Wed, Aug 17, 2022: 5:00 PM-6:30 PM
ESA Exhibit Hall
Background/Question/MethodsEcosystems are most affected by climate change, and due to this, habitat destruction and range reduction of each species in the ecosystem are occurring. Disrupted species destroy the ecosystem balance of the invading area due to their strong adaptability and reproductive ability. Therefore, it is necessary to check the distribution of disturbed species and establish a prevention and management plan suitable for it. This study intends to conduct monitoring and evaluation of potential habitats of disturbed species, focusing on 6 major disturbing species (flora) in Chungcheong Buk-do, South Korea, taking into consideration the geographic and environmental characteristics of the study site. In this study, 11 related influencing factors were established centered on 6 major habitat disturbance species (flora) in Chungcheong Buk-do, Korea, and then potential habitat distribution modeling was carried out through a machine learning algorithm. As a result of the study, it was confirmed that the average accuracy of the model was about 82%, and the model had reliable performance, and that the potential habitat distribution results for each disturbed species were consistent with the ecological characteristics of the disturbed species.
Results/ConclusionsAs a result of verifying the accuracy of the potential habitat distribution map for the six disturbed species samples using the ROC Curve, the average accuracy was about 82%, confirming that the model had reliable performance. As for the classification result, it was confirmed that the distribution result of the potential habitat area was somewhat consistent with the ecological characteristics of the disturbed species through visual comparison and verification with the land cover map and satellite image data of the subject area. The results of this study can carry out a series of spatial ecological evaluations such as the level of ecological threat in the area and the vulnerability of the forest edge, centering on the potential distribution results of disturbed species, which can provide information for environmental management and planning.
Results/ConclusionsAs a result of verifying the accuracy of the potential habitat distribution map for the six disturbed species samples using the ROC Curve, the average accuracy was about 82%, confirming that the model had reliable performance. As for the classification result, it was confirmed that the distribution result of the potential habitat area was somewhat consistent with the ecological characteristics of the disturbed species through visual comparison and verification with the land cover map and satellite image data of the subject area. The results of this study can carry out a series of spatial ecological evaluations such as the level of ecological threat in the area and the vulnerability of the forest edge, centering on the potential distribution results of disturbed species, which can provide information for environmental management and planning.