|
Résumé :
|
Effective sampling of marsh nekton is difficult due to the organisms' use of the marsh-edge and/or marsh surface during high tide. Quantitative sampling approaches currently used are expensive, require permanent structures, and can require a considerable number of personnel for implementation. Our purpose was to assess the use of Breder traps (T) as a sampling method capable of documenting relative abundance of nekton. We sampled marsh habitats (within 1 m of marsh grass) in five bayous using seines at high (HS) and low (LS) tide and compared them with rank abundance and similarity data. Seining (n = 3/tidal stage) was conducted adjacent to each set of traps (n = 4) which were retrieved at low tide. Four transient (Engraulidae, 34.7%; Penaeidae, 12.4%; Portunidae, 6.8%; and Sciaenidae, 1.2%) and four resident families (Palaemonidae, 28.1%; Fundulidae, 9.2%; Atherinidae, 3.2%; and Gobiidae, 1.1%) met our requirements (greater than or equal to 1% of all nekton captured) for analysis and accounted for 96.6% of the total nekton captured. High seine and LS collections were most similar (Jaccard's index, 0.58), followed by T and LS (0.46) and HS and T (0.37). Transient families were captured in greatest numbers and higher rank with seines (LS > NS > T) while two resident families (Palaemonidae and Fundulidae) dominated T collections (T > LS > NS). Our data suggests that Breder traps adequately sample resident nekton which use the marsh surface and should be considered in future studies which require only CPUE estimates of abundance.
|