You will use species distribution models (SDMs) and maps to model and describe the current distribution of a plant species occurring in Madagascar. You will identify and explain factors that might explain the survival of that species and whether the current protected areas system is adequate or not to conserve areas where this species exists.
You will need to:
1. Download the R project, scripts, data and html tutorials from GitHub. Follow the README guidelines to download all the GitHub files here https://github.com/jdmwhite/2022_MSc_Spatial_Course
0yzc/edit?usp=sharinghttps://.com
3. Download species occurrence from GBIF using the R script (1_Species_localities.R).
4. Clean species occurrence records and model their distribution (using R scripts and R markdown provided during the practical session) (1_Species_localities.R and
2_Species_distribution_model.R and associated .Rmd files). Add
5. Incorporate external information to support/complement your results
6. Write up your project report using the structured template provided in R Markdown (3_Student_Report_Template.Rmd), which will be explained in the first afternoon of the module
Project report template
SUMMARY /ABSTRACT (150-200 WORDS)
Question / hypothesis / problem / Malagasy species distribution
Data and method used
Main result / implication / importance / interest / uniqueness
INTRODUCTION (200-300 WORDS)
Malagasy species: information / interest / uses / IUCN rating / why is it important to map this species’ distribution?
METHODS (400-700 WORDS)
This will be assessed on the justification of the parameters chosen for cleaning and modelling the species
a. Pre-processing steps
Data Sourcing and cleaning: GBIF, Coordinate Cleaner, Filtering parameters Justify selection of parameters.
b. Processing steps
Justify selection of spatial thinning thresholds and pseudo-absences/background points (consequences on model evaluation and predictions).
Justify selection of environmental variables and correlation thresholds used. Justify selection of training vs. testing groups.
c. Post-processing steps
Compare and justify different criteria to evaluate the model and thresholds used to convert probability maps into presence-absence maps
Note: there are specific questions and prompts in the R Markdown Template, which will help guide some of your writing for the methods, results and discussion.
RESULTS (FIGURES / MAPS / DESCRIPTION OF RESULTS) (400-700 WORDS) https://.com
1. GBIF data distribution 2. Malagasy modelled species distribution (probability and binary presence-absence output). Compare model outputs: variable importance, model performance, response curves. Add
CONCLUSION / DISCUSSION (400-700 WORDS)
1. Species distribution
2. Would the species benefit from further protection based on its distribution and life history?
3. Are there related topics that this research feeds into?
4. What are the wider subjects that this research could help investigation?
5. How would you research this further beyond the scope of your current analysis?
Sample Literature / Resources:
GBIF
https://www.gbif.org/
https://www.gbif.org/resource/search?contentType=literature
Worldclim https://www.worldclim.org/data/bioclim.html
https://rmets.onlinelibrary.wiley.com/doi/full/10.1002/joc.5086
SDMs
Is my species distribution model fit for purpose? – https://onlinelibrary.wiley.com/doi/full/10.1111/geb.12268
The effect of sample size and species characteristics on performance of different species distribution modeling methods
https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.0906-7590.2006.04700.x CoordinateCleaner: Standardized cleaning of occurrence records from biological collection databases
https://doi.org/10.1111/2041-210X.13152
Spatial thinning and class balancing: Key choices lead to variation in the performance of species distribution models with citizen science data https://doi.org/10.1111/2041-210X.13525
Sample selection bias and presence-only distribution models: implications for background and pseudo-absence data https://doi.org/10.1890/07-2153.1
WDPA https://www.protectedplanet.net/en/thematic-areas/wdpa?tab=WDPA Word limit: Each section has a word limit that you must adhere to. Word counting does not include the code, only the text.
Mark (%) Criteria
85 – 100 Exceptional project report showing very broad understanding of the project area and outstanding knowledge of the relevant literature. Exemplary presentation and analysis of results, logical organisation and ability to evaluate critically and discuss results with insight and originality.
70 – 84 An excellent project report showing evidence of wide reading, with clear presentation and thorough analysis of results and an ability to evaluate critically and discuss research findings. Clear indication of insight, understanding and originality. An extremely competent and well-presented report overall, excellent in most aspects.
60 – 69 A good project report which shows a clear understanding of the problem and sound knowledge of the relevant literature. Sound presentation and analysis but perhaps not exploiting the results to the full. Relevant interpretation and critical evaluation of results, though with some limitations regarding the scope. Good general standard of presentation and organization.
50 – 59 A satisfactory project report which shows some understanding of the problem but limited knowledge and appreciation of the relevant literature. Presentation, analysis
and interpretation of the results at a basic level and showing little originality or critical evaluation. Some weaknesses in the organization and presentation of the report.
https://.com
Add
