Technical Comments

Calculating Forest Biomass Changes in China

See allHide authors and affiliations

Science  24 May 2002:
Vol. 296, Issue 5572, pp. 1359
DOI: 10.1126/science.296.5572.1359a

Fang et al. (1), using forest inventory data to estimate changes in forest biomass carbon storage in China during the past 50 years, demonstrated that China's carbon storage in living forest biomass has increased since the 1980s. We appreciate their valuable effort; however, the equation used to estimate forest biomass [note 20 in (1)] is questionable.

Fang et al. purportedly divided forests into 31 provinces, three site classes in each province, and five age classes in each site class. However, site classifications have not been available in any Forest Resource Statistics of China since the 1950s. It seems impossible for the authors to have performed their calculations without the required site class information. We suspect that the equation actually used did not rely on site classes or even age classes, contrary to the presentation by Fang et al. (1). Because the factor used in the paper, biomass expansion factor (BEF)—defined as the ratio of all stand biomass to living stock volume—did not vary with ages, site class, or province, the equation cited can in fact be simplified toEmbedded Imagewhere Y is the biomass of each forest type,A is the forest area in hectares, and x is the mean stem volume (in m3/ha).

This equation may be what Fang et al. actually used; Fang and Chen (2), in a later paper, used the expression above to get same results as in (1). Unfortunately, Fang et al. (1) failed to describe this detail, and instead presented a more complex expression than was actually used.


Response: Our study (1) developed the BEF method for estimating forest biomass carbon in China using forest inventory data and direct field measurements (1, 2). BEF varies primarily with stand age and site class (2–4); hence, forest biomass for a specific forest type can be calculated [note 20 in (1)] using different BEF values that correspond to stand age and site classes. Those values are summed to produce the national total biomass, Y:Embedded Image(1-1)where xijk ,Aijk , and BEFijk are mean timber volume, area of forest type, and mean BEF for thekth age class, the jth site class, and theith province, respectively.

Because BEF is a function of timber volume that integrates the effects of stand age, site quality, and other biotic and abiotic factors on forest biomass (2, 3, 5), however, and because the function is expressed as BEF = a + b/x for a specific forest type (1), where a andb represent constants for a specific forest type, Eq. 1 can easily be simplified:Embedded Image Embedded Image(1-2) Embedded Image(1-3)where A, x, and BEF are national total forest area, mean timber volume, and average BEF, respectively, and Ai ,xi , and BEFi are forest area, mean stand volume, and mean BEF for the ith province, respectively.

Eqs. 2 and 3 are thus mathematically equivalent to Eq. 1 (6), and it becomes possible to accurately calculate forest biomass for each forest type using forest area, stand volume, and corresponding BEF values without actually using age class, site class, and other information. This is the essence of estimating provincial and national forest biomass for China (1–2), and the lack of a detailed explanation in our paper may have led to the misunderstanding by Zhang and Xu. Although we appreciate their comments, it should be noted that their assertion that BEF does not vary with age, site class, and province is clearly incorrect. It should also be understood from the foregoing that the BEF-volume function remains the same for a specific forest type but that BEF obviously varies with the timber volume.

Finally, we point out that Forest Resource Inventory of China (FRIC) is a huge national forest database with far more information than has been published in the Forest Resource Statistics of China cited by Zhang and Xu. FRIC contains more than 30 variables (7), including the site class data, although it is not available for some of plots. For those plots, the site class can be easily obtained from stand height and age, which are well documented in the FRIC database. That site class can easily be derived from forest height and age is generally understood in forestry (8).


Navigate This Article