Jump to Main Content
Do Gulf of Alaska fish and crustacean populations show synchronous non-stationary responses to climate?
- Puerta, Patricia, Ciannelli, Lorenzo, Rykaczewski, Ryan R., Opiekun, Michael, Litzow, Michael A.
- Progress in oceanography 2019 v.175 pp. 161-170
- Crustacea, climate change, climatic factors, fish, marine resources, models, principal component analysis, surface water temperature, time series analysis, variance, Gulf of Alaska
- Changes in the abundance and productivity of biological populations in the North Pacific have often been associated with large-scale modes of climate variability. The Pacific Decadal Oscillation (PDO), which describes spatio-temporal variability in North Pacific sea surface temperature (SST), correlates with much of this variability. However, since the late 1980s, the North Pacific Gyre Oscillation (NPGO) has explained an increasing proportion of variance in North Pacific climate properties. Ecological responses to this change in the proportion of variance ascribed to the two climate patterns remain poorly understood. Here, we test the hypothesis that relationships between biological time series and climate covariates (SST and the PDO) differ for nine Gulf of Alaska fish and crustacean populations before and after the late 1980s. Additionally, we evaluate whether non-stationary climate-biology relationships arose synchronously across populations as a community response. We used different formulations of Generalized Additive Models in a population and community context and compared results to the classical approach of aggregated population responses based on Principal Component Analysis (PCA). The results showed that climate-biology relationships weakened or reversed for most populations in the late 1980s, coinciding with the increase in variance of the NPGO. However, these non-stationary responses were highly species-specific and did not arise synchronously as a community response. We show that PCA does not represent community dynamics properly when only few species covary in time and exhibit long-term trends. Therefore, this approach might not be always useful to detect synchronous changes among biological time series as a community response. Novel associations among climate variables and novel climate-biology relationships are expected to become increasingly evident with future climate change, and the recognition of switches between different explanatory variable-response relationships may be critical for successful management of marine resources during transitions to these novel climate states.