Establishment Method of Ecological Indicator for Seedling Identification in Arable Land

Eleven characteristics of broadleaves and five characteristics each of grasses and sedges, respectively, were selected as the identification key. We constructed each of key trees using the characteristics and developed a computer program for weed seedlings identification. In order to enhance the identifying an accuracy and feasibility in an identification program based on morphological characteristics, ecological indicators of plants species were also established. Weed occurrence was separately investigated in paddy fields, arable paddy, bank in the paddy field, dry direct seeding, water direct seeding, and transplanting of plane and slope and upland fields, orchards, vegetables, and grains by season patterns for ecological indicator. The number of weed species and of individuals per was higher in arable paddy than in cultivated paddy fields, and was also higher in sloped fields than in plane paddy ones. A total 43 weed species occurred in paddy fields. However, after we divided places and analyzed cropping pattern by season, we found that 5.2, 15.4 and 8.8 weed species occurred on average during given periods (January through April, May through August and after September individually). Generally, the number of weed species and of individuals per was higher in upland fields than in cultivated paddy ones. A total 85 weed species were found in upland fields. However after we divided croplands, we found that 17.2, 14.2, and 14.4 weed species occurred in orchard, grain and vegetable fields respectively. A key tree for identification was simplified by using weed species in specific places by seasons. Therefore, if we develop an identification program based on morphological characteristics which is appended with ecological characteristics, the program will heighten an accuracy rate of seedling identification