AbstractA chronosequence method was used to study plant community changes over successional time in a replicated design over an interval of 56 years. Artificial earthworks (revetments) constructed in 1939 and 1968 within a military manufacturing facility have been maintained under a constant regime of regular cutting and biomass removal. Predictions of species and community level change based on models of Odum, Grime and Peet are tested. Aspect and time are shown to be the major explanatory factors determining vegetational differences between samples. Diversity was seen to be significantly lower in older communities in conflict with Odum's model of community development. Partitioning of community structure by relative abundance suggests that the direction of development is contingent on physical conditions. Change in species types from ruderal to stress tolerators is consistent with Grime's successional model but is supported only weakly and by a minority of the species present. Feet's model of competitive sorting is strongly supported at the l-3m scale
but refuted at larger scales. It is apparent that older communities show a tendency towards divergence in structure. The influence of spatial factors on ecological research methods is examined and discussed and it is suggested that development of predictive community models will require a recognition of the multiple levels of community structure and the multiple scales of interactions between their components.
|Date of Award
- plant community changes
- Landscape archaeology