MLC subjective scale of 1 to 5: 1 unblemished to 5 severely blemished.

The leg joints of 2 pairs (equal sexes) of focal pigs per pen were dissected to expose the joint. The extent of osteochondrosis was scored according to Slevin et al. (2001) on a scale of 0 normal (no gross lesion) to 4 generalised dulling, deep depressions and extensive erosion, ulceration or absence of cartilage.

Microbial status

Gut contents on entry and at slaughter, individual feed ingredients, complete diets (dry and liquid), fresh straw, drinking water, fresh faeces, effluent and dust were sampled during the trial for microbial evaluations by the Veterinary Laboratories Agency (VLA, Bury St Edmunds). Sterilised equipment and sample containers were used to avoid cross contamination. Blood samples were taken from all pigs on entry and during exsanguination at slaughter, plus additional samples were taken from focal pigs at mid-point for Salmonella ELISA testing as described by Heijden (2001).

Sampling

Gut contents (ileum, caecum and colon) were sampled from 2 randomly selected pigs per batch on entry (Tuesday) and from all focal pigs at slaughter. Caecal contents were sampled from all pigs at slaughter. Ligated segments of the ileum, caecum and colon were removed and placed in an insulated box with cool packs. Segments were transported to the VLA for sampling.

All feed ingredients (except soya oil) were sampled during delivery to Stotfold Pig Development Unit. Pelleted dry compound diets were sampled from unopened bags from each new batch of delivered feed. Weekly samples of dry pelleted diets from unopened bags and liquid diets (grower and finisher mix) from the feed tanks (Tanks 1 and 2, see below) were taken.

SID 5 (2/05) Page 26 of 47

Bales used for bedding the straw based system were sampled on 9 separate occasions during the course of the trial.

Each room was fitted with a water tap from which samples were taken on 3 separate occasions (on stocking with pigs, mid-point and on emptying). Taps were flushed and cleaned using 70% alcohol prior to sampling.

Pens were sampled for fresh faeces on 3 separate occasions (on stocking with pigs, mid-point and on emptying). Each pen was sampled from 4 locations and these sub-samples were pooled for analysis. Samples were taken from the dunging area and slatted flooring of the straw based fully slatted systems.

Effluent samples were taken weekly from each room within the fully slatted system. The pit under each pen was sampled from 4 locations and pen sub-samples were pooled to represent the room sample for analysis.

Dust samples were taken weekly from each room. Samples were taken from the windowsills along the outer walls of the room using a sterile Petri dish exposed for 2 hours.

Microbial evaluations

The samples were evaluated for the following microbial organisms.

Table 4 Summary of microbial evaluations of samplesS

alm

onel

la

Tota

l Aer

obic

Via

ble

Cou

nt

Tota

l Ana

erob

ic V

iabl

e C

ount

Lact

ic a

cid

bact

eria

Ent

erob

acte

raci

ae

coun

t

Col

iform

cou

nt

Yea

st c

ount

Law

soni

a an

d B

rach

yspi

ra

Gut contents at entry (8 pigs)Ileum

Caecum Colon

Gut contents at slaughterIleum (focal pigs only)

Caecum (focal pigs only) Colon (focal pigs only)

Caecum (all pigs) Feed ingredients Complete feeds Fresh straw Drinking water Fresh faeces Effluent Dust

Table 5 Media and conditions for microbial counts performed on samples

Microbial organism Media and inoculation Supplement Incubation Reference

Salmonella

Buffered peptone,Diassalm semi-solid agar, Rambach agar, Columbia agar, MacConkey agar.

37oC for 18h; 30oC/41.5oC for 24/48h; 37oC for 24h serial incubations

Netten et al. (1991)

Total aerobic bacteria viable count

Plate count agar, pour plate

5% defibrinated sheep blood for gut samples only

Aerobic, for 72 h

ISO (1991)

Total anaerobic bacteria viable count

Wilkins-Chalgren Anaerobe agar, spread plate

5% defibrinated sheep blood

Anaerobic, 37oC for 48 h

SID 5 (2/05) Page 27 of 47

Lactic acid bacteria MRS agar, spread plate

5% CO2, 30oC for 48 h

Mackie and MacCartney (1965)Enterobacteraciae count Violet red bile glucose

agar, pour plateAerobic, 37oC for 24 h

Coliform count VRBL agar, MacConkey broth

Aerobic, 37oC for 18 to 24 h

Yeast Rose Bengal agar Chloramphenicol, Oxoid SR78

Aerobic, 25oC for 4 days

Lawsonia and Brachyspira PCR

Jones et al. (1993), Moller et al. (1998) and Leser et al. (1997).

Limits of detection; for pour plates, 10 cfu per g: for spread plates, 100 per g.

Environmental impact

The table below summarises systems monitoring for the assessment of environmental impact of housing and feeding system during Production Trials 1, 2 and 4.

Table 6 Summary of environmental monitoringParameter monitored Frequencya

Ammonia concentration at the entry to the exhaust fan in each room

Once per hour per room, continuously

Dust concentration at one position in each room

14 occasions, each nominally of 3 days duration, nominally once a week

Ventilation rate of each room Hourly averages, computed from a large number of basic readings

Ammonia emission rate from each room One value per hour per room, continuously. Dust emission rate from each room 14 occasions, each nominally of 3 days duration, nominally

once a week Effluent volume from each slatted room Each occasion of emptying slurry pitEffluent composition from each slatted room Each occasion of emptying slurry pit

Farm Yard Manure production from whole straw based building

Each occasion of cleaning out the straw based building

FYM composition from whole straw based building

Analysis will be carried out on a pooled sample, once every 14 days

aIn trials 2 and 4 dust concentration will be measured at a reduced frequency of once per fortnight during the period when both buildings are fully stocked.

Ventilation rate

Ventilation rate was continuously measured for each room using a fan-wheel anemometer installed in the exhaust duct, between the inlet damper and the exhaust fan (Demmers et al., 1999). The rate of rotation of the fan wheel as well as the opening angle of the inlet damper were continuously monitored electronically, to give an accurate log of the instantaneous ventilation rate of each room. Each fan-wheel anemometer was calibrated by installation in an identical ventilation chimney fitted to a fan test rig at Silsoe Research Institute before the start of Trial 1 (Moulsley and Randall, 1990). This calibration will be checked at the end of trial 4. Any variation in the calibration curve and hence ventilation rate will be accounted for after trial 4.

Ammonia

Ammonia concentration was measured using a chemiluminescence-type nitric oxide analyzer (Demmers et al., 1999), following catalytic conversion of ammonia to nitric oxide at 750˚C. Measurements were taken at 12 locations, at entry to the exhaust fan within each room of the straw based and fully slatted housing and at four placed immediately outside the buildings, to correct for ambient ammonia entering the buildings. The analyzer was calibrated regularly using certified standard gas mixtures.

Net ammonia emission rate for each room was estimated as a function of hourly ammonia concentration and hourly ventilation rate, corrected for any incoming concentration of ammonia. The ammonia emission was

SID 5 (2/05) Page 28 of 47

normalised to the live weight of pigs in each room. The ammonia emission factor (g NH3-N per live weight unit per hour was calculated from the cumulative emission. One live weight unit corresponds to 500 kg live weight.

Dust

Dust concentration was measured at entry to the exhaust fan within each room of the straw based and fully slatted housing (Takai et al., 1998). Dust concentration was also measured at one of the air inlets in the sidewalls of one room to correct for dust entering the buildings.

The average ventilation rate over the period of exposure of each set of dust concentration samplers (normally 3 days) was multiplied by the average dust concentration (corrected for any incoming dust), to give average dust emission rate for each room.

Waste

The volume and weight of waste (effluent and manure) produced were monitored as described on page 36.

Slurry samples were taken immediately before pit emptying for the determination of dry matter, total nitrogen and ammoniacal nitrogen and phosphorus content according to APHA (1985). Analysis was carried out on a pooled sample representing the content of the slurry pits within each room. Four grab samples of manure were taken from the trailer, at randomly selected points, at 14 day intervals for the determination of dry matter, total nitrogen, ammoniacal nitrogen and phosphorus content. The samples were combined, coned and quartered to give a 1 kg sample for analysis.

Meat quality

Carcase measurements and sampling

Pigs were transported to the abattoir and slaughtered (day 1) under commercial conditions and the carcases were conditioned using achilles suspension.

The carcases were graded using the Hennessy Grading Probe for the estimation of carcase lean content using the following EU approved (88/234/EEC) equation:

Y = 62.5 – 0.62X1 – 0.46X3 + 0.16X4

where,

Y = the estimated percentage of lean meat in the carcase,X1 = the thickness of backfat (including rind) in mm, measured at 6 cm off the midline of the carcase at the last rib (commonly known as P2),X3 = the thickness of backfat (including rind) in mm, measured at 6 cm off the midline of the carcase between the 3rd and 4th last ribs (known as ‘rib-fat’),X4 = the thickness of muscle in mm, measured at the same time and in the same place as X3 (known as ‘rib-muscle’).

Carcases were chilled overnight under commercial slaughter house conditions (1oC). The left side of the carcase from 4 out of 6 focal pigs (see selection Table 7) were transported (day 2) to MLC Winterhill House, Milton Keynes, under controlled refrigeration and stored in a chiller at 1oC for subsequent evaluation.

On day 3, firmness measurements were taken of the subcutaneous fat by penetrometer over the shoulder and leg. Subjective fat firmness scores, on scale of 1 to 8 (1= very soft (oily), 8 = very hard), over the shoulder and leg were assessed by depressing the subcutaneous fat with the tip of a finger or thumb.

The sides will were butchered to provide the following samples:

Shoulder fat (200g); vacuum packed and frozen for subsequent analysis of skatole and indole levels.

Shoulder fat (100g); vacuum packed and frozen for subsequent analysis for fatty acid profile.

Single steak (20 mm); from the anterior end of the loin for immediate assessment of drip loss according to the following method. The steak was weighed and suspended in a polythene bag for 24 hours at 1oC. The steak was

SID 5 (2/05) Page 29 of 47

removed from the bag and reweighed. The drip was weighed. The eye-muscle was removed from the steak and weighed. Per cent drip loss was calculated as (weight of drip loss/(weight of drip + weight of eye-muscle))*100.

Table 7 Selection of focal pigs for carcase and meat quality evaluations

Housing system Room Pen Meat quality evaluationsa Chemical analysisb Fully slatted 1 32 2 mixed sex pairs 1 male and 1 female

31 2 mixed sex pairs 1 female30 2 mixed sex pairs 1 male29 2 mixed sex pairs 1 female

2 28 2 mixed sex pairs 1 female27 2 mixed sex pairs 1 male26 2 mixed sex pairs 1 female and 1 male25 2 mixed sex pairs 1 male

3 24 2 mixed sex pairs 1 male23 2 mixed sex pairs 1 female and 1 male22 2 mixed sex pairs 1 male21 2 mixed sex pairs 1 female

4 20 2 mixed sex pairs 1 female19 2 mixed sex pairs 1 male18 2 mixed sex pairs 1 female17 2 mixed sex pairs 1 male and 1 female

Straw based 1 1 2 mixed sex pairs 1 male and 1 female2 2 mixed sex pairs 1 female3 2 mixed sex pairs 1 male4 2 mixed sex pairs 1 female

2 5 2 mixed sex pairs 1 female6 2 mixed sex pairs 1 male7 2 mixed sex pairs 1 female and 1 male8 2 mixed sex pairs 1 male

3 9 2 mixed sex pairs 1 male10 2 mixed sex pairs 1 female and 1 male11 2 mixed sex pairs 1 male12 2 mixed sex pairs 1 female

4 13 2 mixed sex pairs 1 female14 2 mixed sex pairs 1 male15 2 mixed sex pairs 1 female16 2 mixed sex pairs 1 male and 1 female

Total 64 pairs = 128 20 males and 20 females = 40a Evaluations include: fat firmness, drip loss, Minolta Chroma Meter readings, simulated retail display and sensory evaluations.b Analysis includes: skatole, indole, fatty acid profile and TBARS.

Remainder of the boneless primal loin was vacuum packed and held in refrigerated storage (0-3ºC) until Day 8.

Minolta Chroma Meter readings were taken on the cut surface of each loin, at the anterior end, following a period of blooming and prior to vacuum packing.

On Day 8 the primal loin was butchered to provide the following samples (from the anterior end):

Three over-wrapped retail packs, each of two steaks (20 mm) for simulated retail display.

Two steaks (20 mm) vacuum packed and frozen for subsequent sensory evaluation.

Chemical analysis

Skatole and indole were determined according to the method described by Whittington et al. (1995). Fatty acid profiles were determined by direct saponification according to Enser et al. (1998).

Simulated retail display

SID 5 (2/05) Page 30 of 47

The three retail packs per side were placed in a chiller for a period of simulated retail display (at 0-3ºC). The packs were opened in sequence at 1, 3, 4, 5 and 7 days. On opening, Minolta Chroma Meter readings were taken and the steaks vacuum packed and frozen. Following the end of the trial, day 1 and day 7 samples were frozen for subsequent TBARS (thiobarbituric acid values) assessment according to Vyncke (1970).

Sensory evaluation

Trained sensory panellists were used to evaluate lean and fat samples from cooked loin steaks for the attributes presented in Table 8.

Panel sessions were structured so that samples from each housing system x feeding treatment were compared in each session. Six samples were allocated to each session in an incomplete block design so that there were male and female pigs represented in each session. (Not all treatments were represented in both sexes). This resulted in a total of 22 single panel sessions for the trial. In each session all samples were assessed by six experienced sensory panellists. Cooking and sampling procedures are documented below.

Vacuum packed steaks (2 per carcase) were selected from the freezer and thawed for 24 hours in a chiller at 3oC.

Steaks were removed from the packs about an hour before cooking and where necessary trimmed to a uniform thickness of 20 mm and weighed. Steaks were placed on a cooking turntable. The grill (Stot Benham Supergrill 600 gas grill) was set at the highest temperature and the turntable was placed under the burners.

Table 8 Sensory evaluation of cooked loin chopsSample Attribute MethodLean

Juiciness TastingTendernessPork flavourAbnormal flavourBoar flavour

FatPork flavourAbnormal flavourBoar flavour

Lean and fatOverall acceptability

FatPork odour SmellingAbnormal odourAndrostenoneSkatole

Steaks were grilled to an internal temperature of 75oC (monitored at the end of cooking using a Comark probe C9003 thermometer, Welwyn Garden City, Herts), about 5 minutes on each side. During cooking, the turntable was rotated approximately 60 degrees every 25 seconds. After cooking, the steaks were removed and placed on coded plates and each steak was re-weighed on a clean plate (one plate per steak to avoid transfer of flavour components between samples) to establish cooking loss. Each steak was loosely wrapped in aluminium foil, returned to its coded plate and held in a hot cabinet (54oC) until sub-sampling for sensory evaluation.

Steaks were removed from the cabinet, one at a time, and placed on a clean chopping board. The fat was separated from the lean. Six cubes, one per trained sensory panellist, about 1.5 to 2.0 cms in size, were cut from the lean, discarding the outside edges. Six unbrowned fat samples, about 1 cm were also cut. One sample of lean and another of fat was wrapped in aluminium foil. Excess fat from each pair of chops was placed in a coded, screw-top, wide-mouth glass jar for odour assessment. The wrapped samples, one from each allocation within sensory session, were placed in individual wells of a circular microwave bun-tray and returned to the hot cabinet until all samples were ready for sensory evaluation.

The bun-trays were transferred to the sensory booths within the panelling room. Each booth was illuminated by a green light (to eliminate bias from sample colour differences) and was equipped with a hot plate onto which a bun-tray was placed. Panellists were provided within each booth with a small white plate, knife, fork, glass of

SID 5 (2/05) Page 31 of 47

warm water and a piece of white toast (to cleanse residual flavour between sampling) and a disposable cup for residues. The jars containing fat samples, were placed on a hot plate, for assessment by all panellists.

The six trained sensory panellists were asked to score wrapped lean and fat samples and fat samples within jars, on a psuedo line 24 point scale (1 = weak/low to 24 = strong/high) for the attributes listed in Table 8. For statistical analysis, individual panellist scores were averaged to give a single score per attribute per pig sampled.

Data processing

Production

Feed intake (kg/pig day) was derived from total intake per pen divided by total pig days per pen specific to the grower and finisher stages and overall (grower plus finisher stages).

Growth rate (g/pig day) was derived from total net weight gain per pen divided by total pig days per pen specific to the grower and finisher stages. The overall growth rate was calculated for only those pigs left in the pens for finishing after the numbers were reduced at the end of the grower stage (week 6).

Feed conversion (intake/gain), was derived from total intake per pen divided by net weight gain per pen specific to the grower and finisher stages and overall (grower plus finisher stages).

All other inputs (labour, medicine, power, water and straw) and farm yard manure production were calculated overall (combined grower and finisher stages) from total input (or manure produced) per housing system divided by total pig days specific to each feeding system. Effluent production was measured for each pen and total output was divided by total pigs days for each pen over the combined grower and finisher stages.

Pig health and welfare

Health, hygiene and lesion scores were averaged across pigs within each pen, and across recording weeks for each pen. Behaviour data were expressed as the proportion of observations at which that behaviour was expressed. These data weresubsequently averaged across pigs within each pen, and across recording weeks for each pen. All of the post-slaughter assessments were averaged across pigs within each pen.

Microbial status

All microbial counts were transformed to log10 before analysis.

Salmonella tests on blood (ELISA) and caecal contents (laboratory incubation) were recorded as either positive or negative and present or absent respectively. These data were used to calculate the proportion of pigs that tested positive/present per pen, by dividing the number of positive/present results by the total recorded samples in each pen.

Meat quality

Sensory scores for individual panellist were averaged to give a single score per attribute per pig sampled for statistical analysis.

Sensory score sample means and all other sample assessments were considered as independent experimental units in the statistical analysis of meat quality data.

STATISTICAL ANALYSIS

Production

Pen means for live weight, feed intake, growth rate, feed conversion ratio, slaughter weight, carcase quality measurements and within pen standard deviation for growth and carcase quality measurements were subjected to Analysis of Variance (ANOVA) using the General Linear Model (GLM) in Minitab Statistical Software (Minitab Inc., State College, PA, U.S.A.). Model inputs included housing system (H), feeding system (F) and an interaction term for feeding and housing (I).

SID 5 (2/05) Page 32 of 47

Pen effluent data were subjected to one-way ANOVA to establish the effect of feeding system on waste production.

Pig health and welfare

Pen means for hygiene, lesion, health, behaviour and post-slaughter measurements were subjected to a two-way ANOVA. Model inputs included housing system (H), feeding system (F) and an interaction term for feeding and housing (I).

Microbial status

All microbial counts from gut samples at the start of the trial were subjected to two-way ANOVA and for colon samples GLM was used. Model inputs included H, F and I. All microbial counts from gut samples at slaughter were subjected to ANOVA using GLM. Model inputs included H, F and I.

Viable counts in complete feed samples were subjected to one-way ANOVA for the effect of feeding system.

Bacterial counts in faecal samples were subjected to two-way ANOVA. Model inputs included H, F and I.

Enterobacteraciae in effluent samples were subjected to ANOVA using GLM for the effect of feeding system.

Data obtained for total aerobic and coliform counts in dust samples were subjected to ANOVA using GLM. Model inputs included H, F and I.

Pen positive rates for Salmonella ELISA and caecal presence were subjected to two-way ANOVA. Model inputs included H, F and I.

Environmental impact

Room results for dust and ammonia were subjected to two-way ANOVA, with H and F as model inputs. Pen results were subjected to ANOVA using GLM for the effect of feeding system (F) on effluent production. Room sample results were subjected to ANOVA for the effect of feeding system (F) on effluent composition (F).

Meat quality

Sample data from physical, chemical and sensory measurements of meat quality were subjected to ANOVA using GLM. Model inputs included pig gender, H, F and I.

SYSTEMS TECHNICAL SPECIFICATIONS

Finishing system

Buildings

The finishing buildings have identical shells constructed of a steel frame (33.8m long, 12.30m wide with 2m eaves height at a 15 degree pitch. External walls are constructed from Durox insulated blocks (215mm, U value of 0.6) finished using Fibrocem plaster coating (3mm). The roofs are constructed with fibre cement with fibreglass insulation giving a U value of 0.7.

Ventilation system

The ventilation and environment in each room is automatically controlled (Euromatic DOL34H, Skov, Denmark) to set maximum and minimum ventilation, relative humidity (RH) and temperature against occupancy day.

Table 1 Settings for the environmental control system in the finishing buildingsDay Max. ventilation

( %)Relative

humidity (%)Straw based system:

Target temperature (oC)Fully slatted system:

Target temperature (oC)1 65 55 24 257 75 60 22 2314 85 - 20 2121 - - 19 2042 100 75 18 20

Minimum ventilation: Summer 20% and Winter 5%

SID 5 (2/05) Page 33 of 47

Each room is equipped with temperature and humidity sensors.

Fresh air is ventilated into each room through three side vents allocated in the external wall above each pen. The vents are fitted with adjustable flaps controlled by the Skov system.

Stale air is extracted from each room by a single fan located centrally in the roof. Fan speed and airflow are controlled automatically by the Skov system through adjustment of the exhaust inlet flap under negative air pressure.

Additional heating is provided by two heaters above each pen. These are also automatically controlled by the Skov system.

If in the event of a power cut the flaps and fan go to a default position and the building operates as an ACNV (Automatically Controlled Natural Ventilation) system.

Environmental data is captured on a computer using Skov software.

SID 5 (2/05) Page 34 of 47

SID 5 (2/05) Page 35 of 47

Internal construction of the fully slatted building (left), each pen with independent shallow effluent pit (10 to 14 days storage) for waste monitoring.

Internal layout of the fully slatted building (right), complete with concrete slats, central passage and internal walls, pen divisions and gates made from recycled high density polypropylene (PP), Panel Plus (51mm).

Internal construction of the straw based building (left), with central passage and scraped dunging areas.

Internal layout of the straw based building (right), with solid concrete for strawed lying and scraped dunging areas. Internal walls, pen divisions and gates (incomplete) made from recycled high density polypropylene (PP), Panel Plus.

Figure 1 Pen layout within each room of the fully slatted building

Figure 2 Pen layout within each room of the straw based building

Location of ad libitum hoppers for pen groups on pelleted dry feed (Study 1)

SID 5 (2/05) Page 36 of 47

Gate

Central Walkway

Ad-Lib trough with step for smaller pigs

X

5.5m 1.0m

3.7m

Central Walkway

Gate

Scrape passage

Ad-Lib troughLaying area

Ad-Lib trough

Liquid feed drop pipe

Elevation detail - feed pipe

Central Walkway

Feed line

X

X

3.7m

5.8m1.0m

Liquid feeding trough

Figure 3 Design and specification of the ad libitum liquid feeding trough

SID 5 (2/05) Page 37 of 47

Sectionshowing detail of divider

Front Elevation1200mm wide with 2 dividers,10mm 'chin bar' to reduce spillage,Front lip folded down to form safe edge,All other exposed edges with 5mm wire as protection.

Side Elevation300mm deep with lip height of 220mmBase and back fixings.

Three dimensional view

Liquid feeding troughs in the straw based (above) and fully slatted (left) buildings with delivery pipes and sensors fitted.

Feed centre

Figure 4 Schematic layout of feed centre (not to scale)

fireexit

fireexit

System ID Capacity DescriptionDry feed storage and milling1 DS1 5 tonnes Reserve cloth silo2 DS2 19 tonnes Cloth silo holding whole grain wheat3 DS3 19 tonnes Cloth silo holding pelleted wheatfeed4 DS4 19 tonnes Cloth silo holding whole grain barley5 DS5 5 tonnes Cloth silo holding rapeseed meal6 DS6 19 tonnes Cloth silo holding soya bean meal7 S7 1.3 tonnes Cloth bin with load cells holding milled wheat8 S8 1.3 tonnes Cloth bin with load cells holding milled barley9 S9 1.3 tonnes Cloth bin with load cells holding milled wheatfeed10 DS15 0.5 tonnes Metal bin holding mineral and vitamin supplement mix (grower)11 DS11 0.5 tonnes Metal bin holding mineral and vitamin supplement mix (finisher)12 DS10 0.5 tonnes Metal bin holding fish meal13 Hammer mill14 Dust valve and filter assemblyLiquid feed processing and delivery system (Meyer Lohne, Germany)15 3000 litres Waste storage tank16 Tank 3 3000 litres Central processing tank with load cells, receiving all feed ingredient components,

with heating and chilling facilities 17 Tank 4 1500 litres Protein tank with load cells18 Tank 5 (L1) 3000 litres Insulated fermentation/storage tank with load cells and heating & chilling facilities19 Tank 6 (L2) 3000 litres Insulated fermentation/storage tank with load cells and heating & chilling facilities20 Tank 1 1000 litres Feeding tank with load cells21 Tank 2 1000 litres Feeding tank with load cells22 S44 1000 litres Soya oil pumped into central process tank23 Positive displacement pumps for liquid feed transfer and deliveryHeating, refrigeration and water supply24 S41 Water tank25 Boiler26 ChillerCo-product storage and delivery27 - 30

37, 38, 39 and 40

22 tonnes each

Tanks for separate storage and delivery of four co-products to central process tank

SID 5 (2/05) Page 38 of 47

13

1 2 3 4

567 8 9

10 11 12

14

15161718192021

24

25

26

22

23

27 28 29 30

REFERENCES

APHA (1985). Standard Methods for the examination of water and wastewater, American Public Health Association. Baltimore, Maryland, USA: Port City Press.

Demmers, T.G.M., Burgess, L.R., Short, J.L., Phillips, V.R., Clark, J.A. and Wathes, C.M. (1999). Ammonia emissions from two mechanically ventilated UK livestock buildings. Atmospheric Environment 33 : 217-227.

Enser, M., Hallet, K.G., Hewitt, B., G A T Fursey, Wood, J.D. and Harrington, G. (1998). The polyunsaturated fatty acid composition of beef and lamb liver. Meat Science 49 : 312-327.

Fussel, R.J. and McCalley, D.V. (1987). Determination of volatile fatty acids (C2-C5) and lactic acid in silage by gas chromatography. Analyst 112 : 1213-1216.

Goodwin, R.F.W. and Whittlestone, P. (1979). Enzootic pneumonia of pigs: immunization attempts inoculating Mycoplasma Suipneumoniae antigen by various routes and with different adjuvants. British Veterinary Journal 129 : 456-464.

Heijden, H.M.J.F.v.d. (2001). First International ring trial of ELISAs for Salmonella antibody detection in swine. Berliner und Munchener Tierauzliche Wochenscrift 114 : 9-10.

ISO (1991). Microbiology - General guidance for the enumeration of micro-organisms - Colony count technique at 30oC. Switzerland: International Organization for Standardization: 5.

ISO (2001). Animal and vegetable fats and oils - Preparation of methyl esters of fatty acids. London: British Standards Institution: 24.

ISO, E. (1995). Animal and vegetable fats and oils - Analysis by gas chromatography of methyl esters of fatty acids. Brussels: European Committee for Standardization: 6.

Jones, G.F., Ward, G.E., Murtaugh, M.P., Lin, G. and Gebhart, C.J. (1993). Enhanced detection of intracellular organism of swine proliferative enteritis, ileal symbiont intracellularis, in feces by polymerase chain reaction. Journal of Clinical Microbiology 31 (10): 2611-2615.

Leser, T.D., Moller, K., Jensen, T.K. and Jorsal, S.E. (1997). Specific detection of Serpulina hyodysenteriae and potentially pathogenic weakly beta-haemolytic porcine intestinal spirochetes by polymerase chain reaction targeting 23S rDNA. Molecular and Cellular Probes 11 (5): 363-372.

Liu, H.J., Chang, B.Y., Yan, H.W., Yu, F.H. and Liu, X.X. (1995). Determination of amino acids in food and feed by derivatization with 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate and reversed-phased liquid chromatographic separation. Jounal of the Association of Official Analytical Chemists (AOAC) International 78 : 736-744.

Lyons, C.A.P., Bruce, J.M., Fowler, V.R. and English, P.R. (1995). A comparison of productivity and welfare of growing pigs in four intensive systems. LIvestock Production Science 43 : 265-274.

Mackie, T.J. and MacCartney, J.E. (1965). Handbook of Practical Bacterioloy: A guide to the laboratory diagnosis and control of infection. Edinburgh and London: E and S Livingstone.

MAFF (1982). The feeding stuffs (sampling and analysis) regulations (Amendment 1985). London: Her Majesty's Stationery Office: 76.

MAFF (1993). Prediction of the energy values of compound feeding stuffs for farm animals. Summary of the recommendations of a working party sponsored by the Ministry of Agriculture, Fisheries and Food. London: Her Majesty's Stationery Office. 1285: 16-17.

Moller, K., Jensen, T.K., Jorsal, S.E., Leser, T.D. and Carstensen, B. (1998). Detection of Lawsonia intracellularis, Serpulina hyodysenteriae, weakly beta-haemolytic intestinal spirochaetes, Salmonella enterica, and haemolytic Escherichia coli from swine herds with and without diarrhoea among growing pigs. Veterinary Microbiology 62 (1): 59-72.

SID 5 (2/05) Page 39 of 47

Moulsley, L.J. and Randall, J.M. (1990). Propeller fans for ventilating livestock buildings: 1. Measurement of performance. Journal of Agricultural Engineering Research 47 : 89-99.

Netten, P.V., Zee, H.V.d. and Moosdijk, A.V.d. (1991). The use of diagnostic semi-solid medium for the isolation of Salmonella enteritidis from poultry. 10th International Symposium on the Quality of Poultry Meat. Spelderholt Beckbergan, Netherlands.

Potkins, Z.V. and Lawrence, T.L.J. (1989). Oesophagogastric parakeratosis in the growing pig: effects of the physical form of barley-based diets and added fibre. Research in Veterinary Science 47 : 60-67.

Slevin, J., Wiseman, J., Parry, M. and Walker, R.M. (2001). Effect of protein nutrition on bone strength and incidence of osteochondrosis. British Society of Animal Science Annual Meeting. University of York. BSAS.

Takai, H., Pedersen, S., Johnsen, J.O., Metz, J.H.M., Koerkamp, P.W.G.G., Uenk, G.H., Phillips, V.R., Holden, M.R., Sneath, R.W., Short, J.L., White, R.P., Hartung, J., Seedorf, J., Schröder, M., Linkert, K.H. and Wathes, C.M. (1998). Concentrations and emissions of airborne dust in livestock buildings in Northern Europe. Journal of Agricultural Engineering Research 70 : 59-77.

Vyncke (1970). Direct determination of the thiobarbituric acid value in trichloroacetic acid extract of fish as a measure of oxidative rancidity. Fette Seinfen Anstrichm 72 : 1084.

Whittington, F.M., Nute, G.R., Warriss, P.D. and Wood, J.D. (1995). Skatole and androstenone measurements in the U.K. EAAP Working Group - Production and utilisation of meat from entire male pigs. Milton Keynes. EAAP.

TECHNOLOGY TRANSFER ACTIVITIES AND DATES

Finishing Pigs - Systems Research (LS 3601). Lead Contractor : Meat and Livestock Commission

Conference Presentations

November-02 Society of Feed Technologists, Pigs Meeting, Coventry.May-03 Bishop Burton Pig Industry Conference.May-03 Pig Veterinary Society, Spring Meeting, Leeds.September-03 Occassional Meeting of the British Society of Animal Science, NottinghamNovember-03 Pig Veterinary Society, Autumn Meeting, Leeds.November-03 Society of Feed Technologists, Pigs Meeting, Coventry.January-04 Society of Feed Technologists, Disease Control Without Medicines, Coventry.April-04 Annual Meeting of the British Society of Animal Science, York.May-04 Pig Veterinary Society, Spring Meeting, Chesire.November-04 Society of Feed Technologists, Pigs Meeting, Coventry.May-05 Pig Veterinary Society, Spring Meeting, Leeds.

Workshops and Seminars

November-00 Liquid Feeders' Forum, Milton KeynesJanuary-01 Liquid Feeders' Forum, Milton KeynesFebruary-05 Liquid Feeders' Forum, Staverton Park, Daventry

Shows

May-02 British Pig & Poultry Fair, StoneleighMay-04 British Pig & Poultry Fair, Stoneleigh

Presentations to producer discussion groups, clubs and associations

October-01 East Anglian Pig Advisers AssociationJanuary-02 Essex Pig Discussion GroupMarch-02 Hampshire 100 Club, North WalthamNovember-02 Driffield Pig Discussion GroupNovember-03 Scotlean AGM Meeting, Scotch CornerJanuary-04 NPA Producer Group Meeting, Milton Keynes

SID 5 (2/05) Page 40 of 47

Presentations at British Pig Executive (BPEX) and National Pig Producer (NPA) Regional Meetings

October-02 Eastern Meeting, Bury St EdmundsOctober-02 North East Meeting, Selby.November-02 South Meeting, Newbury.September-03 South West Meeting, TauntonSeptember-03 Eastern Meeting, Bury St EdmundsSeptember-03 South Meeting, Newbury.

Stotfold Visits

July-01 Prof Ian Gibson (MP)January-02 Defra/BPEXFebruary-02 Finishing Systems Project Participants meeting and visitMarch-02 Defra Chief ScientistMarch-02 Bedfordia GroupJuly-02 Whey Feeds and Tate & LyleJuly-02 East Anglian Pig Advisers AssociationSeptember-02 Finnfeeds and CPO ThailandOctober-02 Defra Policy and Livestock Science Unit, Stotfold.December-02 ACMCFebruary-03 IPPC Pig and Poultry Steering CommitteeMay-03 BPEXMay-03 Pork Chain SolutionsJune-03 LallemandJune-03 Australian Pork LtdJuly-03 ADAS Pig TeamAugust-03 Defra Chief Scientific Adviser (CSA)September-03 BPEXOctober-03 BOCM PaulsNovember-03 ADASMay-04 Thames Valley, Cambac Producer groupOctober-04 Bedfordia GroupDecember-04 Defra Pigs & Poultry Policy BranchMarch-05 Janssen Animal HealthApril-05 Midlands Pig Group

Briefings at NPA Technical and Scientific Committee Meetings

July-02 LondonOctober-02 Sutton BoningtonJanuary-03 LondonJuly-03 LondonOctober-03 ThirskJanuary-04 StotfoldNovember-04 ThirskFebruary-05 Grantham

Other

January-04 NFU Pigs Committee, EdinburghFebruary-04 Defra Review of Research on Livestock Nutrition, Warwick University

Websites

www.mlc.org.ukwww.stotfoldpigs.co.uk

SID 5 (2/05) Page 41 of 47

www.bpex.orgwww.defra.gov.uk

Articles in trade journals

Articles have been published in the following trade journals:

Farmers WeeklyPig WorldPig InternationalInternational Pig TopicsPig Progress

MLC/BPEX Publications

Technology Transfer activity has taken place through various MLC/BPEX publications and leaflets as follows:

BPEX Update (Winter 2002, Spring 2003, Auntumn 2003, Autumn 2004, Winter 2004 and June 2005).Press and news releasesTech Talk (Insert in Pig World)General guidelines on liquid feeding for pigs (leaflet)Liquid feeding systems for pigs (leaflet)Stotfold News

The Finishing Pigs Systems Research project was associated with two complementary but separate contracts, AW0130 and LK0649, respectively funded by Defra under a CSG7 Research Contract and funded through the Sustainable Livestock Production (SLP) LINK Programme. The project titles and lead contractors are given below:

AW0130 Welfare of finishing pigs under different management systems. University of Newcastle.LK0649 Using generalised innate immunity to enhance pig health and welfare. Roslin Institute.

Details of technology transfer activities under these projects is given below.

Project AW0130

Conference Presentations

November-03 Pig Veterinary Science, Autumn Meeting, Leeds.April-04 Annual Meeting of the British Society of Animal Science, York.March-05 5th International Colloquium on Animal Acute Phase Proteins, Dublin. April-05 British Society of Animal Science and International society of Applied Ethology, York.May-05 Pig Veterinary Science, Spring Meeting, Leeds.September-05 3rd International Workshop on the Assessment of Animal Welfare at Farm and Group Level, Vienna.

Work under this project has also been referred to in reviews presented at the following conferences:

November-02 Society of Feed Technologists, Pigs Meeting, Coventry.May-03 Bishop Burton Pig Industry ConferenceSeptember-03 Occassional Meeting of the British Society of Animal Science, NottinghamNovember-03 Society of Feed Technologists, Pigs Meeting, Coventry.November-04 Society of Feed Technologists, Pigs Meeting, Coventry.September-05 3rd International Workshop on the Assessment of Animal Welfare at Farm and Group Level, Vienna.

Presentations to policy makers and industry

SID 5 (2/05) Page 42 of 47

January-03 Producer Study Day, StotfoldApril-03 Producer Study Day, StotfoldJuly-03 Review of ‘On-farm Pig Welfare’, Defra Veterinary Research DivisionMay-04 Defra Veterinary Research Division, StotfoldDecember-04 Stotfold Pig Discussion GroupMarch-05 Defra Veterinary Research Division, Newcastle2002 to 2005 Finishing Pigs - Systems Research Project Participant Group Meetings (9 presentations)Autumn 2003 NPA/BPEX Regional Meetings (6 held)2004 Pig Producer Discussion Groups (5 held)

Project LK0649

Conference Presentations

April-04 Annual Meeting of the British Society of Animal Science, York.July-04 7th International Veterinary Immunology Symposium, Quebec, Canada.December-05 British Society for Immunology Annual Congress, Harrogate.

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.

SID 5 (2/05) Page 43 of 47

Under Project LS 3601 Finishing Pigs – Systems Research

Thompson, J. E. 2003. Nutritional precision and performance benefits: The potential of liquid feeding. Proceedings of the Society of Feed Technologists, Pigs Meeting (Jointly with the MLC). Coventry, November 2003. pp. 9.

SID 5 (2/05) Page 44 of 47

Thompson, J. E., Matthews, K. R., Taylor, L. and Gill, B. P. 2003. Finishing pigs systems research: 1) Production and meat quality (Proceedings of the Pig Veterinary Society meeting, Oulton Hall, Leeds 12th and 13th November 2003). The Pig Journal 55: 223-227.

Thompson, J. E. 2004. Nutrition and management for improving pig health and productivity. Proceedings of the Society of Feed Technologists, Pigs Meeting (Jointly with the HGCA and MLC). Coventry, November 2004. pp. 9.

Thompson, J. E., Matthews, K. R., Taylor, L. and Gill, B. P. 2004. The growth performance, carcase and meat quality of pigs finished under different housing and feeding systems: 1. liquid versus dry feeding in fully slatted and straw-bedded housing. Proceedings of the Annual Meeting of the British Society of Animal Science. York, March 2004. p. 42.

Thompson, J. E., Wiseman, J. and Gill, B. P. 2004. Physico-chemical aspects of liquid feed: the effect on component digestibility in growing/finishing pigs of 1) dietary dry matter concentration and 2) dietary fineness of grind. Proceedings of the Annual Meeting of the British Society of Animal Science. York, March 2004. p. 41.

Under Project AW013 Welfare of finishing pigs under different management systems. University of Newcastle.

Edwards, S. A., Scott, K., Armstrong, D., Chennells, D. J., Eckersall, P. D., Gill, B. P., Hunt, B. and Taylor, L. 2004. Finishing pigs systems research: health and welfare (Proceedings of the Pig Veterinary Society meeting, Oulton Hall, Leeds 12th and 13th November 2003). The Pig Journal 53: 123-127.

Edwards, S. A., Scott, K., Chennells, D. J., Campbell, F., Hunt, B., Armstrong, D., Taylor, L. and Gill, B. P. 2005. Finishing pig systems: health and welfare in straw bedded or slatted housing, (Proceedings of the Pig Veterinary Society Meeting. Oulton Hall, Leeds, 12th and 13th May 2005). The Pig Journal:

Scott, K., Armstrong, D., Chennells, D. J., Eckersall, P. D., Gill, B. P., Hunt, B., Taylor, L. and Edwards, S. A. 2004. The welfare of finishing pigs under different housing and feeding systems: 1. Liquid versus dry feeding in fully-slatted and straw-bedded housing. Proceedings of the Annual Meeting of the British Society of Animal Science. York, March 2004. p. 43.

Scott, K., Campbell, F., Chennells, D. J., Hunt, B., Armstrong, D., Taylor, L. and Gill, B. P. 2005a. Inter-relationships between health status, growth and acute phase proteins in finishing pigs (in preparation).

Scott, K., Campbell, F., Chennells, D. J., Hunt, B., Armstrong, D., Taylor, L., Gill, B. P. and Edwards, S. A. 2005b. Acute phase protein concentrations in finishing pigs under two different management systems. Proceedings of the 5th International Colloquium on Animal Acute Phase Proteins. Dublin, March 2005. pp.

Scott, K., Chennells, D. J., Armstrong, D., Taylor, L., Gill, B. P. and Edwards, S. A. 2005c. The welfare of finishing pigs under different housing and feeding systems: liquid versus dry feeding in fully-slatted and straw-based housing (Draft under approval). Animal Welfare:

Scott, K., Chennells, D. J., Campbell, F., Hunt, B., Armstrong, D., Taylor, L., Gill, B. P. and Edwards, S. A. 2005d. The welfare of finishing pigs in two contrasting housing systems: fully-slatted versus straw-based accommodation (Draft under approval). Livestock Production Science:

Scott, K., Chennells, D. J., Hunt, B., Armstrong, D., Taylor, L., Gill, B. P. and Edwards, S. A. 2005e. Multidisciplinary welfare assessment of pig finishing systems. Proceedings of the 3rd International Workshop on the Assessment of Animal Welfare at Farm and Group Level. Vienna, September 2005. pp.

Scott, K., Taylor, L., Gill, B. P. and Edwards, S. A. 2005. Influence of different types of environmental enrichment on the behaviour of finishing pigs in two different housing systems. Proceedings of the British Society of Animal Science (Joint Meeting of the British Society of Animal Science and International Society of Applied Ethology). York, April 2005. p. 82.

Scott, K., Watson, H., Taylor, L., Gill, B. P. and Edwards, S. A. 2005. Feeding behaviour of pigs offered ad libitum dry or liquid diets in fully slatted or straw based housing (in preparation).

SID 5 (2/05) Page 45 of 47

Scott, K., Taylor, L., Gill, B. P. and Edwards, S. A. 2005g. Influence of different types of environmental enrichment on the behaviour of finishing pigs in two different housing systems: 1. hanging toy v. rootable substrate (Submitted). Applied Animal Behaviour Science:

Scott, K., Taylor, L., Gill, B. P. and Edwards, S. A. 2005h. Influence of different types of environmental enrichment on the behaviour of finishing pigs in two different housing systems: 2. ratio of pigs to enrichment (Draft under approval). Applied Animal Behaviour Science:

Scott, K., Watson, H., Taylor, L., Gill, B. P. and Edwards, S. A. 2005. Feeding behaviour of pigs offered ad libitum dry or liquid diets in fully slatted or straw based housing (in preparation).

Under Project LK0649 Using generalised innate immunity to enhance pig health and welfare. Roslin Institute.

Clapperton, M., Bishop, S. C., Hillman, K., Gill, B. P. and Glass, E. J. 2004. Relationship between leucocyte subsets, performance, diet and bacterial load in Large White cross Landrace pigs. Proceedings of the Annual Meeting of the British Society of Animal Science. York, April 2004. p. 17.

Clapperton, M., Bishop, S. C., Hillman, K., Gill, B. P. and Glass, E. J. 2004. Diet affects bacterial load, performance and leucocyte sub-sets in Large White cross Landrace pigs. Proceedings of the 7th International Veterinary Immunology Symposium. Quebec, Canada, July 2004. p. 94.

Clapperton, M., Bishop, S. C., Hillman, K., Gill, B. P. and Glass, E. J. 2005. Relationship between the immune response towards Salmonella infection and performance in Large White cross Landrace pigs (to be submitted). Proceedings of the British Society for Immunology Annual Congress. Harrogate, December 2005. pp.

SID 5 (2/05) Page 46 of 47

SID 5 (2/05) Page 47 of 47