Relating quality changes to storage time for baled alfalfa

The relationships between storage time and several chemical indices for forage quality were established for alfalfa hay baled at two moisture levels (29.9 and 19.7%) in conventional and laboratory bale s made at 1.0, 1.5, and 2.0 times the density o f parent, conventional bales. Bales were sampled after 0, 4, 11, 22, and 60 days. For the high-moisture bales, most quality indices indicated substantial nutrient loss early in the storage period, particularly between days 4 and 11, with little change after 22 days. A nonlinear mathematical model was constructed to describe how neutral detergent fiber and several other quality indices changed with storage time. Acid detergent fiber was related poorly to storage time. Little change occurred in the low (19.7%) moisture bales.

The autho rs wish to express their thanks to Larry Klocke, Lyle Pfaff, Charles King, Duane 1 Starkey, Shane Faurot, Toby Matthies, Greg Basgall, and Francois Altidor for field assistance in completing this project. Department of Agronomy.

Summary
The relati onships between storage time and several chemical i ndices for forage quality were established for alfal fa hay baled at two moisture levels (29.9 and 19.7%) in conventional and laboratory bale s made at 1.0, 1.5, and 2.0 times the density o f parent, conventional bales. Bales were sampled after 0, 4, 11, 22, and 60 days. For the high-moisture bales, most quality indices indicated substantial nutrient loss early in the storage period, particularly between days 4 and 11, with little change after 22 days. A nonlinear mathematical model was constructed to describe how neutral detergent fiber and several other quality indices changed with storage time. Acid detergent fiber was related poorly to storage time. Littl e change occurred in the low (19.7%) moisture bales.

Introduction
Hay quality degrades during storage when moisture levels at baling exceed about 20.0 percent. These changes, fa cilitated by microbial activity and the subsequent heat generation, include oxidation of carbohydrates, dry matter (DM) loss, mold growth, and increased concentra tions of fiber components and artifact (unavailable) nitro gen. These changes decrease the relative feed value and decrease animal performance.
Previous research designed to s tudy changes in forage quality during bale storage has relied on sampling at a single time, normally after microbial activity has ceased. By that time, internal bale temperature has returned to near ambient, following an initial rise. The quality changes that occur early in storage, i.e., when bales are actively generating heat, has largely been ignored . Such information is important to the mechanism s that cause deterioration in hay quality.
Our objective was to establis h how quality of alfalfa hay in laboratory-scale and conventional bales changed during storage.

Experimental Procedures
A 3-year-old stand of `Kansas Common' alfalfa was harvested (fourth cutting) at 10% bloom with a mower-condition e r on September 9, 1992 near Keats, Kansas. The forage was allowed to dry undisturbed, until the desired high-and low-moisture levels were reached on September 11. Actual m oisture levels averaged out to 29.9 and 19.7%. Densities of conventional bales were 19.4 and 11.7 lb/ft (as-is 3 basis) for the high-an d low-moisture alfalfa, respectively .
Laboratory-scale bales were subsequently made fr om the same alfalfa at 1.0, 1.5, and 2.0 times the density of the conventiona l bales using a method described previously (KAES Report of Pro g ress 678, page 31). Specific bale characteristics for each treatment appear in Table 1. All bales were stored in small haystacks.
Hay was sampled prebaling and a t 4, 11, 22, and 60 days postbaling. Sampling dates were chosen prior the study to approximately coincid e with specific points on typical temperature vs. time curves for alfalfa hay during storage. The se curves are characterized by two prominent temperature maxima. The first is short duration (1 to 4 days) and occurs immediately after baling. The s e cond maximum is normally a broad peak or group of peaks, which occurs after the internal bale temperature s has decreased after the first maximum (i.e., abo ut 5 to 20 days post-baling). Bales were sampled on day 4 to separate effects of the first and second temperature maxima. Addit ional sampling at 11 and 22 days was targeted at the second temperature maximum. By day 60, we considered nutritional quality to be stable.
Samples were analyzed for nitrogen (N), NDF, ADF, and acid detergent insoluble N (ADIN) . Acid detergent-insoluble N concentration s were calculated and reported based on a total DM (ADIN-DM), total N (ADIN-N), and total ADF (N-ADF).

Results and Discussion
Each of these quality indices was regressed against stor age time within each moisture level. Various linear and nonlinear models were used to determine the best relationship. For all indices, quality changes over time were best fit by the nonlinear function;

Y = A -Be -kt 2
where A = asymptotic maximum of the response variable , B and k are constants compute d by fitting the model, an d t = time (days). Comparisons of prestorage alfalfa quality traits at both moisture levels indicated no significant differences between bale types within a given moisture level (Table 2). This establishe s that the laboratory-scale baling process did not affect forage quality, and all treatm ents of the same moisture content had a common starting point before entering storage.
During the storage period, low-moisture (19.7%) bales ex hibited relatively minor changes in alfalfa quality indices (data not shown). This was expected, because 20% moisture has been suggested as the threshold for negative quality changes during storage. At the high (29.9% ) moisture level (Table 3), quality changes for all baling treatments were much more pronounced.
All high-moisture treatments exhibited significant incre ases in NDF during storage, but none of these increases occurred between 22 and 60 days, suggesting that the changes occu rred within the first month of storage. NDF increases ranged from 3.4 to 4.8 percentage units in the high-moi s ture bales over the first 4 days of storage, suggesting that they might be associated with the first temperature maximum.
Increases in ADF with time were less dramatic than those in NDF. Am o ng high moisture treatments, only conventional and laboratory-scal e bales (density factor = 2.0) exhibited statistic ally significant ADF increases.
Acid detergent insoluble N responses were relatively consiste nt across the three methods of expressin g artifact N . Particularly large ADIN increase s occurred between 4 and 11 days of storage , a window of tim e when all bales were intensely generating heat. Artifact N levels usually changed little from 22 to 60 days. This suggests that degradation of n itrogen availability is essentially complete within the first month of storage.
It is unclear why NDF valu es increased with storage so much more than ADF. However, change s in both are thought to be due to losses of soluble or nonstructural components and not to changes in fiber components themselves.