Evaluation of Pedometers for Measuring Dis- tance Traveled by Cattle on Two Grazing Systems

The precision and accuracy of pedometers for measuring distance traveled by cattle on production size grazing systems were studied. Pedometer calibration factors were similar among~cattle, but varied because of differences in the sensitivity of pedometers to movement and/or the tightness of the case around the animal’s leg. Adjusting pedometer readings by their individual calibration factor provided a precise and accurate measure of travel distance. Cattle on a short duration grazing system tended to walk farther, and travel distance was more variable than with animals on a continuous grazing system. Travel influences the energy requirements of free grazing cattle. The energy cost of travel may increase the maintenance energy requirements of grazing livestock by IO to 24% compared to stall fed animals (Ribeiro et al. 1977, Havastad and Malecheck 1982). Anderson and Kothmann ( 1980) and Gammon and Roberts ( 1980) showed that distance traveled by cattle varied among grazing systems. Thus, animal performance may be influenced because of Authors are research associate, associate professor. and research asscriate. Department of Range Science, Texas A&M University Research and Extension Center, P.O. Box 1658. Vernon 76384. Report is published with approval of the Director. Texas Agricultural Exp. Station as TA 19126. Appreciation is expressed to Swen R. Swenson Cattle Company for providing the cattle and research facilities for this study. D.M. Anderson is acknowledged for helpful suggestions on the construction of pedometer cases. Mention of a trademark or a proprietary product does no1 constitute a guarantee or a warranty of the product by the Texas Agricultural Experiment Station and dws not imply its approval to the exclusion of other products that also may be suitable. Manuscript accepted June 7, 1984. 90 differences in energy expended for travel in different grazing systems. However, information pertaining to the effect of grazing systems on travel is scarce. Livestock travel distance has usually been estimated by charting animal movements on paper, which is a laborious and subjective procedure. Pedometers have been investigated as a low cost alternative for measuring travel distance. Powell (1968) evaluated “Suprex”pedometers on sheep and reported that pedometers must be individually calibrated to each animal and that accurate calibration could not be obtained for some animals. Anderson and Kothmann (1977) reported that digital pedometers provided a precise and inexpensive method for monitoring distance traveled by cattle. However, they did not evaluate the accuracy of pedometers and used only 2 animals to estimate precision. The objective of this series of experiments was to evaluate the reliability of pedometers for estimating distance traveled by cattle on production scale short duration and continuous grazing systems. The study was designed to identify sources of variation associated with both calibration and the measurement of distance traveled by grazing cattle. Methods and Materials Five experiments were conducted to test the accuracy and precision of pedometers for measuring distance traveled by cattle. The study was conducted at the Texas Experimental Ranch located in the Rolling Plains Resource Region. The cattle used in this study were mature Hereford X Angus cows, except for Exp. 2 where 3 JOURNAL OF RANGE MANAGEMENT 38(l), January 1985 Hereford cows and 2 Hereford X Simmental steers that had been esophageally list&wed were used and Exp. 5 where a horse was used. The instruments used in this experiment were “Digi-Meter” mometers (manufactured by Edge Mark, Japan) and were the same model pedometer evaluated by Anderson and Kothmann (1977). Vertical movement of the pedometer activates a pendulum mechanism that is mechanically linked to a digital readout. The pace length adjustment was standardized among pedometers using a mechanical shaker. The setting was maintained by securing the pace length adjustment screw with silastic (silicon sealant, General Electric, USA). Pedometers were enclosed in a case constructed of 4.8-mm thick acrylic plastic (Fig, I). The case was attached around the metacarpus of the forelee with a 37.mm wide elastic lee band to mevent . I Pedometers were calibrated on Zconsecutivedays. Calibration was replicated 3 times the first day and 2 times the second day. Pedometer readings were recorded both days. Calibration factors were analyzed for the random effects of animal, pedometer/animal, day, and all two-way interactions. Travel data were analyzed for the random effects of animal and pedometer/animal with days as replicates. Experiment 3 and 4-These experiments were conducted to evaluate thereliabilityof pedometersforestimatingtraveldistance in production scale grazing systems. Travel was monitored concurrently in both treatments. Grazing systems studied were short duration(SDG)andcontinuousgrazing(CG).TheSDG treatment contained I6 pastures fenced in a cell design on 450 ha and stocked with I25 cows. Pastures ranged in size from IO to 30 ha. The CC pasture was 242 ha and was stocked with 41 cows. During both experiments, the cattle in the SDG treatment were rotated through the same 5 pastures (two 30.ha pastures and three IO-ha pastures) that were selected previously to monitor other animal and vegetation parameters. Exp. 3 lasted 6 days and Exp. 4 lasted 9 days. The length of the experiments was determined by the time required to rotate through the SDG pastures, whichvaried inaccordance with the growth rate of the vegetation (V&in 1959). Ten cows in each treatment were equipped with I pedometer on their right foreleg. A different set of cows was used for each experiment. Pedometers werecalibrated 2 times beforeand 2 times after each experiment except for the pedometers on the SDGcattle in Exp. 3, which were calibrated only at the end of the experiment. Pedometers were read at the beginning and end ofeach experiment as well as between pastures in the SDG treatment. Travel in each pasture was averaged to determine daily travel for the SDG cattle. Calibrationfactorswereanalyzed foronlyfor6ofth.z CGcattle in Exp. 3 due to incomplete data caused by lost or broken pedometers. Likewise calibration factors were available from 7 and 5 of the SDG and CG cattle, respectively, in Exp. 4. Calibration factors from Exp. 3 were analyzed with a mixed model for the effects of animal (random), time (i.e., beginning vs end, fixed), and their interaction. The analysis of the calibration factors for Exp. 4 was also a mixed model with treatment and animal nested within Fig. 1. Pedomerer enclosed m r? plexi&Is cnre rhor a oi,oched f0 0 COU~k treatment(animal/ treatment)as randomeffectsand timeas a fixed f0r&g. effect. The effect of grazing treatment on average daily travel distance was tested using a f-test for independent samples with slippage on the animal’s leg. The inside surface of the leg band and case were padded with foam rubber to prevent abrasion. Calibration factors were calculated by dividing the travel distance measured on the pedometer by the actual distance walked. Animals were walked I.6 km along a fence to determine the calibration factor. Pedometer readings and the time of reading were recorded daily in Exp. I and 2 and at the beginning and end of Exp. 3 and 4. Average daily travel distance was calculated by dividing the difference between the beginning and ending pedometer reading by the individual calibration factor and then adjusting this distance to a 24-hr basis. Experiment l-Ten cows were equipped with I pedometer on eachfront leg.The studyareawasa39-hapasturethatwaspart ofa l6-pasture shortdurationgrazingcell. Pedometers were calibrated once a day on 3 consecutive days beginning on the first day of the experiment; distance traveled was recorded daily for 2 days. Three cows were not included in the analysis because their pedometers were either lost or broken during the study. Calibration factors and travel data were analyzed for the effect ofanimals, and pedometers nested within animals (pedometers/animals) with both factors considered as random effects and days considered as replications. Analvsis of variance for this and subseauent experiments were unequal variances (Sned& and Cochran i967). Observations of cattle locations were used to provide an indication of the accuracy of the pedometer measurements for the CG cattle in Exp. 3 and 4. The location, activity and number of animals in each subherd were recorded hourly during daylight on topographical maps. Travel distance was estimated by measuring the change in each subherd location between successive observations. The only estimate of nighttime travel was the change in location between the last observation before dark and the first observation of the next day. This estimate of travel was used only for the CG cattle because the high density of the SDG cattle rendered such measurements impractical. Experiment S.-This experiment was conducted to estimate the effect of an animal’s gait on the calibration factor. Because it was considered impractical to keep cattle in any gait other than a walk over the distances necessary to obtain reliable pedometer readings, a horse was used for this experiment. The horse was ridden for 1.6 km at a walk, a trot, and a lope with one pedometer attached around the metacarpal of each front leg. Calibration was only performed once and the data were not analyzed statistically, perfdrmed using the general linear model; procedire of statistical In Exp. I and 2 variation in the calibration factor could not be analysis system (Helwig and Council 1979). attributed to differences in animals, day of evaluation, or their Experimenf 2-Five docile esophageally fistulated animals (3 interaction (Table I). The only significant source of variation for Hereford cows and 2 Hereford X Simmental steers) were equipped pedometer calibration was caused by the pedometer/animal with a pedometer on each leg. The study area was a 6-ha pasture. source. There were 2 possible causes for this variation: (I) pedomeJOURNAL OF RANGE MANAGEMENT 38(l), January 1985 91 Table 1. Sources of variation, mean squaresand aasoctated probabilities of greater F-value in 4 experiments evaluating correction factors for caliTable 2. Sources of variation, mean squares and associated probabilttks of brattng pedometers on cattle. greater F-values for random model analyses of variance of average datly travel distance of cattle in experiments 1 and 2. Source of variation d.f. MS P Animal Pedometer/ Animal Residual Animal Pedometer/ Animal Day Animal X Day Day X pedometer/animal Residual Animal’ Time* Animal X Time Residual Treatment’ Animal/ treatment’ Time* Treatment X Time Residual ‘Random effect ZFixed effect (beginning vs end) Experiment I (Random Model) 6 .099 .57 7 .I17 .OO 28 .020 Experiment 2 (Random Model) 4 .014 .49 5 .014 .04 1 .ooo .83 4 .004 .27 5 .002 .40 30 .002 Experiment 3 (Mixed Model) 5 445 .06 1 .139 .lO 5 .035 .94 II .150 Experiment 4 (Mixed Model) 1 .121 .63 10 .506 .03 I .657 .I2 I .024 .60 10 .I45 Source of variation Animal Pedometer/animal Residual Experiment 1 Experiment 2