DESCRIPTION
r.mess computes the "Multivariate Environmental
Similarity Surfaces" (MESS) in GRASS using R as backend. The
MESS index was proposed by Elith et al (2010) and is also
implemented in the Maxent software. The MESS approach
can be described as follows (from Elith et al 2010):
"The multivariate environmental similarity surface (MESS) calculation
represents how similar a point is to a reference set of points, with
respect to a set of predictor variables (V1, V2, ...). The values in
the MESS are influenced by the full distribution of the reference points, so that
sites within the environmental range of the reference points but in
relatively unusual environments will have a smaller value than those in
very common environments."
This module will also compute the individual environmental
similarity surfaces (IESS), which represents how similar a point is to
a set of reference points for each of the input variable. MESS
is then simply calculated as the minimum of its similarity with respect
to each variable.
Any sample of interest can be used for the reference set. For example,
it can be occurrence records for the species; a sample of a region, or a sample of
current climate conditions in a given area. The input layer representing the
reference distribution can be a vector point layer or a raster layer.
The IESS can have negative values – these are sites where the variable
has a value that is outside the range in the reference set. A negative MESS
thus represents sites where at least one variable has a value that is
outside the range of environments over the reference set, so these are
novel environments.
In addition to the MESS, which is the minimum(IESS), the r.mess
function also allows to compute mean and medium of the IESS layers.
NOTES
Digits should reflect the level of precision of the input layers. I.e., if
values of the input layers are recorded with three digits behind the comma,
Digits should be set to 0.001 or smaller.
SEE ALSO
There is also a similar function implemented for R using the R package
raster, and which is part of the dismo package in R now. See here.
for more information and the script.
AUTHOR
Paulo van Breugel
REFERENCES
- Elith, J., Kearney, M., & Phillips, S. 2010. The art of
modelling range-shifting species. Methods in Ecology and Evolution 1:
330–342.
Last changed: $Date$