imageHere is another Lean Six Sigma (LSS) tool that I love: the Pareto chart. In Lean, there is constant focus on eliminating the “seven wastes.” Six Sigma is used to apply a mathematical approach to measuring the wastes and identifying ways to eliminate them. In general, problems often have multiple contributors, or root causes. Often it’s not feasible, affordable, or even worthwhile, to address all of them. Sometimes getting rid of 80% or 95% of the contributors is good enough. Often, we can’t achieve perfection, or complete elimination of a problem. So, how do we decide which root causes to tackle first?

Enter the Pareto. This is a two-way graph. The bar graph shows how much each individual root cause contributes to the problem, graphed in descending order. The line graphs the overall contribution, so you can see where, for example, the 80% mark is. If you want to eliminate 80% of the contributors, the combined graphs show you which categories to knock out to achieve this. The remaining outliers may be “in the noise” and not worth addressing.

TripletsI find this approach to be helpful when thinking about lamb loss problems. Though in the chart above, there isn’t a lot of data to work with, it’s still useful to me. Lambing time is a blur. We always expect some losses, so any particular loss isn’t extraordinary in the moment. They happen at different times: some in the womb, some at birth, some in the first few weeks, then some here and there as the lambs are growing. By the time the lambs are weaned, heading into adolescence, and thriving on pasture grass, I know I’ve lost some, and have a vague count in my head. But I can’t remember the details, I have to go back through my records to compile them. I certainly couldn’t say off the top of my head what my biggest problems are, and if I tried, I would probably be wrong. 

This year I had eleven “potential” lambs die, which represents 17% of my crop. This is counting any observed fetuses, some of which are very tiny and barely noticeable. But each of those represents a lamb that could have been; so when I do notice them, I make note of them. At the other end of the spectrum, some lambs thrive for months, then drop dead suddenly in the pasture with no warning.

Where not to spend time

Graphing the losses gives a good overview of where to focus my problem-solving efforts. Though it would  be ideal to save every lamb, that’s not a reasonable goal. The “Clostridium?” category is a prime example. For one, there is a question mark there because my diagnosis is only a best guess. LSS discourages away from doing any guesswork: we should only implement solutions to problems we’ve proven exist for sure; and ideally we should only invest in solutions which are also proven to work. So, before I spent time on this problem, I’d need to get a definitive diagnosis on more lambs to be sure it is a real root cause; i.e. have a veterinarian do some necropsies for me.

Clostridium is also called enterotoxemia, or “overeating disease,” caused by a toxic bacteria in the gut that can run rampant in lambs which are eating very rich food. I vaccinate my ewes for this, so they pass on immunity in colostrum. This should generally serve the lambs during the period where they’re getting very rich milk. As my lambs age, their maternal immunity drops off. At the same time the milk supply is also dropping off, and they are transitioning to a grass diet, where there is less risk of clostridium getting the upper hand.

In operations where lambs are fed very high-protein creep feed for an extended time, it’s often necessary to vaccinate the lambs for clostridium multiple times as they are growing. (You have to hit them more than once, because maternal immunity drops off at a variable rate, and vaccinating before that time renders the vaccine useless. So it’s a roulette game of vaccinating two, three or four times, hoping to hit the sweet spot where each has just lost maternal immunity, but has not gone too many weeks with no immune protection at all.) On my farm, theoretically it should be rare for a lamb to die from clostridium. It would be caused only by the unusual combination of early maternal immunity drop-off, late availability of rich milk, and probably a very piggish lamb and a permissive mother.

If I were to start vaccinating lambs for this, it would cost about $40 per round, plus the labor of vaccinating all the lambs multiple times. There is the interesting twist that vaccine introduces risk, since in rare cases, it can trigger anaphylactic shock. It also has the potential to spread disease, because most of us use the same needle across multiple animals, for the sake of efficiency. All this adds up to my conclusion: it’s not worth introducing this protocol to maybe save one lamb a year.

Biggest bang for the buck

On the flip side, what this graph does tell me is that I could benefit from focusing more on preventing fetal losses and died-at-birth (DAB) instances. More than half my problem lies there. I was actually surprised to learn that when I graphed it, because that wasn’t my perception (again, the chaos of lambing time tends to blur the details, which is why it’s so helpful to keep detailed notes for later review). I’m already vaccinating for the main abortifactants: chlamydia and vibrio. I do worry about Q-fever, listeria and some others for which we cannot vaccinate. Many of these diseases are bird-borne, so one idea would be to switch from feeding grain to alfalfa for late pregnancy nutrition. This would cut down on the number of birds poking around in my feed troughs and  high traffic-areas.

Switching to alfalfa instead of grain would also cut down on the competitive eating and pushing-and-shoving that ensues when grain buckets are poured into feeders. So, it’s something for me to pencil out: the cost delta to achieve the same nutrition from alfalfa instead of corn-barley, plus the extra labor to handle the greater weight and volume of hay versus grain. Then, decide whether this is worth the potential to save a few more fetuses.

Handling is another area where I think I could do better. I already avoid handling the sheep as much as I can while they are pregnant. I trim hooves before breeding and after lambing. But I do still handle them during pregnancy to vaccinate, blood test, do SFCP inspections, and move them into different grazing squares. And we’re back to that paradox where vaccinating is intended to reduce risk, but it also introduces risk. In handling, generally my dogs are good, but sometimes they rush the sheep too much, or “accidentally” split one off and chase it at a dead run (naughty, naughty!). Sometimes I’ve moved sheep over slippery wooden bridges, which I’ve learned almost never goes smoothly, even without a dog involved! Sheep have a propensity to have dramatic, slip-on-a-banana-peel kind of falls; the kinds of falls which happen with such force, they can literally jar a fetus loose from its moorings. The goal would be to always move the sheep gently and smoothly, with no hard cornering or rushing: something which I admit doesn’t always happen here!

Ultimately, I have a goal to build a nice Y-chute sorting system, ala Temple Grandin, to make handling go more smoothly. Thus far, I’ve been making do with bunching the sheep into a corner and having the dog hold them, while I do my thing and exhaust them through a man gate. This has worked, but I’m almost to the point where I have too many sheep to cope with doing it this way. So, a chute system is on my to-do list for the next year or two. It should reduce my labor, increase efficiency, and hopefully improve the gentleness of all handling that must be done on pregnant ewes.

The other big takeaway for me on the graph is to try to be available during births a little more. Often DAB’s can be saved if one is there to pull a lamb that would otherwise be stuck in the birth canal too long, or get it up and nursing if it is weak. I take vacation during lambing and monitor the sheep pretty closely during the day. I also value getting my sleep, and don’t believe in sacrificing health and sanity to be “there” for every single birth at all hours of the night. But I can see that increasing my vigilance as much as is feasible could help catch a few more DAB losses and save them. This graph will be in the back of my mind when I’m tempted to hit the snooze button on the alarm, or skip that before-bedtime check because I’m tired!

Cool tool

This perhaps isn’t the best example of a Pareto, as ideally one would have a lot more data, and definitive data, to make concrete conclusions about what improvement efforts would be most fruitful. But is a good illustration of how using some very basic back-of-a-napkin graphing and data analysis can lead you to more objective conclusions and the best investment of improvement efforts. If I had gone from memory, I think the issues that stand out most for me are finding a big, beautiful lamb dead in the field when he was perfectly well the day before- the clostridium case. Fetal deaths stick in my memory the least, because birthing time is busy, and the attention is on the living. But when I graph it, clearly I can see that focusing on the fetal timeframe has the most potential for increasing my yield in future years; and I should worry less about the few late losses, which are more random in nature.