Published online by Cambridge University Press: 02 September 2010
Data records from 556 beef heifers which included age at first oestrus (AFE) were used to test the null hypotheses: 1) daylength at birth is not a determinant of AFE in beef females; and 2) there is not an association between the day on which first oestrus occurred and the lunar phase.
The effect of daylength at birth was measured by forming a variable, X1 where X1 = sine (X1′ 360/365·25)°, and including it among the set of variables in a multiple linear regression equation with AFE as the dependent variable. The resultant estimated coefficient on X1 was -3·617 with a standard error of 21·32, and led to the conclusion that daylength at birth was not a factor which affected AFE in this group of beef heifers.
The lunar phase hypothesis was tested by forming the frequency distribution of the 556 observations of first oestrus over the 30-day lunar cycie. The chi-square test was applied with the theoretical (hypothesized) distribution being the uniform distribution. The value of the test statistic was such that the null hypothesis would not be rejected even with an a value approaching 0·50. Additionally, coefficients of autocorrelation calculated for lag phases of 7, 14 and 21 days were quite small. Thus it was concluded that the occurrence of first oestrus was uniformly distributed over the lunar cycle.