Species Distribution Modeling using R (SDMR02)
27 July 2020 - 30 July 2020£220.00 – £490.00
If you are interested in gaining the introductory knowledge required to work with SDMs, whether you be a student, postdoc, or practicing scientist, this course is for you.This four-day course will provide participants with the background knowledge and skills needed to get started in the use of species distribution models (SDMs) for applied and basic research. The course will focus on (1) the preparation of required spatial datasets (biological observations and environmental predictors); (2) practical considerations in the development, application, and interpretation of SDMs; and (3) fitting and evaluating SDMs using different statistical approaches – all using R.
Using a combination of lectures, coding exercises in R, and case studies, participants will learn to:
- Understand background theory and model assumptions
- Identify, manipulate and prepare spatial datasets for SDMs
- Fit, interpret, and evaluate SDMs using several statistical methods (e.g., Maxent, Mahalanobis distance, generalized linear models, boosted regression trees)
- Project SDMs to predict climate change impacts, etc.
The course is entirely R-based and while techniques of working with spatial data in R will be covered in detail, prior experience with R is highly recommended. If you are new to R, this course will be of most use to you if you work through a few tutorials to understand the basics of R programming before the start of the course. Students are highly encouraged to bring their own data sets, but this is not required for participation.
Course material will be presented by Matt Fitzpatrick who has published broadly in the use of SDMs for applied and basic science.
Any researchers (PhD and MSc students, postdocs, primary investigators) and management and environmental professionals in government and industry interested in the use of SDMs for conservation, biogeography, spatial ecology, or related disciplines.
Venue – PR statistics head office, 53 Morrison Street, Glasgow, G5 8LB – Google map
Availability – 20 places
Duration – 4 days
Contact hours – Approx. 42 hours
ECT’s – Equal to 3 ECT’s
Language – English
Arrival Sunday the 26th July (between 17:00-21:00) and departure Thursday 30th July (accommodation must be vacated by 09:15am).
To book ‘COURSE ONLY’ with the option to add the additional ‘ACCOMMODATION PACKAGE’ please scroll to the bottom of this page.
Other payment options are available please email firstname.lastname@example.org
PLEASE READ – CANCELLATION POLICY: Cancellations are accepted up to 28 days before the course start date subject to a 25% cancellation fee. Cancellations later than this may be considered, contact email@example.com. Failure to attend will result in the full cost of the course being charged. In the unfortunate event that a course is cancelled due to unforeseen circumstances a full refund of the course fees (and accommodation fees if booked through PR statistics) will be credited. However, PR statistics will not be held responsible/liable for any travel fees, accommodation costs or other expenses incurred to you as a result of the cancellation. Because of this PR statistics strongly recommends any travel and accommodation that is booked by you or your institute is refundable/flexible and to delay booking your travel and accommodation as close the course start date as economical viable.
There will be a combination of lectures and hands-on practicals. Afternoon practicals will be based on the topics covered in the morning lectures. Data sets for computer practicals will be provided, but participants are highly encouraged to bring their own data.
Assumed quantitative knowledge
Familiarity with GIS and geospatial data (i.e., rasters and point occurrence data). A basic understanding of statistical modeling concepts and inference, including regression methods and model validation.
Assumed computer background
Basic proficiency with R, including an ability to import/export and manipulate tabular data, fit basic statistical models, and generate simple exploratory and diagnostic plots.
Equipment and software requirements
A laptop/personal computer with a working version of R and RStudio and the following packages (and dependencies) installed and tested for proper functionality:
R and RStudio are supported by both PC and MAC and can be downloaded for free by following these links:
R – https://cran.r-project.org/
RStudio – https://www.rstudio.com/products/rstudio/download/
UNSURE ABOUT SUITABLILITY THEN PLEASE ASK firstname.lastname@example.org
Meet at Flat 2/1, 43 Cook Street, Glasgow G5 8JN between 17:00-21:00
Monday 27th – Classes from 09:30 to 17:30
1) Overview on modeling and mapping species distributions: Theory, Data, Applications
2) Key steps and concepts in developing SDMs
3) Theory of niches, species distributions, and model assumptions / uncertainties
– Range equilibrium
– Niche conservatism
– Sample size & bias
– Correlation of predictor variables
– Defining the study area
– Model thresholds, validation, and projections
4) Applications of SDMs
5) Data for SDMs
– Biological data
– Predictor variables
6) Practical: Working with spatial data in R
Tuesday 28th – Classes from 09:30 to 17:30
Methods for fitting SDMs I – Overview
1) Overview of methods for fittings SDMs
2) Presence-absence vs. presence-only
– Machine Learning
– Boosting & Bagging
– Maximum entropy / point-process
3) Overview of R packages for SDM
4) Variable selection
5) Practical: Getting your data ready for SDMs
Wednesday 29th – Classes from 09:30 to 17:30
Methods for fitting SDMs II – Presence-absence modeling
1) GLMs and GAMs
2) How to evaluate models
3) Model discrimination
4) Model calibration
5) Model complexity / simplicity
6) Boosted regression trees
7) Practical: Fitting presence-absence SDMs using ‘dismo’ and ‘biomod2’
Thursday 30th – Classes from 09:30 to 17:30
Methods for fitting SDMs III – Presence-only / background modeling, Projecting SDMs
1) Creating background data
3) Evaluating presence-only models
4) Dealing with biases species data
5) Projecting / extrapolating models
6) Working with climate change data
7) Practical: Fitting maxent models using R
8) Practical: Projecting models to new places / times