Summary data on oyster spat survival from split-plot mesocosm experiment to examine predation strengths as function of structure.

Abbreviations of variables are as follows, Treatment: Oyster=Oyster only control; MC = mud crab; BC = blue crab; BH = bonnethead shark; Habitat: on= on oyster reef; off= off reef on sandflat; # Alive: number of oyster spat alive in a habitat at the end of experimental trial; Starting #: number of oyster spat in each habitat at start of experiment. Each tank was divided in half between oyster reef and sandflat habitat, reflecting the study’s split plot design. Hence every tank has two entries in the dataset, one for on reef habitat and one for off-reef (sandflat). Because these data were analyzed at the level of binary responses of individual oysters (live or dead), the data were expanded out to the level of individual oysters (i.e., DO loop in SAS) before analyses with generalized linear mixed effects models to analyze individual oyster spat survival with a binary distribution and logit link function.