FROM CANALIZATION TO CONTINGENCY - HISTORICAL EFFECTS IN A SUCCESSIONAL ROCKY INTERTIDAL COMMUNITY

Many landscapes are characterized by a mosaic of patches in various st ages of succession. Whether successional paths dampen, track, or magni fy extrinsic variation in initial conditions influences how much histo rical and site-specific detail is required to explain variation in pat ch composition. I investigated the patterns and importance of historic al effects in a successional marine rocky intertidal community on the central coast of Oregon, USA. Patches in the mid-intertidal mussel bed (M. californianus) were manually cleared in a way that mimicked natur al disturbances. In four separate blocks (large patches similar to 9 m (2)), three sets of plots were initiated with their starting dates sta ggered by one year. Within each set of plots, I manipulated the presen ce/absence of two groups of early successional sessile species under e ach of three predator densities. This design allowed me to address the following general questions: (1) What are the separate and interactiv e effects of successional age, yearly variation, and initial condition s on the temporal changes observed after disturbance? (2) When do inte ractions between early species act to dampen or magnify natural variat ion between years or starting dates? Succession in mid-intertidal patc hes in the mussel bed displayed complex patterns of historical effects , which varied among species and between different stages of successio n. Embedded in this potential complexity were some consistent and repe atable successional trends. Some potentially important canalizing, or ''noise-dampening'' forces in this system included: (1) physiological and/or life history trade-offs between dispersal ability and competiti ve ability, (2) strong direct biotic interactions, which buffer enviro nmental variability, and (3) compensatory (''buffering'') responses of species within an important functional group. ''Noise-amplifying'' fo rces included: (1) variable indirect effects of predators, (2) prey si ze escapes, and (3) predator saturation (or prey ''swamp'' escapes). U nderstanding the patterns and causes of consistency or contingency in succession will be critical for managing variability in landscapes tha t are increasingly dominated by anthropogenic disturbance regimes.