模拟降水对古尔班通古特沙漠生物结皮表观土壤碳通量的影响

水分是控制干旱区生态过程的重要环境因素,在水分受限制的生态系统中,降水通过改变土壤的干湿状态直接控制地下生物过程。生物结皮作为干旱区主要的地表覆盖物,能利用空气中有限的水分进行光合作用,其自身的碳交换是干旱区土壤碳通量的重要组成部分。通过模拟0(对照)、2、5 mm和15 mm 4个降水梯度,利用红外气体分析仪,对古尔班通古特沙漠中部生物结皮以及裸地表观土壤碳通量进行测量,探讨不同强度降水条件下生物结皮对表观土壤碳通量的影响,结果表明:(1)降水增加了生物结皮表观土壤碳释放量,2、5 mm和15 mm 3种降水处理累积碳释放量分别是对照的151.48%、274.97%、306.44%,并且随着降水后时间的延长,表观土壤碳通量逐渐减小直至达到降水前的水平;(2)生物结皮与裸地的表观土壤碳通量对降水的响应不同,对照和最大降水量下,生物结皮表观土壤碳通量大于裸地,但是2 mm和5 mm降水后,生物结皮表观土壤碳通量小于裸地,并且二者在2 mm降水时差异显著(P<0.05),而在其它降水处理下无显著差异;(3)连续两次降水事件,活性碳在初级降水后的大量释放使得二次降水后释放量下降,其中裸地碳释放量下降速率与降水强度正相关。本研究说明,在探求荒漠地区土壤碳交换对降水的响应规律时,应该考虑生物结皮的影响以及连续降水事件的差异。

Effects of simulated precipitation on apparent carbon flux of biologically crusted soils in the Gurbantunggut Desert in Xinjiang, Northwestern China

Abstract Water is the major driver of ecosystem functions and processes in semiarid and arid regions, Water availability is directly linked to precipitation, and variation in the productivity of desert shrublands and grasslands is well known to correspond to seasonal and inter-annual rainfall patterns. In such water-limited ecosystems, pulsed precipitation directly control the below ground processes through a series of soil drying and rewetting cycles. Precipitation events are infrequent and discrete in semiarid and arid region. Global circulation models predict a shift in precipitation patterns to growing season rainfall events that are larger in size but fewer in number. This "repackaging" of rainfall into large events with long intervening dry intervals could be particularly important in semi-arid grasslands because it is in marked contrast to the frequent but small events that have historically defined this ecosystem. In arid ecosystems, soil microbes respond to very limited precipitation events. However, larger precipitation pulses are required by vascular plants for photosynthetic activities. Biological soil crusts are an integral part of the soil system in arid regions worldwide, stabilizing soil surfaces and aiding vascular plant establishment. They are also significant sources of ecosystem nitrogen and carbon. Biological soil crusts can take advantage of limited moisture in the air for photosynthesis, and carbon exchange involving biological soil crusts is an important component of soil carbon fluxes in arid areas. In the Gurbantunggut Desert, about 28.7% of the area is covered by biological soil crusts dominated by lichen. In this paper, to discuss the effect of biological soil crusts on apparent soil carbon fluxes under different amounts of precipitation, CO2 fluxs in soil covered with moss-lichen crusts and in crust-removed soil from the central Gurbantunggut Desert was measured by infrared gas analyzer after addition of simulated amounts of precipitation of 0 mm, 2 mm, 5 mm, and 15 mm. The results show that: (1)the release of CO2 from biologically crusted soil was trigged by precipitation, with the accumulated carbon eflux under 2 mm, 5 mm, and 15 mm of simulated precipitation being 151.48%, 274.97%, and 306.44%, respectively, of that of the control. As time elapsed after rainfall, the carbon flux decreased until it reached the value equivalent to that before precipitation was applied; (2)Biologically crusted soil and soil with crusts removed had different responses to precipitation with the carbon flux being significantly different between biologically crusted soil and bare soil after addition of 2 mm of simulated water. Biologically crusted soil emitted more carbon than bare soil under 0mm and 15 mm of simulated precipitation, although the differences were not significant; (3)A large release of activated carbon after an initial simulated precipitation event was followed by a decline in the soil carbon efflux after further simulated successive rainfall events, with the decline in the flux release rate from bare soil being positively correlated with rainfall intensity. The study shows that the effect of biological soil crusts and successive rainfall events should be taken into consideration when discussing the response of soil carbon exchange in arid lands to simulated rainfall.