MILK PRODUCTION RESPONSES AND RFI ON HIGH AN LOW STARCH DIETS

Discipline: composition; Key words: diet starch concentration, feed efficiency, variation in treatment response, energy partitioning, residual feed intake.

Although it is well known that energy requirements for milk yield to a large extent determine the response of cows to increasing energy content of the diet, in other words, high milk producers will respond proportionally more by increased yields than low producers if the concentrate level of the diet is increased, it remains uncertain whether the actual composition will always result in this generalization and what would be the effect on milk components. A further question is whether the ranking in feed efficiency as measured by residual feed intake will remain if the dietary energy concentration is changed. This obviously is important, since if it is indeed the case, it means that once cows have been selected on this basis, the diet they would be introduced to largely becomes irrelevant – the better ones on a particular diet will remain the better ones on another diet. There is limited evidence that this could be the case. Authors Boerman, Potts and others investigated these two questions in a comprehensive investigation by varying the starch level in the diet. Their results were published in the Journal of Dairy Science, Volume 98 of 2015, respectively on pages: 4698 to 4706 and 4735 to 4747. The titles of the papers are respectively: Milk production responses to a change in dietary starch concentration vary by production level in dairy cattle and Residual feed intake is repeatable for lactating Holstein dairy cows fed high and low starch diets.

In the trial to investigate the effects of dietary starch concentration on milk yield and milk components, a crossover design was used with 32 Holstein cows at about 115 days in milk. The cows differed vastly in milk yield, from 28 to 62 kg per day. They were assigned randomly within level of milk yield to a treatment sequence where the variable components were diets with either 30% dry ground maize (high starch) or 30% soyhulls (low starch) on a dry matter basis. Other components included maize and lucerne silage and the diets were formulated to contain 16% crude protein, 24% roughage neutral detergent fiber, and either 27 or 44% total neutral detergent fiber and 30 or 12% starch for the ground maize and soyhull diets, respectively. In a preliminary period of 14 days, the cows were fed a diet intermediate to the treatments. Treatment periods were 28 days with measurements taken throughout the period for energy calculations and the final five days used for data and sample collection for production variables.

Compared with soyhulls, ground maize increased dry matter intake and the yields of milk, milk protein, milk fat and energy-corrected milk, as well as milk protein concentration. However, treatment did not affect milk fat concentration, but milk fatty acid yield increased with ground maize. The ground maize treatment also increased energy-corrected milk in higher-producing cows with a lesser response in lower producing cows. Furthermore, the ground maize treatment increased milk:feed (feed efficiency) compared with the soyhull treatment, but not body weight or body condition score. In summary, the higher-producing cows benefited from the high-starch diet and the lower-producing cows not really. However, the lower-producing cows were able to maintain production when most of the starch was replaced with non-roughage fiber, implying under South African conditions the fiber of millers by-products such as hominy chop, wheaten bran, maize gluten etc.

In the second trial, the effect of starch content on the ranking of cows in residual feed intake (RFI) was investigated. RFI is a tool to quantify feed efficiency in livestock and is commonly used to assess feed efficiency independent of production level, body weight, or body weight change. A total of 44 first lactation and 65 multi-lactation lactating Holstein cows, averaging 665kg in body weight, 42kg in milk per day and at 120 days after calving, were fed diets of high or low starch content in four crossover experiments with two 28-day treatment periods. The diet compositions were as in the first experiment, with the low starch diets containing about 40% total neutral detergent fiber (NDF) and 14% starch and the high starch diets about 26% total neutral detergent fiber and 30% starch. Individual dry matter intake (and as a result RFI) of a cow was modelled as a function of milk energy output, metabolic body weight and body energy change. Cows were classified as high, medium, or low RFI.

On average, for the linear model used to determine RFI for individual cows, each unit increase in milk energy output, metabolic body weight and body energy gain was associated with respectively an increase of 0.35, 0.09 and 0.05 kg in dry matter intake. When compared with low starch diets, high starch diets increased energy partitioning to body energy gain and tended to increase dry matter intake. The correlation (r) between RFI when cows were fed high starch diets and RFI when cows were fed low starch diets was 0.73 and was similar across each parity and experiment. Fifty-six percent of cows maintained the same RFI classification (high, medium, or low RFI) and only 4 of 109 cows changed from high RFI to low RFI or vice versa when diets were changed. Milk:feed (feed efficiency), income over feed cost and dry matter intake were also highly repeatable (r = 0.72, 0.84, and 0.92, respectively). Significant changes in milk yield and milk component concentration as well as energy partitioning between high starch and low starch diets were observed and still RFI was repeatable across diets. The authors concluded that RFI is reasonably repeatable for a wide range of dietary starch levels fed to mid-lactation cows, so that cows that have low RFI when fed high maize diets will likely also have low RFI when fed diets high in non-roughage fiber sources. In the South African context non-roughage fiber is obtained from millers by-products as mentioned above. The implications of the study are that, although RFI is not absolutely repeatable as the correlations above show, selection by RFI within TMR systems should be fairly reliable whether performance is required on high concentrate (high starch) or low concentrate (low starch) diets. At this stage though, we do not know whether the repeatability will carry from TMR to pasture-based diets as factors such as palatability, herbage height, grazing time etc affecting dry matter intake enter the equation. Thus, one should rather stick to one system in selection or purchasing cows based on RFI results.