Construction of a health evaluation index system for large-diameter trees on islands: a case study of Dajinshan Island, Shanghai

Large-diameter trees are essential for stabilizing forest community structure. Due to their unique biogeographical environment, islands harbor a significant number of large-diameter trees. In this study, we investigated 17 biotic and abiotic factors affecting 168 large-diameter trees on Dajinshan Island. Using the analytic hierarchy process, we selected 12 high- and medium-sensitivity indices and constructed a three-level health evaluation system for large-diameter trees on islands. Results revealed that: (1) On Dajinshan Island, 25 large-diameter trees were categorized as old trees and 37 were categorized as prospective old tree resources. The largest existing old trees in Shanghai include Celtis sinensis Pers., Morus alba L., Quercus acutissima Carr., and Triadica sebifera (L.) Small. Several species, including Albizia julibrissin Durazz., Euscaphis japonica (Thunb. ex Roem. et Schult.) Kanitz, Machilus thunbergii Siebold et Zucc., Morus australis Poir., Pyrus calleryana Decne., Eurya nitida Korth., and Cyclobalanopsis glauca (Thunb.) Oerst., were identified as the largest old trees not previously included in the list of old trees in Shanghai. (2) For the tree health evaluation system, biological factors ranked as the most crucial, followed by stand condition and growing environment. In the index layer, factors such as leaf disease degree, coverage of climbing liana, degree of tree disease, space compression from neighbors, and dead branch ratio carried higher weights, nearing 10%. (3) Approximately 54% of the large-diameter trees on Dajinshan Island exhibited varying health problems. Pyrus calleryana Decne., Melia azedarach L., Euonymus maackii Rupr., Broussonetia papyrifera (L.) L’Hér. ex Vent., Symplocos paniculata (Thunb.) Miq., Euscaphis japonica (Thunb. ex Roem. et Schult.) Kanitz, Dalbergia hupeana Hance, and Glochidion puberum (L.) Hutch. exhibited relatively poor health conditions.