Last week over 400 ecologists, statisticians and blends of these converged on the University of St. Andrews to present and discuss research at the International Statistical Ecology Conference (ISEC). ISEC meetings have occurred biennially since 2008.
ISEC is a great opportunity for ecologists to hear about the latest developments in ecological statistics and to form collaborations with statisticians and modellers. Popular fields at St Andrews included movement modelling and capture-recapture for estimating abundance, which may not interest many plant ecologists but reflects the need to cope with the data streams from telemetry, biologging devices and camera traps.
Two areas with more interest for plant ecologists are spatial modelling and hierarchical modelling. Point process modelling has really accelerated over the last few years and several talks (Janine Illian, Achmad Choiruddin and others) addressed these and they have the potential to increase our understanding of fine scale dispersal, competitive and facilitative effects.
Hierarchical models in many shapes and forms are widely used and many talks covered their application including community modelling but also their limitations. Ben Bolker made a strong case against multi-model averaging, while several cautionary talks about mixture models were heard.
Pre-conference workshops provided a great opportunity for learning one’s way around various packages and methodology including spatial point process modelling with inlabru (David Borchers, Janine Illian and Finn Lindgren), community modelling with hmisc (Otso Ovaskainen) and mvabund and gllvm (David Warton and Gordana Popovic) and more.
Some of the talks that made distinct impressions on me were Stephen Ellner’s plenary on quality vs luck in effects on lifetime reproductive success. Malcolm Itter presented an interesting talk on ‘ecological memory’ modelling as a way of determining and interpreting lag functions in effects of ecological covariates on boreal forest growth. Furthermore, Andria Dawson presented on modelling prehistoric forest composition from pollen while accounting for process and observational uncertainty.
Citizen science was a theme and Kerrie Mengersen presented a terrific plenary on citizen science and diverse ways of utilizing it to inform models of ecological processes, and Christophe Botella presented a specific plant species distribution modelling application (see Botella et al., 2018). Emily Mitchell, while not actually working in plants but another sessile community, presented a cool pattern analysis on spatial structure in Ediacaran faunal assemblages.
While lots of applications were to animal data, I urge plant ecologists to consider attending ISEC and hearing about the latest in statistical ecology. The next ISEC will be in Sydney, Australia in 2020.
Peter Vesk, University of Melbourne and Journal of Ecology Associate Editor
To celebrate ISEC and the BES Quantitative Ecology Annual Meeting, Laura Graham and Susan Jarvis have compiled a virtual issue celebrating all things statistical and quantitative in ecology. Laura and Susan have chosen papers from all five BES journals, including 3 from Journal of Ecology. All papers are free-to-access for a limited time.