| The semiarid region of loess hilly region of China is characterized by its fragile ecosystems and severe soil erosion. Water resource is limited and precipitation is less but concentrated. Improving the regional ecological environment is one of the primary problems. Estimating forest water consumption, especially the plantation transpiration of exotic fast-growing tree species is important for understanding characteristic of water use, regional vegetation restoration and the choice of suitable afforestation tree species in low water use. The Granier-type thermal dissipation probes(TDP) were applied to measure the tree sap flow dynamics in a natural mixed Quercus liaotungensis forest and a pure Robinia pseudoacacia plantation at Mount Gonglushan located in the southern suburb of Yan’an city of Shaanxi Province in the central part of the Chinese Loess Plateau(36°25.40’N, 109°31.53’E). Air temperature, relative air humidity, solar radiation, wind speed, and soil water content were monitored at the same time. In this paper, the characteristics of water use in two typical forest were investigated, including the diurnal course of sap flow characteristics and the relationships between sap flux density and environmental factors, azimuthal and radial variations of sap flux density on xylem trunk and effects on tree-scale transpiration estimates, estimation stand transpiration of two typical forest during the growing season of 2009, interannual variations of stand transpiration and relationships between stand transpiration and biological factors with environmental factors in 2008-2010. The main conclusions were as follows:(1) Sap flux densities(Fd) in Q. liaotungensis and R. pseudoacacia reached their daily peaks earlier than solar radiation and vapor pressure deficit, usually around 10:00 am and 12:00 am, respectively, though the diurnal courses of sap flux density were generally similar to the daily course of environmental factors. As the season and leaf phenology progressed, the overall performance of sap flux density was relatively low at early stage(April-June), high in the mid and late period(July-September), and rapidly declining in the last stage(October). Exponential saturation function was applied to the data of sap flux density and vapor pressure deficit, and the fitted curves effectively reflected the sap flow characteristics in different months. The differences in fitted curves and parameters among months suggested that the transpiration process in these forest trees was also affected by soil moisture conditions or other environmental factors.(2) Sap flux density at four aspects(north, south, east and west) on the trunk were significantly different, but were highly linearly correlated in two typical species. Daily whole-tree transpiration throughout the growing season(May to October) could be well fitted to the corresponding daily total solar radiation and average daily daytime air vapor pressure deficit using exponential saturation functions. The differences relative to tree transpiration estimates based on sap flux densities for four aspects of Q. liaotungensis were typically 30% and 18% in accordance with the sap flux densities for one and two measurement aspects, respectively. In R. pseudoacacia, the errors of tree transpiration estimates based on one, two and three measurement aspects could be 9.00%, 16.21% and 24.50%, respectively, compared with the sap flux densities for four measurement aspects.The pattern of radial variations of Fd(0-1 m, 1-2 cm and 2-3 cm) of Q. liaotungensis was falling from a maximum just below cambium to the smaller value at the inner sapwood. Different types of sap flow radial patterns were observed in Platycladus orientalis(L.) Franco. Fd values at different depths within an individual were highly correlated with each other both in two species. We evaluated the errors in estimating daily water use based on single-depth(0-1 cm) and multi-depth(i.e., every 1 cm in the sapwood) measurements. Using a single 1-cm sensor and omitting the radial variations, errors of daily tree water use may result in overestimation of 27%-90% of Q. liaotungensis. The outermost sapwood annuli(0-1 cm) contributed water use for up to 74%±6% in trees with about 2-cm sapwood depth and 65%±7% in trees with about 3 cm sapwood depth and may be applied to the estimation of tree water use incorporated with correcting coefficients. On the other hand, omitting the radial variations of sap flux density, errors of tree water use could be between underestimation of 42%±3% and overestimation of 21%±3% and application of relatively long sensors(e.g. 2 cm) may reduce the estimate errors in P. orientalis(L.) Franco.(3) In tree-scale water use, DBH could be well fitted with tree water use of Q. liaotungensis and R. pseudoacacia in different months during the growing season of 2009. Within a monthly timescale, water use by daily stand transpiration was correlated with daily solar radiation and daily daytime average vapor pressure deficit in these two forests. Daily average stand transpiration of Q. liaotungensis and R. pseudoacacia during the growing season of 2009 was 0.53 mm day-1 and 0.33 mm day-1, with the maxima of 0.98 mm day-1 and 0.74 mm day-1, respectively. Total stand canopy transpiration of two species was 97.38 mm and 71.7 mm, respectively. The difference of stand transpiration between these two typical species was 35.68 mm. The relatively low stand transpiration we estimated might be attributed to a relatively low leaf area index(LAI) and low stand sapwood area.(4) During the growing seasons of different years between 2008-2010, stand transpiration of the natural Q. liaotungensis forest were 127.85, 97.37 and 94.86 mm, respectively. Among the stand transpiration, omitting the radial variations of sap flux densities the water use of Q. liaotungensis were 85.32, 69.59 and 77.73 mm, respectively. Considering the radial variations of sap flux densities in Q. liaotungensis, the water use was 71.10, 53.14 and 46.62 mm, respectively. Ignoring the radial variations of Fd, the water use of Q. liaotungensis would be overestimation of 20.0%, 30.95% and 42.44%, respectively. The proportions of the water use of Q. liaotungensis and Armeniaca sibirica in the stand water use were basically at a stable level, indicating that the natural secondary Q. liaotungensis forest was at a stage of stable growth in these recent years. During the growing seasons of different years between 2008-2010, stand transpiration of the R. pseudoacacia plantation were 92.3, 61.7 and 66.1 mm. respectively. Daily average stand transpiration of this plantation during the different years were 0.49, 0.33 and 0.33 mm, with the maximum of daily stand transpiration of 1.07, 0.74 and 0.90 mm over the 3 years, respectively. Monthly stand transpiration of two forests showed a significant correlation with LAI, indicating that LAI could be a good scalar for estimating stand water use in this region. However, there was no obvious relationship between monthly stand transpiration and either soil water content or monthly total precipitation even during the full-leaf periods in these two typical forests. Further to analysis the relationship between precipitation and stand transpiration of Q. liaotungensis forest, there was a significant linear correlation between monthly cumulative precipitation and monthly cumulative stand transpiration(all days before the end day of one month in one year) in 3 years. However, the slopes of the linear were varied in different years, maybe attributed to the temporal and spatial distribution or intensity and amounts of precipitation over these 3 years. In an annual timescale, a positive correlation was detected between yearly total stand transpiration and average soil water content over each whole growing season, although it was not statistically significant in the R. pseudoacacia plantation. |
PDF Full Text Request