Event Date
Monday, November 25th, 2019
Course Format
Pre Recorded
About This Course
Range modeling, also known as niche, distribution, or habitat use modeling, is among the most widely used tools in ecology. Its popularity is evidenced by the wide range of conceptual and statistical approaches that have been developed over the last two decades for quantifying ranges. We will spend much of our time on the most popular approach to range modeling – presence-only models – however we will draw connections to other data types and general principles for any range model throughout. We will emphasize understanding how modeling decisions affect predictions and identify whether biology or statistics can provide insights into those decisions. By surveying a broad spectrum of approaches, you’ll learn how to find the right tool for the job and understand the pros and cons of each. Each day we will reserve time for open work sessions where students can receive mentoring while applying new skills to their own data sets or example data sets provided by the instructors.
Intended Audiences
Researchers interested in advancing their range modeling and statistical skills with a comprehensive survey of available tools in biogeography and habit use.
Course Details
Last Up-Dated 25:10:2019
Duration – Approx. 35 hours
ECT’s – Equal to 3 ECT’s
Language – English
Teaching Format
There will be morning lectures based on the modules outlined in the course timetable. In the afternoon there will be practicals based on the topics covered that morning. Data sets for computer practicals will be provided by the instructors, but participants are welcome to bring their own data.
Assumed quantitative knowledge
Familiarity with one or more common approaches to modeling distributions is expected; these might include using Maxent, biomod2, or GLMs. Some familiarity iwth Biasyesian principles will be helpful, although we’ll include a brief refresher.
Assumed computer background
Experience with R, including regressions, graphics, and manipulating data frames. This experience often corresponds to one or more years of using R regularly.
Equipment and software requirements
A laptop computer with a working version of R or RStudio is required. R and RStudio are both available as free and open source software for PCs, Macs, and Linux computers. R may be downloaded by following the links here https://www.r-project.org/. RStudio may be downloaded by following the links here: https://www.rstudio.com/.
All the R packages that we will use in this course will be possible to download and install during the workshop itself as and when they are needed, and a full list of required packages will be made available to all attendees prior to the course.
A working webcam is desirable for enhanced interactivity during the live sessions, we encourage attendees to keep their cameras on during live zoom sessions.
Although not strictly required, using a large monitor or preferably even a second monitor will improve he learning experience
PLEASE READ – CANCELLATION POLICY
Cancellations/refunds are accepted as long as the course materials have not been accessed,.
There is a 20% cancellation fee to cover administration and possible bank fess.
If you need to discuss cancelling please contact oliverhooker@prstatistics.com.
If you are unsure about course suitability, please get in touch by email to find out more oliverhooker@prstatistics.com
COURSE PROGRAMME
Day 1 – Approx. 7 hours
Day 1 begins with a background on the most commons information used for range modeling, exploratory analysis, and identifying the similaries, differences, and subtleties of each.
1) Data types and their strenths and weaknesses
2) Roles for simplicity and complexity in model design
3) Presence-only models: Connections between Maxent, Point process models and GLMs
4) Coding tips and designing a reproducible workflow
Day 2 – Approx. 7 hours
Day 2, As models using only presence data are by far the most common, webll spend some time investigating the subtleties of different modeling challenges, and the connections between modeling decisions and predictions.
5) Advances in presence-only modeling
6) Modeling sampling bias – the key to robust obtaining models
7) Designing block cross validation
8) A wide range of options for evaluating model performance
9) Thresholding predictions – do you have to?
Day 3 – Approx. 7 hours
Day 3 will focus on advanced challenges with presence-only models.
10) Extrapolation and transfering predictions to new times or locations
11) Integrating presence-only data with other data types
12) Spatial prior information – expert maps, dispersal models, related species
13) Spatially explicit models
Day 4 – Approx. 7 hours
Day 4 will focus on different modeling approaches that are appropriate with different data types. * Occupancy models and detection bias.
14) Emerging algorithms, especially for small sample sizes
15) Machine learning algorithms what’s under the hood, and does it matter?
Day 5 – Approx. 7 hours
Day 5 will conclude with a survey of the next generation of range modeling tools including data fusion approaches and provide ample time for students to receive advice while working with their own data sets.
16) Data fusion models
17) Joint distribution models
18) Masking distributions – when do we really need statistics?
Course Instructor
Dr. Antoine Becker-Scarpitta
Works at – University of Helsink
Teaches – Multivariate analysis of ecological communities in R with the VEGAN package (VGNR03)
Antoine is a plant community ecologist working as a postdoctoral researcher at the University of Helsinki and as a postdoctoral fellow at the Institute of Botany of the Academy of the Czech Republic. Antoine holds a degree in Conservation Biology from the University of Paris-Sud-Orsay, and from the Natural History Museum of Paris, he obtained his PhD in Biology/Ecology from the University of Sherbrooke (Canada). Antoine’s research focuses on the temporal dynamics of biodiversity with a particular focus on the forest and Arctic vegetation. Antoine has taught community ecology, plant ecology and evolution, linear and multivariate statistics assisted on R.