Data were analyzed using one and two-way analysis of variance tests performed with Number Cruncher Statistical Software (NCSS) 2007 (Statistical and Power Analysis Software). Data were checked for normality and homogeneity of variances. Pooling procedures involving elimination of terms from the mean square estimates were implemented if a term was non-significant at p>0.25 [59]. Means for significant factors in the ANOVAs were compared using Tukey-Kramer comparison tests. Repeated measures ANOVA was used to assess differences in sedimentation rates over time. In contrast, the replicate level for all biochemical and physiological comparisons was n = 3 tank (with n = 3 corals of each species pooled per tank). Repeated measures ANOVAs were not applied to the analyses of the biochemical and physiological data since PAM fluorometry was performed on random corals within each treatment, and the destructive techniques required for lipid and chl a analyses were performed only once for a given coral fragment.