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 28

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 3642

2. Project title

Optimising nutrient budgets for livestock systems based on alternative forage crops.

3. Contractororganisation(s)

Institute of Grassland and Environmental ResearchPlas GogerddanAberystwythCeredigionSY23 3EB     

54. Total Defra project costs £ 599,667

5. Project: start date................ 01 July 2002

end date................. 30th June 2007

SID 5 (2/05) Page 2 of 28

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 Policy Commission report on the Future of Farming and Food, published 2002, highlighted the need to optimise the efficiency of livestock production and deliver traceable, environmentally-friendly products that meet consumer demands. In pursuit of this, landowners worldwide are under increasing pressure to maximise their use of home-grown forage-based diets for their livestock. Nutrient budgeting is a widely used tool for managing nutrients on farms as optimising the input and output of nutrients is critical to ensuring both short-term productivity and long-term sustainability. However, there was little information available on the effects of incorporating home-grown high-protein forages on the flow of nutrients within livestock systems. The research reported investigated the benefits and limitations of integrating a range of forages, including high-protein leguminous crops and catch crops such as red clover and kale, into livestock production systems. The aim was to provide a scientific basis to facilitate the best practice for optimising nutrient management when integrating home-grown forages into UK livestock systems. The approach was to study the individual components of a whole-farm system, to provide detailed data on soil-plant-animal-soil interactions, when using alternative forages in a livestock system. There were six specific objectives, with each one addressing one these individual components.

OBJECTIVE 1: The effect of feeding ensiled alternative forages on liveweight gain and excreta losses from growing lambsIn the UK, finishing lambs are typically fed ensiled ryegrass with the addition of concentrates to achieve commercially-viable productivity. Restrictions in the use of fish meal and meat and bone meal and fluctuations in the price of soya beans have led to a demand for alternative, traceable and low-cost protein feed. Effects of offering ensiled red clover (Trifolium pratense), lucerne (Medicago sativa), peas (Pisum sativum), kale (Brassica oleracea) and hybrid ryegrass (Lolium hybridicum) on the productivity and nutrient use efficiency of lambs were investigated (Milestone 01/02). Forages were cut, wilted for 24 h and ensiled as round bales. Suffolk-cross lambs aged eight months were housed and offered grass silage during a 5-week standardisation period and then group housed for 14 d adaptation and offered ad libitum access to a treatment silage. For the measurement period, lambs were split into four replicate blocks of five lambs per treatment. Mesh flooring placed over plastic trays were used in one of the four blocks and the lambs were rotationally moved every 14 days, in their respective blocks, so that faeces and urine were collected from beneath all lambs. Dry matter intake and liveweight were recorded every 7 d over 8 weeks. Chemical composition of the five silages showed significant differences between silage chemistries for all parameters measured. Lambs offered pea silage had a very low liveweight gain and this treatment was discontinued after 21 d. Lambs offered red clover, lucerne or kale silage had a higher liveweight gain than

SID 5 (2/05) Page 3 of 28

lambs offered ryegrass silage (P < 0.001). Feed conversion and N-use efficiency was also improved in lambs offered these alternative silages compared with those offered ensiled ryegrass (P < 0.001). Findings demonstrated the potential for ensiled red clover, lucerne and kale compared with ryegrass to improve the productivity and nutrient use efficiency of livestock systems. N excretion from lambs fed lucerne or red clover silage was higher than from lambs fed on kale or ryegrass silage (P < 0.001). This N source could represent either a substantial gain or loss of nutrients depending on how it was recycled within the system. This was investigated in objective 2.

Objective 2: The effectiveness of manures from livestock fed on alternative forages as organic fertilisers, and the response of alternative forages to manure applications. During the growing season in 2004, an experiment tested the hypothesis that applying slurries from sheep fed ensiled red clover, lucerne or kale would alter the yield of swards of hybrid ryegrass compared with applying slurries from sheep fed ryegrass or inorganic fertiliser N (Milestone 02/01). Thirty-two plots (12 x 2.5 m) of ryegrass (cv. AberExcel) were sown in four blocks. Slurries, collected as described in Objective 1, were applied to plots at a rate of 35 t ha -1 and compared with plots receiving ammonium nitrate at a rate of 0, 100 and 250 kg N ha-1 year-1. Ryegrass DM yield from plots treated with 250 kg N or with slurries from lambs offered lucerne or red clover were higher than from other treatments (P < 0.05). Findings demonstrated that, if managed correctly, slurry from ruminants fed on high-protein legume forages when compared with ryegrass can reduce reliance on artificial fertilisers. A second experiment tested the hypothesis that applying slurry to plots of red clover and lucerne once cut for silage would increase dry matter (DM) yield compared with plots not treated with slurry (Milestone 02/02) as N fixation is reduced in legumes after cutting due to a low photosynthate supply. Replicate plots of each forage were sown in 4 randomised blocks. Slurries from dairy heifers offered ryegrass silage were applied to half of the plots at 120 kg N ha-1 year-1. Plots were harvested on 4 dates. Slurry treatment increased the DM but not the N yield of sown species in all plots. There were no slurry × forage interaction effects on the DM yield of sown species or on total N yield, indicating there was no difference in the response of the forages to the N present in the slurry, despite two of the treatments being leguminous. This suggests that the legumes substituted the process of N2 fixation from the air with N uptake from the slurry in order to assimilate N, thus increasing DM yield, whilst photosynthate supply was low after cutting. However, applying slurry to red clover or lucerne did not maximise N-use efficiency.

Objective 3: Optimising requirements of red clover and lucerne for phosphate (P) and potash (K). Current Defra fertiliser guidelines (RB209) recommend applying phosphate (P) and potash (K) to red clover and lucerne according to those set for pure grass swards which have a different crop growth curve. To rectify this problem, yield data from experiments with red clover and lucerne conducted at IGER were collated and used to develop new guidelines for P and K requirements for these crops over the growing season (Milestone 03/01). There is a need to update the current guidelines on the P and K requirements of these forages for use by the agricultural industry. Objective 4: The effect of growing alternative forages within rotations on the yield, chemical composition and weed infestation of subsequent crops and nutrient leaching from soil. Legumes crops accumulate soil mineral nitrogen, which can be used by a subsequent crop. An experiment tested the hypothesis that the yield of a spring (cv. Riviera) barley (Hordeum vulgare) crop would be higher than a winter (cv. Pearl) barley crop when following 3 year old swards of either red clover or lucerne maintained by harvesting for silage (Milestone 04/01). A ryegrass treatment was included as a control. Four replicate, 15 x 15 m plots of red clover, lucerne, hybrid ryegrass receiving 250 kg N ha -1 annum-1 and hybrid ryegrass receiving 0 g N ha-1 annum-1 were sown. Half of each plot was then ploughed and sown with winter barley and the other half was sown with spring barley. Barley was harvested as whole-crop. The total forage, barley forage and grain DM yield of spring barley was higher than that of winter barley on plots that followed legumes but winter barley forage and grain yields were higher than spring barley yields on plots that followed ryegrass (P < 0.01). Overall, the best practice to optimise the soil mineral nitrogen for use by a subsequent cereal crop following legumes would be to sow spring compared with winter barley. This would improve the yield of the subsequent cereal crop and reduce N leaching from the soil.

Objective 5: The effect of harvesting regime on the quantity and composition of effluent from silages prepared from alternative forages, and its potential as a bio-fertiliser. A series of short-term experiments using lab-scale silos investigated the effect of different ensiling conditions on effluent produced from silages prepared from alternative forages (Milestone 05/01). The addition of silage inoculant to red clover, lucerne and ryegrass forage did not alter effluent production or the effluent N concentration. The effluent produced from ensiling ryegrass, red clover and lucerne without inoculant was added to soil cores of differing soil type to investigate their effects on changes in effluent biochemical oxygen demand (BOD), pH and the mineral status of the soil. The BOD of red clover, lucerne and ryegrass silage effluent was 16,202, 21,977 and 25,644 mg l-1, respectively. Total N in the leachate over 9 weeks from Andover series soil cores treated with lucerne effluent was higher than other treatments (P < 0.05). The BOD of the leachates from soil cores were found to be negligible. Destructive soil samples from cores treated with lucerne effluent had higher total N concentrations than other treatments (P < 0.01). Overall, using effluent from alternative forages compared with ryegrass silage effluent may improve the

SID 5 (2/05) Page 4 of 28

nutrient status of a soil without affecting the BOD of leachates. However, due to higher N concentrations, care needs to be taken to avoid N losses in leachates when applying these silage effluents to free-draining soils. Two other experiments tested the hypothesis that the addition of molassed sugar beet pulp (MSBP) or rolled barley could reduce silage effluent production from forage peas or kale when ensiled without wilting. In experiment one, results showed that unwilted pea silage with 0, 50 and 100 kg MSBP t -1 fresh forage produced, on average, 185 (s.d. 29), 40 (s.d. 9) and 0 ml of silage effluent, respectively. In experiment two, data showed that adding 50 and 75 kg MSBP t -1 fresh forage reduced the amount of effluent produced from unwilted kale silage when compared with incorporating 0, 25, 50 and 75 kg rolled barley t-1 fresh forage (P < 0.001).

Objective 6: The potential of alternative forages to supply minerals and trace elements to ruminant livestock, and associated efficiencies of use. Samples of feed inputs, faecal and urinary outputs from feed evaluation experiments conducted under Defra project LS0301 have been submitted for further chemical analysis to quantify the supply and utilisation of P and K from alternative forages (Milestone 06/01). Losses of minerals, such as P, from slurry are small compared to N losses, resulting in slurry that is mineral-rich relative to plant nutrient needs. Samples from 3 experiments were analysed to test the hypothesis that the excretion of minerals differs in lambs offered different ensiled forages. Experiment 1 compared lambs offered forage peas or field beans, experiment 2 compared lambs offered red clover, lucerne or birdsfoot trefoil and experiment 3 compared lambs offered sainfoin or kale silage. In each experiment, 6 suffolk-cross wethers were allocated to each silage and group housed for 14-d. Lambs were then adapted to metabolism crates for 7-d before data was collected over 7-d. Daily amounts of silage offered, refused and urine and faeces produced were measured and sub-sampled for mineral analysis. Differences in mineral intake or excretion between lambs offered the different silages were determined. For example, faecal P output of lambs offered red clover was lower than lambs offered birdsfoot trefoil or lucerne (P < 0.05) and faecal K output was higher in lambs offered birdsfoot trefoil than red clover or lucerne (P < 0.05) but there were no differences in the amount of urinary P or K lost. These findings indicate that feeding red clover may impact on P budgets within livestock systems by reducing the land area needed to recycle excreted P.

OUTCOMES AND FUTURE RESEARCH REQUIREMENTSThe findings have shown the potential to use home-grown forages to build soil fertility, improve nutrient efficiency in livestock, optimise nutrient requirements and thus, maximise nutrient capture and retention; resulting in a win-win scenario for production and the environment. The work has also emphasised the need for future studies to build on these findings and further our understanding of where, in the plant-animal-soil cycle, we can further improve the sustainability of our livestock systems in the UK through the use of these and other novel home-grown traceable feedstuffs. A future research requirement that is now needed is to use the data generated by this project in a whole farm modelling study. Examples of further studies needed to assist with the integration of these forages include investigating: their effects on whole farm nutrient balances when fed in differing proportions or as part of total mixed rations; their potential to extend the grazing season or for out-wintering livestock; and, strategies that will allow for the use of these forage crops to further mitigate the impact of UK livestock systems on the nutrient cycling of N, P, K and carbon on the wider environment and at a global scale.

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).

INTRODUCTION

SID 5 (2/05) Page 5 of 28

In pursuit of sustainable and economically-viable livestock systems, which meet consumer demands, many farmers worldwide are under increasing pressure to maximise their use of home-grown forage-based diets for their livestock. However, in order to meet these new challenges facing the agricultural industry, scientific studies were needed to assess the impact of nutrient flows from these alternative forage crops, in terms of soil – plant - animal interactions, when they are incorporated into livestock systems. The Policy Commission report on the Future of Farming and Food, published January 2002, highlighted the need for livestock farming in the UK to address the large gap in farm profitability between the bottom and top third of producers. The report also addressed the need to optimise efficiency of livestock production, deliver traceable, environmentally-friendly products that meet consumer demands. The Commission encourages a Regulatory framework that includes whole farm plans/audits to improve the balance between sustainable production and the environment.Nutrient budgeting is now a widely used tool for managing nutrients on farms and as an indicator of the sustainability of farming practices. Balancing the input and output of nutrients within the farm system is critical to ensuring both short-term productivity and long-term sustainability. Maximising the efficiency of use of nutrients within a system is also key to reducing bought-in inputs, which are potentially costly in both economic and environmental terms. However, there was very little information available on the effects of incorporating home-grown forages into production systems on the flow of nutrients within that system. The overall aim of this research proposal was to provide a scientific basis to facilitate the best practice for the sustainable integration of alternative forages within livestock production systems in the UK.

OBJECTIVESThere were six specific objectives. The approaches, results and implications of these are each presented in turn.

1) To investigate the effect of feeding ensiled alternative forages on liveweight gain and excreta losses from growing lambs;

2) To quantify the effectiveness of manures from livestock fed on alternative forages as organic fertilisers, and the response of alternative forages to manure applications;

3) To determine seasonal requirements of red clover and lucerne for P and K fertiliser;4) To investigate the effect of growing alternative forages within rotations on the yield, chemical composition

and weed infestation of subsequent crops and nutrient leaching from soil. 5) To investigate the effect of harvesting regime on the quantity and composition of effluent from silages

prepared from alternative forages, and its potential as a biofertiliser;6) To determine the potential of alternative forages to supply minerals and trace elements to ruminant

livestock, and associated efficiencies of use.

Objective 1: The effect of feeding ensiled alternative forages on liveweight gain and excreta losses from growing lambs

1.1 IntroductionDespite the potential for alternative forages to contribute to livestock systems, there has been relatively little research into their effects on voluntary intake and productivity in growing lambs. Some studies have focused on the effects of different ensiling approaches on the nutritive value of individual forages (Fraser et al., 2001a). Other studies have investigated the voluntary intake and productivity in finishing lambs or twin-bearing ewes and their progeny when fed ensiled red clover or lucerne compared with ryegrass (Speijers et al., 2005). Fraser et al. (2001b) compared the effects of harvest date and inoculation on the yield, fermentation characteristics and feeding value of forage pea and field bean silages and Vipond et al. (1998) studied the effects of feeding ensiled kale on the performance of finishing lambs. However, there are no studies comparing the effects of several contrasting alternative home-grown forages when ensiled on the voluntary intake and productivity of finishing lambs. An experiment was conducted to test the hypothesis that using ensiled high-protein forages compared with ensiled ryegrass in systems of lamb finishing would significantly increase livestock productivity, nutrient use efficiency and total nitrogen excreta losses, thereby, the sustainability of these livestock systems.

1.2 Materials and methodsForages were established on an area of stony, well-drained loam of the Rheidol series at the Institute of Grassland and Environmental Research (IGER), Aberystwyth, Wales (52 27’N, 4 01’W). To achieve an optimal soil pH of 6.5, ground limestone was applied at a rate of 5 t ha -1. Compound fertiliser was applied to achieve phosphate and potash indices of 2+ or 3 (MAFF, 2000). Silages were produced from 0.5 ha plots of a hybrid ryegrass (cv. AberExcel), red clover (cv. Merviot) and lucerne (cv. Vertus), sown on 2 September 2002 at 36, 14.5 and 18.5 kg ha-1, respectively. The lucerne seed was inoculated with Rhizobium meliloti. All plots were treated with the insecticide Dursban 4 (chlorpyrifos 480 g L-1; Dow Agrosciences, Hitchin, Herts.) applied at 1.5 L ha-1 on 4 September 2002 and with Lupus slug pellets (3 % methiocarb; Bayer plc., Bury St Edmunds, Suffolk) applied at 5 kg ha-1 on 9 September 2002. Red clover and lucerne plots were treated with herbicide (Headland Judo, propyzamide 400 g L-1; Headland Agrochemicals Ltd., Saffron Walden, Essex) at the rate of 1.75 L ha-1 on 13 December 2002. On 25 November 2002, the hybrid ryegrass was treated with herbicide (UPL Grassland Herbicide, dicamba, 25 g L-1; MCPA, 200 g L-1; mecoprop-P 200 g ℓ-1; United Phosphorus, Warrington, Cheshire) at 5 L ha-1. Fertilizer treatments consisted of ammonium nitrate applied at 80 kg N ha -1 and 65 kg N ha-1 to the hybrid ryegrass plots on 14 March and 30 May 2003 (after first-cut silage), respectively (to provide a total target N

SID 5 (2/05) Page 6 of 28

application of 250 kg N ha-1 annum-1) .Kale (‘Kaleage’, Sharpes International Seeds Ltd, Sleaford, UK; cv. Pinfold and Keeper in a 2:1 ratio by weight) and forage pea (cv. Espace) were sown on 22 April 2003 at 7.5 and 217 kg ha -1, respectively. Weather conditions did not allow for pre-emergence herbicide applications on the pea and kale crops. On 12 May 2003, Lupus slug pellets (3% methiocarb; Bayer plc, Bury St Edmunds, Suffolk) were applied at 5 kg ha -1 to pea and kale plots. The pea crop was also treated with a herbicide mixture of 4 L ha-1 Pulsar (bentazone 200 g ℓ-1, MCPB 200 g L-1; BASF, Cheadle, Cheshire) and 0.4 L Fortrol (cyanazine 500 g L-1; Feinchemie, Saffron Walden, Essex) on 28 May 2003. Kale plots were also treated with a herbicide, Semeron 25WP (25 % desmetryn; Syngenta, Whittlesford, Cambridgeshire) at 1.1 kg ha-1 on 16 June 2003. Ryegrass, red clover and lucerne were harvested on 29 May 2003 and 13 July 2003 (first- and second-cut silage, respectively). On both occasions, forages were cut at their optimal cutting height for maximum re-growth: 60 mm for ryegrass and 100 mm for red clover and lucerne (Frame et al., 1998). Pea and kale plots were harvested, at a height of 100 mm, on 22 July and 12 August 2003. After wilting for 24 h, all forages were then treated with a silage inoculant (Sil-All 4x4™, Alltech, Stamford, UK), applied at 106 colony-forming units g-1 fresh matter (2 L t-1 fresh forage), using a commercial inoculant applicator (Flojet; Team Sprayers, Ely, Cambridgeshire, UK). Forages were ensiled as large round bales using six layers of 25 μm x 750 mm wide film (Silotite, BPI-Agri., Leominister, UK). The red clover, lucerne and ryegrass or pea and kale bales had been stored for at least 147 days before being opened for the experiment. Weather conditions delayed the date of harvest of the forage pea crop and this led to a high degree of mould contamination resulting in the forage being deemed unsuitable for feeding. As a result, the pea silage used in the experiment was obtained from a commercial farm. The commercial pea crop (cv. Espace) had been sown on 21 April 2003 without the use of herbicide. The pea sward was cut using a disc mower, without a conditioner, on 3 July 2003 and ensiled as large round bales following a 48 h wilt, without the use of a silage inoculant. This forage pea silage had been stored for a minimum of 177 days. Bales of each silage treatment were opened every 7 d during the experiment. Prior to feeding, all silages were chopped to a uniform length stored in baskets at 4 º C. Red clover, lucerne, ryegrass and pea silages were then chopped using a commercial big bale chopper (Reco Bulldog, Storti International, Belfiore, Italy) and the kale silage (due to its low DM content) was chopped using a stationary modified precision-chop harvester. The experiment comprised of a six-week standardisation period, a two-week diet adaptation period and an eight-week measurement period. One hundred Suffolk-cross lambs, aged approximately 8 months, were obtained from the same commercial late-lambing flock. All lambs were treated with an anthelmintic and fluke drench (Levamisole and triclabendazole; Combinex Sheep, Novartis Animal Health UK Ltd., Cambridge, UK) and were offered ryegrass silage whilst at pasture. The lambs were then housed as a single group and offered a single ryegrass silage ad libitum for a standardisation period of six weeks. Live weight was monitored at weekly intervals throughout this period. Following this period, the lambs were grouped within gender according to live weight (mean 30.9, s. d. 2.29) and body condition score (mean 2.3, s.d. 0.154) resulting in four blocking groups, nominally large castrates, large females, small castrates and small females. Within each of these blocking groups, five lambs were allocated at random to each of the five forage treatments. The five treatment groups were segregated and experimental silages were introduced gradually into the diet over the following seven-day period (i.e. proportionately 0.75:0.25, 0.50:0.50, 0.25:0.75 and 1.00 of treatment and ryegrass silage offered). A further seven-day period with ad libitum access to their allocated silage was allowed for dietary adaptation. All lambs were then treated with an anthelmintic wormer (mebendazole) containing selenium and cobalt (Ovitelmin SC™, Janssen Animal Health, UK) and transferred to a sheep-housing facility that was arbitrarily divided into four blocks of five pens. Each blocking group was allocated at random to a set of five adjacent pens and silage treatment groups were allocated at random to pens within this. Mesh flooring placed over plastic trays were used in one of the four blocks and the lambs were rotationally moved every 14 days, in their respective blocks, so that faeces and urine were collected from beneath all lambs during an 8 week experiment (slurries were collected for use in Objective 2). Lambs were bedded on sawdust for the remaining three rotations. Each pen of lambs was offered forage ad libitum, with feeding levels designed to ensure a refusal margin of 0.10 to 0.15 each day. Lambs on red clover, lucerne and ryegrass silage were fed first-cut silage during weeks 1 - 4 and second-cut silage during weeks 5 - 8. Fresh weights and dry matter (DM) contents of feed offered to each replicate group of lambs were recorded. Aliquots of silages offered were retained on a daily basis and bulked within measurement week. Feed refusals were weighed three times each week and sub-samples were retained and bulked within each measurement week. Sub-samples of the silage offered and refused were stored at -20 oC prior to subsequent analysis. Individual lambs were weighed and condition scored at the start of the measurement period and every seven days for the following eight weeks. Body condition score (BCS) was determined by handling the loin of the lambs using a scale of 1 to 5, with quarter scores as intermediate points along the scale. Sub-samples of the excreta collected under lambs on each treatment were collected at the end of each fortnightly, mixed thoroughly, diluted with a known amount of water and sub-sampled for analysis for N and ammonium-N concentrations. Total amounts of faeces and urine excreted were determined as lambs were moved every 14 days.The DM contents of the forages offered, and feed refused, were determined by drying to constant weight at 80 °C in a forced-draught oven for 24 h, and the DM content of the samples for all chemical analyses was determined by freeze-drying. Ash was measured by igniting samples in a muffle furnace at 550 C for 16 h. Total N (TN) concentrations were determined using a Leco FP 428 nitrogen analyser (Leco Corporation, St. Joseph, MI, US) and TN was expressed as crude protein (CP) concentration (TN x 6.25). Neutral-detergent fibre (NDF) and

SID 5 (2/05) Page 7 of 28

modified acid-detergent fibre (MAD-F) concentrations were determined according to the method of Van Soest et al. (1991). Metabolizable energy (ME) concentrations of all silages were estimated from the MAD-F concentration of the forage (Givens et al., 1989). Concentrations of water-soluble carbohydrate (WSC) were determined by the method of Thomas (1977). Water extracts of the silage samples were prepared by adding 250 mL water to a 50 g silage sample, and filtering this suspension (after storage at 4ºC for 24 h) through Whatman no. 1 filter paper. The filtrate samples were subsequently used for determination of pH value and further analysis for volatile fatty acids (VFA), ammonia and lactic acid concentrations. Volatile fatty acid concentrations were determined by gas chromatography as described by Zhu et al. (1996). Ammonia-N concentrations were determined by Hoskins distillation technique according to analysis of agricultural materials (MAFF, 1986). Lactic acid concentrations were determined using spectrophotographic methods as described by Merry et al. (1995), whilst pH was determined as described by Cussen et al. (1995). For each pen, daily intake was calculated as the mean of the daily intakes for each measurement week. Within week and within pen mean live weight raised to the power of 0.75 was used to scale intakes to a metabolic liveweight basis. For each lamb daily liveweight gain was calculated as the slope between weekly liveweight data and days on measurement and determined using the non-parametric regression method of Theil as implemented by Dhanoa (1998). Daily intakes, gross feed conversion efficiencies and within-pen means for liveweight gain and condition score and excreta losses were subjected to analysis of variance assuming a randomised block design and treating blocking group as a random effect. In the case of liveweight gain and condition score data, liveweight at the end of the standardisation period was used as a covariate. Treatment means were compared using the Student-Newman-Keuls test. All data were analysed with the aid of Genstat, Version 8.1 (Payne et al., 2005). One lamb (on red clover) was withdrawn from the experiment during the second week of the measurement period and the data from this lamb was treated as a missing value throughout.

1.3 Results Lambs offered the forage pea silage had extremely low liveweight gains during the first three weeks of the experiment. As a result, after careful monitoring, this treatment was discontinued and the lambs on this treatment were taken off the experiment in order to ensure animal welfare was not compromised. The highest proportion of weeds was found in the ryegrass silage (211 and 124 g kg -1 DM in first and second cut, respectively). Of the other forages, the highest proportions were found within the first-cut red clover and lucerne silages (33 and 67 g kg-1 DM, respectively) and the proportion of weeds present in the other silages was found to be negligible. The chemical composition of the silages differed for all the variables measured (Table 1.1). Lucerne had the highest DM content and the NDF concentration for the ryegrass silage was higher than for the other silages. Kale silage had a lower DM content and NDF concentration than the other silages and the highest ash concentration, more than twice the concentration present in the ryegrass silage. The chemical composition of the pea silage offered to lambs for 21 days indicated that the silage was stable, with pH levels below 4.3 and ammonia-N concentrations below 75 g kg-1 TN. VFA data showed that the acetic acid concentration of the silage was approximately twice that of the lactic acid but butyric acid concentrations were low.

Table 1.1. Mean (s.e. of mean) chemical composition [all values g kg -1 DM, except DM content (g kg-1 FM), ammonia-N (NH3-N) (g kg-1 TN) and estimated metabolisable energy (MJ kg-1 DM)] of different silage treatments as fed to lambs over an 8-week period, except pea silage which was offered for 21 days only.

Ryegrass Red clover Lucerne Kale PeasDry matter 470 (44.5) 410 (21.2) 508 (24.2) 187 (7.0) 344 (12.6)Ash 58 (1.4) 81 (1.6) 87 (2.4) 115 (3.6) 86 (3.4)pH 4.6 (0.25) 4.2 (0.04) 4.6 (0.04) 4.0 (0.03) 4.3 (0.04)Ammonia-N 36 (5.4) 49 (4.7) 46 (6.5) 53 (2.0) 72 (1.8)Crude protein 116 (6.1) 188 (4.7) 222 (2.4) 161 (7.1) 206 (9.4)WSC 133 (10.3) 61 (8.3) 28 (4.5) 43 (3.3) 15 (0.81)NDF 623 (5.2) 421 (4.6) 418 (7.9) 285 (4.8) 436 (4.4)ME 10.0 (0.06) 10.6 (0.05) 10.4 (0.04) 11.2 (0.05) 10.2 (0.17)Lactic acid 17 (3.5) 25 (2.1) 21 (1.2) 42 (2.9) 26 (2.6)Acetic acid 4 (0.6) 6 (0.7) 8 (0.9) 19 (1.4) 40 (4.7)Propionic acid 0.3 (0.06) 0.3 (0.07) 0.3 (0.07) 0.8 (0.11) 1.0 (0.10)Butyric acid 1.7 (0.61) 1.3 (0.48) 0.3 (0.09) 0.2 (0.08) 1.0 (0.38)

WSC, water-soluble carbohydrate; NDF, Neutral-detergent fibre; Metabolisable energy

Voluntary intakes of DM and estimated ME were significantly higher for lambs fed ensiled red clover and lucerne compared with those fed kale or ryegrass (P < 0.001) (Table 1.2). The N intake reflected the N concentration in the silages and differed significantly (P < 0.001) among all silage treatments, with lambs fed on lucerne having the highest and lambs fed on ryegrass the lowest intake of N. Statistical differences among the forages did not alter depending on whether data were compared on a per lamb or a metabolic liveweight basis, except that the DM intake of lambs offered lucerne silage was significantly higher than lambs offered red clover silage when compared on a metabolic liveweight basis (P < 0.05). The mean daily intake of lambs offered pea silage over the first 21 d of the experiment was 0.55 (s.d. 0.07) kg DM lamb-1.

SID 5 (2/05) Page 8 of 28

Table 1.2 Mean intake of dry matter (DM), metabolizable energy (ME) and nitrogen (N ) of silages by lambs over an eight-week period when offered different silages, as determined from the data of five lambs in four replicated groups per treatment.

Ryegrass Red clover Lucerne Kale s.e. of difference Level of significanceIntake (d-1)DM (kg) 0.65b 1.01a 1.05a 0.74b 0.044 ***ME (MJ) 6.5c 10.7a 10.9a 8.2b 0.47 ***N (g) 11.9d 30.3b 37.4a 18.9c 1.16 ***Intake (kg-1 LW0.75 d-1)DM (g) 49.8c 69.3b 75.4a 53.4c 2.28 ***ME (MJ) 0.50c 0.73a 0.78a 0.60b 0.024 ***N (g) 0.92d 2.08b 2.68a 1.37c 0.069 ***

***, P < 0.001. Means within rows with different superscripts are significantly different (P < 0.05).

Lambs fed on the legume silages, red clover and lucerne, had significantly higher final live weights at the end of the eight-week experimental period than lambs fed on kale or ryegrass silage (P < 0.01) (Table 1.3). Lambs fed on ensiled alternative forages had significantly higher daily liveweight gains (P < 0.001) and final body condition scores (P < 0.01) than lambs fed ensiled ryegrass. However, it should be noted that, although lambs on lucerne silage had the same liveweight gain during the measurement period of the experiment as those on red clover, the final liveweight of the lambs fed on ensiled red clover was 1.1 kg higher than that of the lambs fed lucerne silage.Food conversion efficiency [kg liveweight gain (kg DM intake)-1] was significantly higher in lambs fed red clover, lucerne or kale compared with lambs offered ryegrass silage (P < 0.001). Examination of the N-use efficiency [kg liveweight gain (kg N intake)-1] of the lambs showed that lambs on kale silage had a significantly higher nitrogen use efficiency than lambs on all the other silages (P < 0.05) and lambs offered red clover silage had higher nitrogen use efficiency than lambs on ryegrass silage (P < 0.01).

Table 1.3 Effects of different silage treatments on the performance, mean food conversion efficiency (FCE) [kg liveweight gain (kg DM consumed) -1] and nitrogen use efficiency (NiUE) [kg liveweight gain (kg N consumed) -1] of finishing lambs over an eight-week period as determined from four groups of five lambs per treatment of finishing lambs over an eight-week period.

Ryegrass Red Clover

Lucerne Kale s.e. of difference

Level of significance

Liveweight gain (g day-1) 36c 135a 135a 100b 8.9 ***Final liveweight (day 56) (kg) 30.9c 38.3a 37.2a 34.0b 0.87 ***Final body condition score 2.3b 2.6a 2.7a 2.6a 0.08 **FCE 0.053b 0.133a 0.130a 0.136a 0.0081 ***NiUE 2.9c 4.4b 3.7bc 5.3a 0.36 ***

**, P < 0.01; ***, P < 0.001. Means within rows with different superscripts are significantly different (P < 0.05).

Lambs fed kale silage produced more excreta than lambs on the other treatments (P < 0.001). In terms of dry matter output, lambs fed red clover or lucerne silage produced more excreta than lambs on either ryegrass or kale silage (P < 0.01). Total nitrogen excretion was higher from lucerne silage than from red clover silage ( P < 0.05) and both legume silages showed higher rates of nitrogen excretion than either ryegrass or kale (P < 0.001). In absolute terms, lambs fed lucerne excreted the most ammonium-N (P < 0.05). The proportion of nitrogen in the form of ammonium in slurry from lambs fed ryegrass silage was significantly lower (P < 0.05) than from lambs fed lucerne, red clover or kale silages. Nitrogen excretion in the form of nitrate was negligible (<0.1g) and was unaffected by treatment.

Table 1.4 Excreta losses over 14d from groups of five lambs offered different silages.

Ryegrass Red clover Lucerne Kale s.e.d. Level of significance

Total excreta (kg) 65c 99b 118b 191a 11.9 ***Dry matter (kg) 16.8b 30.5a 33.7a 16.6b 2.99 ***Total Nitrogen (TN) (g) 708c 1459b 1995a 769c 120.6 ***Ammonium-N (g) 59d 322b 516a 166c 35.7 ***

(g/kg TN) 89b 225a 259a 221a 19.8 ******, P < 0.001 with 9df for error. Means within rows with different superscripts differ significantly (P < 0.05).

1.4 Discussion and ConclusionsAll silages used in the experiment had acceptable pH values, below pH 4.7, indicative of a good fermentation during the ensiling process. The higher pH and lower lactic acid concentrations of ryegrass, red clover and lucerne silage indicate that these silages underwent a more restricted fermentation during ensiling compared with the kale. This was probably due to their higher DM content compared with the kale silage, which had a DM

SID 5 (2/05) Page 9 of 28

content which was outside the target DM content for baled silage (250 - 400 g kg -1) (Merry et al., 2000). All silage ammonia-N concentrations were in the range of 36 – 53 g kg -1 total N, except for the pea silage at 72 g kg -1 total N, indicating that these forages had not undergone extensive proteolysis during the ensiling process. The relatively low CP concentration of the ryegrass and kale silages was probably due to the delay in harvesting, as the CP concentrations of forages decline with increasing maturity (Merry et al., 2000). The ME concentration (10.0 MJ kg-1 DM) of this silage was also at the lower end of the target ME concentration recommended to farmers when making silage for production feeding (ME target of 10.0 - 10.5 MJ kg -1 DM (Stanley et al., 2004)). However, mean CP and ME concentrations of over 500 big bale ryegrass silages produced on commercial farms in the same year as this experiment was 132 g kg -1 DM and 10.1 MJ kg-1 DM, respectively (Frank Wright Ltd., pers. comm) and, therefore, the ryegrass silage used in this experiment was representative of that used in UK livestock production. It should be noted that the ME values of all forages were calculated using the equation for grass silage, and the values for forage other than ryegrass should, therefore, be treated with some caution in this respect.The original forage pea crop grown was affected by a high degree of mould contamination resulting in the forage being deemed to be unsuitable for feeding. The lack of suitability of either of the two forage pea silages was disappointing as previous work has indicated that this forage has good potential as a protein feed, with higher N-use efficiency compared with field beans (Fraser et al., 2001b). High acetic acid concentrations in silage are associated with reduced feed intake by ruminants and this may have accounted, in part, for the low voluntary intakes recorded for lambs in the study. Voluntary intake of silage has been found to correlate positively with CP concentration and negatively with ammonia-N concentrations (Wilkins et al., 1971). In agreement with findings from other studies, the lambs in the current study that were fed ryegrass silage had a lower voluntary DM intake compared with lambs fed on red clover silages (Speijers et al., 2005). In the study of Vipond et al. (1998), which investigated the effects of kale silage on finishing lambs, of a similar age and liveweight to those in the current study, voluntary intake of both kale and ryegrass silages was lower than found in the current study (0.68 and 0.50 kg DM d-1, respectively). This may have been due to the lambs in the former study being offered kale silage with a supplement of molassed sugar beet pellets at 0.45 kg day-1 lamb-1. A target liveweight gain for finishing lambs is up to 150 g day -1 but lambs fed on high quality ryegrass silage without an additional concentrate would be expected to gain only 50 g day-1 (the target liveweight gain for ‘holding’ store lambs) (Butler, 1985). On this basis, the performance of the lambs in the current study fed on ryegrass silage reflect the liveweight gains that would be achieved if ryegrass silage of a quality that was typical of that produced in the UK, at least for 2002, was used as the sole feed for store lambs. It is not surprising therefore, that in most instances in commercial farming practice in the UK, lambs are fed concentrates to achieve optimal performance targets. The findings of this study show that all lambs on the alternative forage silages attained liveweight gains of over 100 g day-1, without the use of additional concentrate supplementation, highlighting the potential of using these forages to achieve more sustainable and nutrient-efficient lamb-finishing systems. Interestingly, the final liveweight of the lambs fed on ensiled red clover was higher than that of the lambs fed lucerne silage, despite lambs on lucerne silage having the same liveweight gain during the measurement period of the experiment. As all the lambs were balanced for liveweight prior to the standardisation period for diet adaptation, this finding was due to the lambs fed red clover being heavier than those fed lucerne at the end of the adaptation period, indicating that the lambs offered red clover silage adapted more readily to their change in dietary forage than the lambs offered lucerne. This could be an important consideration in lamb-finishing systems which require a rapid production response to dietary changes, to compete with market prices. The findings of this study demonstrate the potential for using ensiled farm-grown high-protein forages compared with ensiled ryegrass to improve nutrient-use efficiency in lamb-finishing systems in the UK. Using data presented in another study investigating the voluntary intake and productivity in finishing lambs when fed ensiled red clover or lucerne compared with ryegrass (Speijers et al., 2005a), the mean feed-conversion efficiency for ensiled red clover, lucerne and ryegrass can be calculated as being 0.143, 0.068 and 0.128 kg liveweight gain (kg DM intake)-1, respectively. These values are comparable to the current study with regard to red clover silage but the values for lucerne and ryegrass silage reflect the differences in quality of these silages between the two experiments. The feed-conversion efficiency of lambs offered kale in the current study is higher than that found by Vipond et al. (1998). The higher N-use efficiency of lambs on kale silage compared to the other silages may be due to kale supplying a more optimal balance of energy to protein in the rumen. Changing the diets of lambs to different forages altered the excreta losses when compared with those from lambs fed on ryegrass silage. Although lambs fed on kale silage produced a higher amount of excreta than lambs on other silages, the total DM losses from lambs fed on kale silage were lower than lambs fed the legume silages, confirming visual observations that the lambs on kale silage produced more urine than those on other treatments. In conclusion, the findings demonstrate the potential for using ensiled alternative forages compared with ensiled ryegrass to improve nutrient-use efficiency, and thus, the sustainability of lamb-finishing systems. However, to determine the effects of incorporating these forages into a livestock system, the implications of these findings with respect to total loss of nutrients by the lamb is needed to determine their full economic and environmental impact within nutrient-budget plan for a farm. Given the high initial CP concentration of the legume forages, this could still represent either a substantial gain or loss of nutrients to the whole system, and has the potential to have a profound economic and environmental impact if this N is not correctly stored and recycled. Further work was undertaken to determine the effect of applying slurries from livestock fed on these alternative forages as plant fertilizer to swards of hybrid ryegrass, when compared with slurries from animals fed on ryegrass and with applications of inorganic N, as described under objective 2.

SID 5 (2/05) Page 10 of 28

Objective 2: The effectiveness of manures from livestock fed on alternative forages (red clover, lucerne or kale) as organic fertilisers, and the response of alternative forages to manure applications.Three experiments were conducted as part of this objective: Experiment 2.1 studied the effects of slurry from sheep fed on alternative forages on swards of hybrid ryegrass; Experiment 2.2 determined the response of legume forages to manure applications after frequent defoliation; and, Experiment 2.3 investigated the N-use efficiency of kale forage when grown under different rates of organic and inorganic N application.

Experiment 2.1. Effects of applying slurries from livestock fed on different forages or inorganic nitrogen on swards of hybrid ryegrass 2.1.1 IntroductionManaged correctly, farmyard manure and slurry can effectively reduce the need for expensive inorganic fertilisers. Fluctuations in world feed prices, the ban on animal protein in ruminant rations following the BSE crisis, and increasing consumer concerns regarding traceability has led to an upsurge in demand for home-grown high-protein forages. However, little is currently known about the efficiency with which crops can utilise the nutrients from slurries derived from these forages. An experiment tested the hypothesis that applying slurries from sheep fed ensiled red clover (Trifolium pratense), lucerne (Medicago sativa) or kale (Brassica oleracea) would alter the yield and weed infestation of swards of hybrid ryegrass (Lolium hybridicum) compared with applying slurries from sheep fed hybrid ryegrass or with inorganic N.

2.1.2 Materials and MethodsThirty-two field plots (12 x 2.5m) of a hybrid ryegrass (cv. AberExcel) were sown at a rate of 36 kg ha -1 on 2 September 2002 in four replicate blocks. The plots were sited on an area of stony, well-drained loam of the Rheidol series (52o 26' 55" N, 4o 1' 27" W). To achieve an optimal soil pH, ground limestone was applied at a rate of 5 t ha-1. Compound fertiliser was applied to achieve phosphate and potash indices of 2+ or 3 (ADAS, 1983). Plots were cut on 5 December 2002 to a height of 6 cm. During the first harvest year (2003), ryegrass plots were maintained by cutting to a height of 6 cm on 12 March, 13 May, 25 June, 8 August, 24 September and 10 December and the harvested material removed. Artificial N fertiliser was added to all plots, as 34.5 % ammonium nitrate, on 5 occasions: 11 March, 28 March and immediately after cuts 2, 3 and 4 to provide a total of 200 kg N ha-1 annum-1. Potassium and phosphate fertiliser were added as potassium oxide and phosphorus pentoxide at a rate of 215 kg K2O ha-1 and 159 kg ha-1, respectively, to maintain phosphate and potash indices of 2+ or 3.Slurries were collected from 80 Suffolk-cross finishing lambs fed on ensiled red clover, lucerne, kale and ryegrass during an eight-week period. A full description of experimental work from which slurries were collected was provided in objective 1. The faeces and urine collected were diluted 1:1 with water (except kale which was sufficiently dilute) and mixed thoroughly to form slurries and then stored in 1 m 3 plastic vessels at 4ºC. Slurry from animals fed on the different silages were applied to the 4 replicate field plots of ryegrass and compared with plots receiving ammonium nitrate at a rate of 0, 100 and 250 kg N ha -1 year-1. Slurries were applied manually using calibrated buckets and watering cans with a piece of bent steel attachments to simulate a splashplate tanker. At application, the slurries were further diluted so that all slurries were applied at a ratio of 1 slurry: 2 parts water to allow the material to be applied evenly to the plot surface. All slurries were kept well mixed and were randomly applied within a set time on the same day at a rate of 35 t ha-1 as a split dressing, with half applied on 26 March 2004 and the remainder applied on 20 May 2004. All plots treated with slurry also received ammonium nitrate at 100 kg N ha-1 year-1 applied as a base application at a rate of 25kg N ha -1 on 18 March, 20 May, 5 July and 26 August using a Gandy plot fertiliser (BLEC Landscaping Equipment Ltd., Spalding, Lincolnshire). Potassium and phosphate fertiliser were added as potassium oxide and phosphorus pentoxide at a rate of 154 kg ha-1, to all experimental plots on 27 May and 31 August 2004.

2.1.3 ResultsThe total DM yield relative to N applied in slurries from sheep fed on different forages are presented in Figure 2.1. Data analysis showed the ryegrass DM yield from plots treated with 250 kg N or with slurries from lambs offered lucerne or red clover were higher than from other treatments (P < 0.05). Treating hybrid ryegrass plots with slurries from animals fed on different forages or with different levels of inorganic N did not alter the yield of unsown species.

2.1.4 Discussion and ConclusionsTreating plots of hybrid ryegrass with slurry from animals fed on lucerne plus 100 kg N ha -1 did not significantly alter the DM yield when compared with plots receiving 250 kg N ha-1 of inorganic nitrogen. Applying slurries from lambs fed on different forages did not alter the DM yield of unsown species present in ryegrass swards.

SID 5 (2/05) Page 11 of 28

Figure 2.1. Total annual yield (t DM ha-1) of ryegrass and unsown species compared to the total N applied (kg N ha-1) to plots of hybrid ryegrass treated with slurries from sheep offered 4 different forage diets or with inorganic nitrogen at a rate of 0, 100 and 250 kg N ha-1 year-1.

Experiment 2.2 Effects of applying slurry on the yield of red clover, lucerne or hybrid ryegrass when cut for silage production 2.2.1 IntroductionPerennial forage legumes are an important source of biologically-fixed N, reducing reliance on inorganic fertiliser and, thereby, improving the sustainability of farming systems. Reported rates of N2 fixation in above-ground plant tissues can be up to 373 and 350 kg N / ha / year for red clover and lucerne, respectively (Carlsson and Huss-Danell, 2003). However, the amount of N fixed by forage legumes varies greatly with species, management, soil N supply and the effects of companion grasses (Mallarino et al., 1990) and N acquisition has a major influence on plant productivity, with a direct linear relationship between relative growth rate and plant nitrogen concentration (Verkroost and Wassen, 2005). Typically, unless large amounts of inorganic N are applied, legumes do not respond to N fertiliser as long as they are capable of fixing N2. However, it is easier and less energy-consuming for a plant to absorb N from the soil than to fix it from the air. Research has shown that after defoliation, by cutting or grazing, N2 fixation was reduced in legumes (including red clover and lucerne) due to a low photosynthate supply (Butler, 1987; Farnham and George, 1994). The aim of the experiment reported here was to test the hypothesis that applying slurry to plots of red clover or lucerne, when frequently cut for silage production, would increase dry matter (DM) yield, and thereby N yield, compared with plots not treated with slurry.

2.2.2 Materials and MethodsTwenty-four field plots (12 x 2.5m) were sown with either red clover (cv. Merviot), lucerne (cv. Vertus) or hybrid ryegrass (cv. AberExcel), on 5 September 2003 at 14.5, 19 and 35 kg/ha respectively, in a randomised design in four blocks. The plots were sited on an area of stony, well-drained loam of the Rheidol series (52 o 26' 55" N, 4o 1' 27" W). To achieve an optimal soil pH, ground limestone was applied at a rate of 5 t ha-1. Compound fertiliser was applied to achieve phosphate and potash indices of 2+ or 3 (MAFF, 2000). During the first harvest year (2004), all plots were maintained by cutting to a height of 6 cm on 1 April and then 10 cm on 26 May, 13 July, 3 September and 18 October. An additional cut was taken at a height of 6 cm from the ryegrass plots only on 10 December. On each occasion, the harvested material was removed. N fertiliser was added to ryegrass plots, as 34.5 % ammonium nitrate, on 4 occasions: 1 April, 28 May, 22 July and 8 Sept to provide a total of 200 kg N / ha / annum using a Gandy applicator (BLEC Landscaping Equipment Ltd., Spalding, Lincolnshire). Potassium and phosphate fertiliser, as potassium oxide and phosphorus pentoxide, were added at a rate of 154 and 140 kg/ha respectively on 28 May and 8 September. Slurry was collected from 12 month old dairy heifers fed on ryegrass / clover silage and stored in 1 m 3 plastic vessels at 4ºC prior to application. During the second harvest year (2005) slurries were applied manually using calibrated buckets and watering cans adapted with a steel attachment to the nozzle to simulate a splash plate tanker. At application the slurries were diluted 1:1 with water to allow the material to be applied evenly to the plot surface. All slurries were kept well mixed and were randomly applied within a set time on the same day at a rate of 70 t/ha as a split dressing, with half applied on 30 March and the remainder on 24 May. This provided a total of 114 kg/ha N. Sub-samples of the slurry as applied to each block were analysed. Plots were harvested by taking 4 cuts to a height of 10 cm on 23 May, 12 July, 5 September and 25 October from all plots using a Haldrup 1500 plot harvester (J. Haldrup a/s, Løgstør, Denmark). Yield was determined by weighing the material cut from an area of 12 m x 1.5 m within each plot. Sub-samples of forage, as harvested, were taken to determine DM and N yield and botanical composition. All samples were stored at - 20 C prior to chemical analysis. The DM of the herbage was determined by drying to constant weight at 80 °C in a

SID 5 (2/05) Page 12 of 28

forced-draught oven, and the DM of the samples for chemical analysis determined by freeze-drying. Total N of the herbage cut and the slurry applied was determined using a Leco FP 428 nitrogen analyser (Leco Corporation, St. Joseph, MI, US). Data were analysed by analysis of variance using Genstat, Version 8.1 (Payne et al., 2005). Treatment means were compared using the Student-Newman-Keul multiple comparison test.

2.2.3 ResultsSlurry treatment increased the DM but not the N yield of sown species in all plots (P < 0.001) (Table 2.2). There were no effects of crop or slurry treatment on the yield of unsown species present in the plots. There were no slurry × forage interaction effects on the DM yield of sown species or on total N yield. The DM yield of red clover and lucerne was higher than that of hybrid ryegrass (P < 0.001). The low yield of hybrid ryegrass in the experiment was typical of ryegrass without the addition of artificial N. The total N yield of lucerne was higher than that of red clover (P < 0.001) which was in turn higher than that of hybrid ryegrass (P < 0.001).

Table 2.2. Total annual yield (t DM/ha) of sown and unsown species and total N yield (kg N/ha) from plots of red clover, lucerne or hybrid ryegrass treated or not treated with slurry.

    Sown species DM yield Unsown species DM yield Nitrogen yield

Red clover No slurry 11.1 1.02 362Slurry 13.0 0.19 381

Lucerne No Slurry 12.4 0.26 434Slurry 13.5 0.36 432

Ryegrass No Slurry 3.1 0.55 52  Slurry 4.4 0.44 71Slurry S.e.d 0.19 0.343 10.1

Effect *** ns nsCrop S.e.d 0.24 0.280 8.3

Effect *** ns ***Slurry x Crop S.e.d 0.34 0.485 14.3  Effect ns ns ns

ns, not significant; ***, P < 0.001.

2.2.4 Discussion and ConclusionsSlurry treatment increased the DM but not the N yield of sown species in all plots. The increase in DM yield due to slurry treatment may have been beneficial economically. However applying slurry to red clover or lucerne, rather than to grass swards or other non-leguminous crops, would not maximise the efficiency of use of nutrients on farms where the aim is to ensure both short-term productivity and long-term sustainability in farming practice. There were no slurry × forage interaction effects on the DM yield of sown species or on total N yield, indicating there was no difference in the response of the forages to the N present in the slurry, despite two of the treatments being leguminous. This suggests that the legume forages may have substituted the process of N2 fixation from the air with N uptake from the slurry in order to assimilate N, and to increase their DM yield, whilst their photosynthate supply was low after cutting. However, this assumption does not account for the fact that ryegrass may not assimilate N at the same rate as red clover and lucerne.

Experiment 2.3. The effects of different rates of organic and inorganic N application on the nitrogen-use efficiency of forage kale. 2.3.1 IntroductionShort-term alternative forage crops require the addition of manures prior to their cultivation rather than during the growing season. This approach presents an opportunity for farmers to apply and incorporate animal manures and slurries that have accumulated over the winter period. However, when organic fertilisers are used, it is typical for the N input to be based on the total N content. Therefore, depending on whether N is applied from slurry or from an inorganic source has the potential to greatly affect the N use efficiency by forage kale, given the differences in plant-available N between these two resources. To investigate this, replicated plots of forage kale (Brassica oleracea) were established to determine the N use efficiency by forage kale treated with different rates of nitrogen in either slurry or inorganic N fertiliser.

2.3.2 Materials and MethodsThirty-six plots (12 x 3 m) were allocated to treatment within four randomised blocks. At application, slurries were diluted with water to alter the rate of N application and to allow the material to be applied evenly to the plot surface. On 4 May 2006, slurry at 40, 80, and 120 kg N ha-1 was applied manually using watering cans and slurry at 160 kg N ha-1 was applied using a calibrated bucket and spread using a yard scraper. All slurries were kept well

SID 5 (2/05) Page 13 of 28

mixed and were randomly applied within a set time on the same day at a rate of 35 t ha -1. Water was added to control plots at the same rate as the slurry treated plots on 5 May 2006. The experimental area was then ploughed to a depth of 100 mm prior to sowing with kale (cv. Maris Kestrel) (British Seedhouses Ltd., Bristol, UK) on 9 May 2006 at 6.18 kg ha-1. Inorganic N was applied by hand, on 10 May 2006. The kale was harvested on 6 November 2006 using a Haldrup 1500 plot harvester (J. Haldrup a/s, Løgstør, Denmark), and cut to a height of 100 mm. Yield was determined by weighing the material cut from an area of 10.5 m x 1.5 m within each plot. A 500 g sub-sample of forage, as harvested from each plot, was placed in a forced-draught oven at 80 °C for dry matter (DM) determination. A second 200 g sub-sample was taken and freeze-dried prior to chemical analysis. Forage samples were analysed for total N, P and K concentrations. The nitrogen-use efficiency was calculated as (N yield of treated plots – N yield of control plots / total N applied). Data were analysed by regression analysis with the aid of Genstat, Version 8.1 (Payne et al., 2005).

2.3.3 ResultsThe soil mineral nitrogen status of the site prior to N applications was determined to be Index 2. Regression analysis showed that nitrogen yield increased linearly with increasing rate of application of either inorganic N (P < 0.001) or slurry N (P < 0.05). Fertiliser inorganic N was used more efficiently (P < 0.05) (0.48 ± 0.090 kg N yield / kg N applied) (r =0.73) compared with the N in the slurry (0.13 ± 0.055 kg N yield / kg N applied) (r =0.50). When the inorganic N in the fertiliser was compared with the total of the ammonium-N and nitrate-N in the slurry (0.25 ± 0.106 kg N yield / kg inorganic N applied in slurry), there was no difference in N use efficiency ( P > 0.05). In the slurry, the total N consisted of 50.4 % ammonium-N, 0.28 % nitrate-N, leaving 49.3 % as organic-N.

2.3.4 Discussion and ConclusionsThe nitrogen yield response of the kale to N application was linear despite the N application rates exceeding the 130 kg N /ha recommended for this forage crop at a soil nitrogen supply index of 0 (Defra, 2000). These results were not as expected, making it difficult to demonstrate optimal rate of inorganic fertiliser N compared with slurry N required to optimise N return. It was anticipated that N-use efficiency would decrease for the fertiliser N treatment at rates of between 120 and 160 kg N/ha given the recommendations and the SNS of the experimental site. These findings suggest that the recommended N applications for this crop may need revising through further studies across years at different sites. Furthermore, the results demonstrate the importance of knowing the available N concentration not just the total N concentration of slurry when applied as a plant fertiliser. This highlights the importance of dietary influences on nutrient planning in systems which alter the ratio of inorganic N to total N in slurry. Typically the inorganic to total N content of cattle slurry is assumed to be 50:50 (Defra, 2000) but, as showed in Objective 1 and 2, the ratio of inorganic N to total N alters when different forages are fed to ruminant animals. Future studies using higher N application rates are needed to determine the point at which N-use efficiency of kale is reduced following applications of inorganic fertiliser N and also slurries from ruminants fed on different forage diets.

Objective 3: Seasonal requirements of red clover and lucerne for phosphate (P) and potash (K) fertiliser.3.1 IntroductionCurrent DEFRA fertiliser guidelines (RB209) recommend applying phosphate and potash to red clover and lucerne according to the guidelines for pure grass swards, and do not take into account the difference in annual growth curve when calculating phosphate and potash requirements.

3.2 Materials and MethodsTo rectify this problem, yield data from experiments with red clover and lucerne conducted at IGER was collated and used to develop new guidelines for P and K requirements for these crops over the growing season. Data was then collated further and used to demonstrate the P and K requirements of red clover or lucerne, relative to soil indices and also number of cuts taken for silage production.

3.3 ResultsThe recommendations devised for red clover and lucerne are presented in Table 3.1 and 3.2, respectively. To minimize luxury uptake of potash, no more than 80 - 90 kg/ha potash should be applied in one application. The balance of the recommended rate should be applied in the previous autumn or during grazing. The data presented in the tables has made allowances for sufficient potash for subsequent grazing after cuts. As in RB209, these calculations assume that the P and K concentrations of red clover and lucerne are similar to grass.

SID 5 (2/05) Page 14 of 28

Table 3.1. Recommendations for phosphate (P2O5) and potash (K2O) (kg ha-1) for red clover over a growing season, depending on soil indices and the number of cuts taken for silage production.

Soil P or K index

Cumulative DM offtake 0 1 2 - 2 + 3 4 +

1 Cut System (4t DM/ha)Phosphate 75 50 25 25 20 0Potash 155 125 95 75 30 02 Cut System (8t DM/ha)Phosphate 105 80 55 55 20 0Potash 280 230 190 140 60 03 Cut System (11t DM/ha)Phosphate 125 100 75 75 20 0Potash 355 305 265 175 75 0

Table 3.2. Recommendations for phosphate (P2O5) and potash (K2O) (kg/ha) for lucerne over a growing season, depending on soil indices and the number of cuts taken for silage production.

Soil P or K index

Cumulative DM offtake 0 1 2 - 2 + 3 4 +

1 Cut System (5t DM/ha)Phosphate 85 60 35 35 20 0Potash 180 150 120 100 30 02 Cut System (10t DM/ha)Phosphate 120 95 70 70 20 0Potash 330 280 240 190 80 03 Cut System (13t DM/ha)Phosphate 140 115 90 90 20 0Potash 400 350 310 220 120 04 Cut System (14t DM/ha)Phosphate 145 120 95 95 20 0Potash 425 375 335 215 125 0

3.4 Discussion and ConclusionsThere is a need to update the current guidelines on the P and K requirements of these forages for use by the agricultural industry. As in RB209, these calculations assume that the P and K concentrations of red clover and lucerne are similar to grass. Further work should be undertaken to determine the P and K concentrations among varieties of red clover and lucerne and factors that affect the uptake of these nutrients by these forage legumes.

Objective 4: The effect of growing alternative forages within rotations on the yield, chemical composition and weed infestation of subsequent crops and nutrient leaching from soil. 4.1 IntroductionThe current cost of using inorganic fertilisers in both monetary and environmental terms, has lead to a renewed interest in the use of legumes to facilitate the development of economically- and environmentally- sustainable farming systems. The ability of legumes, such as red clover, to accumulate substantial soil mineral nitrogen (SMN) by biological fixation and its beneficial residual effects for a subsequent crop has been demonstrated in numerous studies. However, further research was needed to determine the best practice to optimise the capture this labile nitrogen for use by a subsequent cereal crop. The experiment compared the effects of red clover, lucerne or ryegrass (with the latter either treated or not treated with inorganic nitrogen) on the forage and grain yield of a subsequent crop of either autumn sown or spring sown barley (Hordeum vulgare). Effects of red clover on N losses underneath a subsequent crop of either autumn or spring barley were also investigated.

4.2 Materials and MethodsFour replicate, 15 x 15 m plots of red clover (Trifolium pratense), lucerne (Medicago sativa), hybrid ryegrass (HRG) (Lolium hybridicum) receiving 250 kg N ha-1 annum-1 and HRG receiving zero N ha-1 annum-1 were sown on 2 Sept 2002. The plots were sited on an area of stony, well-drained loam of the Rheidol series (52o 26' 55" N, 4o 1' 27" W). In the first and second harvest years, plots were maintained by cutting with a Haldrup plot harvester to a height of 10 cm on four occasions per annum. The 250 N HRG plots received nitrogen at 78 kg, 64 kg, 56 kg and 52 kg N ha-1 prior to each cut, respectively. Phosphate and potash levels were maintained at index 2+ or 3 throughout the experiment. Forage dry matter (DM) yields were recorded and sub-samples were analysed for crude protein (CP) (N x 6.25) concentrations. On 13 October 2004, half of each plot was ploughed and sown with

SID 5 (2/05) Page 15 of 28

winter barley (cv. Pearl) at 195 kg ha-1. The other half of each plot was ploughed on the 17 March 2005 and sown with spring barley (cv. Riviera) at 185 kg ha-1. Barley tiller counts were assessed on 7 March and 6 May 2005 for winter and spring barley, respectively. Both barley crops were harvested as whole-crop at the dough stage by sampling a 3 x 13m area with a Haldrup plot harvester to a height of 6 cm. The winter barley plots were harvested on 28th June 2005 and the spring barley harvested on 25 th July 2005. Prior to harvest, crop height and the number of stems 1 m-1 within eight rows of each split-plot were counted, harvested and thrashed to determine the grain:straw ratio and the total grain yield. Barley whole-crop DM yields were recorded and sub-samples analysed for CP (N x 6.25) concentrations. N leaching losses from underneath winter barley or spring barley following red clover was determined using 3 ceramic cups per split-plot with weekly samples collected every 7 d when soil was at field capacity. Daily drainage volumes were calculated as rainfall minus evaporation and soil moisture deficit. Daily rainfall was measured at the site and potential evapotranspiration calculated from data recorded at the IGER weather station (5 km from the site), using the Penman-Monteith equation for short grass and for bare soil (Hess, 2002). Daily drainage volumes were summed to obtain the total drainage for each sampling interval.

4.3 ResultsThe total DM yield of the legume or ryegrass forages during their first and second harvest years were 4158, 12420, 12288 and 12451 kg/ha for the 0 N HRG, 250 N HRG, lucerne and red clover respectively. Mean crude protein concentrations of these forages during the same period were 94, 111, 225 and 192 g kg -1 DM for 0 N HRG, 250 N HRG, lucerne and red clover, respectively. The total forage, barley forage and grain DM yield of spring barley was higher than that of winter barley on plots that followed legumes but winter barley forage and grain yields were higher than spring barley yields on plots that followed ryegrass (P < 0.01) (Table 4.1). Barley tiller numbers were higher where legumes were previously grown when compared with barley where ryegrass was previously grown (P < 0.001) and higher for winter barley when compared with spring barley (P < 0.05). Winter barley had a higher crop height than spring barley (P < 0.001) and the height of barley grown after legumes was higher than barley grown after ryegrass (P < 0.05). Plots of spring barley that followed legumes had higher numbers of stems than plots of winter barley that followed legumes (P < 0.05) but there was no difference between spring and winter barley stem numbers on plots that followed ryegrass. There was no effect of previous forage on CP yield of winter barley but the CP yield of spring barley that followed legumes was higher than spring barley that followed ryegrass (P < 0.05). Spring barley that followed ryegrass had a higher thousand seed weight (TSW) than winter barley that followed ryegrass (P < 0.05) but there was no difference between spring and winter barley TSW on plots that followed legumes.

Winter Barley Spring Barley0 N

HRG250 N HRG

Luc Redclover

0 N HRG

250 N HRG

Luc Red clover PF B SED

PF x B

Establishment Barley tillers m-2 735 876 1010 1221 366 453 941 1012 *** *** 48.8 **Standing crop Crop height (cm) 74 82 84 88 46 52 63 68 *** *** 2.0 ** Barley stems m-2 387 451 465 502 379 468 659 746 *** *** 36.9 ***DM & CP yield (kg ha-1) Total Forage 7490 8398 9136 9852 5183 6634 10875 11705 *** NS 462.1 *** Barley Forage 7481 8395 9131 9848 4763 6190 10820 11681 *** NS 502.5 *** Crude Protein 539 537 619 614 451 464 879 874 *** ** 52.7 ***Grain Yield (kgDM ha-1) 4017 4281 4662 4844 2621 3547 6087 6492 *** NS 337.7 *** TSW (g) 44.4 44.1 42.8 42.0 49.1 47.3 42.3 42.6 ** ** 1.39 *

Table 4.1. Effect of legumes or hybrid ryegrass (HRG; not treated (0 N) or treated (250 N) with N) on the forage and grain yield of a subsequent crop of winter or spring barley.NS, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001. Luc, Lucerne; TSW, thousand seed weight; PF, previous forage effect; B, barley effect; PF x B, previous forage x barley effect.

Total N losses determined underneath spring or winter barley following red clover showed that sowing plots to winter barley significantly increased the N lost through leaching compared with sowing plots to spring barley (Figure 4.1). The total N leached between 11 November 2004 and 17 March 2005 from under winter barley (56.6 kg N /ha) was higher than from under spring barley (24.8 kg N /ha) when sown following red clover (P < 0.001). There were no further losses of N from underneath winter barley after 17 March 2005. The total N loss under winter barley (56.6 kg N /ha) was higher than from under spring barley (35.4 kg N /ha) when sown following red clover between 11 November 2004 and 5 May 2005 (P < 0.01).

SID 5 (2/05) Page 16 of 28

Figure 4.1. Cumulative nitrate-N loss (kg/ha) underneath a subsequent crop of either winter or spring barley following red clover.

4.4 Discussion and ConclusionsOverall, the results showed that the best practice to optimise the soil mineral nitrogen for use by a subsequent cereal crop following legumes would be to sow spring barley compared with winter barley. The results showed that this would result in an improvement in the yield of the subsequent cereal crop and also reduced the amount of N leached from the soil.

Objective 5: The effect of harvesting regime on the quantity and composition of effluent from silages prepared from alternative forages, and its potential as a bio-fertiliser.A series of short-term experiments using 10 kg bin silos were conducted to investigate the effect of different ensiling conditions on the amount and chemical composition of effluent from different silages.

Experiment 5.1: Effects of a bacterial inoculant on silage effluent and silage quality when ensiling red clover, lucerne or hybrid ryegrass without wilting.

5.1.1 Introduction Silage effluent is one of the worst pollutants produced on farm and farmers are encouraged to reduce water pollution risks when silage making wherever possible. Ensiling forage in dry weather and wilting to a suitable dry matter (DM) (at least 250 g kg-1 DM) will significantly reduce the amount of silage effluent produced. However, in practice, weather conditions may not permit the forage to be wilted prior to ensiling. The Codes of Good Agricultural Practice provide guidance to farmers on the control and handling of silage effluent from predominantly grass silage crops. However, there is little information with which to generate guidelines on potential effluent flows from inherently low dry matter crops such as brassicas, legumes and pulses. Consequently there is a need to study the environmental impact of effluents from alternative forages and to explore strategies such as wilting, additives, harvesting periods and combined cropping in order to provide guidance on the control of effluent from inherent low dry matter crops. In this experiment, the effects of a silage inoculant on the quantity and chemical composition of effluent from legume silages were compared with ryegrass silage.

5.1.2 Materials and MethodsForages were established on an area of stony, well-drained loam of the Rheidol series at the Institute of Grassland and Environmental Research (IGER), Aberystwyth, Wales (52 27’N, 4 01’W). Silages were produced from forage cut from 0.5 ha plots of a hybrid ryegrass (cv. AberExcel), red clover (cv. Merviot) and lucerne (cv. Vertus), sown on 2 September 2002 at 36, 14.5 and 18.5 kg ha-1, respectively. The lucerne seed was inoculated with Rhizobium meliloti. All plots were treated with the insecticide Dursban 4 (chloropyrifos 480 g L -1; Dow Agrosciences, Hitchin, Herts.) applied at 1.5 L ha-1 on 4 September 2002 and with Lupus slug pellets (3 % methiocarb; Bayer plc., Bury St Edmunds, Suffolk) applied at 5 kg ha -1 on 9 September 2002. Red clover and lucerne plots were treated with herbicide (Headland Judo, propyzamide 400 g L -1; Headland Agrochemicals Ltd., Saffron Walden, Essex) at the rate of 1.75 L ha-1 on 13 December 2002. On 25 November 2002, the hybrid ryegrass was treated with herbicide (UPL Grassland Herbicide, dicamba, 25 g L-1; MCPA, 200 g L-1; mecoprop-P 200 g L-1; United Phosphorus, Warrington, Cheshire) at 5 L ha -1. Fertilizer treatments consisted of ammonium nitrate applied at 80 kg N ha-1 and 65 kg N ha-1 to the hybrid ryegrass plots on 14 March and 30 May 2003 (after first-cut silage), respectively (to provide a total target N application of 250 kg N ha -1 annum-1). Ryegrass, red clover and lucerne were harvested on 29 May 2003 (first-cut silage) at their optimal cutting height for maximum re-growth: 60 mm for ryegrass and 100 mm for red clover and lucerne (Frame et al., 1998). A 500 g sub-sample

SID 5 (2/05) Page 17 of 28

of the forage harvested was used to determine botanical composition and a second 500 g sub-sample was analysed for DM, ash, WSC, CP (N x 6.25), acid detergent fibre (ADF) and neutral detergent fibre (NDF) concentrations and in vitro digestibility of the organic matter in the dry matter (DOMD). Forages were then chopped to a uniform length of approximately 20 cm, using a stationary modified precision chop forage harvester. Three replicate silos of approximately 10 kg were used for each treatment. Each forage treatment was mixed thoroughly before being treated either with a Lactobacillus plantarum inoculant (Live System, Genus Ltd., UK), applied at 106 colony forming units / gram fresh matter or an equivalent volume of water. Forages were ensiled in 50 l insulated bins, lined with a 250 gauge polythene bag and fitted with a one-way valve for the collection of silage effluent. The forage within each bin was compacted under a lever and piston system, designed to deliver pressure at 817 kg m-2 and the liner within each bin was sealed at the top using a rubber ring. Silage effluent was collected in sealed containers for 11 days. The volume of effluents collected from each silo was recorded daily and stored at 4ºC, before being bulked and sub-sampled for chemical analysis. Effluent chemical analyses determined were DM (100ºC for 48h) and total N. Following completion of the effluent collection phase, a 5 kg sandbag was placed on top of the forage and the bin sealed with a plastic lid. At the same time, the one-way valve at the base of each bin was removed and replaced by a rubber stopper. After 90 days of ensiling, silos were opened and sub-samples taken for chemical analysis. Silage samples were analysed for DM, pH, CP (N x 6.25), ammonia-N (NH3-N), WSC, lactic, acetic, propionic and butyric acid concentrations. All sample material was stored at -20C prior to subsequent chemical analysis. The DM content of the forages, when fresh and ensiled, were determined by drying to constant weight at 80 °C in a forced draught oven, and the freeze DM of the samples for chemical analysis determined. TN concentrations were determined using a Leco FP 428 nitrogen analyser (Leco Corporation, St. Joseph, MI, US) and expressed as crude protein (CP) (TN x 6.25). NDF analysis was carried out according to the method of Van Soest et al. (1991). Ammonia-N was determined according to the method of Bremner and Keeney (1965), WSC concentrations by the method of Thomas (1977), volatile fatty acids (VFA) were determined by gas chromatography as described by Zhu et al. (1996). Lactic acid concentrations were determined using spectrophotographic methods as described by Merry et al. (1995), whilst pH was determined as described by Cussen et al. (1995). All data were analysed using Genstat version 8.1 (Payne et al. 2005). Silage effluent volumes and chemical composition were compared by a two-way ANOVA.

5.1.4 ResultsThe ryegrass, red clover and lucerne forage harvested for ensiling contained 97 (s.e. 0.1), 97 (s.e. 0.8) and 90 (1.4) per cent of the species sown, respectively and their chemical compositions are presented in Table 5.1.

Table 5.1. Mean chemical composition [all values g kg-1 DM, except DM content (g kg-1 FM)] of different forage treatments prior to ensiling.

Ryegrass Red clover LucerneDry matter 215 151 190Ash 41 80 94Crude protein 76 188 216Water-soluble carbohydrate 215 97 57Neutral-detergent fibre 576 376 366Acid-detergent fibre 341 296 303DOMD 616 662 681

Red clover produced more silage effluent in total than lucerne which, in turn, produced more silage effluent than hybrid ryegrass (P < 0.001), which was due to the amount of effluent these silages produced between day 1-3 (P < 0.001) (Table 5.2). During day 4-5, the hybrid ryegrass silage produced more silage effluent than the legume silages (P < 0.001). The effects of inoculant treatment and forage x inoculant on the amount of effluent produced from the silages were not significant. Analysis show that lucerne silage effluent had a higher DM content than red clover silage effluent (P < 0.001) and treating ryegrass silage with an inoculant increased the effluent DM concentration (P < 0.01). Effluent produced from lucerne silage had a higher N concentration than effluent from ryegrass or red clover silage (P < 0.001) but there was no effect of inoculant on effluent N concentrations. The total DM and N lost in effluent from the legume silages was higher than from ryegrass silage (P < 0.01). Red clover silage treated with inoculant had a higher effluent DM loss than the lucerne or ryegrass silages (P < 0.01). Ryegrass silage treated with inoculant had a higher N loss in effluent than untreated ryegrass silage (P < 0.01).

SID 5 (2/05) Page 18 of 28

Table 5.2. Effect of forage and silage inoculant on the mean volume (litres t -1 fresh herbage) and composition (kg / t fresh herbage) of effluent produced and the total DM and N lost (kg / t fresh herbage) in effluent produced from 10 kg silos during day 1-11 of ensiling.

Ryegrass Red clover Lucerne s.e.d. Effect-I +I -I +I -I +I F I FxI

VolumeDays 1 - 3 0.0 0.0 156.3 161.8 81.2 88.8 4.73 *** ns nsDays 4 - 5 26.3 30.8 4.0 7.3 8.3 3.8 5.02 *** ns nsDays 1 - 11 29.0 39.8 162.8 182.8 93.7 95.3 10.42 *** ns ns

CompositionDM (g/kg) 61 77 51 52 69 74 3.0 *** ** *N (g/kg) 1.8 2.2 2.5 2.5 4.5 4.1 0.30 *** ns ns

Losses DM 1.8 3.0 8.4 9.6 6.5 7.1 0.64 *** * nsN 0.05 0.08 0.41 0.45 0.42 0.39 0.118§ *** * **

§ applicable to data on natural logarithm scale; 12df for error.

Ryegrass silage had a higher DM content than red clover or lucerne silage (P < 0.05) (Table 5.3). Red clover silage treated with inoculant had a higher DM content than untreated red clover silage (P < 0.001). Ryegrass had a lower pH than red clover which, in turn, had a lower pH than lucerne silage (P < 0.01). Silage inoculant treatment reduced the pH of ryegrass and red clover but not lucerne silage (P < 0.05). Ammonia-N concentrations were higher in lucerne silage and lower in ryegrass silage compared with red clover silage and inoculant treatment reduced ammonia-N concentrations (P < 0.001). Legumes silages had higher crude protein concentrations and lower water-soluble carbohydrate concentrations than ryegrass silage (P < 0.001). Silages treated with inoculant had higher WSC and propionic acid concentrations than untreated silages (P < 0.05). Lactic acid was not detected in the lucerne silages. Ryegrass silage treated with inoculant had higher lactic acid concentrations compared with untreated ryegrass or red clover silage (P < 0.05) and treating red clover silage with inoculant resulted in higher lactic acid concentrations compared with untreated red clover silage (P < 0.001). Lucerne silage had higher propionic and butyric acid concentrations than red clover silage (P < 0.05). Propionic and butyric acid concentrations in ryegrass silages were negligible.

Table 5.3. Effect of silage inoculant on the chemical composition of red clover, lucerne or hybrid ryegrass when ensiled in 10 kg silos for 90 days.

Ryegrass Red clover Lucerne s.e.d. Effect-I +I -I +I -I +I F I FxI

Dry matter (g/kg) 201 195 164 178 186 186 3.1 *** ns **pH 3.65 3.47 4.38 3.87 5.19 5.52 0.014§ *** *** ***Crude Protein 92 97 222 210 226 223 0.028§ *** ns nsNH3-N (g/kg TN) 55 20 104 51 238 193 0.052§ *** *** ***WSC 33 49 3 8 6 5 0.112§ *** *** ***Lactic acid 24 42 12 34 nd nd 0.075§‡ *** *** ***Acetic acid 22.3 7.6 29.0 17.2 59.2 32.4 0.074§ *** *** ***Propionic acid nd nd 0.8 0.2 7.1 5.9 0.47‡ *** * nsButyric acid nd nd 11.3 3.8 44.4 45.0 1.75‡ *** * *

F, forage; I, inoculant; ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; § applicable to data on natural logarithm scale; 12df for error; nd not detected; ‡ 8 df for error comparison of two levels of forage only

5.1.5 Discussion and ConclusionsUsing unwilted forage in the experiment was designed to ensure that adequate effluent was produced for the study. DM content at ensiling has long been recognised as a major factor affecting the amount of effluent produced when ensiling a forage crop (Woolford, 1978). Other factors may contribute, such as consolidation, forage chop length as well as the degree of stem to leaf of the ensiled forage, to increased movement of moisture within the silo. In the current experiment, the DM content at ensiling affected the amount of effluent produced as the red clover forage, with the lowest DM content, produced the highest volume of effluent and the ryegrass silage, with the highest DM content, produced the lowest. The effects of ensiling red clover and lucerne silage without wilting are in agreement with previous findings by Owens et al. (2002) who also found it resulted in a poor fermentation and poor quality silage. Overall, the red clover, lucerne and ryegrass forages underwent a moderate, poor and good fermentation during ensiling, respectively. In agreement with other studies, pH values tend to be lower in red clover than lucerne silage. It appears that the lower DM and WSC concentrations of the legume forages resulted in a slower rate of fermentation and produced a higher pH value than in ryegrass in the current experiment, further highlighting the impact of DM content on silage quality when ensiling forage legumes. A rapid decline in pH helps limit the extent of protein degradation in the silo by reducing the activity of plant proteases. It

SID 5 (2/05) Page 19 of 28

would therefore be expected that the lucerne silage would have a higher ammonia-N concentration than the red clover silage as was found. However, red clover also typically undergoes less protein degradation than lucerne during fermentation (Papadopoulos and McKersie, 1983; Marley et al., 2003). One explanation for this is that red clover contains an enzyme, polyphenol oxidase, which inhibits proteolysis in the silo (Jones et al., 1995). Inoculants act by establishing a dominant population of lactic acid bacteria to maximise the efficiency of fermentation of water-soluble carbohydrates to lactic acid, causing a rapid decline in pH. The findings in this study are in agreement with previous studies which showed that inoculant treatment had no effect on the total amount of silage effluent produced (Kennedy, 1990) but did reduce the breakdown of N in silo and improved the quality of the silage that was produced (Winters et al., 2001), as indicated by the lower ammonia-N, acetic acid and butyric acid concentrations and higher WSC and lactic acid concentrations. The combination of the higher initial CP concentration of the legumes silages, and the poor fermentation that occurred in these legume silages, contributed to the differences observed in the total N concentrations in the resultant silage effluent and total N lost in effluent when compared with ryegrass silage. In conclusion, the addition of silage inoculant to red clover, lucerne and ryegrass forage did not alter the amount of silage effluent production or the N concentration of that effluent. This study highlights the importance of wilting prior to ensiling legume forages, such as red clover and lucerne, if good quality silages are to be produced and silage effluent and silage effluent N losses are to be avoided. However, when this is not possible, using an inoculant will improve the fermentation process in the silo.

Experiment 5.2 Effects of silage effluent from red clover, lucerne or ryegrass on leachates and soil nutrient status. 5.2.1 Introduction As silage effluent arises from a combination of surface water and plant juices expelled from the ensiled forage (Arnold et al., 2000), the total amount of effluent produced per tonne of fresh crop ensiled can be up to 500 litres depending on the DM content of the forages. The Codes of Good Agricultural Practice provide guidance to landowners on the control and handling of silage effluent from predominantly grass silage crops (MAFF, 1998). However, there is little information with which to generate guidelines on potential effluent flows from other forage crops, such as red clover or lucerne. Consequently, there is a need to explore the effects of effluent from these ensiled forages when applied to soil and to provide land owners with guidance on their effects on leachates and soil mineral status. This experiment tested the hypothesis that adding of silage effluent from red clover or lucerne would not alter the leachates or mineral status of a soil when compared with effluent from ryegrass silage.

5.2.2 Materials and MethodsSilage effluent was obtained when ryegrass, red clover and lucerne forages ensiled without wilting. The experimental approach was as described in Experiment 5.1. The experiment was a 3 x 4 treatment design with 3 soil types and silage effluent from 3 different forages and deionised water as a control. Three contrasting soil types were used in the experiment:

1. Sandy loam over soft sandstone, porous freely draining, neutral to slightly acidic, (Crediton series) 2. Porous and freely draining, highly calcareous silty clay loam over chalk (Andover series).3. Stony, well-drained medium loam (Rheidol series)

Soil cores were obtained from an area that had not received any manure or been grazed for minimum of 6 months nor had it been subjected to compaction prior to sampling. Intact grassland soils, 19cm diameter x 0.5 m deep, were collected by using a tractor-mounted post-driver and a 20 cm diameter x 0.5 m deep stainless steel tube. Plastic drainage pipe was cut into 0.5 m lengths and filed at one end to give a sharpened edge. The pipe was then placed into a specially designed steel cylindrical case with a sharpened basal cutting edge and steel handles. A block of wood, supported by steel strapping and with shaped steel over one end was used to form a lid. A mechanical excavator was used to extract the core in its steel container. The plastic case, containing the soil core was then removed from the metal casing and both ends of the corer were covered with a secured plastic bag whilst the cores were transported back to the laboratory. In the laboratory, the bottom end of the plastic pipe was sealed with a PVC drainage end cap, filled with 5 mm diameter plastic beads (Distrupol, Surrey, UK). A drainage hole was drilled into the bottom of each cap and fitted with a plastic tube in the middle of each PVC cap were fitted to the drainage holes to transfer the lechate into 5 l plastic containers. For the duration of the experiment, the soil cores were placed on a purpose-built supporting stand, kept in a well ventilated building and the top of each soil core was covered with a loosely fitted plastic bag to reduce losses by evaporation. The soil cores were maintained for 8 weeks whilst an equivalent volume of water to equal 3 pore volumes (calculated from porosity = ((s-d) /s)*100, where s = particle density and d= bulk density) were added to each core. The aim was that each core would be representative of soils that were low in nutrients prior to the application of silage effluent, as they would be in standard farm practice. To simulate typical rainfall that would occur in the field, 407 cm of deionised water was added in two applications (9am and 4.30pm) every 7 d throughout the experiment. Rainfall simulation was based on the average rainfall during May recorded at IGER between 1954 -2000. Silage effluent was added according to standard recommended code of practice by diluting the effluent 1:1 with deionised water prior to application and not exceeding a rate of 50 m3 ha-1. A total of 140 ml of each diluted effluent was thoroughly stirred and applied to the surface of the soil cores at surface height (Day 0). Leactates were collected into vessels that were kept stored in fresh ice placed under each soil core. Leachates were collected immediately

SID 5 (2/05) Page 20 of 28

before and 48 h after rainfall was applied every 7 d and pH determined. Sub-samples to determine the biological oxygen demand (BOD) of the leachates were stored in cooled, sealed containers and determined within 36 h using the standard operating procedure used by the Environment Agency, Llanelli, UK. All data were analysed using Genstat version 8.1 (Payne et al. 2005). Silage effluent volumes and chemical composition were compared by a two-way ANOVA. Data analysis of the total N leached over 9 weeks from the cores showed that there was significant between core variation (P < 0.01) so this data for Andover and Rheidol series soil cores were analysed separately.

5.2.3 ResultsPrior to application, data showed that the BOD of red clover effluent was 16,202 (s.e. 828) mg L -1 which was numerically lower than that of lucerne effluent at 21,977 (s.e. 970) mg L -1 and was, in turn, numerically lower than that of ryegrass effluent 25,644 (s.e. 165) mg L-1. The pH of red clover, lucerne and ryegrass effluent was 4.5, 5.3 and 3.7, respectively. The DM and N concentration of the applied effluents are presented in Table 5.2. The soil cores consisting of the Crediton Series were found to become compacted, despite being collected and maintained under the same conditions as the other cores, and this treatment was discontinued. Results showed the BOD of the leachates from the soil cores were found to be negligible. There was no effect of effluent treatment on the pH of leachate from the columns or the destructive soil pH values. The total N in the leachate over 9 weeks from the Rheidol series soil cores showed no effect of silage effluent treatment. The total N in the leachate over 9 weeks from the Andover series soil cores treated with lucerne effluent was higher than other treatments (P < 0.05). Destructive soil samples taken from soil cores treated with lucerne silage effluent had a higher total N concentration than other treatments (P < 0.01) and a total N concentration of the destructive soil samples differed between the two treatment sites (P < 0.05) but there were no effluent treatment x site interactions.

5.2.4 Discussion and ConclusionsThe finding that the soil cores were able to reduce the concentrations of BOD in the silage effluent in the current study is in agreement with other studies into the effects of applications of farm wastes to soil (Brookman et al., 1996). The higher N concentration of the destructive soil samples from cores treated with lucerne effluent compared with ryegrass or red clover was probably due to the higher N concentration of this effluent. Overall, the results suggest that using effluent from alternative forages when compared with ryegrass silage effluent can improve the nutrient status of a soil without affecting the BOD of leachates. However, due to the risk that silage effluent from high protein forages may contain higher N concentrations, care needs to be taken to avoid N losses in leachates when applying these silage effluents to free-draining soils.

Experiment 5.3: The effects of an in-silo absorbent on the quantity of effluent produced when ensiling forage peas with and without wilting. 5.3.1 Introduction Silage effluent is one of the worst pollutants produced on farm and farmers are encouraged to reduce water pollution risks when silage making. Ensiling forage in dry weather and wilting to a suitable dry matter (DM) (at least 250 g/kg DM) will reduce silage effluent production. However, weather conditions may not permit the wilting of forage prior to ensiling. An experiment tested the hypothesis that the addition of molassed sugar beet pulp (MSBP) could prevent silage effluent production from forage peas (Pisum sativum) when ensiled without wilting.

5.3.2 Materials and MethodsForage peas (cv. Espace) were sown on 22 April 2003 at a rate of 217 kg/ha and were harvested on 22 July 2003. The harvested forage was chopped using a stationary modified precision-chop forage harvester, treated with a silage inoculant and ensiled in mini-silos. Treatments consisted of peas either unwilted or wilted for 24 h with either 0, 50 or 100 kg MSBP / t fresh forage added. The mini-silos were prepared by weighing 10 kg of chopped forage and then spreading this on a 2 m x 3 m polythene sheet. The inoculant (Lactobacillus plantarum) (Live System, Genus Ltd., UK) was applied at 106 colony forming units per gram fresh matter (equivalent to the manufacturer’s recommended rate of 2 l / t) using a pressurised hand sprayer. The forage was then mixed thoroughly before being placed into a 50 L insulated bin, lined with a bottomless 250-gauge polythene bag. The filled bin was weighed, and the weight of the forage calculated by subtraction. The plastic liner within each bin was sealed at the top with a rubber ring, and the forage within each bin was then compacted for 10 d under a lever and piston system at 817 kg/m2 pressure. Effluent traps were fitted to each silo and the effluent produced was collected over 10 days. After 11 d, a 5 kg sandbag was placed on top of the forage and the bin sealed with a plastic lid. The mini-silos were opened after 90 days and representative sub-samples analysed for dry-matter content, ammonia-N (NH3-N), total N, water-soluble carbohydrates (WSC), VFA concentrations and pH.

5.3.3 Results Pea silages that were wilted for 24 h did not produce any silage effluent nor did unwilted silages to which 100 or 150 kg MSBP / t fresh forage were added. Unwilted pea silage with 50 kg MSBP / t fresh forage produced a lower mean volume of silage effluent (40 ml, s.d. 9) compared with unwilted pea silage without MSBP (185 ml, s.d. 29) (P < 0.001). There was no effect of treatment on DM loss from the silages. Prior to ensiling, the DM content of the unwilted and wilted forage was 204 g/kg (s.d. 2.7) and 296 g/kg (s.d. 25.4), respectively and the WSC

SID 5 (2/05) Page 21 of 28

concentration of the unwilted and wilted forage was 131 g/kg DM (s.d. 2.5) and 126 g/kg DM (s.d 5.5). The DM content of the MSBP was 922 g/kg and the WSC concentration was 211 g/kg DM. Silage fermentation was improved with the addition of MSBP as indicated by the higher residual WSC (P < 0.001) and lower ammonia-N (P < 0.001) concentrations compared with silages with no MSBP added (Table 5.3). Silages with MSBP had a lower CP concentration than silages without MSBP (P < 0.001) which was due to the lower CP concentration of the additive compared with the pea silage.

Table 5.3. Chemical composition of pea silage either unwilted or wilted for 24 h with either 0, 50 or 100 kg molassed sugarbeet pulp (MSBP) / t forage.

Unwilted WiltedMSBP Wilting

MSBP x

Wilting

MSBP x

WiltingMSBP 0 50 100 150 0 50 100 150 effect effect s.e.d effectDMWSCCPNH3-N pH LactateAcetate

20513

17292

3.7545.617.7

22926

15872

3.7043.618.0

25844

14861

3.7134.417.4

28963

13659

3.7235.215.8

28818

16967

3.7837.320.1

37036

16149

3.7728.813.3

35052

15447

3.7631.415.6

40577

14745

3.8026.311.9

*************

******

********************

17.12.31.85.6

0.0182.331.33

nsns***nsns***

WSC, water-soluble carbohydrates; CP, crude protein; NH3-N; ammonia-N (as g kg-1 N); ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

5.3.4 Discussion and ConclusionsResults showed that wilting peas for 24 hours prior to ensiling or adding 100 kg MSBP / t fresh forage eliminated the production of silage effluent. Furthermore, the fermentation of pea silages, wilted or not wilted for 24 hours, was significantly improved by the addition of MSBP. Further studies are now needed to determine the exact relationship between the DM of peas at ensiling and the amount of MSBP needed to prevent effluent production.

Experiment 5.4: The effects of two different in-silo absorbents on the quantity of effluent produced when ensiling forage kale without wilting 5.4.1 Introduction Silage effluent is an extremely powerful pollutant, with a biological oxygen demand (BOD) of between 30,000 - 80,000 mg O2/litre (Beck, 1989), which can be degraded four times faster than sewage and, thereby, has the potential to be 1000 times more potent. The consequence of silage effluent reaching a watercourse is often a rapid deoxygenation of the water and a decrease in pH, resulting in the death of aquatic flora and fauna (Environment Agency, 2005). Forage kale can be grown as an early-sown catch-crop or a late-sown main-crop with potential for a high CP concentration and DM yield compared to other brassica crops (Martyn et al., 1997). Studies have shown that forage kale, when offered ensiled, can improve growth rates, food conversion and N-utilisation efficiency when compared with lambs fed ensiled ryegrass (Marley et al., 2007). As silage effluent arises from a combination of surface water and plant juices expelled from the ensiled forage (Arnold et al., 2000), the total amount of effluent produced per tonne of fresh crop ensiled can be up to 500 litres depending on the DM content of the forages. The Codes of Good Agricultural Practice provide guidance to farmers on the control and handling of silage effluent from predominantly grass silage crops (MAFF, 1998). However, there is little information with which to generate guidelines on potential effluent flows from inherently low dry matter crops, such as forage kale which has a typical DM of 120 g/kg DM. Consequently, there is a need to explore strategies, such as using different absorbents, to provide guidance on the control of effluent from this low DM forage. An experiment tested the hypothesis that the addition of molassed sugar beet pulp (MSBP) or rolled barley would reduce silage effluent production from forage kale when ensiled without wilting.

5.4.2 Materials and Methods Forage kale was established on an area of stony, well-drained loam of the Rheidol series at the Institute of Grassland and Environmental Research (IGER), Aberystwyth, Wales (52 27’N, 4 01’W). To achieve an optimal soil pH of 6.5, ground limestone was applied at a rate of 5 t/ha. Compound fertiliser was applied to achieve phosphate and potash indices of 2+ or 3 (MAFF, 2000). Kale (‘Kaleage’, Sharpes International Seeds Ltd, Sleaford, UK; cv. Pinfold and Keeper in a 2:1 ratio by weight) was sown on 22 April 2003 at 7.5 kg/ha. Weather conditions did not allow for a pre-emergence herbicide application. On 12 May 2003, Lupus slug pellets (3% methiocarb; Bayer plc., Bury St Edmunds, Suffolk) were applied at 5 kg/ha. Herbicide, Semeron 25WP (25 % desmetryn; Syngenta, Whittlesford, Cambridgeshire) was applied at 1.1 kg/ha on 16 June 2003. The kale was harvested at a height of 100 mm on 12 August 2003. The harvested forage was treated with a silage inoculant (Sil-All 4x4™, Alltech, Stamford, UK), applied at 106 colony forming units per gram fresh matter (2 ℓ/t fresh forage), whilst being chopped using a stationary modified precision-chop forage harvester. Treatments consisted of kale unwilted with either 0 (control), 25, 50, 75 kg rolled barley/t fresh forage added or 25, 50, 75 kg MSBP/t fresh forage added. The mini-silos were prepared by spreading 10 kg of chopped forage on a 2 m x 3 m

SID 5 (2/05) Page 22 of 28

polythene sheet and adding the required amount of each absorbent according to treatment. The forage and absorbent were then mixed thoroughly before being placed into a 50 ℓ insulated bin, lined with a bottomless 250-gauge polythene bag. The filled bin was weighed, and the weight of the forage calculated by subtraction. The plastic liner within each bin was sealed at the top with a rubber ring, and the forage within each bin was compacted for 10 d under a lever and piston system at 817 kg/m2 pressure. Effluent traps were fitted to each silo and the effluent produced was collected over 10 days. At the end of the 10 day compaction period, a 5 kg sandbag was placed on top of the forage and the bin sealed with a plastic lid. Data were analysed by regression analysis using Genstat, Version 8.1 (Payne et al., 2005).

5.4.3 Results Prior to ensiling, the DM content of the unwilted kale, the MSBP and the rolled barley was 127 (s.d. 4.0), 923 and 866 g / kg fresh weight, respectively. Regression analysis showed that there was a linear decrease in the amount of effluent produced when there was an increase in the inclusion rate of either MSBP or rolled barley. MSBP was found to reduce the amount of effluent produced more effectively than rolled barley, with the line of best fit for MSBP described by: effluent = 2311 (± 54.3) - 19.36 (± 1.22)*rate of inclusion (r2 =0.98; P < 0.001) compared with a line of best fit for rolled barley as: effluent = 2311 (± 54.3) – 4.17 (± 1.22)*rate of inclusion (r2 = 0.99; P < 0.003).

5.4.4 Discussion and ConclusionsThe findings of the current experiment are in agreement with previous studies conducted on the effects of different silage absorbents on effluent production, although these were typically based on ryegrass which has a higher DM content than kale. Ferris and Mayne (1994) found that the addition of either 40 or 120 kg of unmolassed sugar beet pulp/t fresh forage retained 1.62 and 1.64 ℓ effluent/kg forage, respectively, from ryegrass with a DM content of 140 g/kg fresh weight. With respect to rolled barley, Done (1988) reported a 70 % reduction when rolled barley at 40 kg/t fresh forage was incorporated in an unwilted ryegrass crop with a DM content of 140 g/kg fresh weight. The finding that rolled barley was not as effective as MSBP at reducing silage effluent production is also in agreement with a laboratory experiment comparing the water holding capacity of a range of potential silage effluent absorbents (Dexter, 1961). Under laboratory conditions, sugar beet pulp was found to be more absorbent than cereal grains in general, although the study did not compare MSBP with rolled barley directly. Overall, the findings of this experiment demonstrate that the inclusion of absorbents to silage, made from inherently low dry matter crops, when weather conditions do not permit the forage to be wilted prior to ensiling can reduce the production of silage effluent, thus reducing the risks to the environment. Further work is now needed to determine the effects of these different silage absorbents on silage fermentation, in-silo losses and the nutritive value of the kale silage.

Objective 6: The potential of alternative forages to supply minerals and trace elements to ruminant livestock, and associated efficiencies of use.6.1 Introduction Up to 75% of excreted N can be quickly lost from slurry as ammonia whereas losses of non-volatiles, such as P and K, are small. Due to these changes, slurry becomes increasingly mineral-rich relative to plant fertilizer needs which has implications for farm nutrient budgeting. For example, following N losses from slurry, the land area needed to recycle slurry P is greater than that needed for slurry N and, thus the impact of diet on P excretion is becoming increasingly important in livestock systems (Powers and Horn, 2001). Research has shown that ensiled alternative forages, such as forage kale (Brassica oleracea), red clover (Trifolium pratense) and lucerne (Medicago sativa), can be used as a productive winter feed for ruminants (Marley et al., 2007). Despite the potential for these alternative forages, there was little information on their effects on mineral budgets within a livestock system. Three experiments tested the hypothesis that the excretion of minerals differs in lambs offered different ensiled forages and to examine the findings with respect to their implications for a farm nutrient budget.

6.2 Materials and Methods Three experiments were conducted comparing different silage treatments when offered to growing lambs. Experiment one compared lambs offered ensiled forage peas (Pisum sativum) or field beans (Vicia faba), experiment two compared lambs offered ensiled red clover, lucerne or birdsfoot trefoil (Lotus corniculatus) and experiment three compared lambs offered ensiled sainfoin (Onobrychis viciifolia) or kale. Forages were established on an area of stony, well-drained loam of the Rheidol series at the Institute of Grassland and Environmental Research (IGER), Aberystwyth, Wales (52 27’N, 4 01’W). To achieve an optimal soil pH of 6.5, ground limestone was applied at a rate of 5 t ha -1. Compound fertiliser was applied to achieve phosphate and potash indices of 2+ or 3 (MAFF, 2000). All forages were treated with herbicides and fungicides according to standard farm practice. For experiment one, forage peas (cv. Magnus) and field beans (cv. Maya) were sown on 29 April 1998 at seed rates of 212 kg ha -1 and 280 kg ha-1, respectively. For experiments two and three, the lucerne (cv. Vertus), birdsfoot trefoil (cv. Leo) and sainfoin (unhulled) (cv. Somborne) seed was inoculated with species-specific rhizobia prior to sowing on 1 May 1997 at rates of 25, 18 and 68 kg /ha. The red clover (cv. Merviot) and kale (‘Kaleage’, cv. Pinfold and Keeper in a 2:1 ratio by weight) were sown on the same date at a rate of 20 and 6.4 kg / ha, respectively. Silages were prepared from peas and beans cut 14 weeks after

SID 5 (2/05) Page 23 of 28

sowing, from second-cut lucerne, red clover and birdsfoot trefoil (22 September), from first cut sainfoin (27 July) and from kale cut 17 weeks after sowing. All forages were cut with a crimper mower (Vicon KM 241 Olympus, Vicon UK Ltd, Market Drayton, UK) and left in swaths to wilt for 48 h. Forages were baled using a fixed chamber round baler (Greenland RF 120, Greenland UK ltd., Market Drayton, UK) and wrapped in 6 layers of 25 m x 750 mm wide film (Silotite; bpi.agri, Leominster, UK). No additive treatments were used.In each experiment, six Suffolk-cross castrated male lambs were restrictively randomised to each treatment on the basis of live weight (average 46 kg; s.d. 3.6) and condition score. Lambs were treated for internal parasites, group housed for 14 days and offered ad libitum access to different silage treatments. Lambs were then placed in metabolism crates for a 7-d adaptation period before data was collected over a 7-d measurement period. All silage fed was chopped to a uniform length using a commercial mixer wagon, thoroughly mixed and stored at 4°C in baskets of 25 kg of silage prior to feeding. Lambs were offered the forage ad libitum and the fresh weights and dry matter (DM) contents of the feed offered and refused were recorded daily. Sub-samples of the feed ‘as offered’ were collected and stored at -20C for chemical analysis. Daily output of urine from each animal was preserved by acidification and its volume measured. Total daily production of faeces for each animal was weighed and stored frozen. Daily sub-samples of urine (5 % of total) and faeces from individual animals during the measurement week were bulked, thoroughly mixed, sub-sampled and stored at -20ºC prior to analysis. DM content of feed offered, refused and all faeces were determined by oven-drying at 80 °C and DM of the samples for chemical analysis determined by freeze-drying. Sub-samples of the silages, faeces and urine were analysed for P and K concentrations according to the method of Cavell (1955) and Allen et al. (1974), respectively. Data were analysed by ANOVA, using Genstat, Version 8.1. Treatment means in experiment two were compared using a Student-Newman-Keuls test.

6.3 Results Lambs offered field beans excreted more P in their urine compared with lambs offered forage peas (P < 0.05) but there was no difference between these treatments for faecal P excretion. Lambs offered forage peas consumed 1.4 times more K (P < 0.05) and excreted 7.2 times more K in their faeces than lambs offered field beans (P < 0.001). Lambs fed birdsfoot trefoil consumed more P than lambs offered red clover or lucerne (P < 0.001) but faecal P output from lambs offered red clover was lower than lambs offered ensiled birdsfoot trefoil or lucerne (P < 0.05). Lambs fed lucerne consumed less K than lambs offered red clover (P < 0.05) or birdsfoot trefoil (P < 0.01). Faecal K output was higher in lambs offered birdsfoot trefoil compared with red clover or lucerne (P < 0.05). There was no effect of treatment on the amounts of urinary P or K excreted. Lambs offered sainfoin consumed more P (P < 0.05) and excreted more faecal P (P < 0.01) but less urinary P (P < 0.001) than lambs offered kale. Lambs offered sainfoin consumed less K (P < 0.05) and excreted 5.1 times more faecal K (P < 0.05) and 2.2 times less urinary K (P < 0.001) than lambs offered kale. Table 6.1. Intake, faecal output and urine output (g/d) of P and K by lambs fed on different silages.

Phosphorus (P) Potassium (K)Intake Faecal

outputUrinary output

Intake Faecal output

Urinary output

Experiment 1Peas 3.3 2.9 0.16 16.4 3.6 7.0Beans 2.8 2.5 0.23 11.9 0.5 8.3S.e.d. 0.38 0.24 0.030 1.79 0.66 0.76Effect ns ns * * *** nsExperiment 2Red clover 5.0b 2.9b 0.66 38.9a 2.4b 20.0Lucerne 4.5b 3.8a 0.46 34.7b 2.7b 20.6Birdsfoot trefoil 6.5a 3.7a 0.64 42.0a 4.7a 19.0S.e.d. 0.26 0.34 0.090 1.93 0.93 1.68Effect *** * ns ** * nsExperiment 3Sainfoin 4.0 3.0 0.24 27.3 4.6 12.4Kale 3.4 2.3 0.45 31.2 0.9 27.3S.e.d. 0.23 0.22 0.047 1.75 1.45 2.18Effect * ** *** * * ***

ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; significant differences between treatments in experiment 2 are denoted by different superscripts within columns (P < 0.05).

6.4 Discussion and Conclusions

SID 5 (2/05) Page 24 of 28

The limitations or benefits of feeding field beans compared with forage peas with regards to farm nutrient budgets appears to differ depending on whether P or K is the element under consideration. In summary, experiment one showed that feeding ensiled field beans compared with forage peas to sheep increased the amount of urinary but not faecal P produced whereas feeding forage peas resulted in lambs excreting 7.2 times more K in their faeces despite only consuming 1.4 times more K than lambs offered field beans. Feeding red clover compared with lucerne silage to lambs may have an important effect on the P and K budget within a livestock system as indicated by the finding that lambs offered red clover excreted less faecal P and the same amount of faecal K despite consuming the same amount of P and more K compared with lambs offered lucerne. In addition, lambs offered red clover excreted less faecal K than lambs offered birdsfoot trefoil despite no differences in the amount of K consumed. Whilst these experiments did not account for differences in N excretion, (previous studies showed that 41, 38 and 39 g N / d was excreted by lambs offered red clover, lucerne or birdsfoot trefoil silage, respectively (Fraser et al., 2000)) or potential differences in the rates of N volatilisation and the land area required for nutrient recycling, the findings indicate that the land area needed to recycle excreted P or K is reduced when lambs are offered red clover compared with the other legumes. This may also be true for P recycling in lamb systems feeding birdsfoot trefoil compared with lucerne. The differences in the ratios of faecal to urinary excretion of K in lambs offered beans compared to peas or offered kale compared to sainfoin may impact on K budgets and recycling as urinary K is more readily available to plants than faecal K (Richards and Wolton, 1976). Overall, feeding different ensiled forages to sheep was shown to alter the excretion of non-volatiles, such as P and K, and has potential implications when developing environmentally-acceptable nutrient balances on farms. Further work is needed to determine if these forages can improve the overall efficiency of nutrient use within livestock systems if we are to understand their potential role in providing longer-term sustainability in farming practice.

OUTCOMES AND POSSIBLE FUTURE RESEARCHThis research has demonstrated the potential benefits and limitations of integrating a range of forages, including high-protein leguminous crops and catch crops, into livestock production systems. The findings have shown the potential to use home-grown forages to build soil fertility, improve nutrient efficiency in livestock, optimise nutrient requirements and thus, maximise nutrient capture and retention; resulting in a win-win scenario for sustainable production and the environment. The work has also emphasised the need for future studies to build on these findings and further our understanding of where, in the plant-animal-soil cycle, we can further improve the sustainability of our livestock systems in the UK through the use of these and other novel home-grown traceable feedstuffs. Examples of further studies that are needed now to assist with the integration of these forages include investigating: their effects on whole farm nutrient balances when fed in differing proportions or as part of total mixed rations; their potential to extend the grazing season or for out-wintering livestock; and, strategies that will allow for the use of these forage crops to further mitigate the impact of UK livestock systems on the environment.

CITED REFERENCES ALLEN S.E., GRIMSHAW H.M., PARKINSON J.A. QUARMBY C. (1974) In: Allen S.E. (ed). Chemical analysis of ecological materials. pp. 565. Oxford: Blackwell Scientific Publications.ARNOLD J.L., KNAPP J.S. and JOHNSON C.L. (2000) The use of yeasts to reduce the polluting potential of silage effluent. Water Research, 34, 3699-3708.BECK L. (1989) A review of farm waste pollution. Journal of the institution of water and environmental management, 3, 467-477.BREMNER J.M. and KEENEY D.R. (1965) Steam distilled methods for determination of ammonium, nitrate and nitrite. Analytica Chimica Acta, 32, 485-497.BROOKMAN S.K.E., CHADWICK D.R. and PAIN B.F.(1996) The fate of biochemical oxygen demand following applications of farm wastes to soils. Final Report for MAFF, Project WA0505.BUTLER  J. H. A. (1987) The effect of defoliation on growth and N2 fixation by Medicago sp. grown alone or with ryegrass. Soil Biology and Biochemistry, 19, 273-279.CARLSSON G. and HUSS-DANELL K. (2003) Nitrogen fixation in perennial forage legumes in the field. Plant and Soil, 253, 353-372. CAVELL A.J. (1955) The colorimetric determination of phosphorus in plant materials. Journal of the Science of Food and Agriculture. 6, 479–480.CUSSEN R.F., MERRY R. J., WILLIAMS A.P. and TWEED J.K.S. (1995) The effect of additives on the ensilage of forage of differing perennial ryegrass and white clover content. Grass and Forage Science, 50, 249-258.DHANOA M.S. (1998) A procedure for Theil’s regression method. GenStat Newsletter, 34, 21-26.DEXTER S.T. (1961) Water retaining capacity of various silage additives and silage crops under pressure. Agronomy Journal, 53, 379-381.DONE D.L. (1988) The effect of absorbent additives on silage quality and on effluent production. In: Stark B.A. and Wilkinson J.M. (eds.) Silage effluent. pp.49. Lincoln, UK: Chalcombe publishers. ENVIRONMENT AGENCY (2005) Water pollution incidents in England and Wales 2004. Environment Agency, Bristol, UK.FARNHAM D. E. and GEORGE J. R. (1994) Harvest management effects on dinitrogen fixation and nitrogen transfer in red clover-orchardgrass mixtures. Journal of Production Agriculture, 7, 360-364.  FRASER M.D., FYCHAN R. and JONES R. (2000) Voluntary intake, digestibility and nitrogen utilisation by sheep fed ensiled forage legumes. Grass and Forage Science, 55, 271-279.

SID 5 (2/05) Page 25 of 28

FRASER M.D., FYCHAN R. and JONES R. (2001a) The effect of harvest date and inoculation on the yield, fermentation characteristics and feeding value of kale silage. Grass and Forage Science, 56, 151-161.FRASER M.D., FYCHAN R. and JONES R. (2001b) The effect of harvest date and inoculation on the yield, fermentation characteristics and feeding value of forage pea and field bean silages. Grass and Forage Science, 56, 218-230. FERRIS C.P. and MAYNE C.S. (1994) The effects of incorporating sugar-beet pulp with herbage at ensiling on silage fermentation, effluent output and in-silo losses. Grass and Forage Science, 49, 216-228.FRASER M.D., FYCHAN R. and JONES R. (2000) Voluntary intake, digestibility and nitrogen utilisation by sheep fed ensiled forage legumes. Grass and Forage Science, 55, 271-279.GIVENS D.I., EVERINGTON J.M. and ADAMSON A.H. (1989) The digestibility and ME content of grass silage and their prediction from laboratory measurements. Animal Feed Science and Technology, 24, 27-43.HESS, T (2002) Potential Evapotranspiration Program for Automatic Weather Stations. Version 3.0 - User Manual. Cranfield University, Silsoe, UK, 29pp.JONES B.A., MUCK R.E. and HATFIELD R.D. (1995) Red clover extracts inhibit legume proteolysis. Journal of the Science of Food and Agriculture, 67, 329-333.KENNEDY S. J. (1990) An evaluation of three bacterial inoculants and formic acid as additives for first harvest grass. Grass and Forage Science, 45, 281-288.MAFF (1986) Analysis of agricultural materials. Reference Book 427. London, UK: Ministry of Agriculture, Fisheries and Food, and HMSO.MAFF (1998) Code of good agricultural practice for the protection of water. Ministry of Agriculture, Fisheries and Food Publication (PB 0587), HMSO, London, UK.M.A.F.F (2000) Fertiliser recommendations for Agricultural and Horticultural Crops. Reference Book 209. 7th

edition. Ministry of Agriculture, Fisheries and Food, HMSO, London, UK.MALLARINO A.P., WEDIN W.F., GOYENOLA R.S., PERDOMO C.H. and WEST C.P. (1990) Legume species and proportion effects on symbiotic dinitrogen fixation in legume-grass mixtures. Agronomy Journal, 82, 785-789.MARLEY C.L., FYCHAN, R., FRASER, M.D., SANDERSON R. and JONES, R. (2007) Effects of feeding different ensiled home-grown forages on the productivity and nutrient use efficiency of store finishing lambs. Grass and Forage Science, 62, 1-12.MARLEY C.L., FYCHAN, R., FRASER, WINTERS, A.L. and JONES, R. (2003) The effect of sowing ratio and stage of maturity at harvest on yield, persistency and chemical composition of fresh and ensiled red clover/lucerne. Grass and Forage Science   58 (4)   397-406MARTYN T.M., YOUNG N.E. and BALSDON S.L. (1997) Comparison of the herbage yields of three kale (Brassica oleracea L.) cultivars. Annals of Applied Biology, 130, 44-45.MERRY R.J., DHANOA M.S. and THEODOROU M.K. (1995) Use of freshly cultured lactic acid bacteria as silage inoculants. Grass and Forage Science, 50, 112-123.MERRY R.J., JONES, R. and THEODOROU, M.K. (2000) The conservation of grass. In: Hopkins A. (ed.) Grass. Its production and utilisation. 3rd. edn. Oxford: UK. Blackwell Science Ltd.OWENS V.N. ALBRECHT, K.A. and MUCK, R.E. Protein degradation and fermentation characteristics of unwilted red clover and alfalfa silage harvested at various times during the day. Grass and Forage Science, 57, 329-341.PAYNE R.W., MURRAY D.A., HARDING S.A., BAIRD D.B. and SOUTAR D.M. (2005) Genstat® for WindowsTM

(8th Edition) Introduction. Oxford: VSN International.PAPADOPOULOS Y.A. and McKERSIE B.D. (1983) A comparison of protein degradation during wilting and ensiling of six forage species. Canadian Journal of Plant Science, 63, 903-912.POWERS W.J. and HORN H.H. van. (2001) Nutritional implications for manure nutrient management planning. Applied Engineering in Agriculture, 17, 27-39.RICHARDS I.R. and WOLTON K.M. (1976) The spatial distribution of excreta under intensive cattle grazing. Grass and Forage Science, 31, 89 – 92.SPEIJERS M.H.M., FRASER M.D., THEOBALD V.J and HARESIGN W. (2005) Effects of ensiled forage legumes on performance of store finishing lambs. Animal Feed Science and Technology, 120, 203-216.THOMAS T.A. (1977) An automated procedure for the determination of soluble carbohydrates in herbage. Journal of the Science of Food and Agriculture, 28, 639-642.VAN SOEST P.J., ROBERTSON J.B. and LEWIS B.A. (1991) Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. Journal of Dairy Science, 74, 3583-3597.VERKROOST A. W. M. and WASSEN M. J. (2005) A simple model for nitrogen-limited plant growth and nitrogen allocation. Annals of Botany, 96, 871-876.VIPOND J.E., DUNCAN A.J., TURNER D., GODDYN L. and HORGAN G.W. (1998) Effects of feeding ensiled kale (Brassica oleracea) on the performance of finishing lambs. Grass and Forage Science, 53, 346-352.WILKINS R.J., HUTCHINSON K.J., WILSON R.F. and HARRIS C.E. (1971) The voluntary intake of silage by sheep. I. Interrelationships between silage composition and intake. Journal of Agricultural Science, 77, 531-537.WINTERS A.L., FYCHAN, R. and JONES, R. (2001) Effect of formic acid and a bacterial inoculant on the amino acid composition of grass silage and on animal performance. Grass and Forage Science, 56, 181-192.WOOLFORD M.K. (1978) The problem of silage effluent. Herbage Abstracts, 48, 397-403.ZHU W.–Y., THEODOROU M. K., LONGLAND A.C., NIELSON B.B., DIJKSTRA J. and TRINCI A.P.J. (1996) Growth and survival of anaerobic fungi in batch and continuous-flow cultures. Anaerobe, 2, 29-37.

SID 5 (2/05) Page 26 of 28

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.

Marley, C. L. , Fychan, A. R. , Fraser, M. D. , Sanderson, R. , Jones, R. (2007) Effects of feeding different ensiled forages on the productivity and nutrient-use efficiency of finishing lambs Online at: http://www.blackwell- synergy.com Grass and Forage Science   62 (1)   1-12

Cardenas, L. M. , Chadwick, D. R. , Scholefield, D. , Fychan, A. R. , Marley, C. L. , Jones, R.   et al. (2007) The effect of diet manipulation on nitrous oxide and methane emissions from manure application to grassland soils Online at: http://dx.doi.org/doi:10.1016/j.atmosenv.2007.04.055 Atmospheric Environment   41   7096-7107

Marley, C. L. (2007) Better forages for finishing lambs Grass and Forage Farmer (89 (Summer))   4

Marley, C. L. , Fraser, M. D. , Fisher, W. J. , Forbes, A. B. , Jones, R. , Moorby, J. M.   et al. (2007) Effects of continuous or rotational grazing of two perennial ryegrass varieties on the chemical composition of the herbage and the performance of finishing lambs Online at: //Wpnetapp1a\protrak\PUBLICATIONS\ 5177\7481-marley-2007.pdf Grass and Forage Science   62 (3)   255-264

Marley, C. L. , Fychan, A. R. , Roberts, J. E. , Theobald, V. J. , Jones, R. (2007) The effects of two different in-silo absorbents on the quantity of effluent produced when ensiling forage kale without wilting 'High Value Grassland: providing biodiversity, a clean environment and premium products': Proceedings BGS/BES/BSAS Conference, Keele University, Staffs, 17-19 April 2007. British Grassland Society Occasional Symposium, 38   Hopkins, J. J. , Duncan, A. J. , McCracken, D. I. , Peel, S. , Tallowin, J. R. B., eds.   269-272

Fychan, A. R. , Marley, C. L. , Fraser, M. D. , Theobald, V. J. , Jones, R. (2007) The effects of feeding different ensiled forages on the minerals excreted from growing lambs - implications for farm nutrient budgets 'High Value Grassland: providing biodiversity, a clean environment and premium products': Proceedings BGS/BES/BSAS Conference, Keele University, Staffs, 17-19 April 2007. British Grassland Society Occasional Symposium, 38   Hopkins, J. J. , Duncan, A. J. , McCracken, D. I. , Peel, S. , Tallowin, J. R. B., eds.   217-220  

Fychan, A. R. , Marley, C. L. , Roberts, J. E. , Lewis, G. G. , Theobald, V. J. , Jones, R. (2007) Effects of applying slurry on the yield of red clover, lucerne or hybrid ryegrass when cut for silage production 'High Value Grassland: providing biodiversity, a clean environment and premium products': Proceedings BGS/BES/BSAS Conference, Keele University, Staffs, 17-19 April 2007. British Grassland Society Occasional Symposium, 38   Hopkins, J. J. , Duncan, A. J. , McCracken, D. I. , Peel, S. , Tallowin, J. R. B., eds.   213-216  

Hobbs, P. J. , Wade, M. J. , Marley, C. L. , Williams, J. S. , Dhanoa, M. S. , Fraser, M. D.   et al. The potential of metabolic profiling to investigate dietary intake and animal performance Submitted to Journal of Agricultural and Food Chemistry

Marley, C. L. , Fychan, A. R. , Jones, R. (2006) Yield, persistency and chemical composition of Lotus species and varieties (birdsfoot trefoil and greater birdsfoot trefoil) when harvested for silage in the UK Online at: http://www.blackwell-synergy.com/doi/pdf/10.1111/ j.1365-2494.2006.00516.x Grass and Forage Science   61 (2) 134-145

Marley, C. L. , Fraser, M. D. , Roberts, J. E. , Fychan, A. R. , Jones, R. (2006) Effects of legume forages on ovine gastrointestinal parasite development, migration and survival Online at: http://dx.doi.org/doi:10.1016/j.vetpar.2006.02.001> Veterinary Parasitology   138 (3-4)   308-317

Merry, R. J. , Lee, M. R. F. , Davies, D. R. , Dewhurst, R. J. , Moorby, J. M. , Scollan, N. D.   et al. (2006) Effects of high-sugar ryegrass silage and mixtures with red clover silage on ruminant digestion. 1. In vitro and in vivo studies of nitrogen utilization Online at: http://dx.doi.org/doi:10.2527/jas.2005-735 Journal of Animal Science   84 (11)   3049-3060

Hobbs, P. J., Wade, M. J., Jones, R., Marley, C. L., Fraser, M. D., Williams, J. S.   et al. (2006) Managing manure by a greater understanding of its metabolic profile? Adopting new technologies Online at: http://www.manure.dk/ramiran/O-18%20Hobbs.pdf 12th RAMIRAN International Conference: "Technology for recycling of manure and organic residues in a whole-farm perspective", Aarhus, Denmark 11-13 September 2006  139-144

SID 5 (2/05) Page 27 of 28

Fychan, A. R. , Marley, C. L. , Roberts, J. E. , Lewis, G. G. , Theobald, V. J. , Jones, R. (2006) Effects of red clover, lucerne or hybrid ryegrass on the yield of a subsequent crop of winter or spring barley Online at: \\ Wpnetapp1a\protrak\PUBLICATIONS\4204\7582-Fychan.2006.pdf 8th British Grassland Society Research Conference, 'Grassland as a Multifunctional Resource', Royal Agricultural College, Cirencester, 4-6 September 2006   Moorby, J. M., eds.   p 3

Marley, C. L. , Fychan, A. R. , Roberts, J. E. , Theobald, V. J. , Jones, R. (2006) The effects of an in-silo absorbent on the quantity of effluent produced when ensiling forage peas with and without wilting 8th British Grassland Society Research Conference, 'Grassland as a Multifunctional Resource', Royal Agricultural College, Cirencester, 4-6 September 2006   Moorby, J. M., eds.   p 51  

Marley, C. L. , Fychan, A. R. , Fraser, M. D. , Roberts, J. E. , Lewis, G. G. , Theobald, V. J.   et al. (2006) Effects of applying slurries from livestock fed on different forages or inorganic nitrogen on swards of hybrid ryegrass Online at: \\Wpnetapp1a\protrak\PUBLICATIONS\4204\7503-Marley.2006.pdf 8th British Grassland Society Research Conference, 'Grassland as a Multifunctional Resource', Royal Agricultural College, Cirencester, 4-6 September 2006   Moorby, J. M., eds.   p 69

Marley, C. L. , Fraser, M. D. , Fychan, A. R. , Theobald, V. J. , Jones, R. (2005) Effect of forage legumes and anthelmintic treatment on the performance, nutritional status and nematode parasites of grazing lambs Online at: http://dx.doi.org/doi:10.1016/j.vetpar.2005.04.037> Veterinary Parasitology   131 (3-4)   267-282

Fychan, A. R. , Marley, C. L. , Fraser, M. D. , Jones, R. (2005) Effect of feeding red clover, lucerne and kale silage on the voluntary intake and liveweight gain of growing lambs Micro-satellite of International Grassland Congress - Silage production and utilisation incorporating the XIV International Silage Conference, Belfast 3-6 July 2005   Park, R. S. , Stronge, M. D., eds.   p155   Wageningen Adademic Publishers

Lopez, S. , Davies, D. R. , Giraldez, F. J. , Dhanoa, M. S. , Dijkstra, J. , France, J. (2005) Assessment of the nutritive value of cereal and legume straws based on chemical composition and in vitro digestibility Online at: http://dx.doi.org/doi:10.1002/jsfa.2136> Journal of the Science of Food and Agriculture   85 (9)   1550-1557

Cardenas, L. M. , Vallejo, A. , Chadwick, D. , Scholefield, D. , Fychan, A. R. , Marley, C. L.   et al. (2005) Effect on soil N emissions of slurry application from sheep fed different diets 'N management in agrosystems in relation to the Water Framework Directive' 14th N-Workshop, Maastricht, Netherlands, 24-26 October 2005   p80

Jaurena, G. , Moorby, J. M. , Davies, D. R. (2005) Efficiency of microbial protein synthesis on red clover and ryegrass silages supplemented with barley by rumen simulation technique (RUSITEC) Online at: http://dx.doi.org/doi:10.1016 /j.anifeedsci.2004.09.008> Animal Feed Science and Technology   118 (1-2)   79-91

Marley, C. L. , Fychan, A. R. , Fraser, M. D. , Winters, A. L. , Jones, R. (2003) The effect of sowing ratio and stage of maturity at harvest on yield, persistency and chemical composition of fresh and ensiled red clover/lucerne bi-crops Online at: http: // www.blackwell-synergy.com/links/doi/10.1111/j.1365- 2494.2003.00392.x/ full/ Grass and Forage Science   58 (4)   397-406

Fothergill, M., Rees, M. E., Marley, C. L. (2003) Lean towards legumes Farmers Weekly   139 (11)   44

SID 5 (2/05) Page 28 of 28