Kulbicki and Wantiez Estimates of fish stocks by shrimp trawl and visual survey off New Caledonia 



671 



Abundance and biomass 



Density estimates frcim tfansects were on average 9.7 

 times larger tlian from trawls, this diffei'ence being 

 highly significant (F test for paired comparison (Sokal 

 and Rohlf 1981) at a = 0.01) (Table 4). Depending on 

 the species, the ratio between the two methods varied 

 from 0.9 to 80 (Table 5). If one considers only the eight 

 most common species mentioned in the previous sec- 

 tion, then the ratio is 7.93. This would indicate a catch- 

 ability of the trawl net of 0.103 for all species and 0.127 

 for the main species. Species for which the ratio is close 

 to 1 (Synodontidae, Cnnthign^ter eompres:sa, Upeneus 

 spp.) are difficult to detect underwater. This can be due 

 to either their behavior (e.g., Synodontidae are usual- 

 ly tnotionless on the bottom, and at times, half buried 

 in the sand) or to their coloration (Canthigatiter com- 



pr-essa being well camouflaged among algae, and 

 Upeneus spp. being able to drastically change their 

 coloration to mimic the bottom). In addition. Table 6 

 indicates that these criptic species are usually found 

 singly or in pairs. 



Species with a high ratio (Apogonidae, Leiognntluis 

 rivulatus. Pristotis jerdo)ti) are not caught by the net 

 for two main reasons. Either they are too small (Apo- 

 gonidae, most Pristotis jerdoni) or they swim too high 

 above the bottom (Leiognathus rivulatus were usually 

 0.5-3 m aliove the bottom) for the net to catch them. 

 These fish are also found in small schools (21-36 fish/ 

 sighting) (Table 6). The correlation between the two 

 methods (Table 5) is some indication of the patchiness 

 of the distribution of these species. Thus a high cor- 

 relation such as for the Synodontidae (?' = 0.75) or 

 Leiognatlius rirulntiis (r = 0.92). indicates that these 



