Seventy-five Years of Breeding Dry Bean of the Western USA

A periodic comparison of cultivars is essential to assess selection gains, determine deficiencies, define objectives, and set breeding priorities. Our objective was to assess the progress, or lack thereof, achieved in improving yield, plant type, maturity, and resistance to major bacterial, fungal, and viral diseases of dry bean of the western USA from 1918 to 1998. Twentyfive great northern, pink, pinto, and red cultivars were evaluated for seed yield at three locations in Idaho and for anthracnose, Bean common mosaic virus, Bean common mosaic necrosis virus, common and halo bacterial blights, Fusarium and Rhizoctonia root rots, Fusarium wilt, and white mold in Colorado, Idaho, and Washington between 1999 and 2006. Yield ranged between 2904 kg ha-1 for pinto 'UI 111' to 3921 kg ha1 for 'Bill Z', which represents a 35% gain in 54 yr. Yield gain in great northern was 587 kg ha-1 , pink 136 kg ha1 , and red 687 kg ha1 . Stability indices ranged from 0.57 for 'Kodiak' to 1.86 for 'UI 3'. Maturity ranged from 90 d for 'UI 320' to 97 d for 'Frontier'. Seed weight ranged from 28 g for 'Viva' to 41 g for UI 320. An acceptable degree of resistance to Rhizoctonia root rot was achieved in most cultivars. All cultivars were susceptible to anthracnose, common bacterial blight, and white mold, and all except 'Chase' to halo blight. Only 'Matterhorn', 'Weihing', and Kodiak combined an upright Type II growth habit with resistance to BCMV and rust. An integrated breeding strategy should be explored for simultaneous improvement of multiple traits in future cultivars. S. Singh, H. Teran, and M. Lema, Univ. of Idaho, Kimberly, ID 83341; D. Webster, Seminis Vegetable Seeds, Filer, ID 83303; C. Strausbaugh, USDA-ARS, Kimberly, ID 83341; P. Miklas, USDA-ARS, Prosser, WA 99350; H. Schwartz and M. Brick, Colorado State Univ., Fort Collins, CO 80523. *Corresponding author ([email protected] ). Abbreviations: BCMV, Bean common mosaic virus; BCMNV, Bean common mosaic necrosis virus; BCTV, Beet curly top virus. N ATIVE AMERICANS grew rain-fed or dryland dry bean (Phaseolus vulgaris L.) before Europeans arrived on the American continents, and dry bean was found in Anasazi and other Native American dwellings in the western USA (Kaplan, 1956; Gentry, 1969; Kaplan and Lynch, 1999). Dryland subsistence dry bean production has been practiced on thousands of hectares in the western USA since time immemorial (Mimms and Zaumeyer, 1947). Popular bean market classes of the western USA include medium-seeded (25-40 g 100 seed weight-') pinto, great northern, red Mexican, pink, and Anasazi. Most of these belong to the race Durango that was domesticated in the semiarid and arid central and northern highlands of Mexico (Singh et al., 1991). These dry bean germplasms are characterized by an indeterminate prostrate Type III growth habit (Singh, 1982). In the western USA, cool nights and long warm dry summer days characterize irrigated production systems. Race Durango Type III cultivars often out yield small-seeded (<25 g 100 seed weight-') race Mesoamerica black and navy cultivars with predominately upright bush Type II growth habit in the USA (Singh and Powers, 2000). Further, both groups of cultivars out yield large-seeded (>40 g 100 seed weight-') Andean light and dark red kidney, white kidney, cranberry, and other market classes with bush Type I growth habit. Published in Crop Sci. 47:981-989 (2007). doi: 10.2135/cropsci2006.05.0322 © 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. CROP SCIENCE, VOL. 47, MAY —JUNE 2007 981 Presently, pinto is by far the most predominant market class in the USA and North America, whereas other market classes are grown more commonly in specific production regions. Breeding and genetics of the western dry bean was initiated by the private sector in southern Idaho in 1918/1919 and at the University of Idaho in 1925 (Dean, 1994, 2000). Subsequently, the USDA started genetic improvement of the western dry bean at Greeley, Colorado, in 1930. In addition to the USDA and Idaho, California, Colorado, Michigan, Nebraska, New York, North Dakota, Oregon, and Washington have ongoing public dry bean breeding programs. Also, there are numerous private breeders dedicated to genetic improvement of several market classes of dry and green (garden, stringless, or snap) bean. Initially, selection for resistance to Bean common mosaic virus (BCMV) (a seed-borne and aphid-vectored potyvirus) within landraces was practiced at the University of Idaho resulting in great northern cultivars such as UI 1 (released in 1930) and UI 59 (released in 1932). Interested readers should refer to Dean (1994, 2000) for the history of breeding, year of release, and details regarding the UI cultivars. Resistance to Beet curly top virus (BCTV) (a leafhopper-vectored geminivirus) from Common Red Mexican landrace, BCMV, and Fusarium root rot [caused by Fusarium solani (Mart.) Sacc. f. sp. phaseoli (Burkholder) W. C. Snyder & H. N. Hans.] was combined subsequently using hybridization among landraces. Expansion of the western dry bean cultivars for production in the midwestern states has occurred since the 1970s. This expansion of production warranted further use of exotic germplasm to change plant growth habit to upright Type II, and resistance to rust [caused by Uromyces appendiculatus (Pers.) Ung.], white mold [caused by Sclerotinia sclerotiorum (Lib.) de Bary], common bacterial blight [(caused by Xanthomonas campestris pv. phaseoli) (Smith) Dye and X. campestris pv. phaseoli var. fuscans] and other diseases. Brick and Grafton (1999) and Miklas (2000) reviewed germplasm use and breeding of race Durango cultivars in the USA, and Singh et al. (1993) emphasized use of intergene pool and interracial hybridization for their improvement in the tropics and subtropics. In the USA alone, several dozen cultivars of race Durango have been developed during the last 80 yr. Concerted efforts were made in 1998 to acquire as many private and public dry bean cultivars as possible. Based on the year of release, traits improved, availability of seed, and their popularity, 25 cultivars representing the four medium-seeded market classes of race Durango cultivated in the USA were selected for a comparative evaluation. Our objective was to assess the progress, or lack thereof; achieved in improving yield, plant type, maturity, and resistance to major bacterial, fungal, and viral diseases from 1918 to 1998. MATERIALS AND METHODS Eight great northern, three pink, nine pinto, and five red cultivars of race Durango were evaluated at the University of Idaho Research and Extension Centers at Kimberly and Parma, and in on-farm trials at Hazelton, ID, from 1999 to 2001. The three sites represent the three major bean production regions in southern Idaho. Kimberly has a mean elevation of 1195 m with a Portneuf silt loam, mixed, mesic, Durixerollic Calciorthids soil, and pH of 7.6. Hazelton, less than 30 km east of Kimberly, is at 1341 m elevation, and soils are similar to those of Kimberly, but no actual data were available. Parma is at an elevation of 703 m with a Greenleaf silt loam (fine silty, mixed, superactive, mesic Zeric Calciargids) soil, and pH of 7.6. A randomized complete block design with four replications was used at each site. Each experimental unit consisted of four rows, each 6.7 m long. The spacing between rows was 0.56 m and an average of 23 seeds linear rn-1 was planted. The Hazelton trials were conducted on-farm and managed by the grower. Nonetheless, test plots at all three sites received fertilizers, preplant herbicides, and cultivation as recommended for commercial production in southern Idaho. Irrigation was applied, as needed to maintain optimum growth at all sites using flat beds and overhead sprinklers at Hazelton and raised beds and gravity flow at Kimberly and Parma. No pesticide for management of any disease or insect was needed. Data were recorded for growth habit during flowering and verified at maturity according to Singh (1982). Days to maturity was recorded when 90% of the pods changed color from green to yellow. The two central rows were harvested at maturity, threshed 10 to 12 d later, cleaned, dried, and seed yield recorded (kg ha') at 12% moisture by weight. Also, weight (g) of 100 seeds taken randomly was recorded. Reaction of 25 cultivars in separate nurseries in the field (on an average of 50 plants per cultivar) to BCTV was determined according to Larsen and Miklas (2004) and to white mold according to Miklas et al. (2001) in Washington. Also, seed yield was measured in a "purgatory field" at the Washington State University—Roza Agricultural Research Unit, where common bean has been planted every other year for the past 40 yr and F. solani predominates. However, F. oxysporum, Pythium ultimum, and Rhizoctonia solani are also present; the soil was purposely compacted to exacerbate the effect of root rots on plant health; nitrogen and phosphorus were limiting; and there was a 35% water deficit. Separate greenhouse evaluations on an average of six plants per cultivar were made in Idaho to determine the reaction of the 25 cultivars to BCMV strains NY15 and US 6 and Bean common mosaic necrosis virus (BCMNV, a potyvirus) strain NL-3K according to Strausbaugh et al. (2003). Also, in the greenhouse in Idaho, common bacterial blight (pathogen isolate obtained from Colorado) was evaluated according to Lema et al. (2006), halo blight (pathogen race 6) according to Taylor et al. (1978), anthracnose [caused by Colletotrichum lindemuthianum (Sacc. & Magn.) Bri. & Cay. race 73] according to Pastor-Corrales et al. (1995), and rust (U. appendiculatus race 53) according to Stavely (1983). Fusarium wilt and Rhizoctonia root rot were evaluated according to Abawi and Pastor-Corrales (1990), and white mold according to Petzoldt and Dickson (1996) in the greenhouse with an average of 24 plants per cultivar in Colorado. For diseases that were scored on a 1 to 9 scale, scores 1 to 3 were considered resistant, 4 to 6 intermediate 982 WWW.CROPS.ORG CROP SCIENCE, VOL. 47, MAY-JUNE 2007 or moderately resistant, and 7 to 9 susceptible. Data for seed yield, seed weight, and days to maturity from each location and year were analyzed separately, homogeneity of variances was tested (Bartlett, 1947), and combined analysis was performed using a mixed model, whereby years and replications were random effects and locations and cultivars were considered fixed (McIntosh, 1983). Subsequently, simple correlation coefficients among the three locations were calculated using the mean seed yield of 25 cultivars over the 3 yr. Stability analysis for 25 cultivars was performed for seed yield according to Eberhart and Russell (1966). Statistical analyses of data were conducted using the SAS (SAS Institute, 2004). RESULTS AND DISCUSSION Mean squares due to year, location, cultivar, and their interactions were highly significant (P 0.01) for seed yield, seed weight, and days to maturity in the combined analysis (Table 1). Because herbicide, fertilizer, and irrigation application at a particular site were relatively consistent over the 3 yr, differences in soil type, climatic factors, seed-bed preparation, and cultivation between the three sites and their interactions with cultivars most likely contributed to significant mean squares. These significant interactions between cultivars, locations, and years further indicated that the response of cultivars changed within and across locations from 1 yr to another. This three-way interaction included both crossover and noncrossover interactions among cultivars with or without changing the rank orders. For example, great northern `UI 465' and `UI 61' had similar yields at Kimberly but yielded differently at Parma when averaged over the 3 yr (Table 2). In contrast, `UI 425' yielded higher than UI 61 at Kimberly, but its yield was comparatively lower at Hazelton. Thus, for maximizing yield within a production region selection of site-specific high-yielding cultivars would be justified. In contrast, identification of and establishing reliable yield estimates for broadly adapted cultivars will require the use of contrasting testing sites representative of the target environments, such as those used in this study for southern Idaho, over a period of three or more years. Mean seed yield among cultivars over the 3 yr at Kimberly was positively correlated with that at Parma (r2 = 0.60 at P 0.01), approximately 300 km west of Kimberly. Yield at neither site was correlated with yield at Hazelton, which is less than 30 km east of Kimberly. These data suggest that dry bean cultivars respond differently to soil, environmental, and other factors. Grower practices at the trials at Hazelton could have contributed to differences from the Kimberly and Parma Research and Extension Centers. Nonetheless, as noted earlier Hazelton is at higher elevation compared to Kimberly, and Parma is at an even lower elevation and is warmer. Because yield of 25 cultivars at Parma tended to be higher in comparison to both Hazelton and Kimberly over the 3 yr (Table 3), it seems that warmer temperature favors higher yield Table 1. Mean squares from a combined analysis of variance for 25 dry bean cultivars evaluated for seed yield, seed weight, and days to maturity at Hazelton, Kimberly, and Parma, Idaho, from 1999 to 2001. Source df Mean square Yield Seed weight Maturityt Year (Y) 2 115268687** 60.0 264.0** Location (L) 2 39776287** 1300.3** 720.0** Y x L 4 19093568** 121.3** 1722.2** Rep (Y x L) 27 539987 4.7 4.6 Cultivar (C) 24 2676423** 382.9** 74.2** Y x C 48 1050075** 19.9** 23.6** L x C 48 868173 8.8** 15.6** YxLxC 96 417760** 6.8** 12.7** Error 647 160904 1.3 5.0 **Significant at the 0.01 probability level. tEvaluated in 2000 and 2001. of medium-seeded cultivars in southern Idaho. Furthermore, because Hazelton and Parma were more contrasting and a positive correlation existed between Kimberly and Parma, future studies should emphasize Hazelton and Parma if resources become limiting. Mean seed yield across locations and years for great northern shows that UI 425 followed by 'Matterhorn' (Kelly et al., 1999b) had the highest, and Weihing' (Coyne et al., 2000) followed by `UI 59' had the lowest yield (Table 2). The 549 kg ha-1 difference in yield between UI 425 and UI 59 represents an 18% yield gain over 52 yr (Table 4). Although, as a group, pink cultivars had relatively higher yield, there were no significant differences among them and only 3% gains were realized between the oldest (Viva) and newest highest-yielding (UI 537) cultivars. There is not a ready explanation for the higher overall yield potential for pink cultivars compared to great northern, pinto, and older red cultivars that have similar seed size, growth habit, and evolutionary origin. The higher yield potential for pink and relatively recently bred red cultivars such as NW 63 (Burke, 1982b) and UI 239 (Myers et al., 1997) was also reported by Mufioz-Perea et al. (2006) in Idaho and has been a common observation in yield trials conducted elsewhere in the western USA. It is noteworthy that pink bean is prominent in the pedigree of Bill Z (Wood et al., 1989) and UI 239, the highest-yielding cultivars, respectively, for the pinto and red market classes. Among pinto cultivars, Bill Z had the highest and UI 111, released in 1944, the lowest yield. The approximately 1000 kg ha -1 of difference between those cultivars represents a 35% yield increase over 43 yr. In the red market class, the 587 kg ha-1 difference between UI 239 with the highest and `UI 3' with the lowest yield represents 18% gain in yield in 55 yr. Interestingly, in the great northern class after the release of UI 425 in 1984, in the pink class after the release of UI 537 in 1990, in the pinto class after the release of WWW.CROPS.ORG 983 CROP SCIENCE, VOL. 47, MAY —JUNE 2007 Table 2. The year of release, growth habit, mean seed yield, stability index, seed weight, and days to maturity for 25 dry bean cultivars evaluated at Hazelton, Kimberly, and Pa rma, Idaho, from 1999 to 2001. Market class and cultivar Year of release Growthhabits Yield Stability index Seed weight Maturity Kimberly Hazelton Parma Mean