The Environment Institute welcomes Professor Fangliang He, who visits Adelaide in October to present a seminar and a short course.
Fangliang is professor and Canada Research Chair in Biodiversity and landscape modelling at the University of Alberta, Edmonton, Canada/Sun Yat-sen University, Guangzhou, China.
His research focusses on understanding how and why species disperse, colonize, and persist in landscapes.
Modelling species distributions in fragmented landscapes
When: Wednesday 22nd October, 3.10 pm
Where: Mawson Lecture Theatre, Adelaide University, North Terrace Campus
The increasing change in land use has profoundly fragmented previously continuous landscapes and degraded the quality of habitats. Understanding the distribution of species in such altered landscapes is essential for effective conservation of species and the preservation of the integrity of biodiversity.
In this talk I will present basic mathematical models that colleagues and I have developed to model the distribution of species in space. My presentation will comprise three coherently related components.
(1) I will first introduce models that directly link two fundamental ecological quantities – spatial distribution and abundance of species.
(2) I will then describe the mathematical relationships unifying key landscape metrics (e.g., the number of patches occupied by a species and the edge length of the occupied patches) that were developed based on these species distribution-abundance models. These relationships predict that species distributions exhibit percolation critical thresholds, i.e., critical population abundances at which the fragmented patches (as measured by the number of patches and edge length) start to coalesce to form large patches.
(3) Lastly, I will apply the species distribution-abundance models to assess species extinction risk following the IUCN Red List criteria. The proposed mathematical models provide a solution to the long-standing problem for IUCN assessments of lack of data on population sizes and are expected to facilitate the quantification of extinction risk worldwide.
I will conclude my talk by proposing the next steps required to understand and test the distribution models in altered landscapes.
Spatial statistics and modelling of species distribution
When: Thursday 23rd October, 9-4 with an hour break at 12pm
Where: Barr Smith South 1063 Computer Lab (use Horace Lamb Lecture Theatre entrance to building for ease of location)
The short course will introduce basic concepts of spatial statistics and instruct methods for analysing spatially structured data. The contents of the course include:
(1) Introduction of spatial statistics (concepts, properties of spatial data)
(2) Measuring autocorrelation
(3) Linear/generalized linear regression models (nonspatial)
(4) Simultaneous autoregressive models for modelling species distribution
Real data of species distributions will be used to illustrate the concepts and methods introduced in this course. R package “spdep” will be used throughout the course.