Monday, October 27, 2014

Economic Value of Manure Sampling and Testing

As you may recall, last week we talked about a bit about manure in relation to precision farming. We said there are really three aspects to precision farming, measuring something, making a decision based on that thing we measured, and then implementing our decision. Today we are going to look at these concepts in terms of manure management, specifically in terms of manure sampling and testing.


To start with we need to understand a little bit about manure. One of the most common questions asked about manure “what is the fertilizer value of manure.” This question usually means, “How much nitrogen is in the manure?” or stated a slightly different way, “how much manure should I apply to fertilize my crop?” This question doesn’t have an easy answer, because unlike commercial fertilizer (anhydrous ammonia, MAP, DAP, or urea) we might usually purchase, manure doesn’t come with a guaranteed composition. Instead, there can be substantial variation in nutrient content from one-farm to the next, year-to-year on the same farm, and to a lesser extent the manure that we are pumping out during a single land application event. This variation makes “book values,” i.e., estimated nutrient contents available in Midwest Plan Services or state extension materials a good starting point for planning purposes, but means that we really need to measure the nutrient content of our own manure to understand what is in it.

One way to think about this is based on probability. This is demonstrated for a swine manure from a deep-pit manure storage in figure 1 below. What you see is that the manure has the greatest probability, or the highest chance of having at a nitrogen content of 7 kg/1000 L (or right around 58 lbs. of Nitrogen per 1000 gallons of manure). However, what’s harder to tell is that there is something like a 20% chance that swine manure might have more than 9 kg N/L (75 lbs./1000 gallons) or a 20% chance that swine manure will have less than 5 kg N/L (41 lbs./1000 gallons). This means if we were basing our manure application off “typical” or average swine manure and trying to applying 150 lbs. N/acre, there is a 20% chance we apply more than 190 lbs. N/acre and a 20% chance we apply less than 110 lbs. of N per acre. So what does this mean to us? Well, it implies there is a pretty good chance we won’t be applying the amount of nitrogen we think we are. This is costing us lost revenue from either not fully utilizing our N (if we end up applying more nitrogen than we need) or from nitrogen limitation in our crop growth (if we are applying less N than we think we are).

Figure 1. Probability or a swine manure from a deep-pit manure storage having different nitrogen contents.

Based on this I think we have an interesting question we can ask ourselves, and that is, what is the return on investment of manure sampling. That is, how much value does manure sampling add to our operation? In determining the value of the manure test, it is important to understand how a farmer can use the information gained from the test results, i.e., how having this information alters the farmer’s nutrient management and affects the farm profit. This is a complex topic, as almost limitless possibilities exist, but to get a general idea I’m going to make a few assumptions and approximation about common cropping and manure management systems in the Midwestern United States. This may mean the results don’t give you an exact value for your operation, but they give you a pretty good starting point.

In this evaluation, I assumed that the manure application method would be either injection or immediate incorporation to maximize N utilization. Additionally, I assumed that best management practices for manure application timing were followed; as a result, the yield response to available N (defined here as the sum of ammonia N and organic N expected to mineralize in the first growing season) would be the same as the yield response to mineral N fertilizer. Finally, we limited crop rotation choices to continuous corn and corn-soybean rotations, as these represent the dominant rotations in the upper Midwestern U.S. The impacts on the value of the manure test of N-limited or P-limited application, as well as when sampling or testing was conducted, were handled by evaluating all cases. Finally, the basis of this effort was that farms intend to use their manure resources to support crop production. In cases where farmers have insufficient land to use all their manure resources, they can only extract the value of the manure test if they can find buyers for the manure nutrients.

The value of the manure test was calculated by estimating the profit that would have been made if the manure was assumed to have a “typical” nutrient composition and then to compare this to the profit generated if the actual nutrient composition was known. The profit graphs for these steps are shown in figures 2, 3, and 4. As a note, I performed this analysis earlier this spring when corn price was $5 per bushel.

Figure 2. Probability of different profits due to different nitrogen contents of manure assuming manure standard rates.

Figure 3. Expected profit from applying manure of a known composition at the maximum return to nitrogen (rate changed to account for measured nutrient content).

Figure 4. Expected value of the manure test, based on graph 3 minus graph 2 (adjusted based for sampled manure nutrient content case minus assumed nutrient content case).

To put these results in perspective I used the model to evaluate what manure sampling would be worth for a 1000-head capacity swine finishing farm using a deep-pit manure storage. On average, the facility generated 4 L of manure per head per day. This farm has collected and tested manure samples every year for the last five years. The first four years of manure sample values were 0.84%, 0.72%, 0.98%, and 0.62% N, with an average and standard deviation of 0.79% ±0.16% N. The N content for the current year was 0.92% N. If no sample was tested, this operation assumed that the manure had an available N content of 0.79%, the average of the previously collected samples. Using pre-application sampling and assuming that manure application was N limited, the value of the manure test would be $30.96 ha-1. Assuming that the manure sample is representative of all the manure from this building, the overall value of the sample was $1,759 (the farm would have applied manure to 56.8 ha). This represents a good return on investment, as the approximate cost of obtaining this information would be $50 for manure testing, $50 for shipping the manure to the testing lab, and $100 for the farmer’s time to collect, label, and ship the sample, giving a return of almost 9:1. If manure application was P limited and manure was sampled during application, the estimated value would be $14.20 ha-1. In this case, the manure was applied to 112 ha, so the actual value of the test would be $1,589

The same analysis was performed for other manure types and estimated values of manure testing are shown in table 1.Overall, the results indicated that sampling and testing the manure provided enough economic value that it would pay for itself.

Table 1. Estimated value of the manure test for different manure type and crop rotations.