Climate forcing and sudden change in marine ecosystems
Rationale—Continental shelf ecosystems occasionally undergo "regime shifts" – abrupt reorganization events that can have deleterious social and economic effects on fishing communities that rely on affected species. Regime shifts are often interpreted as transitions to alternative ecosystem...
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2014
|
| Subjects: | |
| Online Access: | https://eprints.utas.edu.au/22900/ https://eprints.utas.edu.au/22900/1/Litzow_whole_thesis.pdf |
| Summary: | Rationale—Continental shelf ecosystems occasionally undergo "regime shifts" – abrupt reorganization events that can have deleterious social and economic effects on fishing communities that rely on affected species. Regime shifts are often interpreted as transitions to alternative ecosystem states after external perturbations such as fishing pressure cross a critical threshold; they have also been related to shifts in modes of internal climate variability, such as the Pacific Decadal Oscillation and North Atlantic Oscillation. However, hypotheses explaining regime shifts are extremely difficult to test, given the multivariate nature of both stressors and community response, the frequent paucity of data at adequate spatial and temporal scales, possible non-linear and non-stationary relationships, and the use of observational data that preclude strong inference. Furthermore, regime shifts have often been invoked to explain ecological change without the consideration of competing models, such as the accumulation of more gradual change over time. As a result, both the nature of community-level biological change in continental shelf systems (regime shifts vs. more gradual change), and the identity of factors producing observed change, remain poorly resolved. Given increasing anthropogenic disturbance to continental shelf ecosystems globally, there is a need both for better understanding of the dynamics underlying sudden ecological change, and for tools providing early detection of ecosystem change to allow for proactive management measures that might minimize associated socio-economic disruption. Finally, the regime shift concept is largely based on work in North Atlantic and North Pacific ecosystems, and there is a need for studies in other areas to determine if the regime shift model is widely applicable. Approach—In many continental shelf systems, long-term biological observations that are necessary for testing hypotheses concerning decadal-scale ecological change are extremely limited. The first four chapters of this thesis use the northeast Pacific as a model system, as this region is home to one of the best extant datasets of long term, large-spatial scale biological observations globally. I compiled a set of 38 climatic time series (regional climate parameters and large-scale indices) and 78 biological time series (mostly production or abundance estimates for commercially-important fish and invertebrate populations), from the 1960s to the present, covering the continental shelf between 30°N and 65°N. This dataset allowed me to evaluate internal climate variability, commercial fishing and incremental climate change as factors explaining decadal-scale biological variability (Chapter 1); test competing models of gradual change and regime shifts for explaining decadalscale ecological variability (Chapter 2); develop an approach for evaluating possible ecosystem shifts at the ends of time series (Chapter 2); test for non-stationary biological responses to climate perturbations (Chapter 3); and test statistical tools for early detection of ecosystem transitions (Chapter 4). In Chapter 5 I use the methodology developed in the previous chapters to evaluate a data-poor situation in southeast Australian continental shelf ecosystems, using nine climatic and 12 biological time series (seabird reproductive parameters and recruitment estimates for commercially-important fish stocks) for the period 1967-present. This chapter tested competing hypotheses invoking secular change and regime shifts to explain regional climate-biology covariation. Results—Analysis of northeast Pacific data showed that: the ecological change that is frequently related to regime shifts in the literature is in many instances better described as the result of incremental change over time (Chapters 1 and 2); commonly held assumptions concerning the dominant role of internal climate variability as an external driver on the system are poorly supported by data (Chapters 1 and 2); assessment of incipient ecosystem shifts may be supported both by model-derived statistical indicators and by formal evaluations of the length of observation needed to evaluate ecosystem state (Chapters 2 and 4); and assumptions of stationarity are inadequate for understanding community responses to climate variability (Chapter 3). In southeast Australia (Chapter 5), I found that secular change best described decadalscale climatic and biological variability, with little evidence for a role of regime shifttype variability in the system. Conclusions—My findings from the northeast Pacific suggest several revisions to current understanding of decadal-scale variability in continental shelf ecosystems. Even in ecosystems that are generally free of overfishing and heavily loaded by leading modes of global climate variability, anthropogenic drivers (fishing, secular climate change) are at least as important as climate variability as agents of ecological change. Furthermore, gradual change over time is at least as important to consider as more obvious regime shifts when analyzing decadal-scale ecological variability. The tests of early indicators for sudden change included in this thesis are the first application of these proposed methods to actual fisheries management data, and demonstrate their potential usefulness in this context. Finally, while my findings confirmed the importance of secular change in southeast Australian ecosystems, analysis of northeast Pacific data shows that the small sample of biological time series in Chapter 5 is likely to produce an underestimate of the complexity of temporal variability in the system. |
|---|