AN EFFICIENT METHOD FOR ESTIMATING SEAGRASS BIOMASS

Seagrass biomass was estimated in a 37.649 km2 area in Moreton Bay, Qu eensland. The study area was stratified by hand-digitising seven strat a that were identified by photo-interpretation of a colour aerial phot ograph. A pilot study was undertaken to calculate how many samples sho uld be taken at each site and how best to allocate sampling effort to the seven strata. Seagrass samples taken with a modified 'orange-peel' grab and by coring were the same (P > 0.5). As the grab is easier to use, quicker, can be operated by one person from a small dinghy, and d oes not call for diving or wading, we used it for the main study. Four seagrasses were found in the study area. The dominant seagrass was Zo stera capricorni Aschchers., with smaller quantities of Halophila oval is (R.Br.) Hook.f., Halophila spinulosa (R.Br.) Aschers. and Halodule uninervis (Forssk.) Aschers. The total seagrass biomass in the study a rea was estimated at 2145 +/- 568 t (95% confidence interval). Stratif ication improved the precision of the simple random sampling estimate by 68%. The use of a Geographic Information System (GIS) and Computer Aided Drafting (CAD) for sample design, a Global Positioning System (G PS) for locating sample sites in the field and the grab for taking sam ples substantially enhanced research productivity and accuracy. The ex perimental design is statistically robust and provides large samples f or cost-effective and reliable estimates of seagrass biomass.