Changes in genetic variance during advanced cycle breeding in maize

Advanced cycle breeding in crops aims to develop newer, improved versions of inbreds from crosses among current, elite inbreds. Although genetic gains in many crops have been achieved through advanced cycle breeding, this breeding process systematically leads to a narrow germplasm base. Our objectives were to estimate the loss, if any, in genetic variance during advanced cycle breeding in maize (Zea mays L.), and to compare these genetic variance estimates with those expected under a testcross additive model. We evaluated the Iowa Stiff Stalk Synthetic (BSSS), B73 X B37, and Lo904 X Lo916 populations, which have a common genetic background but have different levels of genetic variance expected from the testcross additive model; that is, 100% in BSSS, 82% in B73 X B37, and 45% in Lo904 X Lo916. Contrary to expectations, testcross additive variances (V-TA) for grain yield did not significantly differ among the three populations. As expected, VTA for grain moisture, stalk lodging, plant height, and ear height were significantly lower in B73 X B37 than in BSSS. The decreases in VTA for these traits were not significant from B73 X B37 to Lo904 X Lo916. Fitting a testcross epistatic model did not result in positive estimates of both VTA and testcross additive-by-additive variance (V-TAA) for any trait. In summary, genetic variance, except for grain yield, was lost during advanced cycle breeding and the amount of loss was greater than expected.