Short-term changes in the concentration of chlorophyll.
Diatoms dominate numerically the phytoplankton and Skeletonema Greville, 1865 is the most representative genus (González 1982). Phytoplankton biomass is relatively homogeneously distributed in the water column, with a tendency to higher values between 2 and 6 m depth (Augsburger 1981, Calliari & Antezana 2001). Zooplankton is represented by few copepod species like Acartia tonsa Dana, 1848, Paracalanus parvus Claus, 1863, Centropages brachiatus, Dana 1849, Oithona Baird,
patagoniensis Brady, 1883, and Calanus chilensis Brodsky, 1959 (Peterson & Bellantoni 1987, Peterson et al. 1988, Calliari 1999).
Sampling was performed near the centre of Coliumo Bay (depth ca. 10 m) daily between April 11th and 20th, 1997 (pre-storm period), and subsequently on April 22nd (during the storm), 26 and 30th (post-storm period). On each occasion samples for chlorophyll-a (chl) concentration were taken with a Niskin bottle at 1-meter depth intervals from surface to 9 m (except on April 22nd when samples only from 0 and 9 m could be taken). The concentration of chl and phaeopigments was estimated following Holm-Hansen et al. (1965) from duplicate 50 ml sub samples filtered on GF/F (0.7 μm). Pigments were extracted in 90% acetone at 4ºC and in the dark for 24 h, and the fluorescence of the extract was measured before and after acidification in a Turner Designs 10-005 R fluorometer. Taxonomic analysis and quantification of microplankton was based on 10 or 50 ml aliquots from one formol-preserved daily sample (4 m depth) settled for 24 h (or >48 h for 50 ml chambers) in Utermöhl chambers and examined under an inverted microscope at 200-400 X (Leitz, DM-IL). Minimum individual size of organisms considered in microscopic analyses was ca. 5 µm; thus this dataset includes microplanktonic (20-200 µm) and part of the nanoplanktonic (2-20 µm) size range. Identification was performed to the lowest taxon possible using appropriate references (Rivera & Gebauer 1989, Tomas 1997); nomenclature follows Tomas (1997) and Guiry & Guiry (2006). th
Wind direction and intensity was recorded at 1-minute interval with a Heath Kit 5000C weather station placed on the bay shoreline at Dichato Marine Laboratory (Universidad de Concepción). That record was interrupted on April 23rd owing to severe damages caused by the storm to the wind sensors. Wind for the period April 23rd to April 30th was estimated from data recorded at Bella Vista Meteorological Station (36º47’S, 73º07’W) using transfer functions specifically developed to estimate
the wind velocity at Coliumo Bay from wind data recorded at Bella Vista Station; such functions yield best predictions for the N-S velocity component (r = 0.93) (Calliari & Alfaro 1997) which is the most relevant in the present case (see results, fig. 2). Ancillary temperature and salinity data for the pre-storm period were obtained from daily CTD profiles (Sensordata 202).
A stratification index (S, m-1, equation 1, Cullen & Eppley 1981) for each daily chl profile was calculated as the maximum value of the chl gradient with respect to depth (term in the numerator in equation 1) and normalized by the average chl concentration on that day (denominator):
Z C h l ∑ 9 0 M a x ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ δ Chl δ
n where Z represents depth and n stands for the number of depth intervals (10).
The association between S and wind velocity was assessed by Spearman correlation analysis. The effect of the storm on chl level and microplankton cell concentration was analysed with the Mann-Whithney U test (pre-storm and post- storm conditions). The effect of the storm on the microplankton community composition was assessed by complementary non-parametric multivariate techniques. A hierarchical cluster analysis with group average linkage was employed to group successive sampling dates; this analysis was based on a Bray-Curtis similarity matrix calculated from squared root-transformed species abundance data (Clarke & Warwick 1994), and clusters were defined using a similarity threshold reference value equal to the average of the Bray-Curtis matrix (Arancibia 1988; Rodríguez-Graña & Castro 2003). Differences in the taxonomic composition and species abundance of the microplankton due to the storm were explored by Analysis of Similarity (ANOSIM), a permutation-based test between a priori defined groups (treatments) where generated R statistic values represent the distance scaled to the range 0 - 1 (0 indicating no difference among groups and 1 indicating that all samples are more similar within groups than to any sample from any other group, Clarke & Gorley 2001). A Similarity Percent analysis (SIMPER) was used to identify those taxa that presented largest variation between factor levels in ANOSIM. All non-parametric multivariate analyses were performed using PRIMER v5.2.
Pan-American Journal of Aquatic Sciences (2007), 2 (1): 13-22