Appendix E - Taking a Good Forage Sample for Analysis
Sampling is a major factor affecting the accuracy of forage quality analyses. Chemical analysis is valid only to the extent that the sample analyzed represents the lot of hay or haylage to be fed. The true analysis value of the lot is never known. Thus, the analysis of any sample is an estimate with a defined range within which the true value has a certain likelihood of occurring.
Sampling equipment, sampling method, and sample handling, once the sample has been obtained, will affect estimates of quality. Variation of forage quality within storage packages is always present and will affect the accuracy of any sampling technique. The sample to be analyzed must be representative of the designated lot of hay or silage. One should obtain a small, but representative, amount of hay or silage from many different locations or bales in a lot to adequately represent all the variations present.
Take samples by "lots" of hay or silage. A "lot" is defined as hay or silage which has been made from the same cutting, field, and stage of maturity. A sample should not represent more than 200 tons dry matter. For lots larger than 200 tons, two or more samples should be taken and the average of the results used to represent the lot.
Hay. Hay is harvested and preserved in a number of different packages. It may be in the form of pellets, cubes, small two-wire or string bales, small three-wire bales, large square bales (1 ton), large round bales, or stacked as loose hay.
The most commonly used sampling method for baled or stacked hay employs a hollow tube (probe) to extract core samples from the hay. Use a probe that travels at least 12 to 18 inches into the hay package for most hay packages. The internal diameter of the probe should be at least 3/8 of an inch. Probes with sharpened tips must be kept sharp to cut through hay. A dull tip may reduce the amount of stem material in the sample due to the tip sliding past rather than cutting through the stems.
Baled hay packages are not uniform products because the initial windrows were not uniform and the baling process affects the distribution of leaves and stems (bale structure) within the bale. Based on the structure of the hay package to be sampled, the hay should be probed in such a way as to adequately sample the various concentrations of stems and leaves. At least 20 cores (one core per bale) should be taken and composited to develop one sample per lot. Bales within a lot of hay should be sampled at random. Random means that there should be no prechosen reason for selecting a specific bale to sample (i.e., location, color, leafiness, etc.). Techniques to guard against nonrandom sampling are to sample every fourth or fifth bale going around the stack, truck, or down the row in the field or take at least five random samples from each of the four sides of a stack.
Sample rectangular bales, regardless of size, using a probe centered in the end of a bale and drill horizontally into the bale.
Sample round bales by drilling horizontally into the curved side of the bale. Deteriorated hay from the exterior of the bale should not be sampled if it will not be fed to animals or they can be selective in their feeding. However, if hay to be sold includes the deteriorated exterior, it should be included in the sampling. Bales stored outside should be sampled within 2 to 4 weeks of feeding so that continued deterioration does not significantly lower bale quality from the sample taken for analysis.
For loose hay use a probe at least 30 inches long with 3/4 inch or larger internal diameter and drill at an angle from the side of the stack to the probe's full depth in 20 random locations throughout the stack. In a mow, hold the probe vertically and drill at the spot where the hay is compressed by the weight of the operator. Discard any weather damaged surface layer that would not be included in the part being fed or sold. Hay stored outside should be sampled within 2 to 4 weeks of feeding so that continued deterioration does not significantly lower bale quality from the sample taken for analysis.
Hay cubes or pellets should be sampled by collecting several hay cubes or handfuls of pellets from 15 to 20 locations in each "lot" so that a minimum of 40 cubes or 2 lb of pellets are selected. Each lot should be limited to 200 tons or less.
Silage. Collect a 1- to 2-lb sample from the silo unloader while it is operating or a comparable amount from several sites in a bunker or silo tube. Do not collect a silage sample until at least two weeks after ensiling. Do not collect a silage sample from the top 2 to 3 feet in a top-loading upright silo. Avoid sampling from moldy or spoiled areas in silo, bunker or tube. Also, avoid sampling silage that has been exposed to the air for several hr. Sample bunker silos by sampling 12 to 15 sites from the face of the silage in the silo.
Sampling chopped forage as it is being put into the silo will give an indication of forage quality but will not account for changes occurring during the ensiling process. Fiber changes are usually less than 1 unit and occur primarily because digestible material is lost through respiration or juices leaching out. Protein content and solubility can change significantly during the ensiling process depending on the fermentation process.
Total Mixed Rations (TMR). Total mixed rations are difficult to sample because they are seldom homogeneous or well mixed. When it is unlikely that a sampling method can produce a representative sample, it is recommended that the components of total mixed rations be sampled and analyzed individually. When confident that a representative sample can be obtained, a TMR sample may be analyzed by wet chemistry. NIR calibration on TMR samples has not been successful.
Sampling silages, haylages and total mixed rations may produce a large amount of sample. The sample should not be divided because stems and leaves will separate and settle in the sample. The sample should be taken early in the week, placed in a polyethylene, airtight (e.g. freezer) bag, sealed tightly and immediately mailed or delivered to the laboratory. Perishable samples should be mailed immediately after collection and should be mailed early in the week so they arrive at the laboratory without spending the weekend in shipment. Samples can be frozen before shipment if freezing will not affect analytical results (prussic acid analysis). For samples handled in either manner, the laboratory report of dry matter "as received" will approximate the dry matter content of the lot of hay or silage when sampled.
The sample should be properly labeled for the lot and locations where the cores were taken when it was sampled. The sample should also be labeled for source (area where grown), forage type (species), cutting, stage of maturity, and special conditions (frost, drought, etc.). Further information such as cutting date and interval between cuttings may also help a laboratory make a decision about atypical samples.