Trees, taxonomy & location: mapping phylogeography using Biodiverse Dan Rosauer & Shawn Laffan University of New South Wales & Centre for Plant Biodiversity Research Taxonomic Databases Working Group 22 October 2008

New tools and measures for practical phylogenetic diversity analysis Tools: Biodiverse – a tool for spatial analysis of biodiversity Measures: Phylogenetic Endemism - an index of localised evolutionary history

Summary area cladogram resulting from the analysis of Australian Rhamnaceae, tribe Pomaderreae. From Ladiges et al Why map phylogenies? (1) To study the spatial component of evolution… Dung beetle evolution in Australia’s wet tropics – from Bell et al 2007

To inform biodiversity conservation… Why map phylogenies? (2) Red and yellow areas represent centres of current speciation. They are identified as important areas to protect for maintenance of evolutionary processes Mammalian evolutionary hotspots in California Davis et al, 2008 Molecular Ecology

Because diversity is not distributed equally between taxa… Why map phylogenies? (3) All species are equal, but some are more equal than others… after Rest et al (2003) Snakes & Lizards (several thousand species) Tuatara (2 species)

Part of an ITS tree for tribe Pomaderreae (Rhamnaceae) Kellerman et al (2005) Pomaderris elliptica Phylogenetic diversity

A practical slant on phylogenetic mapping Theory and methods – robust but developing Data (+ ability to access it) – exploding!  Specimen and survey data  Molecular data  Published phylogenies How can we make phylogenetic mapping more accessible and useful for study and management of biodiversity?

Biodiverse –a tool for spatial analysis of biodiversity indices of richness, endemism, compositional dissimilarity, sampling redundancy clustering Monte Carlo randomisations …and now phylogenetic measures Biodiverse has been used to investigate: Endemism in the Australian Flora - Crisp et al (2001), Laffan & Crisp (2003) Taxonomic and genetic patterns in Pultenaea (Fabaceae) - Bickford et al (2004) Scale dependence of relationships between climate and fern species richness - Bickford & Laffan (2006) Mean genetic similarity in Pultenaea Bickford et al (2004)

Load species location records

Define cell size

~25,000 herbarium and survey records for Daviesia (Fabaceae) loaded against chosen cell size (1 degree)

View distributions of individual taxa

Load phylogeny from a nexus or tree file

Display phylogeny D. abnormis D. ovata D. euryloba

Explore phylogeography

View distribution of a clade Explore phylogeography

View species list for selected clade View distribution of a clade Explore phylogeography

View distribution of a clade Explore phylogeography

View distribution of a clade Explore phylogeography

View distribution of a clade Explore phylogeography

View distribution of a clade Explore phylogeography

Select a cell to highlight its taxa on the tree

Choose from a range of analyses Select radius for neighbourhood analysis

Biodiverse Demo

Linking spatial data to the tree via names Name on tree Names in spatial data - Stenanthemum leucophractum - Cryptandra leucophracta - Stenanthemum leucophractum (Schldl.) Reissek - Cryptandra leucophracta Schltdl. - Stenanthemum leucophractum (Schltdl.) Reissek - Cryptanda leuc. STLE Taxon name Stenanthemum leucophractum Names matched using Australian Natural Heritage Assessment Tool over 75,000 formal & informal synonymies for Australian taxa matched with valid names from relevant authorities – eg ABRS, APNI

Two ways to estimate endemism Absolute endemism Estimate the species or PD restricted to a defined area e.g. Sechrest et al (2002) Weighted endemism (for species) For each area: weighted endemism = sum of 1 / range for species present Grid allows continuous variation – the data determines the areas of interest Williams & Humphries (1994) Crisp et al (2001)

Sechrest et al 2002 Only in hotspots Hotspots & elsewhere Not in hotspots Primates in global biodiversity hotspots

Two ways to estimate endemism Absolute endemism (for species or PD) Estimate the species or PD restricted to a defined area e.g. Sechrest et al (2002) Weighted endemism (for species) For each area: weighted endemism = sum of 1 / range for species present Grid allows continuous variation – the data determines the areas of greatest endemism Williams & Humphries (1994), Crisp, Laffan, Linder & Monro (2001), Slatyer, Rosauer and Lemckert (2007)

Phylogenetic endemism a measure of the degree to which elements of evolutionary history are restricted in space Desirable properties of a phylogenetic endemism measure:  Spatially consistent data define areas of endemism allow comparison between equal areas  Independent of level apply consistently to any taxonomic unit  Conceptually simple flexibly applied to different problems and methods

Phylogenetic Endemism A B C Calculate range for each branch, not each species Phylogenetic endemism (PE) = length A / range S.humile + length B / range S.humile & S.reissekii + length C / range S.humile & S.reissekii & S.petraeum & S.argenteum + …

A B C A B C High PE A B C Low PE A B C Amount of shared PD Larger  ================  Smaller Phylogenetic Endemism (PE) contributed by an area with a narrowly distributed taxon A. Rosauer et al. (in prep)

Phylogenetic Endemism (PE) result for Daviesia

Phylogenetic endemism in Australian tree frogs

This project is supported by an ARC Linkage Grant. Partners: Thanks to: Mike Crisp Lyn Cook Marcel Cardillo Steve Donnellan Biodiverse software: Shawn Laffan Eugene Lubarsky (user interface) Dan Rosauer (phylogenetic functions) Dan Rosauer is currently a visiting fellow at the Australian National University Centre for Macroevolution & Macroecology