ASSOCIATIONS 591 



Carapus contains slender eel-like fishes, some of which inhabit crevices, 

 while others are inquiline in habit. The needle fish, C. acus for example, 

 dwells in the intestine of a sea-cucumber, which it enters tail first. The fish 

 appears to react first to a chemical stimulus (probably to a substance 

 carried in the mucus of the holothurian), and then to the anal current 

 (2a, 90). A small fish Apogonichthys utilizes the gastropod S trombus, and 

 many gobies and blennies are inquilines in internal chambers of reef 

 sponges. These fishes are generally small and slender in form and live in 

 tubular sponges or those with large internal cavities (5, 39, 40). 



Most investigations of commensalism have been descriptive in nature, 

 and it is only recently that these associations have been subjected to 

 experimentation. On analysis it appears that three aspects need to be 

 recognized in an appreciation of commensal associations. These are: the 

 economics of the association, i.e. their effect on survival of both associates; 

 their evolutionary origin; and the behaviour patterns concerned with 

 initiating and maintaining the commensal condition. In many respects the 

 same conditions of analysis are applicable to parasitic and symbiotic 

 relationships, described in later sections. 



Precise analyses of the economics of commensal associations have still 

 to be sought, but it is obvious that there is often sharing of food, provision 

 of sheltered environments, etc. It would be interesting to have information 

 establishing ecological parameters, limits of tolerance and partition of 

 foodstuffs in representative associations. 



The behaviour of many commensals is obviously specialized and adapted 

 towards cementing the association. Three problems in commensal behaviour 

 can be distinguished. First, how do the commensals succeed in reaching 

 their host? Second, once the host is reached, what factors bind the com- 

 mensal to the host? Third, how does the reciprocal behaviour of the host 

 affect the association? 



It must be confessed that little is known about how commensals seek out 

 and adopt their hosts. In species with planktonic stages it is suggested that 

 the larva first chooses a suitable substrate, and then reacts to other, more 

 precise, stimuli, in locating the host within that environment. Once the 

 host is reached, other factors intervene to cement the association, among 

 which thigmotaxis and chemotaxis are probably of great importance. The 

 specific responses of pea-crabs and polynoids to certain host species show 

 that the releasing stimuli are likewise highly specific in character. Where 

 the host plays an active part, e.g. in associations between anemones and 

 other animals, the behaviour patterns of both participants need precise 

 analysis. Davenport in a recent review (18) has discussed these factors in 

 terms of established associations and their evolutionary origin. 



PARASITISM 



Parasitism involves modes of life easily recognizable as such, but difficult 

 to define because of their heterogeneity. Ectoparasites and endoparasites 



