Transcript Slides - Indico

Large scale modelling of the distribution
of butterfly biodiversity in Europe
using the BioVeL portal on the EGI infrastructure
Yuliya Fetyukova, Hannu Saarenmaa, University of Eastern Finland, Joensuu, Finland;
Matthias Obst, Sarah Bourlat, University of Gothenburg, Gothenburg, Sweden;
Renato De Giovanni, Reference Center on Environmental Information, Campinas SP, Brazil;
Alan Williams, Norman Morrison, University of Manchester, Manchester, UK,
Francisco Quevedo, Cardiff University, Cardiff, UK
20 May 2014
University of Helsinki, EGI Community Forum 2014
1
Butterfly distribution modelling
Species distribution modelling allows determining variation
among species and their shifts over time as responds to climate
and environmental changes.
Butterflies are good and viable biological indicators:
- well documented;
- easy to identify and monitor;
- butterflies and moths comprise 5% of terrestrial biodiversity;
- react quickly to changes.
Study case steps:
- modelling historical and current species distribution;
- estimate future climatic changes on species;
- computation of habitat suitability shifts
- compute trends in change of abundance,
and estimate Essential Biodiversity Variables
2
Ecological niche modelling
General principles of Ecological Niche Modelling (ENM) method
3
BioVeL Ecological niche modelling
The BioVeL project is creating a set of workflows (series of data analysis steps)
and Web Services for scientists to perform different analyses in the fields of
ecology, taxonomy, phylogenetics and metagenomics.
Data
Species occurrence data of all Butterflies in Europe*
344 species occurrence data sets
GBIF Data Portal http://www.gbif.org/
Environmental layers
WorldClim repository http://www.worldclim.org/
Tools
Data Refinement Workflow, DRF
Ecological niche modelling workflow, ENM
The ENM Statistical Workflow, ESW
BioVeL portal data sweep function for batch processing
https://portal.biovel.eu/
* Karsholt, O. & Razowski, J. (Eds.), 1996. The Lepidoptera of Europe: A distributional checklist, 380 pp,
4
Apollo Books, Stenstrup, Denmark
GBIF species occurrence data
344 GBIF species occurrence data sets have been processed by
BioVeL workflows
Europe mask: decimalLatitude: [32, 72]; decimalLongitude: [-10, 40]
5
Species occurrence data
http://www.gbif.org
Erynnis tages [Dingy Skipper] georeferenced data
BioSTIF: Biodiversity Spatial Temporal Interactive interface
1776 occurrences
Erynnis tages: Species distribution in 1960 – 1990
6
Historic species distribution
Erynnis tages species distribution in 1960 – 1990
Algorithm:
Maximum entropy
Terrestrial layers :
- Annual mean
temperature;
- Annual precipitation;
- Maximum temperature of
warmest month;
- Minimum temperature of
coldest month.
- Altitude in meters
Overall coverage:
74.8%
Overall intensity:
10.7%
7
Habitat suitability shifts by 2050
Erynnis tages: diffLayer = predictionLayer – historicalLayer
Cells with colors from
green to red indicate
an increase
and from green to blue
a decrease of predicted
potential for a species.
8
1
2
Shift vectors
3
SHIFT vectors:
1 Erynnis tages
622.42 km
2 Spialia sertorius
424.31 km
3 Carcharodus alceae
539.04 km
9
Coverage and Intensity statistics
coverage
1960-1990, %
coverage
2050, %
increase(+)/
decrease(-)
in coverage, %
intensity
1960-1990, %
intensity
2050, %
increase(+)/
decrease(-)
in intensity, %
Erynnis
tages
74.77
84.14
9.37
10.75
10.26
-0.48
Spialia
sertorius
63.45
78.69
15.23
6.05
4.71
-1.34
Carcharodus
83.46
97.34
13.89
16.64
20.09
3.45
alceae
...
344 species
Essential Biodiversity Variable of
overall butterfly distribution change
STACK of 344 maps
10
BioVeL portal & ENM wfs
EGI deployment of ENM related services:
- parallelized computation and
hence scalable workflow execution
Portal sweep data functionality:
- batch processing of large amount of species
http://www.biovel.eu/
https://portal.biovel.eu/
11
THANK YOU!
http://www.gbif.org