Wavelet analysis of correlation among Canary Islands octopus captures per unit effort, sea-surface temperatures and the North Atlantic Oscillation

Abstract

Short-term fluctuations in the octopus catches off the Canary Islands have been attributed to changes in the sea surface temperature (SST) and the North Atlantic Oscillation (NAO) pattern. These results have been found using stationarity assumptions. However, the behavior of environmental systems is not always linear, and environmental time series do not always satisfy the statistical property of stationarity. Wavelet spectral analysis is a methodology which can deal with non-linear, non-stationary and noisy time series. Cross wavelet analysis (wavelet coherence) is applied to investigate the environmental effects (SST, NAO) on octopus abundance fluctuations measured as capture per unit of effort (CPUE) from 1989 to 2007 in the waters of the Canary Islands. A slightly positive correlation exists between NAO and CPUE at lags and leads of a few months. Additionally, a good relationship between SST and CPUE exists on the same seasonal scale, but there is a relatively weak relationship between SST anomalies and the NAO. When the analysis is extended to the interannual scale, the wavelet coherence identifies a statistically significant relationship between CPUE and NAO, but this does not happen when the wavelet coherence between CPUE and SST is computed. These results suggest that fluctuations in octopus catches could be the result of SST fluctuations but in synergy with other unknown environmental variables which are also affected by the NAO pattern.