Effect of design modifications and soak time variations on Antillean-Z fish trap performance in a tropical estuary

Design and operational aspects of Antillean-Z fish traps were investigated under tropical estuarine conditions. Catch rates were compared between different tidal states, mesh sizes, entrance funnel designs, soak times, trap volumes, bait container designs and parts of one lunar cycle. The size structures of major fish species trapped were compared between small and large mesh sizes. Catches were found to be reduced at high tide. Traps with the small (12.5 mm) square mesh produced higher catch rates than those with large (42 mm) hexagonal mesh. The smaller mesh traps retained smaller fishes and captured fewer large sparids (two species) than did the larger mesh traps. Higher catch rates occurred in traps with straight, conical entrances during short (2-h) trap soaks than during long (1½-d) soaks, while trap catches with horse-head entrance funnels did not differ significantly between the two soak times. This probably reflects greater ease of trap escapement from straight funnels once the initial high attractiveness of trap bait has declined. For most species, more fishes were captured during short (2-h) than long (1½-d) soaks, indicating that short soaks may be advantageous in many situations. Larger trap sizes (volume approx. 0.92 m^3) produced higher catches than smaller trap sizes (volume approx. 0.40 m^3) for one species only, Lutjanus russelli, suggesting that for many species the logistic advantages of smaller traps may make them a useful sampling option. Of two bait container designs tested one with more numerous but smaller holes was found to be advantageous. With this design there was always bait remaining in the container when the trap was censused, ensuring that all traps were baited throughout the soak. The different responses of particular taxa to various trap design modifications and operation investigated in this study suggests that all aspects of trap design and operation need to be considered when designing particular trap sampling program.