Evolution of foredune texture following dynamic restoration, Doughboy Bay, Stewart Island, New Zealand

• Landscape metrics can be used to track changes in foredune morphology during dune restoration programs.
• This method provides an objective method of evaluating dynamic restoration success.
• Coastal dune restoration programs will benefit by the routine use of reference sites.

Growing concern regarding the geomorphic and associated biotic effects of dune management practises has led to an increase in the number of dune restoration projects globally. Most recent projects aim to enhance the efficiency of aeolian sediment dynamics and increase dune mobility by decreasing vegetation cover, but we lack objective measures to evaluate such projects. Here we demonstrate the use of landscape metrics to quantify the evolution of foredune texture following the removal of vegetation.A long-term program of marram grass (Ammophila arenaria) eradication in southern New Zealand (Doughboy Bay, Stewart Island) is examined. Four metrics: bare sand area, patch adjacency, complexity, and the range of proximity, are used to classify a series of foredune textures beginning with the pre-restoration state through the phases of marram removal, to the current state. Foredune texture at Doughboy Bay has evolved from a semi-stable to an active state as the consequence of restoration. Two metrics, bare sand and adjacency, appear to be particularly good measures of change following marram removal. Patterns and rates of change for these metrics are consistent with ground observations of increased ‘naturalness’ (native plant communities, sand mobility) over the same period.The set of landscape metrics derived for Doughboy Bay were compared to similar sets measured for a nearby foredune system where marram invasion has not occurred, and where conditions presumably represent equilibrium foredune texture. Since the removal of marram at Doughboy Bay and the consequent remobilization of the sand surface, the foredune texture has increased in similarity to that of the reference site, indicating a favourable shift in plant cover as a result of the restoration program. We conclude that landscape metrics can be used to track changes in foredune morphology following restoration. Second, the planning, management, and monitoring of coastal dune restoration programs will benefit by the routine use of reference sites to evaluate the appropriateness and success of restoration actions.