The influences of regional atmospheric precipitation characteristics and its element inputs on the existence and development of Dinghushan forest ecosystems
The influences of regional atmospheric precipitation characteristics and its element inputs on the existence and development of Dinghushan forest ecosystems
复制标题
作者:
Zhou Guo
The region near tropic of cancer is dominated by semi-desert or desert.Dinhushan,being at 23°10′N,112°34′,is located in the region,but is covered by regional monsoon evergreen broad-leaved vegetation.This is undoubtedly resulted from the large rainfall due to monsoon climate,but other facts may complicate this explanation.Some regions near Dinghushan,desertification is unavoidable if the degraded ecosystem is not restored timely.For those kinds of degraded ecosystems,high rainfall will accelerate degradation.This study will examine the characteristics of rainfall,mineral input and their influences on the existence and development of forest ecosystems,and identify the major causes for formation of its typical vegetation. 34 years'rainfall data for the periods 1959 to 1965 and 1973 to 1999 were checked for accuracy using double cumulative rainfall curves with data from Gaoyao and Zhaoqing meteorological observatories as reference.The regression analysis was used to estimate the missing rainfall data of Dinghusha for the periods of 1956 to 1958 and 1966 to 1972. Two rainfall samples over the canopy of forest were taken monthly between 1998 and 1999.Concentrations of ammonia and nitrate were measured using Miniatnvized Kjeldahl and Phenoldisulphonic Acid methods respectively,content of phosphorus was measured using Molybdenum Blue method,and these of potassium,sodium and magnesium were measured using atomic absorption(model Zee man 180-80). More than 80% of annual rainfall falls in wet season(April to September),over which the air temperature is high,and that period coincides with active plant growth.The combination of water and heat inputs influences the development and restoration of ecosystems in this way:(1)Heavy rain in wet season often results in soil erosion and nutrient loss.Sufficient heat input in dry season will loose the washed surface soil,which makes it easy to be flushed in next wet season;and (2)Because of moderate solar radiation input in dry season,the plant growth potential still keeps high and demands for moderate water,but less rainfall in the season can not meet the need,which are detrimental to the recovery of vegetation. The mean rainfall intensity was 8.3mm·h -1 during 1993 to 1997.The medium and heavy rainfall accounts for over 30% of annual rainfall.Low intensity rainfall is much more frequent,but only accounts for 15.3% of the total.As compared with that in Heshan located in the same subtropical zone,the Dinghushan rainfall from the heavy rain and its frequency are much less.Yearly rainfall variability in Dinghushan over many years is 9.3%,much lower than 20% for the regions in the south of Yangtze River.Those results support the assertion that the rainfall and its distribution here have made the vegetation in Dinghushan to recover more easily than that in other regions of the similar latitudinal zones. Over the period of 1956 to 1999,there were 228 storm events with a mean of 5.3 times per year in Dinghushan.About 86.4% of storm rainfall occurred in the wet season,and only 13.6% in the dry season (October to March).There were occasional storms in November,December and January since 80's,which may relate to global climate change,and is worth further study. Over the period of 1956 to 1999,there were 132 typhoons.Typhoon usually occurred between May and November.Using the decadal statistics,we found that the highest number of hurricanes occurred in September in 1950's,August in the 1960's and 1970's,July in 1980's and August and September in 1990's. Following formulae is for autocorrelation analysis:C D=1N-D∑N-Di=1(x\-i-)(x i+d-)In the above equation,C D is he covariance of the rainfall-time series,x\-i is the rainfall at time,i,N is the length of the time series;D has a value of 0,1,2,…,m(mN);x is the mean of x i.The autocorrelation coefficient is equal to the ratio of C D and C 0. The first peak in the auto-covariance function was 7a and the second one was 11a for the rainfall time series in Dinghushan.Global climate oscillation also has a peak period of 7a,and 11a to 13a