Fewer individuals recruited at all sites by the June census and there was no significant
effect of site. While there was a weak positive correlation between recruitment in May
and June, it explained only 10% of the variation between months (Fig. 5). We suspect
the difference between the May and June data reflects different processes at work as the
season progresses. This may include not only changes in the distribution of larvae in the
cove but also changes in the pattern of post-settlement mortality. For example, we might
expect growth of earlier recruits to allow escape from removal by grazing molluscs (e.g.
Miller and Carefoot, 1989). But as grazing increases with spring temperatures, later
recruits may be unable to attain refuge in size