A functional attributes approach to secondary succession pathways in New Zealand.
Thesis DisciplineEnvironmental Sciences
Degree GrantorUniversity of Canterbury
Degree NameMaster of Science
While for many years we have been attempting to conserve the remaining natural areas in New Zealand, interest has recently increased in the possibility of restoring degraded areas back to (predominantly) native vegetation. Knowledge of probable secondary succession pathways for such areas would assist their management. A method to describe, and thus compare, succession pathways would help determine the strongest influences on the succession process. If this method did not use plant species, but species grouped according to shared characteristics, comparison of succession pathways between places with different floras would be possible. This thesis is the first use of the functional attributes method to describe succession pathways for New Zealand vegetation. The method was first tested by a quantitative comparison of species and functional attributes data for two New Zealand sites, using ordination and correlation analyses. The results supported the hypothesis that conversion of the data from species to functional attributes will permit, by use of ordination analysis, recovery of the same rank-order of sites in a successional sequence, using field data from New Zealand vegetation. The second test was a qualitative assessment of the method, using four New Zealand sites, involving comparisons between regeneration sequences in the literature, and those generated by functional attributes information. The regeneration sequence generated from functional attributes information was found to approximately reproduce the sequence predicted or observed by researchers for the study sites. The results were very encouraging, and showed that the functional attributes method has potential for representing succession pathways in New Zealand sites. The potential of the method as a land management tool, including modelling responses to introduced species, is discussed. The next step would be the development of a computer model, allowing multiple sequences to be quickly generated with different disturbance regimes and initial vegetation.