Transpiration Characteristics Of A Black Locust Plantation In Response To Rainfall Reduction In The Sub-humid Region Of Loess Plateau,China

Loess Plateau is one of the major regions in China where vegetation protection and reconstruction have been conducted.Black locust?Robinia pseudoacacia?is the most commonly planted species which accounts for over 300,000 ha of artificial forests in the sub-humid and semiarid areas in Shaanxi Province.However,under the condition of global climate change which is introducing much uncertainty of precipitation patterns,it is of special significance to investigate their responses to precipitation.In this research,by placing waterproof panels within tree rows,partial precipitation was excluded from treated plots since April 2015.Transpiration water use in both treated and control plots were monitored by measuring stem xylem sap flow using Granier-type thermal dissipation probes from 2011 to 2017.Meteorological factors,soil moisture conditions and rainfall were monitored simultaneously.Based on large-sample measurements of stem sap flow,dynamics of transpiration water use and their responses to precipitation and soil water conditions were investigated.Combined with the measurement of physiological and morphological indices related to transpiration,the physiological and ecological mechanism of black locust adapting to short and long-term drought was analyzed as well.The major results and conclusions are as follows:?1?During the experiment,annual precipitation in the study area ranged from516–751mm,and the gross rainfall in the forest was allocated as 78.1%throughfall,17.2%canopy interception,and 4.2%stem flow.The throughfall exclusion experiment reduced the throughfall input by 60.8%,which is about 47.5%of the gross precipitation.The effect of throughfall exclusion treatment on soil water content is significant.Soil water content in treated plots was significantly lower than that in the control plot with the monthly average soil water content in the treatment plot being 76.5%–100%of the control plot,and the annual average soil water content being 81.4%–98.5%of the control plot.The differences in monthly and annual averages of soil water content between the two plots increased with the treatment time.The reduction of precipitation reduced the stand growth rate,with the average annual DBH increment in the treatment plot being significantly lower than that in the control plot.?2?On daily scale,the diurnal change of black locust sap flow was similar to that of solar radiation and water pressure deficit,with a single-peak on bright days.In dry weather condition,the peak time of sap flow was slightly earlier than that in wet weather condition.Daily transpiration of black locust was mainly affected by meteorological factors.The datasets of half-hourly plot-averaged sap flux density and the integrative variable of transpiration?VT?from solar radiation and water pressure deficit were well fitted with the exponential saturation curve,and the fitting parameters could be affected by the changes of soil water content.The daily average sap flux density was significantly positively correlated with solar radiation,air temperature,and water pressure deficit,but negatively correlated with air relative humidity.On monthly and seasonal scales,the monthly average sap flux density showed an increasing trend in early months and then decreased,with those of the control plot being 5.40m L m^-2 s^-1,18.96 m L m^-2 s^-1,23.62 m L m^-2 s^-1,21.63 m L m^-2 s^-1,10.17 m L m^-2 s^-1,and6.06m L m^-2 s^-1 from May to October,respectively.The month when the highest value of the sap flux density appears differed among years,probably because of the difference in rainfall pattern across years.The monthly average sap flux density was only positively correlated with the monthly mean air temperature.On annual scales,the annual average sap flux density of black locust ranged from14.96?21.01 m L m^-2 s^-1,which was significantly positively correlated with the annual average potential evapotranspiration?p<0.05?.?3?Rainfall exclusion treatment decreased the peak values of sap flux density in diurnal courses.The daily average sap flux densities in the treatment plot were lower than those in the control plot by 3.22 m L m^-2 s^-1,9.76 m L m^-2 s^-1,and 13.00 m L m^-2 s^-1 in 2015,2016,and 2017,respectively.The peak time in diurnal courses in the treated plot was 1-1.5 h earlier than in the control plot.Meanwhile,the reduction of rainfall had a negative effect on the response of half-hour average sap flux density to VT,with the parameters?a,b?of treated plots being significantly lower than those of control plots?p<0.001?.The monthly average sap flow density of the black locust in the treatment plot was0.24-19.52 m L m^-2 s^-1 lower than control plot.During the treatment,the monthly average sap flux density in treated plot was significantly positively correlated with soil water content,and the regression slope of monthly averages sap flux density to potential evapotranspiration in treated plot was significantly lower than that in control plot.During the 3 years of treatment,the annual average sap flux densities in treated plot were lower than in the control plot by 9.1%-45.3%.Along with the treatment period,the differences of average sap flux density between the two plots changed from not significant in the first year of treatment?p>0.05?to significant in the second year?p<0.05?,and to extremely significant in the third year?p<0.001?.The annual average sap flux density in treated plot was negatively related with annual average potential evapotranspiration,and positively related with annual average soil water content.?4?Rainfall exclusion resulted in changes of several morphological and physiological indices,which might represent the species strategy for water use regulation.The stomatal densities of black locust trees in treated plot(173.54–215.09 No mm^-2)were 7.3%–14.8%lower than those(203.59–239.96 No.mm^-2)in control plot.The specific leaf area of black locust trees was not significantly different among different months,but those in treated plot(305.31–370.31 cm² g^-1)were significantly higher than those of control plot(255.40–293.24 cm² g^-1).The leaf water potential varied significantly among different months,with small values in July and August and high values in June and September.The predawn leaf water potential in the treated plot was 0.03–0.16 MPa lower than that of control plot,with the largest difference between the two plots appearing in the driest month of July.Meanwhile,no significant difference was observed for the midday leaf water potential between the two plots,which indicated isohydric behavior.There were significant differences in sapflow driving force and whole tree hydraulic conductance among different months.The lowest and highest values for sapflow driving force appeared in August and June,respectively.Meanwhile,the whole tree hydraulic conductance was not significantly different between the two plots in most months,except for the dry period of August when severe drought might cause the water use strategy being anisohydric behavior.The stable carbon isotope abundance in black locust leaves varied significantly among different months.However,significant difference in?¹³C was not detected between plots,probably due to the isohydric behavior of black locust under temporary drought.Although black locust had been considered as an invasive species in regions with ample rainfall or accessible groundwater,considering the strong demand for transpiration water use and the shortage of precipitation supply on the Loess Plateau,the use of this species in reforestation in semiarid areas should be implemented with caution or given necessary management.