Search

Disturbances' role in shaping communities is well documented but highly disputed. We suggest replacing the overused two-trait trade-off approach with a functional group scheme, constructed from combinations of four key traits that represent four classes of species' responses to disturbances. Using model results and field observations from sites affected by two highly different disturbances, we demonstrated that popular dichotomous trade-offs are not sufficient to explain community dynamics, even if some emerge under certain conditions. Without disturbances, competition was only sufficient to predict species survival but not relative success, which required some escape mechanism (e.g., long-term dormancy). With highly predictable and large-scale disturbances, successful species showed a combination of high individual tolerance to disturbance and, more surprisingly, high competitive ability. When disturbances were less predictable, high individual tolerance and long-term seed dormancy were favored, due to higher environmental uncertainty. Our study demonstrates that theories relying on a small number of predefined trade-offs among traits (e.g., competition-colonization trade-off) may lead to unrealistic results. We suggest that the understanding of disturbance-community relationships can be significantly improved by employing sets of relevant trait assemblies instead of the currently common approach in which trade-offs are assumed in advance. [ABSTRACT FROM AUTHOR]

Bet hedging is a means to increase fitness in environments that vary unpredictably in space and time. In such environments, models predict a trade-off between the bet-hedging strategies dispersal and dormancy, while the increasing importance of risk reduction with decreasing predictability should lead to an increase in dispersal and dormancy along gradients of environmental predictability. However, so far there has been no experimental study to test these predictions in the field. Here, we used a set of novel field experiments that enabled us to quantify and separate seedling recruitment from three sources: local reproduction, dormancy, and dispersal. The study included the entire plant community from five environments differing considerably in predictability. Evidence for both the existence of a trade-off between dispersal and dormancy within environments and their increased use in unpredictable environments was very weak. The importance of dispersal for population and community dynamics in our system was extremely low relative to dormancy and local reproduction. This indicates that the role of dispersal for buffering environmental variation may be negligible compared with other risk-reducing strategies. Our findings highlight the urgent need for multispecies and multisite experiments in empirical tests of theoretical predictions. [ABSTRACT FROM AUTHOR]