The same or even better milk performance with non-GMO rations?

Amino acid (AA) balancing continues to be more widely accepted. Reasons include its very positive impact on transitional cow health and early lactation performance, the desire to feed lower protein diets, high prices for protein supplements, continued refinement and improvement of nutrition models and increased availability of rumen-protected amino acids (Schwab, 2015). As Non-GMO labels are on the rise in Europe dairy farmers are forced more and more to offer milk from non-GMO feed which means to replace GM soy products (high in lysine) by non GM alternatives (very much lower in lysine) such as rape seed, peas, beans, lupins and sunflower. Amino acid formulation is a great tool to follow this demand. It helps to keep the required levels of digestible lysine and methionine and ensures the same or even better performance with non-GMO rations.

Two different ways of application
Nutritional models are now available to predict supplies of AA for dairy cows and methionine (Met) is accepted as the most limiting AA. Two different AA strategies can be applied; On Top application or Reformulation. Rumen-protected-AA (RPAA) supplements are added on top of the feed formulation to improve animal performance. This strategy is usually applied with higher milk prices. In Reformulation, RP-AA are utilized to reformulate the diets to obtain the same or better animal performance at lower rumen undegradable protein (RUP) and CP levels with a less expensive raw material mix.

Practical cases and results
The following examples represent actual on-farm cases for both the above mentioned strategies, and are meant to give some practical cases for nutritionists to examine. Every diet was developed by a consultant or nutritionist taking into account local availability of raw material and prices. Moreover, farm characteristics vary significantly not only within a country but between countries, and therefore responses to diet changes are different. As the data relates to commercial farms, it provides a good idea of the impact of AA formulation in dairy diets.

On Top application on German farm
A farm in North-West Germany has 75 Holstein-Friesian dairy cows. Average milk production is 11,714 kg milk with 3.89 % fat and 3.30 % protein. The farm is equipped with a milking robot that allowed individual administration of a specific mix. The feeding system of the farm follows a partial mixed ration (PMR) with two types of compound feed fed in the milking robot station according to stage of lactation and milk yield. The PMR consisted of a combination of grass and corn silage, 3 kg/cow of protein concentrate and 200 g/ cow of mineral feed. A rumen protected methionine (RP-Met) was included at a daily rate of 15 g/cow in the mineral feed for all lactating cows fed in the milking robot. The feeding of RP-Met started in September 2013, after one month of regular milk recording, until April 2014. The test followed the OFF-ON approach. Results from the OFF period were taken from the last milk recording without RP-Met in September 2013. The ON period refers to the following months with the inclusion of RP-Met.

Maximum effect after three months
Feeding of 15 g of the rumen protected methionine in an On Top application increased milk performance of early and middle lactation. The maximum effect occurred after three months of treatment. After a few weeks of treatment, the early lactation cows react with more milk yield, whereas middle lactation cows react with more milk protein content and with more milk yield later. The delayed response from middle and late lactation cows can be explained by better milk curve persistency of the cows that move from fresh to later stages of lactation during the RP-Met feeding period. Under these conditions it is recommended to feed RP-Met for at least three months to observe the maximum effects.

Reformulation on a French and Spanish farm
On a French dairy farm, a typical diet of the area is 15 kg DM/cow/day of corn silage, 2 kg DM/cow/day of grass silage and 3.5 kg of a protein concentrate. Reformulation of the protein concentrate was done using the SmartMILK matrix. The RP-Met source used is a product known for its stability during pelleting. The reformulated protein concentrate had a cost of 314€ /tonne compared to the original at 327€ /tonne. Under the market conditions at the time of the trial, the reformulation approach gave savings of at least 13 € /T in the protein concentrate. Experience at field level over the past four years, indicates the reformulated formula has become very competitive in the market and also helped to reduce N load in the environment while improving milk quality by about 0.3 g/kg of milk protein content. In Spain, the situation was studied on an association of farms located in the North-West part of the country. All these farms had more than 800 dairy cows and the main common characteristic was that they fed the same total mixed ration (TMR). The TMR was produced by a feeding centre that took care of raw material purchasing and silage making. On average the cows produced 10,000 kg milk/year with a milk composition of 3.65% fat and 3.25% protein. The objectives for the reformulation exercise were to maintain milk production during the summer heat and reduce feed costs as at that moment prices for soybean meal were at a record high.

Reduction in feed costs
In the reformulated TMR, soybean meal was significantly reduced and other alternative raw materials, such as rapeseed meal and DDGS wheat, were included. A RP-Met source was used. The crude protein level of the diet declined by 1.1% units and the LysDi/MetDi relationship improved from 3.77:1 to 3.06:1 by keeping grams of LysDi constant and improving the supply of MetDi. The first direct benefit was the reduction in the feeding cost by 84 Eurocents per cow per day. Even though improved milk performance is not one of the key objectives when reformulating diets, in this case, the average milk production went up by 2kg and milk protein % and yield also increased. This resulted in a direct improvement on N efficiency by nearly 3 points.
References are available on request