Problem: estimate coarse root biomass based on species/diameter data.

Approach: Jenkins et al. (Forest Science, 2003) developed component biomass ratios based on species group (hardwood vs. softwood) and diameter. These are general, and do not include an option for ‘stump’ but can be applied consistently to the suite of species we have in the HF C Synthesis, and the ratios can be applied to the good species-specific aboveground biomass we’ve estimated already.

For the few species with published stand-alone, species-specific equations for stump + root, or roots only, we could alternatively use these and the general approach for the rest.

Here is a comparison of the generalized ratio vs. species-specific component approach for a few species that have species-specific component allometries.

alltrees<-read.csv("c:/data/biomass-by-tree-gmr-29nov16.csv") 
alltrees["rootratio"] <- NA
source('C:/Users/aabarker/Dropbox/HF C Synthesis/Data/live trees/live tree data & code/biomasscoarserootshwsw.R')
#Calculate root ratio based on species group and diameter, using the 'rootratio' funcion
for(i in 1:nrow(alltrees)) {
  alltrees$rootratio[i] <- rootratio(alltrees$species[i], alltrees$dbh[i])
}

#take out the mortality
at<-subset(alltrees, alltrees$cond!='M' & alltrees$dbh != 0)

tiny<-subset(at, at$dbh<2.5) #166 or 116937 nondead nonzero (0.1%)
# because the ratios are meant for trees >2.5cm dbh, replace those rootratios with NA
# that is, we're ignoring the root biomass for the 166 trees with dbh<2.5cm, but this is negligible.
at$rootratio<-ifelse(at$dbh<2.5, NA, at$rootratio)


#Make a column for coarse root biomass (kg)
at$RootKgRatioMethod<- at$biomassAF20*at$rootratio

Depending on the tree diameter and species group (hardwood vs. softwood), coarse roots are estimated to be ~19-25% of the aboveground biomass.

## Warning: Removed 166 rows containing non-finite values (stat_bin).

Test 1. Red maple.

This species comprises 30% of the stems in the live tree database. A species-specific allometric equation is available for root+stump. Compare that to the hardwood ratio approach.

rm<-subset(at, species == 'ACRU')
rm$StumpRootComponentEq<- (exp(-0.8947 + (2.4726*(log(rm$dbh/2.54)))))/2.20462
rm$proportion<-rm$RootKgRatioMethod/rm$StumpRootComponentEq

Red maple: species specific equation for stump + root vs. root ratio method

The ratio method gives a smaller estimate than the root+stump species-specific component equation for trees with dbh<8.6 cm, and increasingly larger for trees with dbh>8.6cm. However, for most of the trees, the methods are within +/- 20%.

Overall, the ratio method summed coarse root biomass over all the red maple stems is 88% of the summed root+stump biomass estimated by the red maple component equation.

Test 2. Yellow birch.

This species comprises 4% of the stems in the live tree database, but there are independent species-specific allometric equations available for root+stump, and for roots alone. So, we can compare amongst these three approaches for this species.

yb<-subset(at, species =='BEAL')
yb$stumproot<- (exp(-0.2963 + (2.3223*(log(yb$dbh/2.54)))))/2.20462
yb$rootonly<-(10^(1.6 + (2.3156*(log10(yb$dbh)))))/1000
yb$proportionSR<-yb$RootKgRatioMethod/yb$stumproot
yb$proportionR<-yb$RootKgRatioMethod/yb$rootonly

Root only vs. Stump+root equations for yellow birch

Interestingly, the two different (and independent) component equations give nearly identical results for coarse roots alone, or roots + stump.

General ratio method vs. stump+root component equation for yellow birch

General ratio method vs. root-only component equation for yellow birch

Similar to the red maple comparison, the ratio method gives a smaller estimate than the root+stump or root-only species-specific component equation for trees with dbh<7.6 cm (6.9 for the roots-only component equation), and increasingly larger for larger trees. However, for most of the trees, the methods are within +/- 15%.

And just like the maple, the ratio method summed coarse root biomass over all the yellow birch stems is 88% of the summed root+stump biomass estimated by the yellow birch component equation, or 87% of the root-only equation.