MARINlEi GAMMARIDEAN AMPHIPODA 25 



is substituted an adaptation towards more favorable leverage for 

 burrowing into tissues (plant or animal), maintaining special positions 

 in tubes or channels and in streamlining. Rigidification often occurs 

 without decrease in urosomal or uropodal size. The tube-dwelling 

 Ampehscidae and the inquiUnous Dexaminidae have urosomites 2-3 

 coalesced while the hgnivorous Cheluridae have the urosome greatly 

 increased in size and the uropods greatly enlarged or modified. Jumping 

 ability is very strong in the Cheluridae. 



The ecological linkages to urosomal evolution in the Gammaridea 

 are manifold; no single solution to functional adaptation has been 

 necessary. The segmented urosome is a remarkably stable feature 

 of the Amphipoda; that and its three pairs of uropods are a revolu- 

 tionary feature unique to Amphipoda. That it has undergone only 

 one major reversion (Caprellidea) , yet has imprinted upon it numerous 

 adaptive features whUe there is maintained a minimal structural 

 stability, attests to its creation as a major part of the success of 

 amphipods. Like the mandibles, its basic conservativeness has prob- 

 ably been a factor in the successful dispersal of amphipods into 

 many niches. The paradox of the almost complete loss of abdomen 

 in Caprellidea appears to be ameliorated by the recent discovery 

 of the Caprogammaridae (Kudrjaschov and Vassilenko, 1966) in 

 which a podocerid-like metasome and urosome are maintained and 

 by recall of the caprellidean Cercops in which the abdomen, though 

 distinctly vestigial, is clearly macroscopic. The Caprogammaridae 

 seem to demonstrate that the loss of urosome was not the trigger 

 for the development of caprellids; rather, the extreme development 

 of a tubular body with elongate segments, reduced coxae, terminal 

 migration of various legs, reduction in numbers of gUls and brood 

 plates, all correlated with numerous behavioral changes were more 

 important. The ultimate radiation of caprellids, of course, may have 

 been assisted by the removal of a relatively useless abdomen and the 

 complete amalgamation of head and pereonite 1. 



Development of a dorsal shield in a few Gammaridea has been 

 discussed by Gurjanova (1962). This feature is simply an extreme 

 dorsoventral depression of the body and a splaying of the coxae 

 in the Phliantidae. The abdomen is flexed under the thorax possibly 

 as an additional protection to the ventrum owing to removal of the 

 lateral shield. Many isopods, without flexed abdomen do not appear 

 to require this protection, so there may be other reasons for phliantid 

 flexion. Gurjanova points out possible precursors to phliantids in 

 the caUiopiid genera Chosroes and Sancho but those genera must 

 yet be strongly segregated from phliantids because they have fully 

 developed mouthparts and uropods. They probably should be allo- 

 cated to a new family in order to qualify the development of a 



