Data from 44 adult giraffes (Giraffa camelopardalis) showed that leg weight as a percentage of body weight is sufficiently constant to provide regression equations useful for estimating body or carcass weight. Data from 53 giraffes of various ages show that reliable estimates of body weight can be made from linear body measurements, regardless of sex or age. The best for field use are linear regressions between body weight and chest girth; however, shoulder height is usually the only measurement obtained from predator kills and this can also be used. Measures of the volume of the animals such as length x girth' and log transformations give superior correlations. J. WILDL. MANAGE. 41(4):740-745 A high correlation between dissected leg weight and liveweight in Boran steers (Bos indicus) was reported by Ledger (1965) and used for predicting liveweight in cattle and 8 species of wild African ungulates (Smith and Ledger 1965). This relationship was also found to hold good for African elephants ( Loxodonta africana) by Laws et al. (1967) and was regularly used for estimating body weight of this species (Laws et al. 1975). Relationships between standard body measurements and weight for domestic and wild African animals have also been investigated (Talbot and McCulloch 1965, McCulloch and Talbot 1965, Freeman and King 1969, Hanks 1972, Bertram 1975, Howells and Hanks 1975, Smuts 1975) and expressions have been derived for weight estimations. These methods find ready application in estimating predator or scavenger consumption, studies of meat production, nutrition, growth rates, condition, and estimates of weight for drug dosage rates and the transport of live animals. The purpose of this study was to investigate some of these relationships for the giraffe and to find reliable criteria for weight estimation applicable to live and butchered animals as well as animals killed by lions (Panthera leo). Appreciation is extended to the owners of the Timbavati, Sandringham, and Buffelshoek nature reserves for making the material available for this study. Professor J. D. Skinner is also thanked for his help as are the University of Pretoria and the Council for Scientific and Industrial Research, South Africa, for financial assistance. MATERIAL AND METHODS Material was obtained from culling and research programmes in the woodland/savanna regions of the eastern Transvaal Lowveld of South Africa. Immediately after shooting and bleeding the giraffes were measured as described below, then dissected and weighed in pieces as described by Hall-Martin et al. (1977). As this was done in the field, taking 2-3 hours, it was not possible to make allowance for weight loss due to evaporation nor for blood and body fluids lost during handling. The sum of the weights of all other body components are termed "body weight." The term "foreleg" as used here accords with the definition of Ledger (1963), the "hindleg" is as meant by Laws et al. (1967), while "carcass" is the body less skin, head, tail, feet, viscera, kidneys, and kidney fat (Hall-Martin et al. 1977). A regularly calibrated Salter spring 1 Present Address: Mountain Zebra National Park, Private Bag X66, Cradock 5880, South Africa. 740 J. Wildl. Manage. 41(4):1977 This content downloaded from 157.55.39.181 on Thu, 29 Sep 2016 05:50:36 UTC All use subject to http://about.jstor.org/terms GIRAFFE WEIGHT ESTIMATION * Hall-Martin 741 Table 1. Weights of giraffe hindlegs, forelegs and stomachs (including contents) as percentages of body weight. Stomach plus Hindleg as Foreleg as contents as percent Body weight percent of percent of of body weight Age rangea kg body weight body weight Sex N (years) (range) (Y SD) ( SDo) (t SD) (range) Male 19 8-23 849-1,395 8.8 -+0.42 8.0 ? 0.36 11.5 t 1.3 8.5-13.6 Female 25 6-20 636950 8.9 +. 0.41 7.4 + 0.39 13.9 ? 2.8 8.2-20.7 a Age determination as described by Hall-Martin (1976). balance with a capacity of 200 kg was used. Fifty-seven giraffes were: weighed and linear measurements of 53 of these were recorded. Linear measurements made on animals in lateral recumbency were total body length (L) which included head and tail, shoulder height (H) measured in a straight line from the level of the hoof to the top of the withers, and chest girth (G) taken just caudal to the front legs. Standard correlation and regression procedures (Simpson et al. 1960, Sokal and Rohlf 1969) were used to measure the degree of correlation between variables, to provide predictive functions and to judge the accuracy with which one variable may be predicted from another. For paired comparisons the Students t-test (Simpson et al. 1960) was used. RESULTS AND DISCUSSION Weight Estimation Using Dissected Leg Weights Body weights and dissected leg weights of 19 adult male and 25 adult female giraffes were examined. There was a reasonably wide range of body weights due to age, sex, variations in stomach fill and seasonal influences (Hall-Martin et al. 1977); giraffe lose condition as the quality of their diet declines during the dry season (Hall-Martin 1975). Nevertheless, the leg weights as percentages of body weight showed little variation (Table 1). However, there was a significant decrease in percentage hindleg with age in the males (P < 0.01) but no such difference in the foreleg or among the females. This difference has been explained by HallMartin et al. (1977) as being due, in part, to the disproportionate increase in size of the neck in the older males. It was therefore appropriate to calculate regression equations of leg weights on body weights and carcass weights for males and females separately (Fig. 1). These equations, together with correlation coefficients are presented in Table 2. The mean hindleg or foreleg weight was used in each case as it was shown that there were no statistically