Análise comparativa das armadilhas, utilizadas para a captura do caranguejo africano (Chaceon gordonae, Ingle 1985), no entorno do arquipélago São Pedro e São Paulo, Atlântico Equatorial (Brasil)

Abstract

With the initial aim of investigating the composition and catch rates of different trap models (rectangular large, rectangularmedium and conical) in the São Pedro e São PauloArchipelago region in waters above 200 m, 14 exploratory campaigns were carried out in the period of 6 years (January 2012 to February 2018), comparing the CPUE values for selectivity variables,such as the [First Quarter (FiQ), Second Quarter (SQ), Third Quarter (TQ) and Fourth Quarter (FQ)], depth extract(200 to 700 m), immersion time (I -up to 12h, II -between 13 and 20 h and III -above 20 h) and bait [head of tuna, cangulo, xaréu and mixed (cangulo and xaréu)]provided by the vessel. The catch rates, expressed in terms of CPUE (individuals / trap), were analyzed for each factor separately and among the trap models used. As a result, 1135 specimens were captured, divided into 18 distinct species.Among the species collected, the African crab (Chaceon gordonae) was a more abundant species, with a relative participation of 61.76% of the catches. A significant difference was observed between the LargeRectangular Trap (LRT) and the Conical Trap (CT) (KW = 13.584; GL = 1; p= 0.0002281) Small Rectngular Trap (SRT) and CT(KW = 5.7909; GL = 1; p= 0.01611), whereas the rectangular shapes did not show any static difference between each other (KW = 1.0532; GL = 1;p= 0.3048). In the analysis of the CPUE values for bycatch(the other species captured), there was a significant difference between the LRTand SRT(KW = 22.619, GL = 1;p= 0.975x10-6) and for the ARP and AC (KW = 15.389; GL = 1;p= 6.897x10-5), but between the LRTand SRTthere was no statistical difference (KW = 1.7828; GL = 1;p= 0.2565). When analyzing the data for trap selectivity variables, for Target Species (TS), only a small significantdifference was observed for LRTwhen analyzing CPUE data for month variable (KW = 9.1554; GL = 3; p = 0.02729). In this way, the rectangular models are more efficient in relation to CTand statistically they are equal to each other. Thus SRTis the most suitable for capture in the SPSPAenvironment, due to the lower cost in its construction, because it is lighter and occupy less space in the vessel in relation to LRT.