Dynamics of forest ecosystem properties and resilience in a temperate woodland undergoing dieback.

Abstract

Given that woodland disturbance events are expected to become more frequent and severe in the future, it is crucial to understand how fundamental underlying ecosystem properties (EPs) including biodiversity, ecosystem functions and ecosystem services will be impacted. Currently, this knowledge is severely limited, but is essential for future management of forests at both stand and landscape scales. The impact of such disturbances on EPs can be quantified through the use of ecological resilience and its associated concepts. Resilience relates to either the amount of disturbance an ecosystem can endure and still exist or the degree to which an ecosystem can resist or recover from disturbance. Resilience concepts of ecological thresholds, points of abrupt change in an EP, and resistance, persistence and recovery time are useful metrics to determine disturbance impacts. Moreover, using knowledge of how EPs are affected by disturbance, resilience assessments can enable inference of the current level of resilience that woodland has. The objectives of this thesis were therefore: 1) to determine how biodiversity, ecosystem functions and condition were affected in a forest undergoing dieback; 2) to examine the effectiveness of rapid condition assessment tools as a proxy use for inferring woodland resilience; and 3) to determine whether woodland cover influences resilience of EPs at a landscape scale. The changing condition and extent of the New Forest provided an opportunity to measure EPs across a gradient of changing condition. This was carried out through gradient analysis of dieback, based on basal area decline, at the stand scale. Using the results obtained from the gradient analysis: i) the resilience of important woodland EPs was assessed at the landscape scale using simulations of different intensities of disturbance and woodland cover, which was used as a proxy for connectivity; and ii) commonly-used forest condition assessment tools, specifically airborne lidar and the woodland Common Standards Monitoring condition assessments, were tested to determine how effective they were and whether they could be used to infer resilience at the stand scale. Overall, evidence provided in this thesis suggests that: biodiversity and functional thresholds exist as the forest degrades; current condition assessment tools are not very effective at detecting variation in woodland condition and therefore are not sufficient to infer current resilience; and woodland cover influences the resilience of important woodland EPs, at the landscape scale. All the findings are discussed in context of the New Forest, an ecologically and socio-economically important landscape.