Populations of the tropical limpet Cellana grata Gould were studied on exposed shores around Cape d'Aguilar, Hong Kong. C. grata is a high shore species, the zonation pattern of which varies seasonally, shifting downwards in association with increasing summer temperatures. C. grata is a non-homing species and was active (foraging) whilst awash by the tide, moving up and down the shore with the flood and ebb tides, respectively. Limpets moved up to 1 m in the vertical plane over a tidal cycle. When not foraging limpets took refuge in habitats that reduced the effects of high temperature and desiccation stress. Temperature measurements of individuals and their physical environment showed that the quality of these refuges varied both spatially and temporally. Horizontal rock surfaces were hotter than vertical surfaces; limpet body temperatures and habitat temperatures in refuges (crevices, pool interfaces and east-facing vertical surfaces) were consistantly cooler than adjacent habitats by 3 to 8°C and limpet body temperatures were usually 2°C warmer than the rock they rested on. Experimental manipulations with specimens restrained 0.5 m above their normal resting height or on horizontal rock surfaces and prevented from returning to refuges caused desiccation, osmotic stress and, in many cases, death as a consequence of prolonged emersion In these sub-optimal habitats. Tndividuals on horizontal surfaces lost water more rapidly than those on vertical surfaces and showed higher mortality. C. grata does not adopt many of the behavioural adaptions to reduce heat stress shown by other gastropod species. It does, however, exhibit 'mushrooming' behaviour, lifting its shell from the rock surface, although the exact benefits of this behaviour are unclear. Summer physical extremes may strongly select for limpet activity rhythms entrained by wave splash and the selection of refuges from high temperatures and desiccation when inactive. Variation in the quality of these refuges can lead to selective mortality during very hot periods.

ABSTRACT-The epilithic biofilm of troplcal shores is dominated by cyanobacteria with only a sparse cover of diatoms. Geographical comparison of biofilms is difficult, however, due to variation in enumeration techniques. This paper describes techniques to quantify the biofilm in an attempt to standardize comparative studies. Spec1c.s rlchness and relative abundance were best enumerated using scanning electron microscope (SEM) images of air d r~e d rock chips, as compared to light, epifluorescence and confocal laser mlcroscopy. SEM preparations provided cleal-lmages and allowed identification of most specles Alr d r~e d and Cryo-stage preparations prov~ded the least damaged images of species whereas the harsh dehydration process in crit~cal-point dried specimens damaged morphology and removed loosely dttached species resulting in unclerest~mation of specles richness and abundance A combination of SEh4 techniques (e.g. Cryo-stage, air d r~e d and critical-po~nt dried san~ples) is reconimended for initial species identificdtion. Chlorophyll a extraction from rock chips, as an indirect estlmate of biomass, was more eff~cient using methanol or ethanol a s compared to acetone. Methanol heated for 2 min and cooled for 12 h ylelded 100% chlorophyll a extraction while cold methanol, methanol heated for 2 min and cooled for 3 and 6 h, and ethanol heated for 5 min, extracted greater than 95% chlorophyll a. No significant loss of chlorophyll a was recorded in rock samples stored in moist or dry conditions at 4°C or room temperature for 24 h or at -20°C for a month, but in samples stored at 4°C or room temperature for a week, 60 to 75% loss occurred. Chlorophyll a, once extracted. can b e stored for a week in the dark prior to measurement without significant loss. To standardize techniques for tropical, cyanobacteria-rich epilithic biofilms, chlorophyll a extraction of rock chips using cold methanol 1s recommended.

The tropical rocky shore crab Grapsus albolineatus selectively consumes rare filamentous algae over more abundant foliose algae during the wlnter in Hong Kong. Laboratory experiments have shown that growth of G. albolineatus is enhanced and mortality reduced when given a diet of filamentous algae as opposed to fohose algae. In the laboratory, G. albolineatus consumed filamentous algae (Enteromorpha clathrata, Hincksia mitchelliae and Chaetomorpha antennina) in greater amounts than any foliose algae (Dermonema frappieri, Pterocladia tenuis, Porphyra suborbiculata. Ulva fasciata, Endarachne binghamiae) in both multiple choice and no choice experiments. The most energy-rich alga was Pterocladia, while Porphyra had the highest protein content. Filamentous algae had lower overall nutrient contents than foliose algae. Consumption rates for Enteromorpha and Hincksia were, however, sufficiently higher than for Porphyra, which ensured that the net intake of nutrients per day was greater. Even though protein assimilation efficiency was higher for Porphyra than Hincksia, G. alboljneatus assimilated more protein, per day, from Hjncksia. When the confounding effects of morphology and nutrient value were separated, by offenng the crab choices of commercially available algae (Laminaria sp.) cut into different forms (foliose and filamentous), G. albolineatus showed a strong preference for the filamentous form, despite both forms having the same nutrient value. Preference for filamentous forms is likely to be constrained by the morphology of the chelae, which have delicate tips, and appear unable to tear foliose algae. Algal morphology, therefore, appears to be of prime importance and the nutrient content and digestibility of algae of secondary importance in determining the feeding preferences of G. albolineatus. The high consumption rate of filamentous algae outweighs their relative nutrient deficiencies, indicating that they are better suited to meeting the physiological needs of the crab than foliose algae.