Supplementary material
Potential biases of ingestion estimates
One of the potential biases during our bottle incubation experiments was the ‘food chain effects’ (Nejstgaard et al. 2001). As the incubations proceeded, nanoflagellate growth in the experimental bottles could have be higher than nanoflagellate growth in the controls, because nanoflagellates in these bottles were released from ciliate and dinoflagellates grazing pressure as these grazers were consumed by copepods. This may have significant implications for the validity of copepod grazing rates obtained from traditional bottle incubation studies (e.g. Calbet & Landry 1999 and others). This bias could have produced an apparently low grazing rate on nanoflagellates, even when some copepod grazing did effectively occur. However, ciliates were almost absent during our study and dinoflagellates in the incubation water were scarce and mostly corresponded to the genera Gymnodinium sp. and Alexandrium sp., both genera including autotrophic and heterotrophic species. These cells were present only in the Coastal and Estuarine stations during summer and in the Estuarine station during spring. By considering potential dinoflagellates ingestion (Jakobsen and Strom 2004; Vargas and González 2004), we estimated that uncontrolled intra-guild predation may have reached 1 - 3% of biomass and 3 – 95% of cell abundance.
Tables
Table 1: Mean values of egg diameter (ED), adult body length (BL), egg production (EPR), and ingestion rates (IR)of the transect populations of Acartia tonsa sampled in the study area. Mean values are shown with standard deviation (±).
EstuarineMiddleCoastal
summerspringsummerspringsummerspring
ED (µm)80.0± 3.679.9± 3.880.0± 3.579.9± 3.775.9± 2.475.5± 2.5
BL (mm)1.102± 0.081.045± 0.071.047± 0.071.038± 0.101.055± 0.071.046± 0.09
EPR (egg f-1 d-1) 5 ± 35 ± .69± 426 ± 939 ± 1262 ± 28
IR (µg C f-1 d-1)1.4± 12.4 ± 12.8 ± .96.8 ±2.35.2 ± .72.6 ± .4
Principal component Analysis
Environmental (temperature, salinity, pH, isotope signals), and biological (adult and egg size, food availability, ingestion and reproduction) data was analyzed using multivariate Principal Component Analysis (PCA) to elucidate the segregation degree among sampling stations. The two first component, explaining >80% of the variance, showed a clear segregation of sampling stations where temperature, reproduction and isotope signals contributed greatly to the first component, while salinity and egg size contributed mostly to second component.
Figure 1: Results of PCA analysis showing the two first components that explained >80% of the observed variance on factors and variables, and which clearly separates the three stations in the coastal gradient: Estuarine (E), Meddle (M), and Coastal (C).
Isotope Analysis
A fraction of plankton samples was immediately frozen at -20 °C after sampling, in order to determine the stable isotope (13C and 15N) composition of copepods. Several (> 200) adult individuals of A. tonsa were then ground to a fine powder and subsamples were washed to remove carbonates and redried accordingly to Bunn et al. (1995). The dry ground samples were then analyzed by mass spectrometry (VG Micromass 602C equipment) at the PROFC Laboratory of the ‘‘Universidad de Concepción, Chile’’. The reference materials were the international standards Vienna Pee Dee Belemnite (PDB) for carbon and atmospheric N2 for nitrogen. The relative importance of allochthonous terrestrial C and N contribution to the planktonic assemblages was assessed by applying a two-source mixing model (Bianchi 2007).
.
Figure 2: Comparison of stable isotope composition of copepods sampled during summer campaigns in the upper (<12 m) layer of each coastal station. Bars are mean values plus standard deviation, and significant differences among stations are denoted by X.
Figure 3: Significant relationship between egg production rates (EPR, µg d-1) and egg diameter (ED, µm) estimated and measured, respectively, on each sampling station. Dotted lines denote 95% confidence interval.
References
1