Ch. 7— Maintaining Plant Diversity Offsite • 181 



Seed Storage Laboratory (NSSL), use dehumid- 

 ified rooms to reduce moisture. 



Storage 



Four factors affect seed storage: 1) moisture 

 content, 2) storage temperature, 3] storage 

 atmosphere, and 4) genetic composition of the 

 sample (4,87,88,89). Reduced moisture content 

 is considered the most crucial to maintaining 

 viability. In general, each 1-percent decrease 

 in seed moisture between the 5- and 14-percent 

 range will double the lifespan of a seed sample. 

 Reduction of storage temperature also increases 

 seed longevity. A 5° decrease in temperature 

 between 0° and 50° C doubles longevity (89). 

 Control of storage atmosphere generally does 

 not provide significant advantages over ma- 

 nipulation of moisture and temperature, par- 

 ticularly when the latter is below freezing. 

 Genetics relate to differences between individ- 

 ual accessions or between individuals in a 

 mixed sample and cannot be altered to increase 

 longevity. 



Viability Testing 



Seeds must be tested periodically for viabil- 

 ity. This information helps determine when an 

 accession needs to be grown-out to produce a 

 fresh sample of seeds. The most obvious test 

 is to germinate a portion of the seeds to esti- 

 mate the viable percentage. 



Viability testing involves placing seeds in 

 appropriate conditions (damp blotter paper, 

 agar medium, etc.) and counting the number 

 of seeds that germinate over a period of time. 

 However, if seeds are dormant, obtaining via- 

 bility estimates can be difficult. Citrus species, 

 for example, were thought to have died when 

 prepared for conventional storage but were 

 shown instead to be dormant (82). Heating, cool- 

 ing, lighting, and treatments (e.g., removal or 

 cracking of the seed coat) may be required to 

 overcome dormancy in some species. 



Typically, 200 to 400 seeds are required for 

 viability testing (43). However, a sequential ap- 

 proach reduces the number needed for testing. 

 Forty seeds can be tested to determine whether 

 the accession should be regenerated, stored, or 



whether another 40 seeds are needed (30,42). 

 But a small sample of seeds may need to under- 

 go numerous tests before an answer is reached, 

 which may take longer than testing a single 

 large sample. 



Other tests— involving dyes, physiological 

 tests, or biochemical assays — have been devel- 

 oped to determine seed viability (89). The va- 

 lidity of such tests relies on the ability to dem- 

 onstrate a correlation with actual germination 

 rates. These tests can be useful to determine 

 if dormancy or inappropriate storage condi- 

 tions are producing misleading results (30). 

 Some, such as the tetrazolium dye test for a 

 range of seeds, or X-ray contrast with heavy 

 metals in tree seeds, have been widely used 

 (30,89). Others, such as enzyme tests, provide 

 information useful for the study of seed physi- 

 ology but are more expensive and difficult to 

 perform than standard germination tests. 



Preserving the genetic variability in seed ac- 

 cessions is a major concern in offsite collec- 

 tions. Many accessions, particularly those of 

 primitive landraces and wild species, are ge- 

 netically diverse populations and display con- 

 siderable genetic variation between individuals 

 in a sample. Genetic differences in storage life- 

 span can mean that the genetics of a popula- 

 tion could be altered by decline in viability 

 (86,87,88,108). 



Although seeds generally should be regener- 

 ated when germination drops 15 percent, prac- 

 tical considerations of labor, space, and time 

 can delay this step (32,43,83,108). One recent 

 report stated that NSSL does not regenerate 

 samples until viability has dropped 40 percent 

 (113). This practice, however, may be based 

 more on lack of resources than on scientific 

 considerations. 



Regeneration 



Variations in growth requirements for spe- 

 cies and even for varieties within a species com- 

 plicate the growing-out of seed. In beans, for 

 example, different accessions may require 

 different conditions, e.g., daylength, for grow- 

 ing out. Thus, both subtropical and temperate 

 sites must be used to grow-out a range of bean 



