Response to long-term selection in early maturing maize synthetic varieties

Abstract

Long-term continuous selection is essential for germplasm improvement. However, choice of germplasm for long-term genetic improvement might limit the success of germplasm enhancement programs. The objective of this research was to report the response to long-term selection in early maturing North Dakota (ND) synthetic varieties. We wanted to determine whether the performance of three ND maize synthetic varieties was improved by long-term mass selection (M) and if the performance of one of them was improved by long-term modified ear-to-row (MER) selection. The evaluation of long-term selection response was performed at two plant densities. An experiment in a randomized complete block design with split-plot arrangement was used to evaluate NDSM(M), NDSAB(M), NDSCD(M), and NDSAB(MER) under 75,000 and 42,500 plants per hectare across seven environments. Long-term mass selection for grain yield and stalk lodging resistance in NDSM(M), NDSAB(M), and NDSCD(M) was not successful, since there were no significant changes in grain yield or stalk lodging in these populations at either low or high densities. On the other hand, long-term modified ear-to-row selection was effective for grain yield improvement in NDSAB(MER). Grain yield increased non-linearly from 3.9 Mg ha−1 in cycle 0 to 5.0 Mg ha−1 in cycle 12 at a rate of 2.5% per cycle. Interaction between plant density and genotype was not detected even though selection was performed at relatively low densities (20,000 plants ha−1 for mass selection and 50,000 plants ha−1 for ear-to-row selection). The confirmation of a lack of interaction between plant density and genotype suggests that selection at low plant densities might still be able to provide high-density stress resistance through density-independent genotypes, allowing progeny testing across multiple locations with better accuracy and fewer resources. Selection methods that emphasize both additive and dominance effects such as full-sib recurrent selection are recommended to maximize genetic improvement of advanced population cycles of early maturing synthetics.