BIOdry (version 0.3)

muleMan: Multilevel correlograms

Description

Multilevel Mantel correlograms between two modelFrame objects.

Usage

muleMan(rd, cd, rd.var = NULL, cd.var = NULL, ...)

Arguments

rd

list or dataframe, such as that produced by modelFrame, containing the modeled tree growth.

cd

list or dataframe, such as that produced by modelFrame, with correspondatn modeled aridity (see details).

rd.var

character. Column name of the processed variable in coderd. If NULL then first column in rd is processed.

cd.var

character. Column name of the processed variable in cd. If NULL then its first column is used.

...

Further arguments in mgram

Value

groupedData object) with computed Mantel correlations.

Details

Function mgram in package ecodist is implemented to compare two modelFrame objects, with the first object containing modeled fluctuations of tree growth, and the second one being the modeled fluctuations of aridity. Correspondant aridity model should have at least one level in common with the modeled tree growth (see example).

References

Lara W., F. Bravo, D. Maguire. 2013. Modeling patterns between drought and tree biomass growth from dendrochronological data: A multilevel approach. Agric. For. Meteorol., 178-179:140-151.

Examples

## Fluctuations of tree growh and aridity are modeled and
## compared.

##Multilevel data frame of tree-ring widths:
data(Prings05,envir = environment())
## Radial increments measured on 2003:
data(Pradii03,envir = environment())    
## Monthly precipitations and temperatures:
data(PTclim05,envir = environment())

## Modeling fluctuations of aridity 
cf <- modelFrame(rd=PTclim05,
                 lv = list('year','year'),
                 fn = list('moveYr','wlai'),
                 form = 'lmeForm')
head(cf$resid)
summary(cf$model)

## Modeling fluctuations of tree growth
ar <- modelFrame(Prings05, y = Pradii03,
                 form = 'tdForm', on.time = TRUE,
                 MoreArgs = list(only.dup = TRUE,
                                 mp = c(1,1),un = c('mm','cm'),z = 2003))
head(ar$resid)
summary(ar$model)

## Multilevel correlogram:
mancor <- muleMan(ar,cf,nperm = 10^3)
head(mancor)

## Vector of significances (p < 0.05): 
sig <- with(mancor,ifelse(pval < 0.05,TRUE,FALSE))
## Plotting the multilevel correlograms with correspondent
## significances:
plot(mancor,
     groups = sig,
     pch = c(21,19),
     grid = FALSE,
     abline = list(h = 0, lty = 2, lwd = 0.5),
     layout = c(4,2))