Forage Yield Stability of Fertilized Bermudagrass in the Great Plains.

Our objectives were to document the long-term effects of N, P, and K fertilization on forage yields of bermudagrass (Cynodon dactylon (L.) Pers.).  Fertilizer treatments were applied annually during spring to 3- by 6-m plots on a sandy loam site in south-central Oklahoma from 2002 to 2007.  Plots were harvested two to four times per year for determination of annual forage yields.  Yield stability was assessed through regression of within treatment means on the yearly environment mean (mean among treatments).  Fertilizer treatment by year interactions affected forage yields.  Nitrogen fertilizer rate increased forage yields in 2003 and 2004, decreased yields in 2005, and had no effect in 2002, 2006, and 2007.  Stability analysis revealed that N-fertilized bermudagrass was highly responsive to the environment (P₂O₅ and K₂O held constant at 45 and 112 kg ha^-1, respectively).  Bermudagrass treated with 224 to 336 kg N ha^-1 had slope coefficients of 1.36 to 1.42, indicating high yields when environment was favorable but less-than-average yields when environment was poor.  Yields were most stable at 112 kg N ha^-1.  At this rate, annual forage yields were among the largest, 11.5 Mg ha^-1, compared to a mean across treatments of 10.7 Mg ha^-1, the slope coefficient was 1.04, and deviations from regression were close to zero.  Non-N fertilized bermudagrass did not respond to environment as indicated by a slope coefficient of 0.50.  Deviations from regression were also greatest for 0, 224, and 336 kg N ha^-1 rates.  Neither split-application of N or varying P and K independently of N affected forage yield stability.  Examination of records from 1895 to 2007 revealed average annual precipitation of 954 mm.  From 2002 to 2007, precipitation was 822, 478, 997, 535, 724, and 865 mm, respectively.  Larger responses to fertilizer may be evident in years with better precipitation.