Services

• Interested in the effect of climate and habitat change on your study species?

• Want to know which species will persist / may land in your protected area in the future?

• Need a predictive map to find new populations?

• Need some solid quantitative analysis?

• Need to write an article?

     Here is a sample of what I offer:

Data analysis, detection of patterns, hypothesis testing

I am familiar with large and small datasets. I can perform explanatory analyses (e.g. identifying the drivers of species response to past environmental change) as well as predictive analyses (e.g. predicting species response to future climate change). I can assess temporal trends (either linear or non-linear).

My approaches include the following techniques

• Parametric modelling and variance partitioning (GLM, GLMM, GAM, GAMM, MCMCglmm) 

• Machine learning (MaxEnt, Random Forest down-sampling, XGBoost)

• Cluster analyses (PCA, network analyses)

• Meta-analyses (MCMCglmm)

Mapping predictions of species distribution / future projections

I am specialised in Ecological Niche Modelling (ENM, also erroneously known as Species Distribution Modelling). I use the latest recommended standards for model design. Correlations will always work, which does not mean that they make sense. My methodology involves close interaction with experts to better make the link between data and reality, select environmental predictors with a causal relationship with the species, take into account the socio-economic contexts, and better interpret model output.

I take into account multiple sources of climate data (CHELSA and WORLDCLIM) and presence data (opportunistic points, dedicated standard surveys, polygon expert ranges...). I use the best two algorithms available (Random Forest down-sampled and Maxent) as identified by Valavi et al. (2021).

With all these replicates, conservationists might get lost. Which map to chose for my conservation planning? I provide a unique map, showing the average output, penalised by uncertainty (i.e. the level of disagreement between replicates). This provides a practical tool for practitioners, showing maps of suitable environments for which all models are in agreement. The final map identifies priority areas for intervention in a robust fashion. See below an example of consensus priority map published in Animal Conservation.