Skip to contents

Functions forest_mergeTrees and forest_mergeShrubs merge cohorts of a forest object. Function forest_reduceToDominant performs a strongest simplification of plant cohorts (see details).

Usage

forest_mergeTrees(x, byDBHclass = TRUE, keepCohortsWithObsID = FALSE)

forest_mergeShrubs(x, byHeightclass = TRUE, keepCohortsWithObsID = FALSE)

forest_reduceToDominant(x, SpParams)

Arguments

x

An object of class forest.

byDBHclass

Logical flag to indicate that 5-cm tree DBH classes should be kept separated.

keepCohortsWithObsID

Logical flag to indicate that cohorts with non-missin ObsID should be spared from merging.

byHeightclass

Boolean flag to indicate that 10-cm shrub height classes should be kept separated.

SpParams

A data frame with species parameters (see SpParamsDefinition and SpParamsMED).

Value

Another forest object with simplified structure/composition, depending on the function.

Details

Tree DBH classes are defined in 5-cm intervals, whereas shrub height classes are defined in 10-cm intervals. Tree DBH and shrub height classes are defined up to a specific size (i.e. larger plants are not merged) corresponding to 52.5 cm and 90 cm, respectively.

Function forest_reduceToDominant simplifies the input forest to the tree cohort of highest LAI, among those of the tree species with highest LAI. The leaf area index of the whole tree layer will be attributed to the chosen cohort. The same is performed for the shrub layer.

Author

Miquel De Cáceres Ainsa, CREAF

Examples

# Example forest data
data("exampleforest")

# Reduce to dominant tree and dominant shrub
reduced <- forest_reduceToDominant(exampleforest, SpParamsMED)

# Check that overall LAI does not change
stand_LAI(exampleforest, SpParamsMED)
#> [1] 1.758585
stand_LAI(reduced, SpParamsMED)
#> [1] 1.758585