Geoscience Reference
In-Depth Information
Tool
How it works
What it tells us
Advantages & disadvantages
Species
distribution
models (SDMs)
(also known as
environmental
niche models,
bioclimatic
envelope
models)
Models quantify the
relationship between
the known distribution
of a species and
various environmental
parameters (climate, soils,
topography) to describe
the species' environmental
“envelope” or niche. This
relationship is then used
to project potential species
distributions in the future,
given a speciic climate
projection.
SDMs have been the principal tool to predict future
changes in species distributions. Most modeling
outputs suggest substantial declines in ranges may
occur, even under optimistic assumptions about the
capacity of species to disperse. Estimates of extinction
risk based on model output for multiple species,
indicate substantial increases in the number of species
at risk over next few decades (e.g. (Thomas et al.
2004a), (Warren et al. 2011).
SDMs have many well-discussed limitations
(reviewed in Elith and Leathwick 2009), including
the assumption that species and climate are in
equilibrium, and that other factors, such as species
interactions, are relatively unimportant in setting
distributional boundaries. Despite these limitations,
SDMs have successfully been used to simulate
observed range shifts of some species (Pearman
et al. 2008) and are considered a useful “irst cut”
tool within species risk assessments. SDMs are also
increasingly being coupled with other models based
on more detailed understanding of the demography
or physiology of particular species (where data are
available).
Process-based
models
Include key demographic
(such as growth, birth,
death, dispersal) or
physiological (such as
diapause, germination)
processes to simulate either
community or population
change.
The most widely applied models in the context of
climate change are forest gap models that simulate
long term dynamics of forest structure, biomass and
composition in response to climate and other drivers
(Bugmann 2001). Species-speciic mechanistic models
seek to describe the fundamental niche of a species
(e.g. Kearney and Porter 2004).
Process-based models can provide credible projections
of responses of vegetation stands or individual species
but few have been applied to large numbers of species
because they require either long-term data (in the
case of plant populations) or detailed physiological
and ecological understanding of the species of interest
(Lavergne et al. 2010).
Dynamic
Global
Vegetation
Models
(DGVMs)
Simulate changes in
vegetation and its
associated hydrological and
biogeochemical cycles as
a response to changes in
climate (Prentice, Harrison
and Bartlein 2011).
Models project large-scale changes in biomes (broad
vegetation types) in the future.
DGVMs are a powerful method to investigate
the relationship between biome-level change
and ecosystem services. Their limitations include
poor representation of functional diversity
within
communities, lack of simulation of changes in land use
or disturbances (except ire), and lack of information
at species level. Inclusion of CO
2
effects can reverse
the direction of the predictions generated (Loehle
2011) and differences between outputs of different
models also contribute to uncertainties in their utility
for planning.
