EFFECT OF PLANKTON-NET MESH ON QUANTITATIVE ESTlMATlON OF Oikopleura dioica

CAPITANIO, F.; GONZALEZ, B.; ESNAL, G.

THALASSAS (SANTIAGO DE COMPOSTELA)

UNIV VIGO

Año: 1996 vol. 12 p. 37 - 42

0212-5919

The effectiveness between two plankton nets of different mesh size (90 and 150 um) was compared in relation to the quantitative estimation of Oikopleura dioica and species distribution in the area of Valdes Peninsula and San Jorge Gulf (Argentine, Patagonia) was studied. The frequency distribution of sizes grouped into seven classes was constructed for each net and station, and the densities (n/m3) of each interval were calculated. In order to compare the densities caught by each net, a three way-mixed model ANOVA, with repeated measurements in the net factor was fitted. Model II regression lines were fitted for each mesh size using the geometric-mean method. Parallelism was tested using Fischer´s test. It was found that the 150 ym net underestimates the smallest lenghts (e.g. the 90 ym is a better estimator for sizes less than 0.44 mm). Correction coefficients were calculated as the ratio between the observed and expected densities for each size interval. Densities obtained with the 90 ym net were compared in a randomized block experimental design using Tukey´s method. The immature stage has the highest individual density (size interval 1), intermediate (intervals II and III) and mature stages (0.44 to1.20 mm) ranked in decreasing order. The distribution of O. dioica was studied finding higher abundance off Valdes Peninsula, while it was scarce in the southern stations. This pattern agrees with previous observations in this area.Oikopleura dioica and species distribution in the area of Valdes Peninsula and San Jorge Gulf (Argentine, Patagonia) was studied. The frequency distribution of sizes grouped into seven classes was constructed for each net and station, and the densities (n/m3) of each interval were calculated. In order to compare the densities caught by each net, a three way-mixed model ANOVA, with repeated measurements in the net factor was fitted. Model II regression lines were fitted for each mesh size using the geometric-mean method. Parallelism was tested using Fischer´s test. It was found that the 150 um net underestimates the smallest lenghts (e.g. the 90 um is a better estimator for sizes less than 0.44 mm). Correction coefficients were calculated as the ratio between the observed and expected densities for each size interval. Densities obtained with the 90 ym net were compared in a randomized block experimental design using Tukey´s method. The immature stage has the highest individual density (size interval 1), intermediate (intervals II and III) and mature stages (0.44 to´1.20 mm) ranked in decreasing order. The distribution of O. dioica was studied finding higher abundance off Valdes Peninsula, while it was scarce in the southern stations. This pattern agrees with previous observations in this area.