Nitrogen efficiency in dairy cows: what decades of protein research reveal

Why protein feeding needed rethinking#

Harrison and Karnezos describe ruminant protein nutrition as an area that has produced a great deal of research over recent decades. Their starting point is a problem familiar to anyone who has balanced a dairy ration: diets formulated purely on crude protein content, without regard to protein quality or how much of it survives rumen fermentation, tended toward overfeeding of nitrogen and production that fell short of its potential.

The authors look back at two 1978 textbooks to show how far thinking has shifted. Ensminger and Olentine argued that for heifers, dry cows and lower producers, amino acid quality barely mattered because rumen microbes could manufacture the essential amino acids the cow needed. High producing cows were the exception, since bacterial protein synthesis alone might not keep pace, which is the logic behind rumen undegraded protein and bypass protein feeding. They recommended 14% crude protein for most cows, with their chief worry about overfeeding being cost rather than biology. Kutches, writing the same year, recommended protein levels graded to milk output, from 14 to 15% crude protein below 44 lbs of milk up to 16 to 17% above 66 lbs.

These early frameworks treated protein almost as a single commodity. What changed over subsequent decades, as the paper documents, was recognition that rumen microbes and the cow herself have separate and sometimes competing nitrogen requirements, and that meeting both well is the real route to nitrogen efficiency in dairy cows.

From crude protein to metabolizable protein models#

The 1989 revision of the Nutrient Requirements for Dairy Cattle marked what the authors call a major advance. It split dietary protein into undegraded intake protein and degraded intake protein, and defined the animal's own requirement as absorbed protein reaching the small intestine. This factorial approach distinguished three separate outcomes for dietary protein, covering rumen breakdown, later digestion further along the gut, and whatever passed through undigested. Its weaknesses included a fixed intestinal digestibility value for undegraded protein of 80%, with no allowance for endogenous protein contributions or the actual amino acid makeup of what passed through.

During the 1990s, Cornell University researchers built the Cornell Net Carbohydrate Protein System, a tool that modelled digestion and nutrient metabolism by comparing rates of carbohydrate fermentation against rates of protein breakdown, predicting microbial protein synthesis, ammonia production and the flow of undigested protein to the small intestine.

Santos and co-workers reviewed 88 lactation trials published between 1985 and 1997 and found that swapping soybean meal for protein sources with high rumen undegradable protein content only lifted milk yield in 17% of the comparisons. They concluded the essential amino acid content of these replacement sources was the critical factor, and that the Cornell system tended to overvalue protein supplements rich in undegradable protein. The 2001 edition of the Nutrient Requirements for Dairy Cattle responded by replacing absorbed protein with metabolizable protein, swapping the older degradability labels for the RDP and RUP terms already in use elsewhere, and predicting microbial protein flow from digestible organic matter intake rather than net energy.

How rumen microbes use nitrogen#

Rumen bacteria draw on peptides, free amino acids and ammonia, while protozoa cannot use ammonia at all. Wallace found that ammonia's contribution to microbial nitrogen could range from a minimum of 26%, when peptides and amino acids are abundant, up to a theoretical maximum of 100% when ammonia is the only nitrogen source on offer. Earlier work by Nolan, and by Leng and Nolan, had put the ammonia contribution at 50% or more under typical conditions.

The optimal rumen ammonia concentration for maximising microbial protein yield has itself been argued over for years. Satter and Slyter's widely cited fermentor studies put the figure at 2 to 5 mg/dL, while other in vivo work found microbial protein production kept climbing until ammonia reached 10 mg/dL or even 29 mg/dL. Oba and Allen later reported no clear relationship between rumen ammonia level and efficiency of microbial protein synthesis at all, underlining how unsettled this question remains.

Carbohydrate type also matters. Microbes that ferment non-structural carbohydrates derive roughly two-thirds of their nitrogen from amino acids or peptides, while fibre-digesting bacteria can meet their nitrogen needs entirely from ammonia. Jones and colleagues found that diets high in non-structural carbohydrate paired with excess peptide could actually depress protein digestion and slow fibre-digesting microbes, a reminder that more peptide supply is not automatically better for the whole rumen population.

Testing the synchrony idea#

One appealing idea explored at length in the paper is synchronisation: matching the rate at which nitrogen becomes available in the rumen with the rate at which carbohydrate is fermented, on the theory that this pairing should make microbial protein production more energetically efficient. Russell and Hespell argued that failing to match these rates would waste substrate and reduce microbial growth efficiency.

In practice the results have been mixed. Aldrich and colleagues found that matching degradability of rumen protein and carbohydrate improved bacterial nitrogen production efficiency, yet this did not translate into measurable gains in milk output. Kolver and colleagues found that synchronising carbohydrate release with nitrogen from pasture improved capture of rumen nitrogen but left nitrogen status and milk production unchanged. Altering feeding frequency to achieve synchrony, tested by Robinson and McQueen, produced no benefit either. Henning and colleagues, working with rumen infusions in sheep, likewise found no improvement in microbial yield from synchronised nitrogen and carbohydrate supply.

Hoover and Stokes, reviewing both in vitro and in vivo literature, concluded that the rate of carbohydrate digestion was the dominant factor governing microbial growth, and recommended that nutritionists first secure a stable rumen energy supply before fine-tuning available nitrogen. Stokes and colleagues demonstrated that when adequate carbohydrate and rumen degradable protein were both ensured, in diets containing 31 to 39% non-structural carbohydrate and 11.8 to 13.7% rumen degradable protein, microbial protein flow to the duodenum increased, reducing the amount of undegradable protein and energy needed from other ingredients.

What excess protein costs the cow and the paddock#

Feeding protein above requirement does not just cost money at the mill. Broderick found that dairy cows excrete several times more nitrogen in manure than they capture in milk, which both inflates feed costs and adds nitrogen to the surrounding environment. Elevated blood urea nitrogen and milk urea nitrogen concentrations have each separately been connected with poorer reproductive outcomes in the studies cited, and Sinclair and colleagues associated high rumen degradable protein with altered ovarian and uterine physiology when plasma urea nitrogen exceeded 190 mg/L. Oldick and Ferkins, however, found elevated blood urea nitrogen did not always depress reproduction, so the relationship is not absolute.

Researchers working with high-producing cows on typical mid-western US forages tested this trade-off directly. Broderick formulated diets at 15.1, 16.7 and 18.4% crude protein and found milk and protein yield did not improve above 16.7%, while nitrogen excretion rose in step with dietary protein and proportionally more of it ended up in urine. A follow-up trial by Olmos Colmenero and Broderick, spanning 13.5 to 19.4% crude protein, found peak milk production at 16.5% crude protein, with no further gain above that level and efficiency of nitrogen use for milk protein falling steadily as dietary protein rose.

Brito and Broderick then compared protein sources at a fixed 16.6% crude protein and found soybean meal, cottonseed meal and canola meal supported similar milk yields above 40 kg, with comparable microbial protein flows, while a urea-based diet produced a smaller flow of microbial protein further along the digestive tract. Together these trials point to roughly 16 to 17% crude protein as a practical ceiling beyond which nitrogen is simply wasted rather than converted into milk.

Slow release nitrogen and nitrogen efficiency in dairy cows#

Alongside dietary protein level, the paper surveys three decades of attempts to control the speed at which non-protein nitrogen releases ammonia in the rumen, so that it better matches carbohydrate fermentation. Several named products are listed, ranging from treated and coated forms of urea to compounds such as biuret, with release rates that have ranged from too slow to too fast to be useful.

One polymer-coated urea product, described in the paper by its commercial name, was shown by Akay and colleagues to break down in the rumen at a pace comparable to soybean meal. Its practical appeal lies partly in nitrogen density: the product's crude protein value of 274% compares with 53% for soybean meal on a dry matter basis, meaning a small amount replaces a much larger quantity of soybean meal and frees up physical space in the ration for other ingredients.

In a 220-cow trial, Akay and colleagues used 227 g of this product to reduce a combination of other protein and fibre ingredients in the ration, filling the resulting 1.1 kg of space with wheat middlings and finely ground corn. This lifted rumen degradable protein by 4.12% and starch by 3.06%, and produced a 9.74% increase in milk yield along with a rise in milk protein yield, despite a 5.5% reduction in undegradable protein. A second trial held crude protein and net energy constant between treatments and instead found a 3.73% reduction in dry matter intake on the polymer-coated urea diet, with milk, protein and fat yields unchanged, yielding a 4.2% improvement in efficiency of milk production.

Reading the evidence with care#

The review itself is built from other people's trials rather than new data of its own, and its strength lies in synthesis: setting Santos and co-workers' critique of bypass protein feeding against the National Research Council's successive models, and setting synchrony theory against its patchy experimental record. We would treat the synchrony findings in particular as a pointer rather than a settled result, since several well-designed trials achieved the intended biochemical match in the rumen without translating it into extra milk, which suggests other factors such as a stable energy supply matter more in practice than precise timing.

The authors' own conclusion is that formulating diets around metabolizable protein and amino acid requirements, rather than crude protein alone, has already delivered better performance and more efficient nitrogen use, and that the next gains will come from optimising microbial protein synthesis itself rather than from further tinkering with undegradable protein supplementation.

This paper was presented at the Recent Advances in Animal Nutrition proceedings, part of the broader RAAN Conference Proceedings held alongside other contributions on ruminant nutrition available through the Livestock Library research index.

Practical implications for ration formulation#

For a nutritionist balancing a dairy ration today, the practical message running through this review is that pushing crude protein higher than around 16 to 17% buys little or no extra milk while adding directly to nitrogen excretion and cost. The trials covering crude protein from 13.5% up to 19.4% consistently found production plateauing once rumen and animal protein needs were met, with any surplus nitrogen simply passing through to urine and faeces.

Choice of protein source matters less than its degradability profile and amino acid balance once crude protein is adequate, since soybean meal, cottonseed meal and canola meal performed similarly in Brito and Broderick's comparison. The authors frame microbial protein itself as the most valuable amino acid source available to the cow, closely matching milk protein's amino acid profile, which is why models now emphasise feeding the rumen microbial population well rather than simply maximising bypass protein.

A separate paper, the Short Communication on cloning of the red kangaroo follicle stimulating hormone beta subunit, published in 1998, shares an author with this review and is also catalogued in this index.

Sources and further reading#

• Trove library search: find a library that holds the paper
• Meat & Livestock Australia: red meat industry research and marketing body
• Feedipedia animal feed database: an open-access database of animal feeds
• Dairy Australia: dairy industry research body

Questions#

— The Livestock Library team