Effect of climatic change and afforestation on water yield in the Rocky Mountain Area of North China

Aim of study: We studied effects of climatic variability and afforestation on water yield to make a quantitative assessment of the hydrological effects of afforestation on basin water yield in the Rocky Mountain Area of North China. Area of study: Seven typical forest sub-watersheds in Chaobai River watershed, located near Beijing’s Miyun Reservoir, were selected as our study object. Material and methods: Annual water yield model and Separate evaluation method were applied to quantify the respective contributions of changes in climate and different vegetation types on variations in runoff. Main results: Statistical analysis indicated precipitation did not vary significantly whereas the annual runoff decreased significantly in the past decades. Although forest increased significantly in the late 20th century, climatic variations have the strongest contribution to the reductions in runoff, with the average contribution reaching 63.24%, while the remainder caused by human activities. Afforestation has a more positive impact on the reduction in runoff, with a contribution of 65.5%, which was more than the grassland of 17.6% and the farmland of 16.9%. Research highlights: Compared to the impact of climatic change, we believe the large-scale afforestation may not be the main reason for the reductions in basin water yield.


Introduction
North China, covering an area of more than 1.5 million km 2 , is one of China's six administration regions, and has a very important political, economic and cultural position in China.Since the latter half of the 20th Century, this region has suffered from severe water shortages and soil erosion as a result of several natural and anthropogenic causes (Xia et al., 2007).To control severe soil and water losses, several large-scale forestation programs, such as the Returning farmland to 2 have a lower water yield (25 mm/yr) than adjacent basins with lower forest coverage in northwestern China.Sun et al. (2006) suggests that the average water yield reduction may vary from about 50 mm/yr (50%) in the semi-arid Loess Plateau region in northern China to about 300 mm/yr (30%) in the tropical southern region.Wang et al. (2011a) using a variety of published data showed there is a positive relationship between forest cover and the runoff coefficient (r = 0.77, p < 0.05) in Northeast China.However, contradictory reports on the impact of forest on water yield also exist.For example, Ma (1993) suggested that basins with higher forest coverage generally had higher runoff/rainfall ratios through comparing stream flow from 10 large basins (>100 km 2 ) in the Yangze River basins.Similar positive correlations between forests and water yield for large basins were reported for northern China (Wei et al., 2003).Why are there different conclusions about the relation between forest and runoff?Wei et al. (2008) argued that the effects of afforestation on runoff are not consistent, for there have been only a limited number of paired catchment studies undertaken in China.
The above studies have shown there is uncertainty and variability with the relationship between forestation on potential hydrologic responses across China due to the large differences in climate and watershed characteristics (Sun et al., 2006).Although numerous studies have been conducted in some regions and significant progress has been made in understanding forest-water relations, it is often difficult to reach a consensus on the issue of the relationship between forest and water yield, especially at the large spatial scales, and a comparative multi-basin synthesis on forest-water relationships is still needed.This is especially critical in the Rocky Mountain of northern China that has been experiencing chronic water shortages (Wang et al., 2011a).More importantly, a key issue is that current knowledge about the impacts of forestation on annual runoff at different scale within the Rocky Mountain Area of North China remains too limited to support the regional forestry development and associated water resources management.Therefore, we examined seven case study basins in this region to determine the reason for the reduction of water yield from the mountain areas and provide an objective assessment of the effects of afforestation on catchment runoff.
Hydrologic models and the classic paired-catchments approach have been widely applied for determining changes in water yield caused by land use change through excluding the effects of other environmental variables (Brown et al., 2005;Wagener 2007;Franczyk & Changk, 2009).However, the lack of standard paired watershed experiments has impeded to the development of forest hydrology research in China (Wei et al., 2008).Hydrological models also require a lot of data showed a significant decreasing trend (Xia et al., 2004).Taking the Miyun reservoir as an example, the average annual inflow to the reservoir during the period of 1980-1997 had decreased by 0.4 billion m 3 , compared with the period of 1960-1979(Wang et al., 2000)).Reductions in runoff from the mountains have put tremendous pressure on the ecological environment (Liu & Wei, 1989).Water resources issues in North China have received considerable attention from the Chinese government.At this time, some researchers believe that afforestation was primarily responsible for reducing basin water yield, and was also responsible for accelerating the shortage of water resources (Wang & Zang., 2001;Sun et al., 2006;Wei et al., 2008), which have directly affected the development of the local forestry, and also had a indirectly negative impact on the social and economic development in this region.Hence, it is urgent to understand forest and water relations in the Rocky Mountain area of North China, especially in the water source protection areas, for providing a theoretical reference for basin water resources management and forestry planning in future.
In general, climatic variation and land use change were usually considered as key factors controlling the hydrological behavior of forest watershed by affecting the generation of runoff and soil moisture (Wang et al., 2011a).Climatic variables, especially rainfall, largely determine the runoff volume of a catchment (Wang et al., 2009).Moreover, land use change caused by forestation practices result in significant impacts on hydrology by affecting the characteristics of watershed evapotranspiration, soil moisture, infiltration and groundwater recharge over a range of temporal and spatial scales (Gabris et al., 2003;Farley et al., 2005).During the past few decades, land-use changes vary dramatically in North China, mainly due to the influence of implementation of afforestation project.Increase in forest area will inevitably have an important impact on the local water balance.Furthermore, it will increase ET and reduce the annual water yield at the watershed-scale, which will also exacerbate water shortages in this region.
During the past century, much progress has been made in understanding forests and associated water relations around the world (e.g.Bosch & Hewlett, 1982;Sahin & Hall, 1996;Robinson et al., 2003;Andreassian, 2004).Although there is a large variability due to differences in climate, vegetation, terrain and soils conditions, these studies all suggest that deforestation generally increases water yield, and afforestation reduces the annual runoff for most watersheds.Simultaneously, much progress has also been made in Chinese forest hydrology research (e.g.Yu 1991;Wang & Zhang, 2001;Sun et al., 2006;Wei et al., 2008;Wang et al., 2011b).Liu and Zhong (1978) reported that forest watersheds 3 Did afforestation cause the reduction of water yield in Rocky Mountain Area of North China of most of northern cities' water supply.Miyun Reservoir, located in this region, is one of the major water supplies for Beijing.In recent years, this region has experienced a severe water resource shortage (Xia et al., 2007).Reductions in water yield from the mountains have exacerbated the water shortage problem.To explore the impact of afforestation on basin water yield, seven forested subwatersheds located in the Chaohe and Baihe watershed, the most important water source for Miyun Reservoir, were selected as our study area.No water conservation measures exist in these basins which are located upstream of the Miyun Reservoir.With the implementation of the afforestation project, the watershed forest coverage has changed greatly, and the average forest coverage reached about 60% in the late 20th century.The locations and descriptions of the basins selected for this study are shown in Figure 1 and Table 1, respectively.All of them possess a semi-arid climate.In addition, the field research did not involve endangered or protected species and no specific permits were required for the associated field research.

Data collection
Four satellite images of the study watersheds were obtained from the Institute of Geographic Sciences and which may be unavailable, particularly spatial data representing the landscape structure and its catchment properties.There is little solid, long-term (>40 years) scientific data available from studies in China (Wei et al., 2008).Due to these limitations, a method which has a sound scientific base and is of great practical use is highly desirable.Therefore, the annual water yield model (AWY) developed by Sun et al. (2005) and separate evaluation methods, which have been proved to be applicable in North China (Zhao et al., 2012), were applied to explore the main reason for the reduction of water yield in the typical forested watershed of North China.Moreover, a further objective is to provide sound scientific, quantitative information on the potential hydrologic consequences of forestry management in different regions, and offer guidance for policy making in forestation programs in North China.

Study area
The study region is the earth-rocky mountain area which is located in North China.This area is the source

Annual water yield model
The AWY model was developed by Sun et al. (2005) in southeastern United States.This model is based on actual evapotranspiration (AET) changed by land use.
Regional annual water yield (Y) at a meso-scale can be estimated as the difference between precipitation (P) input and actual evapotranspiration (AET) output: The AET can be described and estimated by the following formula: where, PET is potential evapotranspiration (mm); w is the plant-available water coefficient and represents the relative differences of water use for transpiration.Our previous studies (Zhao et al., 2012) calibrated the w parameter, and found that w coefficients were 1.5 for grass and farmland, 2.8 for forests and 0 for a water body, and settlements provided the best predictions of AET in the Chaobai River.To clarify the effects of afforestation on runoff, the deforestation effects on water yield were simulated by a reduction of the w parameter from 2.8 to 0.0.The runoff restored was simulated after deforestation, and then the percentage of decreased runoff from forests was calculated.The same method was used to estimate the effects of farmland and grass on runoff.
For a watershed with mixed land uses, Where, F i is the percentage of land use; i including forest, farmland, grass and water bodies.

Separate evaluation method
Separation evaluation was used to examine the impacts of climatic variability and land use change on runoff.It is critical to determine the inflection point in the runoff records.According to the inflection point of runoff, the study period is divided into two periods, the ''base period'' influenced slightly by human activities and the ''compared period'' marking significant changes in land use by human activities and takes the inflection point as the cut-off to reflect the effect of land use change on runoff.The method after Wang et al. (2008a) is: where ΔW T is the total change of annual runoff; ΔW H and ΔW C are the amount of annual runoff affected by human activities and climate change, respectively; W HR and W HN are respectively the measured runoff and natural runoff in human activities over the affected period.W B is the runoff in the base period.η H and η C are the percentages of hydrological variables affected respectively by human activities and climate change.

The regional characteristics of land use changes
Based on ArcGIS 9.3, land uses in different periods in the selected watersheds were obtained (Table 2) and shows that the watersheds selected in our study were all dominated by forest land, and the average forest cover generally reached 60% in the late 20th century.In addition, forest areas in the watersheds all exhibited an increasing trend to different degrees, especially around 1998, showing that vegetation restoration pro-Did afforestation cause the reduction of water yield in Rocky Mountain Area of North China higher annual variation of C V (0.59, 0.65) in the Chaohe and Baihe watersheds, respectively.
To remove the noise of large variability on trend detection, Nonparametric Kendall's trend test was used to detect the inflection point of data record of P, T and R as it has the capability of handling unusual data records (Wang et al., 2008b).Nonparametric Kendall's trend test indicated that P did not show a significant trend at the 0.05 significance level.The results were consistent with that of the linear regression tests.Compared to the Nonparametric Kendall's test for P, the annual runoff in Chaohe and Baihe watershed both showed a remarkable decreasing trend at the 0.01significance level.One inflection point in 1998 was identified.With this result, the study period may be divided into two parts: a base period  and a compared period (1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008).The results were consistent with the variation trend of R/P (Figure 3). Figure 3 shows that the significant decline in R amounts to a change from a runoff ratio (R/P) of approximately 0.11 in the first 24 years of the record to a ratio of approximately 0.04 in the last few years of the record at the 0.05 significance level.ceeded well in North China.At the same time, the reduction of water body area and increase of residential land was another distinctive feature of the land use change in this region.

Dynamics of precipitation, temperature and runoff
Average annual precipitation (P) and air temperature (T) from 1978 to 2008 in the Chaobai River was calculated by using the Kriging interpolation method, and trend analysis was also carried out for the P, T and runoff (R) in this region (Figure 2).This indicates that the long-term average annual precipitation was 437.4mm, ranging from 313 mm to 563 mm.General linear regression analysis found that the average annual precipitation showed a slightly decrease during the research period (p = 0.415) while the mean temperature markedly increased (p < 0.001), with a much lower coefficient of variation (C V = 0.08).Due to the warmer and drier climate, annual runoff showed a statistical significant decreasing trend with a much

Effects of climate variations and human activities on runoff
Climate variation and land use change caused by human activities were usually responsible for the hydrological change.As mentioned in the above section, the study period was divided into two phases: base period  and effect of human activities period (1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008).Our previous studies (Zhao et al., 2012) concluded that the AWY model performed well in estimating the contribution of changes in climate and land use on runoff after calibrating the w parameter in North China.So, combing the previous studies, the calibrated w values for grassland, farmland, forest and settlements were 1.5, 1.5, 2.8, and 0, respectively.
Model parameters in the calibration and the land use in 1998 were kept constant, whereas the meteorological data during the effect of human activities phase were inputted to the model.The results of the measured and simulated annual runoff during the evaluation period are shown in Table 3.The separate evaluation method was used to determine the quantitative effect of climatic variation and land use change on annual runoff.Table 4 indicates that climatic variation was the strongest contributor to the reduction in mean annual runoff of the Chaohe and Baihe watersheds, and the contribution reached 57.33% and 69.16%, respectively, while the remaining was caused by human activities.There are many reasons for the differences.Referring to Zheng et al. (2013), we find that precipitation and other climate factors such as temperature, wind, and solar radiation in the Chao River Basin are not very different from those in the Bai River Basin because the two basins are contiguous and their shapes are similar.Therefore, under the condition that the total annual amount of precipitation and other climate factors differed slightly, changes in the precipitation pattern are mostly assumed to be responsible for the differences in the effect of climate change on water yield, for runoff is usually caused by a few erosive rainfall events that are short and intense (Angulo-Martinez & Begueria, 2009 ).Did afforestation cause the reduction of water yield in Rocky Mountain Area of North China fects of different types of vegetation on the reduction in annual runoff were quite dissimilar.The results for the seven basins indicated that the average contribution of forest land was about 65.5%, which was more than the grassland at 17.6 and the farmland at 16.9%.On one hand, there was a larger proportion of forestland in the selected basins, which were generally > 60% in the later periods.On the other hand, the forest cover increased the surface roughness and so intercepted more precipitation, leading to a reduction of surface runoff.Combined with field surveys, the farmland is usually located in a relatively flat terrain where it is easy to conserve soil and water.However, the grassland was present as hillside meadow where it is more conducive to the generation of runoff.Therefore, we predict that runoff generation is much more sensitive to grassland than farmland.Zhao et al. (2012) reported that forest has the biggest impact on runoff in all the vegetation types in Chaobai River Watershed by intercepting precipitation and affecting the characteristics of watershed evapotranspiration and infiltration.This conclusion is also consistent with the result by Zheng et al. (2013).Zheng et al. (2013) augured that the influence of forests on annual runoff depth was significant and increased gradually from 1978 to 2008 in Chaobai River Watershed.

Conclusions
Climatic variability and land-use change in North China can 'tip' the water balance, resulting in serious social and ecological consequences.North China has experienced severe water shortages due to the combined impacts of climate change and human activities during the past decades (Li & Li, 2008).Widespread Combined with the evaluation results of the two watersheds, we assumed that the decreases in runoff between the two periods can be attributed to 63.24% from climate variations in the ChaoBai River watershed.Still, almost 36.76% of human influence is an important anthropic effect.Hence, climate change was considered to the main reason for the reduction of water yield.This conclusion is consistent with the result by Zhan et al. (2011).

Effects of different vegetation types on runoff
As mentioned above, the major causes of annual runoff reduction associated with climatic variation can be greatly accelerated by human activities such as afforestation and farming, etc.Based on the AWY model, the effects of different vegetation types on runoff were evaluated (see Table 5).Table 5 indicates that the ef-   1975-1998-35.53 -15.16 -20.37 42.67 57.33 1999-2008BAIHE 1975-1998-35.85 -11.06 -24.79 30.84 69.16 1999-2008 AVERAGE 36.76 63.24 8 afforestation has been proposed as one means of addressing the increasing dry land and stream salinity problem in North China.However, watershed hydrologic effects of forestation have not been well studied in this region.There is an urgent need to study the effects of afforestation on watershed hydrologic processes to fully understand the magnitude of water quantity responses at multiple spatial and temporal scales across China.Such eco-hydrological studies are essential to guide the recent massive afforestation and ecological restoration campaigns.
Our study suggested that climate change should be responsible for the reduction of water yield from the mountains in the past 30 years, with a contribution of 63.24%.Although afforestation plays an important role in the reduction of runoff, it was not the major reason for the reduction of water yield from mountain areas.The quantification of annual runoff response to afforestation in our study is just a first step towards a better understanding of the impacts of land cover change on water resources in North China.Further research should focus on the effect of change in climate and land use on baseflow and stormflow components at the regional to continental scale.

Figure 1 .
Figure 1.Location of the study region.

Figure 2 .
Figure 2. Annual Variation of P, T, and R with their respective linear regression line.

Figure 3 .
Figure 3. Annual variation of rainfall/precipitation ratio(R/P) in the Chaohe and Baihe watersheds.

Table 2 .
Area of land use types in each of four periods in the Chaobai River watershed hm 2 .

Table 3 .
Annual water yield of the observation and simulation over a decade evaluation period.

Table 4 .
Contributions of climate change and land-use changes to runoff.