CROP SCIENCE, VOL. 49, NOVEMBER–DECEMBER 2009 1969SCIENTIFIC PERSPECTIVESCorn (Zea mays L.) production in the United States was 53 million Mg annually and corn yield averaged 1518 kg ha–1 in the 1930s, when corn hybrids were fi rst commercially grown. Corn production grew to 76 million Mg annually in the 1950s, to 150 million Mg annually in the 1970s, and to 219 million Mg in the 1990s (USDA National Agricultural Statistics Service, 2007). In 2001 corn became the highest tonnage cereal crop worldwide: 557.6 million Mg of corn, 542.4 million Mg of paddy rice (Oryza sativa L.), and 535.6 million Mg of wheat (Triticum aestivum L.) (FAO, 2002). In 2004 the U.S. record corn yield was 10,059 kg ha–1, and in 2007 the U.S. record corn production was 332.7 million Mg (USDA National Agricultural Statistics Service, 2007).World corn production provides feed, food, and fuel for more than 6,000 million humans. World population continues to increase in the face of higher food costs, less arable land, water scarcity, and the threat of global warming with rising temperatures and carbon diox-ide levels. Nobel Laureate Dr. Norman Borlaug (2007) predicted the demand for cereals worldwide will probably grow by 50% during the next 20 yr. Monsanto Co. (St. Louis, MO) has recently announced a “sustainable yield initiative,” which includes the goal of doubling corn yields by the year 2030 (Lohuis et al. (2008), www.monsanto.com [verifi ed 31 July 2009]). Studies by international organizations such as the United Nations, the World Bank, and other global experts emphasize unprecedented growth in global demand for cereals.Heterosis Decreasing in Hybrids: Yield Test InbredsA. Forrest Troyer* and Eric J. WellinABSTRACTYield testing fi nished inbreds to replace pre-liminary single-cross corn (Zea mays L.) yield tests will increase rate of commercial hybrid yield gains. Studies have shown that heterosis decreased 25%/50 yr, 10%/60 yr, and 35%/100 yr. Natural selection and artifi cial selection by plant breeders for adaptedness have increased parental inbred and hybrid seed yields, whereas percentage heterosis decreased. Four studies have shown inbred yields increased 1.9 to 3.5 times faster than heterosis yields. Pioneer Hi-Bred generates 700 new inbreds tested in 6000 single-cross hybrids at 15 to 20 locations annu-ally. Predicted, untested, newer hybrids are then made and tested extensively with commercial hybrids. Parental inbred yield testing is the next to last of several steps in hybrid development. Commercial hybrid development costs have increased logarithmically, whereas performance has increased linearly. Replacing preliminary testcross trials with fi nished-inbred yield trials is more effi cient. About 12,000 new fi nished inbreds can be evaluated annually with no tes-ters and at least 50% fewer locations per inbred with the same testing effort as 700 new inbreds with testers. A calendar year per breeding cycle and annual production costs for 6000 hybrids will be saved. Corn yield trials detect stress sus-ceptibility, which is more apparent in inbreds than in hybrids. Evaluation of more new inbreds will be conducive to increased genetic diversity that produces higher-yielding hybrids.A.F. Troyer, Dep. of Crop Sciences, Univ. of Illinois, Urbana, IL 61801; E.J. Wellin, 111 Presidio Ct., #304, Schaumburg, IL. Received 6 Apr. 2009. *Corresponding author ([email protected]).Abbreviations: PHBI, Pioneer Hi-Bred International; PVP, Plant Variety Protection.Published in Crop Sci. 49:1969–1976 (2009).doi: 10.2135/cropsci2009.04.0170© Crop Science Society of America677 S. Segoe Rd., Madison, WI 53711 USAAll rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.1970 WWW.CROPS.ORG CROP SCIENCE, VOL. 49, NOVEMBER–DECEMBER 2009We can expect increasing yields of U.S. hybrid corn, but corn breeders need to accelerate the rate of develop-ment of higher-yielding, fi nished corn inbreds to increase the occurrence of higher-yielding corn hybrids. Yield testing fi nished inbreds to replace preliminary balanced single-cross yield tests to evaluate inbreds will increase the yield gains of commercial hybrids.More fi nished inbreds are conducive to more genetic diversity that increases percentage heterosis.MATERIALS AND METHODSWe examined four, independent, comprehensive, heterosis studies comparing parental fi nished inbreds to their hybrids across 29-, 53-, 60-, and 100-yr time periods. Schnell (1974) summarized 17 corn heterosis experiments grown in the U.S. Corn Belt from 1916 to 1969. In each experiment, a number of fi nished inbred lines were evaluated together with a complete or a balanced set of single-cross hybrids made from those lines. We simplifi ed Schnell’s intricate fi gure to the four basic regressions across time for each of the three studies; we regressed the means for inbred or cultivar yields, heterosis yields, and hybrid yields along with percentage heterosis across years of time periods.Duvick (1984, 1999) and Duvick et al. (2004) summa-rized data on Pioneer Hi-Bred International (PHBI; Johnston, IA) commercial hybrids introduced in central Iowa across fi ve decades. The tests included 47 commercial hybrids and their inbred parents together in bordered plots at three plant densi-ties in a total of nine locations in 3 yr (~5000 plots). He also summarized data on six sets of 10 single-cross hybrids, each set made from the fi ve most widely used unrelated inbreds in PHBI for central Iowa in each of the six decades. These 60 hybrids and their parent inbreds were tested together in bordered plots in three plant densities at two locations for 2 yr (~1500 plots). We averaged Duvick’s two estimates by decades for the fi rst fi ve decades and used only the single-cross data for the sixth decade.Campbell et al. (2008) summarized data on fi ve modern and fi ve obsolete cotton (Gossypium hirsutum L.) cultivars test crossed on a modern cotton cultivar. In 2005 10 hybrids and 11 cultivars, including the tester, were grown together in four replication, randomized block fi eld trials in Alabama, North Carolina, and South Carolina using two-row entry plots 10.7 m long.