ECOLOGY AND TAXONOMY OF Holimeda 301 



(c) Inter-island channels. In the passes between the islets the flow of 

 water across the algal ridge and back-reef continues into the lagoon 

 uninterrupted by land. Most of the passes are shallow, and the current 

 usually moves strongly in only one direction. Odum and Odum (1955), 

 working in one of these channels on the windward side of Enewetak, 

 obtained a maximum current of 1-44 m sec"^ during high- water neap 

 tide, and suggested that currents were probably twice this velocity 

 during incoming springs. 



The lushest growth of opuntia I encountered at Enewetak was 

 towards the lagoon end of one of these channels, the pass between 

 Lojwa and Alembel on the windward side of the atoll (Fig. 97). In its 

 shallow ( — 2 m to — 5 m), well-lighted, fast-moving waters almost 

 every crevice, especially those exposed to full sunlight, seemed filled 

 with this species. Segment shape was particularly variable, and some- 

 times the segments were tiny, but the characteristics were unmistakably 

 "opuntia" . Such thalli, presented separately to the taxonomist, could 

 well provoke at least a form epithet or two for the literature. Yet such 

 names, under the circumstances, would have no ecological, and probably 

 no real taxonomic significance. Flow-respirometry of one of these 

 Halimeda channels might be expected to reveal productivities as high 

 as any found for the Halimeda-free reef flat. 



The Halimeda species lacunalis f. lata, taenicola, fragilis and micro- 

 nesica also lived in these "streams", but in more sheltered locations 

 than opuntia. The relatively abundant lacunalis f. lata grew on the sea- 

 ward and lee side of corals, often under overhangs provided by the 

 undercut coral bases, or under promontories provided by the surface. 

 Its segments, too, were sometimes dwarfed, and many of the apical 

 segments grazed. In this inter -island channel it was the second most 

 abundant Halimeda species; only opuntia was commoner. Other 

 Udoteaceae were present too, but never prominent in any of the regions 

 visited. This inter-island channel, lagoonward of midpoint, was one of 

 the two sites with the greatest Halimeda species diversity encountered 

 at Enewetak. 



(d) Atom bomb craters. There are six nuclear craters underwater or in 

 the reef flats of Enewetak which provide sites for Halimeda colonists, 

 and I examined two of them, CACTUS and LACROSSE, in December 

 1975. Both are at the north end of Runit (Fig. 97). The larger, 

 LACROSSE, formed in May 1956, is in the back-reef, whereas 

 CACTUS, formed almost exactly two years later, is more an inpocket- 

 ing of an inter-island channel (Fig. 85). Both craters are open to the sea, 

 but CACTUS had most of the crater rim submerged, and so was 



