Miller et al. have put together a review of the use of metrics and null models in community phylogenetics. They review a total of 278 metric/null model combinations and find that these are in fact 72 true unique combinations. They also find that the choice of null model is more influential over type I error rates than the choice of metric. The paper comes with an accompanying R package – metricTester – which allows for testing of method performance.
Lefcheck presents a new R package – piecewiseSEM – which is a structural equation modelling package designed with its uses in ecology, evoluation and systematics in mind. The paper contains two worked examples including code.
Statistical Ecology Research Festival. 7th June, University of Kent, Canterbury.
IALE UK 2016 Conference: landscape characterisation – methods and applications in landscape ecology. 7-9th September, University of Reading.
Genetic data analysis using R. 16-20th August, Isle of Cumbrae, Scotland.
The use of phylogenies in the study of macroevolution. 19-23rd August, Facilities of the CRIP at Els Hostalets de Pierola, Barcelona.
There’s also still time to sign up for our Data Management in R course on Monday (23rd May) at Charles Darwin House in London.
Postdoctoral research assistant at University of Oxford. Human modified tropical forests.
Postdoc at CIBIO-UE, University of Évora. Modelling food web architectures under climate change.
Research Associate at CEH, Wallingford. Ecological modelling of pathogens threatening humans, animals, plants and ecosystems.