E. Cocheret de la Morinière

Diet shifts of Caribbean grunts (Haemulidae) and snappers (Lutjanidae) and the relation with nursery-to-coral reef migrations

Abstract The spatial size distribution of grunts and snappers have previously indicated the separation of juveniles in nursery habitats from the adults on the coral reef. This implies life cycle migrations from nursery habitats (such as seagrass beds and mangroves) to the coral reef. If diet shifts are related to such migrations, then the diets of these fish must change before or around the fish size at which such migrations take place. A wide size range of juveniles of two grunt species ( Haemulon sciurus and Haemulon flavolineatum ) and of two snapper species ( Lutjanus apodus and Ocyurus chrysurus ) were caught in seagrass beds and mangroves, and their gut contents identified and quantified. Regression analysis between fish size and dietary importance of small crustaceans showed a negative relationship in all four species. Positive relations were found for H. sciurus , L. apodus and O. chrysurus between fish length and the dietary importance of decapods, and for L. apodus and O. chrysurus between fish length and prey fish importance. Critical changes in the fish diets with fish size were examined by application of a Canonical Correspondence Analysis (CCA). The CCA yielded three clusters of size-classes of fishes with similar diets, and application of a Mantel test showed that each of these clusters had significantly different diets, and that each cluster diet was significantly specialised. The size at which a fish species ‘switched’ from one cluster to another was compared with size-at-maturity data and with the typical size at which these species migrate from the nursery habitats to the coral reef. H. sciurus and H. flavolineatum may be prompted to migrate from the nursery habitats to coral reef habitats because of dietary changes, or because of the development of the gonads. For L. apodus and O. chrysurus , a dietary changeover forms a more likely explanation for nursery-to-reef migrations than does sexual maturation because these species reach maturity at sizes much larger than the maximum size of individuals found in nursery habitats. Although other factors may theoretically initiate or promote the migration patterns, the results of this study indicate that ontogenetic dietary changes may crucially influence the nursery-to-coral reef migrations of these reef fish species.

Leaf litter removal by the snail Terebralia palustris (Linnaeus) and sesarmid crabs in an East African mangrove forest (Gazi Bay, Kenya)

Quantitative data on leaf litter removal activity of macrozoobenthic organisms in the mangrove forests of East Africa are virtually non-existent. In the present study, litter removal activity was determined in two contrasting types of mangrove stands in Gazi Bay (Kenya). In the relatively elevated Ceriops tagal vegetation, which is only flooded during spring tides, the detritivorous snail Terebralia palustris (Linnaeus) was the major macrobenthic organism responsible for litter removal. Analysis of the delta(13)C value of the foot tissue of the snail indicated a segregation in the food consumed by individuals below and above a size of 50 mm, in agreement with the observation that only larger individuals were feeding on the leaf litter. In the low lying Rhizophora mucronata stand, which is flooded by each high tide, the crab Sesarma guttatum (H. Milne Edwards) was responsible for most of the litter removal (consumption and burial). The availability of water in the C. tagal stand, caused by tidal inundation or by rainfall, was a determining factor in the amount of litter being removed. When the stand remained dry around neap tides, the median litter removal, as a percentage of the litter fail, was only 0.8%. Under wet conditions around spring tide this percentage was much higher: 41.6% by night and 25.2% by day, respectively. These figures reflect the behaviour of T. palustris, which is inactive under dry conditions in order to avoid desiccation. Median litter removal in the R. mucronata vegetation, expressed as a percentage of the litter fall, was 40.3% by day and 21.7% by night. No relation was observed between lunar cycle and activity of the litter processing crabs. Taking into consideration differences in inundation frequency and duration, and in litter removal activity by benthic animals as related to tidal height and day/night cycles, we estimate that in this East African mangrove, on average. 11.2% and 18.6% of the fallen litter is processed by macrobenthic animals in the C. tagal and in the R. mucronata vegetation, respectively. Our results indicate that removal of fallen leaf litter in mangrove forests is not effected by benthic communities dominated by crabs only, but that activities of litter feeding snails may also be significant. [KEYWORDS: mangroves; leaf litter; litter processing; sesarmid crabs; snails; Terebralia palustris Australia; decapoda; queensland; crustacea; grapsidae; ecology]

Tidal exchange of macrolitter between a mangrove forest and adjacent seagrass beds (Gazi Bay, Kenya)

In the present study the tidal transport of macrolitter between the mangrove forest in Gazi bay (Kenya) and the adjacent seagrass meadows in the bay was investigated, by deploying large standing nets, which extended over the entire height of the water column, in the transition zone between both ecosystems. In addition, the presence of macrolitter on the floor ofRhizophora mucronata andCeriops tagal stands was studied. The macromaterial (>2 mm) that was collected with the nets consisted of mangrove material (26%, mostly leaf material), seagrass leaves (60%) and macroalgae (14%). Transport was bidirectional, indicating shuttle movements of the litter, driven by the opposite flow direction of flood and ebb tides. Litter from the mangrove species consisted mainly of leaves from species occurring at the outer zone of the forest,i.e., Rhizophora mucronata andSonneratia alba. This finding suggests that the complex spatial structure of the forest hampers outflow of macrolitter from the more inner parts. Consequently, this material remains trapped within the forest. The dominant presence of seagrass litter in the macromaterial transported with the tidal water, and the conspicuous and persistent presence of seagrass litter in the low lying, peripheralR. mucronata plots (but not in the more elevatedC. tagal plots) suggest that the mangrove forest of Gazi bay is the recipient of carbon and nutrients from the seagrass system. It is hypothesized that the element cycling of the inner parts of the mangrove forest proceeds as that of a rather closed system, whereas element cycling in the outer parts has conspicuous reciprocal connections with the adjacent seagrass meadows.