COS 56-3 - Prediction of unknown potential natural vegetation in heavily disturbed area based on a community assembly rule

Tuesday, August 7, 2007: 2:10 PM
Willow Glen II, San Jose Marriott
Fumito Koike, Yoshiko Kobayashi and Koji Kato, Graduate School of Environment and Information Sciences, Yokohama National University, Yokohama, Japan
Human activities have completely destroyed natural vegetation in many regions in the world. Fragmented forests in urban landscapes often lack climax forest species with limited dispersal ability. Predicting potential natural vegetation (climax forest) is the important ecological issue in these areas. Studies on community assembly rules make it possible to predict climax forest based on plant traits of the local flora, and was conducted in two areas in Japan: historically heavily disturbed inland area of the main island, and fragmented suburban forest patches of Yokohama. Maximum height and shade tolerance are known to be the key traits in predicting dominance of woody species in climax forests, and were measured for local woody flora including alien species. The previously obtained linear regression equation was used to predict relative basal area in climax forests. In suburban fragmented forests, predicted dominance of species in patches with rich climax species showed significant correlation with an actual old-growth forest, suggesting reliability of the assembly rule approach. Climax vegetation for patches without evergreen oaks was predicted as forests dominated by tall deciduous species (hackberry, Celtis sinensis) associated with alien species in some cases. The predicted potential natural vegetation in the historically disturbed inland area was a mixed forest of a temperate evergreen conifer (Cryptomeria japonica) and a tall deciduous species (Zerkova serrata). No such forest can be found currently in Japan, but similar combinations of pollen assemblage have been found in sediments by geological studies. Community assembly rules can be used to predict unknown potential natural vegetation.
Copyright © . All rights reserved.
Banner photo by Flickr user greg westfall.