Harvestingandharnessingdataforbiogeographicalresearch

A.Species TreeGeography TreeB.

20406080100

2 1

3

– inventories and surveys• natural areas, preserves, state

forests, private properties• development permits

– herbarium/museum specimens• collectors often target specific

places or taxa• records of what, where, when,

and by whom

How do we know what grows where?

How do we know what grows where?

∑ Local observations = State Flora

blackbulrush

How do we know what grows where?

∑ State Floras = National Flora

blackbulrush

How do we know what grows where?

∑ National Floras = Global Flora

blackbulrush

Who manages these data?

Benefits of Digitization• Preservation of rare/fragile specimens

• Easier to query a database than physical collections (e.g., WISFLORA)

• Efficient data sharing

• Collation yields powerful data repositories

Digitization of Distribution Data

Examples of data repositories1. Wisconsin Flora – more on this later2. North America Digital Flora3. Biota of North America Project4. Global Biodiversity Information Facility

North America Digital Flora

• Pros:– Great for generating species lists– Can query by species, attribute, or geography– Incorporates interesting GIS layers

• Cons:– Limited to wetland species– Geography search lacks precision

BONAP

• Pros:– Comprehensive of NA diversity– Great, easy to use distribution maps

• Cons:– Can’t query the data– Can’t download data– North America only– Poor resolution in Canada

GBIF

• Pros– Global data of plants, animals, fungi– Database is fully searchable– Data are downloadable

• Cons– Data are incomplete (WIS not in yet)– Some records are suspect

Whatcanyoudowiththesedata?

• TestbiogeographicpatternsinWiscosnin• Historicalbiogeography• Characterizespeciesdistributions• Obtainspecies“climaticenvelope”• Identifyareasexhibitinghighlevelsofendemism• Predictresponsestoglobalclimatechange• Trackandpredictspreadofinvasivespecies

Allwithoutspending$1

Wisflora

WisconsinFloraMappingProject

WisconsinFloraMappingProject

Wisconsinisnotequallyinventoried

>10,000specimens

8,000- 6,000

6,000- 4,000

2,000- 4,000

<2,000

Thetensionzoneseparatingthetwoprovincesisbasedontheupperandlowerlimitsofthesouthernandnorthernspecies,respectively.

Goldencassia Wildindigo Ram’sheadladyslipper

Stemlessladyslipper

WisconsinFloraMappingProject

Mean Phylogenetic Distance (MY)

114-200201-300301-400400-500

Phylogenetic Conservatism

Strongly ConservedConservedOverdispersedStrongly Overdispersed

HistoricalBiogeography

• Ifyouwanted,youcouldbuildaphylogenyof300,000organismsusing187billionbasepairsfromGenbank

e.g.,familyCyperaceae (sedges,bulrushes)– Spalink etal.2016

HistoricalBiogeography

• Buildaphylogeny– Datethephylogeny

20406080100

DatafromGenbank,analyzedusingBEAST

HistoricalBiogeography

• Buildaphylogeny– Datethephylogeny– Reconstructareasinhabitedbyancestors

20406080100

DatafromGenbank+eMonocotDatafromGenbank+eMonocot,analyzedusingBEAST andBioGeoBEARS

HistoricalBiogeography

• Buildaphylogeny– Datethephylogeny– Reconstructareasinhabitedbyancestors

– Measureratesofdiversification

– Trackevolutionofmorphologicaltraitsacrosstimeandspace

20406080100

2 1

3

Diversification Rate

Den

sity

DatafromGenbank+eMonocot,analyzedusingBEAST,BioGeoBEARS andBAMM

HighResolutionSpeciesDistributions

• Characterizedistributionsofspeciesfor1.45millionspeciesusing460milliongeo-referencedoccurrences

HighResolutionSpeciesDistributions

• Characterizedistributionsofspeciesfor1.45millionspeciesusing460milliongeo-referencedoccurrences– Identifyspeciesrichareas

DatafromGBIF,analyzedinQGIS

HighResolutionSpeciesDistributions

• Characterizedistributionsofspeciesfor1.45millionspeciesusing460milliongeo-referencedoccurrences– Identifyspeciesrichareas

– Defineregionswithhighlevelsofrareorendemictaxa

DatafromGBIF,analyzedinQGIS andPAUP* - availableinsystematicslab

Parsimony: typically used in phylogenetics…

ExampledatasetfromSRDownie

SpeciesCharacters

Parsimony Analysis of Endemicity(PAE)

…but, can be applied to PAE

Species A Species B Species C Species D Species E Species F Species G Species HArea 1 1 0 1 0 0 0 0 0Area 2 1 0 1 0 0 0 0 0Area 3 0 1 0 1 0 0 0 0Area 4 0 1 0 1 0 0 0 0Area 5 0 0 0 0 0 0 0 0Area 6 0 0 0 0 1 1 0 0Area 7 0 0 0 0 1 1 0 0Area 8 0 0 0 0 1 1 0 0Area 9 0 0 0 0 1 1 0 0Area 10 0 0 0 0 1 1 0 0Area 11 0 0 0 0 0 0 0 0Area 12 1 0 0 0 0 0 0 0Area 13 1 0 0 0 0 0 0 0Area 14 1 0 0 0 0 0 0 0Area 15 0 1 0 1 0 0 0 0Area 16 0 1 0 1 0 0 0 0Area 17 0 1 0 1 0 0 0 0Area 18 0 0 0 0 0 0 0 0

wheregeographicAreasaretreatedastaxa,andthepresence/absenceofthespeciesserveasCharacters

Parsimony Analysis of Endemicity(PAE)

Case Study: finding areas of endemism in North American sedges

Step 1: Get distributional data of all species in N Am

Step 2: Define geographic region and divide into quadrats

Case Study: finding areas of endemism in North American sedges

Step 3: Construct data matrix: absence = 0, presence = 1

Case Study: finding areas of endemism in North American sedges

Step 4: Conduct parsimony analysis

Case Study: finding areas of endemism in North American sedges

Step 5: Find clades consisting of 2+ endemic taxa with congruent distributions.

Case Study: finding areas of endemism in North American sedges

EcologicalNicheModeling

• Developecologicalnichemodelsusingwaypointandclimatedata(BioClim)

Scirpus longii

EcologicalNicheModeling

• Developecologicalnichemodelsusingwaypointandclimatedata– Identifysuitableareasforendangeredspecies

DatafromGBIF andBioClim,analyzedwithMaxent

Scirpus longii

EcologicalNicheModeling

• Developecologicalnichemodelsusingwaypointandclimatedata– Identifysuitableareasforendangeredspecies

– ReconstructdistributionsduringLastGlacialMaximum

DatafromGBIF andBioClim,analyzedwithMaxent

Scirpus longii

Withoutspending$1,youcan:

• Developecologicalnichemodelsusingwaypointandclimatedata– Identifysuitableareasforendangeredspecies

– ReconstructdistributionsduringLastGlacialMaximum

– Predictspeciesdistributionsin50years

DatafromGBIF andBioClim,analyzedwithMaxent

Scirpus longii

Trackspreadofinvasivespecies

• Wheredidtheyoriginate?

• Whendidtheygethere?

• Wherearetheygoing?

GarlicMustardandPurpleLoosestrife:1870

DatafromGBIF,analyzedinQGIS

GarlicMustardandPurpleLoosestrife:1900

DatafromGBIF,analyzedinQGIS

GarlicMustardandPurpleLoosestrife:1910

DatafromGBIF,analyzedinQGIS

GarlicMustardandPurpleLoosestrife:1920

DatafromGBIF,analyzedinQGIS

GarlicMustardandPurpleLoosestrife:1930

DatafromGBIF,analyzedinQGIS

1927Roadmap

Copyright©2010CartographyAssociates

GarlicMustardandPurpleLoosestrife:1940

DatafromGBIF,analyzedinQGIS

GarlicMustardandPurpleLoosestrife:1950

DatafromGBIF,analyzedinQGIS

GarlicMustardandPurpleLoosestrife:1960

DatafromGBIF,analyzedinQGIS

GarlicMustardandPurpleLoosestrife:1970

DatafromGBIF,analyzedinQGIS

GarlicMustardandPurpleLoosestrife:1980

DatafromGBIF,analyzedinQGIS

GarlicMustardandPurpleLoosestrife:1990

DatafromGBIF,analyzedinQGIS

GarlicMustardandPurpleLoosestrife:Today

DatafromGBIF,analyzedinQGIS

Usingecologicalnichemodeling,youcouldpredictwhichregionsmightbeunderthreatofinvasion.

Some Cautions:Distribution of Alliaria petiolata according to:

GBIF WISFlora

Some Cautions:How do these maps of the distribution of Scirpus atrovirens

differ? Why do they differ?

AdditionalResources

• panbiogeography analysis(martitracks)• simulateoriginandspreadofspecies(biogeosim)• forecastspeciesdistributions(biomod)• spatialanalysisofdiversity(biodiverse)• listoffreephylogeneticsoftwarehere• simulatehistoricalislandbiogeography(shiba)• spatialanalysisinmacroecology (SAM)• predictandanalyzedistributions(GARP)