Research at China Regional Climate Centre
Site description
The China RCC is located in Gutianshan National Nature Reserve in the extreme east of Zhejiang Province. The forest is subtropical evergreen broad-leaved forest; although 25% of China is covered by subtropical forest, over 95% of the forests are secondary (re-grown after major human disturbance) or plantations. Gutianshan is an area of high biodiversity; for example, over 1400 species of seed plants have been found there. The site presents particular challenges as the average gradient of the terrain is 30o.
Climate Champions censusing trees
Research partner institution and Principal Investigator
Here Earthwatch is working with the Institute of Botany at the Chinese Academy of Sciences (IBCAS). IBCAS is a renowned research centre for plant sciences in China, having been actively engaged in research on terrestrial ecosystems for more than 20 years; much of their work is used as a direct resource base for the Chinese Government’s environmental protection and management policies.
Volunteers from HSBC, termed 'Climate Champions', assist with data collection in the field and data entry, overseen by scientists from IBCAS.
The principal investigator of the project is Ma Keping, the Director General of IBCAS. His primary interests have been focused on various aspects of climate change and biodiversity, from the community to global level including carbon dioxide recycling in forest ecosystems, responses of forest ecosystems to global climate change, forest fragmentation and human disturbance, global patterns in biodiversity and community assembly. He has published over 170 papers, is a member of the IUCN Council and sits on numerous scientific committees.
Sampling design
Data collection methods are as described in the Research Introduction. Work started here in July 2008 and will continue to December 2011. There are 12 permanent sample plots of one hectare (100m x 100m), in varying stages of recovery from human disturbance (Table 1). As in the other RCCs, initial work has been to tag all individual trees in each of the 12 plots.
The project began in July 2009 and will run until December 2011. Data collection methods are as described in the Research Introduction. There are 12 permanent plots of 1 ha each, located in old growth, secondary re-growth and plantation (Cunninghamia lanceolata, a conifer) forests.
Approximately 80 trees with a range of diameters were selected from each plot to be fitted with dendrometers. Data is regularly collected from dendrometers. There are 12 leaf litter traps per plot, from which litter is collected, dried and sorted on a monthly basis.
There is additional research into survival of woody seedlings, carbon within the soil and the rate of leaf litter decomposition. For the seedlings, the same information is collected as for trees (species, height, diameter, location) from 1x1 m subplots within each of the 1 ha plots; surveys are conducted twice a year.
Soil carbon is measured by collecting soil from a range of depths at 13 systematically located points within each plot. In the lab the soil is dried and roots and other debris removed before being analysed for carbon content.
Leaf litter decomposition rates are assessed by burying mesh bags of leaf litter at several locations in each plot. Leaves from four species ARE used, as well as a mixture of all four species (Figure 1). Bags are of fine or large mesh, thus excluding or allowing entry of macro invertebrates, to assess their role in leaf litter decomposition. Replicate bags are buried and retrieved at three month intervals over a year. The weight loss of litter is then measured in the laboratory, and amounts of carbon and nitrogen present are analysed. Rate of leaf litter decomposition is thus studied, and the relative contribution of soil macro-invertebrates assessed. The decomposition of dead plant material is an important component of the carbon cycle.
Figure 1 Leaf litter decomposition experiment
Data quality
As large quantities of data are collected by volunteers, each of whom normally stay no longer than two weeks, it is essential to monitor the quality of data collected, and understand where errors might exist. Four 1-ha plots were randomly selected, and in which 20 randomly selected quadrats of 20 x 20 m were chosen. All trees with DBH >= 10 mm were re-measured, including species identification, DBH and coordinate measurement. The criteria of correct measurement are that the error for DBH is within 2 mm, and that for coordinate is within 50 cm.
Percentage of measurements that were incorrect were 0.98%, 1% and 2.1% for species identification, DBH and coordinate respectively, after removing the maxima of 11.9%, 7.6% and 7.6%.
Data were tracked back and it was found that the maximum errors were in data collected by an elderly local helper with poor eyesight, who had problems reading instruments in the dark forest on very cloudy days.