Predicting goat body weight using morphometric measurements: A regression approach

Abstract

The study was conducted local farms in Pabna district, Bangladesh, to
develop an accurate regression-based model for predicting goat body
weight. A total of 49 castrated male goats from four available breeds
were measured to assess the relationship between live body weight and
various body dimensions. The objective was to identify the most reliable
body measurements for predicting live weight and to formulate both
simple and multivariate regression models suitable for field use. Body
weight was recorded using a digital weighing scale, and 13
morphometric traits, including Forehead Diameter, Muzzle Length,
Neck Length, Neck Diameter, Heart Girth Diameter, Point of Shoulder
to Knee length, Knee to Metacarpal length, Hip Bone to Hock Joint
length, Hock joint to Metatarsal length, Wither Height, Rump Height,
Hook to Hook length, Pinto Pin length were collected. Simple linear
regression was performed for each trait, followed by multivariate
regression to determine the best-fitting prediction equations. Among all
measurements, wither height, rump height (R2 = 0.97), and point of
shoulder to knee length (R2 = 0.94) showed the strongest individual
relationships with body weight. A full multivariate model using all traits
achieved an adjusted R2 of 0.98, while a simplified model using only the
three best predictors produced an adjusted R2 of 0.98, making it more
practical for field conditions. The study concluded that live body weight
of goats can be accurately estimated using selected morphometric traits,
particularly wither height, rump height, and length from point of
shoulder to knee length, offering reliable alternative of weighing scales
in traditional livestock markets.