I reviewed the August Miner Institute Farm Report today
http://whminer.org/dairy/farm-report.php
Rick Grant reports on a study where they have been looking at levels of peNDF and NDFu240 in diets. One place Miner Institute and I have disagreed is on using total diet NDFu240 versus my focus on forage NDFu30. They are developing a term called “physically effective uNDF240” (pef x uNDF240) which I suspect is very close to forage NDFu30.
There is also an interesting study into sending forage samples for molds and yeast. They found you should not freeze samples and there were less of a change in plastic bags versus paper bags. The most important point is that shipping conditions after sampling may significantly change the final reported results
Does anybody know where the calculation for peNDF in the Cornell model comes from?
I do not understand why the feed library assigns a higher peNDF value to corn silage than haylage.
With peNDF having a big impact on the model’s prediction for BF and rumen ph, the model actually gives a better BF prediction in high corn silage diets than high haylage diets.
NDS has most of the Haylages at 26.6% peNDF (%DM) and 75% peNDF (%NDF)
on Corn Silages i am showing 29.4% peNDF (%DM) and 80-85% peNDF (%NDF)
the library also has a higher value for wet gluten feed.
Are there any guidelines on what these numbers should be based on DM, NDF levels, length of cut (conventional vs shredlage) etc.
Marcelo:
I believe the original peNDF values in the CNCPS library were taken from Mertens, 1997:
Mertens, D. R. 1997. Creating a system for meeting the fiber requirements of dairy cows. J. Dairy Sci. 80:1463–1481.
https:// doi .org/ 10 .3168/ jds .S00220302(97)76075 -2.
He had a couple of tables in this paper – one based on studies measuring chewing time of various feeds: I tried to cut and paste the table but it didn’t work – but see Table 4 in the paper – Corn silage had a Standardized pef of 0.80 (fine chopped), 0.85 (medium chopped) and 0.90 (coarsely chopped) and alfalfa silage had a standardized pef of 0.70 (fine chopped), 0.80 (medium chopped), and 0.85 coarsely chopped. . For a lab measure it was suggested to use the proportion of material retained on a 1.18 mm screen with vertical shaking as the pef factor. Table 7 list values for various feeds with this method: corn silage – 0.81 and fine chop alfalfa silage – 0.67 and coarse chop alfalfa silage – 0.82. He also lists some pef values for some grains and byproducts. He suggested a dietary peNDf of 22% to maintain mean rumen pH at 6.0. ruminal pH = 6.67 – 0.143(1/peNDF) More recently others have suggested using the Penn State Particle Separator to estimate pef of TMR.
Jim;
Thank you for the quick response.
I was aware of this paper, however, I thought that Dr. Van Amburgh had updated this information in recent years.
I am doing as you say using the 3 sieve PS Shaker box (4mm bottom sieve) to assess peNDF.
NDS has a tab where we can enter this information and compare the predicted peNDF vs the observed peNDF and in every case the estimated is higher than the observed on farm.
This summer, QLF did tmr audits in several of my customer’s dairies and i entered their values in my NDS and came out with the same conclusion.
NDS also allows us to modify the peNDF of the ingredients to match the predicted with the observed and my instinct tells me i need to reduce the values in the corn silage, but i don’t know if what i am doing is correct (most times lowering the values in corn silage and wet gluten feed)
Based on Mertens work, I am probably wrong.
But after I adjust the observed peNDF the predictions of the program seem more accurate based on the cows performance.
I spoke to Larry yesterday and perhaps we can get an update from Dr. Van Amburgh on what are the latest thoughts on this subject.
How would you suggest the approach of adjusting the observed vs predicted peNDF when the observed is lower than the predicted based on Mertens information?
Thank you for the input Jim.
Marcelo
Marcelo, good discussion. I too have tinkered with the P-Size tab in NDS trying to match the estimated and observed pef by arbitrarily lowering the peNDF(%NDF) of the forages in the upper right hand corner of the P-size tab screen. Two things I’ve noticed. 1. – After doing this the TMR peNDF will be 17-18% and fits nicely with the model’s prediction of rumen pH, time below 5.5 pH and acidosis risk. If I use Mertens books values for peNDF the diet peNDF is inflated high and gives a false sense of security related to risk of acidosis. 2. – When you change the peNDF in the P-Size tab it does not change the peNDF of these feeds in your feed library. So when I have the same feed in multiple rations on the same farm I must remember to deflate the peNDF on the constants tab of the feed ingredient analysis. Hope this makes sense.
I added Dave’s paper to the Consultant File List.
Mertens D R 1997 Creating a system for meeting the fiber requirements of dairy cows J Dairy Sci 80 1463 1481
In CWT we need to be careful about copyright material. JDS has a policy of not sharing papers to non-members. This particular paper is old enough that it has been highly cited and circulated so I decided it was worthy of uploading.
I think pef estimated on TMR using the as is weight proportions of material on the 3 screen (19, 8, and 4 mm) will, more than likely, be lower lower than using ‘book’ pef values (mostly derived from the aformentioned Merten’s paper). Therefore our targets for adequacy are going to be lower than the 21-22% peNDF suggested. An illustration using Miner guidelines for PSPS: 5% >19mm; 50% > 8mm, and midrange 15% > 4 mm = pef of 70%. This is probably on the high end of high ration TMRs. Multiply by probably high end total NDF in high diet (30%) = 30% x 0.70 = 21%. This is almost best case scenario for high rations. More likely values are going to fall in the 18-19% range or even lower as Dr. Fry has observed. If I remember Karen Beauchemin had a paper where she measured rumen pH (and maybe chewing) and related it to the DM proportions on top two sieves of the PSPS and came up with minimums to maintain a healthy rumen that were much lower than 21%. (BTW she has an excellent review in a recent: Beauchemin, K.A. 2018. Invited Review: Current perspectives on eating and rumination activity in dairy cows. J. Dairy Sci. 101:4762) Remember the initial guidelines may have been mostly based on aNDF rather than aNDFom (I think), where total ration NDF’s were 1-2 or more units higher. I’m not sure if the equations using peNDF (along with other factors) used in NDS to predict fiber adequacy or milk fat were developed using the Merten’s method or are adjusted now using PSPS values. Keep in mind that when using PSPS we are measuring ‘as is’ weights of material retained on the various screens. The assumption is that dry matter and NDF on a dry matter basis are equally distributed across the sieves. Also weights are proportional which can skew things if we add water or have small particle wet material (more weight on the bottom pan dilutes the percentages on the sieves). Or pellets (or corn kernels) that are retained on the 8 or 4 mm screen that aren’t effective fiber. I think most of us recognize these limitations and can make make some adjustments (mentally if not actually) when accessing PSPS values. The other assumption is that there is minimal sorting. One other comment is that targets for peNDF are not fixed but dependent on other dietary factors e.g. fermentable carbohydrate.
One other thing that we should think about is that when Dave was doing his work the NDF might not have been ash corrected. I raised a question a couple of times if the model is based on ash corrected NDF or not. Does anyone know how this issue might affect the model?
Jim, thank you for your explanation.
Good point Larry, not sure if Merten’s ndf could have been ash corrected, i doubt it.
I agree with Dr Fry and Dr Aldrich on that most of my diets have peNDF’s around 17-18 and the cows are perfectly healthy.