Submitted to: American Forage and Grassland Council Conference Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: 11/15/2015
Publication Date: 1/10/2016
Citation: Soder, K.J., Hafla, A., Rubano, M.D., Stout, R.C. 2016. Application of a rising plate meter to estimate forage yield on dairy farms in Pennsylvania[Abstract]. American Forage and Grassland Council Conference Proceedings Annual meeting, January 10-13, 2016, Baton Rouge, LA. p. 1.
Interpretive Summary: An interpretive summary is not required.
Technical Abstract: Accurately assessing pasture forage yield is necessary for producers who want to budget feed expenses and make informed pasture management decisions. Clipping and weighing forage from a known area is a direct method to measure pasture forage yield, however it is time consuming. The rising plate meter (RPM) is a rapid and indirect method of estimating standing forage yield. Readings from the RPM are converted to dry matter (DM) forage yield using an equation either provided by the manufacturer or from user calibration (with forage clippings). The RPM requires frequent calibrations and default equations provided by manufactures may be unreliable. Therefore, the objective of this study was to evaluate the ability of a RPM to accurately estimate pasture forage yield on dairy farms with multiple plant species and determine the best calibration equation for these conditions. The three Pennsylvania farms used in this study utilized rotational grazing with lactating dairy cows. Forage mass was estimated in each pasture one day prior to grazing between August 23 and November 16, 2012 using 45 measurements, from a FILIPS RPM (n=180-225 readings per farm). To obtain an objective measure of forage DM production, 15 clippings of 3 foot by 4 inches were collected at corresponding measurements of the RPM in each pasture (n=60-75 clips per farm). Visual estimates of botanical composition were made within each pasture to assess species diversity. Equations for estimating pasture forage mass were determined by regressing measured DM yield on the corresponding RPM value to produce a linear equation. Four calibration equations were evaluated: (1) all measurements pooled from all farms; (2) measurements adjusted by farm; (3) measurements adjusted by season (summer or fall, based on calendar date); and (4) measurements adjusted for farm and season. Additionally, a default equation and an equation that considered season, both provided by the manufacturer of the RPM, were evaluated. Equations were evaluated by regression procedures (PROC REG; SAS Inst., 1998) and the estimated standard error of prediction was calculated. Average measured DM forage yield was 1187, 880, and 2205 lb/acre for farms 1, 2 and 3, respectively. Grasses made up 23, 44, and 63% of pasture composition, while legumes composed 48, 25, and 20% of pasture composition, respectively, for farms 1, 2 and 3. Weeds made up 21% of total composition on each of the farms. Equations provided by the manufacturer (default and seasonal) resulted in the greatest level of error for estimating forage yield (error of 32 and 38% of mean forage mass measured, respectively) and low predictability (R2=0.58 and 0.51, respectively). Estimating forage yield using the calibration equations greatly reduced error. Adjusting for individual farm produced an error of 21%, and the lowest R2 (0.49) of all equations evaluated. Using all measurements across farms resulted in a smaller error of 12%, but overestimated forage yield by 161 and 98 lb/acre and underestimated yield by 185 lb/acre, on the three farms and had a low R2 (0.58). Including farm and season in the equation resulted in the greatest R2 (0.78) of all the equations evaluated, however still produced an error of 13%. The equation which considered season only resulted in the smallest error (9%), a high R2 (0.76) but overestimated forage yield by only 13 and 85 lb/acre and underestimated forage yield by 228 lb/acre on farms 1, 2 and 3, respectively. Previous research has indicated that error levels of < 10% were necessary to justify a farmer’s investment in labor and tools for measuring forage mass on pastures, and regressions with very low R2 (0.31) were not sufficient to accurately predict pasture forage yield. The results of this study indicate that manufacturer equations were unreliable for estimating DM forage yield on farms in Pennsylvania when pa