Sunday, December 06, 2015

Garbage In, Garbage Out: Low Quality Feed Produces Low Quality Crickets

Crickets are a hot ingredient for some sustainability advocates and start-ups. In marketing materials and other articles, you'll see claims about how these "mini-livestock" are a low-impact food with enormous potential: they can eat almost anything, they don't need much land, and so on (you'll find plenty of examples in my two round-ups about insects as food: 1 and 2). Sometimes it can even seem like magical thinking of South Park's famous underpants gnomes:
  1. Create a cricket farm/energy bar/snack company
  2. ???
  3. 100% sustainable protein source
With so much hype about insects as food, I was pleased to see a well-designed study on cricket farming from the University of California at Davis (h/t Alissa Walker at Gizmodo) that reminds us of a reality of raising living creatures: they need high-quality food to thrive. This is valuable research, because if insects are going to be an important food in the future, insect husbandry needs be given significant attention.

Garbage In, Garbage Out

"Crickets Are Not a Free Lunch: Protein Capture from Scalable Organic Side-Streams via High-Density Populations of Acheta domesticus", by Mark E. Lundy and Michael P. Parrella examines how diet affects the growth of domestic crickets, with special attention given to diets of processed food waste and crop residues. The researchers raised production-scale populations of crickets on five different feeds and measured the growth rate over the course of their lives (with three replications for each feed, each test was about 60 days). The feeds were commercial poultry feed plus rice bran, two types of processed food waste, and two types of crop residue*. The crickets that received the poultry feed thrived, the crickets that ate the first food waste had a satisfactory growth rate, and those that ate the second food waste or the crop residue showed poor weight gain and significant mortality. In other words, the study demonstrated their title — crickets are not a free lunch — and also the classic computer science proverb: garbage in, garbage out.

The chart below shows the growth rate for each feed treatment (it comes directly from the paper, shared here in compliance with the Creative Commons license of the article). The x-axis shows the number of days since hatching, the left-hand y-axis shows the mass per individual. The filled symbols are for poultry feed (PF) and the first food waste (FW1), which both allowed the crickets to grow steadily. The empty symbols are for the other three waste feeds (food waste FW2 and the two crop residues), which resulted in terrible performance: high mortality and certainly not a commercially viable crop.

Figure 1 from Lundy and Parrella, PLOS ONE 10(4), used under CC license

To gain better understanding of how the inputs affected the outputs, the authors fed the results into an analysis tool. The tool generated a model that has a primary input that the authors called the "feed quality index." The index depends on the feed nitrogen (N, a proxy for protein), acid detergent fiber (ADF) and crude fat (CF) as (N / ADF) + CF (a detailed explanation of terms can be found in this PDF). 

The next figure shows the cricket mass at day 30 versus the feed quality index. The relationship is clear: you need good feed to raise healthy crickets. To some extent this is not surprising, but its importance has been somewhat neglected in the excitement about insects as food.

Figure 2 from Lundy and Parrella, PLOS ONE 10(4), used under CC license

The primary lesson I take away from the article is that insect husbandry needs to be taken seriously, not ignored with an "bugs will eat anything" attitude.  With proper focus, it might be possible to find blends of low-grade waste streams that crickets and other insects will thrive on. Or, to use selective breeding to find insects that thrive on waste streams.  Consider farmed salmon:  in just a few decades, the salmon farming industry has made huge improvements in feed conversion efficiency, reducing the amount of feed protein needed to produce a quantity of salmon protein by more than a factor of three between 1972 and 2009 [2].

Other Waste Streams

Out of curiosity, I looked up the properties of some other food stuffs to see how they would fare on the feed index. I found properties for a few major waste streams — spent grain from beer brewing, olive cake leftover from olive oil production, okara from tofu and soymilk production — and also for a seaweed, which is a food with high promise for sustainability.

Shoveling spent grain from the mash tun
"Spent grain" is the dry material leftover from beer brewing and is produced in huge quantities. I suspect that most of it goes into livestock feed or is composted, but there are also some less common uses of the grains in human food, like in granola bars and or in bread (Peter Reinhart's Whole Grain Breads book has a recipe). I found the properties for spent grains in a Nigerian study about using spent grain to enhance the nutritional value of cookies [3], which had the following properties for spent brewery grain: N = 0.0372, ADF = 0.246, and CF = 0.0279, for a feed quality index of 0.179 (note: all feed properties in this post are as fraction of dry weight).

I also thought that the olive cake leftover from the oil milling process was worth looking into since there are large quantities of olive oil produced around the world. As I looked for the properties, I began to suspect that olive cake would be a bust. Olive Processing Waste Management [4] has this to say about olive cake: "Olive cake is not attractive as an animal feed ...The nutritive value of olive cake is also very low...This nutritional value is close to that of straw."  An animal feeding wiki called Feedipidia  had properties for the residue that remains after the three-phase processing of olives and stone removal [5]:  N = 0.015, ADF = 0.463, CF = 0.237, for a feed quality index of 0.27, which isn't good but not quite as terrible as I had expected.

Centuries old olive press at Mission San Jose de Guadalupe

Okara, the fibrous byproduct from tofu and soy milk manufacturing, fares much better since it is high in protein, low in fiber, and high in fat. A Canadian paper about feeding ducks with food wastes [6] had N = 0.053, ADF = 0.12, and CF = 0.12 for a feed index of 0.57, which falls between the two most successful feeds in the study.

One of the food wastes in the duck study was something called pogo, "a weiner on a stick enveloped in dough and deep fried, discarded because of over- or undercooking, being misshapen, or held beyond the expiration date." Pogo is another name for "corn dog," and not surprisingly this food was quite high in protein and fat, and quite low in fiber, so its feed index was more than 4.5! Being so far off the chart, I suspect that the model formulated by Lundy and Parrella would not work for pogo.

Since they can be cultivated readily in the ocean without any additional input, seaweeds could be sustainable food (a recent article by Dana Goodyear in the New Yorker explores the topic; Goodyear and chef Brooks Headley talked to Evan Kleiman on KCRW's Good Food about seaweed). Seaweeds are typically quite high in fiber and low in fat, though, so the feed quality index is low. Feedipedia has a nutritional table for giant kelp (Macrocystis pyrifera) with N = 0.016, ADF = 0.13, and CF = 0.04, for a feed index of 0.16 [7]. One of the most important points in Goodyear's article is that there are thousands of species of seaweeds and so far there has been little selective breeding to create varieties that would be better for human or livestock consumption.

The figure below rounds up the results. On top of the original black and white figure from the article, I placed the results for spent grain (SG), olive cake (OC), okara (O) and giant kelp (GK).

Figure 2 from Lundy and Parrella, PLOS ONE 10(4), with four new data points (green circles), adapted under CC license.  New points: BG = brewery grains, GK = giant kelp, OC = olive cake, O = okara.
To be sure, looking at single waste streams isn't the most practical exercise because commercial cricket operations will use a blend of ingredients that are chosen to meet meet financial, logistical, and sustainability targets.  Perhaps an operation launches in the Sacramento area to take advantage of the streams of okara after Hodo Soy Beanery revives the recently closed Sacramento Tofu Company factory, spent grain from local microbreweries, olive cake from the olive megafarms to the north, and agricultural waste from every direction.  Or the companies currently supplying feed for the companies that raise crickets for the pet food industry will start working on new feed blends. If a serious market appears for insects as food for humans and livestock, optimization of feed will follow.

* The five feeds: PF = 5:1 ratio of poultry starter feed and rice bran; FW1 = solid, pasteurized, post-process filtrate from an aerobic enzymatic digestion process that converts grocery store food waste; FW2 = minimally processed post-consumer food waste from local municipalities; CR1 = 1:1 ratio of wheat and maize silage (dairy cow feed) with approximately 50% straw; CR2 = 2:1:1 ratio of poultry manure, wheat straw, and rice straw silage

[1] M.E. Lundy, M.P. Parrella, Crickets Are Not a Free Lunch: Protein Capture from Scalable Organic Side-Streams via High-Density Populations of Acheta domesticus. PLoS ONE 10(4): e0118785, 2015. doi: 10.1371/journal.pone.0118785
[2] Ole Torrissen , Rolf Erik Olsen , Reidar Toresen , Gro Ingunn Hemre , Albert G.J. Tacon , Frank Asche , Ronald W. Hardy & Santosh Lall (2011) Atlantic Salmon (Salmo salar): The “Super-Chicken” of the Sea?, Reviews in Fisheries Science, 19:3, 257-278, DOI: 10.1080/10641262.2011.597890
[3] K.O. Ajanaku, F.A. Dawodu, C.O. Ajanaku and O.C. Nwinyi (2011). Functional and Nutritional Properties of Spent Grain Enhanced Cookies, American Journal of Food Technology, 6: 763-771, 2011. DOI: 10.3923/ajft.2011.763.771
[4] M. Niaounakis and C.P. Halvadakis, Chapter 10 ("Uses") in Olive Processing Waste Management: Literature Review and Patent Survey, Elsevier (London), 2006.
[5]  HeuzĂ© V., Tran G., Gomez Cabrera A., Lebas F., 2015. Olive oil cake and by-products. Feedipedia, a programme by INRA, CIRAD, AFZ and FAO.  Last updated on May 11, 2015, 14:32.  Nutritional table
[6]  A. Farhat, L. Normand, E.R. Chavez, and S.P. Touchburn, Nutrient Digestibility in Food Waste Ingredients for Pekin and Muscovy Ducks, Poultry Science 77:1371–1376, 1998, via ResearchGate
[7] Heuzé V., Tran G., Giger-Reverdin S., Lessire M., Lebas F., 2015. Seaweeds (marine macroalgae). Feedipedia, a programme by INRA, CIRAD, AFZ and FAO. Last updated on October 14, 2015, 16:26

Image Credits

Random link from the archive: Charting the chocolate chip cookie

No comments: