Effects of sea urchin grazing on seagrass (Thalassodendron ciliatum) beds of a Kenyan lagoon

We present the results of experimental and descriptive field studies on the effects of dense sea urchin aggregations on seagrass beds in the Mombasa lagoon (Kenya). The study area was dominated by the slow-growing seagrass Thalassodendron ciliatum (49.6% cover), and supported an average sea urchin density of 1.6 m(-2), mostly Tripneustes gratilla (90%). In the T ciliatum meadows, 39% of the cover was heavily grazed by the sea urchins (more than 75 010 dead shoots), 23.4% was moderately grazed (more than 50% dead shoots) and 38.5% was slightly grazed (19.8% dead shoots). We observed 5 aggregations (fronts) of T gratilla in the study area (5000 m(2)), 4 of these within the T ciliatum meadows. These aggregations were linear in structure, with mean densities of 10.4 m(-2) and left in their wake a 'trailing edge' area of defoliated seagrass rhizomes (85% dead shoots). Grazing rates were measured in 2 ways: marked permanent quadrats along the fronts, and sea urchin addition experiments. The first method produced grazing rates of 1.8 +/- 0.43 shoots m(-2) d(-1), and the second produced slightly higher values of 5 +/- 0.86 shoots m(-2) d(-1). New shoot recruitment was estimated from the marked permanent quadrats in the fronts as 0.32 shoots m(-2) d(-1), Simple models indicated that the return interval (i.e, the average frequency of the passage of sea urchin fronts through a seagrass patch) for the 'sea urchin grazing fronts' was 99 or 34 mo, depending on the method used, and that T ciliatum recovery time was 44 mo. We conclude that the sea urchin aggregations observed in the Mombasa lagoon control T ciliatum density by grazing its exposed apical tips.