Diversity in the tropical multipurpose shrub legumes Cratylia argentea (Desv.) O. Kuntze and Flemingia macrophylla (Willd.) Merrill.

摘要

Cratylia argentea (Desv.) O. Kuntze and Flemingia macrophylla (Willd.) Merrill are promising tropical multipurpose shrub legumes. Both are drought-tolerant, well adapted to low-fertility, acid soils, and especially suited for low-input smallholder production systems in the sub-humid and humid tropics. They can be used e.g. as dry season forage supplementation, live soil cover or mulch, erosion barrier hedges, and shade-providing shrubs in young coffee and cocoa plantations. Germplasm collections were assembled from the wild-legume flora in Brazil (C. argentea) and Southeast Asia (F. macrophylla), but research and development are so far based on only a few accessions. Knowledge about the extent of genetic diversity within these collections is very limited. In addition, the potential utilization of F. macrophylla is so far limited by poor forage quality and acceptability of the few evaluated accessions. The objective of the present study, conducted in a research cooperation with the International Centre for Tropical Agriculture (CIAT), Cali, Colombia, was to assess the diversity in the germplasm collections of C. argentea (38 accessions) and F. macrophylla (69 accessions) in terms of morphological and phenological traits, agronomic and forage quality traits, and molecular markers, and to identify superior genotyes. Based on these different characterization approaches, the objective was furthermore to establish core collections for F. macrophylla, and to compare and validate the different strategies, giving particular consideration to their practical implications (time and cost efficiency) for the application to small collections of perennial wild tropical legumes. Cratylia argentea High diversity in terms of phenological and agronomic as well as forage quality traits was detected in the collection, with scope for plant improvement in terms of higher dry season DM production. Accessions CIAT 18674 and 22406 were identified as promising for further evaluation since they were similar to the commercial cultivar ?Veraniega? in terms of forage quality, and superior in terms of DM production, particularly in the dry season. Molecular marker analysis with RAPDs showed that the genetic diversity in the collection was relatively low and fairly homogeneously distributed. Accessions CIAT 22373, 22378, 22380, 22381 and 22411 were identified as possible duplicates. Flemingia macrophylla High diversity in terms of morphological and agronomic as well as forage quality traits was detected among the 69 accessions. The identification of four morphotypes in the collection probably has taxonomic implications. Scope for plant improvement was identified with respect to forage quality ? one of the species? main limitations. Accessions CIAT 18437, 21083 and 21090 had similar DM production and higher digestibility than the control accession, and were virtually free of extractable condensed tannins. Problems with low palatability and low seed production of these promising accessions need to be further studied. Genetic diversity in F. macrophylla was higher than in C. argentea, and corresponded closely to the four morphotypes revealed by conventional characterization. Various duplicate accessions were identified, and evidence was provided that the non-Asian accessions are not native to their collection site regions, but rather introduced from Southeast Asia. The results have direct applications for plant improvement of these promising multipurpose legumes. The superior genotypes selected in this study will be used in work with farmers in CIAT-research sites in Central America and distributed to partners. It must be recognized, however, that the diversity assessed is influenced by the climatic and edaphic conditions at the site where the studies were conducted. Therefore, multilocational trials should be considered with a selected subset (including the promising accessions) of C. argentea and F. macrophylla i) to assess the extent of genotype x environment interaction, and ii) to identify genotypes with consistently high performance in a range of distinct environments. Research on the reproduction system of both species is urgently required to determine the potential extent and impact of outcrossing. Beyond the immediate application of these species for farmer utilization, the results of the use and comparison of different approaches to assess diversity and to establish core collections can help to improve germplasm management and characterization of wild tropical legume species in general. Random sampling has been identified as a valuable and resource-efficient strategy for the creation of core collections when no additional information about accessions is available, and in the absence of adequate funds. The validation of the findings of this study with a broader range of perennial tropical wild legumes is necessary to assess their applicability to other species.