184 



atypical timing of parturition for the above monachines, combined with an apparent lack 

 of alternative explanations such as competition for pupping sites, suggests a southern origin 

 with a northern re-invasion (see Hendey 1972). In the case of Mirounga angustirostris in 

 particular, this implies that Mirounga (or its ancestors) initially migrated southward along 

 the Atlantic side of South America before rounding Cape Horn and moving northward 

 along the Pacific side, as suggested by Hendey (1972). 



Monachus schauinslandi presents several problems for this re-invasion hypothesis. It 

 possesses a normal spring pupping time, although the observation that pupping may begin 

 as early as December for this species (Kenyon 1981b; King 1983; also references in 

 Hayssen et al. 1993) suggests that M. schauinslandi has perhaps shifted back towards a 

 normal spring pupping time faster than the remaining northern monachines. Of more 

 concern, however, is the requirement of placing M. schauinslandi in the Pacific. Only two 

 routes are possible - through the Central American Seaway, as is universally suggested, 

 or around Cape Horn, as has been postulated for Mirounga spp. (Hendey 1972) - and 

 neither is adequate here. Use of the Central American Seaway possesses a time limit, with 

 its closure to marine dispersal occurring at least three and a half to four MYBP (Ray 

 1976a; de Muizon 1982a). In order to accord with our proposed phylogeny, the northward 

 migration of Monachus spp. from the higher southern latitudes (after their derivation from 

 lobodontine stock) would have had to be very rapid indeed. This scenario may well be 

 impossible when allied with the suggestion that the full adaptation of the lobodontines to 

 the high Antarctic latitudes occurred no more than four million years ago (Ray 1976a). 

 However, combined with the suggestion of multiple invasions of the Antarctic continent 

 by the lobodontines (Hendey 1972; McLaren 1975; de Muizon 1982a), it may be that the 

 divergence of the lobodontines and Monachus spp. occurred in the middle southern 

 latitudes. The only other possibility, that of an early dispersal of Monachus schauinslandi 

 through the Seaway (with little or no previous southward migration), would strongly 

 contradict our cladogram, as it would presumably render this species as the sister group 

 to the remaining monk seals, and very likely to the remaining monachines as well. This 

 is the accepted route however, with the invasion occurring between 8.5 to 13 MYBP 

 (Hendey 1972) or even earlier (Repenning and Ray 1977). Costa (1993) even goes so far 

 to suggest that M. schauinslandi did not even migrate through the Central American 

 Seaway with the remaining ancient phocids in the first place. An important point to keep 

 in mind with such early dispersal hypotheses for M. schauinslandi is that the main 

 Hawaiian islands are only about six million years old (although the more westerly islands 

 of the chain such as Laysan and Midway do date from 20 to 28 MYBP respectively) 

 (Clague & Dalrymple 1989). Therefore, if such a scenario holds, it is more than likely 

 that the ancestors of M. schauinslandi remained tied to the American coast for some time 

 before a founder population reached the Hawaiian islands. 



The second route, encircling Cape Horn, is not a viable alternative either. Again, it implies 

 an early separation of Monachus schauinslandi from the Monachus lineage, leaving M. 

 monachus and M. tropicalis either as sister taxa, or requiring them to migrate northward 

 in parallel, possibly on either side of the Atlantic. This latter option does have the 

 advantage of agreeing with the current (or historical for M. tropicalis) distributions of the 



