Multiphasic analysis of growth curves in chickens.

A multiphasic function that considers body weight to result from an accumulation from more than one phase of growth was used to describe growth curves for Rhode Island Red (RIR) and White Leghorn (WL) males and females, from hatching to 45 wk of age. Mean body weight gains were fitted by iteratively reweighted nonlinear regression using the multiphasic function: (formula; see text) where yt is mean gain (grams) at age t; n is number of phases; tanh is hyperbolic tangent; for each phase i, ai is half asymptotic weight; bi is growth rate relative to ai (weeks-1) and ci is age at maximum gain (weeks). For each phase, maximum gain is aibi and duration (days required to attain about 75% of asymptotic yield during that phase) is 2bi-1. Estimates of parameters clearly point to the diphasic nature of growth and to differences between phases of sexes. First and second phases accounted for 97% of total asymptotic weight. For the first phase, males attained 70% of their asymptotic weight, whereas females attained 85%. Duration of the first phase was 15 wk. For the second phase, it was 12.5 wk for RIR and 10 for WL males, whereas it was 5 wk for RIR and 6 for WL females. Maximum gain during the first phase averaged 144.4 g for males, 108.8 g for females, 156.2 g for RIR, and 97.0 g for WL. During the second phase, it was 68.0 g for RIR males and females but 71.5 and 55.9 g for WL males and females, respectively. Age at maximum gain during the first phase was 12.0 wk. During the second phase, it was 24.2 wk for RIR females, 2.6 wk later than for males, whereas it was 27.3 wk for WL females, 8.3 wk later than for males. For a fixed total asymptotic weight, partitioned between two phases of growth, a higher, longer, and later first phase was associated with a lower, shorter, and later second phase; the association was greater for males than for females.