Regression formulas show close relationships between weight and linear measurements of mountain goats (Oreamnos americanus) captured in the Sapphire Mountains of Montana. Residual mean squares were used to compare six equations correlating weight and linear measurements. The most dependable estimates are (1) log weight = 1.2574 (log length + log girth) 2.1531, and (2) weight = 0.0011 (length X girth2) + 11.5 (measurements in English system). These formulas appear to estimate weight accurately regardless of sex or age. Girth should be recorded regularly in future studies of North American ungulates to permit calculation of similar formulas for other species. J. WILDL. MANAGE. 39(4):705-708 The mountain goat has been the subject of relatively few wildlife studies, largely due to the inaccessibility of its habitat. Very few of these wildlife studies have employed any capture methods; those which did generally were associated with mountain goat transplant operations (Cooney 1946, McDowell 1949, P-R Job Prog. Rep., Proj. 5-D-7, Montana Fish Game Dept., Helena, Stockstad 1959, Rogers 1960, Charles 1961, Wadkins 1967, P-R Job Prog. Rep., Proj. W66-R-6, Washington Dept. Game, Olympia), and measurements seldom were taken. Whole weights of both live and dead animals were recorded in Montana (Casebeer et al. 1950, Brandborg 1955, Lentfer 1955, Foss 1962) and in Alaska (Klein 1953); hog-dressed weights were compared with weights of live mountain goats in Montana (Lentfer 1955). Since linear measurements are much easier to take than whole weights, the objective of this paper is to provide an accurate method for estimating mountain goat weights from body measurements. Similar weigh estimation formulas have been used for domestic livestock (Manning and Will ams 1950, Ross 1955, 1958) and for Afric n game animals (McCulloch and Talbot 1965, Talbot and McCulloch 1965, Freeman and King 1969), but not for North American game animals. Suitable regression equations for other big game species would permit reliable whole weight estimates from field measurements. We thank G. D. Maclean, P. M. Smith, and J. T. Watkins for assisting the senior author in the trapping program and C. Smith, Montana Fish and Game Department, for providing advice and supplies. We also are indebted to N. A. Slade, Museum of Natural History, University of Kansas, for statistical advice and to R. S. Hoffman, Museum of Natural History, University of Kansas, for reviewing the manuscript. This study was supported by the National Rifle Association, Theodore Roosevelt Memorial Fund, Watkins Fund of the University of Kansas, National Science Foundation, National Aeronautics and Space Administration, Montana Fish and Game Department, American Forest Institute, and University of Kansas Department of Systematics and Ecology. ' Present address: Biology Department, Indiana University Northwest, 3400 Broadway, Gary 46408. 2 Present address: UMC 37, Utah State University Computer Center, Logan 84322 J. Wildl. Manage. 39 (4):1975 705 This content downloaded from 157.55.39.70 on Tue, 06 Sep 2016 05:25:37 UTC All use subject to http://about.jstor.org/terms 706 ESTIMATING MOUNTAIN GOAT WEIGHTS * Rideout and Worthen Table 1. Residual mean square values of six weight estimation formulas for the mountain goat. EquaResidual tion mean number Formula square (1) W = 5.2458 G 95.5 105.526 (2) W = 4.2394 L 118.1 114.201 (3) W = 0.0609 LG 19.7 47.572 (4) W = 0.0011 LG2 + 11.5 43.337 (5) log W = 1.2574 (log L + log G) 2.1531 42.264 (6) log W = 0.8243 (log L + 2 log G) 2.0248 46.725 MATERIALS AND METHODS Mountain goats were captured in a pen trap in the Sapphire Mountains of Montana during the period 1971-73 (Rideout 1974a, b) for instrumentation in a radiotelemetry study of mountain goat movements (Rideout 1973, 1974b). Linear measurements, in inches, were obtained with a flexible steel tape, and weights were taken with a Hanson spring scale with 200-pound capacity and recorded to the nearest pound. Measurements were taken of 27 goats captured from June to November. Total handling time, during which the animals were marked, measured, and released, was 30-45 minutes. The mountain goats used in this study were taken from one herd and included a fairly complete cross-section of the population with regard to both sex and age. The measurements considered in this paper include (1) total length (L), measured from the tip of the nose to the tip of the tail along body contours, (2) girth (G), measured around the chest directly behind the forelegs, and (3) whole weight (W). Linear regression techniques (Sokal and Rohlf 1969) were used for weight estimation formulas. Calculations were carried out with the Honeywell 635 computer at the University of Kansas with programs written primarily by the junior author. Six equations were compared for best estimates of the weight by examining the residual mean square values (unexplained sums of squares /degrees of freedom). In the case of log transformed values, Y values were con180 -
Read full abstract