Performs climate-niche factor analysis using climate raster data and species presence data.
cnfa(x, s.dat, ...)# S4 method for GLcenfa,Raster
cnfa(
x,
s.dat,
filename = "",
progress = FALSE,
parallel = FALSE,
n = 1,
cl = NULL,
keep.open = FALSE,
...
)
# S4 method for GLcenfa,Spatial
cnfa(
x,
s.dat,
field,
fun = "last",
filename = "",
progress = FALSE,
parallel = FALSE,
n = 1,
cl = NULL,
keep.open = FALSE,
...
)
# S4 method for Raster,Raster
cnfa(
x,
s.dat,
scale = TRUE,
filename = "",
progress = FALSE,
parallel = FALSE,
n = 1,
cl = NULL,
keep.open = keep.open,
...
)
# S4 method for Raster,Spatial
cnfa(
x,
s.dat,
field,
fun = "last",
scale = TRUE,
filename = "",
progress = FALSE,
parallel = FALSE,
n = 1,
cl = NULL,
keep.open = FALSE,
...
)
Raster* object, typically a brick or stack with p climate
raster layers, or a GLcenfa object
RasterLayer, SpatialPolygons*, or SpatialPoints* object indicating species presence or abundance
Additional arguments for writeRaster
character. Optional filename to save the Raster* output to
file. If this is not provided, a temporary file will be created for large x
logical. If TRUE, messages and progress bar will be
printed
logical. If TRUE then multiple cores are utilized for the
calculation of the covariance matrices
numeric. Number of CPU cores to utilize for parallel processing
optional cluster object
logical. If TRUE and parallel = TRUE, the
cluster object will not be closed after the function has finished
field of s.dat that specifies presence or abundance. This
is equivalent to the field argument in rasterize
function or character. Determines what values to assign to cells
with multiple spatial features, similar to the fun argument in
rasterize. Options are 'first', 'last' (default),
and 'count' (see Details)
logical. If TRUE then the values of x will get
centered and scaled. Depending on the resolution of the climate data and
the extent of the study area, this can be quite time consuming. If running
this function for multiple species, it is recommended that the data be
scaled beforehand using the GLcenfa function
Returns an S4 object of class cnfa with the following components:
Original function call
Marginality factor. Vector of length p that describes the location of the species Hutchinsonian niche relative to the global niche
Magnitude of the marginality factor
Sensitivity factor. Vector of length p that describes the amount of sensitivity for each climate variable
Square root of the mean of the sensitivity factor
Named vector of eigenvalues of specialization for each CNFA factor
A p x p matrix describing the amount of marginality and specialization in each CNFA factor.
p x p species covariance matrix
p x p global covariance matrix
RasterBrick of transformed climate values, with p layers
Raster layer of weights used for CNFA calculation
The cnfa function is not to be confused with the
enfa function. enfa performs ENFA as described by Hirzel
et al. (2002) and Basille et al. (2008), and is offered as an alternative to
the enfa function in the adehabitatHS package. There are
several key differences between ENFA and CNFA.
Whereas ENFA returns a specialization factor that describes
the specialization in each ENFA factor, CNFA returns a
sensitivity factor sf that describes the sensitivity in each
environmental variable. This makes the sensitivity factor more
directly comparable to the marginality factor mf, because their
dimensions are identical. Sensitivity is calculated by a weighted sum
of the amount of specialization found in each CNFA factor, including
the marginality factor. As such, the sensitivity factor offers a more complete
measure of specialization than ENFA's specialization factor, which does
not calculate the amount of specialization found in the marginality factor.
As such, CNFA's overall sensitivity (found in the slot sensitivity)
offers a more complete measure of niche specialization than ENFA's overall
specialization (found in the slot specialization).
The default fun = 'last' gives equal weight to each occupied cell.
If multiple species observations occur in the same cell, the cell will only
be counted once. fun = 'count' will weight the cells by the number
of observations.
If there is too much correlation between the layers of x, the global
covariance matrix will be singular, and the overall marginality and overall
sensitivity will not be meaningful. In this case, a warning is issued,
and marginality and sensitivity are both returned as NA.
Rinnan, D. Scott and Lawler, Joshua. Climate-niche factor analysis: a spatial approach to quantifying species vulnerability to climate change. Ecography (2019): <doi:10.1111/ecog.03937>.
Basille, Mathieu, et al. Assessing habitat selection using multivariate statistics: Some refinements of the ecological-niche factor analysis. Ecological Modelling 211.1 (2008): 233-240.
Hirzel, Alexandre H., et al. Ecological-niche factor analysis: how to compute habitat-suitability maps without absence data?. Ecology 83.7 (2002): 2027-2036.
# NOT RUN {
mod1 <- cnfa(x = climdat.hist, s.dat = ABPR, field = "CODE")
# using GLcenfa as an initial step
# for multi-species comparison
glc <- GLcenfa(x = climdat.hist)
mod2 <- cnfa(x = glc, s.dat = ABPR, field = "CODE")
# same results either way
all.equal(m.factor(mod1), m.factor(mod2))
all.equal(s.factor(mod1), s.factor(mod2))
# }
Run the code above in your browser using DataLab