Tuesday, July 5, 2016

Bread for my Horses, IMC Leeds 2016

This is the paper I presented at IMC Leeds.
The current research on horse diets between the end of the Roman Empire and the foundation of Purina Mills in 1894 is very sparse. Most of the available research is focused on military animals, both because of the popularity of medieval and early modern military history in the last century, and the simple availability of records. What is remarkably absent from these works is any discussion (or even mention) of “horsebread,” which was a closely regulated commercial horse feed in England from the mid-thirteenth to the mid-seventeenth centuries. Horsebread also appears to have been common on the continent, and may have remained in regular use as late as the nineteenth century in some areas. It is possible that modern practices of feeding horses bread to “fatten” them have survived, despite decades of veterinary research showing this to be not only ineffectual but potentially lethal, because of the long use of horsebread in Europe. Horses are very efficient at digesting fiber, but have a limited capacity for simple starches. An overload, such as from a large quantity of modern white bread, causes excessive fermentation in the gut. This can lead to colic or laminitis, either of which can be fatal.[1]
The nutritional failures of modern bread for equine diets may be why scholars have been loathe to approach the subject. There are no books that deal with the topic, despite a recent surge in equestrian histories (as todays panels exemplify), and only one article. That article, English Horse-bread, 1590-1800 by William Rubel, provides a hint about why bread, a substance today known to cause colic in horses, was commonly fed for so long. While Rubel doesn’t address this issue, he does detail several recipes from the sixteenth century. These recipes are wildly different from modern bread, and not only due to coarser milling or older strains of grain. The primary difference is that horsebreads were made with some form of ground legume, most often fava beans­– commonly enough that favas were known as horsebeans. Many of the available recipes favor rye, even when horsebread started to be refined for feeding to racehorses. Interestingly, rye has a much better calcium to phosphorous ratio than wheat or even oats. Excess phosphorous in the diet causes a number of disorders, which for a time were referred to as “bran disease” due to their association with feeding wheat bran mashes.[2] When wheat was used for horsebread (and it usually was), it would be sifted and the finer flour would be reserved for more expensive breads for human consumption. The finer flour contains more simple starch, while the coarser (which includes most of the bran) contains more fiber and more protein, all of which is better for the horse is the calcium deficiency is otherwise addressed.  These horsebreads would not only have been “a denser source of calories and protein,” as Rubel points out, but also have a better vitamin and mineral balance, and (critically) be more digestible.
One of Rubel’s arguments, despite his focus on what he calls “elite” horsebreads, is that horsebreads were made from the byproducts of human food productions, and as such were very cheap. Though this is superficially true, this is what I would like to challenge. Even in the late medieval period, when horsebreads were used to feed the poor and their livestock alike, these breads provided a shockingly balanced diet. For livestock, economy alone would suggest feeding these by-products as mashes rather than taking the time to bake them. Even the peas are still found, dried whole, in commercial grain mixes for horses. Due to the nutritional soundness of horsebreads, I believe the ingredients were chosen equally for their effectiveness, even before the development of the bloodhorse (the early Thoroughbred). In some cases, especially before this, as racehorses were elite animals and needed to be fed in elite ways. Brown bread was common, and unfit for consumption by the aristocracy. This sentiment extended to their horses (and possibly even their dogs) so that horsebreads fed to racehorses needed to look different than horsebreads fed to peasant’s animals.
Given that horsebread seems to have become common by the middle of the thirteenth century, it seems likely that this innovation in livestock feeding occurred as a result of two factors. The first is the increased regulation on breads in general, being a major taxable item (there is a paper on this in session 824 this afternoon). The second is the fairly recent proliferation of farm horses, due to the horse collar and the heavy plough. This first would make horsebread production valuable to anyone in position to collect taxes. The second means horses were suddenly in longer, harder work, which requires more bioavailable nutrition, because of the need for meal density without causing any of the assorted digestive issues horses are prone to.
What is in Horsebread?
An ongoing study by Pennsylvania State University has “revealed surprising similarities in the statistics for bread consumption across class and geographic boundaries, confirming the centrality of this staple food in [all medieval diets]…differences of social status and era present themselves in the composition of that bread.”[3] In this light I would like to provide an overview of the “composition” of horsebreads, their change over time, and how each ingredient fit in to human social ideology.
My apologies, this section is rather dense, but is necessary to understanding the critical nutrition differences between horsebread and other breads. I look for three main nutritional factors: protein, necessary for heavy work. Fiber, which is critical to equine digestion. And the calcium to phosphorous ratio, which for a horses total diet needs to be about 1:1. All grains, including in todays diets, have excessive phosphorous.

First the Legumes (Pulses), which are the primary difference between horsebreads and other coarse breads.
            Fava beans
            Also called faba beans, broad beans, or horse beans. They are high in fiber and have an approximately 1:3.5 calcium to phosphorous ratio.[4] Fava beans seem to become more popular in livestock feed (cattle as well as horses) and less popular on human tables towards the end of the medieval period, but more research is needed to confirm this trend and, if it is confirmed, to consider possible reasons. While fava beans are higher in fiber than fine wheat flour, they are lower in protein, a nutritionally deficiency that may have helped lead to the downfall of horsebread after the ‘inferior’ pea fell out of fashion.
            Peas usually meant peas, but not always. Peas could also refer to other legumes such as kidney beans, and sometimes what we call peas would generically be termed pulses. However, Thomas de Grey’s 1639 condescending description of how horsebread of the common sort used to be made provided this description: “loaves or Rouls are rowled in ground or rather bruiſed peaſe.”[5] Bruised peas will stick very well to the outside of a loaf, while bruised kidney beans would only make a mess. Others refer to loaves being rolled in spelted peas. Spelting, a process of shelling and splitting, is used for both peas and fava beans, but not applicable to what we call beans. Peas have about a 1:4 calcium to phosphorous ratio and almost twice the protein density of wheat.

And the Grains
    Wheat has, of course, long been a staple of bread making, due both to its long domestication and its higher gluten content compared to other grains. When horsebreads first began to be used, if wheat flour was used at all it was the bran or the chisel (a coarse, mostly bran flour). These were effectively the by-products of bread making, and as such ideal for making lower class breads, including both peasant breads and horse breads. Usually rye was planted over the winter, but in the late middle ages late planting of wheat became more common.[6] It is unclear if this was due to increased demand, improvements in agricultural technology making soft wheat less risky, or the importation of a hard wheat variety.[7] Wheat was not commonly used for livestock feed because of its value in human diets. Whole wheat also has a very problematic calcium to phosphorous ratio of 1:10.[8] Wheat bran is slightly worse as 1:13, and in both cases much of the calcium is not absorbable.[9]
            Next, Rye
       Rye was common in northern Europe because of its ability to withstand cold, damp climates better than wheat. Rubel discusses a number of literary references to rye bread, or dark breads, being suitable for animal feed but not for a respected human.[10] Thomas de Grey’s accounts of “ordinary horsebread” says they used “bran and chiſel for the moſt part with a little courſe Rye-meal, to make it ick together.” Bran and chisel have a lower gluten content than finer flour made from the endosperm, and thus the rye was added to “make it stick together.” Rye flour was less expensive than wheat, due to it being a heartier crop with a lower sowing ratio than wheat, as well as rye being disdained by the aristocracy. Rye grain was (and is) not generally used for livestock feed, but the left over stalks and chaff were used as forage. Like bran, these were effectively byproducts and that significantly reduced the cost. The use of rye hay would also help balance high phosphorous grain diets: rye hay has a calcium to phosphorous ratio of about 2:1.[11] Though better than wheat, Rye grain is as imbalanced as the rest at 1 to about eight & a half, though Rye’s calcium is more bioavailable than wheat’s.[12]
       The use of oats in medieval England was highly regional, based on local climate and economy. Wetter, less populated areas (which tended to grow more livestock) planted more oats.[14] They were not likely commonly used in early horsebreads, though both horsebreads and peasant breads were known as “breads of many grains” and could legally contain oats.[15] However, they did become more common in elite horse breads by the seventeenth century. This may simply be because of the commonality of feeding oats to horses. Oats, as a whole grain, are the safest thing to feed horses due to their fibrous outer hull, which slows digestions and prevents disorders caused by sudden large amounts of starch or masses of food in the hindgut.[16] Oats are also supremely palatable, and oat hay (like rye hay) can be used as forage. Oats have a calcium to phosphorous ratio of about 1:4.25.[17]
       Barley was the most common grain crop in pre-conquest England, likely a holdover from Roman times, but was largely supplanted by wheat.[18] Barley was still used in field rotations, and was valued above rye and below wheat. Although fairly nutritious and very traditional for livestock, the primary use for barley was in brewing beer. Like oats, barley could legally be used in peasant breads and horsebreads, but this would be based on locally available surplus. Barley has a calcium to phosphorous ratio of a bit better than 1:7.[19]
      For the medieval period, the only other likely ingredients were water and ale barm, a brewing byproduct that would provide yeast. When horsebreads later were altered for feeding to early thoroughbreds, milk and eggs or egg whites were also used, which provided both quality protein and perceived richness. At that point, specialized herbs for palatability, medicinal, or supposed nutritional value were also added.

What is in a Strike?

            Most of the available recipes give measurements in terms of bushels, strikes, and pecks; clearly these recipes are meant to make large volumes. In determining the nutritional makeup of each recipe, the ratios of ingredients are critical. Unfortunately, the term ‘strike,’ as Rubel points out, is a “unit of measure often used for beans that varied in size from half a bushel to four bushels.”[20] This is problematic when other measurements in the recipe are given in pecks, and, in the case of Gervase Markham, when there is no mention of a bushel for comparison.[21] Rubel suggest that a strike of beans is equal to two bushels. This seems entirely reasonable, considering the range of measurements a strike was used for, from two pecks to four bushels. I, however, am accustomed to a strike being one level bushel. A “strike” is the tool used to level a bushel, and a strike was also known as a “London bushel” at the time Markham was writing.[22] Markham spent a great deal of time in London and was buried there. I believe Rubel chose the two bushel measure in order to line Markham's recipes up with John Halfpenny's (1696), who he also discusses. Halfpenny appears to have based his recipes on Markham's. Thomas de Grey (1639), unlike Halfpenny, clearly favors a large amount of grain compared to legumes, and in general there appears to be a trend towards using more flour, more refined flours, and more wheat specifically, all couched in class terms.
            For Rubel's purposes, it doesn't make a large difference. However, since many of my questions revolve around nutritional content (specifically calories, crude protein, and certain mineral ratios), that is a critical difference. And, looking at the recipes, I thought it unusual that Gervase Markham's (1607) first two recipes called for more than twice as much, by volume, of beans than of grain, while his third called for three times more grain than beans (if the measurement of a strike as two bushels were used). If you consider a strike to instead be a single bushel, it is a much less drastic change (being then 4:3 to 3:1, rather than 8:3 to 3:1). Few things will make a horse sick faster than a drastic change in diet.
            I was also convinced that legumes couldn't bake into a loaf without more grain flour. In this, I was wrong. Using the same ratios as Gervase Markham’s “first bread,” with Rubel’s suggestion of the strike being equal to two bushels, I baked a small loaf. It was very hard on the outside, though with almost no defined crust and a complete lack of crumb (the inside was solid, no air pockets). The pea flour to high bran wheat flour at a ration of 8 to 3 had absolutely no trouble maintaining a loaf shape, even without a bread pan (which would not have been used). If anything, the pea flour was too sticky. However, a few things still suggest that the single level bushel measure may be accurate. First is the lack of crumb, a textual quality that is mentioned both in the Assize of Bread and in man of the recipes mentioned. Second is De Grey’s mention of the necessity of whole rye flour to make the loaf stick together. If peas and bran are the only other ingredients, the whole rye flour would only be needed as a binder if the bran was a larger percentage. And lastly, the loaf I baked using the strike as two bushels measurement was a very bright green color. Fava beans are similarly colored, if sometimes paler. Horsebreads, by contrast, are consistently called brown or dark breads.
As a final note
            In cases where traditional horse keeping has not been severed from its roots– i.e., passed from one stablemaster to the next, rather than someone remembering what their grandparents said they did– those that still feed bread feed it stale.[23] This is something that every one of the early modern treatises repeatedly insist on. Fresh bread is less palatable, and apt to cause intestinal upset. Likewise, sourdoughs are preferred, which are closer to traditional horsebreads. Both cases have an effect on both the gluten and the simple starches in the bread, increasing digestibility for the horse.

[1] Laminitis is an inflammation of the tissue within the hoof capsule, which can permanently damage critical structures. For more information see: Lon D. Lewis et al., Equine Clinical Nutrition: Feedings and Care, 1 edition (Baltimore: Wiley-Blackwell, 1995); Lon D. Lewis, Feeding and Care of the Horse, 2 edition (Baltimore: Wiley-Blackwell, 1996); Melyni Worth Ph.D, Storey’s Guide to Feeding Horses: Lifelong Nutrition, Feed Storage, Feeding Tips, Pasture Management (North Adams, MA: Storey Publishing, LLC, 2004).

[2] Worth, Storey’s Guide to Feeding Horses.
[3] “Medieval Technology and American History - In-Depth Articles - The Flower of Wheat: Bread in the Middle and Colonial Ages,” accessed June 10, 2016, http://www.engr.psu.edu/mtah/articles/flower_of_wheat.htm.
[4] “Nutrition Facts and Analysis for Broadbeans (fava Beans), Mature Seeds, Cooked, Boiled, without Salt,” accessed June 10, 2016, http://nutritiondata.self.com/facts/legumes-and-legume-products/4322/2.
[5] Thomas De Grey, The Complete Horse-Man, and Expert Ferrier (London, 1639).
[6] Ronald H. Fritze and William B. Robison, eds., Historical Dictionary of Late Medieval England, 1272-1485, First Edition edition (Westport, Conn: Greenwood, 2002).
[7] Hard wheats are also called winter wheats, because of their ability to survive over winter.
[8] “Nutrition Facts and Analysis for Wheat Flour, Whole-Grain,” accessed June 10, 2016, http://nutritiondata.self.com/facts/cereal-grains-and-pasta/5744/2.; this is problematic for humans as well: “Oats, Wheat, Calcium and Phosphorus | The Well Fed Homestead,” accessed June 10, 2016, http://www.wellfedhomestead.com/oats-wheat-calcium-and-phosphorus.
[9]“Nutrition Facts and Analysis for Wheat Bran, Crude,” accessed June 10, 2016, http://nutritiondata.self.com/facts/cereal-grains-and-pasta/5742/2.; “Keep an Eye on Calcium Balance,” Equinews, May 15, 2012, http://www.equinews.com/article/keep-eye-calcium-balance.
[10] William Rubel, “English Horse-Bread, 1590?1800,” Gastronomica 6, no. 3 (August 2006): 40–51, doi:10.1525/gfc.2006.6.3.40.
[11] Worth, Storey’s Guide to Feeding Horses.
[12] “Nutrition Facts and Analysis for Wheat Bran, Crude.”
[13] An interesting etymological argument could be made based on wheat and rye being singular, but oats being plural.
[14] For more on this, see Charles Gladitz, Horse Breeding in the Medieval World (Dublin ; Portland OR: Four Courts Press, 1997).
[15] Assize of Bread
[16] Worth, Storey’s Guide to Feeding Horses.
[17] Ibid., pg. 204-205.
[18] H. E. Hallam, Joan Thirsk. The Agrarian History of England and Wales:, Volume 2, Cambridge University Press, 1988, pg. 41.
[19] Worth, Storey’s Guide to Feeding Horses., pg. 204-205.
[20] Rubel, “English Horse-Bread, 1590?”
[21] Gervase Markham, Markhams Faithful Farrier: Wherein the Depth of His Skill Is Laid Open in All Those Principal and Approved Secrets of Horsemanship: Which the Author Never Published,but Hath Kept in His Breast, and Hath Been the Glory of His Practise: To Which Is Added Divers Choice Receipts Found in the Authors Closet since His Decease (London? Printed by J. Deacon and are to be sold by W.Thackeray, 1687).
[22] See OED
[23] Pers. Comm. Suzie Peacock, USDF Silver Medalist, et al.


  1. Wonderful paper! I find that when I make the bran-based horse bread -- the bread for the horse masses -- that it always the favorite bread.

    1. Thank you! That means a lot from you. I'm supposed to make another batch and send it for "taste testing" to Four Oaks Arabians. The bran was certainly my favorite! The pea/mixed grain leftover batter also made a spectacular porridge, and at some point I'd like to try cooking a porridge from the already baked bread.