|
|
|
© 2007 Plant Management Network. Automating Measurement of Forage Mass in Pasture J. L. Moyer, Southeast Agricultural Research Center, Kansas State University, Parsons 67357; and M. D. Schrock, Department of Biological & Agricultural Engineering, Kansas State University, Manhattan 66506 Corresponding author: J. L. Moyer. jmoyer@ksu.edu Moyer, J. L., and Schrock, M. D. 2007. Automating measurement of forage mass in pasture. Online. Forage and Grazinglands doi:10.1094/FG-2007-0419-01-RS. Abstract Measuring forage mass (FM) is crucial in grazed pastures, but the process is labor intensive. Several indirect methods of estimating standing FM are used, but all require a large number of observations for an adequate sample. One popular technique relates FM to height of a disk or plate compressing the forage beneath. This report describes and evaluates a forage estimator "sled" that continuously records the height of a pivoting plate and the corresponding global positioning system (GPS) readings as it is pulled over forage. Tall fescue (Festuca arundinacea Schreb.) pasture FM was correlated with plate height (r = 0.85, P < 0.01). Calibrations relating disk meter and plate height data from tall fescue and smooth bromegrass (Bromus inermis L.) pastures were developed. Tall fescue varietal differences in FM of pastures were similarly detected by disk meter and the forage estimator sled. Introduction Measuring forage mass (FM) is often crucial in evaluating the impact of grazing strategies. Further, determination of an appropriate stocking rate depends almost entirely on accurately estimating FM. Sanderson et al. (14) demonstrated the importance to pasture dairy operators of accurately estimating FM, but its measurement can be a labor-intensive process that may result in data of questionable accuracy. Many methods have been used to estimate the amount of FM in a paddock. The basis of most is to remove and weigh the forage from sample areas in the paddock (i.e., to take a destructive sample). Nondestructive estimates of various kinds have been employed, but they must eventually be calibrated, a process known as double sampling (16). The accuracy of all methods depends on the calibration process, but the precision depends on the number of samples taken, because those samples must represent the paddock sampled in order to describe it. Within paddock variability often differs among paddocks, requiring different numbers of samples for accurate measurement. One approach is to estimate variability during measurement (6), but typically the number of samples is based on the maximum variability that might be encountered. Estimating Forage Mass Nondestructive estimates that can be taken rapidly are preferred so that they are practical for obtaining large numbers of observations. Visual estimates are among the most rapid, can be precise when the operator is consistent, and accurate if the operator’s eye is well trained with a sward of known mass (16). A forage "stick" to relate average forage height to biomass is often used because more objective numeric data are obtained. A more sophisticated visual method uses a graduated "obstruction pole" (12). This method proved more accurate than the stick height method (5). Weighting height measurements with cover estimates (10) increased the correlation with herbage weight in some situations. More technically sophisticated methods of indirectly measuring FM have been developed, including the use of radio frequency capacitance (8), light attenuation (2), and most recently spectral reflectance (9,15). Of these methods, the capacitance meter has been most widely evaluated to this point. Murphy et al. (7) found the capacitance meter comparable with other indirect methods, but Sanderson et al. (14) found it and a pasture ruler highly inaccurate in estimating herbage mass. Other techniques use physical pressure from a type of plate or disk to compress the forage, and relate the compressed height to FM. Phillips and Clarke (11) calibrated a weighted disk against FM, and Bransby et al. (1) simplified the instrument, which has been used in some form by many researchers and practitioners. Such instruments are inexpensive and simple to build, and measurements can be obtained easily and rapidly. Disk or plate meters have been shown to be similar or superior to other indirect methods of measuring available forage (4,5,7,14), but recalibration may be needed with different instruments and operators, or as vegetation type, stage, or growth habit change. As indicated earlier, greater within pasture variability requires more indirect samples to adequately determine the pasture mean (6). Increasing the number of samples improves the precision of any measurement. Sanderson and Rotz (13) recommended that plate meter readings be taken at 30 to 50 locations in a typical 1- to 2-acre paddock in the northeastern USA. A rising-plate meter with the capability of logging georeferenced data to a personal data assistant (3) facilitates rapid data collection, but still relies on quadrat readings. This report describes a forage estimator apparatus that was designed to objectively measure compressed forage height on a relatively continuous basis across a paddock area, and to periodically record its readings in an electronic digital format. An Automated, Continuous Forage Estimator The forage estimator "sled" (Figs. 1 and 2) is an apparatus based on a pressure plate attached to a pivoting arm. Mounting on sled runners rather than wheels or tracks maintains a more accurate, uniform reference to the soil surface. The runners were each constructed from 6¼ ft of ½-inch steel rod, with a curved radius of 24 inches in the front. The sensing plate was constructed of a 2- × 4-ft sheet of 1/8-inch high-density polyethylene mounted to a curved metal frame that was attached at the front to a shaft 1½ ft above the ground. The force required to raise the plate was designed to approximate the pressure of a disk meter plate (1). Overall mass was about 60 lb and an additional 60 lb was added to a ballast tray near the rear of the sled. The plate’s pivoting arm was mechanically connected to a standard automotive throttle plate angle sensor which is a potentiometer that converts its position to voltage. The potentiometer reports to computer software (Labview, National Instruments, Austin, TX) via a USB port. A GPS unit (Garmin Model GPS 18 PC; WAAS differential, with a manufacturer claim of 1.7-m root mean square) was programmed to collect 10 readings per second and report the average each second to the computer software, when time, location, speed, and distance from it and plate height from the potentiometer are recorded on a spreadsheet. The measurements of plate height can be converted to estimates of FM by calibration with harvested forage samples, or by cross calibration with the disk meter or other indirect methods of measurement.
Comparing Measurements of Tall Fescue Comparisons of FM, disk meter readings, and sled plate height were made on 16 August 2005 in grazed tall fescue pastures at the Mound Valley Unit of the Kansas State University Southeast Agricultural Research Center (SEARC). Four replicated pastures of each of two cultivars in a grazing study were sampled along 10-ft transects. Three disk meter readings were taken along each of two transects per pasture, and the forage estimator sled was pulled beside each disk meter transect and recorded 10 readings. The disk meter was similar to one described earlier (1), except that ours was constructed of plexiglass (6). All four transects from each pasture were cut with a 3-ft flail harvester at a 2-inch height. Actual dry FM harvested, disk meter readings, and plate height readings for the forage estimator were averaged for each pasture. Forage mass is reported as measured by the flail harvester, disk meter readings, and the forage estimator (Table 1). Average FM harvested from pastures of the two cultivars differed by 18.5%, whereas the forage estimator indicated a difference of 13.9%, but the —1% difference indicated by disk meter readings was not significant (P > 0.05). Disk meter sample variance from the analysis of pasture means (data not shown) indicated that the four replicates should have detected a difference of 17.8%, according to the formula from Moyer and Higgins (6). Sample variances within pastures differed widely, however, providing theoretical detection limits of from 8 to 61% from the six observations per pasture unit. Coefficients of variation (CV, Table 1) indicated that the two estimation methods were at least as precise as the direct FM measurements. Simple correlation coefficients (r) calculated from pairwise comparisons among the three measurements showed them to be positively related, but only the correlation between FM and sled plate height were statistically significant (P < 0.01). The relationship between harvested FM and plate height of the forage estimator is shown in Figure 3. Table 1. Forage mass comparisons of KY 31 tall fescue pastures
Comparisons of disk meter readings and sled plate height were made on subsequent monthly dates (September to December) at the same location. Besides the pastures measured earlier, four replicates of each of the two remaining cultivars at the location were included. Disk meter readings were taken from four random triplets (measurements at the points of an equilateral triangle spaced about 10 ft apart) per 5-acre pasture. Readings were averaged from each of the 16 pastures, and compared with pasture averages obtained with the forage estimator sled. The relationship, which could be used for cross calibration between the two methods, is shown in Figure 4. The means at different dates appeared clustered, with reduced FM as the fall progressed. Curve coefficients may have varied with date, but could not be demonstrated with the limited number of comparisons. Recalibration at different growth stages to improve accuracy may be necessary.
Analysis of variance of FM by date indicated similar comparisons (P > 0.05) among the four cultivars for the two methods of estimation (data not shown). September rankings of the four cultivars were the same for both estimates, but the disk meter separated the highest-ranking cultivar from the other three, while the sled separated only the highest from the lowest. The least significant difference (LSD, α = 0.05) was 10.5% of the mean by using the disk meter forage estimates, whereas the LSD for the sled estimates was 8.8% of the mean. In November, the ranking and the separations of the four cultivars were identical for the two methods, but LSD values were 23.0 and 42.9% of the means for the disk meter and sled, respectively. Comparing Measurements of Smooth Bromegrass Nine lightly grazed pastures of smooth bromegrass at the SEARC Parsons Unit were measured in a manner similar to measurements of the tall fescue pastures at Mound Valley. The last two monthly measurements for FM were performed by using both the disk meter and the forage estimator sled. The relationship of pasture averages obtained with the disk meter, compared with those with the forage estimator, is shown in Figure 5. No separations of treatment means were made by either method. The LSD for bromegrass FM calculated from disk meter readings was 28.8% of the mean for September measurements and 18.8% for October. The LSD for plate-height measurements from the sled in September was 28.6% of the mean, and for October was 45.8% of the mean. It should be mentioned, however, that LSD values for raw, uncalibrated disk meter readings were 52.5 and 36.5% of their means for September and October, respectively. Thus, calibration and curve-fitting of bromegrass FM may also improve accuracy of sled estimates. The forage estimator sled was useful for rapidly obtaining relatively continuous readings of forage plate height across a pasture. The comparisons were similar to those obtained from disk meter readings that could be obtained in about the same amount of time. Combined with the GPS coordinates recorded with each reading, the data can also be used to describe spatial variability of FM across a landscape, either graphically or statistically.
Acknowledgments The authors acknowledge the contributions to the design and construction of the forage estimator of the following students: Brian Bretz, Rex Schertz, Ross Rieschick, Kyle Shaffer, and Eric Bussen. This is contribution no. 07-59-1 from the Kansas Agricultural Experiment Station, Manhattan, Kansas. Literature Cited 1. Bransby, D. I., Matches, A. G., and Krause, G. F. 1977. Disk meter for rapid estimation of herbage yield in grazing trials. Agron. J. 69:393-396. 2. Engel, R. K., Moser, L. E., Stubbendieck, J., and Lowry, S. R. 1987. Yield accumulation, leaf area index, and light interception of smooth bromegrass. Crop Sci. 27:316-321. 3. Flynn, E. S., Dougherty, C. T., and Koostra, B. K. 2006. GPS-enabled rising plate meter with data logging capability. Online. Forage and Grazinglands doi:10.1094/FG-2006-825-01-BR. 4. Ganguli, A. C., Vermeire, L. T., Mitchell, R. B., and Wallace, M. C. 2000. Comparison of four nondestructive techniques for estimating standing crop in shortgrass plains. Agron. J. 1211-1215. 5. Harmoney, K. R., Moore, K. J., George, J. R., Brummer, E. C., and Russell, J. R. 1997. Determination of pasture biomass using four indirect methods. Agron. J. 89:665-672. 6. Moyer, J. L., and Higgins, J. J. 2005. Calculation of sample number to accurately measure available pasture forage. Online. Forage and Grazinglands doi:10.1094/FG-2005-1123-01-BR. 7. Murphy, W. M., Silman, J. P., and Mena-Barreto, A. D. 1995. A comparison of quadrat, capacitance meter, HFRO sward stick, and rising plate for estimating herbage mass in a smooth-stalked, meadowgrass-dominant white clover sward. Grass Forage Sci. 50:452-455. 8. Neal, D. L., and Neal, J. L. 1973. Uses and capabilities of electronic capacitance instruments for estimating standing herbage. I. History and development. J. Brit. Grassl. Soc. 28:81-89. 9. Olson, K. C., and Cochran, R. C. 1998. Radiometry for predicting tallgrass prairie biomass using regression and neural models. J. Range Manage. 51:186-192. 10. Pasto, J. K., Allison, J. R., and Washko, J. B. 1957. Ground cover and height of sward as a means of estimating pasture production. Agron. J. 49:407-409. 11. Phillips, S. M., and Clarke, S. F. 1971. The calibration of the weighted disk meter against pasture dry matter yields. Proc. NZ Grassl. Assoc. 33:68-75. 12. Robel, R. J., Briggs, J. N., Dayton, A. D., and Hulbert, L. C. 1970. Relationships between visual obstruction measurements and weight of grassland vegetation. J. Range Manage. 23:295-297. 13. Sanderson, M. A., and Rotz, C. A. 2004. Precision pays. The Forage Leader 9:14. Amer. Forage Grassl. Coun., Abilene, TX. 14. Sanderson, M. A., Rotz, C. A., Fultz, S. W., and Rayburn E. B. 2001. Estimating forage mass with a commercial capacitance meter, rising plate meter, and pasture ruler. Agron. J. 93:1281-1286. 15. Starks, P. J., Zhao, D., Phillips, W. A., and Coleman, S. W. 2006. Herbage mass, nutritive value and canopy spectral reflectance of bermudagrass pastures. Grass Forage Sci. 61:101-111. 16. Wilm, H. G., Costello, D. F., and Klipple, G. E. 1944. Estimating forage yield by the double-sampling method. Agron. J. 36:194-196. |
||||||||||||||||||||||||||||||||||||||||
