This study aimed to develop tools for biomass estimation and quantify carbon stored
in mangrove forests of Tanzania mainland. The study was carried out in four sites
along the Tanzanian coastline; Pangani, Bagamoyo, Rufiji and Lindi-Mtwara. A total
of 120 plots were measured along transects running perpendicular to sea/rivers. From
each plot, one tree was destructively sampled for aboveground biomass. Thirty
among 120 trees were sampled for belowground biomass. Data analysis was carried
out in R software. Procedures for quantification of belowground biomass for
Avicennia marina (Forssk.) Vierh, Sonneratia alba J. Smith and Rhizophora
mucronata Lam. were documented in detail. Root sampling is recommended for A.
marina and S. alba while for R. mucronata, total root excavation method may be
applied. The methods are more comprehensive than previously reported methods,
therefore they should be applied in quantification of BGB. The study found an overall
mean tree aboveground basic density of 0.60±0.00 (SE) g cm -3 , 0.54 ± 0.01 (SE) g
cm -3 and 0.69 ± 0.01 (SE) g cm -3 for A. marina, S. alba and R. mucronata,
respectively. Similarly, the overall mean tree belowground basic density was 0.57 ±
0.02 (SE) g cm -3 , 0.32 ± 0.01 (SE) g cm -3 and 0.53 ± 0.02 (SE) g cm -3 for A. marina,
S. alba and R. mucronata, respectively. The study also showed that basic density
varied between species, tree sizes and tree components. Accordingly, if properly
determined and applied, basic density may be useful as a conversion factor and yield
accurate biomass estimates. Otherwise they are likely to be a source of uncertainties
in biomass estimation. Common (multi-species) and species-specific above- and
belowground biomass models for the three mangrove species were developed.ii
Species-specific models had better fit than common models. Evaluation of existing
biomass models on data from this study generally showed large and significant
prediction errors. Possibly this may be due to application of the models beyond data
size ranges, geographical locations, and differences in forest structure and tree
architecture. Species-specific models from this study are therefore recommended.
The use models to unrepresented species is not recommended, where necessary
however a conservativeness principle (i.e. when accuracy of estimates cannot be
achieved, the risk of over- or under-estimation should be minimised) need to be
applied. Using biomass models from this study and forest inventory data collected by
National Forest Resources Monitoring and Assessment (NAFORMA) of Tanzania,
the study quantified aboveground carbon (AGC), belowground carbon (BGC) and
total carbon (TC) stored in mangrove forests of Tanzania mainland. Results showed
that, AGC, BGC and TC were 33.5 ± 5.8 Mg C ha -1 (53% of TC), 30.0 ± 4.5 Mg C
ha -1 (47% of TC) and 63.5 ± 8.4 Mg C ha -1 respectively. Given that, mangroves of
Tanzania mainland cover approximately 158, 100 ha, a total of 10.0 millions Mg C
(i.e. 37.2 millions Mg CO 2 e) is stored in mangrove forests of Tanzania. Results from
this study are essential for REDD+ initiatives and provides useful input in
management of mangrove forests in the country.
Climate Change Impacts and Mitigation Programme (CCIAM)and the Kingdom of Norway