| This study uses the data on transpiration of apple (Malus pumila cv. Starkrimson) trees measured with thermal dissipation sap flow velocity probe at a 2-year cycle and a 10-minute measuring interval in combination with the values of tree canopy micrometeorology elements and soil water content measured at fixed position and irregular time to analyze the transpiration characteristics and its influenced mechanism of 10-year old apple trees with a spacing of 3m×4 m in the Hilly area of TaiHang Mountain s in Taihangshan Mountain Range. The results further improve the theories of water consumption by transpiration of apple trees and provide theoretical basis of water ecology for apple production in the semi-arid region.In order to further explore the complexity of the rule of the apple trees transpiration (TR), the fractal dimensions of TR and micrometeorology on the canopy of the apple trees were estimated according to the fractal theory and the differences between them were quantified, hence further enriching the application of fractal theories in forestry science and meteorology.A model for prediction of soil moisture was developed with use of canopy and air temperature differential AT to diagnose or monitor the water stress of the apple trees. The model provides necessary theoretical basis for decision on appropriate irrigation in the areas lacking water resources. It provides useful additions to the core component of the water-saving forestry technologies.1. Temporal variation of transpirationThe diurnal variation pattern of transpiration was evident no matter whether it is during the major growing period (April-September) or the minor growing period (January-March, October-December). Transpiration appeared to be a single-peak curve in clear and cloudy days in both major and minor growing period. Thetranspiration peaked during 12:00-13:00, and the bottomed between 5:00-6:00. During the major growing period, in clear and cloudy days the maximal transpiration of individual trees was 3.50 - 4.97L/h and 1.99 - 4.18L/h in 2003 and 2004 respectively, and the average transpiration rate during daytime (7:00-19:00) was 2.60L/h and 2.34 L/h respectively; in overcast days, transpiration displayed a multi-peak curve during the major growing period, bi-peak curve during the minor growing period except October, and a single-peak curve during all other months. These were mainly attributable to the differences of climate conditions, soil moisture and physiological conditions in the apple trees that made the extent of transpiration, starting and ending time, maximal value and appearing time different.Daily or seasonal variation of the transpiration increased rapidly from April and peaked in May and June.,since then the transpiration decreased slightly. From October, leaves of apple trees started to fall and the transpiration decreases gradually, during November to December, all leaves have fallen off, only a small amount of water is needed to maintain physiological activities, therefore the water consumption by transpiration was very low. The whole-year transpiration was 600.9 mm and 468.0 mm respectively in 2003 and 2004. The transpiration in the major growing period (April-September) and the minor growing period (January-March, October-December) was respectively 502.6 mm and 98.3 mm in 2003, accounting for 83.6% and 16.4% respectively. In 2004 the transpiration was 408.3 mm and 59.7 mm in the major and the minor growing periods accounting for 87.2% and 12.8% respectively. The total transpiration differed 132.9 mm between 2003 and 2004, which was related to physiological characteristics such as the abundant and lean years of fruit production, in addition, the difference was also attributable to the climate differences between different years, particularly the differences of rainfall and hours of sunlight.Analysis based on phenology indicated that 2 peaks appeared during the period from budding to fruit-maturing (early April - mid October), the first peak appeared at the spring bud burst, just from early May to mid June, the second peak appeared at theperiod of rapid fruit enlargement, i.e. the second growing period of new shoots or the growing period of the Autumn shoots, just from early July to later August. Statistic analysis indicated that the transpiration in 2003 and 2004 during the period from budding to fruit-maturing were 513.6 mm and 413.0 mm respectively, with a difference of 100.6 mm between the 2 years. However, the difference expressed as percentage of the total yearly transpiration, respectively 85.5% and 88.3%, were very small. And the distribution patterns among different stages during the budding-maturing period were also similar between the 2 years. Results indicated that although the total yearly transpiration and major-growing period transpiration largely different between years, the proportions of transpiration during the major-growing period to the yearly total were similar.2. Relations between transpiration and canopy micrometeology and soil moistureDuring the major-growing period, daily transpiration per tree was significantly(a=0.01) related to canopy net radiation (Rn), air temperature (Ta), relative humidity(RH), wind velocity (V), and the canopy net radiation (Rn) was the most importantparameter affecting transpiration in each month. The regression equations for 2003and 2004 were respectivelyTR = 0.274 + 0.042V + 0.067Ta - 0.014RH + 0.006Rn (n=26352, r=0.8461),TR = 0.213 + 0.053V + 0.064Ta - 0.017RH + 0.007Rn (n=26352, r=0.8122)where the units for TR, V, Ta, Rn are respectively L/h, m/s, ℃ and w/m2, RH is inpercentage (%)In clear and cloudy days, during the major-growing period, the total daily daytime transpiration (STR) is highly related to the sum of daytime canopy net radiation, average temperature (AVETa), average relative humidity (AVERH), average wind velocity (AVEV) and soil water availability in 0-80 cm soil, the complex correlation coefficient is r=0.733, the fitted regression equation is:STR=0.998+5.367SW+0.163AVEV+0.0395AVETa-0.74AVERH+7.8×10-4SRn Where units for STR, SRn, AVETa, AVEV, SW are respectively mm, w/m2, ℃, m/s. RH is expressed in percentage (%). Statistic analysis indicated that the correlation was significant at 0.01 level and that the sum of daytime canopy net radiation was the most influential environmental factor affecting transpiration.3 Water demand and supply in apple treesIn terms of an entire year, there was no water supply and demand conflict, or water shortage and water stress in the experiment area. However, imbalance of water supply and demand was more prominent during seasonal and phonological or water sensitive period, the imbalance occurs during period from budding to rapid growing (early April-mid June), particularly during budding in earl April and rapid growing in early June, the conflict becomes more prominent.The ratio of rainfall to transpiration (R/TR) during budding period and rapid growth period in 2003 was 0.05 and 0.82 respectively.The ratios in 2004 were 0.14 and 0.63 respectively. The results indicated that in 2003 the water supply and demand in the apple trees was basically balanced, but the imbalance was prominent during the flowering period. Apple trees were sensitive during periods of budding, flowering and rapid shoot growth, therefore water management appears to be more important. From July to September or from ending of shoot growth to rapid fruit enlargement, although rainfall is still sufficient, attention still needs to be paid to water stress or water shortage in the apple orchard due to the poor water conservation capacity in the hilly area of Taihang Mountains where soils are shallow.4. Fractal characteristics of the apple trees transpirationThe transpiration of apple trees has chaotic characteristics, this factor or system is a deterministic system. However, when the analyses of time series or internal randomness and chaotic characteristics were based on data collected with a time interval (Dt) larger than 1 hour, a fictitious phenomenon could appear. Therefore, useof data with a time interval of data collection less than 1 hour for the analysis of transpiration characteristics is recommended.During the major-growing period of apple trees, when the lag time(τ) is 10 min., the fractal dimension (D) of transpiration of 2003 and 2004 were respectively 1.3851 and 1.4639, differing 5.4% from each other, and the embedding dimensions (m) were respectively 6 and 9, it will require at least 6-9 independent variables under such conditions to describe the transpiration dynamics equation or to conduct random simulations of the transpiration process sufficiently. When(?) =30 min., the values of D were 1.6908 and 1.8214 for 2003 and 2004 respectively, 7.2% different with each other, meanwhile the values of m were respectively 10 and 11, under such conditions, at least 10 independent variables will be required to describe the transpiration dynamics equation or to conduct random simulation of the transpiration process sufficiently. When the time intervals of data collection were identical, the differences of fractal dimension of transpiration among different years were not evident, indicating no significant differences in fractal characteristics. Therefore, although transpiration in different years was different in quantity or external appearance, there was neither difference in natural characteristics, nor in internal patterns and internal characteristics.No matter in 2003 or in 2004, all the fractal curves of transpiration (TR) with canopy net radiation (Rn), air temperature (Ta), relative humidity (RH) and wince velocity (V) had 2 non-scaling regions. The circumscription time values of the inflexions and D of TR were mostly close to that of Rn, further proving the result implied by the partial correlation coefficient fitted for the equation of TR with Rn, Ta, RH and V. Rn is the most influential micrometeorology element affecting transpiration5. Temporal variation of canopy leaf temperature(Tc) measured with infra red and its relation to micrometereologyDuring budding to rapid fruit enlargement (April-August), no matter it is clear day orovercast day, the diurnal variation of Tc was in a multi-peak tendency, with the highest value occurred around 12:00-13:00. By comparing with the canopy air temperature (Ta), in clear days the Tc curve displayed larger fluctuations, while the Ta curve was relatively smoother. Tc and Ta did not rise and decline simultaneously, the daytime peak time of Tc and Ta were different with 1-2 hours earlier for Tc. For all clear days, the average relative variation of Tc was 11.50% which was 20.68% higher than Ta (9.53%). In contrast, in overcast days, the diurnal variation of Tc and Ta was relatively consistent.During different development periods (budding, flowering, rapid shoot growth, end of shoot growth, rapid fruit enlargement), correlations of Tc to canopy air temperature (Ta), canopy net radiation (Rn), relative humidity (RH) and canopy wind velocity (V) were slightly different, the main micrometerology affecting daytime Tc were the canopy temperature (Ta) and canopy net radiation (Rn) for all the development periods. During the period from budding to rapid fruit enlargement (early April-later August) ,the regression equation of daytime Tc in clear days to Ta, Rn, RH and V for 2003 and 2004 were respectively Tc = -1.658 + 0.855Ta+0.005878Rn (r = 0.847) and Tc = -4.24+0.76Ta+0.00528Rn (r=0.98), the equations were fitted at 0.01 significance level.6. temporal variation of canopy leaf and air temperature difference(△T) and its relation to environment elementsDuring the periods of budding, flowering, rapid shoot growth, end of shoots growth and rapid fruit enlargement, for both clear and overcast days, the daytime AT displayed a multi-peak curve. In clear days, the maximal values appeared between 12:00-13:00. However, due to the different climate conditions in different period or months, and differences of soil moisture and physiological characteristics of apple trees, the AT value differs among different development periods, the absolute values of AT were significantly smaller in overcast days than in clear and cloudy days, indicating that in overcast days the difference between canopy leaf temperature andair temperature was not as evident as that in clear and cloudy days.In the clear day ,the main micrometeorology affecting on the daytime △T was Rn, the next was RH and then the V, △T was positively correlated with Rn and RH.But correlation with V was sometimes positive and sometimes negative. This result was in consistence with the physical indications of elements of the theoretic equation of AT (Jackson, 1981). The correlation coefficient of the equation was no larger than 0.50, mainly attributable to the impacts of soil moisture and difference impacts of Rn, RH and V at various development periods.Gradual regression analysis indicated that at 14:00 in clear days the AT was significantly correlated with Rn and soil water (SW) in the soil layer between 0-80cm. During the period from budding to rapid fruit enlargement in 2003, the regression equation between AT and Rn and SW at 14:00 was:△T = 7.71-0.43SW + 0.00307Rn-2.91RH+0.18V.The correlation coefficient of the equation was r=0.826 and the F value was F=18.814, significant at 0.01 level.7 Diagnosis of water stress in apple trees and model for prediction of soil waterDuring the period from budding to rapid fruit enlargement, the regression equation of relative available water (RAW) in the root soil to AT and Rn at 14:00 was:RAW=: 0.473-0.064△T+0.00035RnThe complex correlation coefficient was r=0.811, F=23.960> F2,25 0.01 =5.57, the correlation was significant at 0.01 level. Therefore the equation can be used as a model to predict soil water. The regression and correlation coefficients between RAW and △T, Rn differed among different periods. In the model, RAW contains field water capacity (FW), wilting water coefficient (RW). This study determined the values of FW and RW through real measurement and theoretical calculations. The model was tested with measured soil water, results indicated that the simulated RAW was significantly correlated at 0.01 level with the measured values.This study used the canopy net radiation, which is not only related to canopy micrometerology, but also affected by soil water and physiological characteristics of the apple trees. Therefore it can be used to insightfully reveal the relations of soil water to plants and the atmosphere, and to improve the accuracy of predictions. In order to indicate the possibility of using soil water as the indicator of water stress of apple trees, this paper uses the formula by Idson (1981) of calculating the crop water stress index (CWSI) with AT as the independent variable to calculate the correlation coefficient between the apple tree water stress index (TWSI) and the RAW, result indicated that during the period from budding to rapid fruit enlargement, TWSI and RAW were significantly correlated negatively, the extent of water .stress in apple trees decreased with the increase of soil water. Therefore, it is to certain extent feasible to use RAW as indicator of apple tree water stress index or extent. This implies that a key technical issue of establishing irrigation decision system of apple orchard is to understand available soil water on a regular basis.In order to use AT to predict soil water, identify water stress and develop irrigation system, when developing specific models of soil water prediction, the measured canopy net radiation (Rn) can be substituted into the regression equation, according to the threshold values of relative available water (RAW) in soil around the roots for irrigation, the threshold value of can be obtained, the measured and the calculated values of AT were compared, if measured AT less than threshold value, it indicates that no irrigation needed; if larger than threshold value, then irrigation is needed.Considering the inconvenience of measuring canopy net radiation in practice, from statistics and fractal aspects, the relations of total solar radiation (Q) and canopy net radiation (Rn) during major-growing period was analyzed, the possibility of using total Q to estimate canopy net radiation was discussed. Results indicated that apple tree canopy Rn was significantly linearly correlated with Q, the correlation coefficientr was 0.9833 (n=26062), and F=758350.9>>F2,26062 0.01=6.63. The regression equationwas Rn=0.7404Q-31.14. The fractal curves of both Q and Rn had 2 no-scaling regions, the circumscription time value of the inflexion were respectively 438.5 minutes and 420.9 minutes. The fractal dimensions were 1.11 and 1.13 in the first scaling region, and 1.91 and 1.90 in the second scaling region. The fractal dimensions of Q and Rn were very similar. This further proved that Q and Rn were significantly correlated based on the fractal theory. Therefore Q can be used to estimate Rn.This study has the following innovative aspects:1) Through 2 successive years of continuous observations on transpiration and canopy micrometeology of apple trees such such as Rn, Ta RH ,V, in combination of permanently positioned measurements of soil water, this study conducted systematic analyses of the features of apple tree transpiration, overcoming the incompleteness of many other studies in China and abroad, which provided insufficient continuous experimental data. The study further enriched and improved theories of apple tree transpiration.2) The transpiration characteristics of apple trees based quantitative statistics or statistic model was external impression, in order to further deeply reveal the internal complex pattern of transpiration, this paper took advantage of the fractal theories which provided theoretical basis for prediction by random simulation, further enriching the application of fractal theories in forestry science and meteorology.3) Based on the leaf and air temperature differential (△T) on tree canopy, significant progresses have been made in monitoring farmland soil water, and in identifying extent of crop water stress. This study for the first time conducted analyses of temporal variation patterns of apple tree canopy temperature and its impacts. The possibility of using AT to predict soil water and to diagnose water stress was studied, the results provided theoretical basis and technical support to the development of irrigation decision system for the apple orchard, making further additions to the...
|
PDF Full Text Request