Search

The data and R code provided are part of a publication (currently under review). Once published, material and methods from the paper will be available open access. The files are: RData files: - winglength_modeling_df: a list of two ($data: containing all the response and predictor variables needed for the modeling / $phy: the phylegenetic covariance matrix based on a consensus tree of 1000 phylegentic trees, downloaded fromhttp://vertlife.org/phylosubsets/ for our 24 species). Note that the df has been been cleaned and filtered after receiving the raw data from the Finnish and Swedish Natural History Museums. (Cleaning and filtering process are described in material and methods). - bodyweight_modeling_df:a list of two ($data: containing all the response and predictor variables needed for the modeling / $phy: the phylegenetic covariance matrix based on a consensus tree of 1000 phylegentic trees, downloaded fromhttp://vertlife.org/phylosubsets/ for our 24 species). Note that the df has been been cleaned and filtered after receiving the raw data from the Finnish and Swedish Natural History Museums. (Cleaning and filtering process are described in material and methods). This df is smaller than the winglength df, as the raw data contained many NAs for body weight and it has further been filtered to only include daytime (7am-5pm) body mass measurements. R scripts: - bodymass_pglmm_models & winglength_pglmm_models: pglmm modeling process for both response variables following the same protocol - Plots_pglmm: code for generating the forest plots based on model 1 & 2 - plots_singlespecies_regressions: code for generating the four plots of the four example species (based on lmer models but following the results from pglmm model 1 & 2)&nbsp

It contains describtive plot for data, supplementary methods and results

Additional file 6: Figure S5. Partial dependency plots for (a) I. ricinus and (b) I. persulcatus solely based on environmental data.

Additional file 4: Figure S4. The relative contributions of the explanatory variables in the data set of (a) host only, (b) environment only based on the mean ensemble model.

Additional file 7: Figure S6. Partial dependency plots for (a) I. ricinus and (b) I. persulcatus solely based on host data.

Additional file 5: Table S1. The number of times each model contributed to the final ensemble in different data sets.

Additional file 3: Figure S3. The range (lines) and mean (dots) of model performances over 50 model runs in each model algorithm estimating habitat suitabilities for I. persulcatus in different variable compositions: (a) environmental only, (b) host only, (c) environmental and host, and (d) environmental, host, and suitability for I. ricinus.

Additional file 8: Figure S7. Partial dependency plots for (a) I. ricinus and (b) I. persulcatus based on combined host and environmental data, and habitat suitability data for the other tick species.

Additional file 2: Figure S2. The range (lines) and mean (dots) of model performances over 50 model runs in each model algorithm estimating habitat suitabilities for I. ricinus in different variable compositions: (a) environmental only, (b) host only, (c) environmental and host, and (d) environmental, host, and suitability for I. ricinus.

Additional file 1: Figure S1. (a) The sampling strategy for new collections in 2021 was created based on the following criteria. Subdivisions of landscape areas (Area1–Area4), CORINE land cover 2018, a 5-km buffer around existing I. persulcatus occurrences (grey circles), and a 500-m buffer around roads were used to delimit the four sampling areas (light grey lines). For each sampling area, a random sample of 25 collection locations was created depending on the relative shares of forest and meadow categories in each area. (b) The map showing the 2021 results indicates the locations where I. ricinus was found with B. burgdorferi (s.l.)-positive locations.