Water-use efficiency, nitrogen availability and tree growth of boreal forest ecosystem in northern China in response to climate change and major wildfires

Abstract

Climate change has become a catastrophic event that has altered Earth’s ecosystems. Global warming and increasing atmospheric greenhouse gases (GHGs), such as CO2 and methane, have caused a drier climate, thereby affecting the water availability and its impacts on carbon (C) and nitrogen (N) cycles within terrestrial ecosystems. Wildfires also play an important role in C and N cycling in terrestrial ecosystems. Extreme wildfires are known to increase both CO2 and N emissions into the atmosphere and can alter C and N cycling processes. The objectives of this study were to quantify the effects of climate change on long-term tree water-use efficiency, N availability and growth in boreal forest ecosystems of northern China, and to evaluate the signatures and intensities of wildfires and climate extremes (such as drought and flooding) by using 15N natural abundance in tree rings in the context of climate change. Tree rings were extracted from 4 Larix gmelinii sample trees and 4 Pinus sylvestris sample trees, located in a boreal plantation forest of Mohe City, Heilongjiang Province, China. Each set of sample trees had a range of ages, and for L. gmelinii, tree ring samples for tree 1 (54 years old in 2018); tree 2 (54 years old in 2018); tree 3 (57 years old in 2018) and tree 4 (48 years old in 2018). For P. sylvestris, tree ring samples for tree 1 (81 years old in 2018); tree 2 (57 years old in 2018); tree 3 (57 years old in 2018); and tree 4 (57 years old in 2018). Tree rings were measured with a mean annual basal area increment (BAI), while tree ring stable C isotope composition (δ13C) and N isotopic composition (δ15N) as well as total C and N concentrations were measured on mass spectrometer at three-year intervals. Tree intrinsic water-use efficiency (iWUE) was calculated using tree ring δ13C and atmospheric δ13C data. Both tree iWUE and growth data were related to the rising atmospheric CO2 concentration and other climatic data. The tree ring δ15N values were related to post- and pre-wildfire events, total N and atmospheric CO2 concentration. Multiple regression analysis was used to quantify the relationships among tree ring δ15N, BAI, WUE, atmospheric CO2-e (GHG) concentration, temperature, precipitation and humidity of the study site. Tree ring total C and N, δ13C and δ15N were determined on mass spectrometer at Griffith University, Nathan campus. Tree ring samples were cut every three years to obtain the isotopic results, and to effectively analyse the wildfire event that occurred at the study site in 1987 and climate extremes (particularly drought and flooding events) in the last 50-80 years. Overall, these results showed a quadratic decrease in relative humidity over the past 60 years with a continuous increase in average surface temperature, indicating an initial increase in water availability, peaking when it has reached the critical threshold water availability, and a decrease thereafter. As atmospheric CO2 (aCO2) continued to increase, iWUE also increased. BAI showed a quadratical relationship with aCO2, increasing initially but peaking at a critical threshold and decreasing thereafter. Tree ring δ15N, an index of N availability, decreased either linearly and non-linearly with the rising aCO2. Tree ring δ15N of L. geminlii four sample trees responded non-linearly, initially increasing with aCO2, but peaking and then decreasing therafter with the rising aCO2. While for P. sylvestris, two of the four sample trees decreased linearly and the other two trees decreased non-linearly, perhaps due to the aging of the sample trees, with Pinus species to have a higher significant value than those of Larix species trees. Thus, there was decreasing N availability in the last 20-30 years as the aCO2 continued to rise. The result showed that during the major wildfire event of 1987, there was a peak of tree ring δ15N. The wildfire resulted in the increasing δ15N intensity which lasted for about 20 – 30 years, depending on the sample trees. Tree ring δ15N in the P. sylvestris trees was highly sensitive to the major forest wildfires. Once it reached its critical threshold; tree growth would decrease as tree ring δ15N increased. However, tree ring δ15N in the L. gmelinii trees was not as sensitive to the major wildfire of 1987. This study represents the first attempt in the world to highlight that tree ring δ15N could be used to fingerprint both frequency and intensity of major wildfires (such as that occurred in 1987) and climate extremes (such as drought with lower δ15N and flash flooding events with very high δ15N). In conclusion, tree ring δ15N can be used as a sensitive indicator of N availability in the Boreal forest ecosystems, and could be used to fingerprint both frequency and intensity of major wildfires and climate extremes recorded in both L. gemilinii and P. Sylvestris sample trees. Unfortunately, tree growth has decreased with the rising atmospheric CO2 once the CO2 tipping points have been passed in the Boreal forest ecosystems, leading to the positive feedback to climate change, due to increasing water and N limitations in the boreal forest ecosystems under intensifying climate change and wildfires in the last 20-30 years.