PELAGIC BIRO FAUNAS OF THE INDO-PACIFIC OCEANS 479 



Can one map out well-marked "faunal zones" on the bases of these 

 pelagic bird distributions, demarcating the zones, to use the definition 

 of Ekman (1953, p. 80), "where the most clearly defined change in 

 fauna occurs"? 



Such breaks in distribution are readily discernible. Murphy (1936, 

 pp. 65-80) has recapitulated his former well-known thesis that the con- 

 vergences between water masses of different temperatures are powerful 

 barriers in pelagic bird distributions. His recognition of Tropical, 

 Sub-tropical, Sub-antarctic (Antiboreal) and Antarctic Zones is abund- 

 antly demonstrated by the facts. These temperature-determined zones 

 agree in a broad sense with the boundaries of the three main faunas 

 as dealt with in this paper. In the southern hemisphere the sub-tropical 

 convergence is the absolute southern boundary of the Pan-Tropical 

 Fauna, and the northern boundary of the Southern Fauna (cf. the 

 illuminating map of complementary distributions of type birds of these 

 faunas in Murphy, 1936, p. 165). In southern Australia there are 

 anomalies in correspondence between birds and hydrological factors 

 ^\'hich are referred to in part by Fleming (1941 a). 



However, on a broad canvas the picture drawn is a real one. Such 

 concordance between faunas and physical factors emphasises the ancient 

 character of the relationship. Changes in climate which have led to 

 oscillations north and south of the convergences would have resulted 

 in a corresponding ebb and flow of faunal boundaries, and in particular 

 a waxing and waning of the Pan-Tropical Fauna belt. Subdivisions 

 within the areas of main occurrence of the three faunas (such as that 

 between the tropics and subtropics and between the antarctic and sub- 

 antarctic) would have tended to split the faunas into corresponding 

 units, but these appear never to be as important as the greater break 

 between the Pan-Tropical and the two cool-water faunas north and south 

 of it. 



Past oscillations of the convergences have cut off portions of faunas 

 into areas not lived in by their ancestors, or provided temporary bridge- 

 heads across unfavourable stretches of water, over which transgressions 

 were possible. Later when times changed probably most such ad- 

 venturers failed to survive but those that did now form the immigrant 

 fauna to w^hich reference has been repeatedly made in this paper. 



IV. Some Special Problems 



1. THE INDO-MALAYAN BARRIER BETWEEN THE INDIAN AND PACIFIC OCEANS. 



Though there are continuous watenvays through the maze of islands 

 in the east Indian archipelagoes, between the Indian and Pacific Oceans, 

 a significantly large number of species, characteristic of the tropical 

 zones of both oceans, are absent from them. 



