Nutrient Digestibility in Horses Using Numerous Diets

Many equine nutrient requirements are based on certain assumptions about how well horses digest and absorb nutrients. Digestibility varies somewhat by source, and there are also interactions between certain nutrients that affect how well each is digested. Kentucky Equine Research (KER) has conducted a number of digestibility trials to evaluate different types of feeds and feed ingredients for horses.

KER uses a standardized experimental design to measure nutrient digestibility in horses. In this design, four mature horses are fed different experimental diets in a Latin square design (either 2×2 replicated or 4×4) over four collection periods. Each period consists of a three-week adjustment period followed by a five-day total fecal collection. The week before and during the five-day collection period, the horses wear a harness with specially designed pouches that allow the complete and separate collection of urine and feces. During the collection period, daily feed intake and total fecal output are measured. Subsamples of daily feed and feces are taken and frozen. These subsamples are dried and composited for chemical analysis. Both feed and feces are analyzed for dry matter, crude protein, ADF, NDF, crude fiber, fat, calcium, phosphorus, magnesium, potassium, zinc, copper, manganese, and ash. Based on the results, digestibility is calculated for each nutrient measured.

For the purpose of this review, 30 different diets and 120 observations have been evaluated using this design. These diets have ranged from pure alfalfa hay to a combination of sweet feed and fescue hay to pelleted concentrates fed with timothy hay. The majority of horses used in these studies have been Thoroughbreds weighing between 1,100 and 1,320 lb (500-600 kg). Because of the tight range of body weights of the experimental horses, all data are presented as total daily intakes rather than as a function of body weight. Each experiment was conducted for a specific purpose, such as to evaluate the effect of adding distillers dried grains to horse feeds or to determine the effect that pelleting has on alfalfa hay, but for the purpose of the present study, all data have been combined into one data base. While a variety of different feed ingredients were used in these studies, these data are only representative of forages and concentrates from a temperate environment. Tropical forages may behave differently because they contain inhibitory substances not found in temperate forages.

Nutrient digestibility can be expressed in two different ways. One way is as apparent digestibility. Using this calculation, the amount of a specific nutrient that is recovered in the feces is subtracted from the total daily intake of that nutrient. The amount that disappeared (intake minus feces) is divided by the total daily intake to produce a percentage of intake.

Apparent digestibility is a fairly crude way to evaluate digestibility because it only measures the total amount of a particular nutrient in the feces. There are two possible sources of these fecal nutrients. Some of the nutrient could be the undigested residue left from the feed, but some may have actually been excreted into the digestive tract from the horse’s system or it might have sloughed off the intestinal wall. The fecal substances that originate from inside the horse are considered endogenous in nature, and they result in an underestimation of true nutrient digestibility. To overcome the interference of endogenous losses in the determination of digestibility, a statistical procedure called a Lucas test can be utilized. In this test, a range of nutrient intakes is considered. The amount of nutrient that is digested is regressed against its corresponding level of intake. If there are real endogenous losses associated with a particular nutrient, then the calculated level of nutrient digested at a nutrient intake of zero will be a negative number.

Because there are endogenous fecal losses for most of the minerals measured, true digestibility is quite different from apparent digestibility. Calcium and potassium are readily digested from most equine diets, with estimated true digestibilities of about 75%. Magnesium true digestibility equals about 52%, while the estimated true digestibility of phosphorus is considerably lower (25%). The true digestibilities of the trace elements zinc, manganese and copper average 20.8%, 28.5% and 40.0%, respectively. The sources of trace minerals used in these diets were either from the naturally occurring mineral in the feed ingredients or added inorganic sources.

The true digestibility of protein in these various diets averaged 71%. Most of the diets studied had protein digestibilities close to this level with the exception of a high-quality alfalfa hay from Arizona that averaged a protein digestibility of around 80%. Regression analysis of fiber digestibility showed that these components did not produce endogenous losses. Estimated true digestibilities of ADF, NDF, and hemicellulose were 40.2%, 45.0%, and 54.2%, respectively. Interestingly, low fiber intakes resulted in lower fiber digestibility than when high levels of fiber were being fed. Perhaps the fiber from grains (which represented a greater proportion of total at the low fiber intakes) was less digestible than the fiber from forages.

The energy content of the various rations was calculated as percent total digestible nutrients (TDN %). TDN was calculated as: digestible crude protein + (digestible crude fat x 2.25) + digestible neutral detergent fiber + digestible soluble carbohydrate. Soluble carbohydrate (CHO) = 100 – crude protein (CP) – ether extract (EE) – neutral detergent fiber (NDF) – ash. TDN averaged 61.4% for all of the diets measured with a range from 46.9% to 61.0%.

TDN and digestible energy (DE) are useful values to calculate for a horse ration. Because these values are by nature biological assays, it would be impossible to directly determine these values for every feed. From the present study, regression equations have been developed that allow the calculation of TDN (or DE) from analyses routinely performed on horse feed and forage. TDN and DE are really interchangeable units of digestible energy since the heat of combustion of protein and carbohydrate are similar and the heat of combustion of fat is already accounted for in the equation used to calculate TDN. Therefore, to calculate DE as kcal/kg, simply multiply TDN % x 4,400. To find DE as kcal/lb, multiply TDN x 2,000.

The most complete equation to estimate DE is: DE (kcal/kg DM) = 2,118 + 12.18 (CP %) -9.37 (ADF %) – 3.83 (hemicelluloses %) + 47.18 (fat %) + 20.35 (NSC) – 26.3 (Ash %) (R2=0.88). Notice that as protein, fat and soluble CHO increase, DE increases. As fiber and ash increase, DE decreases. This agrees well with what is known about the energy density and digestibility of these various components in horse feeds. It should also be remembered that these equations were developed with either all-forage diets or mixtures of forage and concentrate. These equations tend to overestimate the energy density of straight grains, possibly because of the differences in fiber digestibility mentioned above.

The most recently revised Nutrient Requirements of Horses, published in 2007 by the National Research Council (NRC), based a number of its nutrition recommendations on the results of research conducted by KER.

This article is based on a paper in Advances in Equine Nutrition published by Kentucky Equine Research.