Re-examining the past to inform future trait-based predictions
Ecology in the Anthropocene is characterized by the rise of cumulative effects on ecosystems, and thus there is an urgent need to synthesize current trajectories of ecological change, predict future ecological outcomes in relation to multiple drivers of change, and importantly, account for the naturally large number of components affected and effecting change. The persistence or loss of species in novel Anthropocene ecosystems will depend on several factors that may be predicted using traits: (i) species’ potential responses to environmental forcing (dispersal, establishment, persistence), (ii) the capacity of species to affect community dynamics (i.e. interactions strengths), and (iii) the combined effect of multiple anthropogenic forces on organisms’ interactions with the environment and one another ( i.e. either additive, antagonistic, synergistic), and (iv) the type and duration of stressors (e.g. Hillebrand and Kunze 2020). Further applications of trait-based approaches to modelling and predicting ecological change are needed, as well as validations of such models. Ample testing grounds exist ideally where ecological processes are reasonably well understood and where the impacts of environmental change can be investigated in relation to that prior understanding of ecosystem form and function. For example, numerous examples of community reassembly and varying degrees of ecosystem modification are documented in the literature, particularly on trophic cascades and ecosystem engineers. These examples have helped to shape our understanding of the processes and factors involved in generating ecological outcomes of change; but hindsight also supplies much of what we know of the role of focal species’ traits in driving ecological outcomes.
Using the growing body of existing trait data, ecologists are positioned to build and test trait-based predictions through hindcasting ecological outcomes in systems that continue to face rapid community reassembly. Transition zone ecosystems— for example, regions at the boundaries between tropical and temperate coastal ocean reef ecosystems in Australia, Japan, the Eastern Pacific and Western Atlantic (see Vergés et al. 2014), where prominent coastal warming and ecological mixing zones persist—are areas where suites of species are readily being redistributed due to environmental forcing. These boundary systems provide excellent opportunities to construct and test trait-based hypotheses of rapid ecological change. For example, the ‘tropicalization’ of temperate reef systems is occurring due to multiple and interacting environmental drivers of change (notably climate change, overfishing and the protection of ecosystems through marine park designation) afford researchers expansive opportunities to propose and test hypotheses across gradients in natural experimental setting (Text Box 1; Figure 8). Studies of key filtering traits such as thermal tolerance, larval starvation resistance, and predator avoidance strategies for many marine species in these systems have explained shifts in distribution and abundances of range expanders (Text Box 1; Figure 8). Additionally, the warming of tropical systems will likely lead to transitions to new ecological states not observed in living human time scales. These systems represent a need for generalisable, and trait-based predictive tools to forecast ecological outcomes beyond recent ecological states. We suggest that focusing research effort in this field on incorporating suites of traits into spatially-explicit process models of transition-zone ecosystems are likely to yield the most fruitful tests and validations of trait-based global change predictions.