The Hydrological Processes Of Typical Vegetation And Their Slope Variations At The Diediegou Of Liupan Mountains

The hydrological impacts of vegetation and its change exhibit a significant scale effect, and this must be considered when evaluating the hydrological impacts of vegetation and managing water resources. However, the current studies were conducted on the scale of either plot or watershed. The lacking study on the key scale of slope limits the up-scaling of study results at plot scale up to the watershed scale. So it is urgent to strengthen the studies at slope scale. In this work, field measurements were performed at a grass slope and a forest(Larix principis-rupprechtii) slope within at the semi-arid small watershed of Diediegou of Liupan Mountains, Ningxia, China. The variation of vegetation and soil structure was investigated, and their hydrological impacts were compared among slope positions. As a key point, the reasons of forest transpiration difference among slope positions were examined, and a stand transpiration model which can reflecting the influences of many factors was developed. Using the absolute values of the mean change of the downwards moving average of interested parameters within the horizontal distance of 100 m, the slope scale effect of the hydrological impacts within slope scale was evaluated. The study results can promote the development of forest-water interaction theory and integrated forest-water management technologies.1. The variation of vegetation and soil structure on slopesOn forested slope, both forest canopy LAI in the middle growing season, the average tree height and DBH showed a clear difference among slope positions. The LAI differences among slope positions showed a year difference because of the different precipitation among years. In the rainfall-rich year of 2013 and the rainfall-poor year of 2015, the LAI at plot and its downwards moving average showed a tendency of firstly increase, then a decrease and a gradual decrease, with a scale effect of 0.37/100 m and 0.33/100 m, indicating a higher scale effect in wet year than in dry year. The mean DBH of trees and its downwards moving average showed a tendency of gradual decrease, with a scale effect of 0.32 cm/100 m. The mean tree height showed a tendency of firstly increase and then decrease; the downwards moving average showed a tendency of firstly increase and then tending to leveled off, with a scale effect of 0.31 m/100 m.On grass slope, the grass height and grass coverage showed a clear difference among slope positions and years. The grass height in both 2014 and 2015 showed a waving changing tendency of ‘increase-decrease-increase-decrease’; the downwards moving average of grass height showed a change of ‘decrease-increase-decrease-stabilization’ in 2014 and ‘decrease and then level-off’ in 2015, with a scale effect of 3.1 and 1.4 cm/100 m respectively. The grass coverage showed a waving decreasing tendency in 2014 and a tendency of “firstly increase and then decrease” in 2015, with a scale effect of 3.2%/100 m and 4.8%/100 m respectively.The soil physical properties on forest and grass slopes presented a different pattern of slope position difference, and the scale effect on forest slope was stronger than that on grass slope. Form slope top downwards, the soil bulk density of 0-100 cm layer showed a tendency of ‘increase-decrease-increase’ on forest slope and ‘firstly increase and then decrease’ on grass slope. The total soil porosity showed a tendency of gradually increase on both slopes. The saturated hydraulic conductivity showed a tendency of gradual increase on forest slope and ‘increase-decrease-increase-decrease’ on grass slope. The slope scale effects of soil physical properties were different among the two slopes. With increasing horizontal length from slope top, the 0-100 cm soil bulk density showed a tendency of firstly fast increase and then level-off gradually on forest slope, while a gradual decrease on grass slope, with a scale effect 0.05 and 0.02 g·cm-3/100 m. The total soil porosity showed a tendency of firstly decrease and then increase on forest slope, but a gradual increase tendency on grass slope, with a scale effect of 1.77%/100 m and 1.30%/100 m respectively. The saturated soil hydraulic conductivity performed a linear increase on both slopes, with a scale effect of 0.36 and 0.27 mm·min-1/100 m on forest slope and grass slope respectively.2. Differences of hydrological impacts among vegetation types and slope positionsIn the growing seasons of 2011, 2012, 2014 and 2015, the mean precipitation was 476.0 mm, while the mean total evapotranspiration was 532.0 mm for a larch plantation stand located at the slope foot of a north-facing slope, 350.9 and 338.5 mm for a sunny and a semi-sunny grassland plot. The calculated water budget item was-74.5 mm for forest plot, far less than that on the sunny grassland plot(115.4 mm) and semi-sunny grassland plot(155.4 mm). In the drought year of 2015, normal year of 2011 and wet year of 2014, the growing season precipitation was 332, 472 and 617 mm, the evapotranspiration were 570.0, 450.6, and 567.8 mm for forest plot, 327.7, 309.2, and 381.94 mm for the sunny grassland plot, and 296.4, 297.1 and 405.3 mm for the semi-sunny grass plot, respectively. This means forests consumes generally more water from the inflow of upslope runoff or from previously stored soil water.Influenced by the slope position differences in vegetation and soil properties, a difference in water budget components among slope positions existed on both forest slope and grass slope. Downwards from slope tope of forest slope, the canopy interception decreased gradually in growing season of 2015, the forest transpiration varied in the way of ‘decrease-increase-decrease’, the forest floor evapotranspiration firstly increased and then decreased, the total evapotranspiration decreased gradually. The soil water within the 0-100 cm soil layer showed a waving trend of ‘decrease-increase-decrease-increase’ in both 2014 and 2015. The soil water potential of 0-60 cm layer in the period of Jun-Sep. of 2015 was down slope > middle slope > upper slope, and the mean daily sap flow velocity of trees was down slope > upper slope > middle slope, with a significant difference. Downwards from the slope tope of the grass slope, no clear tendency of increase or decrease was observed for the evapotranspiration, but the soil water content of 0-100 cm layer showed a waving variation of multiple “increase-decrease” in the growing season of both 2014 and 2015.3. The slope position difference in forest transpiration and influencing factorsThe difference in sap flow velocity of trees among slope positions on forest slope was an integrated result of soil water potential and meteorological factors along slope positions. The correlation and regression analysis among the mean daily sap flow velocity, soil water potential of 0-60 cm layer(?0-60), the meteorological factors throughout the growing season(Jun. to Sep.) in 2015 showed, that the main factors influencing sap flow velocity changed with slope positions. At the down slope position with higher soil water potential, the first most important factors were the mean daily solar radiation density(R) and ?0-60, while it was ?0-60 and the mean daily saturated vapor pressure deficit(VPD) at the upper slope with lower soil water potential, and VPD and ?0-60 at the middle slope position. The regression equations were obtained between the daily mean sap slow velocity and main environmental factors.The contributions of different factors on the difference in daily transpiration between the upper slope and down slope position forest plots on forest slope were quantitatively analyzed. Within the 59 days with complete observation during the period from July to September of 2015, the total transpiration at the down slope position was only 4.5 mm higher than that at the upper slope position. However, the single factor’s impact can differ greatly. Taking the upper slope plot as the reference, the difference in soil water potential between the 2 plots performed a promoting effect to transpiration(+16.4 mm), while the difference in sapwood area and terrain shade performed a limiting effect(-6.6 mm and-2.9 mm), and the interaction among the three factors performed also a limiting effect(-2.4 mm). This means that when scaling up the value of transpiration from a plot scale up to a slope scale at the semiarid area, it is necessary not only to consider the stand structure difference(e.g., sap wood area) along slope positions, but also focus on the difference in soil moisture affected by the rainwater redistribution on slope, because this can have a big contribution higher than the sum of all other factors.4. A forest transpiration model reflecting the impacts of multiple environmental factorsAll factors influencing forest transpiration were attributed to three aspects: the evapotranspiration pull from air(potential evapotranspiration, PET), the water supply capacity from root zone soil(0-60 cm relative soil water content, REW), and the water conveyance capacity of vegetation(canopy LAI). Then a transpiration model which is capable to reflect the impact of the three aspects was developed. Firstly, the function types of transpiration(T) response to PET, REW and LAI were determined based on measured data. T respond to PET conformed to the logistic equation, when PET < 4 mm?day-1, T increased rapidly with PET, but when PET > 4 mm?day-1, T increased slowly, and when PET > 5.4 mm?day-1, T began to decrease. T response to the REW and LAI were accorded with the saturated exponential equation, when REW < 0.4, T increased fast with increasing REW, but when REW > 0.4, T increased slowly. Because the measured maximum LAI was only 5.35, T increased nearly linearly with rising LAI, until LAI > 4 where T increased gradually slowly. A multiple multiplication function among the T and the three aspects was set up, and then calibrated by the measured data in 2010, 2012 and validated by the measured data in 2014. The simulation was very good, with a Nash coefficient of 0.88 and 0.87 in 2010 and 2012 and 0.79 in 2014 respectively. This indicates that the model is able to accurately estimate the daily transpiration of Larix principis-rupprechtii plantation under changing factors of PET, REW and LAI.5. The scale effect of hydrological impacts by vegetation on slopesOn the grass slope, the hydrological impacts by vegetation variation along slope positions showed a clear slope scale effect and years’ difference. For the growing season of rainfall-rich 2014 and the drought 2015, along with the horizontal slope length increase from slope top, the downwards moving average of evapotranspiration firstly decreased gradually and then tended to be stable, with a slope scale effect of 12.3 and 3.5 mm/100 m; the moving average of soil water storage of 0-100 cm layer firstly increased weakly and then tended to be stable, with a slope scale effect of 11.6 and 11.3 mm/100 m; the moving average of water yield firstly decreased gradually and then tended to be stable, with a slope scale effect of 21.2 and 19.0 mm/100 m. All theses indicated a stronger slope scale effect of hydrological impacts by the vegetation on grass slope in wet year than in dry year.On the forest slope, the hydrological impacts by vegetation variation along slope positions showed also a clear slope scale effect. For the growing season in 2015, along with the horizontal slope length increases from slope top, the moving average of evapotranspiration firstly decreased gradually and then tended to be stable, with a slope scale effect of 20.1 mm/100 m; the moving average of soil water storage in 0-100 cm soil layer basicly decreased gradually, with a slope scale effect of 8.3 mm/100 m; the moving average of water yield firstly increased and then tended to be stable, with a slope scale effect of 30.4 mm/100 m.