Ch. 7— Maintaining Plant Diversity Offsite • 191 



of some traits may require repeating tests over 

 several years and in different regions (81). 



Evaluation has been perceived as a serious 

 deficiency in the overall effort to maintain crop 

 germplasm (23,37,78,108,117). Insufficient in- 

 formation has meant accessions have been un- 

 derused. However, the situation has been im- 

 proving (81,108). International Agricultural 

 Research Centers have evaluated many of their 

 accessions for important traits, chiefly disease 

 and pest resistance, yield, and quality factors 

 (13,50,108). In the United States, the four re- 

 gional plant introduction stations (see ch. 9) 

 have included examination for several agricul- 

 turally important traits in their preliminary 

 characterizations. This is not, however, suffi- 

 cient to meet all the needs of users, and more 

 extensive efforts are necessary (108). 



Although the usefulness of a collection may 

 depend on its evaluations, questions remain as 

 to who has responsibility for this task. Collec- 

 tion managers might be able to gather morpho- 

 logical data, but they may not be able to per- 

 form the lengthy and detailed trials needed to 

 evaluate traits. Further, it has been argued that 

 such evaluations are best done under the con- 

 ditions in which they will be used because ex- 

 pression may be altered by environmental dif- 

 ferences (36). It would seem, therefore, that 

 evaluation trials for specific traits are best per- 

 formed by the breeders who require those traits. 

 Duplicating efforts could be minimized by put- 

 ting results into centralized database systems — 

 a proposed function of the GRIN system in the 

 United States. 



Traditional Breeding 



Traditional breeding typically involves iden- 

 tifying particular genes or characteristics and 

 incorporating them into existing varieties. Crop 

 development through breeding, a major con- 

 tributor to modern gains in agricultural pro- 

 duction, is a time-consuming process: It may 

 take 10 to 15 years to develop a single new va- 

 riety (box 7-C). 



Traditional breeding has provided as much 

 as 60 percent of the production increases of 

 many agricultural crops (22,24,35,77,124). Never- 



theless, the process must continue in order to 

 sustain agricultural yields — pests and diseases 

 adapt to new varieties, and the needs of growers 

 and consumers constantly change {77). 



Traditional breeding involves several basic 

 steps: 1) locate a genetically stable trait (e.g., 

 yield, pest resistance, or stress tolerance); 2) iso- 

 late plants with the most desired expression of 

 the trait; 3) breed genes into breeding lines of 

 plants similar to those that will be improved 

 to provide more usable material; and 4) cross 

 these plants with other breeding lines to pro- 

 duce plants from which improved crop vari- 

 eties can be selected (41,81). 



The third step, called developmental breed- 

 ing, is important because the desired trait may 

 be located in a wild species or variety that is 

 difficult to cross with domesticated ones. This 

 is the case, for example, with genetic resistance 

 to some 27 serious diseases of the tomato (81). 

 Wild species or landraces may have different 

 growth requirements that make crossing them 

 with other varieties difficult. Developmental 

 breeding overcomes such differences but may 

 require growing plants at multiple locations. 

 Incorporation of genes from over 500 exotic 

 sorghums, for example, required growth in two 

 locations because the exotics required shorter 

 days to flower than commercial U.S. varieties. 

 A cooperative effort, therefore, was established 

 between the Texas Agricultural Experiment Sta- 

 tion and the USDA Federal Station in Mayaguez, 

 Puerto Rico, to perform the crosses and test the 

 progeny (48,81). 



The major constraint to traditional breeding 

 is its dependence on the sexual process of 

 plants. Multiple crossings and testing of off- 

 spring may take years. Molecular biological 

 techniques to locate and map genes in plants 

 may greatly shorten the time needed for breed- 

 ing improved varieties (6). 



Bioteciinoiogicai Improvement 



Biotechnology provides greater precision and 

 speed in the manipulation of genes by avoid- 

 ing the sexual reproductive process (24,41,75). 

 Three general areas have potential impact on 

 the use of stored plant diversity: 1) somaclonal 



