26 U.S. NATIONAL MUSEUM BULLETIN 258 



compress the area covered and make it comparable to that covered 

 in California. 



The numbers of known amphipodan species in various regions of 

 the world are compared in table 31. The tropical data are not equiva- 

 lent to the other data. No tropical continental shore has been explored 

 adequately, although several insular faunas have been rather well 

 studied. All known Indo-Pacific intertidal species have been tallied 

 with the presumption that numerous insular endemics are included. 

 There is no evidence in the literature that Amphipoda are more 

 diversified in the tropics than they are in higher latitudes. For in- 

 stance, the sums of all known littoral-sublittoral Amphipoda from 

 either Norway, southern California, or of the tropical Indo-Pacific 

 are approximately equal: 300 ±5%. Tropical islands are known to 

 have a poorly diverse intertidal amphipodan fauna (Schellenberg, 1938 

 and J. L. Barnard, 1965). The only other evidence in the matter is 

 the writer's examination of the Caribbean Panamic intertidal reef 

 fauna (in preparation). There, the diversity of Amphipoda is poor, 

 although several species occur in very high frequencies of individuals. 

 A coral reef may afford numerous nestling sites for organisms such 

 as Amphipoda, but the paucity of fleshy algae, debris, and detritus 

 may be a restrictive factor. The real test of the latitudinal diversity 

 hypothesis will come when a continental shelf fauna in the tropics 

 can be studied. To be comparable to those areas already studied in 

 higher latitudes the tropical shelf must be composed of relatively 

 fine-grained sediments, generally in the range of 30 to 125 microns 

 of median diameter with an average between 60 and 80 microns. 

 Thus, the survey of the Gold Coast continental shelf by Buchanan 

 (1957) revealed few Amphipoda, but the sediments were extremely 

 coarse. J. L. Barnard (1963) gave evidence showing the impoverish- 

 ment of amphipodan faunas on coarse sediments. 



Differences among tallies of Amphipoda in those regions listed 

 in table 31 are not sufficiently extreme to be significant, except in 

 the case of Mediterranean France. Atlantic France and the southern 

 half of California have approximately equal numbers of species, 

 even though the median latitude of France is nearly 15° farther 

 north than the median latitude of that portion of California. According 

 to Ekman (1953) the warm-temperate region of the northeastern 

 Atlantic extends northward approximately to the western end of the 

 English Channel; thus, warm-temperate waters occur about 15° 

 farther north in the eastern Atlantic than they do in the eastern 

 Pacific. Indeed, Ekman did not recognize a warm-temperate north- 

 eastern Pacific region but this erroneous conclusion was rectified by 

 Garth (1955). The faunas of the southern half of California and 

 Atlantic France are comparable in their equal division between 



