I have recommended fall harvesting of alfalfa following a killing frost for many years, but an observation in 1999 indicated that alfalfa might be harvested anytime in the fall if the plant was ready. I classify the plant ready when it reaches 40 to 50% bloom or regrowth is initiating from the crown (at leasted 2 to 3 inches of growth). This experiment was initiated in 2004 to determine if this hypothesis was correct.
The experiment was located at Fargo, ND, on a Fargo clay soil. The treatments included two inoculated alfalfa cultivars, AmeriStand 201 + Z and WSI 3.0, and two cutting managements, 3 or 4 annual harvests with the fourth taken in the fall (during September or October). The treatments were laid in the field as a 2 by 2 factorial with two replicates. Each experimental unit was 10 by 40 feet. Seed of each cultivar was seeded with a double-disc-opener drill at about 13 lb/acre during May 2004. Weed control was obtained by applying 3.0 fl oz/acre of Pursuit mixed with MSO when the alfalfa was in the second trifoliolate leaf stage. Weed control was excellent.
The 3-cut treatment was harvested once on August 5 and the 4-cut treatment was harvested twice on August 5 and October 7 in 2004. Harvest dates in 2005 were June 22, July 19, August 29, and September 28 and were targeted at very early bud, mid bud, 30% bloom, and regrowth initiating. Harvest dates in 2006 were June 1, June 30, July 31, and September 13 also targeting the same maturities as in 2005.
Forage yields for the 4 years of this experiment are presented in Table 1. Forage yield was nearly doubled by taking the fall harvest in 2004 with no indication of winter injury in 2005. Forage yield in 2005 was 1.52 tons/acre greater with the fall harvest and was primarily due to the fourth-harvest yield.
Forage yield in 2006 was 0.52 tons/acre greater in the fall-harvested plots than non fall-harvested plots. There was some indication that one replicate in the first harvest of the 4-cut WSI 3.0 plot had experienced some winter injury/winter kill, but the forage yield in the second, third, and fourth harvests were equal to the forage yield of the 3-cut WSI 3.0 and 4-cut AmeriStand 201 + Z.
Forage yield in 2007 was 0.33 tons/acre greater in the fall-harvested plots than non fall-harvested plots. Forage yield in the first and third harvests were significantly lower yielding in the fall-harvested plots for both cultivars than in the non fall-harvested plots. However, the fourth harvest produced 1.25 tons/acre, which resulted in greater forage yield from the fall-harvested plots for both cultivars. Forage quality was not influenced by taking the fall harvest in 2007 (Table 2).
Forage yield in 2008 was 0.71 tons/acre greater in fall-harvested plots than non-fall-harvested plots, similar to previous years (Table 1). However, none of the comparisons between fall- and non-fall-harvest treatments by cut were significant. Fall-harvested plots were injured in 2007-08 winter, which delayed the first harvest 3 to 4 days compared with the non-fall-harvested plots. Forage yield of Ameristand 201+Z in the fourth harvest was reduced some by standing water from above-normal rainfall.
Results to date indicate that fall harvest has increased forage yield by 4.78 tons/acre over the five years of this experiment (Table 3). Even though some injury occurred to falll-harvested plots in 2007-08, forage yields were 0.71 tons/acre greater from fall-harvested plots than non-harvested plots. This experiment clearly shows that when harvested at the correct stage in the fall, forage yields are increased compared to non-harvested plots. Harvest during the fall in this experiment occurred when the regrowth shoot reached about 2 to 3 inches in height in 4 of 5 years. Harvest in 2008 would have occurred following a killing frost had this experiment not been terminated.