Biomod.Turnover: Estimation of the Turnover for a given location and...
In BIOMOD: Ensemble platform for species distribution modeling

Description Usage Arguments Value Author(s) See Also Examples

This function allows to estimate the species lost and gained by pixel for the time slice and scenario considered, and the turnover induced by these (predicted) modifications.

1 2	Biomod.Turnover(CurrentPred = NULL, FutureProj = NULL, Turnover.Save = NULL) Biomod.Turnover.raster(CurrentPred = NULL, FutureProj = NULL, Turnover.Save = NULL)

`CurrentPred`	a matrix of n (number of species) columns OR a rasterstack giving the current state of the species in binary
`FutureProj`	a matrix of n (number of species) columns OR a rasterstack giving the future predictions of the species in binary
`Turnover.Save`	the name given to the new object storing the results

A 10 columns database or a 10 layers rasterstack is created (and saved under the name you gave it). The first four are relative numbers: - Disa represents the number of species predicted to disappear from the given pixel. - Stable0 is the number of species which are currently not in the given pixel and not predicted to migrate into it. - Stable1 represents the number of species currently occurring in the given pixel, and predicted to remains into the future. - Gain represents the number of species which are currently absent but predicted to migrate in the given pixel.

PercLoss, PercGain and Turnover are the relative percentages estimated as the following: - PercLoss = 100 x L/(SR) - PercGain = 100 x G/(SR) - Turnover = 100 x (L+G)/(SR+G)

Where SR is the current species richness, L the number of species lost and G those gained.

- CurrentSR represents the current modelled species richness in the given pixel. - FutureSR0Disp represents the future modelled species richness assuming no migration of species - FutureSR1Disp represents the future modelled species richness assuming migration (depending on the datasets given in input, if Migration has been used or not).

Wilfried Thuiller, Bruno Lafourcade

Biomod.RangeSize, Migration

## Not run:  
data(Sp.Env)
data(CoorXY)

#This command is necessary for the run of BIOMOD as a new dataframe is produced for the Models function
Initial.State(Response=Sp.Env[,11:18], Explanatory=Sp.Env[,4:10], 
IndependentResponse=NULL, IndependentExplanatory=NULL)

#Here are done 2 PA runs and 2 repetitions for each run. This will hence take several minutes.
Models(GLM = FALSE, TypeGLM = "quad", Test = "AIC", CTA = TRUE, CV.tree = 50, RF = TRUE, CV.ann = 2, 
   NbRunEval = 2, DataSplit = 80, Roc=TRUE, Optimized.Threshold.Roc=TRUE, Kappa=TRUE, TSS=TRUE, VarImport=0,
   NbRepPA=0, strategy="circles", coor=CoorXY, distance=2, nb.absences=1000)

Projection(Proj = Sp.Env[, 4:10], Proj.name="Current", GLM=TRUE, CTA=TRUE, RF=TRUE, BinKappa=TRUE, FiltKappa=TRUE, repetition.models=TRUE)

#Render Future projections under climate change scenario
data(Future1)
Projection(Proj = Future1[,4:10], Proj.name="Future1", GLM=TRUE, CTA=TRUE, RF=TRUE, BinKappa=TRUE, FiltKappa=TRUE, repetition.models=TRUE)

Ensemble.Forecasting(Proj.name= "Current", weight.method='Roc', PCA.median=TRUE, 
binary=TRUE, bin.method='Roc', Test=FALSE, decay=1.6, repetition.models=TRUE)

Ensemble.Forecasting(Proj.name= "Future1", weight.method='Roc', PCA.median=TRUE, 
binary=TRUE, bin.method='Roc', Test=FALSE, decay=1.6, repetition.models=TRUE)

load("proj.Future1/Total_consensus_Future1_Bin")
load("proj.Current/Total_consensus_Current_Bin")

Biomod.Turnover(CurrentPred = Total_consensus_Current_Bin[,,1], FutureProj = Total_consensus_Future1_Bin[,,1], Turnover.Save="Future1.Turnover.consensus")

#Visualise the results
Future1.Turnover.consensus[1150:1200,]


## End(Not run)