Circumventing chronological uncertainty in attempts to detect and understand non-linear ecosystem responses in shallow lake paleorecords

Abstract:

Shallow lakes can undergo rapid changes in key biotic components. These phenomena, which include loss of submerged macrophytes, fish kills and algal blooms, can occur at sub-seasonal timescales and are often reported to be non-linear, threshold responses to a gradual intensification of an external driver and reflective of a change in state. Although such threshold responses are widely reported, a recent meta-analysis found that most such changes cannot be unequivocally confirmed as true threshold responses. This is because clear records of system stability in the face of a gradual increase in external driver intensity followed by rapid system change are lacking, as are records of post threshold stability in the new state following release of external driver pressure. That threshold responses were not confirmed often reflects insufficient time series of before or after data to establish driver variability and ecosystem stability. In this context, paleo studies provide a means to clearly identify non-linear, threshold responses in shallow lake ecosystems.

The challenge of detecting evidence of non-linear responses in shallow lake ecosystems is often seen as a chronological one. Highly resolved and accurate sediment chronologies coupled with historical records of external driver intensity do provide a means to detect non-linear, threshold responses, but such chronologies are rare in shallow lakes. Fortunately, the ‘tight chronology-historical record of external driver’ approach is not the only, or even the most direct, way to detect non-linear ecosystem responses in paleo records. An alternative, more direct approach is ecosystem response and external driver intensity to be preserved in the same sedimentary record. Theoretically, it is arguable whether any chronological control is needed at all to determine if a non-linear response has occurred, for the key is not how quickly an ecosystem response may occur or if it is linear with respect to time, it is whether it is linear with respect to driver intensity. Accordingly, it is proposed that researchers seeking to detect and understand non-linear ecosystem responses in paleorecords should focus on deriving sedimentary records of external driver intensity rather than expensive and likely futile efforts to derive high resolution chronologies.