You are here: Home Results Task 5.5.2. Testing model by back projection

Task 5.5.2. Testing model by back projection

Task lead: Antoine Guisan, Université de Lausanne, CH

We tested model transferability in time for six tree species in Europe across six Millennia (Pearman et al. 2008) and for one frog species across 50 years (Ficetola et al. 2010), providing good support for the possibility of projecting species distribution in the past. Moreover, further work was required to expand back projection testing across a larger spatio-temporal scale (13 Millennia over the entire European continent) with the aim of gaining insight on future projections of species distributions.

Past climate reconstructions, together with the necessary data on fossil pollen as well as macro-fossils remains were used to test an approach that is relatively new for SDMs. Traditionally SDMs are built considering only the current species distribution, and theoretically they should underestimate the potential distribution when projected in time and/or space. A multi-temporal model calibration approach has been suggested as an option to solve this problem. We considered presences for Picea abies, Abies alba and Fagus sylvatica and 6 climatic variables covering from 13,000 to 1,000 years BP. Using the most updated protocols available we measured the contribution of each 1000 years time step to the total niche of each species and we tested if the partial niches (i.e. those obtained considering only one time frame) are more similar to the total niche (i.e. the niche obtained considering all time frames) than random. Using an ensemble forecasting approach, we calibrated a SDM for each time frame and for the pooled database. We projected each model onto the current climate and we evaluated the results against the current species distribution. We also projected all models into the future. Niche similarity was almost always significant, and it increased through time. The SDMs calibrated considering a single time frame gave very different results when projected over the current climate, providing the evidence for a change in the species’ realized niches. Moreover, they predicted limited climate suitability when compared to the total SDMs. The same results were obtained with projections into future. Basically we demonstrated that the realized climatic niche of the species considered provided differing results in current and future projections of SDMs. Thus, building the niche through time represents a way forward towards a better understanding of species distribution and ecology in a changing climate.


  • Ficetola, G. F., Maiorano, L., Falcucci, A., Dendoncker, N., Boitani, L., Padoa-Schioppa, E., Miaud, C. & Thuiller, W. (2010) Knowing the past to predict the future: land-use change and the distribution of invasive bullfrogs. Global Change Biology, 16, 528-537.
  • Maiorano, L., Guisan, A., Araujo, M. B., Barbet-Massin, M., Broennimann, O., Dubuis, A., Dullinger, S., Edwards, T. C., Engler, R., Pearman, P., Pellissier, L., Petitpierre, B., Pio, D., Pottier, J., Randin, C., Thuiller, W., Yoccoz, N. G. & Zimmermann, N. E. (In prep) Evaluating Species Distribution Models: Why, What and How.
  • Pearman, P. B., Randin, C. F., Broennimann, O., Vittoz, P., Van Der Knaap, W. O., Engler, R., Le Lay, G., Zimmermann, N. E. & Guisan, A. (2008) Prediction of plant species distributions across six millennia. Ecology Letters, 11, 357-369.
  • Maiorano L., Cheddadi R., Zimmerman N.E., Pellissier L., Petitpierre B., Pottier J., Laborde H., Hurdu B.I., Pearman P.B., Psomas A., Singarayer J.S., Broennimann O., Vittoz P., Dubuis A., Edwards M.E., Binney H.A. and A. Guisan. Building the niche through time: using 13,000 years of data to predict the effects of climate change on three tree species in Europe. Global Ecology and Biogeography, in evaluation.
Document Actions