The Applications of Precision Feeding - Gestal Swine Summit

good morning and thanks everyone for joining us for the second week of the just all swine summit my name is amanda udemark and i'll be your host today for our discussion with dr jeff knott from ideal animal nutrition dr knott has over 17 years industry experience and as a specialist in providing technical nutrition support to swine producers around the world jeff has a bachelor's in animal science from north dakota state university and a master's in phd from the university of minnesota in animal science and swine nutrition he has vast domestic and international swine nutrition and business experience jeff spent 10 years with a national nutrition company before founding ideal animal nutrition he is a solution based big picture thinker that can communicate the details jeff is very good at implementing technical nutrition and technologies into practical field applications i'd invite you to please join us today after the presentation for a live panel discussion so feel free to drop questions in the chat area during the presentation to be answered afterwards by either dr naught or our panelist with that i'll turn it over to dr jeff knott thank you amanda for the kind introduction and good to see you and thank you to just all for allowing me to speak on the applications of precision feeding my outline for this presentation today is an introduction going over some of the current interim pork in nrc 2012 model and modeling how i see the application of precision feeding as it is today technology that's needed into the future to advance our understanding and execution of precision feeding and then concluding with a few challenges to the industry like amanda talked about jeff knott masters and phd from the university of minnesota in the early 2000's focusing on ddg research and in 2014 i founded ideal animal nutrition and we really pride ourselves at being at the slat level and doing swan nutrition feeding and consulting and we believe that to do the best job for pork farmers and pig farmers and our clients you have to be in the barn you can't do an adequate job sitting behind a desk or in meeting rooms you have to let the pigs tell you what they're seeing and make applications based on on data it's a team of three of us myself simon kern came on in 2018 he has a master's degree from south dakota state jessica jasper has her master's degree from iowa state focusing on purrs and nutrition interactions and she joined ideal in 20 2020. we have research capabilities where we do ween to finish nutrition and production research and we'll share some of those insights with you as we go through this presentation so what is precision feeding this is kind of a big definition but as i define precision feeding is meeting the daily nutritional requirements contained in the right feed provided at the right time based on individual production not population means so like i said this is a very big definition and questions that always come up when you're talking about precision feeding is what is the correct level of nutrient supplementation and where do we get our estimates from what is proper feed and how it's formulated how do we know what individual production levels are and is it possible to provide and execute precision feeding in a commercial setting i think understanding and having a good nutrient estimates is paramount to precision feeding otherwise we're going to continue to use conservative estimates of requirements and putting a large safety margin on top of formulated feed this costs money to the producers and is not necessarily sustainable in terms of economics and also the environment i hope by the time i get done with this presentation we can all agree that we have a long ways to go in terms of precision feeding there's a couple models that i really focused on for this presentation that i believe give good estimates and there's other models that have been created but the two i want to focus on for this this presentation is the in-report model in nrc 2012 model in gestation you can put inputs for body composition estimates of growth of conception conceptus and protein pools partitioning of energy and amino acid requirements in lactation again body composition milk production partitioning of energy and amino acid requirements i've used both of the both of these models extensively in this presentation and i i must give good credit and a lot of respect to the authors of both of these models as they put a lot of time and effort into trying to carve out what the estimates of requirements are for sows and gestation and lactation starting with the in-report model i really like this dynamic partitioning over time you can see along the bottom access here we have days of gestation and we have lysine in this case lysine digestible lysine estimate requirements in grams per day and you can see maintenance requirements remain relatively stable across the whole gestational period starting at about day 20 we start seeing some litter requirements for amino acids but it's not until we get to mid to later gestation that we see the amino acid requirements for for the sow and her litter increase she has reserves that fluctuate throughout the gestational period and represented in green is the excess nrc 2012 also has very good graphs in my opinion it gives a little more detail in terms of what the protein deposition is based on mammary tissue the uterus over time placental and fluids and also products of conception the fetus as it increases through gestation based on feed intake it will give you an estimate of what lysine requirement is in grams per day and then calculated into a percent of the diet i'm a very visual person visual person and i like to see these sorts of graphs grasped in order to help conceptualize what's happening over time through the gestational period so i'll start with gestation so like i indicated previously these models have been have been created but they give different results and sometimes very different results based on uh the production data that's the inputs that are in and then the results that are achieved through these models and one question i always have is how do we reconcile differences between models especially in this case the in-report model and nrc 2012 models and then throw on top of that how do we reconcile the differences among what genetic companies are recommending uh for us as well i haven't studied a lot between why there's differences between in-report and nrc 2012 i don't know if it's geographical or related to ingredients but someone needs to spend some time reconciling these and pointing out the differences on why we get different results when we put the we put the same input into them and overall why is this important i think it's very important because precision feeding really starts with having good estimates of nutrient requirements as a starting point and if we under feed there's going to be dramatic uh decreases in production but if we over feed nutrients we're also going to cost the producer a lot of money and also may have some negative effects on production as well to kind of illustrate a couple of the differences that i see between the two models i took a world real world example from a client of mine it's a 3800 sow fare to wean facility and currently they're doing about 30.6 pigs per cell per year with an average parity of 3.3 i use the in-report model to estimate uh their body weight at breeding this is in kilograms uh actual production data on total number borne so we can see that parodies one and two are lower we get into the sweet spot of our parity distribution in a 16-4 or a little over 17 piglets born in in parodies four four and five birth weights are estimated at about one point four four kilograms per pig and then actual production data again on pig's wing per litter gilts they load up their guilt litters pretty heavily and then we can see what the effects of aging sows are on overall pigs weaned actual weaning weights are captured and then using that data to estimate what the body weight after farrowing is from the inner model so i had these inputs put into both models the inreport and the nrc 2012 models we've got very different results and i'm going to illustrate those to you as we continue through the gestation models starting off with the inreport model you can see that parodies across the bottom here and then the average parity when i added the added the average parity into the model with the average production of the farm you can see that in parities one and two we have a higher lysine requirement and three for that matter in early gestation and what we term early gestation in my mind is from day zero to 90 and then intra gives a nice estimate of grams of lysine requirement at day 105 that's entered and that's represented in in the gray so as younger animals are still growing through parodies one and two especially there's a higher digestible lysine requirement as compared to older cells that plateaus and you can see that requirement in late gestation for gilts and parity one or higher and then it again plateaus across the bottom if we compare and contrast that with the nrc model it follows about the same um gilts parody ones have a higher lysine early and then again at plateaus in early gestation for the older sows and then in late gestation we see that plateau again that in late gestation older parity cells don't have the estimated requirement for digestible lysine as an early saw would have it's interesting to me to put both of these models on the same line and i apologize i used a line graph instead of a bar chart in this case but you can see parity again here and this this line here represents intra pork sid lysing requirement in early gestation the line above it is nrc 2012 early gestational requirement in report late gestation or day 105 what the estimated requirement is and then nrc uh 2012 uh in the top line and you can see there's vast differences even though we we entered the same way at breeding and the same total born uh very different estimates of digestible lysine or amino acid requirements uh in early and late gestation especially for parodies one two and slightly for three at parity four uh both of these models line up very nicely to each other and the estimates are very close and then as we get to the older cells they plateau and again remain very close um as you look at the different parities for early and late gestational uh requirements the concerning thing in my mind is that in in parities one and two there's roughly twenty percent difference there's twenty percent difference between early estimates from inra and early estimates from nrc 2012 and also from late gestation requirements and how do we reconcile these it gets to be pretty interesting and then i see that as we get to parity after parity four nrc estimates are actually lower than what in reports are and i hope you're not starting to think that this talk is just going to turn into a model discussion it isn't i just really need to drive home the importance and the differences of these different models and how they estimate nutrient requirements of these different animals so to put it in more practical terms taking both of these models what do diets look like if gilt feed allowances is five pounds per day if we go back here we can know how many grams per day the estimate is we know that we're going to feed these gilts about five pounds of feed so we can calculate a percent sid lysine of the diet and formulate an appropriate diet that includes 0.45 digestible lysine to match what the early gestation estimates are that diet today would cost 228.59

we can calculate or formulate a 0.54 lysine digestible lysine diet to match what the nrc 2012 early gestation requirement or estimates are and then as we get to later gestation we can see that both the inra and the nrc models estimate that based on a five pound inclusion or five pound daily feed allowance that we need to increase the digestible lysine of both of those diets in order to meet uh the growing needs of the products of conception or litter birth weights as they go on and for all these estimates or for all these diet formulations i use the cost based on 5.50 per bushel corn 360 soybean meal and 170 uh dollar per ton soy hauls what we can calculate from this example is for every 0.01 percent sid lysine change

in the diet that costs about 68 cents per 0.01 so to put it in more simple terms if our estimate is off by 0.1 percent that could be up to 6.80 per ton

or would equate to [Music] cents 4.65 dollars and sixty-five cents per sale uh per year so this has huge economic uh implications on it but also if we're over feeding a lysine that has that economic implication but if we're under feeding lysine uh it has production implications too where we might have low vigor low birth weight pigs at birth and that those pigs don't do as well uh in the lactational period for in terms of preventing mortality so again it's very very important in my mind that we have good estimates and we understand what good estimates are and we formulate diets appropriate to appropriately so now let's throw on top of this some of the some of the genetic companies estimates into the mix and i took the most recent pic from their pic feeding and nutritional guidelines 2021 they're recommending for parity one or gilts and then parody two plus 11 grams of digestible lysine per day and they have a an interesting website the dynamic female pig at pic where you can enter these estimates in and and and get digestible lysine requirements as well or estimates as well dna breaks it into two phases from day zero to ninety and then day ninety to one fourteen parity one animals uh estimate requirement of thirteen grams per day in late gestation 16 grams per day parity 2 plus 12 grams per day and then 15 grams per day late topics norrisville from their 2017 tn 70 manual breaks it into three phases day 0 to 49 day 50 to 84 and day 85 to 115 again about 13 grams of lysine early a slight increase in mid gestation and then a significant increase in digestible lysine estimate requirements in late gestation for gilts you can see for parodies twos and threes 12.8 drops down a little bit as those animals aren't growing as much or they're reaching uh their mature body weights and then a slight bump of digestible lysine and late gestation and parity four plus lower early lower in the middle a slight bump at the end in order to meet the growing needs of the of the products of conception and birth weight fast as well has requirements or estimates for p1s of 13 grams per day p2s about 12 p3s 10 p4 8.7 and then genesis is a little over 13 for gilts and 12.2 uh

for older cells so let's let's try to look at this in another way again i'm a i'm a visual person so across the bottom i entered what pics recommendations are dna topics norrisvin fast genesis in report and nrc 2012 and you can see for and this is all based on guilt estimate requirements you can see early there's a wide variety of what recommendations these different models and or genetic companies are recommending and then as we proceed through gestation we see that bump or increased estimate requirements starting about day 85 to day 90 where we start increasing digestible lysine uh needs of the of the litter in order to optimize birth weight and vigor at birth there's still a lot of variation amongst these models about 32 percent variation in early gestation and then a huge variation in estimate requirements of about 85 percent in late gestation so i think you can tell from from this example uh we think that pigs are pigs but based on production levels and also genetic potential uh different companies and different groups think differently on what the requirements or estimated requirements of those pigs are or those sows are for digestible amino acid if we look at a parody 2 again we still see quite a bit of variation in early gestation and we see less than what we saw in gilts in the previous slide but there's still a fair bit of variation so this really doesn't give us a very clear path on how nutritionists should formulate diets and what the actual estimated requirements are so what do we do today just to put it into an applied sense of things i took a blended example and i averaged the in-report in nrc 2012 estimates for early and late gestation and then applied some economics to it versus a 0.58 percent sid lysine single feed that's fed for the whole period of time and fed at five pounds so you can see again i entered the production data for the real world farm that we work with and then i took the average between inra and nrc not saying that's right or wrong uh but that's the example that i use for uh gilts in early gestation and then late gestation and based on a five pound feeding rate we could feed a half point five percent sid lysine early to 0.71 and then across all the parities uh at a five pound gestation uh feed allowance what the percent lysing would be in those diets so if we look at applying cost to this over time if we feed a one feed 0.58 percent sid lysine across here that's this level right here right under 0.6 for one cycle or 114 days that'll cost about 67.48 cents per female if we do a blended feed for parody ones and parody twos where we're feeding them according to this line about 0.5 early

and then we do a late increase in in digestible lysine late we can save roughly 56 cents a pig 58 cents a pig if we do a p3 blended represented by this kind of reddish line that's hidden well lower and then we go above to meet the nutrient requirements of those pigs in late gestation six 65.36 parity four plus uh which would be your your gray line here is sixty four dollars and twenty nine cents and then if we do an average for the whole farm of of uh parity three point three that cost would be 64.94 so it shows that there's a huge advantage to not over feeding digestible lysine in early gestation and we can still save money by going higher to meet what the actual requirements of those animals are in late gestation to make sure that we get optimal birth weights in all cases the blends were less cost than the single feed and more more important or most important is making sure we don't under feed gilts um and p2s especially in late gestation when they're still continuing to grow and if we do that we could have poor birth weights and poorer vigor and vitality um at birth and then there's also the economic implications of that as well so like i said i didn't really want this to turn into a model talk but uh it seems the jury is still out on how we should feed and or how many grams of sid lysine females need each day and i'm not i'm not picking on anyone when i point this this out it's the truth of it cell research is very labor intensive it's expensive and it's very variable in terms of the responses that we get so i'd like to see us get more more information and more validation of the current models going forward um i think we can all agree that amino acid requirements change by day 90. i

know there's been a lot of research that's been done that shows that late bump feeding and gestation does it just it doesn't pay and sometimes i wonder if some of these trials that are set forth are already over feeding lysine in early gestation that we don't get the bump that we would see naturally if we were feeding to the sow's requirement in early gestation and then meeting her litter needs in late gestation i think if possible we should try to feed two feeds if we could feed an early gestation diet from day zero to ninety that's lower amino acid higher energy that'll help regain body condition that might be lost during lactation and then if we could feed a late gestation diet that would be higher in amino acids that would focus on fetal development and maternal tissue development in late gestation would be very good if we have the ability in our ess or electronic style feeding stations if we can blend we could do two feeds and feed by average parity i think you could see from the real world example that i showed that the requirements for gilts um and priority uh two styles are very different especially early and late compared to older sows and if we have two feeds we could really do a good job and feed by average parity production if we know that our our gilts are going to have 14.8 pigs born and our parity twos are going to have 14.3 pigs born we can match what those lysing requirements are based on average production for that parity i'd say if nothing else try to get gilts and p2s on their own feed line and really identify gilts and p2s and don't feed them like older sows and we can if we can get gilts and parody twos on their own feed line we can really do a better job by not over feeding amino acids to paired before plus sows or underfeeding our gilts and and parody 2 cells but i think we can all agree that um the amino acid requirements change and that we have a long ways to go in terms of gathering more information and validating our current model so we have good nutrient estimates going forward moving into lactation a question that i've always had as a nutritionist is in most phases of production we formulate to a digestible lysine or an amino acid to net energy ratio we know that it takes x grams of of lysine and x calories of net energy in order to put on a pound of gain in growing and finishing pigs to get the optimal average daily gain and optimal feed conversion but it feels like lactation feeds are really uncoupled meaning that we focus on grams of digestible lysine per day and then we kind of are takers of whatever calories we get from the feed i don't know that this is the right approach we're starting to see some genetic companies in their recommendations list digestible lysine to net energy ratio for lactation for gilts and sows which i think is a is a great step uh moving things forward in the past i believe based on some of the data that's out there from denmark peter thiel's data that in the united states we especially over feed lysine to net energy ratio and i'd like to see some more of that data be conducted in the u.s we know that body condition loss is a major concern during lactation and net energy balance is a big driver of what that body condition can be we know that over feeding amino acids can be negative towards energy balance and lactation feed intake and that higher heat increment of digestion exists from excess soybean meal because it takes more energy to break down and excrete excess crude protein and amino acids jeff i'll interject with just a quick question could you for for our audience here and also for myself could you go into just a little more detail about heat increment from digestion how soybean meal relates to this and why that's such an important discussion in itself sure no that's a great question and thanks for thanks for asking uh basically heat incremental digestion uh expresses the amount of heat that's given off when a certain ingredient is digested so we know that we start out with gross energy which represents 100 of the energy that's available to the pig and as we go through digestible energy and metabolizable energy we excrete some of that energy through fecal energy and urinary energy loss the difference between metabolizable and net energy is what's called heat incremental digestion and if we look at ingredients that have high fiber and high crude protein those typically have higher heat increment than your cereal grains like corn or wheat or barley are some of the more carbohydrate fat rich ingredients that we have so anytime there's heat loss in a system that's energy that's being wasted if you will so the more heat that's given off the less net energy that'll reach uh productive properties like milk production in this case and what's interesting to me if you look at what the gross energy between soybean meal and corn is soybean meal has higher gross energy than corn but on a net energy basis it's at a significant discount to corn because of a lot of that excess energy is being given off as heat and we know that pigs are homeotherms and sows so what a homeotherm means is that a an animal will keep similar to humans they'll keep their body temperature higher than what the environmental temperature is in most cases and the way to balance heat production is by feed intake so if we're already in a outside our thermal neutral zone and suffering from some heat stress the first thing that we're going to do is we're going to drink more water obviously but then the next thing is we'll reduce our feed intake in order to reduce the internal heat of combustion of of digesting that feed in order to balance with our environmental temperature so [Music] in lactation the sow is already outside of her thermal neutral zone and if we have a high a diet that's uh high in soybean meal we're putting more heat onto her or if we're feeding excess soybean meal we need some obviously to meet the amino acid requirements but if we're over feeding it and formulating a diet that's too high in digestible icing we're using more soybean meal this could have a negative effect on her overall heat balance and she'll she'll balance that out by reducing her feed intake which has again trickle-down effects in terms of intake of certain nutrients energy amino acids calcium phosphorus etc etc does that kind of answer your question amanda it does it seems like that could honestly be an entire discussion all in itself so thanks for that that extra background there sure could sorry if i went along on that one you're fine thanks so i guess when i when i plug this production data into the inreport and the nrc 2012 models they're very close in most cases the estimates uh based on the production from the real world example that i cited earlier they're within about three grams per day which in my mind is acceptable in report and nrc are recommended or estimating that for the this level of production that gilts are consumed 46 grams per day parody 2 is 49 grams per day parity 3 plus about 44 grams per day now again to look at what the genetic companies are recommending based on gilts and older sows you can see the pic is is higher dna based on the level of production of 30.6 pigs per cell per year i estimated they're recommending around 63.5 grams

per day topic's norrison bases their estimates or requirements on uh litter gain parity ones based on a low level of of litter gain would be 52 grams at the highest level about 68. what i estimated from this farm would be right around 59 grams based on their litter gain and parity 2 plus 51 to 67 grams per day so again very much higher than what the in-report and nrc models are saying fast 60 grams for gilts 63 grams for parity 2 genesis higher in gilts lower in two so really to summarize this difference genetic companies are higher are estimating higher requirements of amino acids than both the in-report and nrc models so the question becomes can we blend two lactation feeds in order to match the curve based on the the estimated energy and amino acid requirements so i took for parity one i i plugged the data into the model it's estimating you know in early just in early lactation excuse me relatively low levels of digestible lysine requirement and then as we get into peak milk production uh the estimates for digestible icing increase based on the model and what we know for feed intake we modeled 5.8 pound per day or 5.8 kilos or 11.8 pound per day lactation feed intake to get what the diet should be formulated like so then i took that data and i formulated two diets um to match the curve based on these requirements so the high lysine diet would be a 1.05 digestible lysine diet and to be honest with you there's many farms that i feed today that we feed a 1.05 sid lysine throughout all lactation right or wrong that's just kind of where we're at based on what some of the estimates are and trying to meet those requirements so the 1.05 glycine diet will cost about 276

dollars a ton and then the low uh digestible lysine diets formulated at 0.59 that'll cost about 229.46 and again based on the same uh price points that we are we're talking in the gestation example as well so for guilt litters you can see the blue line represents the 1.05 digestible icing diet and then the gray on top represents the 0.59 so we're blending in this case about 85 percent of the 1.05

lysine diet with 15 of the lower lysine diet in order to meet what that guilt requirement is in the first day of lactation and you can see as feed intake goes up we blend more of the lower lysine diet and then as amino acid requirements increase through later lactation during peak milk production we start increasing the amount of the 1.05 lysine diet back in in order to meet the daily requirement in grams of lysine per day so per lactation or per 21 day lactation i'm estimating 248 pounds of lactation for a gilt that'll cost around 32.63 in lactation feed and that guilt will consume roughly 1035 grams of digestible icing per day which is an average of 49 grams and she's going to consume over this lactation period about 277 mega cals in that energy if we feed the single 1.05 lysine diet across all these and without a blend that's going to cost about a dollar 65 more or 34.28 per lactation cycle she's going to consume significantly more grams of sid lysine 56 grams versus at 49 and real a small reduction in in net energy intake i think it's really interesting that in guilt that this difference is so low and i don't know that economics tells the whole tells the whole story here i think there could be some negative effects on overfeeding digestible lysine in terms of overall feed intake i've set them equal here but we really don't know that looking at a parity three just to show a different example on how uh maybe robust a blended feed could be to meet the requirements of older cells or different parity styles again you can see in day one we're going to feed mostly 105 diet but then as feed intake goes up we use more of the lower lysine diet compared to the higher lysine diet so when i simulated these uh this parity three the estimates were about a little over 15 pounds per day intake the blended feed like shown above would cost uh 39.15 uh per sale per lactation uh cycle

and she'd consume roughly 1100 grams of sid lysine which comes out to be 52.4 grams per day which is close to some of the genetic company estimates that we we looked at previously she's going to consume 355 mega cals in that energy but if we feed the single 1.05 digestible lysine that'll cost uh roughly four dollars and what's that i think i have it written four dollars and 52 cents per sale higher to do that to feed a single base a 105 and she's going to consume 1500 a little over 1500 grams of digestible lysine per day which comes out to be 71.7 grams of sid lysine that's a huge difference in my mind in terms of digestible lysine intake i wonder if we did that and we are currently doing that like i said there's many lactation feeds where i'm feeding a 1.05 are we over feeding digestible lysine with a huge intake like a 15 pound per day intake and what's the environment in that fairway house when we do that there's many times i've walked into fairone crates or fairwin rooms and been kind of overwhelmed with ammonia smell i wonder if some of this overfeeding well i do i know it does some of this overfeeding of amino acids and cream protein are definitely reducing the acceptable environment in those lactation rooms also by over feeding digestible lysine and crew protein and amino acids we could be hurting our overall body condition or weight loss increasing weight loss as she's coming out of lactation because we know if we're over feeding amino acids it's going to take a fair bit of energy to deaminate and break down those amino acids and get them excreted so i kind of briefly talked here about some of the potential for production implications as well besides the economic you know in my mind the two big production um implications could be you know the potential for lower feed intake because the 1.05 lysine diet has significantly

higher soybean meal concentration than the 0.59 and then also that could lead to poor energy balance throughout the whole system i don't believe that the economics of a dollar 65 per gil cycle and 452 per sale cycle it doesn't tell the whole story if you look at some of these data that he presented just last month at the iowa state international conference in swan nutrition he looked at one component feeding versus two component meeting feeding based on amino acid requirements and you can see that by feeding one component their data would suggest in 2016 that they got 12 and a half kilograms of milk yield if you feed two and better match what her requirements are and provide more energy and not over feeding amino acids you'll get about 13.3 a significant improvement in milk yield weaning weights also go up with higher milk yield and then also that cell is taking less fat off her own condition in order to support milk production so i guess my take homes in lactation are don't under don't over feed amino acids and under feed energy during lactation this results in excess nitrogen excretion higher heat production through heat incremental digestion potential for poor subsequent feed intake and if we lose a lot more body condition coming out of lactation there could be implications also on subsequent reproduction and subsequent parodies down the road jeff before we move on i have to ask one of my favorite questions for nutritionists so in your opinion and and based on your approach to this is there a particular place on the farm where you think it it maybe makes the most sense to implement precision feeding first whether that be the gestation barn lactation or something like the gdu do you have any opinions on that it's a good question i always have an opinion on something right i think um you know most of the farms that are built today have created gestation you know created lactation we're obviously seeing a move more towards group housing and gestation but in order to execute precision feeding and gestation in my mind is is quite difficult because we have a single gestation line that runs over drop boxes over crates that we're dropping feet into every day and we're adjusting uh feed allowance based on generally a tab or some sort of way to adjust the feed box i think if you could probably be more cost effective to retro a lactation or a fair one room today and the economics uh to using a blended feed would probably dictate that as well there's a higher return per cycle according to this model that i ran a dollar 65 in gilts and four dollars and 52 cents in sows to do a blended feed in lactation to meet the estimated amino acid requirements than would be in gestation i think it'd also be better for the sow and better for maintaining body condition and also subsequent reproduction if we could focus on getting amino acids right based on her daily needs not over feeding them and then the cost standpoint as well absolutely agree thanks jeff carry on okay i'm gonna shift gears here a little bit and and talk about some of the things that we're learning from agronomy i think in agriculture agronomy especially agronomy and followed somewhat closely by the dairy industry are the leaders in developing and applying technology development of satellite imaging has grown exponentially over the past few years i call it agronomists in their toys but they can really look at imaging from a thermal standpoint a hyperspectral or color and then thermal infrared to look at heat from satellites and what they what the agronomists can do is they're using this data in order to help make a host of decisions in terms of is there a disease in this field can we estimate yield what's the maturity of this of this crop what's the biomass and what's the overall health of of crop or fields based on some of this imaging and i think some we have a long ways to go in the swine world in terms of learning from agronomy and the dairy industry for that matter and applying some of these things which i don't know if you know but i like to do a little farming on the side too so here's here's one of my fields called the gravel pit and every week i get these these nice graphs if you will or pictures from from space looking at field health and you can see wherever there's red that means there's problems in this field this field this field is a gravel pit so it's rocky it's hilly there's some some spots where the fertility is or the soil is not as good we can see that from space it points out where the problem problems of this field are you can see in the deep green that means this has high field health that means things are really really good in those cases and we also get a vegetative map along with it that kind of mimics what we see over here on the field health side where we see these problem areas here this is a really wet drowned out spot that has some tile that got fixed but then at the end of the year we can take our our yield maps from our combine and so this represents the green as high yielding areas for corn last year and the red means it didn't produce very well so we can really reconcile these problem areas of the field with what we get from our yield monitor in terms of production and so going forward with this data in mind i can work with my agronomist to come up with a variable variable rate fertilizer application scheme so we'll we'll put more fertilizer in these areas that are high producing we'll put less here because historically and year after year these areas of these fields will be poor producing also we can do variable rate planning in these spots plant more dense areas here of corn and soybean meal and plant less here to give those plants or those seeds a better opportunity to produce the goal in all this you know what i i call this precision agronomy is we save resources for the higher producing ground and we take resources away from the lower producing ground so that we can optimize our production and also econo optimize our economics that we're that we're doing with this thing and the question i always have is can we do the same thing with pigs because we know that on sal farms not not every cell is going to produce equally we have sows down here that have a lower total born they wean fewer pigs that are lighter weight compare that with the girl in the crate next door that's a really high producing stop and right now we don't really know and we don't really treat these animals any differently and i think we really should so again the question is why aren't we doing this in south farm especially electronic south feeding stations farms that have esfs in them because we have a female that's going into an area every day to to get fed why can't we capture uh feeding capture data on her daily as she walks into that feeding system and that we use that information to feed back into a model that could that could help us better uh estimate what her nutrient requirements are i know this if this doesn't exist today but i think this is something that's going to come really quick so what i what i what i foresee into the future is when a when a sauer guild enters the esf that we might have estimates of litter size every time she walks in there every day she walks in there there'll be a real-time ultrasound uh her back fat through digital calipers you know a lot of farms right now use the manual calipers to estimate body condition why couldn't we develop a digital caliper based on laser a digital back fat probe where we can estimate what her body condition is body weights either a scale in the esf or cameras to predict what her body weight is doing as she's progressing through gestation and i think it'd be really cool that all this information would flow back into a model that would estimate what her energy and amino acid requirements are on her individual female data and then a individual feeding program is tailored to her every day based on these changes in litter size if she loses pigs during gestation what her back fat and body condition is doing and what her body weight is doing in farrowing um i think we're going to get to a point where we're going to use the there'll be thermal cameras or thermal imaging above rows of sows that we can use to get body temperatures to indicate six hours that need treatment i think we're all feeling a little bit of a labor crunch uh in farms today and we've got to get better at utilizing technology in order to help us make decisions on farms and one thing that i see that currently gets so overlooked a lot is finding sick sows taking temperatures of animals that are off feed and trying to understand if they need treatment or what's going on with them today body imaging to estimate litter growth rates and soil body condition loss during lactation i think it'd be very very important again if we could get blended uh two two feed lines over uh every sow and then do blended feeds um to better meet what her requirements are and and also to better measure what her daily feed disappearance is really helps the nutritionist do a better job at formulating diets and then blending diets lactation i think we can all agree or we can at least have the conversation that the the requirements in early lactation are very different than when she's in peak milk production in late lactation and we really need to have the ability to feed two lactation diets in order to meet the energy and amino acid requirements based on her feed disappearance and also her litter growth rates that we see and i think technology over the top of sows and cameras and imaging will really help us do a better job in estimating what what the true requirement of these sows are rather than just feeding a single feed and letting her eat all she wants and focusing on grams of lysine and taking the calories that come with it so my challenge to the the swine industry is i think we have a lot to learn and we should really take some notes from the agronomy side the agronomists and their toys to advance our technology that we can utilize in sow farms apply these resources where they're needed to high producing sows and withhold resources or nutrition where they're not needed based on lower producing cells that might not have the amino acids and energy requirements that the high producing cells would have we need to be able to apply this technology any new style farm that's being built should have multiple feed lines uh to feed by production level and parity as we move forward into gathering more of this imaging that that will be coming down the line i guess my my major point is don't skimp on feeding system bins and feed lines that are going into these new sow farms we need to continue to improve and make more robust changes to the gestation and lactation models spend some time reconciling the differences between in report nrc 2012 and then what the genetic companies are recommending and then i think if we continue to improve our models and capture real-time data as i've kind of outlined into the future then we can truly embrace precision feeding and do what's best for the sow and optimize economics on these farms as well with that thank you very much for your attention and i'm happy to answer any questions you might have thanks so much jeff i think i can speak for everyone as most of us probably expected you made some really great points and left us with more great questions and more ideas as well so i think it's a really great place for us to start off our panelists discussion so to our audience please stick around as we transition over to that discussion and then please continue dropping us questions in the chat box for both our speaker and our panelists we'll see you there good morning everybody and thanks for joining us for the live portion of question and answer i hope that you guys really enjoyed dr dr knott's presentation um and so what i would like to do at this point is just briefly introduce our panelists that we have today for the live session um and then what i'll have them do is just go through and briefly tell us a little bit about themselves and why they would be a really good fit for this panel so as you'll see dr jeff knott has also joined us for the live panel discussion he will be able to also kind of join in on this conversation as well so i'll just briefly begin again with introducing everybody and letting them say just a little bit about themselves first we have uh francis simard hello amanda hello everyone my name is francis i'm based in the quebec city area in canada i've been a swine nutritionist for over 20 years now with a special interest in cell nutrition i'm also coordinating our research here in canada and recently we have moved into the precision feeding area especially for gestating sounds by installing gestalt multi in order to understand a bit more what dr knott was telling us previously thanks francis glad to have you uh dr bruno silva hi good morning everybody um it's a pleasure to be here with you with you all um first of all congratulate jeff on his presentation i think there's quite a lot of nice information that we can discuss here um so um i'm a university professor here in brazil um at the federal university of minas gerais basically i've been working with with pigs the past 18 years dedicated towards understanding uh nutritional requirements and also environmental adaptation and um and especially the past years we we've been working a lot uh trying to understand the basics of uh self-feeding behavior um mainly uh uh understanding how do these cells eat feed intake patterns and how can we bring this information into models into precision feeding and of course everything associated with technology we are users of the gestalt technology here perfect thanks bruno uh raphael got the air please hi everyone uh thank you and thank you jeff for your great speech um i've just completed my phd a few months ago and this phd we was with the doctor and christine and we were working on the precision feeding of baking soda and during this phd i was we built a new system that is able to feed lactate themselves in real time according to precision feeding principles and i hope we will discuss about it and what the jeff introduced during the speech awesome so again thanks uh we all heard jeff's introduction earlier so i won't make him kind of reintroduce himself now but again thanks jeff for sitting in with us in case anybody has specific questions about any any comments or maybe recommendations that you made while while you were doing your presentation so uh just to kind of move into some of the the questions and again the discussion what i'd really like for you guys to do is if you if you continue to develop questions please drop those in the question and answer section we've had some really good ones come through and then obviously i've i've got some questions of my own as a lot of people do probably so please just feel free to to keep letting those slide in so one of the questions i guess that i'd really like to start off the discussion with and what i would like to do maybe for our panelists is just to kind of direct that to one person at first but again please feel free to jump in if you have comments or anything to add to the discussion so uh one thing that i the the most recent question that i saw come in that's just right on track is what's holding the swine industry back from embracing precision feeding so let's start with maybe uh bruno if you don't mind to take off with that question what you think's holding the industry back um well it's a good question um i think two aspects one is knowledge lack of knowledge of understanding um how to use this technology how can we implement this intel into our barns into our farms and the second probably cost i think the cost would be what would be a main issue but as jeff very nicely presented to us i mean a precision feeding has this great advantage of working with more efficiency so at the end of the day um if you do your maths correctly it does pay off so um i think it's it's it's really something worth looking into investing but probably cost and and lack of knowledge let's put it like that sure if i can add amanda up until uh recently maybe uh after the technology to apply that in farm was not uh all there the knowledge was developing but i think just recently the few two three years uh there are technologies on there that are more cost effective and also can be simply an application so i think we're getting there and then all the knowledge gaps that doctor not mentioned are being studied as we speak so we're really close to being able to have cost-effective application of precision feeding on farm very good anybody else have anything to add yes i agree with the with francis and i would like to say that now data is becoming more and more available in farms and we have now developed some new techniques to analyze this data and use this data uh with the to to unders use this data so it's a new point and maybe now we have to think also how to use and manage the feed stuff on palm because precision figuring is obviously requiring more and more of this step maybe two maybe three feet step so this is another point very good kind of moving on to a second question i know one of our big focuses here in the u.s is just focusing really on feed

costs they've been a little bit high this year starting to decrease just a little bit now but that would be kind of a big feather in the cap of most equipment providers working on precision technologies you know for just reducing the amount of feed cost and not not having as much feed wastage but i'd be really curious for your perspective especially raphael just uh and maybe bruno as well for different different parts of the world i guess um as far as maybe things like nutrient excretion what do you think are some of the really big uh big ticket items that we could also accomplish with precision feeding shall i start or fail yeah okay you can run you can go i'm sorry okay okay okay great then great um well it's okay it's a good question um it's a good opportunity to share a bit our experience in this um we we recently we recently just uh ended a big study on on precision feeding with gestating cells here at the university uh generally the phd thesis uh we're still now writing papers and everything but um what we did uh during the study was um the thing is the thing with with precision feeding of course it's it you basically uh you have to consider that you always you need a good data set you need a good model i mean you could you need a good uh uh you need good tools to actually to um to to improve and enhance these these daily the knowledge on these daily requirements so what we did here is um we use just our multi at our at our research facilities here at university so you have the flexibility of working with two different diets and and make a blend on a daily basis but the difference here is that we we integrated into the gestalt system we integrated weighing scales so each time the cell goes into the feeding station the system would capture her body weight and together with the feed intake pattern this would be thrown into the gestation model into our gestation model and this would give me the output for the next day so what would be the real um what would be the real ratio between lysine and energy for the next day based on metabolical on the metabolic body weight and based on the feed intake pattern of the animal because this is something that we have been investigating so um what we saw from the study is that if you can allow this animal to improve its its efficiency in terms of of growth rates throughout gestation so according to the different metabolical stages uh throughout gestation you will improve the efficiency of nutrient utilization and this of course results in in several benefits in terms of productivity so we saw the benefits in terms of total body weight growth of this animal fetal development mammary gland development and the interesting thing from the study is that what we saw is that there is a direct correlation between on how you feed these animals so how precise you are in terms of feeding these animals according to parity order according to gestation stage according to its its productivity uh uh potential there is a direct correlation with its efficiency during lactation what we saw is that the cells that are fed um on the precision feeding curve so every day you have a different uh different estimates different nutritional needs or nutritional demands and you can attend this via via the the the feeding stations um during lactation these animals showed uh um a maternal transformation index much lower than the other two strategies that we evaluated just to put everybody on on board here we evaluated three curves so the the precision feeding so every day it would change based on the metabolic body weight of the animal we evaluated the the flat scheme from two uh published uh researchers um so gonsalves and malmon 2016 2018 so we use this research to establish the the the flat schemes and we worked with uh based on on the genetics uh on the genetics that we were working which is topics normally in tn70 we use the the topics model so very similar to what uh what uh jeff shared with us so the high level high so when we compare a precision feeding program with these other two strategies we see that um when you allow the cell to improve its its intake the efficiency of of nutrient usage understanding that you have these dynamics in terms of ratios between lysine and energy you allow also a better efficiency during lactation so what we saw is the cell nut only produces more piglets heavier piglets but she also produces more milk at a cost of using less feed during lactation so this was something very interesting the cells are capable of producing more eating less during lactation because we allow them to recover or to develop their body condition within what is the expected or what the animal needed at these different stages so precision feeding in my understanding is ideal if we want to work with efficiency reduce nutrient expression and if we want to to improve the the maximum output in terms of genetics from the cameras go ahead uh thank you for this study looks very great i would like to hear more maybe another time but if i follow with the lactation precision feeding trial we have made during our work the idea was to combine some scientific knowledge about nutrition and combine it with the data machine and data machine learning and data science so those algorithms were there to to predict feed intake the living intake of the individual cells but also the little growth so we have kind of we are we we add on top of nutrition machine learning algorithm and all of them some part of us accuracy and inaccuracy i mean we can't predict every kilogram of integral exactly but or every kilogram of intake exactly but it's still important to predict those parameters uh in real time and what we have seen during a trial here in canada it that was pretty grim picking was able to reduce both nitrogen and phosphorus expression compared to a conventional system uh conventional opening system and also to reduce the cost by about 12 percent that obviously it depends on what type of pit steps you use as an input and everything and again now according to the performance our performance for the various mirror in terms of liter performance growth etc and sole performance and all of these study was based on the data that are already available on farm so thanks to the gestalt patrol we were able to collect and manage feed intake data every day so that was a very great point and for the nutritional part and the prediction of litter growth we use kind of data that are already available such as the little daily little size uh variety of style and the and body weight of south before powering so uh very basic data but they show very great capability to predict and we we guess that we could do more with more and more data maybe i can add on that down the cast side if i go back in 2016 uh there was some work with the cdbq here in quebec and we had it was a trial ran on a farm applying precision feeding from the 30 in gestation all the way to the end it was uh group housing and uh at the end on the cost side that there was a saving at that time of around three dollars for sale we had some trends on the improving birth weight and survivability of p1's especially [Music] piglet but we didn't it was a multi-parity sound so multi-parity heard so we didn't have enough enough kills you know to verify these trends but uh right now there's the same institute that's running a study on 600 cells p1 so a brand new barn so i guess we'll have more more data on the the uh would say the effect of uh especially on p1 of precision feeding but one thing i want to say is since 2016 the cost of ingredients have increased especially in canada and if we do the same math that we did four or five years ago we're now at around five five dollars and a half percent only on cost of feeding not taking into account the potential improvement in performance so i think uh that is a very very interesting uh in the future and we're really looking forward to see how this high number of free ones will perform with uh precision feeding scheme and gestation and uh and then we're only applying the s4 from day 30 all the way to the end if you're able to apply it right away from the start then i think there's probably some savings because as jeff showed us the requirements of just hitting sal is quite low from day zero to day 30. so there are some saving there that we don't make if we are following the rules in canada and putting sales in group housing from day 35 absolutely i was thinking you know while jeff was going through his presentation and a 4.50 return or cost savings on on a sale eating you know approximately 14 to 16 days of lactation so eating 15 pounds a day i mean that's just almost it seems like a no-brainer i know that there are multitude of factors that go in there that are just kind of creating these barriers for implementing things like precision feeding on these sal farms but i mean especially with feed costs these days it seems like it seems like definitely a frontier you know hopefully that we're moving towards uh to provide all of these different benefits as well because again that's something that none of us have really touched on mostly working in the nutrition arenas but what are these performance benefits that we can that we can also obtain from precision feeding and getting those applications right on the farm so really good discussion the next question uh that i really wanted to kind of focus back to back to jeff he's our boots on the ground nutritionist um what are your thoughts jeff as far as just actually from an implementation keeping it as simple as you know producers are actually able to manage and get get accomplished on the farm if we're if we're looking at trying to feed at least a couple of different diets in gestation what are your thoughts on something as simple as maybe feeding a gestation diet and then after say day 90 a lactation diet you know that your females already producing that you already have on site is already getting delivered what do you think about that or is that is that something that if you're if you're going you know through the trouble of creating a different or feeding a different diet you need a different formulation it's a good question i think a lot of it gets back to logistics and execution you know in a lot of the older farms that we have it's still one one feed line over every gestating cell so that means if you're going to try to top dress something that means a person with a feed cart with a feed scoop snack feeding her which we have uh clients that do that very successfully to enhance the amino acid allowance per day in late gestation but i think as we move forward into more group housing type situations we're going to have two blended feeds where you could have either you know you bring up a good point a lactation feed and gestation feed that you could blend in late gestation in order to meet those requirements i think i think that's the frontier that we're heading towards it's just changing the mindset and the paradigm of how we've always done things in the us especially for the last uh 100 years to moving more towards the technology and implementation one of your really good recommendations was just not to skimp on feed lines and feed bins so hopefully uh hopefully some people will take that into consideration i guess does anybody have anything to add to that um well if i if i can just uh jump in um i i think that besides these aspects is also a very important factor that we have to consider and most systems do not consider this most feeding companies normally don't look into this which is feeding intake behavior so feed intake pattern because at the end of the day you can have an optimal feeding strategy i mean you can have all the the best feed additives you can have all the best nutritional concepts into the feed per kilo but if you don't understand how your animal eats when does this animal eat and how does your feed your feed curve or your feeding management impact on the efficiency of the usage of these diets at the end of the day you're not able of optimizing this type of of strategy and um we've been looking into this in in the past years and it's clear to us and it's nice that i've been sharing this information with francis also uh we've been discussing this also in the past weeks and um for instance we just published a paper where we indicated that uh during lactation a feed intake pattern of cells it's nocturnal 80 or total feed intake happens between midnight and eight o'clock in the morning and then if you look at at feeding strategies on a f

2021-09-25 09:03

Other news