The Responses Of Soil Respiration In Subalpine Meadow To Grazing Methods In Tianlaochi Catchment,Qilian Mountains

Alpine grassland is sensitive to climate change because of its fragile ecological environment.Under the background of global warming,studying the soil respiration of alpine meadow can not only predict the effect of global warmingon the ecosystem,but also provide managers with effective advices.The grass ecosystem in Qilian mountains is a dominant one and alpine grassland area is 3.56 x 105hm2,accounting for 36.1% of the total grassland area.The study on soil respiration of alpine meadow in Qilian mountains can indicate the level of soil C cycle in the grassland of the northwest China to some extent.In this paper,we select the Tianlaochi catchment in the Qilian Mountains as our study area.Soil respiration(Rs)and ecosystem respiration(Re)of alpine meadow were observed using an LI-8100 automatic instrument(LI-Cor,Inc,NE,USA)from 2014 to 2016.Simultaneously,we measured the environmental factors(i.e.,soil temperature Ts,soil surface temperature Ta,soilvolumetric water content VWC and soil surface relative humidity RH)? spectral indices(i.e.,NDVI,EVI,CIrededge)and vegetation physiological and ecological parameters(i.e.,aboveground biomass,grass height,specific leaf area,single leaf area and leaf water content).We analyze the temporal variation of respiration and relationships with environmental factors,spectral indices and vegetation physiological and ecological parameters.In addition,we study the influence of yak excretal,grazing intensity and grazing duration on respiration through stimulated methods including adding yak dung in different quantity(0,1000,1500,3000 g·m-2)to soil surface,clipping treatments(mowed grass amount account for 0%,45—50%,60—75% of the total grass biomass)and clipping duration(1 year,2 years and 3 years).The main conclusions are as follows:(1)Re and Rs have temporal variations with unimodal distribution pattern inmonthly and diurnal scales in growing seasons.The daily minimum Re occurs at0800 h and maximum value at 1300h(Re=18.31±6.26 ?mol·m-2·s-1),the daily minimum Rs happens at 0800 h and maximum value at 1400h(Rs=5.82±1.18?mol·m-2·s-1);The monthly maximum Re is at the end of July(In 2014,Re=21.39±1.43?mol·m-2·s-1)or at the beginning of August(In 2015,Re=20.95±2.66 ?mol·m-2·s-1),The monthly maximum Rs occurs at the beginning of August(In 2014,Rs=8.09±0.28?mol·m-2·s-1,In 2015,Rs=6.13±0.58 ?mol·m-2·s-1).The average daily Re and Rs are9.89±0.69 ?mol·m-2·s-1and 3.78±0.21 ?mol·m-2·s-1,respectively;The average monthly Re and Rs are 11.06±0.06 ?mol·m-2·s-1 and 4.23±0.62 ?mol·m-2·s-1,respectively.The optimum measuring time for Re is from 0900 h to1100h,while the optimum measuring time for Rs is from 0900 h to 1200 h.(2)Respiration has significant exponential correlation with soil surface temperature and soil temperature(P<0.05),and soil temperaturehas higher explanatory power than soil surface temperature.The relationship between respiration and soil surface humidity can be described by a linear function;There is no relationship between respiration and volumetric water content at 5,10,and 20 cm soil depth.The compound model consisting of temperature and moisture shows that: Re is mainly controlled by soil temperature maily,while Rs is controlled by soil temperature and soil moisture together.R=aebTWc andR=aebT+cW are the best model for Re and Rs,respectively.(3)The relationships between respiration and spectral indices depend on the grass disturbance status.In banning grazing grassland,CIrededge has the highest correlation with Re and can explain 72% variation of Re.In grazing grassland,NDVI has the highest correlation with Re and can explain 66% variation of Re.There is no relationship between Rs and spectral indices.(4)Aboveground biomass,grass height,specific leaf area,single leaf area and leaf water content have significant linear correlation with Re,they can explain 44%,60%,64%,48% and 47% variation of Re.(5)Yak dung has positive effect on Rs,the more amount of dung is put,the higher Rs is,they have a significant correlation(P<0.05).The average daily Rs in the soil added yak dung as amount of 0,1000,1500,and 3000 g·m-2 are 4.12,4.32,4.38,and 5.35 ?mol·m-2·s-1,respectively.The average monthly Rs in soil added yak dung as amount of 0,1000,1500,and 3000 g·m-2 are 4.53,4.70,5.47,and 6.60 ?mol·m-2·s-1,respectively.(6)Grass clipping decreases the Re,but the effect is not significant.The average daily Re of non-clipping,moderate clipping and heavy clipping are 11.70,10.21,and8.14 umol·m-2·s-1,respectively.The average monthly Re of non-clipping,moderate clipping and heavy clipping are 11.87,11.36,and 9.91 umol·m-2·s-1,respectively.Clipping duration has negative effect on Rs,that is,the longer duration is,the lower Rs,but the effect is not significant.