Integrating permanent plot and palaeoecological data to determine subalpine post‐fire succession, recovery and convergence over 128 years

• Published in: Journal of vegetation science 2020-09, Vol.31 (5), p.755-767
• Main Authors: Burge, Olivia R., Bellingham, Peter J., Arnst, Elise A., Bonner, Karen I., Burrows, Larry E., Richardson, Sarah J., Wiser, Susan K., Wood, Jamie R., Wilmshurst, Janet M., Hallett, Lauren
• Format: Article
• Language: English
• Subjects: chart quadrats; Convergence; Cores; Ecological studies; Ecological succession; Ecosystems; ecotone; Environmental changes; fire; functional traits; global change; Historic buildings & sites; Historic sites; Modernization; New Zealand; palaeoecology; permanent plot; Pollen; Productivity; Reduction; Taxonomy
• ISSN: 1100-9233; 1654-1103
• DOI: 10.1111/jvs.12887

Aims Post‐fire succession in low‐productivity ecosystems is rarely directly measured. We test whether taxonomic, functional, or baseline convergence has occurred 128 years after a major fire in 1890 in a low‐productivity, fire‐susceptible subalpine ecosystem. The historical plots represent one of the longest‐running quantitative ecological studies in the world. Location Arthur's Pass, South Island, New Zealand. Methods We remeasured historical plots at sites established 42 years after the fire, and modernised previous chart quadrat data to percentage frequency. We tested whether taxonomic and functional dissimilarity within the site, and within environmentally similar transect groups, reduced over time. We used pollen data preserved in soil cores taken on the plots to test whether the site had returned to the reference pre‐fire baseline. Results There is little evidence of taxonomic convergence across the site and dissimilarity remains high. Functional dissimilarity showed little reduction or evidence of convergence. Within maximally similar groups of transects, taxonomic and functional dissimilarity declined over time. Tree frequency increased significantly year‐on‐year. There is a lack of convergence with the pre‐fire baseline, accompanied by a reduction in the proportion of tree pollen represented on most transects. Conclusions The taxonomic rate of successional change has yet to slow, and a steady state has yet to be reached, suggesting it is too early to consider the site has attained maximal convergence. Palaeoecological data have allowed quantification of the effect of a novel fire regime to which the ecosystems are poorly adapted: we found compositional, structural, and dominant‐species differences between pre‐ and post‐fire vegetation. Combining permanent transects and palaeoecological data allows for a synthesis of data at multiple time scales, permitting investigations of long‐duration successions. Following a fire in 1890 in Arthur's Pass (New Zealand), vegetation recovery and change were monitored from 1932 to 2018. Vegetation change was predicted to slow, and composition was expected to become more similar among transects, with increased time since fire. However, we found an increase in the rate of vegetation change in the most recent measure, and little evidence of increasing similarity among transects.