As the largest continent on Earth, Asia has the most diverse climates, and the Asian monsoon dominates climates over a large part of Asia. Potential effects of global climate change on the above- and below-ground allocation of plant photosynthates and its potential feedback on the climate system remain largely unexplored, particularly under the footprint of the Asian Monsoon.
Within Asia and globally, the subtropical evergreen forest type has largely been converted to human use. The Ailaoshan Forest Reserve in fact represents the largest intact example of this forest type in Asia. Therefore it is a unique resource for studying interactions among climate, plants, and soil ecology.
Previous studies in managed forests have manipulated supply of plant-derived carbon to soils by stem girdling or root trenching, and thereby have provided extensive information about below-ground interactions between plants and the soil community. At Ailaoshan, we combined these techniques with manipulations of above-ground plant litter input to study interactions between the plant community and the forest carbon cycle.
In February 2004, eight 20 by 20 meter square plots were permanently surveyed and established on a forested hillslope near the Ailaoshan Field Station, with their perimeters trenched to a depth of 40 cm to exclude effects of external tree roots. Half of the eight major plots were randomly selected for tree girdling, with more than 400 tree stems girdled with a 5 cm band at approximately 1.5 meters high. This stops the supply of photosynthetic carbon from the stem to the roots, without immediately affecting root uptake of water and nutrients.
Within each of the eight major plots, 4 rectangular subplots of 2 by 3 meters were established. They were randomly assigned to control (untreated), root trenching, leaf litter removal, or a combination of the last two treatments. This yielded 8 treatments with 4 replicates each. The 8 treatments were un-manipulated control and the 3 manipulations (girdling, trenching, and litter removal) applied singly and in combination.
Results of the first three years of this project have already been published (three papers in Soil Biology & Biochemistry; one in Biogeochemistry), and several additional manuscripts are currently being prepared. Some highlights of that research are presented below:
Removing aboveground litterfall and the humus layer reduced soil respiration by about 38% more than the litter-C input. In other words, one gram of carbon as litter falling on the forest floor stimulates more than one gram of CO2 –carbon respired, suggesting a respiration priming effect, but the effect was absent from the trenching and girdling treatments. In contrast to many other studies, the trenching and girdling treatments did not significantly reduce annual soil respiration, thus the role of a functioning root system in soil respiration priming remains unclear.
The litter layer had highest temperature sensitivity of respiration (Q10 ≈ 3.5), the intact soil + humus intermediate (Q10 ≈ 3.3), and bare mineral soil the lowest (Q10 ≈ 2.8). Disrupting root processes did not significantly alter these values.
Soil concentrations of sugars were also regulated by aboveground litter input, and sugar concentrations lagged behind litterfall by about 2 months. Temperature and above ground litter input were both linked with increased concentrations of soil amino acids, but the treatments disrupting root processes reduced these effects.
Microbial biomass carbon (MBC) in the humus layer remained constant throughout seasonal changes in soil temperature and water content. Mineral soil MBC steadily decreased through the experimental period, even in untreated plots, and was positively correlated with soil temperature and water content. Litter removal reduced mineral soil MBC by 19 % in the non-girdled plots, but by only 4 % in girdled plots. Neither root trenching nor girdling significantly influenced mineral soil microbial C. These results suggest that belowground carbon from shoots and already present in soil organic matter, rather than above ground plant litter inputs, determine seasonal fluctuations of mineral soil MBC.
Labile organic carbon (LOC) in the mineral soil increased substantially in the litter-removal and stem-girdling treatments in 2006, but not in 2007. Humus had higher LOC and MBC than did mineral soil. All manipulations slowed the potential turnover rate of LOC in both humus and mineral soil. Mineral soil LOC increased in the litter-removal treatments, even though there was no overlying litter to act as a source. Slower potential turnover rates of LOC in humus than mineral soil suggest that plant-derived LOC may be less decomposable than LOC derived from microbial cells in the mineral soil.
Our experiment demonstrated that shifts in plant carbon allocation can substantially alter the dynamics of soil organic carbon. Our results suggest that global warming alone can not predict future soil organic carbon and changes in plant litter and root inputs can profoundly influence the temperature sensitivity of soil respiration.