188 • Technologies To Maintain Biological Diversity 



interest of one person or a few individuals. 

 When these efforts cease, a valuable collection 

 can rapidly deteriorate. Information on the fo- 

 cus and extent of various collections can aid 

 coordination of duplication and minimize the 

 potential for loss (34,112). 



Assessing Diversity in a Collection 



The diversity of a collection can be assessed 

 by collecting morphologic, biochemical, or 

 phytochemical information, frequently called 

 characterization data. 



Most characterization data can help distin- 

 guish one accession from another but not as- 

 sess potentially useful traits. This is particularly 

 true for assays of proteins or DNA, which give 

 little indication of such traits as crop yield, dis- 

 ease, or stress resistance. 



Morphological Assessment. — Assessing the 

 morphology of an accession is the first step to 

 developing accurate characterization. Morpho- 

 logical information for wild species is impor- 

 tant for taxonomic identification and forms the 

 essential baseline from which all other data are 

 related (68). The information on agricultural 

 crops can be used to distinguish individual ac- 

 cessions, as well as identify them taxonomically 

 (114). However, data are on gross appearance 

 and do not reflect the full genetic composition 

 of an accession. 



Care must be taken to ensure reliable results 

 when plants are grown-out for morphological 

 assessment (10). Spacing of plants must be ade- 

 quate to ensure that results do not reflect over- 

 crowding, for example. Samples thought to be 

 duplicates are frequently grown side by side 

 for comparison (13). Biological factors, such as 

 the potential for cross-pollination among ac- 

 cessions, must be taken into consideration. 



The major constraints to assessing morphol- 

 ogy are adequate space, funds, and trained per- 

 sonnel. Though not technically difficult, such 

 assessments require attention to possible envi- 

 ronmental effects and may take a significant 

 amount of time to perform, analyze, and record. 



Biochemical Analysis.— Analysis of proteins 

 or DNA using electrophoretic techniques is 

 another way to assess diversity (94). Isoenzymes, 

 the protein products of individual genes, can 

 change in number or chemical structure when 

 the genes for them are altered, and these 

 changes can be detected by an electrophoretic 

 assay. Examination of DNA can allow compar- 

 ison of the entire genetic composition of ac- 

 cessions. 



Isoenzyme analysis on either starch or poly- 

 acrylamide gels has probably been the most 

 popular technique for assessing genetic diver- 

 sity. Surveys of isoenzyme polymorphism have 

 been performed for maize, wheat, tomato, pea, 

 and barley (94,114). In addition, surveys have 

 been done on hundreds of other cultivars and 

 wild species (94,104,105,114). 



A potential application of this technology is 

 the development of isoenzyme "fingerprints" 

 to permit reliable identification of specific plant 

 varieties to certify breeding materials, to iso- 

 late genetically similar cultivars, or to monitor 

 otherwise undetected genetic changes in acces- 

 sions. Electrophoretic analysis has been used 

 to detect duplication in some offsite collections, 

 such as the CIP collection of potato germplasm 

 (50) and is increasing in application at NPGS 

 facilities (19). However, since the data are gen- 

 erally restricted to a few biochemical charac- 

 teristics and do not reflect performance data 

 or the full genetic composition of an accession, 

 such analysis has been considered risky (39). 



Two-dimensional electrophoresis is used to 

 separate complex protein mixtures such as 

 those found in seed or leaf extracts so that sev- 

 eral hundred proteins can be distinguished in 

 a single gel (9,17,114). The results can be diffi- 

 cult to reproduce, however. The technique re- 

 quires specialized equipment and may be too 

 lengthy for routine use because only one sam- 

 ple can be evaluated at a time. Managers of off- 

 site collections are unlikely to have the time, 

 expertise, or resources to use this technique rou- 

 tinely. 



Restriction fragment length polymorphisms 

 (RFLPs) have been used to directly examine 



