Belowground biodiversity has been linked to many ecosystem functions and services related again to human health and wellbeing, but faces threats similar to the aboveground system. Acting to halt biodiversity loss requires transformative change and effective conservation of all ecosystems. Indeed, soil biodiversity has been neglected in several global biodiversity assessments and conservation actions. To better understand how and why soil biodiversity and related ecosystem functions have lesser priority for nature conservation, it is important to get an overview of how society, and particularly policy-makers, have addressed it in the past and present.
With this analysis, we aimed to review the current status of soil protection by investigating the effect of current conservation efforts in Europe as represented by nature conservation areas.
Analysis are done in:
- R version 3.6.3
- RStudio version 1.2.5033
- ArcMap version 10.7.1
Attached R packages:
- dplyr_1.0.2
- ggplot2_3.3.3
- ggpubr_0.4.0
- MBESS_4.8.0
- psych_2.0.12
- raster_3.4-5
- reshape2_1.4.4
- sp_1.4-5
If R and ArcGIS are already installed, you only have to download this repository. To run the same analysis, you need to request the dataset Full_LUCAS2018-iDiv_dataset_02072021.csv from the authors. However, you may also adapt the code to your own dataset.
After installing and loading all packages necessary, you can start with the actual analysis (see Instructions).
To run the analysis, you have to open the script 0_define_PA_nonPA.R.
Typical install time on a "normal" desktop computer: ~5min
To see if the installation works, it is recommended to run the analysis on a subset of randomized runs (i.e., 10 or 100 instead of 1000) first. This can be done by replacing if(times > 1000) with if(times > 10) or something similar.
To fix error messages among others, you can try to run the code within loops line by line with i (or j or k) defined by hand according to the necessary type of object.
The output should be saved according to the R script. The expected run time for demo on a "normal" desktop computer depends on the subset of data used, but should not exceed 2h.
The analysis includes the following steps, which are also descriped in the script 0_define_PA_nonPA.R:
Take the dataset Full_LUCAS2018-iDiv_dataset_02072021.csv and define a new column "PA" that has "0" as values. This column will have 0 for non-protected, and 1 for protected sites. To fill the column, you need to download the shapefile of the Natura 2000 nature conservation areas here. If you added this shapefile to your ArcGIS project, use the Select by location function in ArcGIS to filter for protected areas. Change the value of the "PA" column of the selected sites to "1". Finally, save the information about the points and their classification (i.e., ID and PA column) to LUCAS_points_withPA.txt.
Run the script 0_define_PA_nonPA.R.
- Add PA column to the LUCAS data set
- Define functions, environmental variables, land cover types, and protected areas.
- Create empty R objects to be filled during the loop.
- Repeat the pairing of sites & the statistical test for their differences 1000 times.
- Merge the output of all the 1000 runs into one table per estimated parameter (i.e., p-value, effect size, and confidence intervalls).
- Plot the estimated parameters subdivided into the three (four) land use cover types.
To reproduce the analysis right from the beginning, you can to define the protected sites (e.g., sites within nature conservation areas) by your own (see Instructions).
Associated publication: [1] Zeiss et al. (2021): Challenges of and opportunities for protecting European soil biodiversity. Conservation Biology. https://doi.org/10.1111/cobi.13930
Related repository: JeMaNd_r/ManagementPlans