Drivers of pelagic plankton diversity across space and time

Direct observational records are generally too short to capture biodiversity dynamics on decadal and longer time scales, making the fossil record essential for investigating the long-term ecological response to climate change. Although past biodiversity changes are typically inferred from sedimentary time series, most of our knowledge about the underlying environmental drivers is based on spatial biodiversity patterns. However, the extent to which spatially derived models can be used to understand and/or predict temporal change remains poorly known. In addition, sedimentary archives provide temporally integrated records of biodiversity. Because such temporal integration is known to smooth short-term dynamics, it is essential to determine whether the environmental variables that govern biodiversity dynamics across short time scales also explain smoothed changes as recorded in the sediment. Therefore, improving our ability to derive meaningful ecological insights from the fossil record requires addressing both the validity of time-for-space substitution and the effects of temporal integration.

Here, we investigate these topics using planktonic foraminifera, a globally distributed group of unicellular plankton with an exceptional fossil record often used as a model system for pelagic biodiversity dynamics. We analyse 96 time series (average resolution: 17 days) of species composition from moored sediment traps spanning over 150 years, together with a large spatial dataset of ocean-floor sediment species assemblages (n ~ 4,000) from the pre-industrial era. These datasets enable comparison of community turnover in time across temporal scales of integration from months to multiple years and in space with temporal integration ranging from months to centuries. 

Preliminary results indicate that sea surface temperature is the best predictor of spatial turnover across integration times ranging from months to centuries, with its explanatory power increasing  as time integration increases. In contrast, predictors of temporal turnover are less consistent, and temperature only emerges as a dominant predictor when biodiversity is integrated over longer (annual) periods. Together, these findings suggest that we can predict past and future patterns of spatial biodiversity change with confidence, but that spatial models currently used to predict temporal biodiversity change may fail to capture the full extent of (local) temporal community change.