Testing the relative sensitivity of 102 ecological variables as indicators of woodland condition in the New Forest, UK

Paul M. Evans, Adrian C. Newton, E. Cantarello, Neil Sanderson, Davey L. Jones, N. Barsoum, Joan E. Cottrell, Stuart W. A'Hara, Lauren Fuller

Research output: Contribution to journalArticle

Abstract

Forests globally are facing an increasing number of threats from modified disturbance regimes, novel stressors and changing environmental conditions. This has ultimately resulted in declines in the ecological condition of many forest and woodland ecosystems, leading to widespread tree mortality and stand dieback. Effective indicators of overall woodland ecological condition are therefore needed for environmental monitoring and to support management responses. To test the effectiveness of different variables that could potentially be used as indicators of woodland condition, 102 variables that describe woodland structure, composition, functioning, edaphic conditions and disturbance regimes were assessed along 12 replicate gradients of beech stand dieback. Results indicated that 35 variables differed significantly between at least two stages of the dieback gradient, indicating their sensitivity to stand dieback. Seven of these indicators related to woodland species composition, two to functional processes, 20 to structural features, four to edaphic conditions, and two to disturbance regimes. These results demonstrate that effective indicators can potentially be identified for each of the ecological categories. Effective composition indicators included species richness of ectomycorrhizal fungi, ground flora and epiphytic lichens; functional indicators were soil respiration rate and net nitrification rate; edaphic conditions included soil Na:Ca ratio, exchangeable sodium, total carbon, Ca:Al ratio; structural indicators included canopy openness, litter cover, sward height, and volume of deadwood, and for disturbance the indicator was Equus dung density. Other measures, such as shrub cover and species richness of carabid beetles and spiders, were not found to vary significantly along the dieback gradients, and were therefore not identified as effective indicators. These results demonstrate the value of gradient analysis for evaluating indicators of woodland condition, but also highlight the need for multi-site studies to identify indicators with widescale applicability.

Original languageEnglish
Article number105575
JournalEcological Indicators
Volume107
DOIs
Publication statusPublished - 1 Dec 2019

Fingerprint Dive into the research topics of 'Testing the relative sensitivity of 102 ecological variables as indicators of woodland condition in the New Forest, UK'. Together they form a unique fingerprint.

Cite this