On-Farm Versus Agricultural Experiment Station Evaluation and Improvement of Intrinsic Characteristics of Landrace Common Bean Cultivars for Sustainable Farming Systems in the Twenty-First Century

Final Report for SW02-038

Project Type: Research and Education
Funds awarded in 2002: $167,717.00
Projected End Date: 12/31/2005
Matching Non-Federal Funds: $23,383.00
Region: Western
State: Idaho
Principal Investigator:
Shree Singh
University of Idaho
Expand All

Project Information

Abstract:

Common Red Mexican, Matterhorn, Mesa, and Othello were drought resistant and competed well with weeds. Common Red Mexican’s resistance was lowest within cultivars. Yield and plant and seed uptake and concentration of nutrients were low in stressful environments. Forty-four F5:7 breeding lines selected from each of two populations in each of four production systems were evaluated in their respective system. Seed of 262 F5:7 breeding lines from UI 320 x Common Red Mexican was produced. Two meetings were held and five field-tours conducted to promote adoption. Six presentations were made and two articles prepared. C. Muñoz earned a Ph.D. degree.

Project Objectives:

This project focuses on identification, genetic improvement, and adoption of the most promising dry bean landraces and cultivars and management practices for low-input sustainable organic and conventional production systems for Idaho in particular and the Western U.S. at large for the 21st century. The specific objectives in 2005 were to (1) complete agronomic characterization and data analyses for three dry bean landraces and 13 cultivars evaluated in seven production systems in 2003 and 2004, (2) identify the most promising landraces and cultivars within and across production systems for their subsequent on-farm adoption through field tours and other means, (3) evaluate 44 high-yielding F5-derived F7 (F5:7) breeding lines and five parents from each of two populations selected independently in each of four production systems, (4) multiply seed of 262 F5:7 breeding lines from a cross of drought-susceptible pinto cultivar UI 320 x highly drought-resistant Common Red Mexican landrace for subsequent identification and mapping of favorable alleles and quantitative trait loci (QTL) imparting drought resistance to facilitate their transfer from landrace into new cultivars, and (5) train students and young researchers in breeding and management practices for sustainable dry bean production systems.

Introduction:

Dry bean (Phaseolus vulgaris L.) production in the Western U.S. mostly occurs under intensively managed high-input production systems. Consequently the production costs are high and farmers unable to compete are being forced out of production. Thus, there is increasing pressure to reduce production costs. Moreover, in recent years, there has been an increasing awareness for cleaner environments, need to conserve nonrenewable resources, and adopt low-input sustainable production systems. Also, there has been a gradual increase in organic bean production. Thus, there is high demand and strong justification for dry bean cultivars that are less dependent on fertilizer, pesticide, and water and have reduced production costs for the growers. However, high-yielding drought- and low soil fertility-resistant dry bean cultivars adapted to sustainable organic and conventional production systems are not currently available in the U.S. Since the beginning (>75 years), in the Western U.S. and the rest of the nation, dry bean cultivars were only bred for resistance to diseases, better seed quality, plant type, and maturity under high-input irrigated conditions at Agricultural Experiment Stations (Dean, 2000). Breeding for soil compaction, weed competition, nutrient and water use efficiency, and sustainable organic and conventional production systems were not practiced. Similarly, on-farm dry bean breeding with farmers’ participation was not carried out. Consequently, resistance to drought and nutrient deficiency, soil compaction, and ability to compete with weeds, all of which were intrinsic characteristics of dry bean landraces grown by Native Americans in low-input subsistence production systems in the Western U.S. for millennia, appear to have been inadvertently lost in most modern cultivars (Singh et al., 2004). Although, modern cultivars require reduced fungicides, they depend on heavy use of synthetic fertilizers, irrigation water, and pre-and post-emergence herbicides.

Research

Materials and methods:

Experiment # 1. Identification of Intrinsic Characteristics of Dry Bean Landraces and Cultivars.
Three dry bean landraces and 13 cultivars released between 1932 and 1998 belonging to Durango race and representing great northern, pinto, and red market classes were evaluated on-farm in four production systems, namely high-input organic (OGH), low-input organic (OGL), high-input conventional (OFC), and low soil fertility (OLF) in 2003 and 2004. They were also evaluated at University of Idaho-Kimberly Research and Extension Center in high-input conventional (RFC), drought-stressed (RDS), and >50 years of continual bean (RCB) production systems. Each market class had a representative landrace or selection thereof. Most of the 13 cultivars were selected based on their seed yield in trials carried out between 1999 and 2001 in southern Idaho (Singh et al., 2001a, b). They were arranged in a randomized complete block design with four replications. Each plot consisted of four or eight rows of 25 or 50 ft length and spaced 22 inches apart. An average of 7 seeds / linear ft were planted.
Pre-plant soil samples were analyzed. In addition, fertilizer and irrigation was applied and management of weeds was handled according to the standard practices of each production system determined by the participating farmers. Nonetheless, OGL and OLF were not fertilized at all and RCB soils were heavily compacted. In the RDS production system, three irrigations were skipped in each year. The amount of irrigation water applied in RFC and RDS was recorded. In two landraces and four cultivars, the gravimetric water content was estimated by taking soil samples up to 2 m depth after planting, one day before and two days after each irrigation, and one day before harvest. Also, soil water potential was estimated in centibars at 0.23, 0.46, and 0.92 m depths, using soil moisture sensors or watermarks (Irrometer Company, Inc, Riverside, California) connected to AM400 data-loggers (Hansen Company, East Wenatchee, Washington). The mean daily precipitation; minimum, maximum, and mean temperature; solar radiation; evapotranspiration Kimberly-Penmann; mean humidity; and average wind speed were recorded from the Twin Falls Agrimet Station (http://www.usbr.gov/gp/agrimet/index.cfm), which is located at latitude 42o 32’ 46” and longitude 114o 20’ 43” (<1,000 m from the experiments) at the USDA-ARS- Northwest Soil and Irrigation Research Laboratory in Kimberly.
Growth habit was recorded during flowering and verified at the end of pod filling. Days to maturity was recorded when 90% of the pods changed from green to yellow color. Dry matter weight was determined by cutting 10 plants/plot above ground at maturity and drying them at 60 0C for five days. Biomass yield was estimated by multiplying dry matter weight by plant density. The two or six central rows (25.60 m2) were cut at maturity and threshed 8 to 12 days later. Harvest index was determined as the ratio between seed and biomass yield. Weight of 100 seeds was recorded. Drought intensity index, percent reduction in seed yield due to drought, and drought susceptibility index were calculated according to Fischer and Maurer (1978).
Plant uptake and concentration of 15 nutrients were determined from the 10 plants cut above ground at maturity. Similarly, 100 seeds taken randomly from each plot were used for assessing seed uptake and concentration of 15 nutrients. The dry ash procedure was used for determining Al, B, Ca, Cu, Fe, K, Mg, Mn, Na, P, S, Si, and Zn. The dynamic flash combustion technique was used for determining C and N. In both procedures, plant and seed samples were dried, weighed, and ground through a cyclone mill. In the dry ash procedure, 0.5 g of each plant or seed sample was placed in 100 ml low form beakers and then put in a furnace at 500 oC for 4 to 6 hours. The ash samples were cooled and weighed. Then 10 ml of 1N HNO3 were added and heated on a hot plate until steam formed on beaker sides. Water was added to 50 ml. The solutions were stirred using a Policeman Teflon stirrer and filtered through #50 Whatman filter paper. The solutions were analyzed with an ICP-OES 4300 DV (inductively coupled plasma optical emission spectrometry), (Perkin Elmer, Inc., Wellesley, Massachussetts). Multi-elements mixed from standard solutions were used to calibrate each element. The concentration of each element was determined comparing the sample intensity with the calibration plots. For further details readers should refer to Muñoz (2005) and three or four refereed articles to be published within the next several months.
A mixed model (McIntosh, 1983) was used whereby years and replications were considered random effects, and production systems and landraces and cultivars were fixed effects. Data for each year and production system were analyzed separately and the homogeneity of variances was tested for the combined analyses (Bartlett, 1947). Also, simple correlation coefficients among traits were calculated using the mean values. All data were analyzed using the SAS (v 9.1.3) GLM procedure (SAS, 2004). Subsequently, ANOVA, summary of results tables, and figures were prepared.

Experiment # 2. On-Farm Adoption of Promising Dry Bean Cultivars.
Four to eight rows wide and 100 to 300 ft long strip-plots of high yielding drought and low soil fertility resistant pinto Mesa and Othello and red NW 63 and UI 239 were planted in all seven production systems except RCB (lack of space) without replication in 2005. Common Red Mexican, due to its susceptibility to the Bean common mosaic virus (BCMV, a potyvirus endemic in the Western U.S.), and Matterhorn, due to its poor seed quality, were not included in the strip-plots. Strip-plots and semi-commercial plots (minimum of 1 acre plots) of these and other promising breeding lines and cultivars would need to be planted by participating farmers in all production systems in 2006 and beyond to maximize their adoption and impact on reducing irrigation water, fertilizer, and pesticide inputs.

Experiment 3. Breeding strategies for high- and low-input organic and conventional production systems.
Drought- and low soil fertility-resistant Common Red Mexican, Mesa, Matterhorn, and NW 63 were crossed with contrasting cultivars or breeding lines such as Buster, LeBaron, Topaz, and VAX 3 to produce two multiple-parent populations, namely Topaz /// Matterhorn / Mesa // Buster / Common Red Mexican and LeBaron /// VAX 3 / Common Red Mexican // Matterhorn / NW 63. Independent selection for seed yield among 220 F2–derived F3 (F2:3) to F2:5 families in each population was practiced with farmers’ participation in OFC, OGH, OGL, and RFC production systems between 2002 and 2004. Forty-four F5:7 breeding lines from each population and production system, developed using the single-seed descent method (Urrea and Singh, 1994) from the 44 highest yielding F5 families (i.e., 20% selection), and their five parents were evaluated for seed yield in their respective production systems in 2005. A 7 x 7 partially balanced lattice design with three replications was used.

Experiment # 4. Identification and mapping of favorable alleles and quantitative trait loci (QTL) imparting drought resistance in Common Red Mexican landrace.
Drought-susceptible pinto cultivar UI 320 was crossed with drought-resistant Common Red Mexican landrace in 2002. Two hundred sixty-two F5:7 breeding lines were developed, using the single-seed descent method from the F2 to F5. Each breeding line was planted in a four-row plot, 10 ft long without replication at Kimberly, Idaho in 2005.

Research results and discussion:

Experiment # 1. Identification of Intrinsic Characteristics of Dry Bean Landraces and Cultivars. All data collected on three landraces and 13 cultivars were analyzed individually and combined for 2003 and 2004. The detailed results will be published in a series of refereed papers within the next several months. Only the highlights will be provided here.
Effects of production systems, landraces and cultivars, years, and their interactions were large and significant (P<0.01) for biomass and seed yield, 100-seed weight, and number of days to maturity. They were also significant for plant and seed uptake and concentration of most of the 15 elements within and across seven production systems. Production systems x landrace and cultivar interactions were significant for most traits, indicating that the rank order of landraces and cultivars differed in different production systems over the two years. The mean seed yield of three dry bean landraces and 13 cultivars in OGL was not associated with yields in any other production system. The mean seed yield and other traits pooled over two years and three dry bean landraces and 13 cultivars for each of the seven production systems is given in Tables 1 to 3. Large differences among three dry bean landraces and 13 cultivars for all traits occurred in each production system. As an example, we will briefly discuss the results of the RDS and RFC production systems here. Details can be found in a Ph.D. dissertation “Differences among dry bean landraces and cultivars for seed yield, water use efficiency, and nutrient concentration in drought-stressed and non-stressed environments” of C.G. Muñoz, University of Idaho, 2005.
The non-stressed (RFC) plots received seven irrigations in 2003 and five in 2004, and drought-stressed (RDS) only four in 2003 and two in 2004. Most water use occurred within the top 0.5 m soil in both the RFC and RDS, and differences in water potential among landraces and cultivars below 50 cm depth were non-consequential even under the more severe RDS (0.62 drought intensity index) production system in 2003 (Figure 1). Thus, dry bean cultivars mostly take up water and nutrients from the upper 50 cm soil profile even in stressful production systems. Drought reduced mean seed yield by 62% in 2003 and by 27% in 2004. Mean seed yield for three dry bean landraces and 13 cultivars over the two years was 2739 kg ha-1 in RFC and 1683 kg ha-1 in RDS. Reduction in seed weight due to drought stress ranged from 0 to 22 % in 2003 and from –3 to 10 % in 2004. However, in both years seed weight of Common Red Mexican was not affected by drought stress. Drought-susceptible UI 59, Bill Z, Common Pinto, Topaz, UI 259, and UI 320 had higher reduction in seed weight due to drought stress. In contrast, in addition to Common Red Mexican, other drought-resistant cultivars, namely Mesa, Matterhorn, and Othello, had the lowest reduction in seed weight over the two years. In 2003, all landraces and cultivars except UI 259 took longer to mature in RDS than in RFC production system. These differences ranged from 1 day for Mesa and Common Red Mexican to 14 days for drought- susceptible Bill Z, Common Pinto, and Topaz to reach maturity in RDS. Under moderate drought in 2004, all landraces and cultivars except UI 465 and Topaz either matured the same day in RDS and RFC or took 1 to 6 days longer in RFC than in RDS.
Considering drought intensity index, mean seed yield in RFC and RDS, percent reduction in seed yield due to drought stress, and drought susceptibility index values, three dry bean landraces and 13 cultivars could be classified into three groups in 2003 (Figure 2). Common Red Mexican and Mesa, which yielded high in both RDS and RFC and had a below average reduction due to drought stress, formed the first group. While Othello and Matterhorn also possessed high and NW 63 and UI 239 moderate level of drought resistance, they yielded moderately in RFC production system. Buster and UI 259 yielded well in RFC but they were susceptible to drought. Drought resistance inadvertently reduced from Common Red Mexican landrace to intermediate level in ‘NW 63’ and ‘UI 239’ released in 1979 and 1983, respectively, and more recently released ‘UI 259’ (1996) and UI 320 (1996) were susceptible to drought stress (Table 4). All early maturing cultivars except Othello (e.g., UI 59, US 1140, Common Pinto, Topaz, UI 320, and LeBaron) were also drought susceptible (Figure 2).
Several broad- and narrow-leafed weed species were the most severe problem from early growth stages in OGH and OGL production systems. It was very encouraging to note that Matterhorn and USPT-CBB-1 (without ability to climb) and Mesa (with ability to climb on weeds), among others, competed well with mixed weed populations in both production systems. In general, they also had high biomass and seed yields across other production systems.
Plant and seed uptake and concentration of most elements in dry bean landraces and cultivars largely depended upon the soil nutrient and moisture status, and not the production system per se. Nonetheless, in general, values for most nutrients in stressful production systems (e.g., OGL, OLF, RCB, RDS) were lower than in non-stressful systems. For example, mean values for N, P, Zn, Cu, Fe, and Mn, given in Tables 2 and 3, indicate that values for some nutrients (e.g., N, P, and K) decreased, whereas those of others (e.g., Cu, Mn, and Zn) increased due to drought stress. Plant and seed Zn concentration had the most marked increase in RDS in both drought susceptible and resistant landraces and cultivars and its translocation efficiency [i.e., the ratio of Zn uptake ratio (seed/whole plant) to harvest index (seed yield/whole plant weight)] was positively associated with harvest index across production systems (Figure 3). Nonetheless, plant and seed concentration for most nutrients for three dry bean landraces and 13 cultivars differed significantly across production systems. An example is given for two cultivars with high and two low seed Zn concentration for the four on-farm production systems in Table 4. Bill Z had high seed Zn concentration in OFC and low in OLF production system. In contrast, Buster had low in OFC and high in OLF, OGH, and OGL production systems.

Experiment # 2.
For researchers as well as participating farmers and others attending the field tours, it was much easier to visualize the virtues of the four most promising cultivars planted for the first time in the strip-plots in all production systems except RCB in 2005. There was high demand for their seed, thus justifying accelerated seed production and additional strip-plots (and semi-commercial plantings) in 2006 and beyond.

Experiment # 3. Breeding strategies for high- and low-input organic and conventional production systems.
Harvested seed of 44 F5:7 breeding lines selected from each of two multiple-parent populations and each of four production systems and their five parents evaluated in their respective production system in a partially balanced 7 x 7 lattice design with three replications in 2005 is being cleaned, and data for yield, 100-seed weight, and seed color will be recorded. The trial needs to be repeated at least one more year to identify the 8 to 10 highest yielding breeding lines within each production system for their subsequent evaluation in comparative yield trials across the four production systems. Breeding efficiency of each production system will be determined and high-yielding cultivars identified for adoption within and across production systems.

Experiment # 4. Identification and mapping of favorable alleles and quantitative trait loci (QTL) imparting drought resistance in Common Red Mexican landrace.
An average of 1 kg seed of each of 262 F5:7 breeding lines from drought-susceptible pinto cultivar UI 320 x drought-resistant Common Red Mexican population was produced at Kimberly in 2005. The breeding lines along with the two parents will need to be evaluated in RFC and RDS production systems at two contrasting locations (e.g., Kimberly and Parma, Idaho) for at least two years to identify and map favorable alleles and QTL controlling drought resistance in Common Red Mexican landrace.

Research conclusions:

The four most promising dry bean cultivars resistant to drought and low soil fertility that also competed well with mixed weed populations in the OGH and OGL production systems, namely Mesa, Othello, NW 63, and UI 239 were identified.
The eight to ten highest yielding F5:7 breeding lines from each of the two populations and in each of the four production systems will be identified for comparative evaluations across production systems after the evaluation of the 44 breeding lines and five parents in 2006 as described in Experiment # 3.
An average of 1 kg seed of each of the 262 F5:7 breeding lines from drought susceptible pinto cultivar UI 320 x drought resistant Common Red Mexican population was produced at Kimberly in 2005.

Participation Summary

Educational & Outreach Activities

Participation Summary:

Education/outreach description:

One non-refereed article and one article for publication in a refereed journal were prepared in 2005. Meeting of all participants was held in the spring to discuss results of 2004 and make the work plan for 2005. In addition, five field-tours (three organized by participating farmers) were carried out in August-September to discuss results and promote on-farm adoption of the four most promising cultivars resistant to drought and low soil fertility that also competed well with mixed weed populations in the OGH and OGL production systems, namely Mesa, Othello, NW 63, and UI 239. Four (three invited) national and two international presentations were made.

Project Outcomes

Project outcomes:

Farmer Adoption

As a result of participating in the project since 2002 and seeing the performance of selected dry bean cultivars in the strip-plots in 2005, there was a high demand for seed of Mesa, Othello, NW 63, UI 239, and USPT-CBB-1. While sufficient seed of each of these is available (harvested from the yield trails and strip plots) for experiments and for use by the participating farmers in 2006 and beyond, Breeder and Foundation seed of pinto Mesa and USPT-CBB-1 will need to be produced in 2006 and subsequent years. The two cultivars of red market class, namely NW 63 was released in 1979 and UI 239 in 1993, hence their Breeder, Foundation, Registered and Certified seeds are available from the Idaho and Washington Agricultural Experiment Stations. On-farm adoption of the four cultivars could be accelerated by additional strip-plots and semi-commercial plantings in 2006 and beyond.

Recommendations:

Areas needing additional study

Three more years of research is needed to successfully complete Experiments # 3 and 4. This will produce the first group of dry bean cultivars ever developed and compared on-farm with farmers participation in each of OGL, OGH, OFC, and RFC production systems. New dry bean cultivars would be expected to reduce dependence on irrigation water, fertilizer, pesticides, and manual labor, and facilitate sustainable organic and conventional production of dry bean in the Western U.S. Also, favorable alleles and QTL imparting drought resistance in Common Red Mexican landrace will be identified and mapped to facilitate their subsequent transfer into new cultivars.
The frequency and timing of irrigation and fertilizer requirement for the most promising cultivars identified in Experiment # 1 could be determined and breeding, genetics, and physiology research intensified to maximize yield and water and nutrient use efficiency, and reduce cost for sustainable dry bean production in the Western U.S.
In addition to participating farmers, other industry personnel, students, and county Extension Educators of agricultural crops could be educated through Field Tours, Bean Schools, and other means.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.