Yield and Agronomic Traits of Waxy Proso in the Central Great Plains

Proso (Panicum miliaceum L.) is a summer annual grass capable of producing seed in 60 to 90 d. This characteristic, and its effi cient use of water, makes it well suited to the short, and often hot and dry, growing season in the high plains of the central Great Plains. The introduction of novel end-use characteristics such as waxy starch can stimulate an increased market for proso. We evaluated 18 experimental F5 waxy lines derived from a cross of ‘Huntsman’ and PI436626 across seven locations. Genotype × environment variation in waxy proso was mostly a matter of changes in magnitude and not crossover interaction. When crossover interaction was implicated, it was generally slight and occurred at lower environmental means—at locations with low mean response to any given variable. Waxy progeny mean yield was lower than Huntsman but signifi cantly higher than PI436626. Except for test weight, waxy progeny mean response for most traits was similar to check cultivars. Mean yield of one experimental line did not differ signifi cantly from Huntsman, and 14 did not differ signifi cantly from ‘Horizon’, the second highest yielding cultivar. In addition, regression analysis suggests that top-yielding waxy lines responded well to high-yield environments. Seed sizes for all waxy lines were smaller than the check lines, but most were signifi cantly larger than PI436626. Waxy lines generally headed at a similar time to Huntsman and the other nonwaxy checks, and most were signifi cantly earlier than PI436626. Late maturity of PI436626 was the main factor limiting its culture in the High Plains region. R.F. Heyduck, Univ. of Nebraska-Lincoln, Dep. of Agronomy and Horticulture, 321 Keim Hall, Lincoln, NE 68583; R.F. Heyduck, current address, New Mexico State Univ., Farmington Agricultural Science Center, P.O. Box 1018, Farmington, NM 87499; D.D. Baltensperger, Univ. of Nebraska-Lincoln, Panhandle Research and Extension Center, 4502 Avenue I, Scottsbluff , NE 69361; L.A. Nelson, Univ. of Nebraska-Lincoln, Dep. of Agronomy and Horticulture, 364 Keim Hall, Lincoln, NE 68583; R.A. Graybosch, USDA-ARS, 368 Keim Hall, Lincoln, NE 68583. A contribution of the University of Nebraska Agricultural Research Division, supported in part by funds provided through Anna Elliott Grant. Received 13 Feb. 2007. *Corresponding author ([email protected]). Abbreviations: DOY, day of year; G×E, genotype × environment. Published in Crop Sci. 48:741–748 (2008). doi: 10.2135/cropsci2007.02.0081 © Crop Science Society of America 677 S. Segoe Rd., Madison, WI 53711 USA All 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. R e p ro d u c e d fr o m C ro p S c ie n c e . P u b lis h e d b y C ro p S c ie n c e S o c ie ty o f A m e ri c a . A ll c o p y ri g h ts re s e rv e d . 742 WWW.CROPS.ORG CROP SCIENCE, VOL. 48, MARCH–APRIL 2008 Of crucial importance, however, is a market for this crop of rising signifi cance. Most proso grain sold in cash trade goes to local elevators, where it is cleaned and processed before entering into the birdseed market (Baltensperger, 1996). Some proso is dehulled and is marketed for human, poultry, and animal consumption. While proso prices have historically been higher than maize or grain sorghum (Sorghum bicolor L. Moench), price levels can fl uctuate dramatically. In any given year, if the birdseed and human consumption markets are saturated, the price for proso drops to feed grain levels. Price ranges from $66 to $485 per tonne over a fi ve-year period are common (Bergener, 2002). A year or two following a price crash, the price will again rise to “specialty crop” levels, and proso again becomes very popular. Due to this vast fl uctuation, proso has an unstable market footing in this regional economy and in rotations and cropping systems in the region. The introduction of novel end-use characteristics, such as “waxy” starch, can stimulate a more stable market for proso. Most common cereal starches contain 20 to 30% amylose and 70 to 80% amylopectin (Jane et al., 1999). Waxy starches contain more than 95% amylopectin (Demeke et al., 1997). Waxy starch will gelatinize at lower temperatures (Lineback, 1999) and results in cooked grains with a sticky surface, allowing it to be easily eaten with chopsticks. This is a preferred characteristic for Asian culinary markets. Graybosch and Baltensperger (2008) evaluated the USDA-ARS proso collection for accessions with waxy endosperm starch and identifi ed fi ve accessions, four of which were from mainland China. The trait was found to be under the control of duplicate recessive alleles designated wx-1b and wx-2b. Earlier evaluations of these waxy accessions had shown that they are too late to mature reliably in the High Plains; because of this, yields are very low (Heyduck et al., 2002). Waxy accessions and numerous check cultivars were crossed in a greenhouse in winter 1999 to integrate the waxy trait with agronomic traits suited for the central Great Plains. Four rounds of selection based on waxy starch characteristic, maturity, panicle architecture, seed color, and plant height produced 47 F4 experimental waxy lines, all descending from a cross of PI436626 and ‘Huntsman’. In this study, 30 lines were evaluated: 18 experimental waxy lines, both parents, an additional waxy line of Chinese origin, and a battery of nine check cultivars. Six agronomic traits were analyzed: days to heading, plant height, lodging, grain yield, test weight, and seed size. The goals were threefold: (i) to assess the genotype × environment (G×E) interactions of the waxy lines, parental, and check proso lines; (ii) to assess the agronomic performance and stability of experimental waxy lines relative to nine cultivars that are commonly grown in the region; and (iii) To identify one or more waxy lines suitable for release as cultivars. MATERIALS AND METHODS Plant Materials Thirty entries were evaluated during the 2002 and 2003 fi eld seasons. Of these, 18 were F5 experimental “waxy” lines developed from a cross of PI436626 and Huntsman. These lines were the result of four rounds of selection for waxy starch character, plant height, panicle shape, and early maturity. The remaining lines in the trial were the parents, as described above, another waxy accession, PI436625, and nine nonwaxy, adapted lines as a control group for the evaluation of agronomic traits: ‘Horizon’, ‘Sunup’, ‘Sunrise’, ‘Earlybird’, ‘Dawn’, 9668-17, 9213, 9308, and 9217-L. A list of entries in this trial is shown in Table 1. Table 1. Description of 30 lines entered in proso millet variety trial, 2002 and 2003. Includes 18 experimental lines, both adapted and waxy parental lines, nine regional check lines, and an additional waxy line of Chinese origin. Entry Starch (waxy/non) Year of release Note Horizon non 2004 Regional check Huntsman non 1995 Adapted parent of waxy test lines, regional check Earlybird non 1995 Regional check Sunrise non 1994 Regional check Sunup non 1989 Regional check Dawn non 1976 Regional check 9213 non Regional check, line under development 9668-17 non Regional check, line under development 9308 non Regional check, line under development 9217-L non Regional check, line under development 172-2-9 waxy Waxy test line 172-2-B waxy Waxy test line 174-7-13 waxy Waxy test line 175-5 waxy Waxy test line 177-3-13 waxy Waxy test line 177-7-5 waxy Waxy test line 177-8 waxy Waxy test line 177-9-2 waxy Waxy test line 177-9-12 waxy Waxy test line 177-9-13 waxy Waxy test line 182-4-24 waxy Waxy test line 182-5-18 waxy Waxy test line 182-7-20 waxy Waxy test line 10097 waxy Waxy test line 10107 waxy Waxy test line 10110 waxy Waxy test line 10127 waxy Waxy test line 10135 waxy Waxy test line PI436625 waxy Waxy accession of Chinese origin PI436626 waxy Waxy parent of waxy test lines in the trial R e p ro d u c e d fr o m C ro p S c ie n c e . P u b lis h e d b y C ro p S c ie n c e S o c ie ty o f A m e ri c a . A ll c o p y ri g h ts re s e rv e d . CROP SCIENCE, VOL. 48, MARCH–APRIL 2008 WWW.CROPS.ORG 743 conducted using SAS (SAS Institute, 2003). PROC MIXED was used for analysis of variance, and Fisher’s LSD was used to compare individual line means. Stability of agronomic traits was determined by regressing genotype mean response on an index of environmental mean response (Eberhart and Russell, 1966). Location means served as the X coordinates for locations along the environmental index. The line within location means were then plotted on the Y-axis above their location mean. Regression lines were then fi tted to these points and the coeffi cients tested for signifi cant diff erence from b = 1.0 by use of an F test. Correlation of trait means with one another, and of means with their b-values, was done using the CoStat package (CoHort Software, 2001). Yield is the predominant factor in selecting a line or lines for release. In this study, we examined the overall mean for a given line and its regression slope. Theoretically, a slope of 1 demonstrates stability; that is, as the environment improves, the response of a given line is relative to the body of entries tested in those environments. For yield, this is straightforward, and we could expect a cultivar with a regression slope of 1.3 to respond very well to a high yield environment. Conversely, a cultivar with a regression slope of 0.6 may yield relatively similarly across environments (and years) but may not utilize benefi cial site and/or climate factors to increase yield. Plot Management and Data Collection Trials were planted 2002 and 2003 at seven locations in Nebraska, Colorado, and Wyoming. All Nebraska trials were planted in plots 1.2 m wide and 5.2 m long (6.3 m2) with four rows at 30-cm spacing. At Wyoming and Colorado locations, plots were planted in plots 1.5 m wide and 4.5 m long (6.8 m2) in six rows at 25-cm spacing. Regardless of planting technique or seed size, the target seeding rate was 16.8 kg ha–1. Heading notes were taken at weekly intervals at four environments in 2002. Heading date was defi ned as the date at which 50% of heads within a plot were fully emerged from the boot and panicle stem was visible above the fl ag leaf. This date was then converted to day of year (DOY), from which the DOY of planting was subtracted to give days from planting to heading. Plant height was recorded at seven locations, fi ve from the 2002 season and two from the 2003 season. Height was measured from the ground to the highest point of a plant determined by visual inspection to be of representative height in that plot. Bent panicles were not straightened out for this measurement. Lodging was recorded only at the four locations when and where it occurred, in late August of the 2002 season. Lodging was determined by visual inspection and rated from 0 to 10 as severity increased. All plots were directly harvested using a small plot combine. In Nebraska, harvested plot area was 4.5 m2 (70% of plot); at the Wyoming and Colorado plots, harvested area was 3.6 m2 (48% of plot). Onboard electronic data collection included grain yield, grain moisture (%), and test weight. In addition, a small subsample was taken from each sample and stored for further evaluation, including seed size measurements and starch characterization. Planting and harvest dates, as well as soil information for the sites, are found in Table 2. Experimental Design and Statistical Analysis The experiment comprised 30 entries and was conducted in a randomized complete block design with four replicates at each of the seven environments. The trial consisted of three locations in 2002 and four locations in 2003. Because the range of climatic patterns between locations in a given year can vary as much as between years at the same location, the year and location eff ects were analyzed together as the environmental eff ect. Environment and replication were analyzed as random eff ects, while genotype was analyzed as a fi xed eff ect. Analyses of days to heading, plant height, lodging, yield, grain volume weight, and seed size were Table 2. Summary of seven location-year environments for 2-yr proso trial. Year Location Irrigation Soil Planting date Harvest date 2002 Scottsbluff, NE Irrigated Keith silt loam 23 May 30 Aug. Scottsbluff, NE Semi-irrigated Keith silt loam 23 May 29 Aug. Torrington, WY Irrigated Vetal fi ne sandy loam 10 June 11 Sept. 2003 Akron, CO Dryland Weld silt loam 9 June 22 Sept. Sidney, NE Dryland Alliance loam 30 May 29 Aug. Scottsbluff, NE Irrigated Keith silt loam 27 May 20 Aug. Torrington, WY Irrigated Vetal fi ne sandy loam 4 June 11 Sept. Figure 1. Overall mean yields (kg ha–1) for selected proso lines (10 of 30) from the 2002–2003 trial. Selected lines represent top eight lines and waxy Chinese plant introduction. Lighter gray bars denote waxy lines, capital P denotes parents of experimental line. R e p ro d u c e d fr o m C ro p S c ie n c e . P u b lis h e d b y C ro p S c ie n c e S o c ie ty o f A m e ri c a . A ll c o p y ri g h ts re s e rv e d . 744 WWW.CROPS.ORG CROP SCIENCE, VOL. 48, MARCH–APRIL 2008 For the regression plots (Fig. 1; see results), the waxy lines with both the highest and lowest mean values for each variable measured are plotted along with the parents, the mean of the 10 checks, and the mean of all waxy lines simply to explore the range of response in the waxy lines. Any choice of a cultivar for release would be guided primarily by mean values across environments and only secondarily based on any evaluation of stability across environments. RESULTS AND DISCUSSION