Bulk Tank Milk Fatty Acids via IR

Dave Barbano on October 11, 2018 at 1:10 am

HI Everyone!
SHORT ANSWER: the models for milk fatty acid testing have not existed on Foss equipment and labs with Foss equipment producing data for these parameters have simply been trying to estimate the denovo, mixed, performed fatty acids and double bonds per fatty from other parameters measured by their instrument. They were doing this with no calibration. This approach does not work. I have tied to do this myself with data from Foss models and could not make it work. I have made laboratories with Foss equipment aware that I can provide milk fatty calibration samples and work with them to evaluate performance of Foss models on their instruments. However, the first step is that Foss needs to develop their own prediction models. Foss is finally starting to respond to this new industry need, but it will still take some time and there will be an ongoing commitment need to by Foss and any other mid infrared milk analyzer companies, as new herd management milk metrics are developed. Development of new metrics is on-going in my laboratory in collaboration with Delta Instruments and Miner Institute. An update will be presented at the Cornell Nutrition Conference in mid October 2018. I suspect that at some point, other researchers may start to work and innovate in this same area because of the interest in the dairy industry.

LONG ANSWER: Foss was the first company to have fatty acid models on an mid infrared milk analyzer but the specific fatty acid models produced by Foss were designed for food processors, not for farm management. Those models don’t seem to have been used much by food processors and have had very little external validation. The development of the first farm management fatty acid models for farm management milk fatty acid testing were done jointly by myself and Delta Instruments. These models measure different groups of fatty acids (de novo, mixed and preformed) and some structural characteristics of milk fatty acids (chain length and double bonds per fatty acid). These parameters are measured directly by the models and not calculated as derived parameters from other fatty acid models running on an MIR instrument. In addition to developing these models jointly with Delta Instruments, we developed milk fatty acid calibration samples that are available for purchase from my laboratory at Cornell University. The farm management milk fatty acid models that were developed jointly by Delta Instruments and my lab at Cornell are available on Delta Instruments and not Foss Instruments. The information of how we developed the models is published in a series of papers in Journal of Dairy Science and Foss or any other mid infrared milk analyzer manufacturer could develop their own models, but there is cost and a substantial amount of work involved in this. Normal ranges for all of these parameters for Holstein cows in TMR feeding management systems have been published in collaboration with my colleagues at Cornell and Miner Institute.

Foss did not have direct models to measure de novo, mixed origin, or preformed fatty acids and they do not have models for milk estimated blood NEFA, chain length and unsaturation. As the interest in this fatty acid information increased some laboratories with Foss equipment tried to calculate the herd management fatty acid metrics from pre-existing Foss fatty models designed for other purposes. I have tried this myself using data from Foss instruments and it does not work. I think this is what the Lancaster lab has been doing and results produced that way are not correct. I have communicated this to Lancaster and other labs doing the same with Foss equipment. I have also told Foss about this. With increasing pressure for this type of measurement, Foss has recently announced (at the North American Lab Managers (NALMA) Meeting) that they will be releasing (maybe November 2018) new mid infrared milk fatty acid models to measure de novo, mixed, and preformed fatty groups directly as is being done on the Delta Instruments. I am aware that a couple of labs have purchased beta test versions of these models from Foss. These models, like all other MIR milk analysis metrics, need to be routinely calibrated with milk samples. Currently, I am not aware of any other milk fatty acid calibration standards for routine calibration of fatty models other than those produced every 4 weeks in my lab at Cornell. We have been producing these samples for 5 years. All Delta instruments measuring farm management milk fatty acids in the field are calibrated with the Cornell calibration samples and at the NALMA meeting, I presented the first comparison of how well Delta Instrument agree from instrument to instrument on milk fatty acid testing from lab to lab for 10 laboratories. No Foss instrument are using calibration samples for fatty acids. I have made Foss aware of the existence of the milk fatty acid calibration samples that are produced at Cornell and available to anyone for purchase.

To my knowledge Foss has not developed models for milk estimated blood NEFA, fatty chain length, or double bonds per fatty acid. The methodology for the development of the milk fatty chain length and double bonds per fatty acid has been published in JDS. At this point the methodology for milk estimated blood NEFA has not been published but the plan is to publish it in the future.

In my opinion, for the good of the dairy industry, Foss needs to develop farm management milk analysis models equivalent to those that I have developed in collaboration with Delta Instruments, if they plan to continue to serve the milk testing farm management sector of the industry.

It has been a great pleasure for me personally to work together with my colleagues at Cornell and Miner Institute, staff at many milk testing laboratories within and outside the US, and field nutritionists and feed suppliers to bring new milk analysis farm management tools to the industry and I thank everyone for the support and encouragement. The work continues and I will be providing an update at the Cornell Nutrition Conference in Syracuse, NY in October 2018!

1 Comment

Dave Barbano on October 11, 2018 at 1:10 am

HI Everyone!
SHORT ANSWER: the models for milk fatty acid testing have not existed on Foss equipment and labs with Foss equipment producing data for these parameters have simply been trying to estimate the denovo, mixed, performed fatty acids and double bonds per fatty from other parameters measured by their instrument. They were doing this with no calibration. This approach does not work. I have tied to do this myself with data from Foss models and could not make it work. I have made laboratories with Foss equipment aware that I can provide milk fatty calibration samples and work with them to evaluate performance of Foss models on their instruments. However, the first step is that Foss needs to develop their own prediction models. Foss is finally starting to respond to this new industry need, but it will still take some time and there will be an ongoing commitment need to by Foss and any other mid infrared milk analyzer companies, as new herd management milk metrics are developed. Development of new metrics is on-going in my laboratory in collaboration with Delta Instruments and Miner Institute. An update will be presented at the Cornell Nutrition Conference in mid October 2018. I suspect that at some point, other researchers may start to work and innovate in this same area because of the interest in the dairy industry.

LONG ANSWER: Foss was the first company to have fatty acid models on an mid infrared milk analyzer but the specific fatty acid models produced by Foss were designed for food processors, not for farm management. Those models don’t seem to have been used much by food processors and have had very little external validation. The development of the first farm management fatty acid models for farm management milk fatty acid testing were done jointly by myself and Delta Instruments. These models measure different groups of fatty acids (de novo, mixed and preformed) and some structural characteristics of milk fatty acids (chain length and double bonds per fatty acid). These parameters are measured directly by the models and not calculated as derived parameters from other fatty acid models running on an MIR instrument. In addition to developing these models jointly with Delta Instruments, we developed milk fatty acid calibration samples that are available for purchase from my laboratory at Cornell University. The farm management milk fatty acid models that were developed jointly by Delta Instruments and my lab at Cornell are available on Delta Instruments and not Foss Instruments. The information of how we developed the models is published in a series of papers in Journal of Dairy Science and Foss or any other mid infrared milk analyzer manufacturer could develop their own models, but there is cost and a substantial amount of work involved in this. Normal ranges for all of these parameters for Holstein cows in TMR feeding management systems have been published in collaboration with my colleagues at Cornell and Miner Institute.

Foss did not have direct models to measure de novo, mixed origin, or preformed fatty acids and they do not have models for milk estimated blood NEFA, chain length and unsaturation. As the interest in this fatty acid information increased some laboratories with Foss equipment tried to calculate the herd management fatty acid metrics from pre-existing Foss fatty models designed for other purposes. I have tried this myself using data from Foss instruments and it does not work. I think this is what the Lancaster lab has been doing and results produced that way are not correct. I have communicated this to Lancaster and other labs doing the same with Foss equipment. I have also told Foss about this. With increasing pressure for this type of measurement, Foss has recently announced (at the North American Lab Managers (NALMA) Meeting) that they will be releasing (maybe November 2018) new mid infrared milk fatty acid models to measure de novo, mixed, and preformed fatty groups directly as is being done on the Delta Instruments. I am aware that a couple of labs have purchased beta test versions of these models from Foss. These models, like all other MIR milk analysis metrics, need to be routinely calibrated with milk samples. Currently, I am not aware of any other milk fatty acid calibration standards for routine calibration of fatty models other than those produced every 4 weeks in my lab at Cornell. We have been producing these samples for 5 years. All Delta instruments measuring farm management milk fatty acids in the field are calibrated with the Cornell calibration samples and at the NALMA meeting, I presented the first comparison of how well Delta Instrument agree from instrument to instrument on milk fatty acid testing from lab to lab for 10 laboratories. No Foss instrument are using calibration samples for fatty acids. I have made Foss aware of the existence of the milk fatty acid calibration samples that are produced at Cornell and available to anyone for purchase.

To my knowledge Foss has not developed models for milk estimated blood NEFA, fatty chain length, or double bonds per fatty acid. The methodology for the development of the milk fatty chain length and double bonds per fatty acid has been published in JDS. At this point the methodology for milk estimated blood NEFA has not been published but the plan is to publish it in the future.

In my opinion, for the good of the dairy industry, Foss needs to develop farm management milk analysis models equivalent to those that I have developed in collaboration with Delta Instruments, if they plan to continue to serve the milk testing farm management sector of the industry.

It has been a great pleasure for me personally to work together with my colleagues at Cornell and Miner Institute, staff at many milk testing laboratories within and outside the US, and field nutritionists and feed suppliers to bring new milk analysis farm management tools to the industry and I thank everyone for the support and encouragement. The work continues and I will be providing an update at the Cornell Nutrition Conference in Syracuse, NY in October 2018!