NCDC212: Metric Definition of the Physical Component of Dairy Diets

(Multistate Research Coordinating Committee and Information Exchange Group)

Status: Inactive/Terminating

Project Menu

SAES-422 Reports

Annual/Termination Reports:

[09/17/2007]

Date of Annual Report: 09/17/2007

Report Information

Annual Meeting Dates: 06/07/2007 - 06/08/2007
Period the Report Covers: 10/01/2006 - 09/01/2007

Participants

Mike Murphy, University of Illinois
Jud Heinrichs, The Pennsylvania Sate University
Jeff Firkins, The Ohio State University
Paul Kononoff, University of Nebraska-Lincoln
David Mertens, ARS-USDA
Ric Grummer, University of Wisconsin (administrative Advisor)
Lou Armentano, University of Wisconsin

Brief Summary of Minutes

NCDC212: Metric Definition of the Physical Component of Dairy Diets
Meeting Minutes and annual report
June 7 and 8
Candlewood Suites 4021 N Mannheim Rd.
Chicago, IL
Present:
Mike Murphy, University of Illinois
Jud Heinrichs, The Pennsylvania Sate University
Jeff Firkins, The Ohio State University
Paul Kononoff, University of Nebraska-Lincoln
David Mertens, ARS-USDA
Ric Grummer, University of Wisconsin (administrative Advisor)
Lou Armentano, University of Wisconsin

Meeting Commenced at 10 AM June 7, 2007

Introduction:
Armentano welcomed the group and introduced administrator Ric Grummer. Dr. Grummer introduced himself and stating his interest in helping the group while also briefly outlining the aim and purpose of NCDC meetings. The NCDC group may be formed for a period of two years with the ability to collectively submit a grant to the appropriate funding agency.
Kononoff was appointed secretary.

Armentano requested that each member of the group take 30-45 minutes to share their observations and opinions on recommendations related to the particle size of dairy cattle rations. The meeting proceeded with experiment station presenting in alphabetical order.

Meeting adjourned at ~ 6:00 PM with agreement to meet in the AM. Discussions continued over supper.

Friday June 6, 2007

- The group met over breakfast before departing. The objective of this meeting was to determine a plan of action.
The meeting closed with a decision to meet again within the framework of the current approved project. It was decided that near the 1str or second week on June 2008 would be a logical time meet and the same site appeared agreeable for all.

Accomplishments

note for the sake of this report we have used the term mean particle length (MPL) for results based on horizontal shaking of feeds obtained with the Wisconsin or Penn State shakers using the diagonal measurements for square openings and diameter for round openings for the former, Mean Particle Girth (MPG) for the rotap vertical or similar vertical sieving based on the side dimension of the sieves, and mean particle size (MPS) for generic reference. It is not the intent to introduce a standard use of Girth as a new term but simply for clarity in this report. I. Illinois - Murphy opened by outlining the importance of presenting particle size data in published papers, specifically reporting the proportion of material being retained on each sieve used (in addition to a mean particle size). This allows other investigators to understand dispersion in feed particle size. - Research (Murphy and Zhu, 1997) has indicated that in measuring particle size of feed, a greater amount of the observed variation is associated with the feedstuff (78%) than compared to the method of measurement (7%). - IL reviewed research that has investigated the effects of mastication on digestion. Although some research suggested mastication alone improves digestion of the feed particle (Olubobokun et al.,1990) results across studies are inconsistent. - Other research (Playne et al. (1978) has indicated that mastication promotes digestion. These observations support Bailey (1962) who suggested that fractional degradation rate of swallowed grass is 42% greater than unswallowed grass. - IL outlined a detailed reanalysis of the data from Kennedy (1985) and Suzuki et al. (2001), this research has been accepted for publication and was designed to evaluate the dynamics of particle size distributions through rumination. o Generally results of Suzuki suggests that mean particle size (MPS) of the tail bolus and down bolus is lower the retained bolus and the retained bolus is greater than the MPS of the up bolus. o Data that may be transposable to mixed diets are limiting. - IL reviewed research that has investigated a possible synergy between rumination and Digestion. NDF disappearance is greater for the down bolus compared to the retained bolus, suggesting rumination positively increases fiber digestion. - Factors affecting passage of particles were also reviewed, namely; o Mastication o Rumination o Fermentation o Intake o Specific gravity o Dilution rate - A model of the dynamics of particle size distribution was presented o 20 papers used o Bolus particle size could be estimated using data from 5 studies. o Particle size of rumen digesta was estimated - Presentation concluded noting that understanding factors involved in combination and the kinetics of their interactions may allow digestion and passage to be manipulated to maximized nutrient supply - Further work should be designed to: o Contribute information of mixed diets o Contribute information of other types of forages such as silages. II. University of Nebraska-Lincoln - Particle size and effective fiber recommendations of the NRC (2001) committee was discussed o Lack of standard, validated methods to measure effective fiber of feeds or to establish requirements for effective fiber limits application of this concept (page 39, NRC, 2001). o 25% NDF with 19% NDF from forage (NRC, 1989) - The results of the Symposium at the 1995 ADSA national meeting were discussed: Meeting the Fiber Requirements of Dairy Cows. - Given this information was available to the NREC (2001) it was surprising more of it was not included in the final NRC (2001) publication - The Penn State Particle Separator (Kononoff and Heinrichs, 2002) was discussed. - Results of foundation experiments were presented. These results were used to formulate PA recommendations for haylages, corn silage and TMRs. - UNL work evaluating the use of co-product feeds was also presented - Data from experiments feeding corn gluten feed, corn bran and DDGS suggest that these feeds could be included in high amounts of diet fed to a dairy cow however some limitations may exist on their ability to stimulate normal rumen function. - Recommendations: o The PSPS should continue to prove to be a useful device in future recommendations. o Current PS recommendations have been useful o An index for effective fiber may not be needed, to simply assess effective fiber as estimates of NDF and particle size independently is useful and highly informative o An index for effective fiber may be useful when integration into models that seek to predict rumen fermentation and rumen pH but accurate models should also seek to account for diet fermentability o Forage is often an expensive ingredient required to meet the physically effective fiber needs of cows. Given the growth of the US ethanol industry starch will become a more expensive ingredient, while igh fiber byproducts may become cheaper relative to both corn (as an anergy source) and forage (as an energy and fiber source). - The group discussed of the usefulness of the third screen (1.18-mm). Murphy suggested that three sieves are the minimum needed to estimate the dispersion of particle size distribution. III. The Ohio State University - Suggested that physical effective NDF is well estimated by measuring chewing behavior. - Data from Erdman (1988) was presented demonstrating that in diet containing 70 to 30% DM as forage only small differences are observed in saliva secretion. - Rumen microbes are likely quite sensitive to actually pH differences that are reflective of very large differences in [H+]. - The point that feedstuff particle size affects intake was made and recommended that any system design to outline particle size recommendations also encompass and account for the associative effects of increasing intakes. In addition, in trials designed to measure impact of particle size, intake may differ among treatments due to both random effects and due to treatment effect. Impact of different DM intake as well as direct effect of particle size on response functions needs to be delineated (see below). - In conducting experiments designed to evaluate diet effects on lactational performance it was suggested that long term experiments such as collecting data from a randomized complete block may be more reflective of field changes when compared to experiments based on much shorter cross over designs (i.e. Latin square). The use of longer treatment periods was encouraged. However, there is also concern about the precision of experiments that characterize feed properties based on animal response (see below) - Mean particle size is a single value that affects a multitude of physiological processes and that the mean is a representation of a distribution (i.e. a greater mean can be used to infer more large particles and fewer smaller particles). - Data from an experiment evaluating the effects of wet by-products on sorting activity was presented. o After 8 hrs sorting was observed given that the proportion of particles > 19 mm increased from 9.8 to 14.8 % while the proportion of particles > 8 mm decreased from 46.3 to 31.7%. o In the same experiment the addition of molasses did not result in any differences in particles > 19.0mm but the proportion of particles > 8mm increased from 57.3 to 45.6 %. o In the same experiment the addition of monensin increased the amount of particles > 19 mm and decreased the proportion of material < 8-mm in diets containing alfalfa hay. - The ability to measure and estimate effective fiber on farms was also discussed o Suggested that there is a gap between what is applied on farms and what is supported by research. o The PSPS is useful in making some recommendations as well as evaluating feeding behavior and sorting. - Information of the importance of the rumen mat was also presented. The mat is likely to retained particles that have the potential to escape (< 1.2 mm) to improve digestibility of slower digester material such as fiber. - The entrapment of byproduct feeds may also be entrapped and stimulate rumination but may also limit intake - These effects have been demonstrated through a series of experiments conducted at OSU using whole cottonseed (WCS). o Firstly the replacement of alfalfa haylage with cottonseed decreased both the fractional passage and digestion rates of potentially degradable NDF when included at 5-15% of the diet DM. Therefore, increasing substitution of alfalfa NDF with cottonseed NDF maintained the rumen pool of NDF, which maintained rumen mat consistency and stimulation of rumination. o When included in a low forage diet (16% fNDF) cottonseed did not affect rumen and NDF fill but reduced ki. Increasing the proportion of forage also increased ki. (note: ki is the sum of kd plus kp or the intake in kg/d divided by the rumen pool size in kg) o When included in a low forage diet (16% fNDF) cottonseed did not affect kp but did reduce kd and ki. Increasing the proportion of forage did not affect kd. o Total rumen NDFD was negatively affected by feeding "easyflow" cottonseed and was also decreased by feeding more forage. o When expressed as min/kg NDF intake easy flow cottonseed and increasing the proportion of forage, increased total time chewing - The importance of processing changes affects rumen dynamics through changes in rumen fill and entanglement in the rumen mat was also discussed. o In the case of fat containing cottonseed rolling the particle may result in an increase in rate of release and pelleting may result in rapid release this result in a reduction of effectiveness. IV. The Pennsylvania State University - PA is currently about to commence a series of studies that evaluate effects of particle size on feeding behavior. - These studies will occur both in a controlled research setting when animals are fed individually as well as a field study. - The aim of the field study is to identify farms experiencing challenged in herd health (i.e. DAs or low milk fat) and to evaluate potential factors that may negatively affect feeding behavior. - In designing the field study the investigators are still determining what response variable should be measured. The group was asked if particle size of feed remaining in the bunk should be measured over a 24 h period or just at the beginning and end of feeding. - D. Mertens suggested that is was important to attempt to measure sorting as a total over the course of an entire day because it is his labs observation that daily one day is more important than "hourly" observations, especially in cases where bunk space is a limiting factor to meal size and time. - It was also suggested that in addition to particle size the determining of NDF of this material may indicate sorting behavior. - One limitation of the field study was discussed, knowing the total mass of feed remaining in the bunk over time. It was agreed that such a limitation is inherent in such a feeding trial. Composition of orts is a reflection of both the amount of sorting that has occurred as well as the orts remaining as a fraction of the original feed offered. Both a reduction in sorting with constant fraction of orts or more orts as a fraction of feed offered with the same total sorting can cause orts to look more like the diet offered. V. USDFRC - A considerable time was used for discussion of how to measure effective fiber. - Dave Mertens stated that in his 1997 symposiums publication effective NDF was suggested to be measured using a vertical horizontal shaking sieving method. In using this method some material > than the sieve size would fall through because during the sieving process the particle may bounce through. - The justification for this estimate was an MS thesis from Georgia (Cardoza) in which 1.18 appeared to be the inflection point of particles not appearing in the feces. - Again a considerable amount of discussion of how peNDF was estimated in the 1997 publication. Dave Mertens offered to share his procedure but briefly outlined that vertical shaking was used and the sample was dried. - Mertens shared data in which several sieving methods were compared with this aim of correcting the estimation of particle > 1.18-mm as estimate by the USDA-ARS method. - Preliminary evidence exists suggests that corn silage material > 3.9 mm for the ASAE method and 4.5 mm for the Penn state would be equivalent to the proportion of particles > 1.18-mm by rotap vertical sieving. - Preliminary evidence suggests that other forage material > 2.6 mm for the ASAE method and 4.4 mm Penn State would be equivalent to the proportion of particles > 1.18-mm. - Preliminary evidence exists suggests that for TMR material > 2.5 mm for the ASAE method and 3.2 mm Penn State would be equivalent to the proportion of particles > 1.18-mm. - all of the values above are based on side dimensions or diameters - Each of the above indicates that compared to both methods used in the field, the USDA-ARS methods sieve material to a greater degree and that potentially some particle longer than the sieve size may escape the sieve. - Additional discussion occurred on how the ASAE standards integrate particle sieving estimates to determine mean particle length. For square holed sieve, these equations seem to suggest that particles retained on the sieve are no longer than the diagonal of the individual pore. Although this may be correct Mertens suggested that based on probability this would be the largest particle retained on the sieve and not indicative of the average size. VI. University of Wisconsin - The hypothesis that the degree of effectiveness likely depends upon the length of the particle. In attempt to get a different particle distribution with the same mean an experiment was conducted at UW (Leonardi 2005). - Results from this experiment suggested that mean particle length is a good predictor of ruminating activity and independent of the particle size distribution. However, for eating chews longer particles may be more effective than indicated by the mean. - Results also suggested that particles longer than 26.9 mm may be selected against by most cows and extensively so by some cows. - Data from a published technical note was also presented (Armentano and Taysom (2005) in which the ability of a single screen to estimate mean particle length was explored. A considerable amount of discussion of how these predictions could be tested. The relationship between Mean Particle length and fraction of feed aboe one of the intermediate screens is similar information as long as the distribution shape and standar deviation are similar. This appears to be true for silages and TMR in the data set used, presumably because the harvesting machines use similar chopping methods. For feeds with narrower distributions of particle size (more homogenous) the use of a single screen could give very less useful results and would tend to either assign a pef value of 0 or 1 to the feed when an intermediate value might be appropriate. - The distinction between measuring the NDF above a certain screen as a non-animal measure of peNDF (method 1) vs. determining a pef coefficient (from mass distribution from either a single screen or mps) multiplied by total NDF (Method 2) was discussed. These approaches differ conceptually in that the first approach ignores the NDF that is below a certain screen. There is consensus that fine NDF is an important component of any system to provide useful recommendations, therefore if method 1 is used a separate way to credit the finer fiber must be added. An approach where alternative sets of requirements for peNDF (by method 1) and total NDF could be combined would be similar to the existing approach in the NRC. In addition to NDF, NDF digestibility and starch availability and perhaps other measures of dietary carbohydrate digestion impact the feasibility of different diets. The advantage of method 2 is that it combined particle size and total NDF and gives a more holistic approach in a single number and this is the approach use in CPM. It is noted that this combines some physical and non-physical factors in one measurement. A distinct advantage of method 2 (or any other technique) with a pef based on mass distribution is the ease of converting values obtained with different sieve sizes and techniques, negating the need to standardize on any one sieve size. Conversely if NDF above a screen is used, both the screen and method must be standardized before hand or NDF must be determined on multiple fractions or with multiple sieving techniques. For research purposes it would be useful to restrict the candidates to a limited number of screen sizes and sieving techniques. - In discussion the point was made that to obtain good separation multiple screens must be used, so the data for MPS and alternative screen 'cutoffs' is available. It is only if NDF (or other) analysis is needed that it becomes attractive to pool screens prior to analysis. It was also noted that if screens below a critical point are eliminated, this will not affect sieving of the coarser material and a combined single fraction can be obtained on the pan for use in analysis to provide a division of NDF into two fractions (pan, and total diet - pan). - Further discussion ensured on how peNDF should be measured in animal trials; the group seems to agree that chewing activity is a valuable index. Whether it should be expressed at sum of eating and ruminating or ruminating alone was a point of disagreement. Also it is common for chewing data to be expressed as total time, time per unit NDF consumed and time per unit DM consumed. - General: Before concluding the meeting the group appeared to agree that the current effective fiber recommendations by the NRC (2001) publication can be improved given the body of research in this area and the research and field tools available to describe diets and feeds. Meeting adjourned at ~ 6:00 PM with agreement to meet in the AM. Discussions continued over supper. Friday June 6, 2007 - The group met over breakfast before departing. The objective of this meeting was to determine a plan of action. Firstly the group felt that given the amount of confusion over the topic that there was a need to clearly define what effective fiber is. Some discussion followed on whether the group should be responsible for outlining a recommendation or simply outline a broad range. - Heinrichs suggested that the group co-author a technical note that first clearly outlines recommended methods to measure particle size and that given the body of research using the technique that the rotap sieve system be used in this procedure. It was suggested the publication also recommend a method to estimate mean particle size and should specify a maximum amount (i.e. 25%) to be retained on any one screen. The sieve profile should also include a sieve measuring 1.18-mm the technique to outline methods to present data on a DM basis. Writing such a note would force consensus building in the group ad would reduce confusion in the field. - The group supported this suggestion and Dave Mertens volunteered to take the lead in circulating a first draft of the procedure. While there is disagreement over the best ultimate procedure there is agreement over the need for labs to compare data and therefore one consistent technique that can be used across labs in exploring alternative techniques is needed. -The meeting closed with a decision to meet again within the framework of the current approved project. It was decided that near the 1str or second week on June 2008 would be a logical time meet and the same site appeared agreeable for all. Other points: Particle size or length is not the only 'physical' attribute of feeds that is important in rumen function and animal performance. Jeff Firkins pointed out that specific gravity is an important physical factor. Rick Grant submitted comments relating to 'friability' of feeds as determined by ball milling which may also be an important additional physical factor. There was no disagreement on these points but discussion of the ease of routinely acquiring this data. Both points lie well within the scope of this committee and deserve consideration. Summary: Points of consensus: There is considerable confusion in the industry due to different measuring systems, terminology, and procedures for making recommendations. The existing NRC system does not fully utilize the data currently available. There is value in fine fiber, such as that from byproduct feeds. These feeds are often cost effective relative to both forages and standard concentrates and are likely to become even more available as a result of expanded biofuels production. Any system of formulation should ideally do the following: Account for value of fine fiber Account for variation in particle size of forages and other feeds with significant size Integrate with other aspects of carbohydrates (especially rate of digestion of different carbohydrate fractions) Consider the effect of adding small amounts of chopped straw or other fairly feeds of significant particle length but narrow particle size distribution Sieve results expressed as side of a square opening vs. the diagonal can be converted by a factor of 1.4 Physical screening results among methods are highly correlated and can be converted with some reliability. Differences in as fed, DM and NDF distribution among screens and methods complicate this. Points where consensus is lacking: The physical effectiveness value of finer feeds. Whether a screen based index (with screen specific coefficients), mean particle girth or length, or a simple bivariate split is the most reasonable approach to take. If a bivariate split, which screen size? Should data be expressed base on diagonal or side The nature of the measurements and systems to be used. How much to restrict the system to allow on farm evaluation vs. off site analysis and the related questions of parallel systems where more comprehensive research data and less comprehensive field data can be related. Points for discussion at next meeting: How to account for differences in DM intake that occurs in a trial, options are: Fixing DM intake Expressing results, such as chewing time per kg DM intake Expressing results, such as chewing time per kg NDF intake Reporting results for chewing etc. as ls means adjusted for DMI as a covariate In case of the latter, should unadjusted means also be reported Resolving this issue does not resolve the issue of how feeds that change DMI should be handled in a feeding system that uses physical parameters. Proposal to use diagonal or diameter openings for horizontal shaking results reported as mean particle length, and side openings of square sieves with vertical shaking as particle size. Note the PSPS is a hybrid system as the fine screen is a sieve but the sieving action is horizontal and the support spreadsheets and graphs for this device are consistent with this approach. Variability of bioassay results is large - do these coefficients have any real value, how do we weigh, report this variability. How do we best deal with the need for "precision' favored by short term studies, vs. accuracy favored by long term studies. Also health effects are likely obvious in longer term studies.

Publications

C. Leonardi and L. E. Armentano Short Communication: Feed Selection by Dairy Cows Fed Individually in a Tie-Stall or as a Group in a Free-Stall Barn J Dairy Sci 2007 90: 2386-2389. T. F. Gressley and L. E. Armentano Effects of Low Rumen-Degradable Protein or Abomasal Fructan Infusion on Diet Digestibility and Urinary Nitrogen Excretion in Lactating Dairy Cows J Dairy Sci 2007 90: 1340-1353. P. J. Kononoff, S. K. Ivan, W. Matzke, R. J. Grant, R. A. Stock, and T. J. Klopfenstein Milk Production of Dairy Cows Fed Wet Corn Gluten Feed During the Dry Period and Lactation J Dairy Sci 2006 89: 2608-2617. B. N. Janicek, P. J. Kononoff, A. M. Gehman, K. Karges, and M. L. Gibson Short Communication: Effect of Increasing Levels of Corn Bran on Milk Yield and Composition J Dairy Sci 2007 90: 4313-4316. P. J. Kononoff, S. K. Ivan, and T. J. Klopfenstein Estimation of the Proportion of Feed Protein Digested in the Small Intestine of Cattle Consuming Wet Corn Gluten Feed J Dairy Sci 2007 90: 2377-2385. M. A. Shah and M. R. Murphy Development and Evaluation of Models to Predict the Feed Intake of Dairy Cows in Early Lactation J Dairy Sci 2006 89: 294-306. J. L. Firkins, A. N. Hristov, M. B. Hall, G. A. Varga, and N. R. St-Pierre Integration of Ruminal Metabolism in Dairy Cattle J Dairy Sci 2006 89: E31-51E.

Impact Statements

The physical attributes of diets for lactating cows are widely accepted as important and this importance is clearly demonstrated in the literature and acknowledged in the NRC feeding guidelines. However, consensus on how to best measure dietary physical characteristics and integrate this measurement with chemical information on the diet is not standardized. The goal of this committee is to provide a forum to build the best consensus possible given the data available and to suggest new approaches needed to choose among options where consensus remains elusive.