TMR Sampling: Valuable Exercise or a Random Number Generator?

Sampling and analyzing the total mixed ration (TMR) has several potential uses. It can be used to identify nutritional deficiencies or surpluses in the diet that was actually fed to the cows. It can be used to estimate manure excretion of nutrients via mass balance calculations. The consistency of ration delivery can be evaluated by sampling the TMR, and it can be used to determine whether the ration that is delivered to the cows is the same as the diet that was formulated. However, for any of these uses to be valid, the TMR sample must accurately reflect the diet that was actually fed. Previously, we found that sampling variation was substantial for TMR samples. This was investigated further by sampling three different TMR (one had silages and concentrate; one had silages, concentrate and hay, and one had silages, hay, whole cottonseed, and concentrate) using two different sampling protocols. One protocol was simple and consisted of taking several handfuls of TMR across the feed bunk. The other protocol consisted of putting trays in the feedbunk prior to feed delivery and then removing the trays filled with TMR, mixing, and sampling from the trays. Sampling protocol had very little effect on sampling variation or on the accuracy of the sample. Samples of TMR did not accurately estimate the true mineral concentrations (sodium, phosphorus, and copper) of the TMR. A single sample of TMR (using either protocol), however, generally gave an accurate estimate of the true concentration for dry matter (DM) and crude protein (CP) in the TMR. For neutral detergent fiber (NDF), a single sample had a high risk of being wrong (i.e., inaccurate), but taking duplicate samples and averaging the analytical results were generally accurate. TMR sampling can be accurate for macronutrients but care must be taken when sampling and often duplicate samples will be required.