Chinese Pangolin Changes Local Vertebrate Assemblages and Contributes to Their Interspecific Interactions by Burrowing and Revisitation

ABSTRACT

The burrow microhabitats created by burrowing mammals, as a hotspot for biodiversity distribution in ecosystems, provide multiple critical resources for many other sympatric species. However, the cascading effects of burrow resources on sympatric animal community assemblages and interspecific interactions are largely unknown. During 2020–2023, we monitored 184 Chinese pangolin (Manis pentadactyla) burrows using camera traps to reveal the burrow utilization patterns of commensal species. We totally recorded up to 57 species, with 19 mammal species, 32 bird species, and 1 reptile species recorded in the burrows revisited by Chinese pangolin, with 19 mammal species and 25 bird species in the non‐revisited burrows. Among them, most bird species as peripheral species primarily utilize soil mounds while most mammal species as burrow‐used species utilize burrow tunnels. The structure of animal communities in the burrows revisited by Chinese pangolins is more complex than that in the burrows not revisited. Furthermore, the positive correlation between community species in pangolin‐revisited burrows is also stronger. Our results demonstrate that the presence and repeated visitation by Chinese pangolins could enhance positive interactions (i.e., the emergence of one species promotes the emergence of another) among species that utilize the burrow resources (particularly, burrow‐used species). Our study provides the first evidence that the ecological role of the Chinese pangolin and its associated burrow microhabitats in promoting the coexistence of burrowing commensals and the restoration of Chinese pangolin populations may potentially contribute to the restoration of local biodiversity and ecological processes.

In this study, we systematically analyzed the utilization patterns of Chinese pangolin burrows by sympatric species in Guangdong Province, China, including differences in species composition using burrow mounds and burrow tunnels. We found that repeated visits to burrows by Chinese pangolin promoted the use of burrows by sympatric species, suggesting that this behavior effectively maintained the availability of burrow resources; at the same time, the burrow resources of Chinese pangolin also promoted multispecies coexistence through enhanced interactions among burrow‐using species, which was conducive to the maintenance of biodiversity. This study explores the role of the Chinese pangolin as an ecosystem engineer in forest ecosystems from the perspective of its provision of burrow resources for sympatric species and emphasizes the potential benefits of conservation efforts for other sympatric species.

1. Introduction

At the fine scale, the diversity and heterogeneity of microhabitat resources determine distribution patterns in a diversity of ecosystems (Brown et al. 2020; Alaniz et al. 2023; Gorczynski et al. 2023). Ecosystem engineers can frequently create or modify microhabitats through physical behaviors, which alter habitat resource availability and thus indirectly affect the survival and distribution of other species (Gálvez Bravo, Belliure, and Rebollo 2008; Di Blanco et al. 2020; Orazi et al. 2022; Alaniz et al. 2023), such as the water ponds built by Eurasian beaver (Castor fiber Linnaeus, 1758) (Orazi et al. 2022) and tree cavities formed by woodpeckers (Alaniz et al. 2023). Burrowing mammals are known to create complex burrow systems as essential components of microhabitat resources for a range of activities, as well as providing shelter, thermal refuges, foraging grounds, and other resources for other sympatric species (Davidson, Lightfoot, and McIntyre 2008; Dawson et al. 2019; Murphy et al. 2021). When natural refuges are limited due to anthropogenic disturbances or natural conditions, the emergence of new animal burrows may increase the supply–demand (DeGregorio, Veon, and Massey 2022; Alaniz et al. 2023). For example, the Pallas's Cat [Otocolobus manul (Pallas, 1776)] often utilizes burrows of Tarbagan marmot [Marmota sibirica (Radde, 1862)] for breeding and escaping predators (Ross et al. 2010), and plateau pika [Ochotona curzoniae (Hodgson, 1858)] burrows alter the structure of alpine grassland bird community (Lambert et al. 2022). Animal burrows are currently one of the most biologically utilized microhabitats, and a reduction in burrow resources may lead to increased inter‐ or intraspecific competition and even have some negative impacts on the conservation and restoration of local biodiversity (Murphy et al. 2021; DeGregorio, Veon, and Massey 2022; Alaniz et al. 2023).

The co‐utilization of the burrows and associated microhabitats by multiple sympatric species has received much attention and may provide insights into the restoration of ecological processes. The species co‐utilization of burrows can support a diversity of burrow‐dependent biological communities, but it also raises questions about the ecological interactions that occur in these habitats. The species interactions in the community are important carriers for ecosystem energy flow and material cycle and thus play a key role in enhancing ecosystem stability (Karakoc, Clark, and Chatzinotas 2020; Kim, Dolph, and Terui 2022; Luo et al. 2023). However, most studies focus on descriptive, for example, records of utilizing species and simple delineation of spatiotemporal niches (Di Blanco et al. 2020; Murphy et al. 2021; DeGregorio, Veon, and Massey 2022). By analyzing the community structure and species relationships of burrowing co‐exploiting organisms, we can demonstrate the aggregation effect of burrowing microhabitats on sympatric fauna and the modification of multi‐species coexistence relationships, which could contribute to revealing the key role of ecosystem engineers in ecological restoration.

Pangolins are the only mammals that are covered with scales and belong to the order Pholidota and the family Manidae. There are nine species (four in Asia: Manis crassicaudata J.E.Gray, 1827; Manis pentadactyla Linnaeus, 1758; Manis javanica Desmarest, 1822; Manis culionensis (de Elera, 1915); four in Africa: Smutsia gigantea Illiger, 1815; Smutsia temminckii Smuts, 1832; Phataginus tetradactyla (Linnaeus, 1766); Phataginus tricuspis Rafinesque, 1821; the recently discovered species: Manis mysteria, its origin is still unclear) and all of which are mainly feeding with termites and ants (Challender, Nash, and Waterman 2019; Gu et al. 2023). With the exception of the P. tetradactyla, the other seven species of pangolins dig underground burrows to access subterranean termite/ant nests for foraging (the habits of M. mysteria have not yet been reported), and use these burrows as long‐term shelters for resting, breeding, and raising their offspring (Challender, Nash, and Waterman 2019). Among them, the Chinese pangolin (M. pentadactyla) as a typical burrowing mammal in forest ecosystems, can excavate many burrows within its home range each year for foraging and resting (Wu, Sun, and Zhang 2020; Sabin et al. 2024). Some of these burrows can be revisited and modified by primary excavators or secondary utilized animals, while others are completely abandoned after physical damage (Figure 1). Their burrowing activities frequently alter habitat heterogeneity by creating surficial bare mounds and expanding underground space (Sun et al. 2024), and the burrows have been monitored for use by a wide range of animals (Sun et al. 2021). Human activities such as poaching, urban expansion, deforestation, soil excavation, and degradation of natural water sources, have led to a sharp decline in their population sizes (Yang et al. 2018; Gao et al. 2022; Dhami et al. 2023). Therefore, more research on the ecological role of Chinese pangolin excavation is extremely urgent and can also provide practical support for the conservation, population restoration, and even reintroduction of the Chinese pangolin.

FIGURE 1.

Diagram illustrating the evolution of Chinese pangolin burrows and the secondary users of burrow resources. In the experimental study system, the colored Chinese pangolins indicate multiple visits to the burrows, while the gray Chinese pangolins represent those that will no longer use the burrow, which has been completely abandoned. We hypothesize that repeated visitation of resting or foraging burrows by the Chinese pangolin would enhance burrow resource availability, thereby exerting positive effects on community species composition and multi‐species coexistence. Burrows without Chinese pangolin visits may have limited species utilizing them, or may not be utilized at all. Burrow photo credit: Sun Song; Chinese pangolin illustrations credit: Pangolin Conservation Research Center of the National Forestry and Grassland Administration; other species illustrations credit: https://www.phylopic.org/

The Chinese pangolin is a newly defined ecosystem engineer, and how its burrow resources affect the co‐utilization by sympatric species remains largely unexplored (Sun et al. 2021). In addition, the use of burrow resources by sympatric species is not limited to the exploration of burrow tunnels, and many birds are also found to peck, sandbathe, or sunbathe on bare soil mounds (accumulation of subsurface soils overturned by burrowing mammals) outside of burrows (Waterman and Roth 2007; Davidson, Lightfoot, and McIntyre 2008). Animals' use of different parts of the microhabitat reflects their resource needs, and similarity in resource needs is a key driver of animal community assemblage (Kraft, Valencia, and Ackerly 2009). Therefore, by analyzing the characteristics and differences in community structure formed by species co‐using burrow tunnels or exposed mounds, we can explore the patterns of burrow utilization by commensals. We can also examine the role of pangolin burrows in promoting ecological niche differentiation and multi‐species coexistence in biological communities.

As a typical narrow‐feeding species (Ma et al. 2024), the Chinese pangolin has little direct trophic relationship with other species (no natural enemies observed) (Tamang, Sharma, and Belant 2022). Nevertheless, there may be complex potential non‐trophic relationships (facilitation) between pangolin and other burrow utilizers that contrast with the lack of trophic relationships. Monitoring has shown that the Chinese pangolin often re‐excavates its resting burrows and revisits some of the foraging burrows for food or resting. Such repeated visits and excavating activities can contribute to maintaining the integrity of the burrow tunnel and the bareness of the soil mound, further positively influencing the availability of burrow resources (Figure 1; Figure S1). Thus, the co‐utilization of burrows by pangolin and other associated species may produce spatially convergent interactions. By exploring the impact of Chinese pangolin on the utilization of pangolin burrows by other sympatric species, we can expand our understanding of the non‐trophic interactions and conservation between Chinese pangolins and local animal communities (Yeakel et al. 2020; Losapio et al. 2021).

In this study, we monitored 184 burrows excavated by Chinese pangolins using camera traps during 2020–2023 in three subtropical forests in Guangdong Province, Southern China, to reveal the burrow utilization patterns of commensal species. We primarily examined whether Chinese pangolins’ revisits to burrows could affect the secondary users of their burrow resources (Figure 1). We predicted that the revisitation by Chinese pangolin could facilitate burrow utilization of commensal species and increase positive interactions among co‐utilizers, thus we assessed (1) the impact of Chinese pangolin repeated visits and burrowing microhabitats on the community composition of secondary users of its burrows, and (2) the effects of burrow microhabitats and repeated visit by Chinese pangolins on the interrelationships among other microhabitat‐utilizing species.

2. Materials and Methods

2.1. Study Area

We conducted research in three regions in Guangdong Province, China, where the population of Chinese pangolin is relatively abundant and their range of activity is stable (Figure 2). Heping Forest in Heyuan City, which is characterized by low rolling hills, with a mean annual temperature (MAT) of 19.7°C, a mean annual precipitation (MAP) of approximately 1717.1 mm, and a forest coverage rate of 73.4%. Tianjingshan Forest in Shaoguan City is located at the southern foot of the Nanling Mountains. The MAT is around 18.5°C, the MAP is approximately 2800 mm, and the forest coverage rate is as high as 97.6%. Wuqinzhang Forest in Huizhou City, which belongs to the Lotus Mountain branch, the main peak has an elevation of 1186.2 m. The natural vegetation is mainly subtropical secondary evergreen broad‐leaved forest. The MAT is around 20°C, and the MAP is 1950 mm. Three research areas show no significant differences in climate characteristics and vegetation composition, making them suitable habitats for the Chinese pangolin.

FIGURE 2.

Map of the study area in the Guangdong province, China, and the 184 burrow (camera trap) locations. Among them, 90 burrows (revisited: 67, non‐revisited: 23) were distributed in Heping forests, 39 burrows (revisited: 23, non‐revisited: 16) were distributed in Tianjingshan forests, and 55 burrows (revisited: 16, non‐revisited: 39) were distributed in Wuqinzhang forests.

2.2. Burrow Searching and Camera Trapping

We searched for burrows within the pangolin distribution area using the transect method. We selected relatively fresh pangolin burrows during the survey because this allows us to monitor the ruined process of abandoned burrows by pangolins. To avoid spatial autocorrelation due to the close proximity between burrows, a single burrow was selected within a 100 m range. We distinguish pangolin burrows from those excavated by other fossil animals and differentiate between resting burrows and foraging burrows within pangolin burrows, based on the identification methods described by Sun et al. (2024). All purposed burrows are a mix of both resting and foraging burrows in this study and are divided into non‐revisited and revisited burrows by pangolins based on camera trap results. A total of 184 burrows were selected and unevenly distributed across the entire study area (Figure 2). We used cameras (OWL UVL4 and UVL5) to monitor these Chinese pangolin burrows. Cameras were mounted on trees approximately 150 cm distant from the burrow entrances and at a height of around 80–100 cm, angled toward the burrow entrances to capture all wildlife interactions. This can minimize the omission rate of capturing images of pangolins and other species, and fully monitor animal behavior (Zhang et al. 2023).

The camera sensitivity was adjusted to the highest level to ensure the capture rate of animals, and the camera setup typically includes 1 photo + a 15‐s video segment. No interval was set between different shooting groups to enhance the frequency of burrow utilization incidents by other species in the Chinese pangolin's habitat. Batteries and memory cards were typically replaced approximately every 3 months to ensure proper camera functioning. However, due to the impact of COVID‐19 during the monitoring period, some camera replacements were delayed to intervals ranging from 6 months to slightly over a year. This delay, along with camera malfunctions and heavy rainfall, resulted in cameras being deployed in the field for a duration varying from a minimum of 32 days up to 715 days.

2.3. Monitoring Data Analysis

Collected data are uploaded to the monitoring system of the Pangolin Conservation Research Center of the National Forestry and Grassland Administration. Species identification is first performed using Artificial Intelligence (AI), with its algorithm based on a neural network model using a fine‐grained deep learning framework. It performs real‐time depth model library comparisons on video data through a front‐end integrated GPU computing module, combined with edge high‐performance processing animal intelligence recognition algorithms, to identify, analyze, and count wildlife. Currently, it can recognize over 1400 bird species and 100 mammal species, with an accuracy rate of over 90% for the Chinese pangolin and the majority of other animals. Subsequent manual review is conducted to ensure the accuracy of species identification, referencing “China's Mammalian Diversity (2nd ed.)” (Jiang et al. 2017), “A Checklist on the Classification and Distribution of the Birds of China (3rd ed.)” (Zheng 2017), and consulting with experts. Animal behavior is classified into three categories: (1) burrow‐using behavior (entering and exiting burrows, staying at the burrow entrance); (2) soil mound‐using behavior (foraging, resting, sand bathing, marking, fighting, and preening at burrow mound); and (3) non‐burrow‐using behavior (walking, running, and passing by; Desbiez and Kluyber 2013; Dawson et al. 2019).

After the review process, the system exported a table with behavioral profiles and time for further analysis to determine the number of camera trap working days and independent effective photos. A single camera operating continuously for 24 h is counted as one camera day. A set of independent effective photos (IP) is defined as a group of consecutive photos captured by the same camera location with different species or with a time gap exceeding 30 min between photos of the same individual of the same species (O'Brien, Kinnaird, and Wibisono 2003). The relative abundance index (RAI) is employed as an indicator of the relative population size of species and is calculated using the formula as follows (O'Brien, Kinnaird, and Wibisono 2003): RAI = Ai/T × 100, where Ai represents the number of independent effective photos for the i‐th species and T represents the total working days of all infrared cameras.

2.4. Burrow and Species Define

In all monitored burrows, we categorized burrows into revisit and non‐revisit burrows based on the number of Chinese pangolin occurrences each camera captured (Table 1). The revisit burrows are characterized by intact entrances (the smoothness is related to the frequency of pangolin use) and not blocked tunnels, and their soil mounds are bare and have clear outlines; the burrows that pangolins revisit with dilapidated entrances and blocked tunnels and their soil mounds are covered by litters and have no clear outlines (Figure S1).

TABLE 1.

Burrow type and species identification.

Category		Description
Burrow	With Chinese pangolins revisitation	Burrows where the number of Chinese pangolin occurrences was ≥ 2.
	Without Chinese pangolins revisitation	Burrows where Chinese pangolins were never seen or were seen only once.
Species	Other species	Identified species with IP < 5 or only non‐burrow‐using behaviors were displayed.
	Burrow‐used species	Identified species with IP ≥ 5 and with more than five occurrences of burrow‐using behaviors, each lasting over 5 s.
	Peripheral species	Identified species with IP ≥ 5 and displayed soil mound‐using behaviors, each lasting over 5 s, but with less than five occurrences of burrow‐using behaviors.

Note: IP refers to “a set of independent effective photos.”

We conducted species identification at the species level, except for some nocturnal rodents (including mice and rats) that proved difficult to differentiate. We collectively referred to these as “Rodent species.” We also used “Soricidae species” as a collective term for some difficult‐to‐differentiate shrew family members. We classified all identified species based on their utilization of different burrow resources, as presented in Table 1. To clearly identify species dependent on resources generated by Chinese pangolin revisitation within burrows, we selected species utilizing burrow resources for our analysis (Table S1).

2.5. Statistical Analyses

We constructed a classification relative abundance matrix and calculated diversity indices of the burrow‐used species and peripheral species for each study burrow using the “hill taxa” function in the “hillR” package (Li 2018). When q = 0, this corresponds to species richness, considering only the presence or absence of species without considering relative abundance. When q = 1, this corresponds to the Shannon diversity index, incorporating species abundance, with relative abundance calculated using RAI. When q = 2, this corresponds to the Simpson diversity index, focusing primarily on the most common species. We compared the final results for significance using the Wilcoxon test. Additionally, we analyzed the sampling intensity to ensure that community comparisons were made at the same level of sampling effort (Figure S2).

We employed principal coordinates analysis (PCoA) with the Bray–Curtis dissimilarity index to compare the differences in community composition between specific taxonomic units. Negative eigenvalues were observed in the calculation of PCoA eigenvectors, and using the broken‐stick model test, it was found that their absolute values were small and had no impact on the subsequent PCoA analysis. Therefore, there was no need to transform the original distance metrics used in the calculations. The inter‐group differences in community structure were tested using permutational multivariate analysis of variance (PERMANOVA). To further ascertain whether these inter‐group differences might be influenced by the dispersion of the data, a permutation multivariate dispersion (PERMDISP) test was conducted using the “betadisper” function, with p value estimations derived from 999 permutations. The above analysis was conducted using the “vegan” package (Oksanen et al. 2013).

To further elucidate the interspecific associations among species utilizing burrow resources, as well as the interspecific associations with Chinese pangolins inside burrows visited by it, we employed the variance ratio (VR) method as proposed by Schluter (1984) to investigate the overall association among community species. This was accomplished by utilizing the “sp.assoc” function in the “spaa” package (Zhang and Ma 2013). If VR = 1, there is no overall association among species; if VR>1, there is a positive overall association among species; if VR<1, there is a negative overall association among species. To determine the significance of the overall association, the statistic W was utilized. When $x_{0.95}^2(N)<W<x_{0.05}^2(N)$, the overall association is deemed not significant. Second, we examined the association between different species using the “psych” package, calculating the strength of interspecies association with Spearman rank correlation. When the correlation coefficient (r) is greater than 0, it indicates a positive association, with a higher value closer to 1 signifying a stronger correlation, and a smaller p value in significance testing. Conversely, when it is less than 0, the association is interpreted as negative. Concerning the peripheral species, we classified them into different functional groups based on their activity patterns and trophic levels, including diurnal predators, diurnal prey, nocturnal predators, and nocturnal prey. Nonetheless, sympatric carnivores are not sufficient to pose a threat to some large herbivorous animals; thus, these species were separately grouped as a distinct group. Finally, we analyzed the interspecific associations between burrow‐used species and peripheral species in the two types of burrows.

All analyses in this study were conducted using R version 4.0.3 (R core team 2021).

3. Results

We deployed a total of 184 cameras to monitor burrows, yielding a total of 39 941 camera‐days of effective work. A total of 106 burrows were observed to be revisited by Chinese pangolins, while 78 burrows were found to be devoid by Chinese pangolin visits (see Table S1 for detailed information). Throughout the monitoring period, we identified a total of 57 species that utilize burrow resources, including 36 bird species, 20 mammal species, and 1 reptile species. The peripheral species and burrow‐used species in the revisited and not revisited burrows by Chinese pangolins, as well as the activity patterns and trophic levels of each species, were different (Table S2).

3.1. Species Diversity and Community Composition of Burrow Microhabitat Utilizers

The diversity of peripheral and burrow‐used species exhibited a consistent pattern between burrows visited by Chinese pangolins and those left unvisited. Wilcox test showed that, regardless of the value of q, burrows with Chinese pangolin visitation had significantly higher species diversity compared to burrows without Chinese pangolin visitation (Figure 3a,b). According to PCoA ordination plots and PERMANOVA tests, significant differences in community composition were observed between the two types (Figure 3c,d). The PERMDISP analyses revealed significant differences in community composition among peripheral species between the two types of burrows (F = 10.602, p = 0.002), while no differences were detected among species utilizing burrows (F = 0.093, p = 0.741). The burrows revisited and non‐revisited by Chinese pangolins, respectively, hosted 14 and 11 unique peripheral species, sharing a total of 26 peripheral species (Table S2); they hosted 6 and 1 unique burrow‐utilizing species, respectively, and sharing a total of 6 species (Table S2).

FIGURE 3.

Comparison of species diversity and community composition. Panels (a) and (b), respectively, represent the species diversity of peripheral and burrow‐used species for each study burrow, while panels (c) and (d), respectively, illustrate the community composition of peripheral and burrow‐used species. Here, q = 0: Species richness, q = 1: Shannon–Winner index, and q = 2: Simpson index. “***”: p < 0.001, “**”: p < 0.01, “*”: p < 0.05.

3.2. Interspecies Associations

The VR of both visited and unvisited burrows by the Chinese pangolin is less than 1 from Table 2, indicating a negative overall association among community species. The statistic W of the variance test proportion coefficient is outside the interval $[x_{0.95}^2(\mathrm{df}),x_{0.05}^0(\mathrm{df})]$, suggesting that this negative association is significant.

TABLE 2.

Interspecific overall association.

Burrow	Species	VR	W	df	$[x_{0.95}^2(\mathrm{df}),x_{0.05}^0(\mathrm{df})]$
With CP revisitation	Burrow‐used species	0.133	14.144	106	[83.240, 131.031]
	Peripheral species	0.130	13.831	106	[83.240, 131.031]
Without CP revisitation	Burrow‐used species	0.296	23.099	78	[58.654, 99.617]
	Peripheral species	0.351	27.398	78	[58.654, 99.617]

Abbreviations: CP: Chinese pangolin; VR: variance ratio; W: statistic; df: degrees of freedom.

In the burrows revisited by Chinese pangolins, there are statistically significant positive associations between Chinese pangolins and the various burrow‐used species, with a total of 27 pairs (including 12 pairs with p < 0.05, 10 pairs with p < 0.01, and 5 pairs with p < 0.001), and 2 pairs showing significant negative correlations (Figure 4a). For the species utilizing burrows without pangolin visits, six pairs exhibit significant positive correlations (Figure 4b). Among the peripheral species, species pairs are generally more positively correlated (Figure 4c,d; Figure S3). Furthermore, we also found that despite variations in the activity peaks of burrow‐used species, there is a high degree of overlap in their overall activity patterns (Figure S4).

FIGURE 4.

Network graph depicting species interactions. The interspecific association of burrow‐used species: (a) with Chinese pangolin revisitation and (b) without Chinese pangolin revisitation, and the association between different functional groups of peripheral species: (c) with Chinese pangolin visitation and (d) without Chinese pangolin visitation. MP: Manis pentadactyla (Chinese pangolin); CE: Callosciurus erythraeus; DP: Dremomys pyrrhomerus; LE: Leopoldamys edwardsi; MC: Myophonus caeruleus; MK: Mustela kathiah; MM: Melogale moschata; RS: Rhizomys sinensis; TM: Tamiops maritimus; TS: Tamiops swinhoei; PL: Paguma larvata; PP: Prionodon pardicolor; Rod: rodent species (rice and rats); Sor: Soricidae species. D_prey: diurnal prey; D_pred: diurnal predator; N_prey: nocturnal prey; N_pred: nocturnal predator; Herbi: herbivore.

The interaction results among all species indicate that there are 480 and 259 pairs of interspecific associations between burrow‐used species and peripheral species at burrows revisited by Chinese pangolin and those not revisited, respectively (Figure 5; Table S3). Regardless of whether pangolins revisit the burrows or not, the number of positive associations between burrow‐used species and peripheral species is greater than the number of negative associations (Figure 5). In addition, the burrows revisited by Chinese pangolins (Figure 5a, 74.79%) have a higher proportion of positive associations between burrow‐used species and peripheral species compared to those not revisited (Figure 5b, 67.18%).

FIGURE 5.

Interspecific associations between peripheral and burrow‐used species. The larger loop on the left represents peripheral species, while the smaller loop on the right represents species that utilize the burrows. (a) With Chinese pangolin revisitation; (b) without Chinese pangolin revisitation.

4. Discussions

The revisitation of burrows by Chinese pangolins promotes the burrow co‐utilization by other species.

Our results suggest that the revisitation of Chinese pangolins to burrows enhances the burrow resource utilization by other species, resulting in higher diversity in community species. This supports our first hypothesis, emphasizing the importance of Chinese pangolins as ecosystem engineers in excavating burrows and preserving them by revisitation and re‐utilization, thus contributing to improving habitat resource heterogeneity and maintaining biodiversity (Dhital et al. 2020; Sun et al. 2021).

For abandoned burrows that have not been revisited by pangolins for a long time, the burrow entrance can collapse quickly due to rainfall and the filling of litter and soil (personal observation). Tunnels are the most important component of burrows, providing a stable microclimate, food, and shelter resources. Chinese pangolin burrows can significantly buffer ambient temperature fluctuations (daily temperature fluctuations inside burrow <0.2°C), thus as thermal refuges for many burrow‐used species (Sun 2022). For example, small‐toothed ferret‐badger [Melogale moschata (Gray, 1831)] and Soricidae species have been monitored and usually use burrows for rearing offspring. Non‐digging species usually enter burrows to seek food (such as escaped termites), or shelter from dangers (prey evading predators) in the meanwhile (Di Blanco et al. 2020; Symes et al. 2019). As reported by Dhital et al. (2020), masked palm civet (Paguma larvata) and yellow‐throated marten (Martes flavigula) were often seen to enter or try to enter the active burrows of pangolins in Nepal.

Additionally, many spiders would weave webs at the burrow entrance to catch flies or insects, which in turn attracted birds to peck at the insects stuck in the webs. Carnivorous animals always search for prey within burrows or use burrows for hunting, such as the yellow‐bellied weasel (Mustela kathiah Hodgson,1835) preying on rodents at burrows. This behavior may be a general outcome aimed at concentrating search efforts in areas with higher prey activity (Sillero‐Zubiri and Gottelli 1995; Agha et al. 2017; Hofstede and Dziminski 2017). Thus, the visit to and reuse of burrows by pangolins safeguards the openness of the burrow entrance and the integrity of the burrow tunnel and indirectly maintains the use of the burrow by other animals and the resulting trophic cascade. We found that the RAI of species in burrows revisited by Chinese pangolins was higher than those in burrows not revisited (Figure S5).

The revisiting or re‐modifying of burrows by pangolins also contributes to maintaining the integrity and bareness of the mounds due to the habit of scraping grass into the burrows of pangolins. The bare mounds can create sand‐bathing environments for some bird species (Andersen, Bennett, and Holbrook 2021). For example, silver pheasant [Lophura nycthemera (Linnaeus, 1758)] and Chinese bamboo partridge [Bambusicola thoracicus (Temminck, 1815)] exploiting mounds, potentially for removing lipids from their feathers, dealing with ectoparasites, or regulating body temperature (Olsson and Keeling 2005). Additionally, it provides new food sources that may be harder to access on undisturbed soil surfaces (Desbiez and Kluyber 2013), including the seeds in the deep soil overturned to the surface by pangolins and the wind‐carried seeds intercepted by a mound (Grinath et al. 2018). For old abandoned burrows, however, mounds tend to disappear after 1–2 years, mainly due to rainfall and animal trampling. Therefore, the variation in resource availability between revisited and non‐revisited pangolin burrows is the main reason for the significant difference in the diversity of species utilizing the two types of burrows (Figure 3).

According to the combined analysis results of PERMANOVA and PERMDISP, there may actually be no significant difference in the community composition of burrow‐used species between the two types of burrows. This may be because the species of burrow‐used animals is relatively few, and the species that use the burrow tunnels of pangolin‐abandoned burrows essentially belong to a subset of those that use the tunnels of revisited pangolin burrows. On the contrary, the results of both analyses show that there are significant differences in the community composition of peripheral species between the two types of burrows. This indicates that the changes in burrow resources caused by pangolins' revisits may have attracted more species of birds to utilize them, resulting in significant changes in the composition of the peripheral species community.

In the case of pangolin burrows distributed in forests, which are characterized by small burrow entrances (entrance diameter <20 cm) and deep tunnels, abundant litter can quickly cover the soil mound and block the burrow opening if without pangolin's revisitation (Sun et al. 2024). Few burrows were encountered in the pangolin's historical range, despite local feedback that pangolins disappeared in less than 5 years (Gao et al. 2022). For species that are highly dependent on pangolin burrows, the loss of burrows may trigger the absence of key survival resources and ultimately lead to migration or even disappearance (Linley et al. 2024). Given the habitat destruction and continuous reduction of burrow resources, the importance of the function of Chinese pangolins deserves more attention.

4.1. Population Decline and Anthropogenic Disturbance Worsen the Community Structure of Pangolin Burrow Commensals

Community association reflects the strength of interactions among community species (Götzenberger et al. 2012; Yuan et al. 2018; Luo et al. 2023). In this study, VR and statistic W demonstrated a significant negative correlation in overall interspecies associations of the community, indicating an unstable community structure of pangolin burrow commensal species. This could be because: (1) the severe decline of wild pangolin populations in mainland China, which leads to the number of burrows in the existing pangolin range being small, and the disappearance of burrows in the pangolin's historical range (Lin 2011; Shrestha et al. 2021); the poor burrow resources cannot afford the long‐term exploitation of the entire commensal biome, and this lead to the potential competition among species (Sebastián‐González et al. 2016); (2) the burrow resources in unvisited burrows may decrease or disappear, leading to changes in the composition of these species and causing community instability; (3) excessive anthropogenic disturbance affects the use of pangolin burrows by commensals, such as the fear effect caused by the frequent passage of people and domestic animals and predation by dogs or cats (Yang et al. 2018; Shrestha et al. 2021).

4.2. Chinese Pangolin Promotes Multispecies Coexistence by Enhancing Interactions Among Burrow Commensal Species

In this study, we analyze for the first time the effects of pangolin on the interspecific relationships of burrow commensal species and argue for a key ecological role of pangolin in promoting multi‐species coexistence. We found a commensal community with a higher proportion of positive associations among species always formed in burrows revisited by Chinese pangolins compared with burrows non‐visited. A positive association between two species means that the presence of one species potentially promotes the emergence of the other. Therefore, the more positive associations between community species, the more favorable the coexistence of multispecies (Luo et al. 2023). This result partially explains why the burrows revisited by pangolins attracted more exploiters but still maintained a relatively stable community of commensal species. There is also a more positive association between pangolins and other burrow utilizers, indicating that the pangolin is not a lone player in the ecosystem but interacts indirectly with many species through burrows (Sun et al. 2021).

Overall, revisits to burrows by pangolins promoted both more positive and negative associations between burrow‐used species (burrow tunnel utilization) than peripheral species (mound utilization) (Figure 4). This may be because the more predator–prey and competition trophic relationship that exists between burrow‐used species and the structural features of pangolin burrows (only one entrance) increase the probability of successful hunting (Zhang et al. 2024). Predators are also attracted to wait for an opportunity to catch prey (Dawson et al. 2019). For example, leopard cats [Prionailurus bengalensis (Kerr, 1792)] have been observed crouching at the burrow entrance to wait for rodents (most abundant pangolin burrow utilizers; Figure S4), and yellow‐bellied weasels (M. kathiah) have been observed carrying rodents out of burrows. As a result, predators will tend to utilize the burrows which are utilized by prey, thus constituting a positive association (Johnston 2017).

For peripheral species, predators may be difficult to capture successfully on mounds (mounds are open on all sides), but we also found burrows visited by Chinese pangolins promote more positive associations among species compared to burrows not revisited by Chinese pangolins (Figure 4c,d; Figure S1), and Chinese pangolins show more positive correlations with various species. This could be due to the fact that resources provided by bare mounds are greater than old mounds covered with litter.

In addition, we also evaluated associations between burrow‐used species and peripheral species. Results showed that there are more positive associations between burrow‐used species and peripheral species at burrows with pangolin revisited. This indicates that pangolins also promote the coexistence of two functional groups of co‐utilized species in burrows. In addition, regardless of whether the pangolin revisited the burrow or not, the number of positive associations between burrow‐used species and peripheral species was greater than that of negative associations. This indicates that the utilization of peripheral species benefits not only from the revisitation by the Chinese pangolin but is also related to the presence of burrow‐used species. We speculate that some burrow‐using species also modify the burrows by clearing and expanding tunnels, thus creating usable resources. Similarly, carnivorous animals among the burrow‐using species (such as Prionodon pardicolor and P. bengalensis) use the burrows for hunting, which may explain the negative connections between the two species groups. Therefore, the burrow acts as a medium that facilitates stable interspecific interactions between burrow‐using and peripheral species (Figure 5).

Analyzing dynamic interactions between species can effectively improve our understanding of the spatial and temporal distribution patterns of species (Wisz et al. 2013; Astarloa et al. 2019), as well as the impact on ecosystem stability and biodiversity (Blois et al. 2013; Nieto‐Lugilde et al. 2018; Astarloa et al. 2019). For example, the nocturnal carnivores (P. pardicolor, M. moschata, and P. larvata) are interspecific competitors with each other due to a high degree of ecological niche overlap (food, habitat selection, and activity rhythms; Banerji 2021). They are therefore difficult to coexist in resource‐poor areas, such as plantation forests, where habitat heterogeneity is relatively low (Sun et al. 2024). However, pangolins reconcile this tense interspecific competition by providing burrowing resources to feed a large number of primary consumers (like rodents) and maintain coexistence among secondary consumers (Figures S3 and S4). This safeguards the structural integrity of the habitat food web and avoids excessive resource flows to primary consumers, thus increasing the ecosystem's resilience to disturbance.

4.3. Implications for Chinese Pangolin Conservation

We believe that the Chinese pangolin should be protected more for its functional importance and uniqueness than for its current endangered status (Yang et al. 2018; Khatiwada et al. 2022; Dhami et al. 2023; Panta et al. 2023). Burrows excavated by the Chinese pangolin attract multiple species to explore and utilize and revisits by the pangolin perpetuate the availability of burrow resources, which promotes multi‐species coexistence (Sun et al. 2021). Therefore, the conservation, restoration, and even reintroduction of wild populations of pangolin in China will improve the survival of many burrow‐dependent species and indirectly contribute to the restoration of biodiversity and multiple ecological processes.

However, the effects of the Chinese pangolin on biome and ecological processes are still largely unknown, including the effects of the burrow microclimate on thermoregulation in reptiles and invertebrates, and the ecological role of soil mounds in the formation of seed banks and the germination of pioneer plants (Coggan, Hayward, and Gibb 2018; Di Blanco et al. 2020). We suggest that pangolin burrow resources contribute to the population construction of producers and primary consumers, which in turn affects trophic relationships and other ecological processes from the bottom up (Lundgren et al. 2021; Mittelman et al. 2022). More extensive and in‐depth studies on the role of ecosystem engineers played by the Chinese pangolin thus are necessary and urgent.

In addition, serious anthropogenic disturbances remain in the pangolin range, particularly threats to the Chinese pangolin from domestic dogs and damage to burrows by other livestock. Therefore, there is a growing need for legal education and enforcement, coupled with support to change the livelihoods in these areas (Yang et al. 2018), to protect this environmentally sensitive and indicative species.

Conflicts of Interest

The authors have no conflicts of interest.

Supporting information

Table S1 Camera days and number of species

Table S2 Vertebrate species using burrows with and without the visit of Chinese pangolins by camera trapping

Table S3 Interspecific associations between peripheral species and burrow‐used species

Figure S1 Images of two types of pangolin burrows with burrow age more than one year. Among them, a‐d are pangolin burrows with revisitation by Chinese pangolin; e‐h are pangolin burrows without revisitation by Chinese pangolin.

Figure S2 Sampling effort and species diversity estimates for peripheral species (a, b) and burrow‐used species (c, d). (a, c) Species sampling coverage curves based on individuals, (b, d) rarefaction and extrapolation curves based on individuals, with shaded areas indicating the 95% confidence intervals.

Figure S3 Network graph of species interactions among peripheral species. The interspecific association of peripheral species (a: With Chinese pangolin visitation, b: Without Chinese pangolin visitation).

Figure S4 Comparative analysis of activity patterns among burrow‐used species. The upper triangle represents the activity patterns between different species pairs when Chinese pangolins (CP) visit burrows, while the lower triangle represents the activity patterns between different species pairs when Chinese pangolins do not visit burrows.

Figure S5 Comparison of relative abundance of burrow‐used species.

Sun, S. , Mao R., Dou H., et al. 2026. “Chinese Pangolin Changes Local Vertebrate Assemblages and Contributes to Their Interspecific Interactions by Burrowing and Revisitation.” Integrative Zoology 21, no. 3: 499–510. 10.1111/1749-4877.12946