Feed Efficiency – Bovidiva

My Twitter feed just exploded. Yet another study has been released claiming that if we all just gave up beef, the planet would be saved, Elvis would come back from the dead, and rainbow-belching unicorns would graze the Northern Great Plains. I may have exaggerated a little with the latter two claims, but the extent of media coverage related to the paper “Land, irrigation water, greenhouse gas and reactive nitrogen burdens of meat, eggs and dairy production in the United States” seems to suggest that the results within are as exciting as seeing Elvis riding one of those unicorns…but they’re also about as believable.

Much as we’d all like to stick our fingers in our ears and sing “La la la la” whenever anybody mentions greenhouse gases or water footprints, we cannot deny that beef has an environmental impact. Yet, here’s the rub – so does every single thing we eat. From apples to zucchini; Twinkies to organically-grown, hand-harvested, polished-by-mountain-virgins, heirloom tomatoes. Some impacts are positive (providing habitat for wildlife and birds), some are negative (nutrient run-off into water courses), but all foods use natural resources (land, water, fossil fuels) and are associated with greenhouse gas emissions.

So is this simply another attack on the beef industry from vegetarian authors out to promote an agenda? Possibly. The inclusion of multiple phrases suggesting that we should replace beef with other protein sources seems to indicate so. But regardless of whether it’s part of the big bad vegan agenda, or simply a paper from a scientist whose dietary choices happen to complement the topic of his scientific papers, the fact remains that it’s been published in a world-renowned journal and should therefore be seen as an example of good science.

Or should it?

I’m the first to rely on scientific, peer-reviewed papers as being the holy grail for facts and figures, but there’s a distressing trend for authors to excuse poor scientific analysis by stating that high-quality data was not available. It’s simple. Just like a recipe – if you put junk in, you get junk out. So if one of the major data inputs to your analysis (in this case, feed efficiency data) is less than reliable, the accuracy of your conclusions is….? Yep. As useful as a chocolate teapot.

Feed efficiency is the cut-and-paste, go-to argument for activist groups opposed to animal agriculture. Claims that beef uses 10, 20 or even 30 lbs of corn per lb of beef are commonly used (as in this paper) as justification for abolishing beef production. However, in this case, the argument falls flat, because, rather than using modern feed efficiency data, the authors employed USDA data, which has not been updated for 30 years. That’s rather like assuming a computer from the early 1980’s (I used to play “donkey” on such a black/green screened behemoth) is as efficient as a modern laptop, or that the original brick-sized “car phones” were equal to modern iPhones. If we look back at the environmental impact of the beef industry 30 years ago, we see that modern beef production uses 30% fewer animals, 19% less feed, 12% less water, 33% less land and has a 16% lower carbon footprint. Given the archaic data used, is it really surprising that this latest paper overestimates beef’s environmental impact?

The authors also seem to assume that feed comes in a big sack labeled “Animal Feed” (from the Roadrunner cartoon ACME Feed Co?) and is fed interchangeably to pigs, poultry and cattle. As I’ve blogged about before, we can’t simply examine feed efficiency as a basis for whether we should choose the steak or the chicken breast for dinner, we also have to examine the potential competition between animal feed and human food. When we look at the proportion of ingredients in livestock diets that are human-edible (e.g. corn, soy) vs. inedible (e.g. grass, other forages, by-products), milk and beef are better choices than pork and poultry due to the heavy reliance of monogastric animals on concentrate feeds. By-product feeds are also completely excluded from the analysis, which makes me wonder precisely what the authors think happens to the millions of tons of cottonseed meal, citrus pulp, distillers grains, sunflower seed meal etc, produced in the USA each year.

Finally, the authors claim that cattle use 28x more land than pigs or poultry – although they acknowledge that cattle are raised on pasture, it’s not included in the calculations, which assume that cattle are fed feedlot diets for the majority of their life. This is a gross error and underlines their complete ignorance of the U.S. beef industry. Without cow-calf operations, the U.S. beef industry simply would not exist – efficient use of rangeland upon which we cannot grow human food crops both provides the foundation for the beef industry and creates and maintains habitats for many rare and endangered species of plants, insects, birds and animals.

Want to know how to reduce the environmental impact of food production overnight? It’s very simple – and it doesn’t involve giving up beef. Globally we waste 30% of food – and in developed countries that’s almost always avoidable at the consumer level. Buy the right amount, don’t leave it in the fridge to go moldy, and learn to use odd bits of food in soups or stews. Our parents and grandparents did it out of necessity – we can do it to reduce resource use and greenhouse gas emissions; and take the wind out of the sails of bean-eating anti-beef activists.

Can we feed up to 10 billion people in 2100 by improving crop yields, reducing deforestation, and reducing meat and dairy consumption? These solutions are among those suggested by Jonathan Foley at the University of Minnesota’s Institute of the Environment to enable the increase in food production required by the future global population. These are logical suggestions, yet the proposal that meat and dairy consumption should be reduced is likely to be the most-debated, particularly as livestock industry stakeholders may regard this as yet another attack on animal agriculture.

The futility of the “Meatless Mondays” campaign has been discussed ad infinitum, yet in contrast to the EWG’s recent report, Foley does not attempt to promote a vegetarian or vegan ideology or to suggest that climate change could be reversed if only we all ate humanely-certified or organic meat. Instead, the report concludes that resources could be saved if we shifted to meat consumption towards pork and poultry production as:

…it takes 30 kilos [66 lb] of grain to produce one kilo [2.2lb] of boneless beef… We’re better off producing grass-fed beef or more chicken and pork, which requires far less grain feed

Based on those data, Foley’s conclusion is entirely logical. However, as Carl Sagan said, “Extraordinary claims require extraordinary evidence” – and here the evidence is lacking. A recent review of feed efficiency by Wilkinson reports that monogastric animals require 4.0 kg (swine) or 2.3 kg (poultry) of feed per kg of gain. Monogastrics are indeed considerably more efficient than their ruminant counterparts as beef animals require 8.8 kg feed per kg gain – considerably more than swine or poultry, but far less than Foley’s estimate.

It would be convenient to argue that the errors in Foley’s feed efficiency data (not to mention religious limitations on pork consumption) negate the report’s conclusions. But isn’t it logical to argue that we should eat meat produced in systems that use fewer resources to produce animal protein? Personally, I spend more than half my time traveling to present precisely that message to the animal industry and to encourage livestock producers to improve efficiency. I absolutely believe that we need to improve productivity and efficiency in order to feed the growing population. However, traditional feed efficiency data have a major flaw – it’s assumed that all animal feed could otherwise be used to feed humans.

Wilkinson suggests that the traditional concept of feed efficiency be re-examined to reflect the quantity of human-edible crop inputs that are used to produce a unit of energy or protein from animal products. For example, humans cannot digest pasture, only 20% of the nutritional value of oilseed meals can be utilized for human food and yet 80% of nutrients within cereals, pulses and soybean meal are human-edible. By partitioning out the human-edible component of animal feed, Wilkinson demonstrates that the human-edible energy feed efficiency ratios for pork and cereal beef are similar (Figure 1*) and that dairy production actually produces twice the amount of human-edible energy than it uses (input:output ratio of 0.5). On a protein basis, cereal beef has a higher human-edible protein feed efficiency ratio (3.0) than pork (2.6), but suckler beef systems where cattle are grazed on pasture again produce more human-edible protein than they consume (input:output ratio of 0.9, Figure 2*). Not only are these revised feed efficiency estimates considerably lower than those quoted by Foley, but they underline the importance of herbivorous grazing animals in utilizing human-inedible forage to produce animal protein.

Numbers have power – it’s always tempting to base a suggestion around a single data point that “proves” the argument. Feed efficiency is a useful metric, but as we face an ever-increasing challenge in balancing food demand, resource availability and consumer expectations, it’s critical that we examine the bigger picture. The ruminant animal has a major evolutionary advantage in its ability to digest forages – we may be better acquainted with the human dietary advantages of probiotic bacteria than our ancestors, but until we are equipped with human rumens (humens?) we cannot hope to effectively make use of all crop resources.

*The importance of acknowledging the human-edible component of feed efficiency was part of my presentation at the Alltech Ruminant Solutions Seminar in Ireland this week – to go to a PDF copy of my presentation please click here.