MEPS 375:25-39 (2009)  -  DOI: https://doi.org/10.3354/meps07758

ABSTRACT: The rocky subtidal ecosystem of the Atlantic coast of Nova Scotia is characterised by 2 community states, kelp beds and urchin barrens that alternate on a decadal time scale. While the shift from barrens to a kelp bed requires a large perturbation, namely a disease outbreak causing urchin mass mortality, the reverse shift is more gradual and mediated by the formation and propagation of destructive grazing aggregations (or fronts) of sea urchins Strongylocentrotus droebachiensis. We have developed a spatial model for the re-establishment of kelp beds and the formation of a feeding front of urchins in order to describe transitions (phase shifts) between kelp bed and barrens states following an urchin mass mortality event. Our model includes size-specific movement, growth, and survival of urchins, as well as growth of kelp and grazing by urchins. The position of the kelp bed–barrens interface (or grazing front) predicted by our model is in general agreement with field observations at 2 sites and for up to 7 yr after a mass mortality event. Elasticity analysis showed that daily foraging movements by individual urchins have the greatest effect on the position of the feeding front and the density of the urchins there. Survival and recruitment rates of urchins also have large effects on urchin density at the front. The density of urchin populations in deep-water thermal refuges (from a disease agent) and the distance between these source populations and remnant or emergent kelp beds have large effects on the time required for the re-formation of a feeding front after an event of urchin mass mortality. These findings suggest that site-specific characteristics should be taken into account in the development of sustainable harvesting strategies and habitat-based management of the urchin fishery.

KEY WORDS: Alternative states · Feeding front · Sea urchins · Strongylocentrotus droebachiensis · Kelp beds · Phase shifts · Population dynamics · Mathematical modelling