The database itself, with source data and the build for every release. Explore it through the data portal or the HTTP API.
Trait Ecology & Evolution · UNSW Sydney
We are the Cornsters: the joint labs of Profs Daniel Falster and Will Cornwell at UNSW Sydney. Our mission is to understand how plant traits shape ecology and evolution, from single leaves to whole floras. We build the open data, models and software that work needs, and share them so that others can use them too.
01Plant trait data
AusTraits is an open, curated database of traits for the Australian flora, harmonised from hundreds of sources. Around it sit the tools that build it, define its traits, resolve its plant names and serve its data.
The database itself, with source data and the build for every release. Explore it through the data portal or the HTTP API.
Access the AusTraits database from R and wrangle data from any traits.build database.
A workflow for harmonising trait data from diverse sources into a documented, standard structure. AusTraits is built with it.
The AusTraits Plant Dictionary, with reviewed definitions for nearly 500 plant traits and machine-readable endpoints.
Resolve and update the names of Australian plants against the Australian Plant Census.
Align taxon names to any taxonomic reference, and generate a reference list for any taxon group in any region.
Species- and community-level trait means for the Australian flora, built from AusTraits with climate and occurrence data.
An offshoot of traits.build for physiological response-curve data, such as leaf gas exchange, hydraulic vulnerability and fluorescence curves.
02Vegetation models
The plant model simulates how traits shape the growth, competition and evolution of plants in forests. The family around it adds documentation, community assembly, site and species set-up, leaf physiology, fire and visualisation.
Trait-driven models of size-, patch- and trait-structured forest demography, ecology and evolution.
A field guide to the plant model, with user guides, theory and worked notebooks.
Extends plant to community assembly and trait evolution, with solvers for resident equilibria and invasion fitness.
Sets up plant runs by bundling a site’s environmental drivers with species’ trait trade-offs and sourced parameter sets.
A fast, header-only C++ ODE solver with reverse-mode automatic differentiation and an R interface.
A leaf gas-exchange model in which stomatal behaviour emerges from plant hydraulics through gain–risk optimisation.
Forest flammability using FRaME fire mechanics in C++, generalised from discrete strata to continuous size distributions.
Interactive 3D views of forest stands simulated with plant, drawn as individual trees growing through time.
03Vegetation plot data
HAVPlot, the Harmonised Australian Vegetation Plot dataset, brings vegetation plot surveys from across Australia into one compilation of more than 200,000 plots. We are rebuilding it as an open, documented workflow that does for plot data what AusTraits does for trait data. The repositories will open as the work matures.
A harmonised compilation of Australian vegetation plot data, built from many sources into a single documented database.
PaperNot yet public
A workflow for harmonising vegetation plot data from diverse sources. It is the plot-data counterpart to traits.build.
Not yet public
04Nature sounds
Tools for finding species in sound recordings: gathering recordings from citizen-science archives, running and combining automated detectors such as BirdNET and Google Perch, and turning their detections into data for analysis.
Process and visualise the large volumes of detections produced by BirdNET.
Download and analyse ecoacoustic recordings from iNaturalist and the Atlas of Living Australia.
Train and run custom classifier heads on Google Perch embeddings, for recognising species beyond Perch’s own vocabulary.
Parameter-free, score-level ensembling of two acoustic detectors.
05Who lives where, and what has changed
Tools for building species lists for any place from herbarium collections and citizen-science records, and for tracking how those lists change: new discoveries, species long gone unobserved, and the curation of “all time” floras for national parks.
Generate place-based species lists for anywhere in the world, from collection data and iNaturalist records in the ALA and GBIF.
The “all time” plant species list for Royal National Park, NSW, with a script that checks the ALA for candidate new discoveries since the list was made.
The “all time” plant species list for Yosemite National Park, California, with a script that checks GBIF for candidate new discoveries since the list was made.
Compares Yosemite and Royal National Parks to track which species have gone long unobserved and which are new discoveries, through time.
An honours thesis on plant species in Royal National Park that have long gone unrecorded, with the data, code and figures behind it.
06Everything else
Standalone apps, R packages and data packages for statistics, remote sensing, animal movement and more.
Estimation and inference for allometric lines by (standardised) major axis regression.
Hierarchical Bayesian models that estimate differential-equation parameters, such as growth rates, from repeated measurements.
Fit and compare alternative models for single-cohort litter decomposition data.
Process drone imagery to calculate spectral diversity, linking diversity seen from the sky with diversity on the ground.
Derive a colour palette from satellite imagery of any location on Earth.
The Biomass And Allometry Database (BAAD) for woody plants, delivered as a versioned R package.
A global dataset of bulk leaf δ¹³C values.
Bayesian multilevel fits of standardised major axis lines to bivariate data.
Standardise raw tri-axial accelerometer data and extract metrics of movement dynamics.
Still public and citable, but no longer actively developed.
07Reproducible research
Papers whose analyses you can reproduce from the code and data in a public repository. For full publication lists, see our Google Scholar profiles.
traitecoevo is the shared GitHub home of the Cornsters (Cornwell + Falster): the labs of Daniel Falster and Will Cornwell at UNSW Sydney, with our students, postdocs and collaborators. The tools here grew out of our research, and many are now used well beyond it.
Current and past lab members are listed on our lab websites:
Most of our R packages are built on R-universe, so you can install them without extra tools:
install.packages(
"austraits",
repos = c("https://traitecoevo.r-universe.dev",
"https://cloud.r-project.org")
)