ensemble.novel: Mapping of novel environmental conditions (areas where some of the environmental conditions are outside the range of environmental conditions of a reference area).

Description

Function ensemble.novel creates the map with novel conditions. Function ensemble.novel.object provides the reference values used by the prediction function used by predict .

Usage

ensemble.novel(x = NULL, novel.object = NULL,
    RASTER.object.name = novel.object$name, RASTER.stack.name = x@title,
    RASTER.format = "raster", RASTER.datatype = "INT1S", RASTER.NAflag = -127, 
    KML.out = FALSE, KML.maxpixels = 100000, KML.blur = 10,
    CATCH.OFF = FALSE)
ensemble.novel.object(x = NULL, name = "reference1", mask.raster = NULL,
    quantiles = FALSE, probs = c(0.05, 0.95), factors = NULL)

Arguments

RasterStack object (stack) containing all environmental layers for which novel conditions should be calculated. With ensemble.novel.object, x can also be a data.frame.

novel.object

Object listing minima and maxima for the environmental layers, used by the prediction function that is used internally by predict. This object is created with ensemble.novel.object.

RASTER.object.name

First part of the names of the raster file that will be generated, expected to identify the area and time period for which ranges were calculated

RASTER.stack.name

Last part of the names of the raster file that will be generated, expected to identify the predictor stack used

RASTER.format

Format of the raster files that will be generated. See writeFormats and writeRaster.

RASTER.datatype

Format of the raster files that will be generated. See dataType and writeRaster.

RASTER.NAflag

Value that is used to store missing data. See writeRaster.

KML.out

If TRUE, then kml files will be saved in a subfolder 'kml/zones'.

KML.maxpixels

Maximum number of pixels for the PNG image that will be displayed in Google Earth. See also KML.

KML.blur

Integer that results in increasing the size of the PNG image by KML.blur^2, which may help avoid blurring of isolated pixels. See also KML.

CATCH.OFF

Disable calls to function tryCatch.

name

Name of the object, expect to expected to identify the area and time period for which ranges were calculated and where no novel conditions will be detected

mask.raster

RasterLayer object (raster) that can be used to select the area for which reference values are obtained (see mask)

quantiles

If TRUE, then replace minima and maxima with quantile values. See also quantile and quantile)

probs

Numeric vector of probabilities [0, 1] as used by quantile and quantile)

factors

vector that indicates which variables are factors; these variables will be ignored for novel conditions

Value

Function ensemble.novel.object returns a list with following objects:

minima

minima of the reference environmental conditions

maxima

maxima of the reference environmental conditions

name

name for the reference area and time period

Details

Function ensemble.novel maps zones (coded '1') that are novel (outside the minimum-maximum range) relative to the range provided by function ensemble.novel.object. Values that are not novel (inside the range of minimum-maximum values) are coded '0'. In theory, the maps show the same areas that have negative Multivariate Environmental Similarity Surface (MESS) values ((mess))

Examples

Run this code

# NOT RUN {
# }
# NOT RUN {
# get predictor variables
library(dismo)
predictor.files <- list.files(path=paste(system.file(package="dismo"), '/ex', sep=''),
    pattern='grd', full.names=TRUE)
predictors <- stack(predictor.files)
predictors <- subset(predictors, subset=c("bio1", "bio5", "bio6", "bio7", "bio8", 
    "bio12", "bio16", "bio17"))
predictors
predictors@title <- "base"

# reference area to calculate environmental ranges
ext <- extent(-70, -50, -10, 10)
extent.values2 <- c(-70, -50, -10, 10)
predictors.current <- crop(predictors, y=ext)
predictors.current <- stack(predictors.current)

novel.test <- ensemble.novel.object(predictors.current, name="noveltest")
novel.test
novel.raster <- ensemble.novel(x=predictors, novel.object=novel.test, KML.out=T)
novel.raster

plot(novel.raster)
# no novel conditions within reference area
rect(extent.values2[1], extent.values2[3], extent.values2[2], extent.values2[4])

# use novel conditions as a simple species suitability mapping method
# presence points
presence_file <- paste(system.file(package="dismo"), '/ex/bradypus.csv', sep='')
pres <- read.table(presence_file, header=TRUE, sep=',')[,-1]
pres.data <- data.frame(extract(predictors, y=pres))

# ranges and maps
Bradypus.ranges1 <- ensemble.novel.object(pres.data, name="Bradypus", quantiles=F)
Bradypus.ranges1
Bradypus.novel1 <- ensemble.novel(x=predictors, novel.object=Bradypus.ranges1, KML.out=T)
Bradypus.novel1

par.old <- graphics::par(no.readonly=T)
graphics::par(mfrow=c(1,2))

# suitable where there are no novel conditions
raster::plot(Bradypus.novel1, breaks=c(-0.1, 0, 1), col=c("green", "grey"), 
    main="Suitability mapping using minimum to maximum range")
points(pres[, 2] ~ pres[, 1], pch=1, col="red", cex=0.8)

# use 90 percent intervals similar to BIOCLIM methodology
Bradypus.ranges2 <- ensemble.novel.object(pres.data, name="BradypusQuantiles", quantiles=T)
Bradypus.ranges2
Bradypus.novel2 <- ensemble.novel(x=predictors, novel.object=Bradypus.ranges2, KML.out=T)
Bradypus.novel2
raster::plot(Bradypus.novel2, breaks=c(-0.1, 0, 1), col=c("green", "grey"), 
    main="Suitability mapping using quantile range")
points(pres[, 2] ~ pres[, 1], pch=1, col="red", cex=0.8)

graphics::par(par.old)

# deal with novel factor levels through dummy variables
predictors <- stack(predictor.files)
biome.layer <- predictors[["biome"]]
biome.layer
ensemble.dummy.variables(xcat=biome.layer, most.frequent=0, freq.min=1,
    overwrite=TRUE)

predictors.dummy <- stack(predictor.files)
predictors.dummy <- subset(predictors.dummy, subset=c("biome_1", "biome_2",  "biome_3",  
    "biome_4", "biome_5", "biome_7",  "biome_8",  "biome_9", 
    "biome_10", "biome_12", "biome_13", "biome_14"))
predictors.dummy
predictors.dummy@title <- "base_dummy"

predictors.dummy.current <- crop(predictors.dummy, y=ext)
predictors.dummy.current <- stack(predictors.dummy.current)

novel.levels <- ensemble.novel.object(predictors.dummy.current, name="novellevels")
novel.levels
novel.levels.raster <- ensemble.novel(x=predictors.dummy, novel.object=novel.levels, 
    KML.out=T)
novel.levels.raster

novel.levels.quantiles <- ensemble.novel.object(predictors.dummy.current, quantiles=TRUE,
    name="novellevels_quantiles")
novel.levels.quantiles
novel.levels.quantiles.raster <- ensemble.novel(x=predictors.dummy, 
    novel.object=novel.levels.quantiles, KML.out=T)
novel.levels.quantiles.raster

# difference in ranges for variables with low frequencies
background <- dismo::randomPoints(predictors.dummy.current, n=10000, p=NULL, excludep=F)
extract.data <- extract(predictors.dummy.current, y=background)
colSums(extract.data)/sum(extract.data)*100
novel.levels
novel.levels.quantiles

par.old <- graphics::par(no.readonly=T)
graphics::par(mfrow=c(1,2))
raster::plot(novel.levels.raster, breaks=c(-0.1, 0, 1), col=c("grey", "green"), 
    main="novel outside minimum to maximum range")
rect(extent.values2[1], extent.values2[3], extent.values2[2], extent.values2[4])
raster::plot(novel.levels.quantiles.raster, breaks=c(-0.1, 0, 1), col=c("grey", "green"), 
    main="novel outside quantile range")
rect(extent.values2[1], extent.values2[3], extent.values2[2], extent.values2[4])
graphics::par(par.old)

# }

Run the code above in your browser using DataLab