A Comparative Study on the Spatio-Temporal Evolution and Driving Factors of Oases in the Tarim River Basin and the Heihe River Basin During the Historical Period

Abstract

Oases are the core carriers of societal development in arid regions, and their spatial patterns have changed significantly, driven by climate change and anthropogenic activities. This study integrates historical documents, archeological materials, maps, and remote sensing data. The changes in the temperature, precipitation, settlements, war frequency, and oasis area were identified by combining quantitative and qualitative methods, and the partial least squares path model (PLS-PM) was utilized to quantify the natural and human driving factors. The results show that the oasis development in the Tarim and Heihe River Basins exhibits distinct spatio-temporal variability and phased characteristics and is comprehensively shaped by both natural and anthropogenic drivers. The Tarim Basin’s natural oases demonstrate a “fluctuating recovery” pattern. The cultivated oases gradually expanded. The natural oases within the Heihe River Basin have persistently decreased, and cultivated oases show a “U”-shaped evolution pattern. This reflects the strong intervention of human reclamation in the cultivated oases. The introverted social ecosystem has endowed the Tarim River Basin with the ability to self-repair and achieve a periodic recovery. The Heihe River Basin serves as a strategic corridor for national external engagement, relying on regime stability. A regime collapse led to its lack of a stable recovery period. The PLS-PM reveals that the Tarim River Basin oasis evolution is predominantly driven by climate fluctuations. The path coefficient of natural factors for artificial oases is 0.63, and extreme drought leads to natural oasis contraction. The human influence dominates the Heihe River Basin, with a −0.93 path coefficient linking the cultivated oasis area to human factors. The frequency of wars (load 0.74) and changes in settlements (load −0.92) are the key factors. This study provides a powerful case for the analysis of the evolution and driving mechanism of future oases in drylands.

1. Introduction

Drylands constitute 41% of the global land area and provide livelihoods for 38% of the global population. Consequently, the ecological and environmental stability of these regions is imperative for the sustainable development of human societies [1,2]. An oasis constitutes a distinctive eco-geographical landscape that has emerged under arid climatic conditions, seamlessly integrating natural elements such as water, soil, and life with social elements including human beings and industries. It serves as the fundamental environment for human survival, activity and development in drylands. Within western China, the oasis area constitutes a mere 3–5% of the total area yet supports 90% of the populace while generating 95% of the regional wealth [3,4]. The oasis plays a pivotal role in human activities and economic and social development in the Northwest Arid Zone. Oasis stability is therefore a key determinant for sustainable regional socio-economic development [5]. However, the oasis is a fragile ecosystem, susceptible to the constant changes in external natural and human factors [6,7,8].

The potential for irreparable harm to the oasis ecosystem can have deleterious consequences for the quality of life of local residents once a critical juncture is reached. Under the background of the synergistic influence of human activities and the natural environment, deeply investigating the dynamic evolutionary laws governing oases is crucial scientifically. The formulation of environmentally sustainable land management strategies and the execution of ecological restoration measures become paramount when confronted with potential hazards.

Encompassing much of the Tarim Basin, the Tarim River Basin represents a major inland drainage system in China. Its primary watercourse, the Tarim River, serves as the core of this basin’s hydrological network and plays a crucial role in ecological, economic, and social functions, sustaining oasis ecosystem stability, regional economic growth, and resident livelihoods [9]. Ranking as China’s second-largest inland drainage system, the Heihe River Basin boasts an extensive development history. Functioning as a key breadbasket for western China, the basin’s middle reaches bear critical responsibility for regional food security. The downstream Ejina area serves as a key pastoral economy hub in the west. The ecosystem is extremely fragile. Climate change and various activities carried out by humans have both had significant impacts on local ecosystems. These areas face critical ecological issues such as the degradation of natural forests and grasslands, soil salinization, and lake desiccation [10,11], and the stability of oases is under threat. The stability of oases directly affects regional social stability and ecological security. Examining basin green oasis dynamics and their driving factors is critically important for advancing oasis ecological restoration and meeting regional sustainable development goals.

Cui Weiguo [12] employed remote sensing technology combined with a GIS to reconstruct the dynamic changes in the areas of various landscape types within the Manas Oasis over the past 40 years. Li Xiaoning [13] applied landscape ecology theory and “3S” technologies (remote sensing, GIS, and global positioning system) to reconstruct the landscape pattern of the Aksu Oasis from 1990 to 2008, concluding that human activities are the primary driving force behind oasis pattern changes. Yuerqianguli Kasimu [14], supported by remote sensing (RS) and GIS technologies, analyzed the correlation between the expansion of the cultivated land area and the expansion of the oasis area, revealing the temporal and spatial characteristics of oasis expansion in the Charchan River Basin in the past 40 plus years. Numerous retrospective studies have reconstructed oasis development dynamics through the qualitative analysis of multi-source historical evidence [15,16,17]. Jie et al. employed a variety of sources, including historical documents, archeology, map information, and remote sensing data, to ascertain the distribution of oases during specific periods. Their findings suggest that the Minqin oasis has undergone a west-to-east migration process, indicating a general evolutionary trajectory [18]. Some inter-decadal-scale studies have utilized remote sensing data to quantitatively analyze oasis evolution, investigating interactions between oasis dynamics and natural–human drivers [19]. This underscores population growth and the farm policy orientation as core drivers of the Jinta Oasis (Gansu Province) expansion [20]. However, these qualitative studies, characterized by their broad descriptions, fell short of providing actionable guidelines for the sustainable management of oases [21].

However, there is a paucity of comparative analytical studies of a similar duration and scope in the historical period. This article takes nearly 2000 years as the research period, focusing on analyzing the key driving factors of oasis dynamics in the Tarim and the Heihe River Basin. The analysis is supported by a diverse range of data sources, specifically including various literature records, historical atlas materials, and the published literature in the academic field. This research endeavors to address the existing gap in knowledge by synthesizing the findings from recent studies and offering a comprehensive case study for the systematic investigation of oasis ecosystems under both natural disturbances and human interventions by comparing and analyzing the similarities and differences in oasis spatio-temporal patterns across both basins and their drivers and exploring the relative roles of climatic versus anthropogenic influences governing oasis evolution. On this basis, we propose targeted ecological protection and resource management strategies. This study’s conclusions offer a theoretical foundation supporting sustainable oasis ecosystems across northwestern China, and at the same time offer reference examples for the formulation of protection strategies and the optimization of management models of oases in other drylands.

2. Materials and Methods

2.1. The Study Area

The Tarim River Basin, located in the Xinjiang Uygur Autonomous Region in northwestern China (Figure 1), is the largest inland river basin. It extends 1321 km in total length and encompasses an area of approximately 1.02 million square kilometers. The region’s climate is a typical continental arid climate, with receives under 100 mm of annual rainfall. The annual average runoff of surface water of the Tarim River is estimated at 398 × 10⁸ m³, the average annual sunshine duration is 2400–3250 h, and annual temperature range is 10.6–11.5 °C. Tarim River Basin’s water resources are predominantly reliant on alpine ice and snowmelt, as well as precipitation recharge, resulting in uneven water distribution. The oasis along the river constitutes the primary agricultural production area and population concentration point. Straddling Gansu Province and Inner Mongolia in northwestern China, the Heihe Basin ranks as Chinese second-largest inland drainage system. The Heihe River Basin extends approximately 540 km in length and encompasses an area of approximately 142,900 square kilometers. The climate of the Heihe River Basin is also a continental arid climate. However, due to the influence of the Qilian Mountains, the upper reaches of the basin are characterized by higher precipitation levels, averaging more than 200 mm annually. The Heihe Basin’s water supply is primarily sourced from Qilian Mountain glacio-pluvial recharge.

2.2. Data Collection

The study period encompasses eight ancient dynasties of China, the Republic of China, and the People’s Republic of China (

Table 1. Different periods in the last 2000 years.

Dynasty	Period	Length of Dynasty (Years)
Han Dynasty	206 BC–220 AD	426
Wei-Jin-era Northern and Southern Dynasties	220 AD–589 AD	309
Sui Dynasty	589 AD–618 AD	29
Tang Dynasty	618 AD–907 AD	289
Song Dynasty	960 AD–1279 AD	319
Yuan Dynasty	1279 AD–1368 AD	89
Ming Dynasty	1368 AD–1644 AD	276
Qing Dynasty	1644 AD–1912 AD	267
Republic of China	1912 AD–1949 AD	38
The People’s Republic of China	Since 1949 AD–	Since 1949 AD–

). Historical materials were acquired primarily from ancient classics and various types of literature, and we also fully draw on the fundamental research work carried out by predecessors in related fields (

Table 2. Dataset information in the Tarim and Heihe River Basin.

Dataset	Timescale	Reference
Settlement	206 BC–2000 AD	Records of the Grand Historian; Han History; The Classic of Mountains and Rivers; Book of Wei; Book of Later Han Dynasty; Tang History; Song History; Yuan History; Ming History; Qing Dynasty National Chorography
Cultivated oasis	206 BC–2000 AD
Natural oasis	206 BC–2000 AD
Population	206 BC–2000 AD	(Liu et al., 2014) [22]
Temperature	206 BC–2000 AD	(Jia et al., 2017) [23]
War frequency	206 BC–1912 AD	https://www.minzushi.org/minzushi/11547.html (accessed on 21 August 2025).
Precipitation standardization	1450 AD–2000 AD	(Zhong et al., 2004) [24]
Datasets information in the Heihe River Basin
Settlement	0–2000 AD	(Shi, 2017) [25]
Cultivated oasis	0–2000 AD	(Lu, 2015) [26]
Natural oasis	0–2000 AD	(Lu, 2015) [26]
Population	0–2000 AD	(Shi, 2017) [25]
Temperature	0–1949 AD	(Wang, 2014) [27]
War frequency	0–1949 AD	(Tang et al., 2018) [28]
Precipitation	0–1949 AD	(Wang, 2014) [27]

).

Temperature and precipitation have been reconstructed for the historical period. These reconstructions utilized observational records from the contemporary era and changes in paleoclimatic conditions. Settlement and war frequency have also been studied. These data were obtained from historical document collections and the basis of previous work.

2.3. Research Methodology

2.3.1. Segmented Linear Regression (PLR)

This study uses the segmented linear regression (PLR) method to analyze the mutation points in the historical period. Segmented linear regression is a statistical method that provides separate results by switching regression to produce several parts of independent variables [29]. Linear regression uses PLR and is divided into two segments based on time. Under the assumption of the constant relation, the segment boundary marks the sudden change point [30]:

$$\mathrm{y}=\left\{\begin{array}{l}{k}_{1}x+m_1,t\le t_1\\ k_{2}x+m_2,t_1<t_{\le }{t}_2\\ \dots \dots \\ k_{n}x+m_n,t_{n-1}<t\le t_n\end{array}\right.$$

(1)

In this study, let x and y denote the independent and dependent variables, respectively. The rate of change in the linear segment is indicated by k₁, k₂, …, and k_n, whereas m₁, m₂, …, and m_n denote the intercept of the linear segment. The mutation points, t₁, t₂, …, and t_n, are determined by the sum of the squares of the minimum residuals of the regression lines, and the two regression lines fitted before and after the turning point have corresponding p-values less than 0.05. In a regression model, R² represents the proportion of the degree of variation in the dependent variable that can be explained by the independent variable. It is the square of the correlation coefficient (R), reflecting the degree to which the model fits the data. p-values quantify how likely it is to obtain an observed sample result that is as or more divergent than actual data, assuming the null hypothesis (typically no association between predictor and outcome variables) holds.

2.3.2. Mann–Kendall Trend Test

For time series X, the statistics of the Mann–Kendall trend test are as follows:

$$S={\displaystyle \sum _{i=1}^{n-1}{\displaystyle \sum _{j=i+1}^{n}sgn(x_j-x_i)}}$$

(2)

where x_j is the jth data value of the time series; n is the length of the data sample; and sgn is the sign function, and its definition is as follows:

$$sgn(\theta )=\left\{\begin{array}{ll}1& (\theta >0)\\ 0& (\theta =0)\\ -1& (\theta <0)\end{array}\right.$$

(3)

Mann and Kendall proved that when n ≥ 8, the statistic S roughly follows a normal distribution, with a mean of 0 and a variance of

$$Var(S)={\displaystyle \frac{n(n-1)(2n+5)-{\displaystyle \sum _{i=1}^n{t}_i(i-1)(2i+5)}}{18}}$$

(4)

where t_i is the number of data points in group i.

The standardized statistic Zc is calculated according to the following formula:

$$Zc=\left\{\begin{array}{ll}{\displaystyle \frac{S-1}{\sqrt{Var(S)}}}& (S>0)\\ 0& (S=0)\\ {\displaystyle \frac{S+1}{\sqrt{Var(S)}}}& (S<0)\end{array}\right.$$

(5)

That is, Zc follows a standard normal distribution.

The measure of the magnitude of the trend is

$$\beta =Median\left({\displaystyle \frac{x_i-x_j}{i-j}}\right)$$

(6)

where 1 < j < i < n, the median represents finding the median of the sequence. A positive β-value indicates an “upward trend”, while a negative β-value indicates a “downward trend”.

The method of the Mann–Kendall trend test is as follows: the null hypothesis is

H0: 

β = 0. When |Zc| > Z(1 − α)/2, the null hypothesis is rejected. Here, Z(1 − α)/2 represents the standard normal variance, and α is the significance test level.

2.3.3. Analysis of Partial Least Squares Path Model (PLS-PM) Construction

This study analyzes factors affecting oasis area in Tarim and Heihe Basins using partial least squares path modeling (PLS-PM). PLS-PM was first proposed by the Swedish statistician Herman Wold in 1975 for solving the complex relationship between latent variables and manifest variables in Structural Equation Modeling (SEM) [31]. Building upon this foundation, Zhao employed the PLS-PM model to assess the synergistic impacts of various factors on the variations in phytoplankton communities within the Han River’s middle and lower reaches. In their study, Zhao et al. utilized environmental and spatial factors as explanatory variables, thereby expanding the application of PLS-PM in environmental research [32]. In practice, the plspm package (version 0.5.1) within the R statistical (version 4.4.2) environment is employed to construct a partial least squares path model for data analysis. In the constructed PLS-PM model, the observed variables (temperature, precipitation, settlement, and war frequency) were categorized into the latent variables representing environmental and anthropogenic influences. The oasis extent was employed as the response variable. The internal influences of natural and human factors, as well as their impacts on oasis area, were investigated, respectively.

3. Results

3.1. Dynamics of Watershed Oases and Settlements over the Past 2000 Years

The Heihe River Basin has undergone several key phases of development over the past two millennia. These developments are evident in the region’s archeological investigations, historical documents, and site distributions. The basin has experienced the evolution of settlements, natural oases, and a cultivated oasis. The following is a timeline of key phases (Figure 2): the reign of 1 A.D.—early 3rd century A.D., the late 6th to early 10th centuries AD, and the mid-17th to early 20th centuries. These periods were characterized by significant unification and prosperity in Chinese history, during which the Central Plains’ dynastic governance over the Heihe and Tarim watersheds underwent substantial expansion. Concurrently, the agricultural population and area of cultivated oases in the Heihe and Tarim Basin underwent substantial expansion, accompanied by the emergence of numerous new settlements.

In contrast, settlement construction in the Heihe River Basin underwent a period of decline between the early 3rd and early 5th centuries A.D. and between the 10th and mid-15th centuries A.D. Similarly, settlement construction in the Tarim River Basin experienced a decline between the early 3rd and early 5th centuries A.D. and between the 10th and mid-17th centuries A.D. This period was characterized by a significant decline in population, with a notable exodus of the Central Plains farming peoples from the Heihe and Tarim River Valleys. This exodus, often occurring before the formal division of the dynasty, resulted in the abandonment of numerous settlements and cultivated oasis areas. Concurrently, the geographical expanse of natural oases within the Heihe Valley underwent a decline, persisting from the 8th to 16th centuries, when it underwent a slight stabilization. However, following this period, with the transition to a new dynasty, a renewed decline in the area of natural oases became evident. The natural oasis in the Tarim River Basin exhibited an initial growth followed by a decline between the 1st and 10th centuries AD., after which it demonstrated a decreasing trend. However, it subsequently demonstrated a rising growth trend after the 17th century. In contrast, the Heihe Basin natural oases exhibited a persistent pattern of diminution, and the Tarim River Basin oases demonstrated an oscillatory trend, with an initial growth, a subsequent decline, and renewed growth.

Employing the segmented linear regression analysis, we divided the oasis evolution across the Heihe and Tarim River Basins into three stages (Figure 3). During the first stage, settlement numbers rose from 143 in the first year AD to a peak of 246 during the Han Dynasty and then decreased to 36 during the Sui Dynasty in the seventh century. Concurrently, the cultivated oasis area decreased from 1741 km² to 600 km², while the natural oasis area decreased from 6342 km² to 6014 km². Overall, the settlement number, cultivation area, and natural oases all manifested a declining trend chronologically. Settlement totals displayed a significant positive correlation with the cultivation area, as well as a positive correlation with the area of natural oases. In addition, a positive correlation was also shown between natural oases and the cultivated oasis area. In the first stage of oasis evolution in the Tarim River Basin, the number of settlements increased from 70 in 206 BC to 95 in the third century, and the cultivated oasis area increased from 7.21 × 10³ km² to 19.22 × 10³ km². The natural oasis area increased from 49.11 × 10³ km² to 64.64 × 10³ km². The number of settlements, cultivated oases, and the natural oasis area increased over time. The settlement number was positively correlated with the cultivated oasis area and also with the natural oasis area. The areas of natural oases and cultivated oases were positively correlated. Between the mid-third century and the late seventh century, the settlement number declined from 99 to 75, and the cultivated oasis area shrank (19.23 × 10³ km² to 6.25 × 10³ km²). The natural oasis area contracted (70.61 × 10³ km² to 53.57 × 10³ km²). The settlement number and the cultivated oasis and natural oasis areas decreased with time. A significant positive relationship existed between settlements and the cultivation area, as well as with the natural oasis area. There was a positive correlation between natural oases and the cultivated oasis area.

The second period (7th to 16th century AD) is delineated on the basis of the segmented linear regression analysis. During this period, the Heihe River settlements increased from 36 in the 7th century to 87 in the 16th century. The cultivated oasis area increased from 595 km² to 821 km², while the natural oasis area decreased from 6016 km² to 5950 km². The area of cultivated oases decreased first and then increased during this period, and the natural oasis area was relatively stable. The amount of settlements exhibited significant positive associations with both cultivated and natural oases. A positive correlation was also shown between the natural oasis area and the cultivated oasis area. In the second phase of the Tarim River Basin (late 7th century to 1850s), the settlements increased from 75 to 237 during the late 7th to 10th centuries. The cultivated oasis area increased from 6.26 × 10³ km² to 19.21 × 10³ km², and the natural oasis area increased from 53.57 × 10³ km² to 64.75 × 10³ km². The settlement number exhibited a positive relationship with the cultivated and natural oasis area. The area of natural and cultivated oases was positively correlated. During the 10th century to the 1850s, the number of settlements dropped from 237 to 41. Cultivated oases dropped from 19.2 × 10³ km² to 16.79 × 10³ km². The natural oasis area decreased from 64.75 × 10³ km² in the 10th century to 13.2 × 10³ km² in the mid-17th century. The number of settlements and the cultivated and natural oasis area varied with time, and the overall tendency was a decreasing trend. The number of settlements and the cultivated oasis area were positively correlated, as was the natural oasis area. The natural and cultivated oasis areas were positively correlated.

In the third stage, the population of the Heihe Basin grew from 74,000 in the 16th century to more than 1.92 million at the beginning of the 21st century. At the same time, the cultivated area of oases increased from 821 km² to 5544 km², while the natural oasis area decreased from 5950 km² to 1777 km². The cultivated oasis area is significantly positively correlated with the amount of settlements, while there was a negative correlation between cultivated oases and natural oases. In addition, the number of settlements was also negatively correlated with natural oases. In the third stage (1850s to the beginning of the 21st century), although the settlement number decreased, the population in the Tarim River Basin increased from more than 1.5 million to more than 7.48 million. The area of cultivated oasis increased from 16.79 × 10³ km² to 53.12 × 10³ km², and the area of natural oases increased from 21.29 × 10³ km² to 63.25 × 10³ km².

3.2. Analysis of Drivers at Different Stages

The driving factors of historical periods are sorted by time (Figure 4), and the driving factors of different stages are analyzed. In phase I (Yuan-Yi to seventh century AD), except for the period of the Wei, Jin, and North and South Dynasties, the temperature of the Heihe River Basin was mostly in a warmer phase. At this time, the temperature was higher. The mountainous areas of the Heihe River Basin had more ice and snow melt water, while in the Wei, Jin, and North and South Dynasties, the temperature dropped, and the natural oasis area shrunk. During the same period, the Tarim River Basin was also in a warmer climate, with higher temperatures, which led the meltwater from the ice and snow to increase, and the water of the Tarim, Peacock, Hetian Rivers, etc., increased. The natural oasis area was increasing during this period, and it was in a stage of the restoration of the natural oasis area. Following the unification of the Western Regions through warfare during the Han Dynasty, the country entered a phase of pronounced prosperity and ascendancy. The establishment of capitals in these regions played a crucial role between the mainland and the Western Regions in facilitating trade and cultural communications. The Han Dynasty established numerous military institutions across the Heihe Basin to enhance oasis development, leading to a substantial expansion of the area dedicated to manmade oases. Concurrently, the Tarim River Basin witnessed intensified military defense measures. The favorable climate and the robust central government’s jurisdiction contributed to the development and utilization of artificial oases, which expanded at a rate greater than that of natural oases. In the latter part of this period, numerous wars transpired, resulting in the abandonment of numerous cities and ponds. In the Wei, Jin, and North–South Dynasties, the Heihe River Basin experienced cold, humid climatic conditions, with declining temperatures and precipitation levels reaching approximately 500 mm. Despite an increase in precipitation, the Qilian Mountain ice melt water was also lower compared to the warm period, leading to a downward trend in the Heihe Basin water flow. Its cultivated oasis was reduced, especially in the downstream area. Concurrently, the Tarim Basin experienced a transition to a cold, arid climate, accompanied by precipitation and temperature drops. Snow and ice accumulation in the mountainous regions reached unprecedented levels, exceeding 350 mm. The Tarim River and Lop Nur’s main stream volume declined and resulted in the recurrent migration of river channels, leading to the disruption of the connection between the Peacock River and the northern mainstem of the Tarim River. This disruption resulted in a substantial decrease in the flow into Lop Nor. Concurrently, desertification escalated dramatically and with an increased frequency, leading to a substantial retreat of the natural oasis area. The Central Plains experienced persistent turmoil and chronic division, resulting in weakened governance over the Western Regions. The frequent warfare and social instability caused by the Heihe River Basin multi-regime struggle, frequent regime change, and war led to significant population depletion and displacement. This triggered demographic contraction, an agricultural standstill, and the desertion of numerous settlements and anthropogenic oases.

The second stage, which occurred from the 7th to the 16th century, was characterized by temperature fluctuations in the Heihe Basin. During the 10th century, the region experienced a period of lower temperatures, while the final years of the eleventh century and the sixteenth century marked the lowest recorded temperatures. With the exception of the late Tang Dynasty, which experienced a surge in precipitation, the remainder of the period under consideration was relatively stable. In the early Sui and late Tang Dynasties, numerous wars and conflicts occurred, and during the early Sui period, agricultural development was not prioritized in the western region of the river. It was only during the Tang Dynasty that the policy of cantonization was implemented. The Tarim Basin region underwent a transition in its climate, becoming warmer and more humid. This shift resulted in frequent flooding, triggered by higher temperatures and increased snow and ice melt. Consequently, river channels underwent changes in their course. The natural oasis area expanded during this period. The Tang Dynasty initiated the garrisoning and fortification of the Heihe River Basin, establishing sectional commissars along the border to defend the north against the threat of the Turks and Tubo. The An-Shi Rebellion, which erupted within the Tang Dynasty, prompted the Heshi army to embark on a campaign into the interior of the country to subdue the uprising. This event allowed the Tubo to seize the Hexi Corridor and hold it for close to a century. This period marked a reversal of the oasis settlement, which was not fully restored until the nineth century. The ongoing wars of this era led to population decline, urban and rural abandonment, and a significant contraction of the oasis’s territory. During this interval, the Tang Dynasty in the Tarim Basin successfully reclaimed the Western Region through the implementation of administrative structures and garrison armies. A substantial number of settlements and artificial oases were established during this period. However, the Tang Dynasty faced significant challenges from Arabia and other regions in the nineth century, leading to a gradual decline in its control over the Western Region, which was subsequently divided and fragmented once more. The 10th to 16th centuries, specifically the 11th, 13th, and 16th centuries, were characterized by lower temperatures. Concurrently, precipitation levels were also at their lowest. The climate during this period was marked by a dry and cold spell. The presence of low temperatures and minimal precipitation contributed to the emergence of this arid climate. Precipitation and low temperatures led to a reduction in the meltwater from ice and snow, and at this time, the Heihe River Basin was under the jurisdiction of the Xixia Dynasty. The local ethnic groups primarily engaged in livestock production, while the agricultural output was primarily allocated to the military. The oasis underwent a restoration, though its development remained constrained. In the 15th century, precipitation levels increased, resulting in a transition to a cold and humid climate. During the Yuan and Ming Dynasties, the Mongol Yuan occupied the Hexi region and implemented a policy of massacres, leading to a significant decline in the population and the abandonment of towns and farmland. During the Ming Dynasty, the Ming government established guard posts and military cantonments to consolidate its rule over the Heihe River Basin. Then the oasis was restored and developed. The 10th century marked the onset of a 500-year period of alternating dry and cold periods in the Tarim River Basin, followed by the Little Ice Age in the 15th century. This period is characterized by the contraction of the Tarim River’s main streams and tributaries, accompanied by the expansion of the desert, driving the population to migrate to safe areas and causing the original ecological settlements to fall into a state of abandonment. Concurrently, in the Yuan and Ming Dynasties, the region experienced recurrent military conflicts and governance fragmentation, leading to profound socio-economic and cultural devastation. Protracted religious factional conflicts between the Kashgargar Kharakhan dynasty and the Ü-Tsang and Western Zhou Hui migratory birds marked the end of the Buddhist domination in the Western Regions, resulting in the destruction of its prosperity and flourishment [33].

The third stage of the Heihe Basin’s evolution, spanning from the 16th to the 21st century, is characterized by a decline in the temperature and precipitation. During this period, the climate remained arid and frigid. After this initial phase, temperature variations became more erratic, while precipitation exhibited fluctuations and underwent changes relative to the preceding period, resulting in a more stable state. Noteworthy peaks in precipitation occurred during the Ming, Qing, and the Republic of China eras. The Little Ice Age culminated in the Tarim River Basin during the 17th century, then persisted for about a century thereafter [22,34]. In the 19th century, the Tarim Basin river system’s structural evolution demonstrated a “four-source mainstream” pattern that persists to the present day. Prior to the 19th century, environmental changes in the Heihe and the Tarim Basins were primarily influenced by natural factors. The transfer of oases can be regarded as being dominated by environmental climate change. However, following the 19th century, the population of the Heihe and Tarim River Basins experienced a substantial increase, as did the area dedicated to cultivated oases.

Conduct a trend analysis of the driving factors of the two river basins (

Table 3. Mann–Kendall test for driving factor trends in Heihe and Tarim River Basins.

Variable	Period	β	Zc	Significance
Temperature	0–600s AD	−0.02301196	−7.3819	Significant
	600s AD–1500s AD	−0.005018685	−2.8142	Significant
	1500s AD–2000s AD	0.05456576	6.3243	Significant
Population	0–600s AD	−0.5454545	−1.0445	Insignificant
	600s AD–1500s AD	0.3461538	1.019	Insignificant
	1500s AD–2000s AD	0.5454545	1.0445	Insignificant
War frequency	0–600s AD	0.07697947	1.6414	Insignificant
	600s AD–1500s AD	0.05206379	3.6721	Significant
	1500s AD–2000s AD	−0.02430556	−0.41248	Insignificant
Precipitation	0–600s AD	0.00212024	1.1512	Insignificant
	600s AD–1500s AD	−0.001959552	−1.1812	Insignificant
	1500s AD–2000s AD	0.001220903	0.49167	Insignificant
Mann–Kendall test for driving factor trends in Tarim River Basin
Variable	Period	β	Zc	Significance
Temperature	200s BC–350s AD	0.003949863	4.3392	Significant
	350s AD–700s AD	0.006108254	7.5459	Significant
	700s AD–900s AD	0.01403525	3.1277	Significant
	900s AD–1850s AD	−0.003248803	−7.3497	Significant
	1850s AD–2000s AD	0.03609626	3.5595	Significant
Population	1760s AD–2000s AD	0.4511278	3.0038	Significant
War frequency	200s BC–350s AD	−0.03386514	−1.5937	Insignificant
	350s AD–700s AD	−0.00995174	−0.33086	Insignificant
	700s AD–900s AD	−0.1269231	−1.1816	Insignificant
	900s AD–1850s AD	−0.03028464	−2.3542	Significant
	1850s AD–2000s AD	0.375	1.6202	Insignificant
Precipitation	1450s AD–1850s AD	0.002270788	1.5969	Insignificant
	1850s AD–2000s AD	0.01847167	2.9683	Significant

Note: The significance level is α = 0.05.

), a notable increase in temperature was observed in both basins during the period spanning from 1500 to 2000. The Z value of the Heihe Basin was 6.3243, whereas the Z values of the Tarim Basin were 7.5459 (350s AD–700s AD), 3.1277 (700s AD–900s AD), and 0.03609626 (1850s AD–2000s AD). These findings suggest a generalized warming of the global climate during this period. The Heihe Basin exhibited a discernible downward trend in the 0–600s and 600–1500s stages, while the Tarim Basin manifested an upward trajectory spanning from 200s BC to 900s AD. The Tarim River Basin exhibited a substantial population growth trend during the 1760s AD–2000s AD period, while the Heihe Basin did not demonstrate a significant trend in the population change. The observed population growth may be associated with improved environmental conditions, technological advances in agriculture, and enhanced living conditions, with these factors potentially being more significant in the Tarim River Basin. The Heihe Basin experienced an increase in the frequency of wars from the 6th to the 15th centuries, while the Tarim River Basin did not exhibit a significant trend in war frequency for the majority of the period from the 200s BCE to the 2000s CE, with a notable decrease in war frequency only from the 9th to the 18th centuries. Both basins exhibited an upward trend in precipitation during the 1850s to 2000s AD period. The Heihe River Basin demonstrated a non-significant increase, while the Tarim River Basin exhibited a significant increase, with a Z value of 2.9683. This increase may be related to global climate change and alterations in regional climate patterns.

3.3. Partial Least Squares Path Model (PLS-PM) Construction Analysis

This study was conducted to analyze the factors influencing the oasis area in the Tarim River and Heihe River Basins. Partial least squares path modeling (PLS-PM) was employed to achieve this objective. The analysis entailed the quantification of the factor loadings of the observed variables (temperature, precipitation, settlement, war frequency) on the latent variables (natural and human factors). It also entailed the quantification of the path coefficients of the latent variables on the oasis area. This approach enabled the revelation of the strength and direction of the influence of each factor on the oasis area in the two basins. The analysis revealed that both natural and human factors exert significant influence on the two watersheds, though the extent and direction of their impact vary.

As demonstrated in Figure 5, the numerical values represent the factor loadings (loadings) of the observed variables (temperature, precipitation, settlement, war frequency) on the latent variables (natural factors, human factors). Factor loadings are pivotal parameters in PLS-PM, as they signify the magnitude of the relationship between observed variables and latent variables. As demonstrated in the figure displaying the observed variables versus latent variables for natural oases in the Heihe River Basin, the temperature (−0.99) has a robust negative impact on natural factors. Conversely, precipitation (0.22) has been observed to have a positive effect on natural factors. Settlement (−0.93) exhibits a negative effect on human factors, while the war frequency has a positive impact on human factors (latent variables). The Heihe River Basin’s natural oasis area is negatively impacted by natural factors (−0.11), while human factors (0.95) exert a positive influence. The temperature (0.34) had a positive effect on natural factors in the Tarim River Basin’s natural oasis area, while precipitation (−0.99) had a negative effect on natural factors. Among human factors, the settlement (0.85) and frequency of wars (0.55) had a positive effect. Conversely, natural factors exhibited a negative influence on the Tarim River Basin’s natural oasis area, while human factors demonstrated a positive effect within this same area.

The Heihe River Basin’s cultivated oasis has been identified as an important influence on the region’s environmental dynamics. Specifically, the temperature (−0.99) has been found to exert a substantial negative impact on natural factors, while precipitation (0.18) has been observed to exhibit a positive effect on these factors. Settlement (−0.92) exerts a robust negative influence on human factors, while the war frequency (0.74) demonstrates a pronounced positive impact on human factors. The analysis further reveals that natural factors (0.06) exert a less positive influence on changes in the Heihe River Basin’s cultivated oasis area, while human factors (−0.93) demonstrate a strong negative impact. In the Tarim River Basin, the temperature (0.003) exhibited a modest positive influence on the natural factors, while precipitation (0.98) demonstrated a pronounced positive impact. Conversely, among the human factors, settlement (0.06) exhibited a minor positive effect, while the war frequency (−0.99) demonstrated a strong negative effect. Collectively, natural and human factors drive the expansion of the Tarim River Basin cultivation area, with natural drivers predominating.

4. Discussion

4.1. The Phased Differences in the Evolution of the Tarim and the Heihe River Basin Oases

This study explores the historical and environmental factors that have shaped the evolution of the Tarim and the Heihe River Basin oases over two millennia. The analysis reveals that climate change and war frequency have been the predominant forces influencing these changes. The Tarim River Basin’s oasis evolution is marked by high evolutionary volatility, cyclical growth and decline, and a dynamic interplay between environmental and human factors. The Tarim River Basin’s natural oasis area exhibited a fluctuating pattern, characterized by an initial rise, a subsequent decline, and an eventual resurgence. The area showed a slight increase in the first stage, a notable decline during the second stage, and a sharp increase during the third stage. After the end of the Little Ice Age during third stage, coupled with policy support and protection for human activities, the natural oasis area increased sharply [35]. The cultivated oasis area decreased slightly in the first stage but increased significantly in the second and third stages. The Heihe River Basin’s natural oasis area cumulatively decreased by 72.0% over the past 2000 years and was in a continuous decline mode. The degradation accelerated in the later period from the 16th to the 21st century, during which the area of natural oases decreased by nearly 70.1%. The transformation of the cultivated oasis’s reclamation, from the first stage of shrinkage to a partial recovery, the second stage of recovery and increases, and finally to explosive growth, is similar to a U shape. The Heihe River Basin oases exhibited discrepancies in their recovery cycle, diverging from the trajectory experienced in the Tarim River Basin, particularly during the decline phase under the Central Plains dynastic governance. This deviation in the recovery cycle has led to the progressive contraction of the Heihe Basin oasis area, with some regions undergoing prolonged periods of no recovery. The Tarim River Basin oases exhibit a “recovery cycle” because they form an inward-oriented, largely self-contained social–ecological system. Situated deep within Asia, the Tarim Basin lies far from the political heartlands of the Central Plains. When external political pressures subside, the system retains both the spatial latitude and the ecological potential to restore and reorganize itself internally [22]. In contrast, the Heihe River Basin oases show no recovery cycle during periods of dynastic decline; instead, they continue to contract. As an outward-oriented national strategic corridor, their very survival and prosperity hinge on the stability of the central state [28]. Once the central regime collapsed, the military shield, economic lifeline, and water infrastructure network along the Heihe corridor unraveled in a systemic, cascading fashion. Because these large, state-built hydraulic works required continuous centralized maintenance, their failure was abrupt and irreversible. Once this highly integrated water lifeline was cut, the entire oasis agro-system ground to a halt, and the fragmented local communities lacked both the technical capacity and the social organization to repair it on their own [36].

4.2. The Differences in Influencing Factors of Oasis Evolution Between the Tarim and the Heihe River Basins

For the changes in the natural oasis area, the natural factors in both the Heihe and the Tarim River Basins have had a negative impact, while human factors have played a positive role and have had a greater influence. Regarding the changes in the cultivated oasis area, the natural factors in both the Heihe and Tarim River Basin are positive. However, in terms of human factors, the human factors in the Heihe River Basin have a negative impact, while those in the Tarim River Basin exert favorable impacts. The difference lies in the human factors—the negative proportion of the settlement in the Heihe River Basin is larger, while the negative proportion of the war frequency in the Tarim Basin is larger.

In terms of the driving factors under consideration, the changes in the Heihe Basin oasis area were found to be more influenced by human factors. By contrast, the Tarim River Basin was found to be influenced by both natural and human factors. A comparative analysis reveals that the predominant drivers of the Heihe Basin oasis evolution are human factors (0.95 and −0.93). While in the Tarim River Basin, the cultivated oasis is more affected by natural factors (0.63). The Tarim Basin is located in one of the most arid areas. Its water resources are heavily dependent on the alpine snow and ice melt, as well as precipitation recharge. The degree of drought and the uneven distribution of water resources have greatly affected the formation and oases evolution. The restrictions imposed by climate conditions on oases in the Tarim Basin manifest chiefly in periods of extreme drought. Water scarcity curtails the expansion and stabilization of oases, and natural oasis evolution closely correlates to water resources [37]. In periods of declining precipitation or uncertain seasonal water sources, fluctuations in the climate, particularly in precipitation patterns, can lead to changes in desert and oasis areas throughout history. These changes can result in the degradation of oasis areas [38]. The oases of the Tarim Basin have demonstrated significant fluctuations throughout history, particularly in the context of climatic anomalies or prolonged droughts, and the changes in the oasis area have exhibited clear cyclical patterns.

In contrast to the Tarim Basin, the Heihe River Basin receives precipitation sourced from the Qilian Mountains [39]. The upstream areas of this basin are characterized by relatively abundant precipitation and relatively more stable water resources. Consequently, climate change is predicted to have a lesser impact on oasis evolution in the Heihe River Basin, particularly in the upstream areas with higher precipitation, which may alleviate the downstream water resource’s tension problem to a certain extent [40,41]. Consequently, the Heihe River Basin oases are subject to greater anthropogenic drivers, with both the settlements and war frequency among the anthropogenic influences playing an important role. The correlation between the number of settlements and the population is such that an increase in the number of settlements leads to an increased demand for food and other foodstuffs. This expands cultivation and increases the cropland area, which is typically located near settlements. The outbreak of war has been shown to lead to a decrease in the population size, with residents fleeing and moving away from the war. Conversely, periods of war have been observed to result in an increase in the number of abandoned settlements, leading to the abandonment of once-living cultivated oases and a subsequent decrease in the cultivated oasis area. This finding is consistent with previous results that political factors can be viewed as central human drivers of the evolution of oasis distribution patterns [27]. Furthermore, the population density and changes in cultivated oases are significantly correlated [42]. In relation to the changes in oasis areas within the Heihe River Basin, the natural oasis is undergoing a transition towards a cultivated oasis. The impacts of human activities within the Heihe River Basin can, in certain historical periods, exceed the impacts of natural factors on the oasis [43,44]. Consequently, the optimization of resource allocation policies is imperative to ensure the harmonious interplay between natural and human factors, thereby fostering the balanced development of both natural and cultivated oases and propelling the sustainable advancement of the region [45].

Our research, which focused on the historical oasis evolution in Tarim and Heihe River Basins over the past two millennia (206 BC–2000 AD), revealed that some of the measured data were incomplete, with some collected and counted exclusively based on dynasties. In subsequent research, carrying out the excavation and verification of historical materials should be pursued to improve their usability. A particular focus should be placed on the exploration of historical documents. Such as local chronicles, travelogs, and ancient historical records. These materials may contain valuable information about ancient oases, the distribution of water systems, and various natural and human activities in history. Conversely, samples with environmental indicators, such as soil and sediment, can be analyzed to explore ancient environmental changes during the historical period. This analysis can verify and supplement historical documents, thereby enhancing our understanding of the past. To this end, interdisciplinary collaboration is imperative, as the changes in the historical environment are complex and multifactorial. Future research must therefore combine the research methods and results of many disciplines, including archeology, history, ecology and sociology. In comparison with the historical period, the influence of human activities in the later period is becoming increasingly significant, and it is the primary driving factor. To achieve a deeper and more precise comprehension of the past environmental transformations in the Tarim and Heihe River Basins, there is an imperative to augment research on human activities.

5. Conclusions

This paper utilizes a comparative study of oasis dynamics and their drivers across the Tarim and Heihe River Basins in the last 2000 years (206 BC to 2000 AD) This research reveals significant spatial and temporal differences and phases characterizing the oasis evolution in both basins. The Tarim River Basin natural oases show the characteristics of a “fluctuating recovery”. There were significant changes in three stages: an increase from the 1st to the 7th century, a decrease from the 7th to the 16th century, and a sharp increase after the 16th century. Its cultivated oases expand in a stepwise manner. The natural oases in the Heihe Basin have been in a continuous decline, with a cumulative reduction of 72.0% over 2000 years, especially accelerating their deterioration after the 16th century. The cultivation of oases shows a “U-shaped” evolution: early shrinkage, a partial recovery in the middle stage, and explosive growth in the later stage. It reflects the strong intervention of human reclamation in the pattern of oases. The introverted society–ecosystem endows the Tarim River Basin with resilience. After the external pressure weakens, it achieves a periodic recovery through self-repair. The Heihe River Basin is a strategic corridor for an outward-oriented country and relies on the central government for stability. The collapse of the regime led to a chain of disintegrated water conservancy systems, thus lacking a recovery period This pattern highlights human activities’ profound influence on oasis evolution. Climate fluctuations predominantly shape the oasis evolution within the Tarim Basin. The path coefficient of natural factors on the cultivated oasis in the PLS-PM model reaches 0.63, with the temperature and precipitation directly affecting oasis changes and extreme drought leading to the drastic retreat of the natural oasis. In contrast, in the Heihe River Basin, human factors play a significant role. The path coefficient of the area of cultivated oases and human factors reaches as high as −0.93. The frequency of wars (loaded with 0.74) and the increase/decrease in settlements (loaded with −0.92) also play a dominant role in oasis evolution. The frequency of war (loading 0.74) and the increase/decrease in settlements (loading −0.92) play a dominant role. Future research should focus on human activities, integrating multi-disciplinary research methods and achievements to elucidate the complex mechanisms underpinning oasis evolution.