• Title/Summary/Keyword: Terrain Classification

Search Result 101, Processing Time 0.016 seconds

Identification of Bird Community Characteristics by Habitat Environment of Jeongmaek Using Self-organizing Map - Case Stuty Area Geumnamhonam and Honam, Hannamgeumbuk and Geumbuk, Naknam Jeongmaek, South Korea - (자기조직화지도를 활용한 정맥의 서식지 환경에 따른 조류 군집 특성 파악 - 금남호남 및 호남정맥, 한남금북 및 금북정맥, 낙남정맥을 대상으로 -)

  • Hwang, Jong-Kyeong;Kang, Te-han;Han, Seung-Woo;Cho, Hae-Jin;Nam, Hyung-Kyu;Kim, Su-Jin;Lee, Joon-Woo
    • Korean Journal of Environment and Ecology
    • /
    • v.35 no.4
    • /
    • pp.377-386
    • /
    • 2021
  • This study was conducted to provide basic data for habitat management and preservation of Jeongmaek. A total of 18 priority research areas were selected with consideration to terrain and habitat environment, and 54 fixed plots were selected for three types of habits: development, valley, and forest road and ridge. The survey was conducted in each season (May, August, and October), excluding the winter season, from 2016 to 2018. The distribution analysis of birds observed in each habitat type using a self-organizing map (SOM) classified them into a total of four groups (MRPP, A=0.12, and p <0.005). The comparative analysis of the number of species, the number of individuals, and the species diversity index for each SOM group showed that they were all the highest in group III (Kruskal-Wallis, the number species: x2 = 13.436, P <0.005; the number of individuals: x2 = 8.229, P <0.05; the species diversity index: x2 = 17.115, P <0.005). Moreover, the analysis by applying the land cover map to the random forest model to examine the index species of each group and identify the characteristics of the habitat environment showed a difference in the ratio of the habitat environment and the indicator species among the four groups. The index species analysis identified a total of 18 bird species as the indicator species in three groups except for group II. When applying the random forest model and indicator species analysis to the results of classification into four groups using the SOM, the composition of the indicator species by the group showed a correlation with the habitat characteristics of each group. Moreover, the distribution patterns and densities of observed species were clearly distinguished according to the dominant habitat for each group. The results of the analysis that applied the SOM, indicator species, and random forest model together can derive useful results for the characterization of bird habitats according to the habitat environment.