[Music] good morning everyone I want to thank at a sale for hosting this meeting this is a great meeting it's always great to be in front of people in the feed industry as well as interacting with suppliers like this this is probably one of the best companies that we get to work with around the world before I get started I want to ask you a couple questions one is how many of you are using a platform based on the Cornell model ok INRA feed into milk the Dutch system the German system we had a smartass up front there long time customers so I know what he's actually doing and he'll say whatever he wants before I actually get into him I could talk about everything that's in the Cornell model how we implement things but there's I'm taking a slightly different track with this well we'll talk some about the model but we're also going to talk about how we actually can improve our implementation of any system by looking at some specific inputs and such but before I before I even go to the next slide I realized last late last night that I forgot a slide and it is probably to me the most important slide that I should have in every presentation so I'm just gonna start drawing it quick here and they're probably gonna yell at me that I'm getting too far away from the microphone but they will live several years ago I did some calculations just to look at how much energy does do these cows metabolize every day versus humans and and if we take a human male at a hundred kilos of body weight at maintenance so not really exercising a lot just kind of a kind of like me kind of a slug okay and when I actually did this like did these calculations I weighed a hundred kilos that was before I quit smoking and spent more time in the office so I would have to change it to higher than that but basically a human male at maintenance is going to metabolize about 25 kcals of energy per kilo of bodyweight okay that's our base if we take they've actually put data loggers on football player and measured their energy metabolism during a football match and if we look at the energy that they're metabolizing during a batch it's about sixty kcals so here we've got these guys out there running like mad for three hours this kind of energy metabolism then they take a couple days off and everyone considers them a hero and they get paid ungodly sums of money but if we look at a cow that's producing thirty seven to forty kilos of milk she's metabolizing 100 K Cal's per kilo of body weight if we put them under different stresses higher temperatures more walking she can be upwards of 110 K Cal's per kilo of body weight folks these animals are amazing athletes we basically expect that cow to run a marathon and play a full football game every day so we really when we think about the efficiency of these cows we really need to keep in mind that we need to do everything we can to make her as comfortable as possible keep her cool and feed her appropriately for a long time we've always looked at amino acids as additives and I think it's it should be pretty clear after the talks yesterday especially Chuck's and Phil's about how amino acids are their required nutrients but for us to really take advantage of how we formulate we have to spend a little bit of extra time on getting some data inputs correct so before we get into that though I always start slides like this because when I tell people I'm from New York the first thing that everyone says is I love that city okay I'm 51 years old I've in upstate New York my entire life with the exception of going through the airports and driving through the city to get to a wedding on Long Island I've never been to the city it is as different where I live as being in Paris versus where we were last night okay so if we look at where I live I live in the red dot and if we zoom in to that red dot you can see we started to have all sorts of lakes we have for true seasons New York states the four third or fourth largest dairy producing state in the country we have about 650 thousand cows and we have lots and lots of water so we're dealing with some pretty especially larger farms we have some pretty strict environmental regulations we zoom in some more there's the dairy I'm a partner and we're dying out of 800 cows and we actually live in a vet the farms in a valley so this this ground down here it's actually on an aquifer and we have we can get 200 millimetres of rainfall and be out doing primary tillage work the next day but that water supply all the water that aquifer also supplies to water for about 60,000 people and our last field is about 50 meters from the will't village wellhead so we're watched pretty closely pretty simple operation of this buttons actually been extended so this is lactating cows oops dry cows and heifers our feed storage system nor storage because of environmental regs there's another big pond here for runoff from the bunks I love complex the original facility was actually built in 1995 I started working with the farm when I was an extension in 1997 and I did most of my PhD actually on this farm implementing commercial quality control system so all based off of Six Sigma on this farm it was it was quite interesting just a quick shot of our weaned calves and then I started generating this graph we've recorded kilos of milk shipped per cow all the way back to 97 I took this graph I started this graph January 1st of 2010 Asian within year 2 X - 3 X and then these drops everyone says it's heat stress but it's not it's when we would start feeding fresh corn silage so watching graphs like this wide using historical data and then using the model we started to see that we were losing a lot of opportunity were losing a lot of milk so if you always diffused the model as part of decision making a decision-making tool to do what ifs if we change this if we change when we start feeding corn silage what what will it do for us in terms of milk production over the next year so that that allowed us to change our feeding started storage strategy some things related that overcrowding poor forage quality you can see all the good the bad and the ugly that goes on on the dairy all of this has been based off the model a lot of the most of these decisions and and I just want to be clear that this is not a new model this model actually started to be developed in the mid to late 1980s and the first real release that people could get their hands on was in 1991 and in 1992 they started to develop CPM dairy to be the first real commercial application to do formulation based on this system I've been working with it since 1990 when I started my masters and that was the same year I got married so we're 29 years later and I'm still involved with both model and my wife and I tell people you know there's days that I wonder if I'm married to my wife and the model is my mistress or if I'm married to my wife and the models my mistress and and the true answer to that question is it depends on the day back in 2005 when everyone was starting to retire the original fathers of the model we were given the opportunity to become a spinoff out of spin out out of Cornell three of us left and started MTX we're still tightly in relationship with with Cornell with especially Mike Van Amburg who's in charge of developing the model at Cornell so as there's new biology new things that are done we communicate back and forth with Mike quite a bit on how we're gonna actually implement this in the field so how do we implement this in the field well one of the first things is communication is critical when we're implementing this on-farm it and it's not just from the model from doing day-to-day formulations but it's also doing these what-if scenarios we can do some really wonderful things for example comparing ground corn versus expanded corn what impact is that change in processing that change and starch availability going to do to our overall formulation strategy and this is both from the standpoint of how we function within amts but also how the farms really have to function to take full advantage of what we do with formulation and that's both basic formulation to get proteins and and and carbohydrates correct but also when we start looking at going to the net going to the next level and implementing amino acids implementing different fatty acid strategies whatever is that these farms have to have this commitment to continuously improve because when we look at when we make a change on farm there are some changes we'll see immediately or within two weeks but to really with most of these nutritional changes these higher-end changes it may take a year for us to see the full benefit of that change that we made I jokingly asked audiences how long does it take for you to determine if this nutritionist is good I've had dairy farmers say haha two weeks I'm like yeah no let's let's be real most people will say six months maybe 12 months but in reality if we're going to touch every system in the dairy it's going to take us three years to really see the true impact of what that nutritionist is doing with that farm that because that's how long it'll take us to start from the baby calf all the way through to those heifers are through their first lactation so they have to there has to be this commitment for consistency and continuous improvement so let's start with this and fill when filled through this slide up yesterday I was about ready to jump up and down and say you sob you stole my slide because we get into some what is a mathematical model and all we're trying to do is take this data that has been studied and then try to do these mathematical relationships so this is this is just one example this is one equation out of the model and that is how do we get to predicting total nitrogen excretion and in this case it's a really good relationship with crude protein where protein intake increases the total amount of nitrogen excretion increases as well so in this case that it's a very tight relationship and this is an acceptable equation but when we throw this one up and we look at the relationship between milk production and crude protein of diets okay yeah we can fit a line to it but when you look at within the 16% diet range they go from 20 kilos all the way to 37 kilos okay so a relationship like this is shows us when we see data like this this will not work this is an unworkable relationship and we actually have to dig deeper into how are we going to model this relationship between protein and milk production and just to back up what happens to all of that nitrogen this is what I consider this a classic study I love this data and it's a real simple design this was I had out of Glenn Broderick's lab at the dairy forage and Research Center at Madison where all they did was increase the amount of diet crude protein so you can see in the red yes as they increased protein the increased nitrogen intake didn't really see a change in milk nitrogen output and it was all related to menorah nitrogen and when we look at that even deeper fecal nitrogen really didn't change but it's all urinary nitrogen now let's talk about that urinary nitrogen in terms of especially here in the EU with the nitrogen directive and that's all that's all related to emissions both air quality and water quality so that this is where we really have that potential to by formulating better and getting most of these diets in this 14 to 16 percent crude protein range if we want to go there that we can really minimize our nitrogen excretion and maximize productivity of these cows so the Cornell model if we look at what a cow is she is this she's this amazing beast besides being this uber athlete she also has this huge fermentation vessel that for years has always been treated kind of as a black box and that's really where the Cornell model comes into is our primary focus is how do we model what's going on in that room and to predict microbial yield because these cows as we've learned over many years we are limited in our productivity on these cows based on how much form edible carbohydrates can we get down her throat and keep her healthy so that gets into all sorts of things fiber quality fiber digestibility starch digestibility because it's all related to how much bacterial fermentation are we gonna get and can we really get what is the maximum amount of amino acid flow to the cow that we can get out of the rumen and in general and when we're talking herds regardless of level of production we're gonna be between 40 and maybe 70% of the total amino acid flow it's going to be coming from microbial yield so that's why there's been such this huge focus by the Cornell group over the years to try and get a handle on that predictability she also has multiple psychological needs and that is that all these different systems are competing for this supply of nutrients at the same time so really when we look at the Cornell model and most of these systems Indra is the same we're basically a giant accounting system we're trying to account for all these different inputs all these different variables and come up with a way to give us a prediction an accurate prediction that we can go out and say this diet this cow is going to give us this level of milk production and be confident in that that is the result we're going to see so when we look at where we start getting into some of the critical control points of using these models for us with the Cornell model there's two primary areas when we talk about the requirement side how we define the animal is critical we're going to go through some examples of that and from the supply side it's getting that feed chemistry right these feed characteristics because really what we're doing with the model is this we describe feeds by their fermentation kinetics okay so we split things into pools I don't care what formulation system you use the single most important input oh wait how many of you are veterinarians okay veterinarians my wife's of that so it's open season how do you determine how much of an antibiotic to administer to a cow do you know body weight yeah bodyweight and all of these formulation systems is the most important input within the Cornell model it touches every section except determining the protein and energy requirements for lactation okay and yet we don't measure it on a daily basis or even on a weekly basis now another unknown little fact is over time because of genetics we have selected for bigger cows okay a couple years ago the Cornell group went and they compared what the Cornell herd body weight was then versus in 1993 same genetic selection program same base forages similar formulation strategy so this is purely what has happened to this herd in terms of genetics 1993 the mature weight of the herd was 668 kilos in 2016 776 folks that's about a 1% per year increase in mature size this is also happening with jerseys hell we've got jerseys in the US now with mature sizes of six hundred six hundred and twenty kilos for me to walk onto a Jersey fireman see a five hundred kilo average jersey size is pretty rare okay we have selected for bigger cows because they give more milk because they could eat more here's a cow this was actually a cow on one of one of the studies she tipped the scale at a thousand kilos and if you look at it it's kind of hard on screen there she's maybe a body condition score three she is just this big deep body beast who can eat and eat and eat and make lots of milk we seen this so even at home we just went through on eight hundred cows we have 77 cows right now that are that lifetime production is greater than than 50 thousand kilos okay these cows are just bigger and if they survive they can produce boatloads of milk we also know and this is great I love this next slide because it shows some of the relationships we forget about heifers there's this awesome relationship between how much milk a first calf heifer is going to give and what as a percentage of the mature herd it's the same relationship as a percentage of mature body size this is actually from a study at Cornell and this is week of trial it's not even week of lactation so I know on the right we've got milk those heifers were averaging 78 percent of the mature herd and they weighed 79 percent of the mature herd one of the first things that I look at when I go on farm is this relationship this tells me everything about how the heifer program is working because if they don't produce and I'm thinking yesterday to some of the report the results from some of these trials where they weren't seeing a milk response in relation to in any treatment in the first calf heifers the probability is those heifers were too small and actually we're diverting all of that knew all those nutrients that should go to milk into increased body growth a heifer is gonna grow to reach her Dex targeted weight before she's going to produce milk the other is and this is I think we're one of the only systems in the world that does this is when we look at distances how far do these cows walk and when we get into grazing systems or dry Lots this number can get really big really quick this is an example this is a 4,200 cow Jersey herd dry lot dairy down in New Mexico and if you look so this is the milking Center each of these is a is a pen and way down here the late lactation cows and you can see all those little black dots that's jerseys okay they look like black black dots from Google Earth I won't make any other comments about those little barn rats they can be very good animals though so this is I just did this with Google Earth okay and you can do this path measurement so these cows going from the middle of the pen go to the parlor 500 meters one-way walking this does not include her walking within the pen going to eat or anything this is just how far she's walking to be milked one way so twice a day milking she's walking 2 kilometers a day back and forth from the parlor we go to a smaller farm we go home 750 cows three times a day milking we're 600 meters total walking distance now why is this important The Australian's gotta love the Australians for doing stuff like this even though they can't play rugby they they put cows on treadmills and they actually determined what is the energy consumption for cows walking and the way it works out is for every kilometer of cow walks on the flat it's equivalent to about 500 mils of milk energy all right let's think about this in terms of pasture farms some of these pasture farms I've been on are those cows are walking 10 12 16 kilometers a day so massive amounts of energy being being diverted from milk to maintenance energy because of activity so we again we include this within the system we also include environmental adjustments for heat stress cold stress and they both impact the energy batad the energy requirements as well as dry matter intake and if we look at some things like how is the animal clean or dry so we look at a dry cow versus a heifer is she in the wind or not if we take it even at zero degrees 100 kilo heifer that's out there she's covered in mud and manure and wet and she's out in the wind her maintenance requirement went up 300% what's gonna happen to her productive energy at this point namely Dean she's not gonna gain she's just gonna survive can we feed around this a little bit but situations like this it's really difficult and we can't forget dry cows that this is this is actually I took this picture a couple years ago in Argentina it was 20 degrees out it was a beautiful day we're standing there looking at the calves on this farm and I turn around and I look and I see these dry cows out there in the Sun panting exhibiting clinical heat stress on a 20 degree day because they were in direct sunlight exposure so I grabbed my little infrared camera that poor cow her skin temperature is 42 degrees okay so we meet when we described these animals appropriately we can try and formulate around this but it also gives us an opportunity to go to the fire management and say look these cows are heat stressed these dry cows here's the data showing what heat stress dry cows end up as let's talk about shade and here's the impact that we can see 10 minutes okay okay there's a few other inputs that are important from minus 21 to plus 21 days we actually include a requirement from a Genesis so these discussions these numbers we talk about in relation to 1,200 1,300 grams of MP in the close up dry cow part of that is that mammal Genesis requirement that we are actually are taking into account if we look at the rumen model this is the key this is the and this is what makes us a nonlinear model we're trying to model things this in a competent competitive function is it it's a competition for a substrate is it going to be digested in the rumen or is it going to pass out of the rumen so that gives us some unique things to consider for example starch and starch quality and if we look at for example fresh grain versus fermented grain we know there's this difference in total tract digestibility and by using something like a 7-hour starch digestibility we can then calculate a rate and come back and say okay this is how much is going to actually degrade in the rumen and it also gives us an opportunity to look at pricing of feeds for example instead of pricing feeds on total starch content what if we price feed on a potential room and degraded starch content we get completely different answers folks if we look at corn you know here's whole corn cracked ground steamrolled steamed flaked and look at what happens to room it'll starch disappearance we increase it now there's a processing cost but if we can ferment more of that corn in the rumen we're actually going to grow more bugs and we're gonna have more amino acid flow to the cow water we could go to something like ground barley now on a per ton basis ground barley may be more expensive than corn but when we look at it this way from a room integrated starch standpoint ground barley might be significantly cheaper if we look in an exam out for time there's been a lot of work by the Cornell group to last 10 years on fiber digestibility and we've implemented this this three time points not a look at fiber digestibility instead of just using a single time point and a relationship with lignin I used to say this was the worst forage I've ever seen the farm then I told the farm this is a farm out of Wales I told them a few weeks ago that I actually have one worse so in this case if we use the old lignin relationship it's saying that three percent of the NDF is in indigestible to the rumen but if we use these new messages which is actually using a 31 20 and a 240 our and DF digestibility we actually find that 62% of the NDF in this forage was indigestible to the cow so that's a huge difference in terms of potential dry matter intake potential amino acid flow from microbial yield from the fermentation of this forage we can also do the same with non forages using different time points but this really allows us to see the this relationship between digestible and indigestible and then how quickly this and this potentially digestible MDF degrades so we use these feed fractions within the model they're similar and concept of what the NRC and and feed into milk and other systems use this concept of ABC where a is rapidly degradable in the room and sees unavailable and be mild moderate slowly degradable in the rumen we just use feed chemistry to break these down so for example on carbohydrates what we call a for sugars b3 digestible NDF and then C is the und f2 240 our o but it doesn't matter because on farm folks if we really want to improve how these farms perform everyday we got to get farms to be doing their own dry matters on silages dama even a 200 Cal farm should be doing doing it at least twice a week larger farms we should be talking three to seven times a week doing dry matters under silages simplest thing in the world this is a mathematical approach to look at what are the important feed components ok so in this case if we look at energy allowable milk we see that MDF MDF MDF lignin lignin ok these these are the most sensitive feed chemistry inputs to predicting energy allowable melt but look at this and the protein side so this is MP allowable milk okay total pool size of protein nd F + EF + EF lignin starch again it comes back to how much carbohydrates can be stuff into her to maximize room and fermentation similar type thing with the degradation rates where it is the fiber digestibility the starch digestibility for both energy and protein that are the most related to what the potential milk production is so regardless we have very similar ranking for what are the most important feeding analytes to use and just as an example of what impact these rates and digest abilities can do this is just a real simple example looking at different types of corn processing so same diet the only thing I changed here was how was were you were we using ground corn or steamed flaked corn so it's the same amount one point two kilos of corn meal or one point two kilos of flaked corn if we look over here we can see there is about a two kilo difference in potential milk production simply by changing starts processing okay again what can we do from a what-if scenario so this you saw this yesterday from chuck with the amino acid composition of tissue milk and bugs and I always like this one because this one really gets us thinking about what are we going to do with amino acid formulation we look at the relationship between milk protein concentration and milk yield there's really no relationship there but if we look at that same relationship between milk yield and milk protein yield was a very strong relationship okay now part of the reason is whey proteins are a secondary osmotic regulator to milk following so as we increased milk protein yield we're either going to slightly change milk protein concentration but we're more likely going to see an increase in volume to keep that osmotic pressure equivalent for the mammary gland so if we look at the relationship between this is total like total grams of lysine metabolizable lysine total grams of defining MP Metheny and with protein yield we see there's a really good relationship so that really shows us and guy it's where we're most likely one of these two amino acids limited as we increase gram total grams even of the tabloids Matheny in or metabolizable lysine we're going to see an increase in milk protein yield now over the last few years Cornell has been evaluating how to look at things a little differently with amino acids how many of you have ever done poultry or swine nutrition okay now if I ask one of you tell me what intake is of a 50 kilo pig what are you going to reply with you're gonna reply with this much energy for years the monogastric industries talked about intake and amino acids in relation to energy in ruminants we were always taught dry matter intake in other in something like amino acids or MP or energy but really when we look at these amino acids and energy are pretty tightly regulated and when Cornell went through and was deriving some of the next generation of the model they started looking at things this way we actually find that when we look at glycine in relationship to energy okay it versus the efficiency of use of those amino acids that there's a really really tight relationship which is great because now we can start formulating around this so to maximize milk protein yield we're talking about of lysine so this is grams of metabolizable lysine for em cal of em ii intake in this case for lysine to maximize milk protein yield we're talking about a three to one relationship when we talk about lysine it's good Metheny it's an even stronger relationship and we want to Maxim Maxim eyes milk protein yield we're gonna be around 1.15 grams of Athenian 4m cal of em e and it's interesting in that I've kind of evolved to that same level in my prepartum diets and I've gotten amazing results now that does mean that I'm supplementing with thiamine and almost every diet and you have to keep in mind though that different methania sources can give us different results this is a little trial I did when I was consulting for a company almost 20 years ago so the control diet and then either product a or product B we can see that both of them reduced mu n we can see that both of them increased milk fat and we can see that with product B we actually got a higher milk protein concentration response than product a okay at this point it's like product day he looks great where product B looks great until we look at this when we look at milk yield product a actually gave us a greater milk volume response and when we look at grams of milk fat and grams of milk protein they're the same so at the end of the day it's a question of how is the farm paid for milk if they're paid per kilo of fat and protein it doesn't matter if they're paid for volume and a concentration response may be a different decision so we have to be aware that some of these products will respond differently now I will say in this case product a and product B are from the same company and there's only one company that has two Metheny products on the market so doesn't take much to figure out what company and say oh it is okay and we have to remember amino acids are required we look at not only the production response but we have to be thinking about immune function and reproduction these are just to some people they're added benefits I think we're getting enough data now that we'll be able to create an immune sub model to really show how much energy and amino acids are being pulled for adequate immune function so first things first with any of these models and I tell people this - I am NOT a ruminant I'm not a cow attrition us first and foremost I'm a microbial nutritionist I'm gonna do everything I can do to maximize room and flows and then I supplement what the cow requires we want to make sure that we have good inputs that were maximizing for amenable carbohydrates want to keep enough PE and you have to keep the room and healthy supply adequate ammonia and some degradable protein to keep the bugs happy and then supplement with appropriate ingredients the bypass proteins amino acids fats whatever to reach the targeted level of production that we want and with that thank you very much for being here and if you have questions [Music]