IUCN shapefiles
To begin, download the species distribution shapefiles from the IUCN
website.
You can import this data using the terra::vect or
sf::st_read functions. For the purpose of this tutorial, I
will be utilizing the distribution data for frogs belonging to the
Phyllomedusa genus, which is conveniently pre-loaded within the
letsR package.
We can plot the data to see how it looks like.
# Plot
## Color settings and assignment
colors <- rainbow(length(unique(Phyllomedusa$binomial)),
alpha = 0.5)
position <- match(Phyllomedusa$binomial,
unique(Phyllomedusa$binomial))
colors <- colors[position]
## Plot call
plot(sf::st_geometry(Phyllomedusa), col = colors, lty = 0,
main = "Spatial polygons of Phyllomedusa")
data("wrld_simpl")
plot(sf::st_geometry(wrld_simpl), add = TRUE)
Quick start
Next step, we can use the function lets.presab to
convert species’ ranges (in shapefile format) into a presence-absence
matrix based on a user-defined grid system. A simple way to do this is
to define the extent and resolution of the grid.
Please be aware that when working with shapefiles containing numerous
species or opting for a high-resolution grid, the function’s execution
may become notably slow. In such instances, it is advisable to monitor
the relative running time of the analysis by enabling the
count = TRUE argument.
The lets.presab function yields a
PresenceAbsence object (unless
show.matrix = TRUE, in which case only a presence-absence
matrix is returned). This object is essentially a list comprising a
presence-absence matrix, a raster containing geographical information,
and the species names. For additional details, refer to
?PresenceAbsence. To obtain summary information about the
generated Presence-Absence Matrix (PAM), the summary
function can be employed.
##
## Class: PresenceAbsence
## _ _
## Number of species: 32
## Number of cells: 1187
## Cells with presence: 1187
## Cells without presence: 0
## Species without presence: 0
## Species with the largest range: Phyllomedusa hypochondrialis
## _ _
## Grid parameters
## Resolution: 1, 1 (x, y)
## Extention: -93, -29, -57, 15 (xmin, xmax, ymin, ymax)
## Coord. Ref.: +proj=longlat +datum=WGS84 +no_defsYou can also use the plot function directly to the PAM
object.
The plot function also allow users to plot specific
species distributions. For example, we can plot the map of
Phyllomedusa hypochondrialis:
As said before, the PAM object contains the actual presence absence matrix, to access it we can use the following code:
The first two columns of the matrix contain the longitude (x) and latitude (y) of the cells’ centroid, the following columns include the species’ presence(1) and absence(0) information.
## Longitude(x) Latitude(y) Phyllomedusa araguari
## [1,] -74.5 11.5 0
## [2,] -69.5 11.5 0
## [3,] -68.5 11.5 0
## [4,] -75.5 10.5 0
## [5,] -74.5 10.5 0Using different projections
Some users may want to use different projections to generate the
presence absence matrix. The lets.presab function allow
users to do it by changing the crs.grid argument. Check the
example using the South America Equidistant Conic projection.
pro <- paste("+proj=eqdc +lat_0=-32 +lon_0=-60 +lat_1=-5",
"+lat_2=-42 +x_0=0 +y_0=0 +ellps=aust_SA",
"+units=m +no_defs")
SA_EC <- terra::crs(pro)
PAM_proj <- lets.presab(Phyllomedusa, xmn = -4135157,
xmx = 4707602,
ymn = -450000, ymx = 5774733,
res = 100000,
crs.grid = SA_EC)##
## Class: PresenceAbsence
## _ _
## Number of species: 32
## Number of cells: 1396
## Cells with presence: 1396
## Cells without presence: 0
## Species without presence: 0
## Species with the largest range: Phyllomedusa hypochondrialis
## _ _
## Grid parameters
## Resolution: 1e+05, 1e+05 (x, y)
## Extention: -4135157, 4664843, -450000, 5750000 (xmin, xmax, ymin, ymax)
## Coord. Ref.: +proj=eqdc +lat_0=-32 +lon_0=-60 +lat_1=-5 +lat_2=-42 +x_0=0 +y_0=0 +ellps=aust_SA +units=m +no_defsplot(PAM_proj)
# Add projected country boundaries
data("wrld_simpl")
plot(sf::st_transform(sf::st_geometry(wrld_simpl), pro), add = TRUE)
Other features
The function lets.presab has some other useful
arguments. For instance, users may wish to exclude regions where species
are extinct or retain only the breeding ranges. The
presence, origin, and seasonal
arguments enable users to filter species distributions based on the IUCN
classification of different parts of a species’ range. To find the
specific values for these arguments, consult the IUCN
metadata files.
In certain scenarios, it proves advantageous to consider a species
present in a cell only if it covers more than a specified percentage
value. Users can customize this threshold using the cover
argument. It’s important to note that initially, this option is
exclusively available when the coordinates are in degrees
(longitude/latitude). However, with the latest update on GitHub, users can now
employ the cover argument with other projections as
well.
# 90% cover
PAM_90 <- lets.presab(Phyllomedusa, xmn = -93,
xmx = -29, ymn = -57,
ymx = 15, res = 1,
cover = 0.9)
plot(PAM_90)
Observing the plot above, it’s evident that cells near the continent’s border no longer reflect the presence of the species.
When generating multiple PresenceAbsence objects for
distinct groups, users might prefer to maintain a consistent grid. To
achieve this, it’s crucial to retain the
remove.cells = FALSE argument, preventing any modification
to the grid. Conversely, setting remove.cells = TRUE
excludes cells with a value of zero in the final matrix, meaning sites
where no species are present won’t be included.
PAM_keep_cells <- lets.presab(Phyllomedusa, xmn = -93,
xmx = -29, ymn = -57,
ymx = 15, res = 1,
remove.cells = FALSE)You can now employ the summary function to confirm
whether the empty cells were retained.
##
## Class: PresenceAbsence
## _ _
## Number of species: 32
## Number of cells: 4608
## Cells with presence: 1187
## Cells without presence: 3421
## Species without presence: 0
## Species with the largest range: Phyllomedusa hypochondrialis
## _ _
## Grid parameters
## Resolution: 1, 1 (x, y)
## Extention: -93, -29, -57, 15 (xmin, xmax, ymin, ymax)
## Coord. Ref.: +proj=longlat +datum=WGS84 +no_defsAdditionally, for users intending to retain species that do not occur
in any cell of the grid, it is essential to configure
remove.sp = FALSE.
