Estimation of above ground biomass of forest trees using dbh as a single parameter

Abstract

Forests in Sri Lanka contribute to the mitigation of climate change through sequestrating a net amountof carhon dioxide and also maintaining carbon stock as biomass. Total tree biomass comprised ofabove ground and below ground biomass of trees. This paper presents a methodology developed toestimate individual tre-e above ground biomass using allometric relationships. The advantage of thismethod is that the above ground biomass per unit area can be determined by estimating these valuesfor different species separately within that area. The present study was conducted in Yagirala Natural Forest Reserve situated in Kalutara District, low country wet zone of Sri Lanka. In this study, theabove ground biomass of different species available in that forest were predicted using the diameterat breast height (dbh).

45 tree species were identified in the forest and models were separately developed for each species.For each species, 3 classes were established using the size, i.e., small, medium, and large. For class 1and 2 trees, the main stem and the average size branch were divided into sections and for eachsection end diameters and length were measured. The volume of each section was estimated usingSmal ian's formula. The total branch volume was estimated by multiplying the average branch volumeby number of branches. Then a volume and biomass relationship was estimated for each species bymeasuring the biomass of a known volume.

The leaf biomass in the crown was estimated by measuring that in I m''of crown volume. Then it wasconverted to the entire crown via crown volume. Finally above ground tree biomass was calculatedby adding branch and crown biomass to stem biomass.

Then allometric relationships were developed to predict the above ground biomass of each of 45species separately using dbh. A strong relationship was found between total tree biomass and dbh ofall species with R2 values over 95%. Tn order to test further improvement of the models, total heightwas added as the second explanatory variable. However, it did not improve the R~ and for mostspecies height became non-significant. Therefore the finally selected basic model structure was logbiomass = a + b log dbh. The model parameters however, indicated different values for a and b forthe different species, varying a from -1.71 (Mastixia tetranda) to - 0.73 (Xylopia parvifolia) and bfrom 2.01 (Putranjiva zeylanica) to 3.29 (Syzygium cordifolium).