Ch. 7— Maintaining Plant Diversity Offsite • 175 



Photo credit: Centra Internacional de ta Papa (CIP) 



Collecting wild potato species in South America, 

 a center of potato diversity. 



other propagation materials (cuttings, tubers, 

 etc.) are only general and may be altered by spe- 

 cific conditions in the field. For example, the 

 Central America and Mexico Coniferous Re- 

 sources Cooperative (CAMCORE) has estab- 

 lished guidelines for environmental and geo- 

 graphic factors to consider, depending on the 

 number of trees to collect from, and the amount 

 of seed to be collected (25,26). But if collectors 

 encounter small populations of tree species, col- 

 lection is made from any tree vifith seed (26). 

 Thus, the guidelines for a collecting expedition 

 depend heavily on the expertise and judgment 

 of the individual collector. 



Scientists and experienced collectors also 

 provide helpful information on collecting sites. 

 Several of the large so-called genetic stock col- 

 lections of crop germplasm in the United States, 

 such as the one for tomato at the University 

 of California-Davis, are overseen by a few in- 

 dividuals with special interests in that crop. The 

 knowledge these people have about origins and 

 distribution of a crop is frequently the result 

 of extensive observations and field collecting 

 experiences. 



Computer-based modeling is another poten- 

 tially useful tool for predicting the location of 

 sites appropriate for collecting. Data from a few 

 key locations may provide information on the 

 distribution of a particular crop trait, such as 

 drought tolerance. A map can then be con- 

 structed by computer-based extrapolation of 

 neighboring regions. Areas likely to contain 

 plants with similar characteristics could be 

 selected (2). However, this technology has its 

 limitations. This kind of analysis requires pre- 

 cise data on latitude, longitude, and elevation 

 for collected plants, for example— information 

 that is not currently available in most crop data- 

 bases (2). And because overall geographic in- 

 formation comes from satellite imagery, it can 

 be prohibitively expensive (2). Political or other 

 (e.g., geographic) restrictions on collecting in 

 some areas may also make acquisition of plant 

 samples difficult. Finally, data for the initial 

 profile are obtained from sites chosen by sta- 

 tistical analysis, and plant distribution may not 

 parallel these mathematically chosen sites. Re- 

 finements in existing databases, collection in- 

 formation, and artificial intelligence systems 

 may someday allow such models to assist in 

 collecting. However, it seems the importance 

 of using existing data for such tasks has been 

 overlooked (46). 



line 



After samples have been collected, compli- 

 cations may arise in transporting them. Move- 

 ment of plants from one region to another al- 

 ways carries some risk that pests (nematodes, 

 snails, insects, etc.) or pathogens (viruses, bac- 

 teria, or fungi) will also be transported (14,59, 



