Appendix A. Size-dependent mortality of Watersipora.
We quantified the size-dependence of overgrowth mortality for Watersipora. We used biweekly photographs to follow a cohort of Watersipora individuals that settled during August 2008. Watersipora colonies are visible to the naked eye at all sizes, allowing us observe overgrowth mortality at the smallest sizes. Photographs were taken every two weeks, and survival probability for each two week interval was modeled as a function of size using a logit regression with binomial error. Mortality was strongly size-dependent, with colonies greater than ~5 mm2 largely resistant to overgrowth mortality (n = 192, Wald test p < 10-3, Fig. A1).
Quantification of Ascidia mortality at small sizes is more difficult due to the small size and transparency of newly settled individuals. Consistent with the results of our competitor removals, overgrowth mortality of individuals large enough to be visible in photographs occurred very rarely.
|FIG. A1. Size-dependent Watersipora mortality. Mortality over two-week intervals was modeled as a function of log colony size, using a generalized linear model with logit link and binomial error.|