Picea maximowiczii (Pinaceae) is listed as a threatened species (Vulnerable) of Japan since the estimated number of mature trees is lower than 2,000. Because there was no aggressive conservation measures so far, CHUBU Regional Forest Office, which manages most of the natural habitats, established the genetic resource forest of P. maximowiczii in Nishidake National Forest at Mts. Yatsugatake in 2010. The grafted trees were propagated from the entire distribution area and planted in the genetic resource forest, which was segmented to the North area from Mts. Yatsugatake (9 localities, 69 clones, 372 trees) and the South area from Mts. Southern Alps (9 localities, 65 clones, 372 trees) by the original locality. In 2017 after 7 years of planting, the accumulated survival rate (80%) and the average tree height (228 cm) showed that it grew almost smoothly. However, in the South area, the tree number was predicted to decrease to 138.7 from 372 in 2050 based on the average mortality rate of 1.9 % y -1 in 2015-2017. In addition to 13 disappeared clones, 5.8 clones will be expected to decrease by 2050. Since there is a negative correlation between the diameter of the picking trees and the graft survival rate, the more higher productivity of grafting seedlings will be expected by using scions collected from small and young trees growing in the genetic resource forests than by using scions from large and old trees growing in the natural forests. Therefore, it is considered that the number of clones in the genetic resource forests will be kept by an additional planting of re-propagated grafted seedlings.
Ex situ seed conservation is one of the most effective ways to conserve wild forest tree species. We present examples of the millennium seed bank in the United Kingdom and Japanese seed bank projects and examine problems with the seed conservation of wild species. To conserve the seeds of wild species, effective seed collection and appropriate preservation processing adapted to seed traits are important. New methods involving prediction models and simple seed desiccation should be effective. The quality of seed banks should be improved by obtaining information on seed traits, such as desiccation tolerance, germination behavior, and seed dormancy. To increase this body of information, it is necessary to share information through collaboration between domestic and foreign seed banks.