Conference Themes

Session Themes SOTM 2023

1. Detection, attribution, & prediction under species on the move
2. Leveraging bioinformatics, big data science, and emerging technologies to study species on the move
3. Understanding ecological and evolutionary mechanisms facilitating or hindering range shifts
4. Impacts of climate change on community structure and patterns of taxonomic, functional, and phylogenetic diversity
5. Ecological responses to climate change from micro to macro scales
6. Invasion risk and range redistribution of non-native species under climate change
7. Conservation paradigms & management strategies for a moving and (dis)connected future
8. Governance, legal, and ethical issues for shifting species and changing ecosystems
9. Cultural, social, and economic dimensions of changes in species distributions
10. Indigenous knowledge and species on the move
11. OneHealth – Implications of species on the move for human and animal health
12. Ecosystem and biome shifts, including geochemical impacts from species on the move

Sunday May 14, 2023
9 am - 4 pm

USING MICROCLIMATE DATA AND MODELS FOR ECOLOGICAL APPLICATIONS

Lead organizer: Dave Klinges, University of Florida, dklinges9@gmail.com

Co-organizers: Mike Kearney, University of Melbourne, m.kearney@unimelb.edu.au; Ilya Maclean, i.m.d.maclean@exeter.ac.uk; Rebecca Senior, Durham University, rebecca.senior@durham.ac.uk

Understanding the impacts of global change on organisms and ecosystems requires data and models that represent ecologically-relevant conditions. Microclimates have long been studied in ecology, and a recent resurgence of interest has produced a proliferation of fine-resolution, large-extent databases, along with models that predict microclimate and organismal responses. This hands-on coding workshop will overview the application of microclimatic information in ecology and introduce several R packages, and may be of interest to any who seek to use quantitative methods in climate change ecology. We will cater to conference attendees with novice or intermediate modeling experience, and there is no prerequisite for familiarity with climate or biophysical modeling.

Climate and the niche: We will review the direct and indirect connections between climate and organisms and how this relates to the ecological niche, especially the distinction between the realized and fundamental niche and between correlative and mechanistic niche modeling and their projection to predict distributions.
Climate databases and access: we will cover several established and new climate data products that are suitable for ecological applications [e.g. WorldClim, ERA5, terraclimate, CHELSA, SoilTemp, ForestTemp, DirtClim], and briefly compare them to shed light on how each can be useful.
Climate downscaling and variable selection: we will demystify several downscaling techniques (methods to convert coarse macroclimate to higher-resolution, more proximal local conditions) via mcera5 and other tools. Furthermore we will discuss best practices for summarizing climate time series data into ecologically-relevant bioclimatic variables.
Microclimate and biophysical modeling: we will then use several mechanistic models that can predict high-resolution microclimates at practically any time and anywhere on earth [NicheMapR, microclima, microclimc, microclimf]. We will also use mechanistic niche models to simulate the physiology and behavior of an organism in response to local microclimate, for understanding its body condition, demography, or other ecological consequences, using RShiny apps for NicheMapR.
Microclimate applications in ecology: we will finish with further approaches for understanding species niches, distributions, and habitat suitability, such as climate-driven Species Distribution Models (SDMs) and connecting microclimate to life history using metabolic theory.

After this workshop, participants will be equipped with the tools and knowledge to employ cutting-edge climate and biophysical modeling techniques. Participants will be provided in advance instructions to install the necessary packages, the R scripts to be used, and the corresponding data (no requirement to bring your data). Participants are expected to be familiar with base R/RStudio, and we will follow tidyverse conventions.

Sunday May 14, 2023
9 am - 5 pm

Important Note: This workshop is a two part workshop with Part 1 workshop being led by Mireia Valle. In the early afternoon of Sunday, Part 2 will be led by Saras Windecker (University of Melbourne) and David Uribe-Rivera (CSIRO Australia) and provide an overview of model-based data integration for modeling species’ distributions.

SDM Part 1

Building species distribution models in accordance with the ecological niche theory

Lead organizer: Mireia Valle, AZTI, Marine Research, Basque Research and Technology Alliance (BRTA), mvalle@azti.es

Co-organizers: Leire Citores, AZTI, Marine Research, Basque Research and Technology Alliance (BRTA), lcitores@azti.es; Maite Erauskin , AZTI, Marine Research, Basque Research and Technology Alliance (BRTA), merauskin@azti.es; Leire Ibaibarriaga, AZTI, Marine Research, Basque Research and Technology Alliance (BRTA), libaibarriaga@azti.es; Guillem Chust, AZTI, Marine Research, Basque Research and Technology Alliance (BRTA), gchust@azti.es

Species distribution models (SDMs) are widely used as a tool for understanding species spatial ecology. They link species occurrence or abundance with environmental features of the location, via statistical modelling. According to ecological niche theory, species response curves are unimodal with respect to environmental gradients. While a variety of statistical methods have been developed for species distribution modelling, a general problem with most of these habitat modelling approaches is that the estimated response curves can display biologically implausible shapes which do not respect ecological niche theory. This is because species response curves are fit statistically with any assumption or restriction, which sometimes do not respect the ecological niche theory. To better understand species response to environmental changes, SDMs should consider theoretical background such as the ecological niche theory and pursue the unimodality of the response curve with respect to environmental gradients. At this short course you will learn to build SDMs under the ecological niche theory framework. We will use shape-constrained generalized additive models (SC-GAMs) that allow imposing concavity constraints in the linear predictor scale and avoid overfitting. SC-GAMs are based on the same statistical framework as GLMs and GAMs regression methods, but they allow us to incorporate monotonicity and concavity shape-constraints in the component functions of the linear predictor of the GAMs. Imposing concavity constraints should be an effective alternative to fitting nonsymmetric parametric response curves, while retaining the unimodality constraint, required by ecological niche theory, for direct variables and limiting factors. We will guide you building SC-GAMs from the beginning retrieving the occurrence data and environmental data from global public datasets. We will clean raw data removing outliers and select the environmental variables using a step-forward function. The course is intended to anyone who wants to broaden their knowledge of SDMs and improve their model’s realism. We will follow a R tutorial developed in AZTI. R is a programming language widely used by data scientists and R basic knowledge is required to follow the course.

Sunday May 14, 2023
9 am - 4 pm

Application of connectivity modeling to identify and predict movement and range redistribution

Workshop leads: Prof. Rob Fletcher (Professor) University of Florida, Andrew Marx (Data scientist, University of Florida), Maru Iezzi (Postdoctoral Associate, University of Florida), Alex Baecher (PhD student, University of Florida)

Required materials:
1) Install R on your computer (if you haven’t already!).
3) Make sure to install / update the following R packages: samc, raster, terra, gdistance, sf, tidyverse, leaflet, htmlwidgets, Matrix
4) We will provide other materials over email (code, data, literature).

This 1 day workshop will distill the basic principles of connectivity modeling and its relevance to predicting movement and species redistribution with environmental change. Participants will be introduced to a circuit theoretic approach as well as a new framework utilizing spatial-absorbing markov chains (SAMC) and how to apply this analytical tool to study landscape connectivity and how it may alter global biodiversity redistribution under rapid environmental change. This workshop will explore how best to model connectivity of both habitat- and climate-analogs. Along the way, we will discuss core principles in defining a species range such as adaptive potential, connectivity and climate exposure, and morphological and physiological traits that vary from contracting to leading edge of a species range. We will also discuss the common problems, assumptions, and oversights occurring in published predictions of species range redistribution.

Time	Topic
9:00-9:10	Welcome and Introductions

Backfround and Motivation

Time	Topic
9:10-10:00	Connectivity and species on the move
10:00-10:15	The state of connectivity modeling using R and other software
10:15-10:30	Coffee break / discussion

Markov chains, connectivity, and the SAMC framework

Time	Topic
10:30-11:00	An introduction to the SAMC and the samc package
11:00-11:15	Relationships with other connectivity frameworks
11:15-11:30	R activity—contrasting the SAMC and circuit theory
11:30-12:00	Lunch break

Building a connectivity model with the SAMC

Time	Topic
12:00-12:30	An introduction to the SAMC and the samc package
12:30-1:00	Building and tuning a SAMC model
1:00-1:30	Spatial networks and the SAMC
1:30-1:45	Coffee break / discussion

Advanced topics

Time	Topic
1:45-2:00	Modeling dispersal kernels with the SAMC
2:00-2:30	Movement ecology and the SAMC
2:30-3	Dynamic models for connectivity and species redistribution
3-3:30	Scaling models to large regions
3:30-4:00	Discussion of applications and conclusions

Friday May 19th
1pm - 6pm
Introductory meeting during the week over lunch (TBD)

Creating your own downscaled future climate data for ecological applications using CHELSA CMIP6

Lead organizer: Dirk Nikolaus Karger, Swiss Federal Institute for Forest, Snow, and Landscape Research WSL.

In order to predict the future response of species to climate change, climate prediction data are essential. With the release of the Coupled Model Intercomparison Project Phase 6 (CMIP6), a large amount of new data has become available. CMIP6 includes a much larger number of climate models, socio-economic scenarios and experiments than its predecessor, CMIP5. However, these data are generally too coarse compared to the spatial detail needed for most ecological studies. Many applications that study the impact of climate on species do not require the full range of variables, models, or SSPs that CMIP6 can provide, but rather a high resolution of 1 km. Furthermore, they often rely on only a limited number of so-called bioclimatic variables, which describe averages or variations in climate over long periods of time. The CHELSA data portal already offers such downscaled CMIP6 data. However, due to the large number of possible combinations, it is very difficult to provide pre-processed, downscaled CMIP6 data for all climate and SSP models. In addition, data are only available for a limited number of time periods, which do not always coincide with the time periods required for some studies.
In this workshop, we will provide an overview of the chelsa-cmip6 Python module, which allows the creation of custom, downscaled CMIP6 bioclimate data for any region. The chelsa-cmip6 package allows you to specifically select the desired geographic extent, climate model, GCM, and time period to facilitate easy delivery of state-of-the-art climate data for ecological climate research.

After this workshop, you will be able to produce your own future 1km bioclimatic data based on the Delta Change method. The workshop will cover the following aspects.

Theory:

Overview about the theory behind downscaling and the delta change method.
Overview about the CMIP6 initiative, numerical climate modelling, the concept GCMs, ESMs, socio-economic pathways etc.
Overview of the chelsa-cmip6 module. Code structure, its classes and functions.

Practical exercise:

Creating your own future bioclimatic data for a specific time in a future.
Creating a monthly timeseries of future climate data.

Prerequisites: Although chelsa-cmip6 is written in Python, no real knowledge of Phyton is required to use it, but some basic knowledge of Python may be beneficial to take full advantage of the module. However, you should know how to use the terminal or command line interpreter of your computer.

Sunday May 14, 2023
9 am - 5 pm

Important Note: This workshop is a two part workshop with Part 1 workshop being led by Mireia Valle. In the early afternoon of Sunday, Part 2 will be led by Saras Windecker (University of Melbourne) and David Uribe-Rivera (CSIRO Australia) and provide an overview of model-based data integration for modeling species’ distributions.

SDM Part 2

Theory and practice of model-based data integration for modeling species’ distributions

Workshop Leads: Saras Windecker, University of Melbourne, sm.windecker@unimelb.edu.au
David Uribe-Rivera, Commonwealth Scientific and Industrial Research Organisation, de.uribe.r@gmail.com

To make the most accurate prediction of a species’ distribution it is important to make use of all relevant data. Although opportunistically collected species records are more readily abundant, they also contain lower quality information and are more likely to be biased. Standardised surveys may suffer less observation bias, but they are expensive to conduct and are therefore more likely to have limited spatial extent. Combining opportunistic and standardised data not only gives more information to a model, but also provides a chance for structured survey data to help correct biases in opportunistic data. Model-based data integration allows us to explicitly model the ecological (species’ occupancy/abundance) and observational processes (detection/sampling effort) simultaneously, propagating information contained in each while accounting for appropriate biases. In this skills showcase we will explain the concepts behind model-based data integration, demonstrate its implementation in a Bayesian framework for modeling species’ distributions, and practice with a number of exercises. Our examples will use broad spatial-scale presence-only data and spatial abundance data from structured surveys, and will illustrate how Point process models can be used to facilitate this integration. We expect that participants will leave more confident in the theory and execution of integrated models.