Dispersal is a key process for survival of patchy populations. It is predominantly described as a fixed trait, typically measured in ectotherms under optimal climatic conditions (i.e. sunny weather, high temperature, few or no wind). However, it is now hypothesized that dispersal should be considered as a plastic trait, depending on environmental conditions.
Optimal climatic conditions are not predominant in temperate regions, and ectotherms like butterflies are known to conserve movement ability under sub-optimal conditions.
Functional connectivity is the way landscape properties affect the behavior of an organism dispersing between habitat patches. We make the hypothesis that functional connectivity varies according to climatic conditions.
The objective of this work is twofold: (1)  investigate if and how the dispersal behavior of a butterfly (Maniola jurtina) varies with climatic conditions. 480 individual trajectories were recorded in an agricultural matrix under varying climatic conditions, and characterised by the probability of leaving a patch, wether individuals adopted a short- or long-distance flight strategy, the distance reached, length and sinuosity of the paths. (2) Simulate dispersal behavior to investigate the response of landscape functional connectivity to spatio-temporal changes in climatic conditions.

Results/Conclusions

Dispersal attempts were observed on a range of sub-optimal to optimal climatic conditions (between 15-28°C, 15-100% sunshine cover and 4-18m/s wind speed).
Probability of dispersal attempts increased by 20% to 40%  with increasing temperature, sunshine, and decreasing wind speed. Dispersal strategy shaped for long-distance was more sensitive to climatic conditions than movements shaped for local dispersal. Trajectories were longer, straighter, and went farther from habitat patch with better climatic conditions. Unexpectedly, female dispersal behavior occurred at lower temperatures than male’s, which may be explained by avoidance of male harassment.
These results emphasise that dispersal estimates based on optimal climatic conditions may be far overestimated, and that plasticity induced by varying climatic conditions should not be neglected. We predict that spatio-temporal variations of climatic conditions have a strong influence on landscapes functional connectivity by mediating structural connectivity and species dispersal ability and behavior.
This hypothesis is tested by using an individual-based dispersal model developped for Maniola jurtina. It is applied to scenarios of virtual “climatic landscapes” representing the spatial variations of climatic conditions caused by the different land cover types. Model parameters of dispersal occurrence and efficiency are adapted to local climate conditions according to the results described above.