2014-15: ECOL: Multivariate Models in Ecology: March 2-4, 2015


This workshop was held at SAMSI in Research Triangle Park, NC.


This workshop arose from the Multivariate Modeling working group of the SAMSI Ecology program. Multivariate modeling for ecological processes encompasses many fundamental ecological challenges and the working group is exploring several of them. A primary focus is on joint species distribution modeling, measured through presence-absence or abundance. Historically, species distribution modeling has only been considered for a single species or for an aggregated species such as functional types, genera, taxa, or communities. Here, we are exploring the joint dependence between species in achieving their distributions on various spaces. These spaces may be climate spaces, yielding climate niches and range limits; they may be geographic, yielding joint spatial distributions. Working with a large number of species requires implementing dimension reduction strategies. Additionally, depending upon scale, spatial modeling may be needed in order to capture not only dependence between species at a given location but also dependence with regard to presence-absence or abundance between locations. An important connected focus is successful application to a variety of datasets, treating very different types of species and reflecting the incidence of various environmental disturbances, e.g., fire or infestation. A related theme is the thorny issue of fusion of presence-absence data with presence-only data. That is, frequently, multiple sources of information are available to learn about species distributions. One type of source is designed experiments where certain plots are selected for their environmental characteristics and are then sampled, in particular, yielding presence-absence data. A second type of source is so-called museum or citizen science data arising from collection which is not designed but rather consists only of observed presences. Clearly both sources inform about species distribution; the challenge is to synthesize them.

Questions: email eco@samsi.info

Schedule and Supporting Media

Participant List
Speaker Titles/Abstracts

Monday, March 2, 2015

Time Description Speaker Slides Videos
8:15-8:45 Registration and Welcome
8:45-9:30 Data Fusion for Ecological Studies Beth Gardner, NCSU
Dave Miller, Penn State
Krishna Pacifici, NCSU
9:30-10:15 Dimension Reduction for Joint Species Distribution Models Daniel Taylor-Rodriguez, SAMSI
Kimberly Kaufeld, SAMSI
10:15-10:30 Break
10:30-11:00 Joint Species Distribution Modeling in the Pacific Northwest Bijan Beyednasrollah, Duke University
Ian Breckheimer, University of Washington
Andrew Latimre, UC-Davis
11:00-12:15 Plenary Discussion to Set Agenda/Goals
12:15-1:30 Lunch
1:30-5:00 Subgroup Meetings
5:00-6:30 Poster Session and Reception

Tuesday, March 3, 2015

Time Description Speaker Slides Videos
8:15-8:30 Registration
8:30-11:30 Subgroup Meetings
11:30 Plenary, Subgroups Report Back
12:30 Lunch
1:30-5:00 Subgroup Meetings

Wednesday, March 4, 2015

Time Description Speaker Slides Videos
8:15-8:30 Registration
8:30-11:30 Plenary Discussions
11:30 Box Lunch and Adjourn