meteSAR {meteR}

Compute METE species area relationship (SAR)

Description

Uses raw data or state variables to calculate METE SAR and EAR (endemics area relatiohsip) as well as compute the observed SAR or EAR from data, if provided

Usage

meteSAR(spp, abund, row, col, x, y, S0 = NULL, N0 = NULL, Amin, A0,
  upscale = FALSE, EAR = FALSE)

Arguments

spp

vector of species identities

abund

numberic vector abundances associated with each record

row

identity of row in a gridded landscape associated with each record, or desired number of rows to divide the landcape into

col

identity of column in a gridded landscape associated with each recod, or desired number of columns to divide the landcape into

x

the x-coordinate of an individual if recorded

y

the y-coordinate of an individual if recorded

S0

total number of species

N0

total abundance

Amin

the smallest area, either the anchor area for upscaling or the desired area to downscale to

A0

the largest area, either the area to upscale to or the total area from which to downscale

upscale

logical, should upscaling or downscaling be carried out

EAR

logical, should the EAR or SAR be computed

Details

Currently only doublings of area are supported. Predictions and comparison to data can be made via several options. If spp and abund are not provided then only theoretical predictions are returned without emperical SAR or EAR results. In this case areas can either be specified by providing Amin and A0 from which a vector of doubling areas is computed, or my providing row, col and A0 in which case row and col are taken to be the number of desired rows and columns used to construct a grid across the landscape. If data are provided in the form of spp and abund then either row and col or x and y must be provided for each data entry (i.e. the length of row and col or x and y must equal the length of spp and abund). If x and y are provided then the landscape is gridded either by specifying Amin (the size of the smallest grid cell) or by providing the number or desired rows and columns via the row and col arguments.

SARs and EARs can be predicted either interatively or non-iteratively. In the non-iterative case the SAD and SSAD (which are used to calculate the SAR or EAR prediction) are derived from state variables at one anchor scale. In the iterative approach state variables are re-calculated at each scale. Currently downscaling and upscaling are done differently ( downscaling is only implemented in the non-iterative approach, whereas upscaling is only implemented in the iterative approach). The reason is largely historical (downscaling as originally done non-iteratively while upscaling was first proposed in an iterative framework). Future implementations in meteR will allow for both iterative and non-iterative approaches to upscaling and downscaling. While iterative and non-iterative methods lead to slightly different predictions these are small in comparison to typical ranges of state variables (see Harte 2011).

Value

an object of class meteRelat with elements

pred

predicted relationship; an object of class sar

obs

observed relationship; an object of classsar

Author(s)

Andy Rominger <[e-mail]>, Cory Merow

References

Harte, J. 2011. Maximum entropy and ecology: a theory of abundance, distribution, and energetics. Oxford University Press.

See Also

sad, meteESF, metePi

Examples

## Not run: 
data(anbo)

## using row and col from anbo dataset
anbo.sar1 <- meteSAR(anbo$spp, anbo$count, anbo$row, anbo$col, Amin=1, A0=16)
plot(anbo.sar1)

## using simulated x, y data
anbo.sar2 <- meteSAR(anbo$spp, anbo$count, x=anbo$x, y=anbo$y, row=4, col=4)
plot(anbo.sar2)

## using just state variable
thr.sar <- meteSAR(Amin=1, A0=16, S0=50, N0=500)

## End(Not run)