I'm trying to model the future potential distribution of an invasive species in Europe. Its native range is in North and South America but has recently spread to Asia, Australia, and parts of Africa. I am wondering what approach is better, using only the distribution data from the native range or the whole distribution of the species? Several papers suggest that depending on the choice, the resulting models could differ greatly, partially due to the niche shift that is common in invasive species.
I would appreciate any insights.
I am currently on almost the same topic. In my approach, I performed two models one in the global range (1) and one only in the invaded range (2). Results are different in fact. I think the most accepted approach is to use data from the native and invaded ranges together. However, I suggest you try the approach (1) and (2) and then check the response curves of the model, to see what the models are doing. You can also consider a niche shift approach (Broennimann), to check which model violates fewer assumptions. Hope it helps, Luigi.
If you go measure the dispersion of the specie in the past you use the original distribution but you go predict the future dispersion you use the actual distribution.
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude