MEPS 366:111-117 (2008)  -  DOI: https://doi.org/10.3354/meps07523

ABSTRACT: The relationship between climate, represented by the North Atlantic Oscillation (NAO), and the calanoid copepod Calanus finmarchicus has been extensively studied. The correlation between NAO and C. finmarchicus has broken down (post-1995). In the present study, we revisit the relationship between C. finmarchicus and the NAO. Our reanalysis shows that previous treatment of this data did not take into account 2 aspects of both the C. finmarchicus and NAO index time-series: (1) the presence of significant trends and (2) significant autocorrelation. Our analysis suggests that previously reported relationships between NAO and C. finmarchicus abundance can be explained largely by the trends in both data series. Removing the trend from both time-series resulted in a decrease in the amount of C. finmarchicus abundance variability explained by the NAO. Trend removal eliminated the autocorrelation from the NAO time-series, but not from the C. finmarchicus time-series. Partial autocorrelation analysis showed that the autocorrelation present in the C. finmarchicus time-series is only found at a lag of 1 yr, suggesting strong, year-to-year connectivity in this population. We included the lagged C. finmarchicus abundance into a regression with the NAO and found that C. finmarchicus variability is explained by the previous year’s abundance and, to a much smaller extent, by NAO variability. Limiting the time-series to the most recent 22 yr period (1981 to 2002) showed that the NAO is no longer correlated to C. finmarchicus abundance, and the autocorrelation in the C. finmarchicus abundance series also appears to be weakening.

KEY WORDS: North Atlantic Oscillation · Calanus finmarchicus · Trend · Autocorrelation · Time-series