Do different production environments justify separate maize breeding programs?

Abstract

DiVerent production environments are being adopted by farmers. Therefore, allocation of resources to breeding research that targets diVerent production environments should be continuously assessed. Agronomists should conduct extensive hybrid £ production environment interaction research before recommending breeders to conduct separate breeding programs for each production environment. The lack of interactions between genotypes and production environments (e.g., tillage) would not justify conducting separate breeding programs and duplicating breeding resources. On the other hand, separate breeding programs would be necessary if cultivar rankings diVer. The purpose of this paper is to review the available literature on experiments designed to test genotype £ tillage interactions (GT) in maize (Zea mays L.). No-till system (NT) and conventional till system (CT) were utilized as examples of diVerent production environments. The majority of experiments reviewed showed that there is no need to develop cultivars speciWc to NT because the cultivars that were developed under CT systems performed relatively the same under NT. The magnitude of GT interactions found was very small to expect better cultivars from breeding under NT. Additional research is needed to conWrm these conclusions, especially when applied to other production environments (e.g., development of cultivars under organic conditions). Scientists should evaluate genotype by tillage interactions before investing additional resources in breeding for those speciWc target environments. Top yielding genotypes seem be consistent across years, locations, inputs; and most of the present evidence suggests that breeding for speciWc till systems is not necessary.

Genotype × tillage interaction in a recurrent selection program in wheat

“…Our results show that stored water at harvest was higher under non-tillage (Table 1) even when the stored water at seeding was the same ( Carena et al (2009) showed that there is no need to develop cultivars specific to non-tillage because the cultivars that were developed under conventional tillage systems performed relatively the same under non-tillage. In line with the above opinion, some authors consider that since the germplasm within breeding programs is continually changing and farming practices change with time, there is no need to drive parallel breeding programs (Sip et al 2009).…”

Germplasm enhancement for adaptation to climate changes

“…We have identified cost/effective, and less input oriented breeding strategies to optimize water under extreme conditions. We intend to identify the ones for N use efficiency under similar conditions based on the principles addressed by Carena et al (2009a). Certain genotypes have demonstrated to be more stable across environments which should be more tolerant to climate change (Carena et al 1998).…”

“…It is, therefore, necessary to evaluate genotype x N and genotype x soil type interactions to determine the need/ lack of need for duplicating cultivar development programs under different environmental conditions. We have previously determined that no interactions are present when testing different plant densities and tillage systems in maize , Hyrkas and Carena 2005, Carena et al 2009a. We have the opportunity to challenge current and past production research by addressing the real facts behind sustainability.…”

Hybrid Maize Seed Production Yield Associations with Inbred Line Performance in Multienvironment Trials

“…Further, PREs El Quebrachal in-season sowing (EQ), Fernandez in-season and late sowings (FE), VD, and Puerta Grande in-season and late sowings (PG) were identified within the subtropical GR (Fig. Carena et al (2009) reported evidence of several genotype ´ tillage or soil type interaction studies conducted on maize in the period 1983 to 2006. Therefore, those locations or planting windows that fell within the same PRE (e.g., Carmen late and Rojas late and very late sowings in CR2) might be treated as replications, giving redundant information (Setimela et al, 2005).…”