Identifying Migration Routes of Wild Asian Elephants in China Based on Ecological Networks Constructed by Circuit Theory Model

Simple Summary

In March 2020, a group of wild Asian elephants suddenly roamed north from China’s **shuangbanna National Nature Reserve and subsequently arrived in Kunming City, Yunnan Province, which has aroused widespread concern in the society. Based on this event, this study attempted to identify the possible migration routes of Asian elephants according to the existing land conditions and the movement habits of Asian elephants if wild Asian elephants in China migrated again. It was found that the areas with a high probability of migration by Asian elephants were mainly concentrated in the gently undulating shrubs, bamboo forests, and grasslands. Moreover, dense forests with steep slopes and high altitudes, cultivated land, and construction land are less likely to be passed by Asian elephants. Our findings can not only provide valuable insights into suggestions and solutions for the current conservation of wild Asian elephant species but also be beneficial in biological protection and biological reserve planning.

Abstract

Humans overlap with Asian elephants, resulting in frequent costly human–elephant conflicts, which disturb and even threaten local residents. In this study, we treat provincial and national nature reserves where Asian elephants still exist and other alternative habitats suitable for Asian elephants in southern Yunnan, China, as ecological patches. By using this approach, we can treat the terrain and surface state factors that hinder the migration of Asian elephants as a form of ecological resistance surface. We can then use a circuit theory model and remote sensing data to construct an ecological network, which allows us to identify ecological corridors and ecological pinch points. Herein, the possible migration routes of wild Asian elephants were identified. The main results are as follows: (1) In the study area, dense forests with steep slopes and high altitudes, cultivated land, and building land have greater migration resistance, while the gently undulating shrubs, bamboo forests, and grasslands far away from the city have less migration resistance. (2) There are three ecological corridor groups in the study area, mainly composed of shrub and grassland. The ecological corridors identified in this paper are the most likely migration routes of wild Asian elephants in China, and areas with higher simulated current densities reflect a higher probability of Asian elephants passing through. (3) According to the analysis, the ecological pinch points in the study area are 602 km² in total, and woodland and grassland account for 89.2% of the total ecological pinch area. The areas where the pinch points are located have a high probability of Asian elephants passing through and a narrow space. Our findings can provide suggestions and solutions for the current conservation of wild Asian elephant species, alleviate human–elephant conflicts, promote the harmonious coexistence between humans and nature, and provide reference for biological protection and biological reserve planning.

1. Introduction

The Asian elephant (Elephas maximus) is a first-class protected animal in China, which can be found in most areas south of the Yellow River Basin, and is listed as an endangered species by the International Union for Conservation of Nature (IUCN). The species is threatened by climate change and various human activities (e.g., deforestation), and consequently, the distribution of Asian elephants in China has retreated southward at a rate of 0.5° latitude per century [1]. Asian elephants are migratory mega herbivores requiring a large space to live—space that has diminished considerably in recent years [2]. The acceleration of urbanization and industrialization has rapidly eaten away at the available habitat, leaving the Asian elephant on the brink of extinction [3,4]. At present, there are only about 300 wild Asian elephants left in China, mainly distributed in the southern part of Yunnan [5]. Habitat loss and fragmentation for Asian elephants are mainly due to massive deforestation [6]. Human production and living spaces have replaced forests in many areas [7]. Traditional migration routes of Asian elephants are now cut off, and human settlements and Asian elephant ranges have begun to gradually overlap over the last two centuries [8]. Understandably, human–elephant conflicts occur frequently, and there were 27 deaths and many injuries from 2011 to 2019 [9]. Human–elephant contact is not only expensive but also interferes with and threatens the daily activities and safety of local residents. Accidents related to Asian elephants have become a serious local problem. Therefore, effective measures to alleviate human–elephant conflicts should strengthen the quality restoration of Asian elephant habitats and foraging areas on the basis of protecting the rights and interests of rural settlements and urban built-up areas. However, the restoration of Asian elephant habitat quality is usually measured on a 10-year scale [10]. Designing available ecological corridors for Asian elephants is critical to guide Asian elephants to suitable areas in case of reoccurrence of migration. Predicting the movement route of the elephant herd is the basis for promoting the protection of Asian elephants and alleviating human–elephant conflicts. However, current studies have focused on the assessment of the habitat quality of Asian elephants and the factors affecting habitat quality or the society’s attitude and awareness of the human–elephant conflict [1,11]. The limited understanding of the prediction of migration routes has hindered the development of mitigating human–elephant conflicts and protecting wild Asian elephants.

Circuit theory, derived from physics, was applied by Mcrae and Beier to study the gene flow in heterogeneous landscapes, arguing that circuit theory can be used to predict the migration path of biological populations and the probability of successful dispersal [12]. The charge in the circuit has the characteristics of a random walk, and the spillover of species is also highly random. The circuit theory model connects circuit theory and motion ecology through random walk theory. At present, the circuit theory is often used in the research on ecological sustainable development and long-term construction and the study of ecological networks based on biodiversity [13,14]. Little research has been conducted leveraging this approach to identify species migration and movement ecology with an eye toward conservation in a human–wildlife context. The “ecological patch-ecological resistance surface-ecological corridor” research paradigm is often aimed at providing effective and systematic means of solving ecological problems in land space planning and the ecological restoration of land space and is committed to the application of the ecological network and ecological security pattern [15,16]. The general ecological significance of this paradigm is reflected in the research, and ecological network construction for specific species is considered slightly narrow and insufficient to date.

In March 2020, a group of wild Asian elephants roamed north from China’s **shuangbanna National Nature Reserve and arrived in southern Kunming in early June 2021. The event has attracted widespread international attention, largely because no such long-distance migration of Asian elephants has occurred in the past half-century [2]. Elephant migration is a complex ecological process that varies in distance, timing, duration, and drivers. At present, there are many controversies about the reasons for elephant migration in academia. This study focuses on the discussion on how to scientifically identify the migration routes of Asian elephants. Based on the existing situation and the premise of promoting the protection of Asian elephants and mitigating human–elephant conflicts, this paper uses a circuit theory approach and adopts the research paradigm of “ecological patch-ecological resistance surface-ecological corridor” to identify the migration routes of Asian elephants.

2. Materials and Methods

2.1. Study Area and Data Sources

The study area is the cities and autonomous prefectures where the main protection objects include Asian elephants or habitats suitable for the survival of Asian elephants in the provincial and national nature reserves in the “National List of Nature Reserves (2015)”. In our survey of these sites, we included Baoshan, Lincang, Pu’er, Dehong, **shuangbanna, Honghe, and Wenshan, a total of 3 cities and 4 autonomous prefectures (Figure 1). These locations are located within the south of Yunnan Province, China, between 21°05′~25°50′ north latitudes and 97°29′~106°14′ east longitudes, bordering Myanmar, Laos, and Vietnam, with a total area of 181,128 km² and an elevation of 72~3759 m. The study area is located in the low-latitude monsoon climate zone, which is the area most affected by the southwest monsoon in China. The study area is rich in habitat types and is considered a global biodiversity hotspot [17].

The data used in this study consisted mainly of land use, terrain, administrative boundary nature reserve, and Asian elephant historical early warning points. Among these data sources, land-use remote sensing monitoring data (30 m × 30 m), digital elevation model (DEM) data (90 m × 90 m), and administrative boundary vector data of the study area were obtained from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (https://www.resdc.cn/, accessed on 6 March 2023). The data for nature reserves were obtained from the website of the Central People’s Government of the People’s Republic of China (https://www.gov.cn/, accessed on 6 March 2023). The historical early warning points of Asian elephants were obtained from the Asian Elephant Early Warning System (http://elephant.noahsark.org.cn/#/, accessed on 1 August 2023).

2.2. Methodology

As described, we focused our research on identifying potential areas where we could connect the “ecological patch-ecological resistance surface-ecological corridor” paradigm. In this context, areas with Asian elephants are considered type I ecological patches, and protected zones with suitable habitats for Asian elephants are taken as type II ecological patches. By synthesizing the terrain and surface state factors that hinder the ecological flow of Asian elephants we could establish an ecological resistance index system. By then applying circuit theory, we can identify an ecological network, set thresholds to extract ecological corridors and ecological pinch points, and produce a model of potential migration pathways (Figure 2).

2.2.1. Identification of Ecological Patches (Step 1)

The facilitation of ecological functions and the preservation of ecological integrity can be enhanced by the presence of high-quality ecological patches [17]. For this purpose, ecological patches were designated as provincial and national nature reserves, as these areas predominantly harbor wild Asian elephants in southern China [5]. Based on previous studies [1,18], this study classifies nature reserves that primarily focus on Asian elephants as ecological patch I, while nature reserves with high-quality Asian elephant habitat evaluation are classified as ecological patch II. This study focuses on wild Asian elephants in China. Initially, migration begins from ecological patch I, and as biological flow and the spread of Asian elephants occur, both ecological patch I and ecological patch II become equally significant. The approach examined in this study aligns with the prevailing circumstances of China’s swift urbanization and the underlying purpose of establishing nature reserves.

2.2.2. Construction of Ecological Resistance Surface (Step 2)

The accessibility and connectivity between ecological patches can be judged by setting the minimum expansion resistance value, since species need to overcome certain resistance for spatial movement [19]. The setting of the resistance value of ecological expansion should consider various factors influencing the dispersal behavior of specific wild species [20]. However, there is currently no accepted standard for setting resistance values [21]. It is simplest to envision the “resistance” value as a holistic estimate of how difficult it is for a given species to move through a given corridor. Much as the electrical resistance of an object is a measure of its resistance to the flow of current, these resistance values are estimates of the difficulty associated with navigating an ecological corridor [22,23]. The construction of a generalized ecological corridor then might be considered a more comprehensive assessment that takes into account the average difficulty any species may have traversing the corridor [14,17]. This study comprehensively analyzes the preferred distribution landscape types of wild Asian elephants in China and the geographical characteristics of spillover and constructs comprehensive resistance factors by synthesizing terrain surface state factors. Resistance values are often estimated using field surveys of topography and terrain, as well as animal tracking data. Estimation of resistance values might also consider niche factor analysis, though we do not attempt that here [24,25,26,27] (

Table 1. Resistance factor and its assignment.

Resistance Factor Type	Factor of Resistance	Class of Classification	Value of Resistance
Terrain	Elevation (m)	<1000	10
		1000–1500	50
		1500–2000	100
		>2000	10,000
	Slope	<10	10
		10–15	50
		15–30	100
		>30	10,000
	Terrain roughness	<1.1	10
		1.1–1.2	50
		1.2–1.3	100
		>1.3	10,000
Surface state	Land-use type	Shrubbery, grassland, sparse woodland	10
		Forest, bare land, dry land	100
		Swamp, river beach, rural settlement	200
		Other woodland, paddy field, bare rock land	400
		Urban land, other construction land, river canal, lake, reservoir, permanent glaciers, snow	10,000

Note: Land-use type as defined by the Chinese National Land Use/Cover Classification System (CNLUCC) based on multi-period remote sensing data.

). Herein, “ecological resistance” is a product of four factors: elevation, slope, terrain roughness, and land-use type. Resistance values reflect the ease with which Asian elephants traverse different terrains and landscapes. The lower the resistance value, the more likely Asian elephants are to diffuse into the environment and the higher the accessibility and circulation of Asian elephant ecological flow. The higher the resistance value, the stronger the resistance to Asian elephant spillover migration. The terrain and landscape constitute rigid constraints on Asian elephant spillover; the potential range of the resistance value is regarded as infinite, and it is generally set to be large enough in the actual assignment process. To optimize computing efficiency, resistance values are capped at 1000 following [25,26,27].

Asian elephants are huge in size, with a female shoulder height of 2.24–2.54 m and a weight of 2720–4160 kg and a larger male shoulder height of 3.2 m and a weight of 5400 kg [28]. The size and physiological structure determine that it is difficult for Asian elephants to traverse steep and undulating areas. Consequently, this study considers both slope and terrain roughness resistance factors. Altitude also has a certain impact on the spillover migration of Asian elephants, but within the study area, altitude is not the dominant factor. Instead, migration may be impacted by habitat suitability. Wild Asian elephants in China prefer to live in bamboo forests, bamboo-broadleaved mixed forests, grasslands, or shrubs. They have strong adaptability to habitats with rich understory vegetation but weak adaptability to cultivated land and impermeable surfaces affected by human activities. Asian elephants often travel near water sources for food, swimming, and bathing. Asian elephants usually choose to cross shoals or small streams. In order to avoid being washed away or submerged by the current, Asian elephants generally do not choose to cross large rivers with rapid currents.

2.2.3. Identification of Ecological Network (Step 3)

(1)

Ecological network identification

Circuit theory connects circuit theory and motion ecology through random walk theory [29]. The model treats the landscape as a conductive surface and simulates the movement process of species as similar to the random flow of electrons. In this fashion, we can calculate the magnitude of the current between each ecological patch and consider all possible paths in the landscape [30]. The formula for calculating the current density between two nodes is

$$I=\frac{V}{R_{eff}}$$

(1)

In the formula, I is the current through the conductor, V is the voltage measured through the conductor, and R_eff is the effective resistance. The effective resistance R_eff can reflect the spatial separation between nodes. The voltage V represents the probability that a random walker leaves any node and successfully reaches a given target node (i.e., the probability of successful diffusion down a corridor). The branch current value I can reflect the ecological flow through the resistance section of the branch, which can be used to predict the movement probability of the ecological flow on the corresponding node or path.

(2)

Ecological corridor extraction

Ecological corridors are channels that ensure energy and material flow between important ecological patches. When the corridor has a certain width and habitat quality, it will ensure the realization of ecological processes and ecological functions [31]. Put simply, a corridor is an area through which organisms may reliably pass. In some areas, the current density is extremely low, and the possibility of Asian elephants passing through is extremely small. Conversely, if a connection is identified where current density in the ecological corridor area is higher, the connectivity between the patches is higher [22,23]. This study employs the natural break (Jenks) classification method to ensure the ecological corridor’s significant connectivity and high current density [32]. The current density is divided into ten categories, with the first two categories (0.003) serving as the threshold for extracting the ecological corridors. These corridors represent the identified migration routes of wild Asian elephants in China. The zonal statistics tool in the ArcMap10.6 toolbox is utilized to conduct extraction queries and obtain area statistics for each land-use type.

(3)

Ecological pinch point identification

Circuit theory also introduces a concept known as an “ecological pinch point”. When an area in the model has a high current density in the ecological network, the possibility of biological movement is high, and it has strong irreplaceability [33,34]. Combining the minimum resistance model and the circuit theory model was used to identify ecological pinch points [35]. However, when there is large resistance value in the surrounding area, biological movement is compressed in a relatively narrow range, resulting in a “pinch point” through which a large amount of traffic is funneled through a relatively small aperture in the model network [36]. In this paper, a 10 km buffer zone is set for ecological patch I. The natural breakpoint method is used to divide the current density into three levels, and the area in the buffer zone with the highest current density is considered an ecological pinch point.

3. Results

3.1. Ecological Patches

There are 28 ecological patches totaling 5612.18 km², accounting for 3.1% of the total surface of the study area. Most sources are distributed in the west of Dehong, the southwest of Lincang, central and southern Pu’er, ** land with higher altitudes and steep slopes have greater resistance to ecological expansion, while shrubs, bamboo forests, and grasslands that are far away from cities and have low absolute and relative elevations have small resistance to ecological expansion. We identified three ecological corridor groups in the study area, mainly composed of shrub and grassland, which conforms to the actual behavioral patterns of Asian elephants and exhibits a mesh-like distributional pattern. The ecological corridors identified in this study represent possible migration routes for wild Asian elephants in China. The areas with higher current densities reflect a higher probability of Asian elephants passing through. We also identified key ecological pinch points representing 602 km² of the total study area; woodland and grassland account for 89.2% of the total ecological pinch area. Pinch points represent narrow areas with a high probability of Asian elephants passing through. Residential areas near the ecological pinch point can be equipped with defensive measures to avoid human–elephant conflicts, and woodland and grassland can be protected to enhance the guidance of the migration route of Asian elephants. The migration routes and ecological pinch points of wild Asian elephants identified in this paper can provide methods and suggestions for the current conservation of the wild Asian elephant, alleviate human–elephant conflicts, promote the harmonious coexistence between humans and nature, and provide a basis for short-term biological protection.