driving forces of land use and land cover change (lucc) in...

http://www.scar.ac.cn

Sciences in Cold and Arid Regions 2012, 4(5): 0422–0430

DOI: 10.3724/SP.J.1226.2012.00422

Driving forces of land use and land cover change (LUCC) in the Zoige Wetland, Qinghai-Tibetan Plateau

GuangYin Hu *, ZhiBao Dong, JunFeng Lu, ChangZhen Yan

Key Laboratory of Desert and Desertification, Cold and Arid Regions Environmental and Engineering Research Institute, Chi-nese Academy of Sciences, Lanzhou, Gansu 730000, China

*Correspondence to: GuangYin Hu, Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences. No. 320, West Donggang Road, Lanzhou, Gansu Province 730000, China. Tel: +86-(0)931-4967496; Email: [email protected]

Received: January 23, 2012 Accepted: June 21, 2012

ABSTRACT

The Zoige Wetland is located in the northeastern Qinghai-Tibetan Plateau, which is highly sensitive to global environment change and human disturbance because of its high elevation and cold environment, thus, it’s a hotspot for land use and land cover change (LUCC) research. We used Landsat MSS images from 1975, Landsat ETM images from 2000, and Landsat TM images from 1990 and 2005 to assess the LUCC in the study area, using GIS techniques, as well as topographic, vegetation, and soil maps combined with field investigations. The monitoring result shows that the study area’s environment degraded rapidly between 1975 and 2005, including wetland shrinkage from 5,308 km2 to 4,980 km2, sandy land expansion from 112 km2 to 137 km2, forest land decreasing from 5,686 km2 to 5,443 km2, and grassland degradation from 12,309 km2 to 10,672 km2. According to the analysis of meteorological data and social-economic statistical data, we concluded that the LUCC in the Zoige Wetland was caused by both natural and anthropogenic factors, but human activities were primarily responsible for the observed LUCC, thereby, we suggest human behaviors must be adjusted to control environmental degradation. Keywords: Zoige Wetland; LUCC; remote sensing; environmental degradation; Qinghai-Tibetan Plateau

1. Introduction

Land use and land cover change (LUCC) is a complex process caused by the interaction between natural and social systems at different temporal and spatial scales (Lambin, 2000), and LUCC dynamics are influenced by the types of land cover involved, the ecological mechanisms of succes-sion and regeneration, the physical components of the envi-ronment, socioeconomic activities and their cultural context, and meteorological phenomena or other natural factors (Dale et al., 1994; Kareiva and Wennergren, 1995; Lin-denmayer and Franklin, 1997). As LUCC studies cover large areas and time periods, making ground-based surveys pro-hibitively difficult, remotely sensed images are generally used in LUCC monitoring and analysis at a range of spatial and temporal scales (Mertens and Lambin, 1997; Rogan et al., 2003; Crews-Meyer, 2004). During recent decades, the

study of LUCC has become a prominent research topic, as LUCC has been recognized as one of the most important factors of environmental modification worldwide, especially since the International Geosphere and Biosphere Programme (IGBP) and the International Human Dimensions Pro-gramme on Global Environmental Change (IHDP) initiated their core project on LUCC in the mid-1990s (Turner II, 1995; Lambin, 2000).

The Zoige Wetland lies in the northeastern Qing-hai-Tibetan Plateau, which is highly sensitive to global en-vironment change and human disturbance because of its high elevation and cold environment. During the last decade, environmental problems in the Qinghai-Tibetan Plateau have been receiving an increasing amount of attention due to the impacts of global warming and intensifying regional development (Wang and Cheng, 2001; Dong and Chen, 2002; Liu et al., 2008). Since the 1960s, the mean annual

GuangYin Hu et al., 2012 / Sciences in Cold and Arid Regions, 4(5): 0422–0430

423

temperature in the source regions of China’s Yellow River (where the Zoige Wetland is located) increased by 0.7 °C. This is one of the largest temperature increases recorded on the Qinghai-Tibetan Plateau, and an order of magnitude higher than the range of global average values (Wang and Zhou, 1998; Yang et al., 2004). The rapid climate change in this region is causing environmental problems such as per-mafrost degradation, desertification, and grassland degrada-tion (Wang et al., 2001). The Zoige Wetland is recognized as a natural reservoir for the Yellow River, because nearly 30% of the Yellow River’s total flow originates from the wetland, and it has the most extensive distribution of high-altitude peat bogs in the world. However, beginning in the 1970s, large-scale drainage of marshes was conducted in the Zoige Wetland to increase grain and livestock production, which caused large areas of marshland degraded to grassland or even to sandy land. To protect the environment of this region, the Chinese government classified this area as a county-level

nature reserve in 1994, and upgraded it to a national nature reserve in 1998 due to its important role in water conserva-tion, bio-diversity protection, and wetland conservation. We therefore conducted this study to reveal the LUCC charac-teristics and understand its driving forces. 2. Study area

The Zoige Wetland is located between 100°45′44.19″E and 103°37′8.26″E, and between 31°51′2.21″N and 34°23′5.01″N, in the northeastern part of the Qinghai-Tibetan Plateau (Figure 1), bordered by the Min Mountains to the east, the Anyemaqen Mountains to the west, Gahai Lake to the north, and the Qionglai Mountains to the south. It covers an area of 38,147 km2 and is under the administration of Zoige, Hongyuan, and Aba counties of Sichuan Province and Maqu County of Gansu Province. The altitude of this region ranges from 2,410 to 4,964 m, with a mean altitude of 3,702 m.

Figure 1 Location of the Zoige Wetland

This region has a humid, frigid continental monsoon

climate due to its high altitude. The mean annual tempera-ture ranges from 0.6 to 1.1 °C, and the mean annual precipi-tation ranges from 654 to 780 mm. The precipitation mainly falls between May and September, accounting for more than

80% of the annual precipitation (Zhang and Lu, 2009). The mean annual evaporation ranges between 1,100 and 1,273 mm, with a mean wind speed of 3.64 m/s in winter and spring and a maximum wind speed of 35 m/s. The tempera-ture does not vary greatly between seasons, but there is a


424

long cold season and a short warm season. The main eco-nomic activity is grazing, and the residents of this region mainly belong to the Tibetan ethnic group. 3. Methodology 3.1. Data sources

We used four sets of remotely sensed data in this study, including Landsat Multispectral Scanner System (MSS) images acquired in 1975, with an 80-m spatial resolution; Enhanced Thematic Mapper Plus (ETM+) images acquired in 2000, with a 30-m resolution; and Thematic Mapper (TM) images acquired in 1990 and 2005, with a 30-m spatial res-olution. We selected images recorded between June and October, when vegetation grew well, and sandy lands are more easily recognized during this period. In addition to these data, we used topographic maps at a scale of 1:100,000 surveyed by the Chinese Mapping Agency in the early 1980s, vegetation and soil maps for ancillary materials to assist our classification of land use and land cover types during the visual interpretation process. 3.2. Image processing and interpretation

We performed geometric corrections using version 9.1 of the ERDAS Imagine software (ERDAS, Norcross, GA). The ETM+ images acquired in 2000 were georeferenced using ground control points derived from the topographic maps. The mean positional errors for georectification of the TM and ETM+ images were controlled to less than 1.5 pix-els in mountainous regions and less than 1 pixel in flat re-gions, which are acceptable for large-scale surveys. The TM images from 1990 and 2005 and the MSS images from 1975 were matched with the geometrically corrected ETM+ im-ages from 2000 by means of an image-to-image matching method provided by the ERDAS Imagine software.

To accurately classify the land cover, we investigated the study area at the beginning of this research by taking a series of photographs of each type of land cover and recording its geographic coordinates using a global positioning system (GPS) receiver, and then finding the corresponding land cover in the remote-sensing images at those geographic co-ordinates to build an interpretation symbol database for use in subsequent image interpretation. In addition, we digitized the topographic maps at a scale of 1:100,000 to provide vector data, then transformed the vegetation and soil maps to the same projection mode and coordinate system so that they could be overlaid on the other images to provide assistance with image interpretation.

Visual interpretation was based on image characteristics such as color, shape, size, shading, texture, structure, and relative spatial distribution of each class of land cover. In addition, the abovementioned ancillary materials were used to improve the accuracy of the classification. This compre-hensive analysis and logical reasoning skills are both neces-

sary during the process of visual interpretation, so the inter-preters must possess abundant geo-science knowledge. Alt-hough the visual interpretation of the TM, ETM+, and MSS images was labor-intensive and time-consuming, the result-ing mapping accuracy is higher than that provided by either unsupervised or supervised classification, which are unable to discern examples of the same land cover with different spectral characteristics or examples of different classes of land cover with the same spectral characteristics.

The polygons with different land cover types in the im-ages acquired in 2000 were labeled according to their cover class. Once the digital map of the land coverage in 2000 was complete, the polygons were copied and the segments that needed modification were changed based on the 2005 and 1990 images; they were updated by adding, deleting, or modifying lines in order to reflect LUCC from 1990 to 2000 and from 2000 to 2005. Similarly, the polygons in the 1990 land coverage map were updated (by adding, deleting, or modifying lines) and the 1990 map was compared with the 1975 land coverage map to determine LUCC from 1975 to 1990. The visual interpretation process was completed using version 9.1 of the ArcMap software (ESRI, Redlands, CA).

The interpretation results on images for the study area were validated in the field in August of 2006. Subsequent corrections were made after field validation to ensure the classification accuracy over 95%. 3.3. Land use and cover type classification system

Based on the classification system of Chinese Resources and Environment Database of the Chinese Academy of Sci-ences, we divided grasslands into three second-level catego-ries (with high, moderate, and low vegetation cover) as grassland is the main land cover in this region. We grouped all subcategories of forest, cultivated land, bodies of water, and building land under the corresponding first-level cate-gory, but we separated sandy land and wetlands from the unused land category as these land use types have special meaning in the studies of environmental change (Table 1). 4. Results

Over the whole study period in the Zoige Wetland, the largest area is occupied by grassland, followed by forest and wetland (Figure 2). In 2005, the area of grassland was 25,367 km2, accounting for 66% of the total area, the area of forest and wetland was 5,443 km2 and 4,980 km2, respec-tively, accounting for 14% and 13% of the total area. From 1975 to 2005, wetland shrank from 5,308 km2 to 4,980 km2, sandy land expanded from 112 km2 to 137 km2, forest de-creased from 5,686 km2 to 5,443 km2, building land ex-panded from 35 km2 to 52 km2, and 1,638 km2 of grassland degraded from higher vegetation coverage to lower ones (Table 2).

Table 3 shows the transition matrix of the LUCC in the sub-periods of 1975–1990, 1990–2000, and 2000–2005.


425

From 1975 to 1990, 121 km2 of the forest changed to high, moderate, and low coverage grasslands. The conversion among the three types of grassland included 206 km2 of high coverage grassland degraded to moderate coverage grass-land, 74 km2 of high coverage grassland degraded to low coverage grassland, and 41 km2 of moderate coverage grassland degraded to low coverage grassland. Wetland

mainly changed to grassland, and the expanded sandy land mainly changed from degraded grassland, besides, there was 0.34 km2 of sandy land changed from degraded wetland. The area of building land increased steadily, mainly by re-placing grassland. However, some forest, grassland, and wetland recovered during this sub-period, but their recover areas were very small.

Table 1 Description of the land use and cover type classes

Class Description Forest Trees, shrubs, bamboo forest, mangroves in coastal regions, and all other kinds of forested land. High coverage grassland

Natural grassland, improved grassland, and mowed grassland with vegetation cover >50%. These grasslands were generally distributed in areas with good water availability and had a good growth status.

Moderate cov-erage grassland

Natural grassland and improved grassland with vegetation cover of 20% to 50%. These grasslands were always dis-tributed in areas with water shortages.

Low coverage grassland

Natural grassland with vegetation cover of 5% to 20%. These grasslands experienced severe water shortages and pro-duced poor grazing conditions.

Bodies of water Natural water and water at conservation facilities. Building land Urban and rural residential areas, industrial land, mining land, and land used for transportation infrastructures. Sandy land Land covered by sand, with vegetation cover of <5%, including deserts but not including sandy beaches beside bodies

of water. Wetland Areas that is flat and low-lying, with poor drainage conditions. As a result, they are always damp, with seasonal water

logging or continuous flooding. Generally covered with hygrophytes. Unused land Land that has not been used or that would be prohibitively difficult to use (e.g., bare rock and extremely steep slopes). Cultivated land Areas cultivated with crops, including old arable land, newly cultivated land, fallow land, and land undergoing crop

rotation, as well as fruit crops and agro-forestry land mainly used to cultivate crops. This also includes beaches and coastal land that have been cultivated for more than three years.

Figure 2 Land use and cover map of the Zoige Wetland in 2005


426

Table 2 Area and percentage of total area for each land use or cover class in 1975, 1990, 2000, and 2005 in the Zoige Wetland

Class 1975 1990 2000 2005

Area (km2) % of total Area (km2) % of

total Area (km2) % of total Area (km2) % of

totalForest 5,685.77 14.9 5,530.59 14.5 5,505.08 14.4 5,442.98 14.3Grassland:

High coverage 12,309.49 32.3 11,488.15 30.1 10,839.25 28.4 10,671.57 28.0Moderate coverage 7,946.13 20.8 8,643.89 22.7 8,865.87 23.2 9,074.51 23.8

Low coverage 4,623.23 12.1 5,127.26 13.4 5,628.81 14.8 5,621.19 14.7Bodies of water 584.88 1.5 575.82 1.5 569.62 1.5 577.84 1.5Building land 34.85 0.1 43.66 0.1 48.08 0.1 52.21 0.1Sandy land 111.62 0.3 119.22 0.3 131.35 0.3 137.03 0.4Wetland 5,308.35 13.9 5,196.66 13.6 5,078.66 13.3 4,980.08 13.1Unused land 1,252.16 3.3 1,143.99 3.0 1,196.41 3.1 1,305.69 3.4Cultivated land Total

291.19 38,147.66

0.8 100

278.3238,147.59

0.7100

284.4738,147.59

0.7 100

284.4838,147.58

0.7100

Comparing with the change rates between 1975 and

1990, the LUCC rates between 1990 and 2000 showed a slower change in forest cover, with only 33 km2 of forest changed to other types of land cover. In contrast, the degra-dation of grassland accelerated, with 904 km2 of high cov-erage grassland changing to other types, including 499 km2 of moderate coverage grassland and 375 km2 of low cover-age grassland. The grassland degradation also included 200 km2 of moderate coverage grassland degraded to low cov-erage grassland. However, recovery of moderate and low coverage grassland also occurred in 292 km2 due to coun-termeasures carried out by local governments. The wetland, sandy land, and building land showed changes similar to those in the sub-period of 1975–1990 (i.e., sandy land and building land increased and wetland decreased) (Table 3).

The trends of LUCC in the sub-period of 2000–2005 were similar to the sub-period of 1990–2000. The environ-mental degradation continued and the wetland shrank stead-ily in the three sub-periods. Between 2000 and 2005, alt-hough the decrease rate of forest slowed down and the re-covery rate of grassland increased, the area of degraded grassland was still much larger than the recovered (Table 3). 5. Driving forces responsible for the LUCC 5.1. Natural factors

We analyzed the mean annual climate changes in the study area based on meteorological data recorded by the Zoige, Maqu, Aba, and Hongyuan meteorological stations from 1975 to 2005. The climate of the Zoige Wetland tended to become increasingly warm and dry between 1975 and 2005 (Figure 3). The mean increase per decade in the mean annual temperature was 0.39, 0.40, 0.47, and 0.54 °C, re-spectively, for Hongyuan, Zoige, Maqu, and Aba counties, which is much greater than the global mean of 0.03 to 0.06 °C per decade (Houghton, 1995). The increased tem-perature led to degradation of frozen soils (Wang, 1997).

The Zoige Wetland is a transitional zone between zones with permafrost and seasonally frozen soils in the northeastern part of the Qinghai-Tibetan Plateau. However, no contem-porary frozen soils could be found in 2005, and permafrost disappeared everywhere except in the surrounding moun-tains due to the continual warming. Permafrost formerly played an important role in maintaining the meadow vegeta-tion of the Zoige Wetland and other areas of the Qing-hai-Tibetan Plateau because permafrost layers maintain wa-ter around plant roots by blocking infiltration of water into deeper soil layers. This decreased the water content in the surface soil, leading to drying of the surface soil and in-creased water stress for plants, contributing to the observed LUCC.

The seasonal distribution of precipitation in the study area is extremely uneven, with nearly 80% of annual precip-itation falls between May and September, which aggravated spring and autumn droughts and resulted in grasses to dry out and turn yellow earlier in the growing seasons (Wang et al., 2005). Furthermore, the mean annual precipitation in the four counties tended to decrease between 1975 and 2005. The mean decreases in annual precipitation per decade were 42, 16, 19, and 34 mm, respectively, for Hongyuan, Zoige, Maqu, and Aba counties (Figure 3). The decrease in precipi-tation, whose effects were exacerbated by the aforemen-tioned loss of permafrost, is likely to have greatly increased moisture stress on the vegetation, leading to degradation of grasslands and marshes. 5.2. Anthropogenic factors 5.2.1 Overgrazing

Livestock husbandry is the primary industry in the Zoige Wetland. The number of livestock increased rapidly between 1975 and 2005 (Figure 4) due to the increasing demand for economic growth, which greatly exceeded the theoretical carrying capacity of this region’s grassland (Sheng and Tian,


427

2006; Niu et al., 2008). The theoretical carrying capacity for Hongyuan, Zoige, Maqu, and Aba counties was 1.29×106, 1.86×106, 1.27×106, and 1.21×106 sheep units, respectively. The number of livestock in Zoige and Maqu counties, where

the most severe LUCC occurred, was continuously greater than the carrying capacity between 1975 and 2005, although the carrying capacity also changed in response to climatic changes and human disturbances.

Figure 3 Changes in mean annual temperature and precipitation between 1975 and 2005 in the ZoigeWetland

Figure 4 Increase in the number of livestock between 1975 and 2005 in the Zoige Wetland

(Livestock data for some counties were not available in some years)


428

Table 3 LUCC transition matrix for the three sub-periods of 1975–1990, 1990–2000, and 2000–2005 (The diagonal of the table represents the areas in which land use or cover type did not change)

Class Change in area (km2)

Total area1 2 3 4 5 6 7 8 9 10 1975 to 1990

1 5,562.25 3.74 2.92 3.22 1.82 — — 0.37 0.69 — 5,575.012 47.4 12,009.95 20.99 15.27 1.66 — — 20.92 0.98 0.43 1,2117.63 50.24 205.82 7,872.94 10.95 2.1 — — 45.73 4.54 — 8,192.324 23.8 74.08 40.86 4,580.99 0.92 — — 28.24 0.03 10.55 4,759.475 0.13 2.45 — — 578.3 — — 2.48 — — 583.366 0.05 3.69 0.45 0.11 — 34.85 — — — 1.57 40.727 — 2.48 1.48 5.45 — — 111.62 0.34 — — 121.378 — 5.63 0.07 0.87 0.09 — — 5,204.53 — — 5,211.199 1.69 1.11 6.33 6.38 — — — 5.76 1,245.92 — 1,267.1910 0.21 0.54 0.09 — — — — — — 278.63 279.47

1990 to 2000 1 5,497.76 0.94 0.47 4.88 0.6 — — 0.29 0.14 — 5,505.082 4.62 10,583.95 188.2 35.91 1.12 — 0.22 23.82 0.75 0.67 10,839.263 13.93 499.21 8,227.05 67.48 0.68 — 0.39 51.51 4.8 0.83 8,865.884 9.9 375.06 200.37 4,988.86 7.45 — 0.12 46.77 — 0.28 5,628.815 0.29 0.98 1.08 0.46 565.44 — — 1.37 — — 569.626 — 2.53 0.64 0.81 — 43.66 — 0.05 — 0.39 48.087 — 6.15 4.37 2.16 — — 118.5 0.17 — — 131.358 0.23 9.77 5.17 16.29 0.53 — — 5,046.56 0.11 — 5,078.669 3.68 5.17 13 10.25 0.01 — — 26.12 1,138.19 — 1,196.4210 0.21 4.38 3.55 0.17 — — — — — 276.15 284.46

2000 to 2005 1 5,441.37 0.01 0.52 0.86 0.11 — — — 0.11 — 5,442.982 1.06 10,524.61 90.72 38.38 0.22 0.04 0.11 13.07 2.6 0.77 10,671.583 3.68 217.66 8,739.86 82.59 0.77 — 1.4 27.71 0.75 0.09 9,074.514 3.91 92.99 117.25 5,353.3 — — 0.25 52.51 0.97 — 5,621.185 — 0.34 0.2 0.44 574.24 — — 2.61 — — 577.836 — 2.71 0.67 0.33 — 48.5 — — — — 52.217 — 0.34 0.09 0.8 — — 135.59 0.17 0.04 — 137.038 — 3.37 1.55 2.33 — — 0.39 4,954.2 18.23 — 4,980.079 2.96 0.77 4.58 3.51 — — — — 1,293.86 — 1,305.6810 — 4.14 0.49 — — — — — — 279.88 284.511: Forest; 2: High coverage grassland; 3: Moderate coverage grassland; 4: Low coverage grassland; 5: Bodies of water; 6: Building land; 7: Sandy land; 8: Wetland; 9: Unused land; 10: Cultivated land.

Figure 5 shows the change in the number of livestock

and in the carrying capacity for Maqu County between 1975 and 2005 (Niu et al., 2008). The number of livestock increased rapidly, while the carrying capacity tended to decrease, suggesting that the over-grazed grassland tended to become degraded. On this basis, the grassland of Maqu County was over-grazed by 16%, 98%, 101%, and 69% in 1975, 1990, 2000, and 2005 respectively. The other coun-ties were also over-grazed, though to different extents, between 1975 and 2005. For example, Hongyuan County was over-grazed by 39% in 1991, Aba County was over-grazed by 64% in 2003, and Zoige County was over-grazed by 32%, 51%, and 63%, respectively, in 1985, 2001, and 2003 (Zhang et al., 2007).

5.2.2 Drainage of water systems

Originally, there were about 4,902 km2 of marshes in the Zoige Wetland, accounting for 4.1% of China’s total area of marshes (Yang and Wang, 2001). The mire and peat soils in the wetland remained essentially undisturbed until at least the 1950s (Cai, 1965). However, beginning in the 1970s, large-scale drainage of marshes was conducted in the Zoige Wetland to increase grain and livestock produc-tion. Nearly 1,000 drainage canals, with a total length of 2,864 km, were built. The total drained area reached 2,000 km2, accounting for 41% of the total marshes and peat lands in the Zoige Wetland. This drainage restrained or even elim-inated the development of marsh and peat soils, which sub-


429

sequently degraded into meadow soils or even aeolian sandy soils. Yang and Wang (2001) noted a successive degradation process in the wetland from mire soil or peat soil to a meadow soil or aeolian sandy soil, accompanied by a marked drop in soil fertility, when these soils were drained, especially in terms of the content of organic matter, total nitrogen, and total phosphorus. The drained soils were then subjected to further degradation when they were over-grazed. 5.2.3 Rodent damages

Rodent damage is another important factor that may have been responsible for grassland degradation in this re-gion. The local rodents include the steppe zokor (Myospalax sp.), pikas and Himalayan marmots (Marmota sp.), and voles (Microtus sp.). All of these animals burrow in the up-per layers of the soil and consume grasses and grass roots, and where these animals live at high density, this can lead to

the destruction of grassland areas (Luo and Zhen, 2006). As their natural enemies (such as the Tibetan fox, Siberian weasel and tawny eagle) have declined, combined with in-effective measures to control these herbivores, the damage they caused has become increasingly serious.

In the Zoige Wetland, nearly 3,478 km2 of grassland has been damaged by rodents. Estimates suggest that there are nearly 1,000,000 zokors in Zoige County and that they con-sume 25 kt of fresh grass annually, the equivalent of 12,500 sheep units. In Maqu County, the area of grassland harmed by rodents has increased continuously, from 700 km2 in 1993 to 1,230 km2 in 1995 and 1,670 km2 in 2001, repre-senting an increase of 970 km2 from 1993 to 2001. The area of damage in Zoige County was 235 km2 in the early 1980s, but had increased to nearly 3,000 km2 in 2003, nearly 13 times the value in the 1980s. In addition, degraded grassland provides a comfortable environment for rodents, leading to increased populations and increasing amounts of damage (Zhou and Li, 2003).

Figure 5 Change in the number of livestock and the carrying capacity of Maqu County between 1975 and 2005 (data from Niu et al., 2008)

6. Discussion and conclusion

We described the LUCC process in the Zoige Wetland using multi-temporal remotely sensed images. It was found that the environment of the Zoige Wetland degraded greatly and continuously from 1975 to 2005. The observed envi-ronment degradation appears to have been caused by a com-bination of climate change (a trend towards warmer and drier) and human activities (overgrazing and drainage of water systems). The patterns of LUCC mainly included changes from forest to grassland, decreases in grassland vegetation coverage, and changes from wetland to sandy land.

Natural factors such as climatic variations constituted the background for the LUCC. However, the range of climatic variation did not appear to exceed the threshold required for the observed LUCC (such as the increase of sandy land). Human disturbances appear to be the most important factors responsible for the LUCC in the Zoige Wetland. Evidence for this is our observation that degraded meadows could restore by themselves once they had been fenced to protect them from grazing animals. Human disturbances leading to the LUCC in the Zoige Wetland mainly include over-grazing, drainage of water systems, land reclamation for agriculture, unsustainable collection of medical herbs and peat, and mining. However, compared with over-grazing and drainage


430

of water systems, the other factors were only significant in some local areas.

The "tragedy of the commons" has led to problems such as overgrazing in common pasture areas throughout the study area, so it will be necessary to adjust the grazing rights of local peoples to prevent further overgrazing. It will be very important to strengthen the knowledge of environmen-tal protection by local peoples so they will understand the need for such changes. In addition, the government must find alternative forms of employment that will be attractive to these peoples and that will sustain them in the long term so that they are not forced to return to their old, unsustaina-ble practices. Acknowledgments: We are grateful for the financial support of the National Natural Science Foundation of China (Grant No. 41201002), Foundation for Excellent Youth Scholars of Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences (Grant No. 51Y184A61), China Postdoctoral Science Foundation funded project (Grant No. 2012M512050), and the National Natural Sci-ence Foundation of China (Grant No. 41130533, 41171010). REFERENCES Cai X, 1965. Swamps of the Zoige Plateau. Science Press, Beijing. Crews-Meyer KA, 2004. Agricultural landscape change and stability in

northeast Thailand: Historical patch-level analysis. Agriculture, Ecosys-tems & Environment, 101: 155–169.

Dale VH, Oneill RV, Southworth F, Pedlowski M, 1994. Modeling effects of land management in the Brazilian Amazonian settlement of Rondônia. Conservation Biology, 8: 196–206.

Dong YX, Chen KL, 2002. Status and driving force of sandy desertification in upper reaches of Yangtze River. Resource and Environment of Yang-tze Basin, 11(1): 84–88.

Houghton J, 1995. Global Warming. Meteorological Press, Beijing. Kareiva P, Wennergren U, 1995. Connecting landscape patterns to ecosystem

and population processes. Nature, 373: 299–302. Lambin EF, 2000. Land-Use and Land-Cover Change (LUCC), Implemen-

tation Strategy. International Geosphere-Biosphere Programme (IGBP) Report 48. International Human Dimensions Programme (IHDP) Report 10. Stockholm, Bonn.

Lindenmayer DB, Franklin JF, 1997. Managing stand structure as part of ecologically sustainable forest management in Australian mountain ash

forest. Conservation Biology, 11: 1053–1068. Liu JY, Xu XL, Shao QQ, 2008. The spatial and temporal characteristics of

grassland degradation in the Three-River Headwaters Region in Qinghai Province. Acta Geographica Sinica, 63(4): 364–376.

Luo JG, Zhen WJ, 2006. Research on the present situation of desertification of grassland in Northwest Sichuan and its control measures. Journal of Sichuan Forest Science and Technology, 27(1): 63–66.

Mertens B, Lambin EF, 1997. Spatial modeling of deforestation in southern Cameroon. Applied Geography, 17: 143–162.

Niu SW, Ma LB, Zeng MM, 2008. Effect of overgrazing on grassland deser-tification in Maqu County. Acta Ecologica Sinica, 28(1): 145–153.

Rogan J, Miller J, Stow D, Franklin J, Levien L, Fischer C, 2003. Land-cover change monitoring with classification trees using Landsat TM and ancillary data. Photogrammetric Engineering and Remote Sens-ing, 69: 793–804.

Sheng HY, Tian L, 2006. Study on grassland desertification in Ruoergai Basin. Journal of Guizhou Normal University, 24(4): 16–20.

Turner II BL, 1995. Land-Use and Land-Cover Change Science/Research Plan. International Geosphere-Biosphere Programme (IGBP) Report No. 35; International Human Dimensions Programme (IHDP) Report No. 7. Stockholm and Geneva.

Wang GX, Cheng GD, 2001. Characteristics of grassland and ecological changes of vegetations in the source regions of Yangtze and Yellow rivers. Journal of Desert Research, 21(2): 101–107.

Wang GX, Li Q, Chen GD, Shen YP, 2001. Climate change and its impact on the eco-environment in the source regions of the Yangtze and Yellow rivers in recent 40 years. Journal of Glaciology and Geocryology, 23(4): 346–352.

Wang QC, Zhou LS, 1998. Climate change analysis of the source regions of the Yangtze River and Yellow River. Environment of Qinghai, 8(2): 73–77.

Wang SL, 1997. Frozen ground and environment in the Zoige Plateau and its surrounding mountains. Journal of Glaciology and Geocryology, 19(1): 39–46.

Wang Y, Zhao ZZ, Qiao YS, Li CZ, 2005. Characteristics of the climatic variatation in Zoige in the past 45 years and its effects on the eco-environment in the area. Journal of Geomechanics, 11(4): 328–333.

Yang JP, Ding YJ, Shen YP, Liu SY, Chen RS, 2004. Climatic features of eco-environment change in the source regions of the Yangtze and Yellow rivers in recent 40 years. Journal of Glaciology and Geocryology, 26(1): 7–16.

Yang Y, Wang S, 2001. Human disturbances on mire and peat soils in the Zoige Plateau. Resources Science, 23(2): 37–41.

Zhang S, Guo H, Luo Y, 2007. Assessment on driving force of climate change and livestock grazing capacity to grassland sanding in Ruoergai. Chinese Journal of Grassland, 29(5): 64–71.

Zhang XY, Lu XG, 2009. Multiple criteria evaluation of ecosystem services for the Zoige Plateau Marshes in southwest China. Ecological Econom-ics. DOI:10.1016/j.ecolecon.2009.05.017.

Zhou H, Li H, 2003. A discussion on the causes and counter measures of the aeolian desertification on the Zoige Grassland. Sichuan Grassland, 1: 35–41.

driving forces of land use and land cover change (lucc) in...

Documents