نبذة مختصرة : International audience ; If individuals have spatially dierentiated preferences for sites or areas im- 8 pacted by an invasive alien species, eective management must take this 9 heterogeneity into account and target sites or areas accordingly. In this 10 paper, we estimate spatially dierentiated preferences for the management 11 of primrose willow (Ludwigia grandiora), an invasive weed spreading in a 12 French regional park. We use an original spatially explicit discrete choice 13 experiment to evaluate individuals' willingness to pay (WTP) to control the 14 invasion in dierent areas of the regional park. Our results indicate that 15 WTP for management highly depends on the area considered, with areas 16 where it is three times higher than others. We analyze the main factors 17 explaining the heterogeneity of preferences and show that the closer respo n-18 dents live to the park, the more they visit and/or practice activities in it, the 19 higher their WTP and spatial preferences. Park residents and regular users 20 have highWTP and unambiguous preferences for targeting control to specic 21 areas. Non-residents and occasional users have much lower WTP and more 22 homogeneous spatial preferences. These results suggest that implementing 23 management strategies that spatially target invasion control according to 24 public preferences is likely to produce signicant utility gains. These gains 25 are all the more important as the preferences taken into account are those of the stakeholders directly concerned by the invasion, the residents and reg-27 ular park users. Ignoring these spatial preferences will lead to sub-optimal 28 invasion management. ; Lorsque les individus ont des préférences spatialement différenciée sur les sites impactés par une espèce invasive, les décideurs doivent prendre en compte cette hétérogénéité spatiale et cibler la préservation en conséquence. Nous estimons ici les préférences, de manière spatialement différenciée, pour la lutte contre la Jussie (Ludwigia grandiflora), une plante ...
No Comments.