Improving the efficiency and sustainability of milk production systems through the production and utilisation of high quality grass silage

Project code: D-89-17

Objectives

The overall objective of this project is to conduct a targeted series of research studies which will address knowledge gaps related to the production and utilisation of high quality grass silage in dairy cow diets. This will be achieved through desk based exercises, grass plot and ensilage studies, and dairy cow feeding studies, thus achieving a holistic approach to the research programme. Specifically, the objectives can be defined as follows:

1. Through a short survey prior to the commencement of the project, to identify barriers to improving silage quality within Northern Ireland.
2. Through an interrogation of the Hillsborough Feeding Information System (HFIS) data base, to identify trends in the composition of silages produced within NI over the last 15 years, and where possible, reasons for the trends. This aspect of the project will also quantify the frequency of production of very high dry matter silages, which may in itself highlight the need to undertake additional research to examine the efficiency with which these silages are utilised.
3. Through a series of plot and ensilage studies, to identify, i) the impact of 'winter growth' on silage composition and fermentation characteristics, ii) the interaction of nitrogen fertiliser type and nitrogen uptake rates over the growing season, on herbage composition and ensilability, iii) the impact of harvesting interval on the total yield of utilisable nutrients harvested over the growing season.
4. Through a series of dairy cow feeding studies, to examine, i) the impact on cow performance of 'winter growth herbage' when harvested and ensiled for the production of first cut silage, ii) the impact of very high quality silage on cow performance and nutrient utilisation, and strategies to mitigate against any detrimental effects, iii) the effects on cow performance of offering herbage fresh (zero grazing) vs offering the same herbage ensiled, v) the production response of high yielding dairy cows to very high quality silages across a range of concentrate levels, and optimum concentrate strategies with very high quality silage.
5. Through this large 'programme based' research project, to raise awareness of issues related to grass silage quality on NI dairy farms, cumulating in a silage open day and the publication of a silage manual.

Research Provider: AFBI

Project Lead: Conrad Ferris

Project Team: David Patterson, Debbie Hynes, Tianhai Yan, Alan Gordon

Start Date: 1st April 2018

End Date: 31st March 2022

Co-Funding: DAERA (Evidence & Innovation Strategy)

Background

Making more use of forage in livestock diets has been identified as a key object of the sustainable Land Use Management Strategy. While 'alternative' forages such as whole crop silage and maize silage are included in dairy cows diets to a limited extent in Northern Ireland (NI), grass silage remains the predominant forage within the diets of housed dairy cows. In addition, while silage feeding was previously largely confined to the 'winter period', an increasing proportion of cows in NI are now housed either all year, or for increasing parts of the year. Consequently, the reliance on grass silage is increasing. Furthermore, while historically the focus on most farms (and most research programmes) was the production of high quality first cut silage, the increasingly confined nature of our production systems means that all silage harvests (1st, 2nd, 3rd and even 4th) make important contributions to the diets of our dairy herds. While the increasing importance of silage in dairy cow diets cannot be denied, there has not been a focused silage research programme within NI for approximately 15 years. Recognising the importance of grass silage to the NI dairy sector, this proposal seeks to develop a targeted research programme to address a number of recently identified knowledge gaps relating to the production of high quality grass silage, throughout the whole growing season, and the utilisation of high quality grass silage in the diets of high yielding dairy cows. The programme is also set within the context of increasing environmental pressures with regards both water and air quality, and recognises that making more use of quality locally produced forages has the potential to reduce concentrate usage on farms, and as a consequence P imports in concentrates, thus aligning the work with the Land Use Management Strategy for Northern Ireland.

While high quality silage offers real opportunities to cut down on the use of expensive concentrate feeds, high quality silage can be difficult to make, especially in latter cuts, and can create challenges when being offered. For example, issues that need to be considered when making high quality silage include the impact of winter carry over grass on silage quality, optimum fertiliser N levels for multi cut silage systems to minimise fermentation problems, optimum cutting stage especially in relation to rate of fall in digestibility over the growing season, and the impact of wilting on feed efficiency. In addition, many farmers have concerns about the inclusion of high quality silage in dairy cow diets. For example, high quality silage offered with high levels of concentrate can result in milk fat depression and increase the risk of sub-acute acidosis. In addition, earlier research examining the response of dairy cows to silages of different qualities involved cows with maximum milk yields of approximately 32 kg/cow/day, while average winter milk yields of 35 - 40 kg are now commonplace. The impact of silage quality on the response of cows to concentrates at these higher milk yields is largely unknown, not just at a cow level, but at a system level. In particular, there is anecdotal evidence that even when high levels of concentrate intakes can be achieved with forage based diets, there is an increased risk of greater levels of body tissue loss with higher yielding cows. If this is true, then fertility may be negatively impacted, and cow welfare impinged. Furthermore, with an increasing move to zero grazing type systems, it is still unclear if similar cow performance can be achieved by offering the same quality of grass either ensiled vs fresh, thus avoiding the need for daily harvesting.

In addressing these issues, the project has potential to help increase efficiency in production systems, while greatly increasing the awareness of the potential and limitations of high quality silage in dairy cow diets. The premise of this project is that high quality silage can and should play an increasing role in dairy cows diets in Northern Ireland, and that AFBI research can help overcome some of the barriers to its adoption. This will advance the overall achievement of producing more milk from forage i.e. grass and grass silage.

 

Existing Knowledge

As already highlighted, Northern Ireland has not had a focused grass silage research programme in place for approximately 15 years. Similarly, few other UK research institutes are currently conducting research into grass silage. With the exception of the Scandinavian countries, most of which do not use perennial ryegrass, few other European countries have specific grass silage research programmes. Consequently, much of the information presented in this review is at least 15 years old.

With regard to silage composition in Northern Ireland, a brief review of the composition of silages analysed within the HFIS laboratory over the last decade was recently undertaken by Yan et al. (2017a and b). In general, DM concentrations have tended to increase, crude protein levels have fallen, and metabolisable energy (ME) concentrations have remained relatively unchanged. However, this review did not involve an in depth analysis of geographical trends in silage composition, or make any attempt to explain the trends observed.

Many previous studies have examined the impact of field wilting on dairy cow performance, including several studies conducted at AFBI (Gordon and Peoples, 1986: Yan et al., 1998). In general, the results of these studies have demonstrated that when grass is subject to rapid wilting, intake is increased and milk yields normally improve. However, the impact on energy use efficiency is less clear, with Gordon et al. (2000), in one of the few studies examining this issue using cows in respiration calorimeters, finding energy use efficiency to decrease with increasing degree of wilting. However, in a separate analysis, Wright et al. (2000) found no reduction in the efficiency of milk production with wilting.

The adverse impact of excessive nitrate levels in herbage on ensilage characteristics is well known, and farmers understand that the crop must be left to mature sufficiently for 'nitrate levels to fall'. While this is generally not an issues within a two or three cut system, this can become more challenging when crops are harvested at a less mature stage. In reality there is little information on the rate of uptake of nitrogen (N) by growing crops or the rate of dilution of N (and fall in nitrate levels) in the growing crop at different stages of the growing season. While many studies have examined the response to fertiliser N in terms of DM yield most studies simply look at the 'end point' (ie at harvest) in terms of herbage N composition, rather than providing information on interim changes taking place within the growing crop or the impact on the ensilability of the grass. However, there is evidence that N levels in herbage tend to increase (at a fixed regrowth interval) as the season progresses (Binnie et al., 1997). From an environmental perspective there is increasing awareness of ammonia emissions from agriculture. The 'Making Ammonia Visible' report (January 2018) stated that 91% of all ammonia emissions in NI come from agriculture. The report recommends the use of treated or stabilised urea fertilisers which are less vulnerable to volatilisation and de-nitrification (releasing ammonia) when spread and recommends farmers switch from CAN-type nitrogen (which is also vulnerable to de-nitrification) to stabilised urea, with the use of straight urea fertiliser to be banned from 2020.

While many studies have examined the response of dairy cows when offered grass silage based diets, research on the performance of dairy cows offered very high quality grass silage diets has received little attention:

With growing herd sizes (difficult to graze large herds in the autumn), and an increasing number of herds housed full time, autumn grazing has become less important on many farms. This is especially true following harvest of third cut silage, which often takes place on land blocks some distance from the farm, and which cannot be easily grazed by dairy cows. As a consequence, herbage which grows following third harvest may remain un-grazed over the winter and is ultimately ensiled with first cut herbage. While a number of studies in the Ireland (for example, Hennessy et al. 2006) have examined the impact of autumn/winter growth (focusing on closing date) on the yield and quality of herbage available for grazing in the spring, the impact of this herbage on silage quality and subsequent cow performance does not appear to have been examined previously. While it is recognised that strategies exist by which this herbage can be removed (for example grazing by sheep and/or young stock), on many farms these strategies are not adopted. Evidence of the impact of this winter herbage on subsequent silage quality many help change these practices.

The impact of declining silage digestibility on cow performance is well known, with a review by Gordon (1989) concluding that, on average, a 10 g/kg reduction in D value resulted in a reduction in milk yield of 0.37 kg/cow/day. In addition, a number of other studies have demonstrated the decline in digestibility with maturity, with for example, Givens et al. (1989) observing a mean decrease of 2.5 g/kg per day with each day decline in digestibility. However, this study (which involved samples from farms), like most other studies focused on first cut silage, and did not examine changes in digestibility of grass pre ear emergence. In addition, as already highlighted, as cows are increasingly housed for significant parts of the 'summer', if not year round, the quality of silage produced during all harvests is important. However, there is surprisingly little work available on changes in digestibility during subsequent harvests, or indeed on the overall impact of harvesting frequency on total yield of utilisable nutrients (not just dry matter). The latter is important as while farmers are aware that more frequent harvests may reduce total DM production over the season, the impact on the yield of utilisable nutrients is less well understood.

The popularity of zero grazing within NI has increased during recent years, and AFBI is currently conducting a research programme into the practice (E&I, 15/01/06). Indeed, early outcomes of this research programme have highlighted that confined cows offered zero grazed grass can outperform confined cows offered grass silage based diets (Debbie McConnell, unpublished). However, in these studies, the silage offered, while of good quality, was made from a sward with a very different digestibility than the zero grazed grass offered. The question has however been asked, that if similar quality of grass as used within zero grazing systems was ensiled and offered, would performance differ from that achieved with zero grazed grass offered directly. A previous AFBI study by Cushnahan and Mayne (1995) provided some insight to this, and demonstrated that fresh grass and silage from the same sward could have similar nutritive value when offered to dairy cows. However, this study involved herbage of a much lower digestibility than typically used while zero grazing.

While the benefits of high quality silage are well known, some farmers feel that very high quality silage will have such a high rate of passage through the rumen that it will be poorly utilised. To address this, straw is often included in dairy cow diets, thus improving the quantity of effective fibre in the diet, but reducing overall diet quality. This was examined in an earlier AFBI study by Ferris et al. (2000), with few beneficial effects of straw inclusion observed. However this study involved moderate yielding cows offered moderate concentrate levels. The impact of concentrate type on the utilisation of very high quality silages does not appear to have been examined.

The role and potential of very high quality silage in the diets of high yielding dairy cows is also poorly understood as most studies with these cows seek to maximise performance through concentrate supplementation. Understanding the response of dairy cows to concentrate supplementation is essential to ensure cows are correctly fed. However, silage quality is one of the key determinants of the intake and milk yield response of cows to concentrate feeding. This was examined previously by Ferris et al. (2001) with moderate yielding cows. The results of this study demonstrated that as concentrate level increased, the difference in performance between cows offered a high and moderate quality silage was substantially reduced. However, the maximum milk yields in this study, even at the highest concentrate levels, were approximately 33 kg/cow/day, considerably lower than typical average milk yields of higher yielding herds today. This is reflected in the data which is used to provide the response to concentrate supplementation within the equations behind the Hillsborough Feeding and Information System (HFIS) at present, and which suggest that irrespective of silage quality, milk yields converge at a concentrate level of approximately 17 kg/cow/day, and a milk yield of approximately 35 kg/cow/day (Figure 1). Understanding the responses of today's higher yielding cows when offered diets comprising high quality silage is essential, as is being able to demonstrate the concentrate sparing effects of high quality silage with high yielding cows. Indeed, there is a perception that irrespective how good the quality of silage is, when cows are offered predominantly forage based diets, body condition score will suffer. There is indeed some evidence to support this suggestion, with cows on high forage systems generally having a lower condition score than those on higher concentrate systems, despite similar levels of milk output (Ferris et al., 2003). However, once again these studies involved lower yielding cows, compared to those on many farms today.

The project team recognise that while a significant amount of scientific literature exists, most of this is not readily available to dairy farmers, or is certainly not available in one location. In addition, given the 15 year absence of a focused grass silage research programme in NI, there has been an 'erosion of independent expertise' in relation to the science of silage making. Given the continuing and increasing importance of grass silage to the local dairy sector, the project team believe that there is a need to collate research knowledge into a 'farmer friendly' book/technical guide for NI dairy (and other ruminant livestock) farmers. This is planned as a key output from this project, and will incorporate the findings of this project, as well as previously published research in a way which presents key facts for farmers. It is proposed that this book/technical guide will be launched at a 'silage open day' at the close of this project.

 

Additional information that the project will provide

This project recognises that a significant amount of scientific literature already exists in relation to making and feeding grass silage. However, the project team, in consultation with CAFRE and Industry representatives, have identified a number of important knowledge gaps, especially in relation to high quality silage for high yielding dairy cows, which this project will seek to address. Through a desk based exercise, the project will begin by providing an accurate assessment of changes in the quality of silage produced in NI over the last 15 - 20 years, and attempt to provide an insight to these changes. This aspect of the project will also identify the frequency of very high DM silages produced in NI. If this is significant, then consideration will be given to identifying if there is evidence that these higher DM silages have a lower utilisation efficiency that lower DM silages. This could either involve a desk based exercise, or a feeding study.

Plot and mini silo studies will seek to quantify changes in nitrate, and nitrogen concentrations in herbage throughout the growing season, and use this to improve the recommendations on N fertiliser used for silage swards so as to minimise the risk of adverse fermentation and reduce denitrification (ammonia loss). Further plot and mini silo studies will seek to quantify how silage digestibility declines throughout the growing season and quantify the total yield of digestible nutrients from multi cut silage systems.

Plot and mini silo studies will seek to quantify changes in nitrate and nitrogen concentrations in herbage throughout the growing season, and use this to improve the recommendations on N application rates for silage swards so as to minimise the risk of adverse fermentation. Additional plot and mini silo studies will seek to quantify how silage digestibility declines, especially pre-ear emergence, and throughout the growing season, rather than focusing mainly on primary growth herbage.

A series of dairy cow studies will seek to address issues which will help improve utilisation of silage, and especially very high quality silage. For example, the impact of straw inclusion and concentrate type offered as a supplement to very high quality grass silage will be examined, an important issue given that some farmers feel that high quality silage is simply utilised inefficiently by dairy cows, and as such not worth making. In addition, the impact of feeding silage 'contaminated' with herbage which grows during the autumn/winter will be examined. Results will allow advice to be given on whether it is better that this autumn/winter herbage in removed either in the autumn or the following spring, or if it can be left for inclusion within the silage with adverse effects on cow performance. This cow study will be associated with detailed plot and ensilage studies. Given the increasing interest in and uptake of zero grazing, the project will seek to clarify if similar cow performance could be obtained by ensiling the herbage on a single day, rather cutting and offering fresh herbage daily throughout the summer.

The project will also examine the response of higher yielding dairy cows to concentrate supplementation, across a range of silage qualities, thus providing information not currently available for these much higher yielding cows. This information will then be used to update the response curves within the HFIS, thus allowing a more accurate prediction of the likely intake and responses to a range of silage qualities for the NI dairy sector. This project will also seek to demonstrate the potential of very high quality silages in the diets of high yielding dairy cows, and clarify if these inevitably result in excessive body tissue loss. These latter studies will involve offering silage made within a 'systems' approach, and the performance levels will be expressed, not just on a cow basis, but per ha basis.

Finally, the project will seek to collate scientific literature which is not readily available to dairy farmers, into a 'farmer friendly' book/technical guide for NI dairy (and other ruminant livestock) farmers. This is planned as a key output from this project, and will incorporate the findings of this project, as well as previously published research in a simple, concise and easily usable reference manual.

Scientific Objectives & Tasks

Objective	Tasks required to achieve objective	Milestones (target date of achievement)
Understanding silage quality in Northern Ireland
To conduct a short survey of NI dairy farmers, to understand key factors limiting improvements in silage quality.	AFBI are currently in the process of undertaking a large scale 'grassland' survey of the NI dairy sector, with a section of this survey covering broader silage issues. However, the results of this will not be known until 2019. Consequently, Objective 11.1 involves undertaking a short survey which specifically seeks to identify the main factors limiting improvements in silage quality on NI dairy farms. The survey will be conducted on the 24th and 25 January 2018 (prior to this project commencing) at two dairy Open Days that will be held at CAFRE - Greenmount College Dairy Centre. The target is to survey approximately 100 farmers. One of the issues that will be addressed within this survey is the extent to which current contractor costing systems (ie charge per ha/acre) impact on silage quality. If this is identified as being a significant issue, then AFBI will explore with CAFRE what information/research is required to develop alternative costing systems (ie yield based systems). Should the development of an alternative costing mechanism require a research component, a CRF will be developed following discussions with DAERA leads to broaden the remit of this project.	March 2019
To identify trends in the quality of silage made in Northern Ireland over the last 15 - 20 years by interrogating the Hillsborough Feeding Information System (HFIS) data base.	Over the last two decades, several thousand grass silage samples from commercial farms in Northern Ireland have been analysed every year in the analytical laboratory of the Agri-Food and Biosciences Institute using the Hillsborough Feeding Information System (HFIS). This task will involve analysing data from this data set to examine: 1) long term trends in silage composition in NI, 2) an evaluation of the effects of ensiling year/ensiling season and other parameters (e.g., weather conditions, long term fertiliser use) on silage nutritive quality. In addition, the HFIS data base will be analysed to identify the proportion of very high DM silages that are made within NI each year. This is important as there is anecdotal evidence that these high DM silages are used inefficiently. The original proposal submitted September 17 (in section 11.2) had proposed research to further investigate this issue. However, that aspect of the research has been removed from this proposal, but may be reconsidered (subject to a CRF) once the frequency of very high DM silages is known.	January 2019
Grass growth and ensilage studies
To examine the impact of fertiliser N type at different stages of the growing season on rate of N uptake by the grass plant (weekly uptake) over a seven week growing period, herbage yield, herbage composition and ensilability. The objective is to develop improved fertiliser recommendations for the production of high quality silage.	This plot study will be conducted over two successive years. This study will be conducted on four occasions over the growing season (starting mid March, mid May, late June and mid August). On each occasion plots will be treated with either CAN or stabilised fertiliser N at appropriate rates based on RB209 Recommendations, and allowed to grow for a 5 - 7 week period. Plots will be destructively harvested at a series of time intervals, i.e. at 2, 3, 4, 5, 6 and 7 weeks post fertiliser application. As well as yield and routine herbage analysis, the proportion of applied fertiliser which has been taken up by the grass plant (less that released from the soil) will be estimated, together with the buffering and nitrate content of the herbage. *The herbage harvested at the end of each growing period will be ensiled in mini silos to examine the impact of N parameters on fermentation characteristics. Preparation of a scientific paper (2 years).	February 2019 (Year 1) April 2020 (Year 2)
To examine the impact of herbage which grows during the late autumn/winter period on the yield and quality of first cut silage	This replicated plot study will be conducted over two successive years. Following harvest of third cut silage in autumn, swards will be managed in one of three ways, as follows: 1) no further harvesting until the following spring, 2) autumn growth herbage removed after 5 weeks (likely mid October), 3) autumn growth herbage removed after 8 weeks (likely early November). Yield and quality of herbage removed will be quantified.	May 2019 (Year 1) May 2020 (Year 2)
	In early May all plots will be harvested, and yield recorded, and herbage samples dissected for live/dead proportions. Herbage from the different treatment plots will be ensiled in pipe silos for an 80 day period. In-silo losses and herbage quality will be assessed at opening.	August 2019 (Year 1) August 2020 (Year 2)
Effect of harvesting interval on the total yield of utilisable nutrients	This five treatment study will seek to examine the impact of a number of harvest intervals on the total yield of utilisable nutrients over the entire growing season. In early Spring (mid-March) a sward will be trimmed off to remove any autumn/winter growth herbage, and a series of replicated plots established. The effect of silage harvest interval (30, 36, 45, 60 and 90 days, representing 6, 5, 4, 3 and 2 harvests/year) on the total yield of utilisable nutrients will be assessed over a 180 day period (until mid September). This represents a total of 20 individual harvests over the season. Total fertiliser N inputs over the season will be the same with each of the five harvesting interval treatments, with fertiliser applications based on RB209 recommendations for the respective cutting intervals.   When each plot is harvested the growth stage of the herbage will be defined, yield will be measured and the chemical composition of the herbage determined. Herbage will then be ensiled (most likely using big bales/box silos). 'Silos' will be opened after an 80 day ensilage period, the silage thoroughly mixed, and sufficient quantities bagged up for use in sheep digestibility studies. This will also provide an opportunity to compare the NIRS composition of the herbage as ensiled (especially ME) vs that of silage removed	December 2020 (Year 1) December 2021 (Year 2)
	The digestibility of each silage will be determined using sheep confined in digestibility crates (