Extruded bioplastic was prepared from cornstarch or poly(-hydroxybutyrate-co--hydroxyvalerate) (PHBV) or blends of cornstarch and PHBV. The blended formulations contained 30 or 50% starch in the presence or absence of polyethylene oxide (PEO), which enhances adherence of starch granules to PHBV. Degradation of these formulations was monitored for 1 year at four stations in coastal water southwest of Puerto Rico. Two stations were within a mangrove stand. The other two were offshore; one of these stations was on a shallow shoulder of a reef, and the other was at a location in deeper water. Microbial enumeration at the four stations revealed considerable flux in the populations over the course of the year. However, in general, the overall population densities were 1 order of magnitude less at the deeper-water station than at the other stations. Starch degraders were 10- to 50-fold more prevalent than PHBV degraders at all of the stations. Accordingly, degradation of the bioplastic, as determined by weight loss and deterioration of tensile properties, correlated with the amount of starch present (100% starch >50% starch > 30% starch > 100% PHBV). Incorporation of PEO into blends slightly retarded the rate of degradation. The rate of loss of starch from the 100% starch samples was about 2%/day, while the rate of loss of PHBV from the 100% PHBV samples was about 0.1%/day. Biphasic weight loss was observed for the starch-PHBV blends at all of the stations. A predictive mathematical model for loss of individual polymers from a 30% starch-70% PHBV formulation was developed and experimentally validated. The model showed that PHBV degradation was delayed 50 days until more than 80% of the starch was consumed and predicted that starch and PHBV in the blend had half-lives of 19 and 158 days, respectively. Consistent with the relatively low microbial populations, bioplastic degradation at the deeper-water station exhibited an initial lag period, after which degradation rates comparable to the degradation rates at the other stations were observed. Presumably, significant biodegradation occurred only after colonization of the plastic, a parameter that was dependent on the resident microbial populations. Therefore, it can be reasonably inferred that extended degradation lags would occur in open ocean water where microbes are sparse.