Variation in recruitment and the establishment of alternative community states.

Mussel beds and rockweed stands (fucoid algae) have been shown to be ilternative states on rocky intertidal shores in New England, and here the hypothesis that variation in recruitment provides opportunity for the development of alternative community states was tested. Disturbance by ice scour opens patches for development of alternative states, and in winter 1996-1997, 60 experimental clearings of differing sizes were established on Swan's Island, Maine, USA. Half of the plots were re-cleared during the winter of 2010-2011. Recruitment data for barnacles, mussels, and fucoid algae collected from 1997 to 2012 were used to (1) test for persistence of scale-dependent thresholds, (2) estimate the magnitudes and sources of variation, (3) fit a surface of alternative states as defined by the cusp catastrophe, and (4) test if 1997 recruitment would predict 2010-2011 recruitment in re-scraped plots (i.e., a test of divergence, which is expected in systems with alternative states). For barnacles and mussels, recruitment varied enormously year to year and among sites, but showed consistent patterns over the long-term with respect to clearing size. Average recruitment prior to re-clearing was a good predictor of recruitment afterwards. In contrast, over 50% of the variance in fucoid recruitment was unexplained with weak effects among years and locations. Past fucoid recruitment was a poor predictor of subsequent recruitment. The cusp analysis indicated that fucoid recruitment defines the alternative states. Fucoid recruitment was largely unpredictable and suggests long-term, small-scale priority effects drive the development of alternative states. These observations strongly reinforce the notion that long-term and well- replicated experiments are necessary to develop robust tests of ecological theory.