Abstract

Collapsing fisheries are significant global problems and what drives these collapses is hotly debated. Early-warning indicators of fish populations in trouble are urgently needed that could be used as indicators of change in, and reference points for, exploited marine populations. Marine ecosystems are characterized by complex non-linear patterns of variability, however these ecosystems often display remarkable quantitative regularities at different spatial and temporal scales. These regularities usually take the form of power law distributions that are well described through scaling patterns under the perspective of fractal theory. Scaling patterns represent self-similar processes – processes which have the same distributions at short and long time and/or spatial scales. In the present work, we applied fractal-based analyses to detect time-invariant scaling relationships in daily catch time series from the smooth pink shrimp (Pandalus jordani) fishery off the west coast of Vancouver Island, Canada. We also compared these with scaling patterns previously identified in daily catches of the southern hake (Merluccius australis) fishery of Chile. Scaling parameters, such as the Hurst coefficient (H) (self-similarity index) and the fractal dimension (D) (an index of complexity), were estimated. More complex scaling models (e.g., a Universal Multifractal Model) were also used to detect and estimate multiple scaling patterns (multifractality) and their parameters. The possibility to use these parameters – and their variability as a function of time – to identify different stages of these fisheries was analyzed. Underlying processes as sources of the observed fractality/multifractality are proposed. proposed.