AFBI Study Reveals Surprising Boost in Milk Yield with Early Lactation Cows from Sustainable Cow Diets

Date published: 10 November 2025

A unique study carried out at AFBI, Hillsborough has uncovered that dairy cows offered concentrates based on by-product feedstuffs (low human-edible ingredients) produced more energy-corrected milk over the first 150 days of their lactation than those on conventional concentrates that included cereals and soyabean meal (high human-edible ingredients).

The use of these alternative feeds reduced the carbon footprint, when soyabean meal was excluded, but removing human-edible ingredients did raise challenges due to their higher phosphorus concentration. This article delves into the findings and their implications for sustainable dairy farming.

Background

Dairy cows have the unique ability to convert fibrous human inedible forages (grass and grass silage) into high-value, nutrient dense milk suitable for human consumption. However, higher yielding cows are normally fed significant quantities of concentrates to support milk production and these concentrates often contain ingredients, such as wheat and maize, which could also be consumed by humans. The use of these “human-edible” ingredients in livestock diets is being increasingly challenged as we seek to feed a growing global population. Furthermore, there are growing concerns about the sustainability of some feed ingredients used in dairy rations, for example, soyabean meal.

Given these challenges it makes sense to consider using ingredients in dairy rations which have a low human-edible content such as rapeseed meal and distillers grains and to reduce the use of cereals and soyabean meal. Results from a series of short-term feeding studies indicated that reducing the inclusion of human-edible ingredients in dairy cow concentrates had little impact on the performance of mid-late lactation dairy cows. However, it might be expected that there could be a negative impact associated with using more by-product ingredients in the diet of cows over a longer feeding period especially during the early stages of lactation. This was examined in a recent study within the DAERA funded SusMilk project which examined the effects of reducing the levels of human-edible feedstuffs (cereals and soyabean meal) in the diets of early lactation dairy cows. Specifically, the study sought to examine if the effects were influenced more by the removal of cereals grains or soyabean meal. 

Research study

The experiment involved 88 autumn calving cows with performance examined from calving through to 150 days-in-milk. Cows were offered a mixture of grass silage and concentrates (50:50 mix on a dry matter (DM) basis) with four types of concentrate being offered. The treatments included:

• High human-edible concentrate (included cereals and soyabean meal)
• Concentrate with no soyabean meal
• Concentrate with no cereals
• Concentrate with no cereals or soyabean meal

The cereal component of the concentrate was replaced with by-product ingredients which included sugar beet pulp and wheat feed, while soyabean meal was replaced with feed ingredients like rapeseed meal, distillers grains and maize gluten. All four concentrates were formulated to have similar metabolizable energy (12.4 MJ/kg DM) and crude protein (155 g/kg DM) concentrations. Daily dry matter (DM) intake, milk yields, milk composition and edible feed conversion rate (eFCR) were measured during the experiment. eFCR was defined as the quantity of human-edible output in milk divided by the amount of human-edible feed used to generate it.  The carbon footprint and phosphorus concentration of the treatment concentrates were also assessed to determine the potential environment impact of offering these feeds.

Key results from the study are summarised in Table 1. Replacing the cereal and soyabean meal component of concentrates with by-product ingredients led to a decrease in the human-edible fraction of the concentrate. For example, the concentrate based on cereal grains and soyabean meal had a human-edible fraction of 51%, while removing both cereals and soyabean meal produced a concentrate with a human-edible fraction of 18%. Removing cereals from the concentrate had a greater impact on human-edible fraction (25%) than removing soyabean meal (42%). Cows offered concentrates with a reduced human-edible fraction (i.e. diets with low cereals and soyabean meal and high level of by-products) had a higher total feed intake, energy corrected milk yield and fat plus protein yield. Removing the cereal component from the concentrate led to a reduction in milk protein concentration but had no effect on milk fat content. Reducing the human-edible fraction of the concentrate led to an increase in eFCR, where cows offered the concentrate with no cereals and soyabean meal produced three times more human-edible product (milk) than the amount of human-edible feed they consumed to produce it. It is also worth noting that the eFCR values recorded across all the treatments were greater than 1.0, which highlighted the ability of the dairy cow to convert human inedible or low human-edible feedstuffs into a nutritious human food. Cows offered the low human-edible concentrate treatments also tended to have a higher Margin Over Concentrate (MOC) value than cows offered the high human-edible concentrate. Reducing the human-edible fraction of the concentrate also reduced the carbon footprint, particularly when the soyabean meal portion was replaced with by-product ingredients. For example, removing the soyabean meal component from the current formulations reduced carbon footprint by up to 55 %. However, it should be noted that the figures were generated on the assumption that the soyabean meal used was sourced from a deforested region. If the soybean meal were sourced from a non deforested area then this would have a significant impact on the carbon footprint of the final concentrate. For example, if soyabean meal from a non deforested area were included in the high human-edible concentrate formulation, then the resulting carbon footprint would change from 1,304 to 667 kgCO2e/ton. 

It is also important to note that reductions in concentrate human-edible fraction with the current formulations also led to an increase in phosphorus concentration of the concentrate by up to 75 % and this has the potential to lead to an increase in farm phosphorus balance, which could have a negative effect on water quality if not managed. The results of this study highlights the need to carry out additional work to investigate the effects of feeding low human-edible concentrates to dairy cows on phosphorus utilisation and indeed determine how cows utilise these feeds to produce more milk than that observed by feeding conventional based concentrates.

Summary

This study provides clear evidence that the performance of early lactation cows can be enhanced through the use of concentrates formulated with low levels of human-edible ingredients. Notably, removing soyabean meal from these diets not only improved milk yield but also significantly reduced the carbon footprint – highlighting a promising pathway toward more sustainable dairy production. However, the elevated phosphorus content in these alternative rations raises important questions about the need to consider the wider impacts of changing diets on the potential environmental impact. As global pressure mounts to reduce reliance on human-edible feed sources, this research lays a critical foundation. Further studies are essential to understand the implications for phosphorus utilisation and to ensure that future feeding strategies are profitable, environmentally responsible and scientifically sound.

Table 1. Impact of reducing the human-edible fraction of concentrates on the performance of early lactation dairy cows

¹ eFCR = edible food conversion rate; ² Figures based on milk price of 38 ppl +/- changes in milk quality. Concentrates costs based on feed prices for October 2025; ³ Values taken from Global Feedstuff Lifetime Assessment Institute Database