Alfalfa hay marketed in California is currently sold on TDN, which is estimated from fiber content. Studies conducted by the University of California-Davis in the 1950s indicated that the single proximate entity most highly correlated with TDN was modified crude fiber. At that point in time, the use of a single entity to estimate quality was understandable.
Since the 1980s, acid detergent fiber (ADF) has been used to estimate TDN. Chemical entities comprising ADF are not constant and vary with season and stage of maturity. Alfalfa hays with the same ADF content can vary dramatically in digestibility. The current system is antiquated due in part to the efforts of geneticists and agronomists who have developed new alfalfa hay varieties.
The California chapter of the American Registry of Professional Animal Scientists (ARPAS), in conjunction with the University of California-Davis, California State Polytechnic University-San Luis Obispo and Sapienza Analytica, has undertaken a project to update the method used in California to better characterize alfalfa hay quality.
A series of studies are planned with the goal of estimating metabolizable energy of alfalfa hay from near-infrared reflectance spectra. These studies employ both in vivo and in vitro methods to measure disappearance of organic matter, dry matter as well as rate, site and extent of digestibility of selected chemical and proximate entities.
Sampling alfalfa hays for both the in vivo and in vitro studies, representing the diversity seen throughout California, began in April and will continue until this fall. It is expected that the in vitro portion of this study will be completed by the middle of 2009 and that the in vivo study will be completed by early 2010.
California ARPAS is funding this project via corporate sponsorships and also by holding a series of raffles.
Current methodology used to evaluate alfalfa hay in California is unchanged from the system developed at UC Davis in the 1950s, with the exception that acid detergent fiber (ADF) is the proximate entity of interest, not modified crude fiber. The goal of this study is to develop a system to calculate metabolizable energy from digestible energy. In turn, digestible energy will be calculated from apparent degradability (de) of each crude protein (CP) and neutral detergent fiber (NDF) times their respective caloric density. Apparent degradability will be measured for each sample from rate, site, and extent of digestion. The initial feedstuff will be pure-stand alfalfa hay and will employ measuring methods based on near infrared spectrophotometry (NIRS).
Samples were collected during 2008 from throughout California and represent the diversity seen in the reported nutrient content of alfalfa hay. Data from laboratory Dairyland Laboratories Inc. (Arcadia, WI) for 3000 alfalfa hay samples from throughout the United States show a range in crude protein content of from 14.7 to 24.6%; our range is from 11.8 to 24.6%. The ADF range was greater for the California ARPAS samples (20.5 to 51% vs. 24 to 39%). The range of NDF was also greater in the California ARPAS samples, 41.4 points of NDF versus 19.1 for Dairyland’s 3000 samples. Of special interest is the range in presumed hemicellulose (calculated as the difference between NDF and ADF). In the California samples, the range is 4 times greater in than in the 3000 samples collected from throughout the United States. It is commonly believed that the presumed lack of variation in hemicellulose makes it a poor candidate for study. Our data show sufficient variability to make it an item of interest.
Dry matter content appears to be normally distributed, whereas CP, ADF, NDF, and TDN (total digestible nutrients calculated by regression from ADF content) were distributed in a biphasic manner. These latter observations are consistent with the current pricing structure in California; premiums are paid for hays containing less then 30% ADF, and hays with a greater ADF content are discounted. Alfalfa that will not be suitable for lactating dairy cows is allowed to grow longer and tons/acre increased.
All samples collected will be analyzed using traditional wet chemistry, NIRS, and in vitro digestion techniques. Ten samples representing the range in diversity will be used in an in vivo study; weaned wether lambs will be fed at levels that approximate maintenance and ad libitum.
Kinetic analyses of in vitro data will use terms similar to those found in current computer models. All estimators will be calculated using least squares models, assuring that these estimators will be unbiased. Initial in vitro data are interesting; alfalfa samples with similar ADF concentrations have similar disappearances at 30 hours. Kinetics of degradation show differences, biological end-points are different, and these are not detectable using current methodologies.
Disappearance of material from the digestive tract is a function of the digestibility of that material and rate of passage. This can be represented mathematically as:
de = kd/(kd + kp),
where de = apparent degradation, kp = degradation rate, and kp = passage rate.
When kp = 0, as is found in an in vitro system, then de = kd. Given that de will be measured at kp = 0, maintenance and ad libitum, it will be simple to estimate kp for in vivo trial conditions. Passage rates may also be estimated from feed and fecal compositions in a dairy or feedlot setting. Seat-of-the-pants estimates may be made from feed intakes and production.
Estimates of metabolizable energy are made by subtracting losses in feces, urine, and methane from gross energy (feed). Given that nitrogen intakes will be high relative to animal requirements, estimates of ME will be low. It is expected that, for most models, NEl will be the property of feed used to express energy content. Given that the efficiency of ME use for milk synthesis is constant, nutritionists may use whatever kl is appropriate for the herd or animal being evaluated.