HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Mebrahtu, T. Articles by Andebrhan, T. Search for Related Content PubMed Articles by Mebrahtu, T. Articles by Andebrhan, T. Agricola Articles by Mebrahtu, T. Articles by Andebrhan, T. Related Collections Other Legumes

CROP BREEDING, GENETICS & CYTOLOGY

Evaluation of Soybean Breeding Lines for Corn Earworm Antibiosis

Agricultural Research Station, Virginia State Univ., P.O. Box 9061, Petersburg, VA 23806

* Corresponding author (tmebraht{at}vsu.edu)

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Corn earworm (CEW) (Helicoverpa zea Boddie) is the most serious insect pest of soybean [Glycine max (L.) Merr.] and often has reduced yield in much of the Mid-Atlantic and southern coastal plain of the USA. The objectives of this study were to determine the magnitude of genotype x year interaction (GYI) of antibiosis to CEW and to identify genotypes with stable performance. Thirty-two soybean breeding lines, including two susceptible and one line resistant to CEW, were planted in single row plots arranged in a randomized complete block design in 1996 through 1998 at Petersburg, VA. Petri dish assays were used to evaluate terminal foliage of field-grown soybean for resistance to CEW. Data were analyzed by means of genotype x year analysis of variance (ANOVA) and cultivar superiority performance measure (CSPM). Even though there was considerable variation among the genotypes tested during the three growing seasons, genotypes that ranked consistently either lower or higher in mean CEW larval weight were found. Among the genotypes tested, V89-2623, VS94-42, and VS94-27 generally had the highest CEW larval weight, whereas VS94-11, VS94-12, and VS94-26 were consistently in the lowest CEW larval weight group. Results suggest that CEW resistance of these latter three lines is relatively less affected by environment. They could serve successfully as genetic sources for breeding resistance to CEW. The rank correlation between the combined over year mean ranks and cultivar superiority performance measure value ranks was highly significant (r = 0.985). This suggests that either method of analysis could be suitable in identifying genotypes with stable resistance to CEW. The CEW larval bioassay technique was reliable in separating resistant and susceptible genotypes.

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
YIELD OF AGRICULTURAL CROPS is often reduced below its potential level because of pests and diseases. Control of these pests could, therefore, increase yield. However, pesticides, which are often very effective in controlling pest populations are generally not effective in totally eliminating pests. Additional pests may immigrate after application and some of the pest population may survive pesticide application.

Corn earworm is the most serious insect pest of soybean in much of the Mid-Atlantic and southern coastal plain of the USA (Stinner et al., 1980). The adult is attracted to soybean at the time of flowering (Johnson et al., 1975) where it prefers to oviposite on developing new foliage (Eckel et al., 1992). Small larvae tend to be found in greater density in young foliage, perhaps because of protection provided by rolled leaves (Kraemer et al., 1997). Late migration of older larvae to the developing pods and pod feeding may cause severe economic loss. Several management schemes to reduce CEW populations during or before the first two generations have been suggested (Fife and Graham, 1966; Knipling and Stadelbacher, 1983; Wiseman et al., 1984; Mueller et al., 1984). Another management approach is to control CEW on the third generation hosts through the use of minimal pesticides in combination with cultural control (Bradley and Van Duyn, 1980), biological control, and insect resistant cultivars. Even though insecticides provide adequate control of CEW, the use of host plant resistance would be an economically and ecologically better management technique. The genetic base from which most soybean breeders have selected for resistance to insect defoliation is narrow. Most lines selected for resistance are derived from one of three plant introductions (PIs), 171451, 227687, or 229358, that were identified by Van Duyn et al. (1971) as resistant to Mexican bean beetle (Epilachna varivestis Mulsant). These lines were later shown to have resistance to many other defoliating insects, including CEW (Hatchett et al., 1976; Luedders and Dickerson, 1977; Turnipseed and Sullivan, 1976; and Clark et al., 1972).

Breeding program decisions commonly rely on knowledge of the genetic structure of breeding populations and an understanding of the relative importance of genotype x environment interactions (GEI). This interaction is a common concern in plant breeding programs and germplasm evaluation trials. For plant breeders, it is often desirable to find genotypes that show little interaction with environments. Such genotypes may be regarded as stable. Stability is the ability of a genotype to avoid substantial fluctuation in injury over a range of environments, and is a breeding objective difficult to achieve. The cause of CEW injury stability could be due to physiological, morphological, and phenological mechanisms that impart stability. The objectives of this study were to determine the magnitude of genotype x year interaction (GYI) of antibiosis to CEW and to identify genotypes with stable performance.

	MATERIALS AND METHODS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
A total of 35 soybean genotypes were used for this study. Twenty-two were breeding lines from Virginia State University (VSU), six were from Virginia Polytechnic Institute and State University (VPI & SU), four lines were from ARS/USDA at North Carolina State University, and three were controls (Table 1) . The two susceptible controls (PI 399055 and Essex) and a resistant control (L76-0049) were selected on the basis of prior results from soybean insect defoliation studies, Kraemer et al. (1988)(1990) and Elden et al. (1982). The resistance found in L76-0049 was derived from the highly resistant PI 171451 (Rufener et al., 1986). The resistant parents of the VSU breeding lines included in this study were selected on the basis of the insect resistance reported by Kraemer et al. (1988)(1990). The VPI lines included in this study were tested during the 1995 growing season at Warsaw, VA, and produced yield equal to the standard check ‘Hutcheson’ (Glenn Buss, personal communication, 1996). The genetic materials from VSU were obtained by selecting several F₂ plants that showed less CEW damage from each population. At seed maturity, the selected F₂s were harvested individually and threshed and advanced to F₃. The genetic materials were advanced to the F₅ through single seed descent (SSD). F₅ single plants were selected at random and threshed individually. Breeding lines that produced high seed yield were identified from the bulk F₆ plant rows.

Foliage was collected when plants had reached R1 development stage (Fehr et al., 1971). Half to fully expanded terminal trifoliolates were excised, placed in plastic self-sealing bags, and transported to the laboratory in a cooler. The trifoliolates were separated into leaflets with petioles removed and placed into 150- by 15-mm plastic Petri dishes lined with moistened #2 filter paper. Four Petri dishes were used per genotype per replication, making a total of 12 Petri dishes per genotype. Two neonate CEW larvae were placed on the foliage in each Petri dish to allow for possible first instar mortality unrelated to leaf antibiosis (Kraemer et al., 1997; Kraemer, 2001). The Petri dishes were incubated in an environmental chamber at 25°C and 14:10 (light:dark) photoperiod. Although relative humidity within the environmental chamber was not controlled (50–60%), the filter paper within each Petri dish was kept moist by applying approximately 1 mL of water every other day. The number of larvae per Petri dish was reduced to one after 4 to 5 d. After 10 d, the larvae were weighed and mortality was determined. When more than one CEW larvae was found in a Petri dish, probably from eggs brought in on the foliage, the largest larva was selected for analysis.

Statistical Analysis
The data for each year were analyzed by (SAS,1996) as a randomized complete block design. After testing the homogeneity of error variance, combined data over years analysis was conducted by a genotype x year interaction (GYI) model. Year was tested for significance using replication within year (rep[yr]) as the error term, genotype was tested with GYI as the error term, and GYI was tested by means of the pooled error term. Means were separated by least significant difference (LSD) at the 5% probability level as described by Steel and Torrie (1980).

The number of replications (R) required to show specific differences between genotypes is given by the following formula modified from Mendenhall and Scheaffer (1973):

where t_/2 is the t-values associated with significance level of the t-test ( = 0.05 for our calculation) and are dependent on the degree of freedom (df) of sample variance. t is the t-value associated with accepting a false null hypothesis based on the t-test (ß = 0.05 for our calculation), ² is the error variance, d is the true difference between two genotypes.

The previous formula can be modified to determine the number of environments (E) required to show specific differences between genotypes with different levels of GYI

where E is the number of environment, R is the number of replications within an environment, and ²_GE is the interaction variance component. The rational for determining the number of replication or environments to detect specific difference between treatment means rather than overall treatment effects has been discussed by Carter et al. (1983) and Reese et al. (1988).

Both GYI and cultivar superiority performance measure (CSPM) methods of analyses were used for the data. The CSPM values estimated are the mean squares of the differences between an entry mean and the maximum mean of a location, summed and divided by twice the number of locations (Lin and Binns, 1998). The genotype CEW larval means generated by GYI were ranked in descending order from 1 to 35. One being with the highest value and 35 being with the lowest CEW larval weight. Similarly the values generated by the CSPM were ranked from 1 to 35. The genotype with the lowest CSPM value was given a score of 1 and the genotype with the highest value was given a rating of 35. A Spearman's rank correlation was used to determine the association of rank scores generated by G x E analysis with the mean ranking of CSPM scores.

	RESULTS AND DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Weather patterns varied in the 1996, 1997, and 1998 growing seasons. In 1997, soil moisture was marginal at planting, and dry conditions persisted through most of the growing season. The combination of prolonged dry conditions combined with high temperatures affected soybean growth and development. Significant differences (P 0.01) among the genotypes tested were found for CEW larval weight for each growing season. In 1996 and 1998, the genotypic variances exceeded that of error variance, while in 1997, they did not. The significant genotypic differences indicated that genetic variation exists for CEW resistance among the tested genotypes. This offers promise for selection and improvement through breeding and selection.

A significant GYI was observed. This interaction was the result of a change in the magnitude and rank of the differences among the genotypes in the different years. The significant GYI observed suggests that the performance or response of the genetic materials used in this study were not stable from one growing season to another. A stable genotype is defined as a genotype able to avoid substantial fluctuation in CEW resistance over a range of environments, a breeding objective that is difficult to achieve. Stability could be due to physiological, morphological, and/or phenological mechanisms.

The sum of squares of the sources of variation expressed as percentages revealed the relative contribution of each source to the total variance. The pooled error variances contributed proportionally the highest (55%) to the total variability, followed by replication within year (19%), and genotype (11%). The proportional contribution of year to the total variation was only 9%. The GYI contributed 6% to the total variation. Among the major components, the interaction component contributed the least. The pooled error sum of squares on the other hand contributed proportionally more than genotype and GYI to the total sum of squares.

The number of replications required to detect a 10, 20, 30, 40, and 50% difference of magnitude between means of genotypes for CEW larval weight ( = 0.05, ß = 0.50) were 181, 45, 20, 11, and 7, respectively. If the power of the test was increased to 0.9 (1 - ß), then the number of replications required to show specific genotypic differences would increase 2.5 fold. The number of environments required to show specified differences in CEW larval weight between two genotypes was calculated assuming ²_GE = 0.1 ² (Table 2) . Detection of 40% difference of the mean between two genotypes would require four and three environments with four and five replications per environment, respectively. However, if the specified difference between two genotypes was 50% of the mean, then four and three environments, respectively for two and three replications are required or two environments with four replications per environment.

The CEW larva mean weight distribution in 1997 was similar to 1996 in that 37% had lower values than the overall mean, 26% were equal to the overall mean, and 37% had higher values than the overall mean. In 1997, lines that registered the lowest CEW larval weight mean values were VS94-05, VS94-11, VS94-12, VS94-44, VS94-45, and V91-0613. The mean larvae weight in 1998 revealed that about 46%, had CEW mean larval weight lower than the overall mean, 26% had weights about equal to the mean, and about 28% of the genotypes had weights higher than the overall mean. The lines resulting in the lowest CEW larval weights were VS94-11, VS94-12, VS94-16, VS94-17, VS94-18, and VS94-26. The mean larval weight was higher in 1996 than 1997 and 1998. The susceptible checks, Essex and PI 399055, had 156 and 170 mg overall mean larval weights, respectively, which were significantly higher than the 81 mg mean of L76-0049, a resistant check.

In comparing genotypic CEW larval weight values averaged over the 3 yr, we found about 29% with lower values than the overall mean, 11% with equal values to the mean and the largest group about, 66%, had weights higher than the overall mean. The overall CEW mean larval weight of the genotypes ranged from 55 mg for VS94-12 to 169 mg for PI 399055 with a mean of 124 mg (Table 3) . The mean of the top 10 high ranking genotypes ranged from 169 to 139 mg. The genotypes with CEW larval weights significantly higher (0.05) than overall mean were PI 399055, VS94-42, V89-2623, Essex, and VS94-13. Over three growing seasons three lines, VS94-11, VS94-12, and VS94-26, consistently showed lower CEW mean larval weights than the other genotypes. These results suggest that the expression of CEW resistance in these three lines is relatively less affected by environmental fluctuations. They could serve as genetic resources in a breeding program aimed at reducing CEW larval foliage damage in soybean. In each of the three growing seasons and in the overall average results, the susceptible checks, Essex and PI 399055, exhibited higher values than the overall genotypic means, whereas the resistant check, L76-0049, was significantly lower than the overall genotypic means.

The Spearman's rank correlation between the combined over years mean ranks and CSPM ranks was 0.985 (0.01), and between the overall rank average was 0.996 (0.01). Moreover, the rank of the overall mean and the rank average was 0.995 (0.01). These high correlation values indicated that either GYI or CSPM methods of analyses are ideal in identifying genotypes that are significantly resistant to CEW and stable.

The mean CEW larval weight of the two susceptible standard genotypes, PI 399055 and Essex, were 169 and 157 mg, respectively. The resistant standard (L76-0049) had a CEW larval weight of 81 mg which was significantly lower than both the susceptible checks, PI 399055 and Essex. In this study, we found VS94-12 to have 33% lower CEW larval weight than L76-0049. The other genotypes that are higher than L76-0049 by less than 5, 14, 17, and 28% were VS94-11, VS94-21, VS94-16, and VS94-26, respectively. This suggested that first, the screening technique used was adequate to make a clear demarcation between the CEW resistant and susceptible lines and second that sufficient variability exists among the genotypes tested to make further improvement possible through selection and hybridization.

The broad range of resistance exhibited by PIs 171451, 227687, and 229358 have enhanced their usefulness in breeding programs. However, they are all from late maturity groups (MGs) VII and VIII and represent a relatively narrow genetic base from which to select for resistance. The addition of the new breeding lines VS94-11, VS94-12, VS94-16, VS94-21, and VS94-26 includes some additional resistant parentage that may increase genetic diversity. Breeding lines VS94-11 and VS94-12 were selected from a L76-0049 x Essex cross. The resistant parent of L76-0049 is from one of the old resistant standards, PI 171451. However, breeding lines VS94-21, VS94-26, and VS94-16 were selected from York x PI 416937, Bay x PI 416937, York x PI 416925 crosses, respectively. Plant introductions 416937 and 416925 are from MG VI and were reported to have resistance to Mexican bean beetle defoliation (Kraemer et al. 1990). PI 416937 is also resistant to drought and tolerant of aluminum (Carter and Ruffy, 1992). The four breeding lines VS94-12, VS94-16, VS94-21, and VS94-26 belong to MG VI and VS94-11 belongs to MG V. All these lines were entered into Mid-Atlantic cooperative yield test during 1994 and produced seed yield equal to the standard checks ‘Brim’ and ‘Twiggs’.

Determining the GEI is common in plant breeding programs and germplasm evaluation trials because it is often desirable to find genotypes that show little interaction with the environment. While some lines had a stable level of resistance over environments, others did not. Because the GYI was significant, tests in multiple years will be required to adequately evaluate resistance to CEW. Selections among these genotypes for CEW resistance would be different in each growing season.

The resistance measured by Petri dish assays is of the antibiosis type (Painter, 1951). Resistance resulting from nonpreference was not evaluated and could also be a factor in the field. However, resistance based on antibiosis is more reliable because nonpreference may vary with the type of alternative plant hosts in the area. Petri dish bioassays have been shown to correlate well with field evaluations (Kraemer 2001). The results of our study also demonstrated that the Petri dish assay technique was reliable to separate susceptible from resistant genotypes. On the basis of the resource allocation estimates, we found that the larval weight differences between genotypes of 40% of the experimental mean can be detected with three environments and five replications per environment using this bioassay. The data for this study could assist researchers in the design and resource allocation for future experiments involving CEW antibiosis study. Moreover, these results suggested that sufficient genetic variation exists among genotypes to make improvement through breeding. The advantage of using host plant resistance to reduce yield losses to CEW over the use of alternative methods generally outweighs its limitation. Economic benefits offered by host resistance are its ease of adaption and reduction of variable costs. From an ecological perspective, the main benefits of host plant resistance are its specificity and compatibility with integrated pest management (IPM).

	ACKNOWLEDGMENTS

 
The authors would like to thank Drs. Glenn Buss from VPI & SU and Thomas E. Carter, Jr., ARS/USDA North Carolina State University, Raleigh, NC for providing soybean breeding lines to be included in the study.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Journal Series no. 226. The use of any trade names and/or vendors does not imply approval to the exclusion of the other products or vendors that may also be suitable.

Received for publication October 1, 2001.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 

• Bradely, J.R., Jr., and J.W. Van Duyn. 1980. Insect pest management in North Carolina soybean. p. 343–354. In F.T. Corbin (ed.) World Soybean Research Conference II. Proc. Westview, Boulder, CO.
• Carter, T.E., Jr., J.W. Burton, J.J. Cappy, D.W. Israel, and H.R. Boerma. 1983. Coefficients of variation, error variances, and resource allocation in soybean growth analysis experiments. Agron. J. 75:691–696.[Abstract/Free Full Text]
• Carter, T.E., Jr., and T.W. Ruffy, Jr. 1992. Soybean Plant Introductions exhibiting drought and aluminum tolerance. p. 335–346. In C.G. Kuo (ed.) Adaptation of food crops to temperature and water stressproceedings of an international symposium, Tawain. 13–18 Aug. 1992. Publ. 93-410. Asian Vegetable Reseach and Development Center, Shanhua, Taiwan.
• Clark, W.J., F.A. Harris, F.G. Maxwell, and E.E. Hartwig. 1972. Resistance of certain soybean cultivars to bean leaf beetle, striped blister beetle, and bollworm. J. Econ. Entomol. 65:1669–1672.
• Eckel, C.S., J.R. Bradley, Jr., and J.W. Van Duyn. 1992. Reduction in soybean yield and quality from corn earworm flower feeding. Agron. J. 84:402–409.[Abstract/Free Full Text]
• Elden, T.C., R.L. Bernard, C.R. Edwards, and M. Kogan. 1982. Notice of release of five lines of soybean resistant to Mexican bean beetle. Plant Resist. Insects Newl. 8:22.
• Fehr, W.E., C.E. Caviness, P.T. Burmood, and J. Pennington. 1971. Stage of development descriptions of soybeans, Glycine max (L.) Merr. Crop Sci. 11:929–931.[Abstract/Free Full Text]
• Fife, L.C., and M. Graham. 1966. Cultural control of overwintering ball worm and tobacco budworm. J. Econ. Entomol. 59:1123–1125.
• Hatchett, J.H., G.L. Beland, and E.E. Hartwig. 1976. Leaf-feeding resistance to bollworm and tobacco budworm in three soybean plant introductions. Crop Sci. 16:277–280.[Abstract/Free Full Text]
• Johnson, M.W., R.E. Stinner, and R.L. Tabb. 1975. Ovipositional response of Heliothis zea (Boddie) to its major hosts in North Carolina. Environ. Entomol. 4:291–297.
• Knipling, E.F., and E.A. Stadelbacher. 1983. The rationale for area wide management of Heliothis (Lepidoptera:Nictuidae) populations. Bull. Entomol. Soc. Am. 29:29–37.
• Kraemer, M.E. 2001. Insect resistance in vegetable and tofu soybeans. J. Entomol. Sci. 36:57–66.
• Kraemer, M.E., M. Rangappa, T. Mebrahtu, and P.S. Benepal. 1988. Field evaluation of soybean for Mexican bean beetle resistance. I. Maturity Groups VI, VII, and VIII. Crop Sci. 28:497–499.[Abstract/Free Full Text]
• Kraemer, M.E., M. Rangappa, T. Mebrahtu, and P.S. Benepal. 1990. Field evaluation of soybean for Mexican bean beetle resistance. II. Maturity Groups III, IV, and V. Crop Sci. 30:374–377.[Abstract/Free Full Text]
• Kraemer, M.E., M. Rangappa, and A. Mohamed. 1997. Evaluation of vegetable soybean genotypes for resistance to corn earworm (Lepidoptera:Noctuidae). J. Entomol. Sci. 32:25–36.
• Lin, C.S., and M.R. Binns. 1988. A superiority measure of cultivar performance for cultivar x location data. Can. J. Plant Sci. 68:193–198.
• Luedders, V.D., and W.A. Dickerson. 1977. Resistance of selected soybean genotypes and segregating populations to cabbage looper feeding. Crop Sci. 17: 295–396.
• Mendenhall, W., and R.L. Scheaffer. 1973. Mathematical statistics with applications. Duxbury Press, North Scituate, MA.
• Mueller, T.F., V.E. Harris, and J.R. Philips. 1984. Theory of Heliothis (Lepidoptera:Noctuidae) management through reduction of the first generation: Critique. Environ. Entomol. 13:625–634.
• Painter, R.H. 1951. Insect resistance in crop plants. MacMillan, New York.
• Reese, P.F., H.R. Boerma, and R.S. Hussey. 1988. Resource allocation in experiments measuring soybean tolerance to soybean cyst nematode. Crop Sci. 28:589–593.[Abstract/Free Full Text]
• Rufener, G.K. II, R.B. Hammond, R.L. Cooper, and S.K St. Martin. 1986. Mexican bean beetle (Coleoptera: Coccinellidae) development on resistant and susceptible soybean lines in the laboratory and relationship to field selection. J. Econ. Entomol. 79:1354–1358.
• SAS Institute. 1996. SAS user's guide: Statistics, 5th ed. SAS Institute, Cary, NC.
• Steel, R.G.B., and J.H. Torrie. 1980. Principles and Procedures of Statistics. McGraw-Hill, New York.
• Stinner, R.E., J.R. Bradley, Jr., and V. Duyn. 1980. Sampling Heliothis sp. on soybean. p. 407–421. In M. Kogan and D.C. Herzog (ed.) Sampling methods in soybean entomology. Springer-Verlog, New York.
• Turnipseed, S.G., and M.J. Sullivan. 1976. Plant resistance in soybean insect management. p. 549–557. In L.D. Hill (ed.) World soybean research. Intgerstate Pringers and Publishers, Danville, IL.
• Van Duyn, J.W., S.G. Turnipseed, and J.D. Maxwell. 1971. Resistance in soybeans to the Mexican bean beetle. I. Sources of Resistance. Crop Sci. 11:572–573.[Abstract/Free Full Text]
• Wiseman, B.R., N.W. Widstrom, and W.W. McMillian. 1984. Increased seasonal losses in field corn to corn earworm. J. Georgia Entomol. Soc. 19:41–43.

This article has been cited by other articles:

				T. Mebrahtu and M. Kraemer Registration of VS94-11, VS94-12, and VS94-21 Soybean Germplasm Lines Resistant to Corn Earworm Foliar Damage Crop Sci., March 1, 2004; 44(2): 701 - 702. [Full Text]

This Article Abstract Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Mebrahtu, T. Articles by Andebrhan, T. Search for Related Content PubMed Articles by Mebrahtu, T. Articles by Andebrhan, T. Agricola Articles by Mebrahtu, T. Articles by Andebrhan, T. Related Collections Other Legumes