Characteristics of Changes in Livestock Numbers and Densities in the Selinco Region of the Qinghai–Tibetan Plateau from 1990 to 2020

Abstract

A thorough understanding of the development process of grazing activities and an elucidation of their complex mechanisms are crucial for the formulation and adjustment of livestock management policies. In the Selinco region of the Qinghai–Tibet Plateau, we conducted a comprehensive analysis of year-end livestock numbers and densities over the past 30 years. The results indicate a gradual decline in overall livestock numbers in the Selinco region during this period, with a notable decrease between 2004 and 2014, followed by stabilization. Notably, the number of yaks has significantly increased, whereas the numbers of sheep, goats, and horses have markedly decreased. Regarding livestock density, there is a spatial pattern of decrease from northwest to southeast, with the density order being Seni District > Bange County > Anduo County > Shenzha County > Nima County > Shuanghu County. Between 2004 and 2014, most counties experienced a significant decline in livestock density, exhibiting a trough–peak pattern. However, after 2014, a complex spatiotemporal dynamic emerged. Concerning driving factors, from 1990 to 2004, rural population and economic development were the primary influences on livestock density. After 2004, forage–livestock balance management policies, snowstorms, and fluctuations in livestock prices likely became the main influencing factors. Further detailed analysis of these factors is essential for developing more effective management strategies.

1. Introduction

The origins of grazing activities on the Qinghai–Tibet Plateau trace back to ancient nomadic societies [1]. Ancient nomads, utilizing yaks and sheep as primary economic assets, sustained an ecological balance through continuous migration in search of suitable pastures and water sources [2]. This nomadic lifestyle was not merely an economic pursuit but also embodied a harmonious coexistence with nature. Over time, with population growth and societal advancement, herding practices transitioned from traditional nomadic forms to more sedentary patterns [3,4]. This shift introduced new challenges, notably overgrazing and excessive pastureland exploitation, resulting in gradual ecological deterioration across the Qinghai–Tibet Plateau [5,6,7]. In response to these challenges, the government implemented a series of forage–livestock balance policies [8,9]. The primary objective of these policies was to preserve the ecological equilibrium of the grasslands and mitigate environmental issues arising from overgrazing. The effectiveness of these policies varies across regions due to the severe climatic conditions and uneven distribution of grassland resources on the Qinghai–Tibet Plateau [9]. In regions primarily devoted to pastoralism in the Selinco region, environmental challenges are particularly pronounced due to its elevated average altitude, which results in harsh climatic conditions. Marked by low temperatures and scant precipitation, this area experiences substantial temperature variations at high altitudes, exacerbating the difficulties faced by livestock and grazing activities [10]. This situation poses a formidable challenge to the successful implementation of the forage–livestock balance policy, necessitating more flexible and location-specific management strategies. In recent years, considerable attention has been focused on understanding the development of grazing activities in the Selinco region and the underlying mechanisms [10,11,12]. In this context, a profound understanding of the evolution of grazing in this region and the identification of its unique driving forces will significantly impact the formulation of livestock management policies and the adjustment of environmental protection measures on the Qinghai–Tibet Plateau.

Livestock numbers and densities are critical indicators of the progression and impact of grazing activities [13,14]. The variability in livestock numbers represents a complex system that delineates the intricate interplay among population dynamics, economic evolution, environmental factors, and regulatory policies [15,16,17]. Historically, changes in livestock numbers have been intricately linked to local population trends and economic development [18,19]. For instance, population growth has correlated with increased demand for livestock products, consequently leading to a rise in livestock numbers and driving economic progress [20,21]. This correlation is particularly pronounced in rural areas, where the demand for food and resources is closely intertwined with livestock activities [22]. Moreover, fluctuations in livestock numbers also reflect human adaptability to environmental changes. Over time, societal development has heightened environmental awareness, prompting policies and proactive environmental protection measures to curtail the excessive surge in livestock numbers and avert grassland degradation due to overgrazing [23,24]. Variations in grazing livestock density are directly correlated with the utilization and management of grassland resources [25,26]. Maintaining moderate livestock density is crucial for the sustainable management of grassland resources [27]. Increased livestock density, potentially leading to overgrazing, exacerbates issues such as grassland degradation and soil erosion [28,29,30]. A nuanced understanding of the spatial and temporal variations in livestock density not only facilitates the assessment of the long-term impacts of grazing activities on grassland resources but also provides a scientific foundation for effective resource management policies. Policymakers, leveraging insights into livestock density trends, can determine when it is necessary to enforce restrictive measures to prevent irreversible environmental damage due to overgrazing.

Amidst concurrent economic development and population expansion, striking a balance between environmental conservation and economic advancement has emerged as a pivotal concern. Research on fluctuations in livestock numbers and densities is particularly vital in this framework, owing to the intricate interactions between livestock and the ecosystem. Therefore, this study will utilize nearly 20 years of livestock statistical data from the Selinco region to analyze the spatiotemporal patterns of livestock numbers and density, along with examining their significant driving factors. This will provide a foundation for future environmental protection and economic development policies in the region.

2. Materials and Methods

2.1. Study Area

The Selinco region is situated in the northwestern Qinghai–Tibet Plateau, China (Figure 1a), specifically in the northern part of the Tibet Autonomous Region. The geographic coordinates of the region range from 29°56′ to 36°28′ N and 85°3′ to 93°1′ E. Encompassing an area of 301,812 km², the elevation of the region ranges from 4139 to 6940 m (Figure 1c). The region consists of six counties within the prefecture-level city of Naqu in Tibet: Seni District, Bange, Shenzha, Nima, Shuanghu, and Anduo (Figure 1c). The Selinco region lies within a sub-cold climate zone, characterized by an annual precipitation of less than 655 mm. Influenced by atmospheric circulation and terrain, precipitation decreases from east to west and from south to north. Winters and springs are cold with strong winds. Temperatures exhibit minimal variation throughout the year, averaging between −1.1 °C and 1.6 °C [10]. Additionally, pasture grasses in the region, due to intricate and challenging environmental conditions, are low-growing and sparse but exhibit remarkable adaptability to adversity [11]. Despite their resilience, yields remain relatively low. However, the extended duration of sunshine and significant temperature differences between day and night create optimal conditions for nutrient synthesis and accumulation in the grass [31]. Given the high altitude and harsh natural conditions, livestock husbandry has emerged as the predominant industry in the region [32]. Grassland husbandry not only provides natural dairy and meat products for the local populace but also significantly impacts the economic, cultural, and ecological development of the region [10]. The region’s grasslands constitute the most vital and expansive natural ecosystem, providing essential conditions for the livelihoods of the majority of herders [12]. Furthermore, the region’s poor livability combined with its vast expanse results in extremely low population densities in most areas [33]. Particularly, the area north of Shuanghu County consists of extensive uninhabited zones. Consequently, the grazing density in the Selincuo region exhibits a high spatial distribution in the southeast and a low spatial distribution in the northwest (Figure 1b).

2.2. Data Acquisition and Pre-Processing

We actively collaborated with the county statistical departments and agricultural and rural bureaus in the Selinco region. This collaboration enabled us to obtain county-level statistical yearbook data from 1990 to 2020. These yearbook records include not only year-end livestock counts but also essential metrics such as rural populations and per capita GDP, providing our study with comprehensive and invaluable data. To assess the overall impact of various livestock on grazing resources, we applied the methodology proposed by Bao et al. [9,10]. Specifically, we standardized livestock types—yaks, sheep, goats, and horses—into sheep units using conversion coefficients of 4.5, 1, 0.8, and 5.5, respectively. Furthermore, we assessed the factors influencing livestock density using data on annual total precipitation, annual average temperature, elevation, and normalized differential vegetation index (NDVI). The precipitation and temperature data were sourced from the Resource and Environment Science and Data Center (https://www.resdc.cn/; accessed on 5 June 2022) at a resolution of 1 km. The NDVI data were extracted from the MOD13A3 data within the MODIS dataset, with a spatial resolution of 1 km [34].

2.3. Methods of Analyzing Grazing Activities

One-dimensional linear regression analyses were conducted on livestock numbers at both regional and county levels to comprehensively characterize temporal changes. Livestock density was calculated using standardized sheep units for each type of livestock at both regional and county levels, along with their respective areas. This approach takes into account significant differences in weight, food requirements, and activity patterns among different livestock species. Direct comparison and calculation based solely on the number of livestock would result in inaccurate and unfair results. Using standard sheep units allows different types of livestock to be converted into equivalent standard units, thereby simplifying calculations and management. The calculation formula is LD = LN/A, where “LD” represents livestock density, “LN” denotes standardized sheep units of livestock numbers, and “A” stands for the area at the regional or county level.

3. Results

3.1. Change in Year-End Stock of Livestock

3.1.1. Various Types of Livestock Changes at the Regional Level

Figure 2 illustrates the changes and trends in livestock numbers from 1990 to 2020. The analysis reveals a significant increase in yak numbers, contrasting with the declining trend observed in all other livestock species, albeit more gradual for goats. Specifically, yak numbers showed erratic fluctuations between 1990 and 1998, followed by a pronounced and sustained upward trajectory from 1998 onward. Notably, the most substantial increase occurred between 2000 and 2007. However, after 2008, the yak numbers gradually stabilized, reaching their peak at 1,103,700 in 2019, with a low of 703,000 recorded in 1998. In contrast, the sheep number exhibits a distinct pattern, characterized by a consistent overall decline, particularly since 2004, and stabilization after 2014. Remarkably, sheep numbers showed a significant surge from 2000 to 2004, peaking at 3,727,900 in 2004 and plummeting to a minimum of 2,169,300 in 2020. Goats, exhibit a gradual decline, with a notable increase from 1996 to 2002, a stabilization period from 2002 to 2011, and a subsequent sharp decline after 2011. The goat number reached its peak in 2004 at 1,425,400 and hit its lowest point in 2020 with only 726,400. From 1990 to 2004, horse numbers remained relatively stable. However, a distinct downward trend has been evident since 2004. While numbers experienced a slight increase from 1998 to 2003, the descending trajectory intensified after 2004, notably between 2011 and 2012, when numbers plummeted from 40,500 in 2011 to 34,400 in 2012.

3.1.2. Changes in Year-End Stock of Various Types of Livestock at the County Level

Data on livestock numbers at the county level were utilized to construct a line graph illustrating changes from 1990 to 2020 (Figure 3). Observations reveal distinct patterns in livestock numbers across all counties within the Selinco region. For county-level yak numbers, Seni District and Anduo County showed no notable patterns during 1990 to 1998 but displayed a clear increasing trend from 1998 onwards. This trend is linked to their southeastern locations in the Selinco region, which feature more favorable climatic conditions conducive to yak survival and reproduction. In contrast, Bange, Nima, and Shuanghu counties experienced relatively stable yak numbers throughout the study period. This stability is attributed to these counties having a smaller overall livestock base, resulting in minimal damage to grasslands by yaks and, consequently, less disruption from the grass balance policy. Under these conditions, these counties have ensured the sustainable use of grassland resources, maintaining relatively stable yak herds. Regarding sheep, goat, and horse numbers at the county level, there has been a significant overall decline since 2004. Notably, there has been a decrease in the number of sheep across all counties. Furthermore, while the number of goats is also decreasing, it is important to note that Nima County has a significantly higher number of goats compared to the other counties. Regarding the number of horses, Nima, Shenzha, and Shuanghu Counties have relatively small numbers with stable changes, whereas the other counties have exhibited a significant downward trend since 2004.

3.2. Changes in Livestock Density

3.2.1. Overall Change Trends and Influencing Factors

Figure 4 illustrates the variations in livestock density, rural population, and per capita GDP in the Selinco region over the past three decades. The survey results indicate a significant and steady increase in the rural population and per capita GDP from 1990 to 2020. However, a gradual decline in overall livestock density was observed over the same period. Specifically, from 1990 to 2000, livestock density exhibited stable fluctuations. Subsequently, there was a substantial increase in livestock density from 2001 to 2004, peaking at 31.5 sheep units/km² in 2004. Following this, livestock density noticeably declined from 2004 to 2014, reaching a low of 24.96 sheep units/km² in 2014. The trend then stabilized, with a minor peak observed in 2019. Conversely, the rural population and per capita GDP in the study area demonstrated a significant increasing trend throughout the entire study period. For instance, the decline in livestock density from 2004 to 2014 may be attributed to local policies such as the “Returning Pasture to Grassland” policy, the “Planning for the Protection and Construction of Ecological Security Barrier in Tibet,” and the grassland ecological subsidy and reward mechanism. Additionally, livestock densities in the region are correlated with snowstorms. For example, the Chinese government website (https://www.xizang.gov.cn/, accessed on 31 May 2024) reported that in 2009, large areas of Shuanghu, Shenzha, Nima, Bange, and Anduo Counties experienced persistent and heavy snowfall, which covered pastures and resulted in the death of a large number of livestock.

3.2.2. Spatial Distribution of Livestock Density for Various Types of Livestock

To understand the geospatial distribution pattern of livestock, we averaged and spatially analyzed the livestock densities of total livestock, yaks, sheep, goats, and horses from 1990 to 2020 (Figure 5). The distribution pattern of total livestock revealed higher livestock densities in the southeast compared to the northwest. Seni District exhibited the highest grazing intensity at 142.90 sheep units/km², while Shuanghu County displayed the lowest livestock density at 4.84 sheep units/km². Notably, the livestock density in Seni District far surpassed that of other counties, with the highest intensity in the remaining counties reaching only 51.40 sheep units/km² (Anduo County). The distribution patterns of yak, sheep, goat, and horse livestock densities showed greater similarity. Seni District consistently recorded the highest livestock density, while Shuanghu County consistently had the lowest. Additionally, yaks and sheep significantly influence livestock density in Seni District, Anduo County, Bange County, and Shenzha County. For instance, in Seni District, the average livestock densities of yaks and sheep were 110.10 and 23.25 sheep units/km², respectively, accounting for 93% of the overall livestock density. Conversely, grazing intensity in Nima and Shuanghu Counties is predominantly attributed to yaks, sheep, and goats. The average livestock density of horses is minimal, making a negligible contribution to the total livestock intensity across all districts and counties.

3.2.3. Characterizing Spatial and Temporal Variations and Trends in Livestock Density at the County Level

By standardizing sheep units across districts and counties in the Selinco region, we calculated livestock density at the county level and conducted a univariate linear trend analysis (Figure 6). The overall fluctuation in livestock density in both Shuanghu and Nima Counties over the past three decades has consistently remained below 10 sheep units/km², indicating relatively stable development despite minor fluctuations. For instance, livestock density in Shuanghu County ranged from 4.06 to 6.51 sheep units/km², while in Nima County, it fluctuated between 16.11 and 22.33 sheep units/km². In contrast, significant variations in livestock density were observed in Shenzha, Bange, Anduo, and Seni. Specifically, Seni District witnessed a significant overall increase, particularly from 2000 to 2007, followed by a marked decline from 2007 to 2014, and ultimately achieving relative stability from 2014 to 2020. Conversely, Bange, Nima, and Shuanghu Counties collectively experienced a significant decrease, particularly from 2005 to 2014, while maintaining relative stability from 2014 to 2020. Shenzha and Anduo Counties generally displayed stability, particularly from 1990 to 2000, with significant declines from 2008 to 2010 and from 2011 to 2014, along with notable increases from 2000 to 2004. Notably, Shenzha County exhibited a significant upward trend from 2014 to 2020, whereas Anduo County demonstrated a significant downward trend during the same period.

3.2.4. Analysis of Influencing Factors on Livestock Density

The results of the correlation analysis, based on the correlation between livestock density and elevation, normalized differential vegetation index (NDVI), temperature, and precipitation at the township level, indicate that precipitation stands out as the primary determinant influencing livestock density in the region (

Table 1. Correlation between livestock density and DEM, NDVI, precipitation, and temperature at the township level.

	Variable	Elevation	Normalized Differential Vegetation Index (NDVI)	Annual Total Precipitation	Annual Average Temperature
Livestock density	Correlation coefficient	−0.5713 *	0.85159 *	0.61491 *	0.16335
	p-value	1.01 × 10−5 < 0.001	9 × 10−18 < 0.001	8 × 10−7 < 0.001	0.20084 > 0.05

Note: “*” denotes significant correlation.

). Precipitation serves as a crucial limiting factor for vegetation growth in grassland ecosystems, directly impacting the availability of forage for livestock [35]. Consequently, regions with ample precipitation tend to maintain higher grassland productivity, supporting a larger livestock population [36]. Additionally, the elevation of the study area emerged as a secondary factor influencing livestock density, with a correlation coefficient of −0.5 (p < 0.05). This correlation is attributed to the fact that lower altitudes generally correspond to higher temperatures and precipitation, creating favorable conditions for human habitation and optimal grassland conditions, thus facilitating livestock rearing to meet demands [37,38].

4. Discussion

4.1. Analysis of Spatial and Temporal Variations in Grazing Activities

Regarding the trends in livestock numbers and livestock densities across the Selinco region, a distinct pattern of change was observed during the period of 2000–2014. Specifically, from 2000 to 2004, livestock numbers experienced a sharp increase attributed to the concurrent trends in population and economic growth. During the period of 2004–2014, livestock densities experienced a dramatic decrease, most probably attributable to a series of forage–livestock balance policies implemented by the Chinese government for the region since 2003. Notably, all counties exhibited a decline in sheep numbers, likely attributable to forage–livestock balance policies primarily targeting sheep due to their relatively large numbers. For instance, in 2004, the Chinese government initiated a program of “Returning Grazing Land to Grassland” in the region to enhance the ecological integrity of the grasslands by alleviating grazing pressure and restoring vegetation cover [39]. Regarding the number of goats at the county level, a gradual decline in their numbers has been observed since 2011. This is, in large part, attributable to the grassland ecological subsidy incentive mechanism policy. Particularly for Seni District, being the county with the largest goat number, its declining trend is quite apparent. As for the number of horses, the grazing pressure is comparatively insignificant, with a peak value of merely 61,600 heads. Furthermore, the demand for horses among pastoralists has markedly decreased due to the widespread availability of modern transportation, leading to a pronounced decline in horse numbers throughout the study period. As the standard of living improved, horses gradually began to be used for tourism, thus their numbers tended to stabilize. For example, according to the statistical yearbook of Naqu City, the number of civilian vehicles in the region increased from 22,644 in 2009 to 63,205 in 2020.

Despite the continuous growth in the rural population and per capita GDP, the overall livestock density in the region has significantly declined. This trend indicates that the region is striving to balance economic growth with sustainable ecological development. Moreover, the livestock density in Shuanghu County and Nima County remains generally low and relatively stable, particularly in Shuanghu County, where it stayed below 7 sheep units/km² throughout the study period. This is primarily due to the relatively low precipitation in Shuanghu and Nima Counties, which directly affects local vegetation growth and subsequently leads to a significant reduction in livestock density. In contrast, other counties exhibit relatively high overall livestock densities with greater variability. Notably, in Seni District, livestock densities ranged from 111.7 to 160.21 sheep units/km² during the study period. This can be attributed to the favorable natural conditions in the area, including better precipitation, abundant vegetation, and relatively low altitude, which collectively enable the region to sustain high livestock density and promote the development of animal husbandry.

4.2. Shortcomings and Prospects

Despite the study encompassing nearly three decades of livestock data, potential interruptions and inaccuracies in data collection could compromise the precision of trend analyses. Remarkably, the Seni District consistently exhibited high livestock density over the study period, warranting particular attention and conservation efforts. Concerning influencing factors, the study solely employed simple correlation methods, neglecting to account for multifactorial influences. Therefore, to elucidate the interdependencies among various influencing factors and livestock density, future research should utilize multivariate techniques such as multiple regression or principal component analysis. Furthermore, apart from snowstorms, droughts rank among the most severe natural disasters impacting livestock production in the region. Historically, the region endured severe droughts in 1956, 1966, 1994, 1998, and 2015, inflicting substantial damage on livestock production [40]. Future investigations should dissect the mechanisms through which natural disasters (e.g., droughts and snowstorms) and anthropogenic disasters (e.g., disease outbreaks) influence livestock density to augment our understanding. This analysis could unveil the coping strategies of local herders or farmers during disasters and the potential impacts of such events on livestock density fluctuations. Furthermore, owing to considerable environmental and socio-economic disparities among regions, simplistic comparisons of livestock density may not adequately reflect the actual scenario. Subsequent research should scrutinize the spatiotemporal dynamics and trend characteristics of varying livestock densities. Additionally, livestock density is modulated by seasonal factors, yet this study did not investigate seasonal variations. Given that this study examined fluctuations in livestock numbers and density at the county level, future research should progressively extend to more granular spatial scales to furnish a more comprehensive and specific analysis of grazing activities. Ultimately, predicated on the research findings, scientific policy recommendations should be advanced to foster the sustainable development of regional livestock husbandry.

5. Conclusions

Over the past three decades, the overall livestock density in the Selinco region has progressively decreased, with the most pronounced decline observed between 2004 and 2014, after which it stabilized. Throughout the study period, livestock densities in Shuanghu and Nima Counties remained consistently low and exhibited stable fluctuations. Conversely, Seni District experienced a substantial overall increase in livestock density, particularly pronounced between 2000 and 2007, followed by a notable decline from 2007 to 2014. Notably, from 2014 to 2020, the livestock densities in Seni District, Bange County, Nima County, and Shuanghu County exhibited stable trends, while Shenzha County showed a significant upward trend and Anduo County demonstrated a significant downward trend. Future research focusing on these variations will enhance our understanding of the evolution of grazing practices, thereby providing a more robust scientific basis for sustainable ecological management.