README.md

In jantumajer/IncrementR: Analysing tree height growth in R

IncrementR package: Analysing height growth of trees in R

This package is designed to be used by members of dendrochronological comunity, especially those, who analyse height growth of trees and lateral elongation of ramets of prostrate shrubs. Specific tools were developed for input data in the form of tree-ring widths extracted from series of successive stem height levels. The main computed parameters provided by the package include height growth along the stem, changes of stem eccentricity, and stem taper. The appropriate determination of average height growth depends on correct estimation of the number of tree-rings in different stem levels, which might be complicated by missing rings in off-pith cores. The presented package therefore contains also functions for common procedures for estimation of the number of missing tree-rings. Graphical visualisation of most results is possible.

Installing instructions

Technical note article describing individual steps of the use of IncrementR package was published in Dendrochronologia Journal. Its full reference is Kašpar J., Tumajer J., Treml V. (2019): IncrementR: analysing height growth of trees and shrubs in R. Dendrochronologia. 53: 48-54. DOI 10.1016/j.dendro.2018.11.001. We invite you to read the article before running the package at https://doi.org/10.1016/j.dendro.2018.11.001

Installing instructions

Recently, the package is available only from GitHub using devtools package. Use following syntax to install and load it in your R.

install.packages("devtools")
library(devtools)

install_github("jantumajer/IncrementR")
library(IncrementR)

Example data

Instalation of IncrementR package contains sample data frames related to apical growth of single tree located in Hruby Jesenik Mts. (Czech Republic). Data frame trw stores cross-dated tree-ring widths obtained by coring tree stem in successive levels from stem base to the top. Data frame po contains estimates of the distance between last measured tree-ring and pith for off-pith cores (in mm). Finally, data frame meta stores information about distance of individual sampling levels from the ground (in cm).

Use folowing syntax to access sample data:

data(IncrementR_data)
View(trw); View(meta); View(po)

Function syntax and descriptions

Eccentricity(trw.series, complete=FALSE)

Function Eccentricity calculates eccentricity indices according to Braam et al. (1987), Alestalo (1971) and Schweingruber (1996) for each tree-ring. This function requires the input of four cross-dated series per tree and/or sampling level (argument trw.series). Argument complete defines whether four samples were obtained for all sampling levels. If some of required samples are missing, then complete=F. The output of this function is a list containing three data frames with values of eccentricity indices for each method and each orientation around the stem circumference.

EMR(trw.series, p.off, nyrs, method="TRW")

Function EMR estimates the number of missing tree-rings between the last measured tree-ring and the pith of each core. Argument trw.series defines data frame with series for which the estimation shall be performed. Argument nyrs defines the number of innermost measured tree-rings, whose average will be used to estimate mean tree-ring width of missing rings. Argument p.off contains data frame with an information about estimated distance in mm of the last known tree-ring to the pith (stored in the column with name “P.OFFSET”). Last parameter is method, which defines the method used for estimation. User can select from “TRW”, “BAI” and “Both”, which is and average value of “BAI” and “TRW” (for detailed description of the approaches see Altman et al. 2016). The output of this function is a data frame with the number of missing tree-rings that should be added to each core.

RMR(trw.series, mr.estimate, nyrs=5, nsph=4)

Function RMR may be applied following the EMR processing, and serves to add the number of missing rings to pith into data frame with tree-ring width series. Arguments of RMR function include trw.series (contains tree-ring series to be modified), mr.etimate (an output of EMR function for trw.series), nyrs (the number of innermost measured tree-rings, whose average will be used to estimate mean tree-ring width of missing rings) and nsph, which defines how many cores were sampled at each level (options are 1, 2 and 4). If nsph>1, the estimation of the number of missing rings from EMR function is corrected so as to have the same number of tree-rings in all series coming from the same level (for the description of approach see apical function). The output of this function is a new data frame, which contains tree ring-series with added “virtual” tree-rings between pith and the first measured tree-ring.

apical(trw.series, meta, mr.estimate)

The function apical first creates datasets containing number of TotalRings (Measured + Missing rings nearby pith as derived by EMR) for each core and each level along the stem. Inputs include trw.series, meta file with sampling heights of each sample in cm (column “Level_cm”) and mr.estimate (the output of EMR function). The output is the list of two data frames; the first one (N.ring_Core) contains series codes with the number of measured, missing and total rings. The other one (N.ring_Level) contains codes of each sampling level with the number of measured, missing and total rings, together with the level height. Moreover, the mean height growth velocity between two subsequent levels (in cm.yr-1) is calculated and stored in “Speed.cmyr” column and error of its estimation (see Figure 1 for description) is stored in “MeanSpeedError.cmyr” column. If the number of estimated tree-rings differs from the other cores of the same height level, two rules are applied to estimate number of tree-rings in a given level. First, cores with pith presence are prioritized; and if the number of tree-rings of such cores differs, median of their TotalRings is used to estimate the number of tree-rings in height level. Secondly, if none of cores contains the pith, median of TotalRings of all series from the same level is used.

taperCalcul(trw.series, meta)

Function taperCalcul calculates taper and taper angle between subsequent sampling heights for all trees. The inputs are trw.series (containing tree-ring series) and meta (metadata containing the sampling heights in cm in column “Level_cm”).

BAICalcul (trw.series)

Function BAIcalculation creates a series of basal area increments for each sampling level. It requires four cores per sampling point (i.e. level). First, it automatically averages all series coming from the same tree and sampling level and, subsequently, applies function BAI.in from dplR package to calculate area increments. The only input argument of this function is trw.series (containing cross-dated tree-ring width series and, optionally, modelled missing tree rings near to the pith using the functions EMR and RMR). The output of this function is a data frame containing a series of basal area increments for each sampling level.

drawEccentricityGraph (trw, ecc, meta, plot, tree, withEccentricity=T, method="Schweingruber")

First graphical function of “IncrementR” package is drawEccentricityGraph. Argument trw represents a data frame with tree-ring width series (possibly) previously corrected for missing rings nearby the pith using RMR. Argument ecc contains output of eccentricity function. Arguments plot and tree specify a tree which will be drawn, argument meta refers to a metadata file with sampling heights of each level in cm (column “Level_cm”). Logical argument withEccentricity specifies whether the graph will show only stem profile or stem profile with eccentricity indices; and, finally, method defines index to be drawn (available options include “Alestalo”, “Braam” and “Schweingruber”). The output consists of two graphs of stem profiles (E-W and S-N) with respective eccentricity indices of each sampling level.

drawCrossSectionProfile (trw.series, plot, tree, level, show.legend=T)

Cross section profile based on approximation of cross section as an ellipse can be drawn for each sampling level. The function drawCrossSectionProfile has arguments specifying the cross section (plot, tree and level) to be plot; trw.series (data frame containing tree-ring width series, possibly previously corrected using RMR); and logical argument show.legend specifying if the legend (colours of different calendar years) will be drawn.

drawTaper(taperFile, plot=1,tree=1, variant="Taper")

Taper between consecutive sampling levels can be plot by taperCalculation function as a line-chart. Argument taperFile is the output of taperCalculation function; arguments plot and tree specify tree to have the taper drawn; and logical argument variant specifies whether taper or taper angle will be drawn (available options are “Taper” or “Angle”).

drawApicalData(trw.series, apicalData, plot=1, tree=1)

Arguments of drawApicalData function include trw.series (data frame containing tree-ring width series); apicalData (output of apical function); and plot and tree (specifying the tree to be visualized). The resulting plot consist of two parts – the left panel shows the number of “Total” tree-rings (i.e., measured tree-rings + estimated missing tree-rings nearby pith) in each sampling level along the stem; the right one shows mean height growth velocity between two consecutive sampling levels (points) together with mean errors of its estimates (error bars).

drawBAI(baiFile, plot=1,tree=1,logscale=F, show.legend=T)

The output of function BAICalcul can be plotted using the drawBAI function as a line chart with the x-axis showing calendar years and the y-axis showing the basal area increments of individual sampling levels from a single tree. The arguments of this function include baiFile (output of the BAIcalculation function); arguments plot and tree specify the tree to be plotted. Logical argument logscale specifies whether the result is presented on a log-scaled y-axis, and logical argument show.legend specifies whether the legend is drawn.

jantumajer/IncrementR documentation built on Dec. 6, 2020, 11:22 a.m.