Genotypic expression at different ages: I. Prolificacy traits of sheep

Genetic parameters for prolificacy traits for Columbia (COLU), Polypay (POL Y), Rambouillet (RAMB), and Targhee (TARG) breeds of sheep were estimated w ith REML using animal models. Traits were number of live births (LAB), litt er size at birth (LSB) and weaning (LSW), and litter weight weaned (LWW). N umbers of observations ranged from 5,140 to 7,095 for prolificacy traits an d from 5,101 to 8,973 for litter weight weaned for the four breeds. For sin gle-trait analyses, ewes were classified as young (1 yr old), middle-aged ( 2 and 3 yr old), or older (> 3 yr old). After single-trait analyses, three- trait analyses were done for each characteristic with traits defined by age class. Generally, heritability estimates from single-trait analyses were l ow and ranged from .01 to .17 for LAB and LSB and from .00 to .10 for LSW. Heritability estimates obtained for LWW ranged from low to moderate (.00 to .25) and were less for older ewes. Heritability estimates from the three-t rait analyses were generally similar to estimates from single-trait analyse s. Heritabilities for LAB and LSB were similar, and, for three-trait analys es, they ranged across age groups from .07 to .13 for COLU,.13 to .16 for P OLY,.10 to .16 for RAMB, and .01 to .16 for TARG. Estimates for LSW from th ree-trait analyses ranged from .07 to .12 for COLU, .04 to .09 for POLY, .0 1 to .11 for RAMP, and .03 to .11 for TARG. For LWW, heritabilities ranged from .00 to .21 for COLU, .05 to .08 for POLY,.12 to .15 for RAMB, and .18 to .29 for TARG. Genetic correlations for LAB, LSB and LSW among age-define d traits ranged from .25 to 1.00. Genetic correlations for LAB and LSB betw een young and middle and between young and older age classes were less than .80 in COLU, POLY, and RAMB breeds. Only genetic correlations between midd le and older age classes for these breeds were greater than .80. For TARG, genetic correlations among all age classes were greater than .80 (.88 to 1. 00) for those traits. All genetic correlations among ages for LSW were grea ter than .80 for POLY and TARG. For RAMB, only the correlation between youn g and older age classes for LSW was less than .80 (.45). None was greater t han .80 for COLU. For LWW, genetic correlations among all age classes in PO LY and RAMB were greater than .80 (.82 to 1.00). For COLU, genetic correlat ion between young and middle was low (.07), between young and older was hig h (.88), and between middle and older classes was moderately high (.54). Fo r TARG, genetic correlations were .49, .65, and .98 for young-middle, young -older, and middle-older age classes, respectively. Results indicate that m ore progress could be made in selection programs for prolificacy traits in some sheep breeds by considering age of ewe as a part of the trait rather t han by simply adjusting for ages of ewes.