Transcript Why model species ranges?

Ecological Niche Modelling
Theoretical Principles &
Practical applications
LECTURE STRUCTURE
Why model species ranges?
What is a niche? – fundamental and realized
Correlative range modelling – background and assumptions
Distribution datasets
Variables and their selection
Models and their selection
Model calibration and evaluation
2
WHY MODEL SPECIES RANGES?
We need to know where species occur and why they occur
where they do:

We want to predict where a particular species occurs.

We want to know more about organism-environment
relationships.
3
USED IN RESPONSE TO





Increase the rates of habitat, and face species loss;
Complete (spatial and temporal) distribution info for a large
number of taxa;
Contrast the existing distribution data mostly collected in an
ad hoc fashion.
Given the rate of species loss, it is unlikely that we will get the
distribution data that we need in time if we rely on
conventional survey techniques.
Atlases are an invaluable data source and cover very few taxa
but they are very important for model development and
calibration.
4
WHAT WE KNOW ABOUT SPECIES DISTRIBUTION?
Despite many decades of investigation our
knowledge is still incomplete and at broad scale!!!
USGS Forest Service
Atlas Florae Europeae - Jalas & Suominen (1965-2013)
5
EUFORGEN Network (2009)
DISTRIBUTION MODELS HAVE BEEN USED TO PREDICT





















species richness
centres of endemism
the occurrence of particular species assemblages
the occurrence of individual species
the location of unknown populations
the location of suitable breeding habitat
breeding success
species abundance
genetic variability of species
help target field surveys
aid in the design of reserves
inform wildlife management outside protected areas
guide mediatory actions in human–wildlife conflicts
monitor declining species
predict range expansions of recovering species
estimate the likelihood of species’ long-term persistence in areas considered for protection
identify locations suitable for introduction
identify locations suitable for reintroductions
identify sites vulnerable to local extinction
identify sites vulnerable to species invasion
explore the potential consequences of climate change
6
MODELLING LITERATURE AT A GLANCE
2,215 scientific products
1,964 published papers
Source: www.scopus.com
7
THEORETICAL PRINCIPLES: FUNDAMENTAL vs REALIZED NICHE
(G. Evelyn Hutchinson, 1957)
Definition: n-dimensional hypervolume described by N environmental and resource
constraints within which a species can maintain a viable* population. In other
terms, it is the combination of conditions and resources required by an
individual species defines the area in which it is able to live.
Definition: the set of conditions actually used by given species/population after
interactions with other species (predation and especially competition) have been
taken into account.
* Viable = Survive + Grow + Reproduce
8
PRINCIPLES: RANGE EDGES
WHAT DETERMINES THE EDGE OF GEOGRAPHIC RANGES?

Populations do not just cease to exist at the edge of their geographic distributions, but rather
taper off gradually. The edges of geographic ranges are thus defined by a change in the local
population dynamics, where net gains in population are reduced to levels lower than the net
losses.

There are changes in local population dynamics at the edge of a distribution, and more net
losses than net gains
THESE POPULATION LEVEL CHANGES ARE DUE TO:

Changes in abiotic factors (physical barriers, climate factors, absence of essential resources) and
biotic factors (impact of competitors, predators or parasites)

Genetic mechanisms that prevent species from becoming more widespread.
Abiotic/biotic factors are only limiting because a species has not evolved the morphological +
physiological + ecological means to overcome them.
9
SPECIES RANGES ON THE BOOKS
Biomes of the World
10
Fagus sylvatica
Corylus avellana
Quercus cerris
Castanea sativa
Quercus cerris
11
SPECIFICS: HOW DOES NICHE-BASED
MODELLING WORKS?
12
NICHE-BASED MODELLING: ASSUMPTIONS






Environmental factors drive species distribution
Species are in equilibrium with their environment
Limiting variables – are they really limiting?
Evidence for species dying/not reproducing due to climate
Collinearity of variables
Static vs dynamic approaches: data snapshot or time series
response?
13
CAUTIONARY NOTE ON MODELLING IN GENERAL

Risk of all models: the results of a model can never be better than the data used to construct and
run it. If the basic input data is incorrect, this error can be multiplied many times over.

Need to understand assumptions, explicit and implicit.

Models are an abstraction of reality, the accuracy and assumptions depend on the scale of the
model, and so although the model will improve our understanding of the processes involved, it
does not necessarily mean that the outputs are 100% truth.
14
SPECIFICS: VARIABLE SELECTION
Direct
Indirect
Definition
Variables with biological
relationship with study species
Variables that correlate with study species
because of correlation with series of
intermediate direct factors rather than
direct relationship
Example
Climate, nesting sites, soil nutrients
(plants), interacting species, site
isolation
Elevation, soil, topography, geology,
soil nutrients (animals)
Strength
Model structure easily interpreted in
biological meaningful terms.
Direct biological relationship should
generalize better to new areas, and
be more effective for climate change
modeling than indirect predictors.
Provides more info for conservation
management
Data sets widely available in GIS
Low cost, ease of collection
Can be effective predictors, i.e. elevation in
mountainous areas
Encompasses a range of correlated
variables so should: result in
parsimonious models if variable selection
applied, recording fewer variables
Weakness
Variables require greater effort to
record
Data sets may need to be estimated
for large spatial extents (using
indirect variables reducing
overall accuracy
Correlation with direct variables tend to be
location specific
Limited interpretation – biological meaning
inferred, resulting in increased
uncertainty
15
VARIABLES AND THEIR SELECTION
 Species only select their habitats in the broadest sense (Heglund 2002), and distribution patterns
are the cumulative result of a large number of fine scale decisions made to maximize resource
acquisition.
 The more accurately these fine-scale resources can be approximated and access quantified, the
better the model should perform if all models were equal.
 Predictions at broad scales can use broader environmental variables, often associated with the
fundamental niche.
 Finer scale predictions need to concern themselves more with those variables that determine the
realized niche.
Source: Pearson & Dawson 2003 16
CRITICALITIES
1. Environmental variables are often direct variables, such as precipitation or temperature.
2. Soil conditions include pH, texture, organic carbon and fertility, but these measures are often
difficult to obtain on an appropriate scale, since there can be considerable variation even within
an area.
3. In general, it is a good idea to avoid indirect measures of a variable, which is obviously a
challenge since much of a country is not monitored, and many such measures are not easily
taken.
4. Features such as slope and altitude allow some degree of projection into the future, since the
lapse rate (extent to which temperature changes with increasing altitude) may closely parallel
changing climatic conditions.
5. Solar radiation and wind, which are essential for plant growth responses and dispersal are both
particularly challenging to obtain accurate measures of.
17
WHAT DATA ARE AVAILABLE AND WHERE?
Museum/Herbarium data
Survey Atlas
Forest Inventories
Field data
Presence / Absence data
Taxonomic updates of older museum
Hydrological Bulletin
Field data
National/Regional projects
Geoportals
WebGIS repositories
Museum/Herbarium data
Survey Atlas
Expert Atlas
Presence / Absence data
Taxonomic updates of older museum
Hydrological Bulletin
Global Climatic repositories
International projects
Geoportals
WebGIS repositories
18
EXAMPLES: SPECIES
Forest Inventory of Sicily (Italy)
http://sif.regione.sicilia.it/
National Forest Inventory of France
http://inventaire-forestier.ign.fr/
National Forest Inventory of Spain
http://www.magrama.gob.es/
Species information system of Spain
http://www.anthos.es/
National Forest Inventory of Portugal
http://www.icnf.pt/portal/florestas/ifn
National Flora Database of Croatia
http://hirc.botanic.hr/fcd/
19
EXAMPLES: SPECIES
Forest Map of Europe
http://www.efi.int/portal/936
Forest Species distribution in Europe
http://forest.jrc.ec.europa.eu/
Ecosystems map of America
http://rmgsc.cr.usgs.gov/ecosystems/dataviewer.shtml
Global species distribution
http://www.gbif.org/
European forest genetic resources
http://www.euforgen.org
20
EXAMPLES: ENVIRONMENTAL VARIABLES
Digital Elevation Model of Italy
http://tinitaly.pi.ingv.it/
National Geoportal of Italy
http://www.pcn.minambiente.it/GN/
US Topographic maps
http://viewer.nationalmap.gov/launch/
National Geoportal of Spain
http://sig.magrama.es/geoportal/
National Climatic Registry of Spain
http://www.aemet.es/es/idi/clima/registros_climaticos
Andalusian Climatic Database
http://www.juntadeandalucia.es/medioambiente/
21
EXAMPLES: ENVIRONMENTAL VARIABLES
Global Digital Elevation Model
http://gdem.ersdac.jspacesystems.or.jp/
European Soil Map
http://eusoils.jrc.ec.europa.eu/
World Environmental Data Explorer
http://geodata.grid.unep.ch/
World Soil Grid Database
http://soilgrids1km.isric.org/index.html
World Climate Data
http://www.worldclim.org/
European Climate Dataset
http://eca.knmi.nl
Global Bioclimatic Data
http://www.climond.org
22
HOW DO WE CHOOSE A MODEL TYPE?












BioClimatic envelope (Bioclim)
Profile Techniques
Ordination (ENFA)
Environmental Distance (DOMAIN, CSM)
Ordinary Regression (Arc-SDM)
Generalised additive models (GAM)
Regression-based Techniques
Generalised linear models (GLM)
Classification and regression trees (CART, RF)
Genetic Algorithm (GARP)
Artificial neural networks (SPECIES)
Machine Learning Techniques
Boosted regression Trees (BRT)
Support Vector Machine (SVM)
Bayesian/Maximum Entropy (MaxEnt)
23
SOFTWARE AVAILABLE?
http://openmodeller.sourceforge.net/
http://mmweb.animal.net.cn/index.jsp
eHabitat Platform
http://dopa.jrc.ec.europa.eu/
https://www.cs.princeton.edu/~schapire/maxent/
http://www.nhm.ku.edu/desktopgarp/
24
Time
WHAT WE CAN DO WITH NICHE MODELLING?
25
STUDY OF PAST TO…
STUDY OF PRESENT TO…
STUDY OF FUTURE TO…
1.
2.
3.
4.
5.
6.
7.
Species ecological behaviour
Inner plasticity and buffering attitude
Response to climatic variables
Range stability and oscillations
Putative glacial refugia
Demographic history
In situ conservation strategies
1. Suitable surfaces for immediate
reforestations
2. Amount of sites suitable for reforestation
3. Human effect on the current landscape
4. Ecological corridors
5. Present vulnerability
1.
2.
3.
4.
5.
6.
Response of plants to changes
Bioclimatic niche reassessment
Extinction risk
Range shift and dynamics
Future conservation strategies
Areas to be selected as refugia
26
PRACTICAL APPLICATION: QUERCUS SUBER L.
27
CURRENT DISTRIBUTION
Vessella F., López-Tirado J., Simeone M.C., Schirone B., Hidalgo P.J. (2015)
28
PAST RECONSTRUCTION (120 KYR – PRESENT)
29
FUTURE PROJECTIONS AND CLIMATE CHANGE
30
31
PRACTICAL APPLICATION: POTENTIAL FORESTS IN ANDALUSIA
Present
32
2011-2040
33
2041-2070
34
2071-2100
35