Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science

EFFECT OF NITROGEN FERTILIZATION AND STORAGE METHOD ON THE AMOUNTS AND PROPORTIONS OF PROTEIN FRACTIONS IN TIFTON 85 BERMUDAGRASS (Cynodon spp.)

By

Stuart J. Rymph

August 1999

Chairman: Dr. E.C. French III

Major Department: Agronomy

Dynamic computer models of rumen digestion and absorption based on the Cornell Net Carbohydrate and Protein System (CNCPS) provide the ability to predict N utilization by ruminants with greater precision than allowed by previous systems. Information required by these models about the protein profiles (fractions) of feeds, especially tropical forages, however, is limited. A study conducted at Ona, FL evaluated the effects of N fertilization and ensiling on these protein fractions. Analyses specified for these computer models were used to measure the protein fractions. Increasing N fertilization and ensiling both increased the proportion of A fraction protein (non-protein nitrogen or NPN) and decreased the proportion of B3 protein (available cell wall N). Forages ensiled at very low dry matter (DM) levels likely experienced clostridial fermentation and exhibited extensive degradation of B3 protein. The degradation products contributed to the observed increase in A fraction protein. The proportions of B1, B2, and C fractions were generally stable regardless of N fertilization rate or type of fermentation. The availability of laboratory analyses for these protein fractions is currently limited. Farmers and consultants using these models must rely on "reference values" provided with the models for protein fraction inputs. This study demonstrated that the proportions of these fractions was not static and could be influenced by management practices. Results of this experiment were used to establish a more dynamic method of estimating protein fractions in bermudagrass silages with the capability of accounting for the effect of management effects. Several methods were developed and their estimates tested against the results of a study conducted at Bell, FL during the summer of 1997. Results of the testing indicate that the use of 2.34%, 19.01%, and 5.32% of total N (the means of from the Ona data) as estimates for the B1, B2, and C fractions, respectively, provided accurate estimates of these three fractions. Use of the regression model

B3 Protein (% of total N) = 0.574(DM)+0.293(CP)+0.720(NDF) -58.1,

based on the DM, CP, and NDF levels of the silage, accounted for over 80% of the variation in the B3 fraction protein (% of total N) observed in the test data set. Estimation of the proportion of A fraction protein as the difference between 100 and the sum of the other four fractions provides the complement to the rest of the estimates and accounts for over 30% of the variation observed in the test data. Laboratory analyses provide the most accurate estimate of the protein fraction profile of a forage. Until such analyses are more commonly available, the suggested system for estimating these fractions will provide a more flexible and accurate representation of these nutritive properties of bermudagrass silage than the best current reference values available – those provided with the CPM-Dairy computer model.