Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Positive ecological effects of wind farms on vegetation in China’s Gobi desert


With the rapid development of wind power, there are increasing concerns about the negative ecological effects of its construction and operation. However, previous studies have mainly focused on the effects of wind farms on flying fauna (i.e., birds and bats) or climate change separately from communities or ecosystems, and little attention has been paid to vegetation during wind farm operation. Furthermore, few studies have referred to vulnerable ecosystems with low biomass and biodiversity. In this research, a field study was conducted to investigate the effects of wind farms on the individual traits, community structures and ecosystem functions of Gobi Desert ecosystems. The effects were measured by comparing interfering areas (IAs, located between 40 m and 90 m in the downstream direction of the wind turbine) with non-interfering areas (NIAs, located over 200 m from the wind turbine matrixes). The results showed that (1) plant individuals in IAs were less stressed and in better physiological states than those in NIAs; (2) for community structures, IA plants tended to be shorter and denser and had a higher coverage condition than that of NIA plants; and (3) ecosystem functions in IAs were significantly improved due to the existence of shrubs and higher biomass. Meanwhile, significant correlations were identified between the wind wake caused by the large spinning blades and the community structures. Constructing wind turbines in the Gobi Desert is a win-win strategy that both contributes to the growth of desert vegetation with a favourable microclimate and sufficiently utilizes wind power to produce clean energy.


Compared to traditional fossil fuels, such as petroleum, coal and natural gas, wind energy is an efficient renewable energy source that can significantly alleviate severe energy shortages, air pollution, and environmental degradation1. Hence, under the current energy policies that encourage low or zero greenhouse gas emissions, wind power has become one of the fastest-growing energy sources, and the global cumulative installed wind capacity nearly doubled from 283 GW in 2012 to 539 GW in 20172. Wind energy supplied 3.7% of the world’s electricity demand in 2015. It is expected to reach 15–18% by 2050 at a capacity of 2.3–2.8 TW, which represents a shift from fossil fuels to renewable energy resources1,3.

The number of wind farms has been increasing rapidly in response to the demand for renewable energy. In contrast to the beginning period of the 1980s, when wind energy was thought to be absolutely “clean” and completely free from having any risk of environmental impact, the rapid development and spatial extent scale of wind farm facilities has raised concerns about their negative effects4. First, the construction of wind farms may result in an increasing risk of collision and habitat loss of flying fauna (especially birds and bats)5,6,7,8. Second, the large number of wind farms may produce noticeable climatic change due to the alteration of kinetic energy fluxes and the encompassed turbulent flow phenomena9. Previous research that focused on global climate change documented that a net near-surface warming of 0.7–1.5 °C, together with zonal precipitation that decreased by approximately 1%, occurred in areas covered by wind farms according both to satellite observations9,10,11 and model simulations12,13,14. However, the effects of wind farms on vegetation have not been explored3. For example, life cycle assessments, which have been widely used to quantify the environmental impacts of wind farms across the life cycle15,16,17,18, have often excluded the impact on vegetation, even though there has been recent evidence of change19.

China has led the global wind market for nine consecutive years, i.e., since 2009, accounting for 35% of the global installed capacity2. However, the development pattern of China’s wind power is different from that of traditional wind energy powerhouses. In most European countries, wind farms are mainly located in areas with high biodiversity, such as in areas along coastlines, while inland wind farms rather than coastal and offshore wind farms dominate in China20. Over one-third of the national installed wind capacity is located in the sparsely populated drylands or in the Gobi Deserts in north-western China; these areas are rich in wind resources and are suitable for massive wind power development21. Guazhou, which is in Gansu Province and is known as the “World Wind Library”22, owns the largest wind farm cohort in the world. Currently, the wind energy generated in Guazhou can reach levels greater than 20 GW, which is roughly equivalent to that of all of Spain (i.e., 23 GW in 2017, ranked fifth in the world after Germany)2,23. Furthermore, we should not overlook the important roles of the desert ecosystems in terms of biodiversity and carbon sinking; additionally, deserts are home to 6% of the world’s population, and together with other dryland ecosystems, harbour almost one-third of the terrestrial global carbon stock24. However, current research on the ecological effects and the evaluation of wind development in deserts has not kept pace with recent progress and is largely dwarfed by the large size of wind farms in these areas.

This study explores the effects of wind farms on vegetation by using the Gobi wind farms in Guazhou as a case study. The differences in the spatial patterns around the wind turbines in terms of individual traits (i.e., metabolic scaling exponent and metabolic level), community structures (i.e., vegetation coverage, community height, and density), and ecosystem functions (aboveground biomass, AGB) as well as their response to the wind wake were investigated in the Gobi Desert in Guazhou. The objectives of this study are to (1) explore whether inland wind farms in deserts have significant effects on vegetation and whether the effects are positive or negative; (2) quantify the vegetation variations caused by the operation of wind farms in ecosystems; and (3) analyse whether the effects are related to local climate change (i.e., the change in wind velocity).

Materials and Methods

Study area

The field study was carried out at the First Guazhou Wind Farm (95°17′12”−95°34′43″E, 40°45′31–40°36′2″N), Guazhou City, Gansu Province, which is in the arid area of north-western China. Guazhou experiences a typical continental climate, with an uneven precipitation and temperature pattern. The mean annual precipitation is 52.6 mm, with over 58.1% occurring from June to August, the mean annual temperature is 9.6 °C, and the highest temperature occurs in July and August. The prevailing wind is in the east direction. Winds predominantly (80%) flowed from the east in the year prior to our field study (from June 2015 to May 2016), which is in accordance with the prevailing wind direction. All climatic data are from the China Meteorological Data Sharing Service Center (2007–2016, The First Guazhou Wind Farm is located 7 km north of the county centre in the Gobi Desert; construction began in 2009, and the wind farm began operating in 2010. No vegetation damage other than infrastructure occupation occurred during the construction period, and no man-made maintenance of the ecosystem was carried out during the operation period. Because it has been in operation for nearly one decade, the ecosystem has adapted to the existence of wind turbines and become stable. The design of the wind turbine generator system is based on FD77-1500, which features a rotor diameter of 77 m and a hub height of 82 m. The foundation is a shallow buried reinforced concrete bucket foundation, which is 16 m in diameter, 3 m in depth and concreted in situ; after excavation, the hole was filled with backfill soil (631 m3), concrete (332.5 m3) and reinforced steel (29.5 t). The study site is a typical shrub-dominated desert that has formed a stable self-organized patchiness pattern25.

Field study design

The experiment was conducted between June and August 2016. Seventy-five interfering area (IA) quadrats were established in the downstream direction (west) of 15 independent wind turbines. The IA quadrats were distributed along a distance gradient of 40 m to 90 m from the centre of a wind turbine (Fig. 1C,D) because each generator requires an area of clear ground with a radius of 30–40 m, and this area represents the transition from the foundation to natural vegetation areas and must be excluded. The identification of the farthest distance was based on the stable wind velocity according to the extended Jensen model26,27, and previous studies have shown that the effect on birds is not significant at distances exceeding 100 m7,28,29,30. A total of 30 non-interfering area (NIA) quadrats were randomly established away from the wind turbines to minimize the spatial heterogeneity (>200 m between wind turbine matrixes, Fig. 1A). The threshold of 200 m was based on the distance between single wind turbines because the design of the wind farm layout must maximize the efficiency and minimize the power loss caused by wind wakes27. All sampling quadrats were 10 × 10 m in size and were selected at sites that were far from transmission lines, roads (Fig. 1B) and ditches (Fig. 1A) to reduce possible disturbances.

Figure 1

The illustration of the field study design. Matrixes of wind turbines in the wind farm (A) and a satellite picture of a single wind turbine (B). White and red squares indicate the different experimental areas. Images (A,B) are from Google Earth (version The pattern of the experimental design is shown in the horizontal view (C) and in the vertical view (D). Dashed circles and ellipses divide the five experimental annuli along the radius to the centre of a wind turbine. Blue arrows show the wind direction during the experiment. Each of the green square represents a 10 × 10 m quadrat, and the light green sector is the area used to establish quadrats.

Sampling and measurements

Due to the harsh environment, only three species grow in the study site, including Bassia dasyphylla (Fisch. et Mey.) O. Kuntze, Nitraria sphaerocarpa Maxim. and Ephedra przewalskii Stapf. The first species is an annual grass, and the latter two species are perennial shrubs. Three individual species were sampled in all quadrats. We measured the crown radius (crowns were viewed as ellipses in a two-dimensional space, and the geometric mean of its two axes was used as an estimation of the radius) and the height of each plant. The density was determined by dividing the number of plants in each quadrat and the area of the quadrat. The 5TE soil moisture, temperature and electrical conductivity sensor was used to measure the volumetric water content and the soil temperature. Soil moisture and temperature were continuously measured in each quadrat during the sampling. The sensor was buried at a depth of 30 centimetres during measuring. To minimize damage, only a tiny fraction of a whole plant was clipped, and the total AGB and leaf mass were calculated based on the proportion of the horizontal section area of the clipped parts over that of the whole plant based on the horizontally captured images. The leaves were separated from the stems to weigh the respective masses, and the green photosynthetic shoots of Nitraria sphaerocarpa Maxim. were classified as leaf parts, while the yellow shoots were treated as stem parts. Dry mass was weighed after oven drying at 70 °C for 48 h in the laboratory31.


Metabolic scaling model

The metabolic theory of ecology (MTE) describes the allometric scaling relation of metabolic rate and body size. Its central equation is based on a power law32,33, which can be expressed as follows:

$$Q={Q}_{0}{M}^{\alpha }$$

where Q is the mass-specific metabolic rate (i.e., respiration rate or photosynthetic rate), Q0 is a normalization constant and can be treated as an indicator of metabolic level in Glazier’s metabolic-level boundary hypothesis34. M is the body size, and α is the metabolic scaling exponent, which has been shown to vary both among taxa35 and among physiological states36.

In theory, the net photosynthetic rate has been shown to be isometric with respect to the total photosynthetic leaf mass (Ml)37. Therefore, we used the allometric relation of leaf mass and AGB

$${M}_{l}={Q}_{0}{{\rm{AGB}}}^{\alpha }$$

to describe the metabolic condition of an individual based on the variation in biomass partitioning.

Wind wake model

We used the well-known extended Jensen model to simulate the wake of wind velocity around a wind turbine at hub height26,27,38. The core of this model considers the wind turbine as having a higher surface roughness. The ratio of the wake downflow v(x) (at the downstream distance x from the wind turbine) over the inflow velocity v0 can be expressed as follows:

$$\{\begin{array}{c}\frac{v(x)}{{v}_{0}}=1-a(1+\frac{\frac{x}{d}}{\sqrt{{(\frac{x}{d})}^{2}+\frac{1}{4}}}),x\le {x}_{0}\\ \frac{v(x)}{{v}_{0}}=(1-exp(-\,\beta \frac{x}{d})[1-(1-\sqrt{1-{C}_{T}}){(\frac{d}{d+2kx})}^{2}],x > {x}_{0}\end{array}$$

where a is the flow induction factor, which is defined as u = (1-a)v0 (u is the wind velocity in the hub plane) according to actuator disc theory; d is the turbine diameter; CT is the thrust coefficient t derived from a: CT = 4a(1 − a); and k is the wake expansion coefficient. In this study, CT = 0.743, a = 0.2465, d = 77 m, β = 1, and k = 0.0939 according to the manual of wind turbine generator system: FD77-1500 and the local wind velocity.

Because the wake is also embedded in the atmospheric boundary layer (the bottom part of the atmosphere, which is directly influenced by its contact with the planetary surface) with its mean velocity shear, we used a popular power law derived from a large cohort of empirical data to evaluate the change in the wind velocity in the vertical profile as follows:

$$\frac{v({h}_{1})}{v({h}_{2})}={(\frac{{h}_{1}}{{h}_{2}})}^{\alpha }$$

where v(h1) and v(h2) are the wind velocities at heights h1 and h2, respectively, and α is the wind shear exponent. Here, we used the one-seventh power law, where α = 0.14, as many investigators have done for fairly flat terrains39.

Statistical analysis

The regressions, ANOVAs and correlation analysis were conducted in R (3.1.3). Ordinary least square regression analyses were used to establish scaling relationships between leaf mass and aboveground biomass. One-way ANOVAs were performed to test the differences between normally distributed groups. To test the differences between different groups of non-normal data, Kruskal-Wallis (KW) post hoc tests were employed using the pgirmess40 library in R (3.1.3). To assess how the wind effects changed the relation between each pair of functional traits, classical Pearson’s correlation tests were performed to verify the relationships of vegetation traits within IA and NIA plots.


Individual physiological states

Regression analyses of the log10-transformed data revealed linear relationships between the leaf biomass and the AGB of individuals (Table 1, Fig. 2). In particular, with P < 0.001 and R2 > 0.983, the slopes (i.e., the metabolic scaling exponents) at distances from 40 m to 90 m were 0.877, 0.979, 0.891, 1.033 and 0.991, respectively (Fig. 2). These values were statistically higher than the value of 0.75 (P < 0.001), i.e., the universal scaling exponent in the MTE, but lower than 1.15 (P < 0.001) for the NIA plants (Fig. 2A–E). Compared with the NIA plots, the metabolic level (Q0) was significantly higher in the IA plots, ranging from −0.18 to −0.08, where the lower end corresponds to a distance of 70–80 m (P < 0.001, Fig. 3A). The metabolic scaling exponents (α) and metabolic levels (Q0) were one-to-one associated and negatively correlated (Fig. 3A,B). For the IA vegetation, the points were concentrated in the interval of low metabolic scaling exponents and high metabolic levels (the dashed-circle area in Fig. 3B).

Table 1 Parameters for regression analyses of leaf mass (Mleaf, g) and aboveground biomass (AGB, g).
Figure 2

Scaling relationships between leaf mass (Mleaf) and aboveground biomass (AGB) of plant individuals at different distances from the wind turbines. The distances in (A) to (E) are 40–50, 50–60, 60–70, 70–80, 80–90 m, respectively, and (F) represents individuals in non-interfering areas (NIAs).

Figure 3

The distributions of the metabolic scaling exponent and metabolic level of non-interfering and interfering plant individuals. (A) Metabolic scaling exponent (α) and log metabolic level (Q0) at different distances around the wind turbines. Empty circles denote α, and solid circles indicate Q0. (B) Metabolic scaling exponent versus log metabolic level (Q0). Orange and black indicate the non-interfering area (NIA) and interfering area (IA) individuals, respectively, in both (A,B). Error bars indicate standard errors.

Community structures and ecosystem functions

For community structures, significant differences were found between the IA and NIA communities (Fig. 4). The coverage of IA vegetation was higher, but not significantly, than that of NIA vegetation (P = 0.588, Fig. 4A). The community coverage at the distances of 50–60 m (mean ± standard error, 0.51 ± 0.098%) and 60–70 m (0.63 ± 0.20%) were significantly higher than those at the distances of 70–80 m (0.14 ± 0.046%), 80–90 m (0.11 ± 0.059%) and those of the NIA communities (0.37 ± 0.15%), and the differences were significant (KW post hoc: P < 0.005; Fig. 5A). The community coverage at the distance of 40–50 m presented a large variance (0.88 ± 0.69%) and was not significantly different from the groups at other distances (KW post hoc: P > 0.05). In contrast, the height of the NIA communities (8.30 ± 1.27 cm) was significantly higher than that of the IA communities (P < 0.005, Fig. 4B). The density and AGB were both significantly higher in IAs than in NIAs (P < 0.01, Fig. 4C,D). The highest density occurred at 50–60 m (6.31 ± 0.84 100 m−2) and 60–70 m (5.31 ± 0.94 100 m−2), and the densest plots (50–60 m) were nearly three times as dense as those in NIAs (2.23 ± 0.96 100 m−2, Fig. 5C). Because the emergence of large shrubs will cause a significant increase in AGB, we analysed the AGB of shrubs (AGBs) and grasses (AGBg) separately. Similar to the distribution density, the highest AGBg occurred at 60–70 m (37.29 ± 6.70 g), followed by 50–60 m (33.76 ± 2.16 g), where values were over four times that of the NIA classes (8.09 ± 1.98 g, Fig. 5D). No shrubs were found in the NIA communities; however, shrubs appeared in the IA communities in the 40–50 m and 60–70 m plots, and the AGBs values reached 3692.70 g and 1238.52 g, respectively, which were well above the values of the AGBg, which ranged from 11.95 g to 33.76 g (KW post hoc: P < 0.005; Fig. 5D).

Figure 4

The differences in the community structures and ecosystem functions between interfering areas (IAs) and non-interfering areas (NIAs). (A) to (D) represent the spatial averages of coverage, height, density, and aboveground biomass of grass (AGBg), respectively. The error bars indicate standard errors. “*”, “**” and “***” indicate the significance levels of 0.05, 0.01 and 0.005, respectively.

Figure 5

The changes in the community structures and ecosystem functions at various distances from the wind turbines. Changes in coverage (A), height (B), density (C), and AGB (D). Solid circles and empty triangles represent the aboveground biomass of shrubs (AGBs) and grass (AGBg), respectively. Orange and black represent the non-interfering area (NIA) and interfering area (IA) plots, respectively. The error bars indicate standard errors.

The results showed that the height of the community was positively correlated with the wind velocity, while the density was negatively correlated with the wind velocity, both at the 0.05 significance level (Table 2). Hence, changes in the wind velocity caused by wind turbines had opposite effects on the height and density, respectively. Wind velocity was not significantly correlated with other functional traits, including the ratio of leaf to stem, leaf water content, stem water content, coverage, and AGB (Table 2). Six paired correlations appeared to be significantly different (underlined values in Table 3). Much higher correlation coefficients were identified for stem water content and leaf water content, height and stem water content, and coverage and height in interfering conditions (Table 3).

Table 2 Correlation analysis among wind velocity and related vegetation traits at different scales.
Table 3 Correlation analysis and comparison of vegetation traits within interfering area and non-interfering area plots.


This study demonstrated that inland wind farms in deserts had positive ecological effects on vegetation during operation, which was contrary to the conclusions in the literature showing that the presence and operation of wind turbines had negative environmental impacts3. The positive effects can be explained at three different scales. At the individual scale, plants experienced a better growth environment and were in a better physiological state under the protective conditions provided by wind turbines. Considering the same or similar species, the metabolic scaling exponent (α) of individuals varies at different physiological states34. The value of α is noticeably larger (>0.75, or even >1) when water is severely limited41 or in the case of less light and low temperatures42,43,44,45. In this study, the significantly lower α in the IAs indicated that wind turbines mitigate drought stress, thus promoting the metabolism of plant individuals (indicated by Q0).

At the scale of community structure, the IA plants grew shorter but at a higher density (Fig. 5B,C), and the promotion of lateral growth over vertical growth resulted in a higher coverage condition than that of the NIA plants (Fig. 5A and Table 3). The difference in coverage was not significant, which may be due to the large variance in the coverage between 40 m and 50 m within the IA (Fig. 4C). Meanwhile, the significant positive correlations between coverage and density and between coverage and biomass indicated that the communities under the influence of wind farms had a stronger emerging trend than those in the NIA regions, where the communities were scattered (Table 3). In this study, the changes in plant communities were significantly correlated with the wind variation. The results were in accordance with the previous conclusions that wind farms could change the local microclimate and have a noticeable impact on the local vegetation and ecosystem46. However, how wind wake affects precipitation, evapotranspiration or soil moisture and further vegetation remains unclear. For example, the postulated effect of wind turbines on evapotranspiration showed an increasing trend of 0.2 mm h−1 during stable operation12; however, there were no field data supporting this hypothesis. Therefore, the quantitative relationship between local climate change and plant variation and the corresponding response of vegetation to this change have yet to be explored. Moreover, the turbine-generated climate change and the natural climate inter-annual variability should be separated during an assessment47.

Under the influence of wind farms, shrub species emerged, and overall AGB increased. The increase in biodiversity showed that the functions of the Gobi ecosystem had been greatly improved, as more diverse systems are expected to be more stable against perturbations and extreme events48,49,50. Shrubs are important modulators of physical conditions, including water availability, microclimates and soil nutrients, which are crucial for other communities in arid ecosystems51,52,53,54. In deserts, shrubs around wind turbines may improve microclimate (moister, Fig. 6) by shading to reduce evapotranspiration and reducing wind speed, thus providing a more suitable habitat for shrubs. It is convenient with previous studies that grasses emerging below shrubs may have a higher survival rate and a better physiological status than those that germinated in the open areas between shrubs55,56,57.

Figure 6

Comparison of the volumetric water content (A) and temperature (B) of soils with and without shrubs in interfering areas (IAs). “*” indicates a significance level of 0.05. Error bars indicate standard errors. Grey and black represent the soil conditions with and without shrubs, respectively.

The influences of wind farms on ecosystems are inevitably dependent on species and location46. In this study, the Gobi Desert is characterized by distinct simple structures, poor productivity and extremely harsh environmental conditions, which are distinct from the characteristics of coastal ecosystems. The desert ecosystem is highly vulnerable to any unfavourable disturbances due to a lack of water and is sensitive to environmental changes24. Using climate models, Li et al.58 found that large-scale wind farms in the Sahara caused a two-fold increase in precipitation, and the resulting increase in vegetation, led to enhanced precipitation. Our results support these simulated results, suggesting that large-scale wind farms in deserts have a positive effect on vegetation by improving microclimate. Therefore, compared with other systems with small-scale, scattered distributions, and low precipitation-vegetation feedback, changes in environmental conditions resulting from wind turbines could have a much greater effect on desert vegetation, which could be the main reason for the positive ecological effects of the Gobi wind farms. In other studies, similar positive effects were also found for specific systems and species. For instance, some land animals (e.g., tortoise) benefit from the enhanced food availability, low traffic and declining predator populations in the new habitats provided by wind power facilities59,60. In some benthic environments, new communities are established, and local fish richness increases because the wind turbine foundations may provide highly heterogeneous rocky-to-sandy habitats over the original homogeneous sandy habitats61,62,63,64. Thus, future research should quantify the effect of local habitat changes on different species so that the construction mode of wind power can be improved by reinforcing positive effects and avoiding negative effects.

Another explanation for the positive effect of wind farms observed in this study may be the strong edge effect of the areas of cleared ground, as well as that of the roads and transmission lines of wind power facilities (Fig. 1B). Surface run-off from precipitation collected along desert roads provides more water for vegetation along the road curbs than for vegetation located away from the road, resulting in enhanced productivity, diversity and perhaps even stability65,66. Port and Thompson67 found that roadside plants benefited from passing cars because they acquired additional nitrogen from exhausted gas. Plants in the vicinity of the desert transmission lines were also found to exhibit similar growth trends, i.e., they were larger, more vigorous and productive. Bolling and Walker68 reported that the decompaction of roads and the amelioration of micro-topography heterogeneity will increase the probability of a more natural community succession and lead to a smooth development trajectory. Although the disturbance of these facilities was avoided as much as possible in our study, the effect of the area of cleared ground within the 40 m radius could not be completely ignored, as these areas were adjacent to the 40–50 m quadrats. This adjacency could be responsible for the large variance in vegetation coverage in these areas closest to the area of cleared ground (Fig. 5A) and contribute to the non-significant difference between IAs and NIAs (Fig. 4A). It should be emphasized that this situation is artificial, which is away from a natural development of the plants in their habitats. However, the potential edge effects of wind turbines are worthy of our attention and should be used in the future design of wind farms to maximize their ecological benefits.

This study is the first to report the effects of wind farms on vegetation at different scales using site experiments. It is true that all facilities used for renewable energy production occupy a large area of land and require roads and transmission lines, thus resulting in habitat loss or change during the facility’s construction phase4,69,70. However, for wind energy, the ecological effects were smaller than those of other renewable energies70, and offsetting schemes could be implemented during the operation phase. The results further indicated that building wind farms in the Gobi (i.e., a good site selection) can utilize abundant wind resources with low ecological effects. Although the current wind development route of China is from inland to coastal and offshore farms20, we found empirical ecological evidence for this win-win development mode for both resources and ecology in an extremely harsh environment. Furthermore, based on our results, shrubs, such as Nitraria sphaerocarpa Maxim. and Ephedra przewalskii Stapf, can be artificially planted to improve the environment around the wind turbines during the operation phase.


Understanding the ecological effects of large-scale wind farms in Gobi ecosystems is one of the key objectives of sustainable development. To our knowledge, this study is the first to report the effects of Gobi Desert wind farms on vegetation. This study indicates that wind farms have positive ecological effects on vegetation in desert ecosystems at different scales. Under the influence of wind turbines, plants were more metabolically efficient, with higher community coverage, density, and AGB. A strong correlation was found between the changes in the community structure and the local climatic variation caused by wind turbines. Thus, building wind farms in desert areas that utilize wind energy is a cleaner production method that simultaneously promotes the development of local vegetation. It is a win-win strategy that both contributed to the growth of desert vegetation with a favourable microclimate and sufficiently utilized wind power to produce clean energy. In the future, long-term studies are required to explore the mechanism of how vegetation changes in response to variations in environmental factors caused by facilities. In practice, the positive effects of wind turbines could be better utilized in the ecological compensation and restoration of the Gobi Desert. Evaluating the impacts of habitat changes caused by wind power facilities on different systems and different animals requires a comprehensive assessment, as it is important for energy policy making. In addition to wind power in the Gobi Desert, studies should further explore whether other renewable energy facilities have similar ecological effects in deserts and other extreme environments.


  1. 1.

    Tanaka, N. Technology roadmap: Wind energy, 2013 edition. (2013).

  2. 2.

    GWEC (Global Wind Energy Council), Global wind statistic 2017, (2018).

  3. 3.

    Tabassum-Abbasi, Premalatha, M., Abbasi, T. & Abbasi, S. A. Wind energy: Increasing deployment, rising environmental concerns. Renew. Sust. Energ. Rev. 31, 270–288 (2014).

    Article  Google Scholar 

  4. 4.

    Abbasi, S. A. & Abbasi, N. The likely adverse environmental impacts of renewable energy sources. Appl. Energy 65, 121–144 (2000).

    Article  Google Scholar 

  5. 5.

    Barrios, L. & Rodriguez, A. Behavioural and environmental correlates of soaring-bird mortality at on-shore wind turbines. J. Appl. Ecol. 41, 72–81 (2004).

    Article  Google Scholar 

  6. 6.

    Mde, L., Gfe, J. & Ferrer, M. The effects of a wind farm on birds in a migration point: the Strait of Gibraltar. Biodivers. Conserv. 13, 395–407 (2004).

    Article  Google Scholar 

  7. 7.

    Drewitt, A. L. & Langston, R. H. W. Assessing the impacts of wind farms on birds. Ibis 148, 29–42 (2006).

    Article  Google Scholar 

  8. 8.

    Farfán, M. A., Vargas, J. M., Duarte, J. & Real, R. What is the impact of wind farms on birds? A case study in southern Spain. Biodivers. Conserv. 18, 3743–3758 (2009).

    Article  Google Scholar 

  9. 9.

    Zhou, L., Tian, Y., Roy, S. B. & Thorncroft, C. Impacts of wind farms on land surface temperature. Nat. Clim. Change 2, 539–543 (2012).

    Article  ADS  Google Scholar 

  10. 10.

    Rajewski, D. A. et al. Crop Wind Energy Experiment (CWEX): Observations of Surface-Layer, Boundary Layer, and Mesoscale Interactions with a Wind Farm. Bull. Am. Meteorol. Soc. 94, 655–672 (2013).

    Article  ADS  Google Scholar 

  11. 11.

    Smith, C. M., Barthelmie, R. J. & Pryor, S. C. In situ observations of the influence of a large onshore wind farm on near-surface temperature, turbulence intensity and wind speed profiles. Environ. Res. Lett. 8, 034006 (2013).

    Article  ADS  Google Scholar 

  12. 12.

    Roy, S. B., Pacala, S. W. & Walko, R. L. Can large wind farms affect local meteorology? J. Geophys. Res.: Atmos. 109, 4099–4107 (2004).

    Google Scholar 

  13. 13.

    Roy, S. B. Simulating impacts of wind farms on local hydrometeorology. J. Wind Eng. Ind. Aerod. 99, 491–498 (2011).

    Article  Google Scholar 

  14. 14.

    Wang, C. & Prinn, R. G. Potential climatic impacts and reliability of very large-scale wind farms. Atmos. Chem. Phys. 9, 2053–2061 (2010).

    Article  ADS  Google Scholar 

  15. 15.

    Chen, G. Q., Yang, Q. & Zhao, Y. H. Renewability of wind power in China: A case study of nonrenewable energy cost and greenhouse gas emission by a plant in Guangxi. Renew. Sust. Energ. Rev. 15, 2322–2329 (2011).

    Article  Google Scholar 

  16. 16.

    Xue, B. et al. A life cycle co-benefits assessment of wind power in China. Renew. Sust. Energ. Rev. 41, 338–346 (2015).

    CAS  Article  Google Scholar 

  17. 17.

    Wang, W.-C. & Teah, H.-Y. Life cycle assessment of small-scale horizontal axis wind turbines in Taiwan. J. Clean. Prod. 141, 492–501 (2017).

    Article  Google Scholar 

  18. 18.

    Yang, J. H. et al. The life-cycle energy and environmental emissions of a typical offshore wind farm in China. J. Clean. Prod. 180, 316–324 (2018).

  19. 19.

    Ji, S., Gao, C., Chen, B. & Li, S. Carbon emission accounting for wind farm based on life cycle assessment. Acta Ecol. Sin. 36, 915–923 (2016).

    Google Scholar 

  20. 20.

    Zhang, J., Zhang, J., Cai, L. & Ma, L. Energy performance of wind power in China: A comparison among inland, coastal and offshore wind farms. J. Clean. Prod. 143, 836–842 (2017).

    Article  Google Scholar 

  21. 21.

    CWEA (Chinese Wind Energy Association). 2016 Statistics of installed capacity of wind energy in China. Chin. J. Wind Energy 02, 32–44 (2017).

  22. 22.

    Chang, R., Zhu, R. & Guo, P. A Case Study of Land-Surface-Temperature Impact from Large-Scale Deployment of Wind Farms in China from Guazhou. Remote Sens. 8, 790 (2016).

    Article  ADS  Google Scholar 

  23. 23.

    Watts, J. Winds of change blow through China as spending on renewable energy soars, The Guardian (2012).

  24. 24.

    Durant, S. M. et al. Forgotten biodiversity in desert ecosystems. Science 336, 1379 (2012).

    CAS  Article  ADS  Google Scholar 

  25. 25.

    Rietkerk, M., Dekker, S. C., de Ruiter, P. C. & van de Koppel, J. Self-organized patchiness and catastrophic shifts in ecosystems. Science 305, 1926–1929 (2004).

    CAS  Article  ADS  Google Scholar 

  26. 26.

    Bastankhah, M. & Porte-Agel, F. A new analytical model for wind-turbine wakes. Renew. Energy 70, 116–123 (2014).

    Article  Google Scholar 

  27. 27.

    Stevens, R. J. A. M. & Meneveau, C. Flow Structure and Turbulence in Wind Farms. Annu. Rev. Fluid Mech. 49(págs), 311–339 (2017).

    MathSciNet  Article  ADS  Google Scholar 

  28. 28.

    Winkelman, J. E. The impact of the Sep wind park near Oosterbierum, the Netherlands on birds 3: flight behaviour during daylight. (1992).

  29. 29.

    Christensen, T. K., Hounisen, J. P., Clausager, I. & Petersen, I. K. Visual and radar observations of birds in relation to collision risk at the Horns Rev offshore wind farm. (2003).

  30. 30.

    Kahlert, J. A. Investigations of birds during construction and operation of Nysted offshore wind farm at Rødsand - Annual status report 2003: Report request. Commissioned by Energi E2 A/S. (2004).

  31. 31.

    Li, Z. T., Yang, J. Y., Drury, C. F. & Hoogenboom, G. Evaluation of the DSSAT-CSM for simulating yield and soil organic C and N of a long-term maize and wheat rotation experiment in the Loess Plateau of Northwestern China. Agr. Syst. 135, 90–104 (2015).

    Article  Google Scholar 

  32. 32.

    Kleiber, M. Body size and metabolism. Hilgardia 6, 315–332 (1932).

    CAS  Article  Google Scholar 

  33. 33.

    West, G. B., Brown, J. H. & Enquist, B. J. The fourth dimension of life: Fractal geometry and allometric scaling of organisms. Science 284, 1677–1679 (1999).

    MathSciNet  CAS  Article  ADS  Google Scholar 

  34. 34.

    Glazier, D. S. A unifying explanation for diverse metabolic scaling in animals and plants. Biol. Rev. 85, 111–138 (2010).

    Article  Google Scholar 

  35. 35.

    Glazier, D. S. Beyond the ‘3/4-power law’: variation in the intra- and interspecific scaling of metabolic rate in animals. Biol. Rev. 80, 611–662 (2005).

    Article  Google Scholar 

  36. 36.

    White, C. R., Cassey, P. & Blackburn, T. M. Allometric exponents do not support a universal metabolic allometry. Ecology 88, 315–323 (2007).

    Article  Google Scholar 

  37. 37.

    Wang, Z. et al. A theoretical framework for whole-plant carbon assimilation efficiency based on metabolic scaling theory: a test case using Picea seedlings. Tree Physiol. 35, 599–607 (2015).

    CAS  Article  Google Scholar 

  38. 38.

    Jensen, N. O. A note on wind generator interaction. Rep. Risø-M-2411, Risø Natl. Lab., Roskilde, Den. (1983).

  39. 39.

    Bechrakis, D. A. & Sparis, P. D. Simulation of the Wind Speed at Different Heights Using Artificial. Neural Networks. Wind Eng. 24, 127–136 (2000).

    Google Scholar 

  40. 40.

    Giraudoux, P. Pgirmess: Data analysis in ecology. R package version 1.5.8 (2013).

  41. 41.

    Deng, J. M. et al. Plant mass-density relationship along a moisture gradient in north-west China. J. Ecol. 94, 953–958 (2006).

    Article  Google Scholar 

  42. 42.

    Lonsdale, W. M. & Watkinson, A. R. Light and Self-Thinning. New Phytol. 90, 431–445 (1982).

    Article  Google Scholar 

  43. 43.

    Dunn, C. P. & Sharitz, R. R. The relationship of light and plant geometry to self-thinning of an aquatic annual herb, Murdannia keisak (Commelinaceae). New Phytol. 115, 559–565 (1990).

    Article  Google Scholar 

  44. 44.

    Glazier, D. S. Scaling of metabolic scaling within physical limits. Systems 2, 425–450 (2014).

    Article  Google Scholar 

  45. 45.

    Carey, N. & Sigwart, J. D. Size matters: plasticity in metabolic scaling shows body-size may modulate responses to climate change. Biol. Lett. 10, 20140408 (2014).

  46. 46.

    Armstrong, A., Waldron, S., Whitaker, J. & Ostle, N. J. Wind farm and solar park effects on plant-soil carbon cycling: uncertain impacts of changes in ground-level microclimate. Global Change Biol. 20, 1699–1706 (2014).

    Article  ADS  Google Scholar 

  47. 47.

    Vautard, R. et al. Regional climate model simulations indicate limited climatic impacts by operational and planned European wind farms. Nat. Commun. 5, 3196 (2014).

    Article  Google Scholar 

  48. 48.

    Yachi, S. & Loreau, M. Biodiversity and ecosystem productivity in a fluctuating environment: The insurance hypothesis. Proc. Natl. Acad. Sci. USA 96, 1463–1468 (1999).

    CAS  Article  ADS  Google Scholar 

  49. 49.

    De Boeck, H. J. et al. Biomass production in experimental grasslands of different species richness during three years of climate warming. Biogeosciences 5, 585–594 (2008).

    Article  ADS  Google Scholar 

  50. 50.

    Downing, A. L., Brown, B. L. & Leibold, M. A. Multiple diversity-stability mechanisms enhance population and community stability in aquatic food webs. Ecology 95, 173–184 (2014).

    Article  Google Scholar 

  51. 51.

    Polis, G. A. The ecology of desert communities. Vol. 42 (The University of Arizona Press, Tucson, 1991).

  52. 52.

    Shachak, M. et al. Woody Species as Landscape Modulators and Their Effect on Biodiversity Patterns. BioScience 58, 209–221 (2008).

    Article  Google Scholar 

  53. 53.

    Bezemer, T. M. et al. Divergent composition but similar function of soil food webs of individual plants: plant species and community effects. Ecology 91, 3027–3036 (2010).

    CAS  Article  Google Scholar 

  54. 54.

    Liu, J. L., Li, F. R., Liu, C. A. & Liu, Q. J. Influences of shrub vegetation on distribution and diversity of a ground beetle community in a Gobi desert ecosystem. Biodivers. Conserv. 21, 2601–2619 (2012).

    Article  Google Scholar 

  55. 55.

    Wrj, D. & Milton, S. J. Patch disturbances in arid grassy dunes: antelope, rodents and annual plants. J. Arid Environ. 20, 231–237 (1991).

    Article  Google Scholar 

  56. 56.

    Tielborger, K. & Kadmon, R. Relationship between shrubs and annual communities in a sandy desert ecosystem: a three-year study. Plant Ecol. 130, 191–201 (1997).

    Article  Google Scholar 

  57. 57.

    Schenk, H. J., Holzapfel, C., Hamilton, J. G. & Mahall, B. E. Spatial Ecology of a Small Desert Shrub on Adjacent Geological Substrates. J. Ecol. 91, 383–395 (2003).

    Article  Google Scholar 

  58. 58.

    Li, Y. et al. Climate model shows large-scale wind and solar farms in the Sahara increase rain and vegetation. Science 361, 1019–1022 (2018).

    CAS  Article  ADS  Google Scholar 

  59. 59.

    Lovich, J. E. & Daniels, R. Environmental characteristics of desert tortoise (Gopherus agassizii) burrow locations in an altered industrial landscape. Chelonian Conserv. Bi. 3, 714–721 (2000).

    Google Scholar 

  60. 60.

    Agha, M. et al. Turbines and Terrestrial Vertebrates: Variation in Tortoise Survivorship Between a Wind Energy Facility and an Adjacent Undisturbed Wildland Area in the Desert Southwest (USA). Environ. Manage. 56, 332–341 (2015).

    Article  ADS  Google Scholar 

  61. 61.

    Maar, M., Bolding, K., Petersen, J. K., Hansen, J. L. S. & Timmermann, K. Local effects of blue mussels around turbine foundations in an ecosystem model of Nysted off-shore wind farm, Denmark. J. Sea Res. 62, 159–174 (2009).

    Article  ADS  Google Scholar 

  62. 62.

    Lindeboom, H. J. et al. Short-term ecological effects of an offshore wind farm in the Dutch coastal zone; a compilation. Environ. Res. Lett. 1341, 35101–35113 (2011).

    Article  Google Scholar 

  63. 63.

    Stenberg, C. et al. Long-term effects of an offshore wind farm in the North Sea on fish communities. Mar. Ecol. Prog. 528, 257–265 (2015).

    Article  Google Scholar 

  64. 64.

    Vandendriessche, S., Derweduwen, J. & Hostens, K. Equivocal effects of offshore wind farms in Belgium on soft substrate epibenthos and fish assemblages. Hydrobiologia 756, 19–35 (2015).

    Article  Google Scholar 

  65. 65.

    Johnson, H. B. Productivity, Diversity and Stability Relationships in Mojave Desert Roadside Vegetation. Bull. Torrey Bot. Club 102, 106 (1975).

    Article  Google Scholar 

  66. 66.

    Lightfoot, D. C. & Whitford, W. G. Productivity of creosotebush foilage and associated canopy arthropods along a desert roadside. Am. Midl. Nat. 125, 310–322 (1991).

    Article  Google Scholar 

  67. 67.

    Port, G. R. & Thompson, J. R. Outbreaks of insect herbivores on plants along motorways in the United Kingdom. J. Appl. Ecol. 17, 649–656 (1980).

    Article  Google Scholar 

  68. 68.

    Bolling, J. D. & Walker, L. R. Plant and soil recovery along a series of abandoned desert roads. J. Arid Environ. 46, 1–24 (2000).

    Article  ADS  Google Scholar 

  69. 69.

    Mcdonald, R. I., Fargione, J., Kiesecker, J., Miller, W. M. & Powell, J. Energy sprawl or energy efficiency: climate policy impacts on natural habitat for the United States of America. Plos One 4, e6802 (2009).

    Article  ADS  Google Scholar 

  70. 70.

    Gasparatos, A., Doll, C. N. H., Esteban, M., Ahmed, A. & Olang, T. A. Renewable energy and biodiversity: Implications for transitioning to a Green Economy. Renew. Sust. Energ. Rev. 70, 161–184 (2017).

    Article  Google Scholar 

Download references


This work is financed by the National Natural Science Fund of China (31330010), the Natural Science Foundation of Zhejiang Province (LZ13C030002), and the China Postdoctoral Science Foundation (2018M640549). We thank Weiming Cao for helpful comments on early version of the manuscript and assistance during manuscript revision.

Author information




K.X., L.H. and G.W. conceived, designed the study. K.X., L.H., H.H., S.L. and Z.W. did the field work and preformed the data collection. K.X. and L.H. coordinated data analysis and interpreted results. K.X. created Fig. 1 in Adobe Illustrator CC. K.X., L.H., Y.D., Y.L., L.L., A.K. and G.W. wrote and revised the main manuscript text.

Corresponding author

Correspondence to Genxuan Wang.

Ethics declarations

Competing Interests

The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Xu, K., He, L., Hu, H. et al. Positive ecological effects of wind farms on vegetation in China’s Gobi desert. Sci Rep 9, 6341 (2019).

Download citation


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing