Effects of landscape context and seasonality on time to recovery

Background/Question/Methods

Populations of many species are stressed due to a plethora of anthropogenic activities. To ensure the persistence of their populations, the correct assessment of the time required for a population to recover from stress is of pivotal importance for the field of ecosystem management. The sole focus on the life-history of the species, and in particular its intrinsic population growth rate r, gives an optimistic predictor of recovery times. In a most simple model, the time an exponentially growing population needs to return to a fraction of the size it had before disturbance is a simple linear function of 1/r.

Recovery times obtained from models incorporating more ecological realism, like spatially-explicit individual-based models (IBM), are not that straightforward to predict. When introducing space, recovery times become dependent on the landscape context, in particular on the choice of spatial scales for modelling population dynamics, for exposure to stress, and for observing recovery. Under temporally varying conditions, recovery times will in particular depend on the choice of constant versus seasonally varying conditions.

In simple IBMs for aquatic and terrestrial species with different life-history (mobility,  dispersal mechanism and generation times), we investigate the impact of spatial dimensions and of seasonality of the environment on recovery times. The role of spatial dimensions is analysed in systems with continuous habitat, of which only a part is exposed to lethal stress, and in which recovery can be measured in sub-systems ranging in size from the exposed part to the whole system. The role of seasonality is analysed by comparing recovery for species with different generation times, in non-seasonal systems (with continuous reproduction) and in seasonal systems, in which seasonality leads to some synchronization (reproduction peaks). 

Results/Conclusions

We explain the underlying mechanisms, identify whether specific choices should be considered worst-case or not, and deduce rules of thumb based on species traits, that aid in selecting an appropriate worst-case setup for model studies at similar scales.