******************************** README ******************************
Radinger J, Wolter C, Kail J Data from: Spatial Scaling of Environmental Variables Improves Species-Habitat Models of Fishes in a Small, Lowland Sand-bed River.


    							 SEPT 2015.

THESE DATA ACCOMPANIES:

Radinger J, Wolter C, Kail J. Spatial Scaling of Environmental Variables Improves Species-Habitat Models of Fishes in a Small, Lowland Sand-bed River. PLOS ONE.

Please, contact me if you have suggestions, find errors, inconsistencies, or
any other bug in the file. As well, please let me know about your uses of this
data and send manuscripts and reprints when available. I'll be happy to help
you in any case, as far as I can.


********************************** CONTENTS ***********************************
1. README.txt - This file. Including a description of the results, dataframes and variables.

The following 4 data files (zip files) provide model results for the respective predictor dataset. The results include a global BRT model (brt.model.global, including all variables) and a final BRT model (brt.model.final, including only statistically relevant variables) for each species. Furthermore, summarizing statistics (brt.stats.final) for each final model (e.g. cross-validated AUC) and associated plots showing the influence of selected selected variables (species_response.pdf) are provided. Models are stored as R objects in the *.rds format and can be loaded with the R command readRDS().

2a. Set_AV - Results for single boosted regression tree (BRT) models for 13 fish species, 4 modelled distance classes (0, 200, 2500, 4000 m) and assessed hydromorphological data (AV) without topological variables.

2b. Set_AV_TV - Results for single boosted regression tree (BRT) models for 13 fish species, 4 modelled distance classes (0, 200, 2500, 4000 m) and assessed hydromorphological data (AV) with topological variables (stream orders, distance from mouth).

2c. Set_MV - Results for single boosted regression tree (BRT) models for 13 fish species, 4 modelled distance classes (0, 200, 2500, 4000 m) and measured hydromorphological data (MV) without topological variables (stream orders, distance from mouth). 

2d. Set_MV_TV - Results for single boosted regression tree (BRT) models for 13 fish species, 4 modelled distance classes (0, 200, 2500, 4000 m) and measured hydromorphological data (MV) with topological variables.

3. GRASS GIS Scripts - 2 scripts for (i) transforming vector based hydromorphological data into GRASS raster format and (ii) calculating distance based average focal predictors for 4 predefined distance classes (0, 200, 2500, 4000 m) using the GRASS rdfilter add-on.

4. R scripts - Scripts for (i) calculating the boosted regression tree models (BRT, script 1), (ii) analyzing (2a) and plotting (2b) the model performance based on AUC and (iii) analyzing and plotting the contribution and rank of single variables (3a-3c).

5. Model performance (AUC) - R-dataframe storing the model performance (AUC) of the single models for 13 species, four modelled distance classes, assessed vs. measured variables and with/without topological variables. The dataframe is stored as R object in the *.rds format and can be loaded with the R command readRDS().

6. Variable contribution - R-dataframe storing the contribution of single variables for each of the final BRT models. The dataframe is stored as R object in the *.rds format and can be loaded with the R command readRDS().

7. Variable contribution rank - R-dataframe storing the importance rank of single variables for each of the final BRT models. The dataframe is stored as R object in the *.rds format and can be loaded with the R command readRDS().

********************************* CONTACTING **********************************
Johannes Radinger
Leibniz-Institute of Freshwater Ecology and Inland Fisheries
Müggelseedamm 310
12587 Berlin
Germany

Tel.: +49 30 641 81 797
Fax: +49 30 641 81 750     
e-mail: jradinger@igb-berlin.de

***************************** ACKNOWLEDGEMENTS ********************************
Fish and hydromorphological data were kindly provided by the LLUR Schleswig-Holstein. This study is part of the IWRM-net project “IMPACT” and has been funded by the German Federal Ministry for Education and Research (grant number 02WM1134).