Because of the large number of samples, only one bay could be sampled each day: Glades Bay (3/3/2003), Tambourine Bay (4/3/2003) and Boronia Park (5/3/2003). In the complete design, Boronia Park was sampled twice (28/2/2003 and 5/3/2003) to measure change in benthos and sediment over the time-course of the experiment, but there was little change over this period (Tolhurst and Chapman, in review). Therefore, we conclude that differences among bays reflect spatial and not temporal variation and the data from the second time of sampling Boronia Park are used here to test for correlation. These habitats are patchily distributed along the edge of these narrow mangrove forests. Three interspersed sites, each approximately 4 m× 4 m and between 20 and 40 m apart, were sampled
in each bay. Replicate samples, approximately 1 m apart,
were taken in each site