Fire occurrence in relation to bamboo dominance in the Qinling Mountains of China: Evidence from phytolith and charcoal records in Holocene sediments

Abstract

The internal genetic clock of semelparous mast-flowering bamboo species creates gregarious regions of flowering and death, thus leading to fuel-load accumulation that potentially promotes fire (Fire-Cycle Hypothesis). Higher abundance of bamboo could be expected to provide greater fuel loads after flowering. Here, we investigated the possible link between bamboo dominance and fire occurrence. As the main food source for the giant panda in the Qinling Mountains of China, Bashania fargesii (Farges Canebrake) and Fargesia qinlingensis (Arrow Bamboo) dominate lower (1000–1900 m) and higher elevations (1800–2700 m), respectively. Four soil profiles (elevation ranging from 1240 to 2170 m) in areas of known recent bamboo flowering events (∼1980) and adjacent non-flowering sites were assessed for phytoliths and charcoal. Characteristic phytoliths were chosen to describe bamboo dynamics, and charcoal particles were extracted from the soil to reconstruct fire history. Carbon 14 dating of soil charcoal samples was conducted using an accelerator mass spectrometry technique. In addition, we chose two depths (28–52 and 90–100 cm) in one Bashania profile (B. fargesii forest) and one Fargesia profile (F. qinlingensis forest) for phytolith dating. Our findings indicate that bamboo has been a dominant component of these ecosystems almost for the entire profile; some 6400 yrs for Bashania and ca. 10000 yrs for the Fargesia. It is also clear from the charcoal and phytolith data that bamboo abundance and recent mast flowering were significantly related to fire occurrence for lower elevation Bashania while the higher elevation Fargesia was not. One of the four profiles had chronological charcoal records throughout and the oldest charcoal aged was ∼1410 yr BP, suggesting an anthropogenic origin. The mixing of phytolith and charcoal made it unrealistic to describe temporal dynamics of bamboo and fire events, and thus our interpretation is cognizant of the dating discrepancies and conservative (less-speculative). Due to the long history of bamboo and much shorter history of fire within the profile, our data lend little evidence to the Fire-Cycle Hypothesis.