Synergy of Empirical Breeding, Marker-Assisted Selection, and Genomics to Increase Crop Yield Potential

indicated that genetic improvement usually accounted for about one-half of the total yield increase, with the This paper was presented as part of the symposium entitled “Postremainder attributed to changes in cultural practices Green Revolution Trends in Crop Yield Potential: Increasing, Stagnant or Greater Resistance to Stress.” In this presentation, we have such as increased rates of mineral fertilizers and the use focused on (i) uses of marker technology in determining the genetic of herbicides for weed control and pesticides for control basis of phenotypic expression and the manipulation of phenotypic of insects and diseases. Duvick (1997) suggested that the variation in plants. This included the use of markers in understanding increased grain yielding ability of these widely successful heterosis, in attempts to improve hybrid predictions, in quantitative hybrids was due primarily to improved tolerance of abitrait locus (QTL) identification and mapping, in marker-assisted selecotic and biotic stresses, coupled with the maintenance tion (MAS), and in enhancing breeding success in the development of the ability to maximize yield per plant under nonof improved lines and hybrids; (ii) the role of genomics in developing stress growing conditions. a precise understanding of the genetic basis of phenotypic expression Opportunities for gains resulting from changes in culwhich will then provide more precision in the manipulation of phenotural practices are limited (particularly in the USA and typic variation; and (iii) some attempts to integrate marker technology and genomics into empirical breeding strategies. In addition, we have other developed countries). Therefore, future gains in focused on what has been successful as well as what has fallen short the productivity of most crops may depend almost enof expectations, and have suggested some of the possible reasons for tirely on genetic improvements. In fact, environmental the lack of success. Because of page limitations, we could not include concerns may cause a reduction in the use of agricultural an exhaustive review of the plant literature and have limited many chemicals and fertilizers. Also, many parts of the world of our examples to investigations in maize (Zea mays L). may have limited supplies of such chemicals and plant nutrients. Therefore, plant breeders will need to develop and apply new technology (such as marker-assisted seT global ability to provide adequate amounts of lection) at a faster pace to more effectively improve the food, feed, and fiber from domesticated crop plants yield potentials of crop plants for the ever increasing has resulted largely from the collective empirical breedglobal human population as well as for the changes in ing efforts of farmers and plant breeders spanning many consumer preferences. millennia. The continued increases in plant productivity have resulted from artificial selection, either conscious Quantitative Traits and QTLs or unconscious, on the phenotypic expressions of the A majority of economically important plant traits, targeted species. Prior to the 20th century, plant breedsuch as grain or forage yield, can be classified as ing was largely an art with little or no knowledge of multigenic or quantitative. Even traits considered to be genetic principles. Although plant improvement since more simply inherited, such as disease resistance, may the rediscovery of Mendel’s principles has involved both be “semi-quantitative” for which trait expression is govart and science, the contributions of science will unerned by several genes (e.g., a major gene plus several doubtedly assume a much greater role as new technolmodifiers). The challenge to use strategically new techogy becomes more widely used and as additional gains nology (such as DNA-based markers) to increase the in agricultural productivity are required to support a contribution of “science” to the “art plus science” equagreater global population. New opportunities to use getion for plant improvement therefore applies to most, if notypic selection, or combinations of genotypic and phenot all, traits of importance in plant breeding programs. notypic selection, to increase yield potentials are emergAlthough the focus of this symposium is on yield potening at an ever-increasing pace. tial, that yield must be harvestable. Therefore, traits On the basis of a comparison of 36 widely grown such as standability (or lodging resistance), disease resishybrids adapted to central Iowa and released at intervals tance, and insect resistance must also be considered. from 1934 to 1991, Duvick (1997) reported that the Historically, early researchers in quantitative genetics increase in maize grain yield during that time span averquestioned whether the inheritance of these continuaged nearly 74 kg ha21 yr21. Hybrid comparisons were ously distributed traits was Mendelian (Comstock, based on side-by-side trials, so all of the gain could 1978). The answer to this question has major implicabe attributed to genetic improvement. Earlier studies tions in the consideration of the use of markers for plant breeding programs. During the past century, both plant C. Stuber, USDA-ARS and Dep. of Genetics, N.C. State Univ., Raleigh, NC 27695-7614; M. Polacco, USDA-ARS, Plant Genetics Res. and animal geneticists have obtained convincing eviUnit, Univ. of Missouri, Columbia, MO 65211; M. Senior, Novartis dence that Mendelian principles apply to quantitative Agribusiness Biotechnology Research, Inc., 3054 Cornwallis Rd., Research Triangle Park, NC 27709. Received 28 Dec. 1998. *CorrespondAbbreviations: BAC, bacterial artificial chromosome; MAS, markering author ([email protected]). assisted selection; QTL, quantitative trait locus, RFLP, restriction fragment length polymorphism; YAC yeast artificial chromosome. Published in Crop Sci. 39:1571–1583 (1999).