Zoonotic Barrier Disruption and the Rise of the Third Plague Pandemic: A One Health Analysis of 19th-Century Yunnan and the Emergence of Yersinia pestis Strain 1.ORI

Simple Summary

Our research investigates how major pandemics can begin when human activity disrupts the natural world. We studied the origins of the Third Plague Pandemic, which killed millions globally after emerging from China’s Yunnan region in the 19th century. Using a “One Health” approach, we analysed the deep connections between the environment, animals, and human society at that time. In Yunnan, a mining and farming boom caused widespread deforestation. This habitat destruction forced wild rodents into closer contact with both common rats and human settlements. We propose that these shifts created new pathways for a dangerous plague strain to circulate and adapt. Once established in rats living near people, the disease could spread more easily to human populations. Contributing factors like migration, decline in real wages, food insecurity, opium use, heavy metal contamination, and social upheaval may have further accelerated the outbreak. By showing how specific environmental and economic changes in Yunnan set the stage for this pandemic, our study provides a historical framework. Understanding these past triggers helps us identify risks and prevent new zoonotic diseases from emerging today.

Abstract

The Third Plague Pandemic originated in 19th-century Yunnan, China, yet the confluence of factors that enabled the pandemic strain Yersinia pestis 1.ORI to emerge and spread globally remains unclear. Using a One Health framework, this study investigates how human-driven ecological and socioeconomic changes disrupted zoonotic barriers in Yunnan. We conduct an interdisciplinary historical analysis, triangulating evidence from Qing dynasty gazetteers, environmental reconstructions, and biological data on plague ecology, including host–vector dynamics, to model conditions for spillover and spread and to build a convergent, validated case. The analysis identifies a mid-19th-century convergence that created a high-risk interface: widespread deforestation from mining and agriculture, rapid population growth, increased synanthropic rat densities, and the turmoil of the Panthay Rebellion. Socioeconomic stressors—labour migration into mining valleys, currency devaluation undermining food security, and comorbidities such as malnutrition, heavy metal contamination, and opium use—may have further increased host susceptibility. This socio-ecological context catalysed spillover and establishment of the 1.ORI strain in commensal rat populations. The findings show the pandemic’s origin reflects spatiotemporal convergence rather than a single cause, while noting uncertainty in quantifying historical ecological and health parameters; the case offers a framework for assessing contemporary pandemic risks. It underscores how layered pressures operate across timescales.

1. Introduction

1.1. Zoonotic Barrier Disruption and the Rise of the Third Plague Pandemic—A One Health Analysis

At the dawn of the 21st century, global health is confronted by emerging infectious diseases driven by rapid socio-economic change, environmental disruption, and intensifying contact between human and animal populations [1]. From SARS to Ebola, the erosion of ecological barriers that once contained pathogens within wildlife reservoirs has repeatedly facilitated zoonotic spillover into human populations [2]. Building resilience against these outbreaks requires a clearer understanding of how pathogens emerge, evolve, and adapt—especially as human activities dismantle the natural defences that have evolutionarily prevented animal diseases from crossing into humans [3]. Addressing this threat requires an integrated approach, one that connects the biological, environmental, and societal drivers of disease emergence.

The One Health framework provides this essential integrative lens. It argues that pathogen evolution, host dynamics, and spillover risk cannot be understood in isolation, but must be studied as interconnected components of a destabilized ecosystem [4,5,6,7,8]. While this interdisciplinary approach is well-positioned to assess the development of zoonotic risk and inform prevention strategies, recent reviews have identified significant weaknesses in its application. These shortcomings diminish the capacity of One Health studies to deliver truly holistic assessments that incorporate the intrinsically interdependent variety of environmental, animal, and human health factors, thereby limiting their ability to develop effective risk prevention strategies.

For instance, a 2025 review by Carvalho et al. found that only 7.4% of 309 models evaluated incorporated the three major components of zoonotic risk: hazard, exposure, and vulnerability [9]. Furthermore, it is widely accepted that socio-economic factors play a critical role in shaping the anthropogenic activities believed to drive the recent increase in zoonoses. The integration of these factors is therefore essential for understanding the development of zoonotic risk and for formulating robust prevention strategies. However, in a separate sampling of 195 studies published between 1986 and 2023, Giacomini et al. (2025) found that only 15% of One Health studies could truly be described as interdisciplinary, simultaneously addressing ecological, economic, and social dimensions [10].

This paper responds directly to these identified gaps. It presents an interdisciplinary One Health analysis that follows the recommended integrated approach but, crucially, also adds a specific biological-microbiological dimension by examining the development of the pandemic Yersinia pestis strain, 1.ORI, as a central part of the assessment.

A useful way to conceptualise microbial activity within these interconnected systems is to view it as a process of microbes overcoming environmental and human barriers that have been progressively eroded across the biosphere- particularly within wildlife domestic landscapes [11]. For a microbe to develop into a disease-causing agent (pathogen) and trigger a zoonotic epidemic or pandemic, it must successfully navigate a series of these barriers [12] (p. 2). As these barriers operate across different spaces and populations, human activities increasingly disrupt natural processes and elevate the risk of zoonotic outbreaks.

1.2. Anthropogenic Environmental Change, Zoonotic Barrier (ZB) Erosion and the Opportunity for Zoonoses

The erosion of these zoonotic barriers (ZBs) is driven by a process of ecological destabilization. Large-scale environmental alterations—including global and regional climate change, as well as anthropogenic deforestation, agricultural expansion, and urbanization—degrade habitats and reduce the availability of ecosystem services for diverse wildlife species [12]. This destabilisation creates an imbalance that favours generalist pathogen hosts and vectors, granting them a competitive advantage and enabling rapid population growth. Consequently, the frequency and intensity of interactions at the wildlife-domestic interface (Figure 1) increase, elevating the probability of spillover. This systemic stress imposed on wildlife populations can induce a state of chronic physiological dysregulation, known as allostatic overload (Figure 2), which further compromises host immune function and disrupts ecological niches. Together, these disruptions create selective pressures that act as evolutionary bottlenecks, amplifying pathogen strains with enhanced transmissibility and greater potential to spill over into new hosts.

Central to this process is the concept of zoonotic barriers (ZBs)—the complex web of ecological, physiological, and behavioural factors that limit cross-species transmission. When large-scale anthropogenic changes such as deforestation, agricultural expansion, and urbanization degrade these barriers, they create evolutionary bottlenecks. These bottlenecks select for pathogen strains with enhanced transmissibility and spillover potential, setting the stage for epidemics.

1.3. The Third Plague Pandemic: A One Health Case Study of Zoonotic Barrier Integrity

The Third Plague Pandemic, originating in 19th-century Yunnan and driven by the Yersinia pestis strain 1.ORI (biovar Orientalis) [14], serves as a profound historical case study of this dynamic. Plague’s well-documented phylogeny and historical context offer a unique opportunity to dissect how socio-ecological disruption can catalyse pandemic emergence.

The initial transmission system in Yunnan is well established, involving the primary host Rattus flavipectus and the efficient flea vector Xenopsylla cheopis [15]. In contrast, the origins of the key pandemic strain (1.ORI) that spread around much of the world remain less certain. Genomic evidence suggests 1.ORI may have evolved from Tatera indica (Indian gerbil)-adapted strains in northern India before transferring to the closely related Rattus rattus and spreading to Yunnan [16], though this remains a hypothesis. Other endemic plague foci in Yunnan (e.g., strains 1.IN3, 1.IN5) are linked to rodents such as the Yunnan vole (Eothenomys miletus) and field mice (Apodemus chevrieri) [16], which occupy ecological niches distinct from the commensal Rattus species.

1.4. Hypothesis and Approach

This article hypothesises that the convergence of two factors precipitated the Third Plague Pandemic: first, the large-scale erosion of ZB in Yunnan due to 19th-century socio-economic and environmental disturbance, and second, the coincident evolution and spread of the Y. pestis 1.ORI strain, which was maintained by commensal Rattus species and efficiently transmitted by the flea Xenopsylla cheopis.

To test this hypothesis, we employ an interdisciplinary One Health methodology. We first trace the evolutionary pathway of the 1.ORI strain from its origins in the Tatera indica reservoir to its establishment in Yunnan’s Rattus populations. We then analyse the concurrent disintegration of Yunnan’s socio-ecological fabric through the integration of Qing dynasty gazetteers, historical medical records, and environmental data. This analysis focuses on how land-use change, real wage decline, opium cultivation, and heavy metal contamination collectively appear to have degraded wildlife health and compressed host populations, creating a demographic bottleneck that strongly correlates with the rapid spread of 1.ORI.

1.5. Paper Structure & Outcome Summary

Following this introduction, we detail our integrated methodological framework. The results are presented in two complementary sections. First, the evolutionary-eco-biological narrative of the 1.ORI strain’s emergence and host adaptation, establishes the strain’s evolutionary pathway, detailing its origins in the Tatera indica-Xenopsylla astia system and its subsequent host dynamics upon entering Yunnan. The second section is the historical-ecological analysis of ZB breakdown in 19th-century Yunnan, where we analyse the interrelated processes in 19th-century Yunnan that drove pandemic emergence. This analysis begins with the socio-economic shifts—including real wage decline, population growth, and deforestation—that triggered widespread ecological change. We then demonstrate how these changes precipitated a biological breakdown of ZBs, exploring the potential roles of heavy metal contamination and opium-driven food insecurity in suppressing immune function and facilitating infection.

Finally, we synthesise these threads to show how displaced Rattus populations, concentrated in human settlements, created a demographic bottleneck for Y. pestis, setting the stage for the Panthay Rebellion to act as the final trigger that unleashed the pandemic. By systematically linking pathogen evolution to socio-ecological context, this study provides a comprehensive model for understanding pandemic origins through a One Health lens.

2. Materials and Methods: An Integrated Historical Case Study

This study employs a structured, interdisciplinary methodology within a One Health framework. Our objective is to reconstruct the multi-causal emergence of the Third Plague Pandemic in 19th-century Yunnan by integrating distinct data streams. Given the historical nature of the inquiry, we use a process of triangulation and convergent validation, as Bendrey et al. advocated [4], where hypotheses are tested against the independent convergence of evidence from palaeoecology, historical demography, primary documents, and modern biomedical science.

This interdisciplinary, case study analysis is a systematic assessment of how the co-evolution of the Yersinia pestis 1.ORI strain and the anthropogenic erosion of ZBs created the essential conditions for the pandemic in Yunnan. We integrate a bio-environmental tracing of the pathogen’s spatiotemporal context with a One Health analysis of the factors compromising barrier integrity. This bio-environmental tracing draws upon both the Ecological-Molecular Genetic approach developed by Suntsov [17,18,19] and Rayfield et al.’s Niche Construction Theory [5], which together illuminate how unsustainable anthropogenic environmental change progressively increased human zoonotic risk over time. If environmental change is disproportionately driven by human activity, then community health and environmental health are not separate concerns but mutually constituting. As Reynolds, Kutz, and Baker [20] argue, a community’s health—including the knowledge systems, values, and institutions that shape its decisions—directly influences the socio-economic structures that, in turn, affect environmental, plant, and animal health. Accordingly, they contend that food security, contaminants, economic health, physical health, ecosystem health, and culture, language, and tradition collectively shape outcomes for both wildlife and local communities [20]. While all these factors may indeed have played a role in the plague pandemic in Yunnan, our analysis is necessarily constrained by the available evidence; we therefore focus on those dimensions for which we can reasonably adduce data: ecosystem health, food security, economic health, and contaminants, as these are traceable through palaeoecological evidence and contemporary historical accounts.

The methodological workflow is presented in four parts: (I) Criteria for Palaeoecological Data, (II) Criteria for Historical-Demographic Data, (III) Criteria for Primary Historical Document Analysis, and (IV) Logic of Interdisciplinary Integration (See Table 1).

2.1. Criteria for Selection and Analysis of Palaeoecological Records

Palaeoecological data provide physical evidence of landscape and biogeochemical change. Records were selected and interpreted based on the following criteria:

2.1.1. Geographic Provenance

Priority was given to lake sediment cores from documented centres of 18th–19th century anthropogenic activity, specifically the copper mining districts of Yunnan (e.g., Kunming Lake region, Dongchuan).

2.1.2. Proxy Evidence

Selected studies must analyse a combination of proxies:

Deforestation: Pollen spectra showing a decline in arboreal pollen and an increase in herbaceous/pioneer species.

Land-Use Intensity: Microscopic charcoal influx rates as a proxy for biomass burning.

Erosion & Contamination: Sediment accumulation rates and geochemical data (e.g., increased measurements of Pb, Cu, Zn) indicating erosional and pollutant flux.

2.1.3. Chronological Control

Included studies must employ robust, published age-depth models.

Temporal Resolution: Primary chronology for the last ~300 years with effective resolution for the 18th–19th centuries. Dating resolution is sub-decadal to decadal (5–30 years) with medium to high confidence, sufficient to correlate environmental shifts with historical phases, which are corroborated with primary historical documentation.

2.2. Criteria for Selection and Analysis of Historical-Demographic Data

Quantitative demographic and land-use data for 18th–19th century Yunnan are derived from Lee’s (1982) [21] foundational historical scholarship. Our use of this data adheres to the following framework:

2.2.1. Source Acknowledgment

We treat synthesized historical figures as best-available estimates illustrating order-of-magnitude trends (e.g., >350% population growth 1750–1850), not as precise statistics.

2.2.2. Awareness of Limitations

Our analysis explicitly acknowledges the documented limitations of Qing dynasty data, including potential under-registration in population registers and the variable use and settlement of land area.

2.2.3. Triangulation to Mitigate Bias

To mitigate source limitations, demographic trends are validated through triangulation with:

Palaeoecological Data: Independent evidence of agricultural expansion and deforestation.

Primary Historical Accounts: Qualitative descriptions of migration, labour, and settlement patterns.

2.3. Criteria for Primary Historical Document Analysis

Qualitative data were extracted from 19th and early-20th century Chinese historical documents, primarily local gazetteers (difangzhi), customs reports, and administrative memoranda.

2.3.1. Sourcing and Selection

Digital Repositories: Documents were identified via keyword searches in authoritative databases, including the Erudition Database of Local Gazetteers (Fangzhi Database, 方志数据库, In Chinese) and CNKI’s Local Chronicles Collection. These digital platforms aggregate county-level and prefectural records from multiple editions, enabling comparative tracking of terminology evolution (shuyi, 鼠疫, In Chinese) and spatial distribution patterns over time.

Printed Historical Compilations: Printed historical compilations, such as the Guangxu-era Yunnan Tongzhi (Yunnan Tongzhi, 云南通志, In Chinese) and selected fascicles of the Qing Veritable Records (Qing Shilu, 清实录, In Chinese), provide authoritative accounts to cross-reference local narratives and identify macro-level policy changes (e.g., granary failures, mining decrees, population relocations) that shaped zoonotic conditions in 19th-century Southwest China [22].

Inclusion Criteria: Passages were included if they contained direct references to: (a) plague symptoms or outbreaks, (b) rodent anomalies, (c) mining, deforestation, or cash-crop cultivation, (d) opium use or malnutrition, (e) famine or flooding, or (f) social unrest in mining districts.

2.3.2. Analysis and Categorization

A systematic inventory of relevant quotations was compiled.

Inductive thematic analysis was conducted: recurring themes were identified and grouped into categories relevant to ZB integrity (e.g., Host Susceptibility; Subcategories: Opium Use, Malnutrition).

Each quotation was coded to fit at least one category and subcategory.

Interpretive Framework: This inventory is treated as qualitative evidence of contemporary perceptions. The prevalence of themes across multiple independent sources is used to identify robust narrative patterns.

2.4. Logic of Interdisciplinary Integration and Causal Modelling

The core methodology is the synthetic integration of the above data streams to build a causal model.

2.4.1. Temporal and Spatial Alignment

We identified spatiotemporal coincidences (e.g., mining peaks, population surges, and deforestation phases in the same region and decade), noting that waves of change across different regions in Yunnan converged during the mid-19th-century Panthay Rebellion.

Convergent Validation for Causal Links: A hypothesized causal chain was constructed (e.g., Mining & Agriculture→ Deforestation + Metal Release → Rodent Displacement + Human Immunosuppression → Spillover Risk). Each link required support from at least two independent evidentiary streams.

Example—Human Immunosuppression: The hypothesis that miners faced elevated immune risk required: (A) Historical evidence of opium use/malnutrition, (B) Palaeoecological evidence of heavy metal exposure, and (C) Biomedical literature confirming the immunosuppressive mechanisms of (A) & (B).

2.4.2. Role of Modern Biomedical Literature: Contemporary Phylogenetic and Clinical Studies Served Two Functions

Mechanistic Plausibility: To establish the biological mechanisms that make our historical-pathogenic links plausible.

Interpretive Framework: To provide the modern scientific context through which fragmented historical and palaeoecological evidence is interpreted within a One Health model.

Table 1. Primary-Data-Categories and-Selection-Criteria.

Data Category	Primary Sources	Inclusion Criteria	Exclusion Criteria	Key Variables Extracted
Palaeoecological Records	Lake sediment cores (e.g., Braun et al., 2015 [23]; Li et al., 2022) [24].	1. Geographic origin within Yunnan or directly adjacent plague-relevant regions. 2. Time coverage includes 1700–1900 CE. 3. Analysis includes proxies for deforestation (pollen, charcoal) and/or heavy metal contamination (XRF, ICP-MS). 4. Published in peer-reviewed journals with clear dating methods (210Pb, 14C).	Records lacking robust chronology, from irrelevant ecological zones, or with no direct proxy link to anthropogenic change (mining, agriculture).	Deforestation rate estimates, heavy metal (Pb, Cu, Zn, Fe) concentration timelines, sedimentation rates.
Historical Demographic Data	Qing dynasty registers, secondary syntheses (e.g., Lee, 1982) [25].	1. Quantitative population or land-use data for Yunnan. 2. Time-series spanning at least 1700–1850. 3. Derived from recognized scholarly reconstructions 4. Provides prefectural-level or finer spatial resolution. 5. Is corroborated by palaeoecological evidence (indicating resource use relevant for population)	Isolated, uncorroborated figures; data without discussion of source limitations; national aggregates lacking regional specificity.	Population estimates, cultivated land (mu) area, population density calculations.
Historical Documentation	Local gazetteers (difangzhi), customs reports, travel logs (e.g., Yunnan Tongzhi) [26].	1. Primary source from the period 1750–1900. 2. Contains direct references to: plague symptoms/outbreaks (shuyi), rodent anomalies, mining or agricultural activity, deforestation, opium cultivation/use, famine, social unrest, & environmental/climate change. 3. Sourced from authoritative digital collections (Erudition Database, CNKI) or critical printed editions.	Vague or allegorical references; sources from unrelated geographic regions; tertiary interpretations without primary text.	Qualitative descriptions of events, economic conditions, and ecological changes. Keyword frequencies.
Phylogenetic & Biomedical Literature	Peer-reviewed journals	1. Studies elucidating Y. pestis 1.ORI lineage evolution, transmission dynamics, or metal acquisition systems. 2. Experimental or review articles detailing the immunosuppressive effects of heavy metals (Pb, Cd) or opioids. 3. Publications from the last 20 years, prioritizing high-impact or frequently cited works.	Purely descriptive phylogenies without ecological context; mechanistic studies on irrelevant pathogens or stressors.	Genetic markers (e.g., gly+/gly−), R. rattus & R. flavipectus Y. pestis hosts (R. rattus (RrC)), virulence factors, mechanisms of immune impairment.

Table 1. Primary-Data-Categories and-Selection-Criteria.

Data Category	Primary Sources	Inclusion Criteria	Exclusion Criteria	Key Variables Extracted
Palaeoecological Records	Lake sediment cores (e.g., Braun et al., 2015 [23]; Li et al., 2022) [24].	1. Geographic origin within Yunnan or directly adjacent plague-relevant regions. 2. Time coverage includes 1700–1900 CE. 3. Analysis includes proxies for deforestation (pollen, charcoal) and/or heavy metal contamination (XRF, ICP-MS). 4. Published in peer-reviewed journals with clear dating methods (210Pb, 14C).	Records lacking robust chronology, from irrelevant ecological zones, or with no direct proxy link to anthropogenic change (mining, agriculture).	Deforestation rate estimates, heavy metal (Pb, Cu, Zn, Fe) concentration timelines, sedimentation rates.
Historical Demographic Data	Qing dynasty registers, secondary syntheses (e.g., Lee, 1982) [25].	1. Quantitative population or land-use data for Yunnan. 2. Time-series spanning at least 1700–1850. 3. Derived from recognized scholarly reconstructions 4. Provides prefectural-level or finer spatial resolution. 5. Is corroborated by palaeoecological evidence (indicating resource use relevant for population)	Isolated, uncorroborated figures; data without discussion of source limitations; national aggregates lacking regional specificity.	Population estimates, cultivated land (mu) area, population density calculations.
Historical Documentation	Local gazetteers (difangzhi), customs reports, travel logs (e.g., Yunnan Tongzhi) [26].	1. Primary source from the period 1750–1900. 2. Contains direct references to: plague symptoms/outbreaks (shuyi), rodent anomalies, mining or agricultural activity, deforestation, opium cultivation/use, famine, social unrest, & environmental/climate change. 3. Sourced from authoritative digital collections (Erudition Database, CNKI) or critical printed editions.	Vague or allegorical references; sources from unrelated geographic regions; tertiary interpretations without primary text.	Qualitative descriptions of events, economic conditions, and ecological changes. Keyword frequencies.
Phylogenetic & Biomedical Literature	Peer-reviewed journals	1. Studies elucidating Y. pestis 1.ORI lineage evolution, transmission dynamics, or metal acquisition systems. 2. Experimental or review articles detailing the immunosuppressive effects of heavy metals (Pb, Cd) or opioids. 3. Publications from the last 20 years, prioritizing high-impact or frequently cited works.	Purely descriptive phylogenies without ecological context; mechanistic studies on irrelevant pathogens or stressors.	Genetic markers (e.g., gly+/gly−), R. rattus & R. flavipectus Y. pestis hosts (R. rattus (RrC)), virulence factors, mechanisms of immune impairment.

3. Results

3.1. Evolutionary Trajectory of Yersinia pestis: Environmental Adaptation and the Emergence of Strain 1.ORI

3.1.1. A One Health Approach to Pathogen Evolution: Contextualizing Genetic Change

Analysing Yersinia pestis through a One Health lens requires integrating its molecular evolution with the environmental contexts that selected for, and thus shaped, its defining traits. Understanding the biological evolution of Y. pestis requires examining how pathogens adapt to hosts, how environmental changes reshape disease dynamics, and how interactions between hosts, vectors, and pathogens can drive cross-species spillovers. Among zoonotic diseases, the plague stands out as one of the most studied due to its historical impact and well-documented evolutionary path.

A combination of modern molecular, genetic and environmental data has made it possible to reliably establish that (1) the direct ancestor of the plague pathogen is the pathogen of the Far Eastern scarlet-like fever Yersinia pseudotuberculosis O:1b [27]; (2) the transformation of the pseudotuberculosis population into the population of the plague pathogen took place in the recent historical past, no earlier than 30 thousand years ago, in the Late Pleistocene or Holocene [28,29] and (3) the speciation of Y. pestis occurred in natural conditions without any human influence in Central Asia in the parasitic system of the Mongolian marmot (Marmota sibirica) and its flea (Oropsylla silantiewi) [16].

From the settlements of the Mongolian marmot, the plague pathogen spread through parasitic contacts into populations of sympatric burrowing rodents (Rodentia) and pikas (Ochotona) and forming an area of primary natural foci in Asia and the far south-east of Europe (Figure 3). The plague existed exclusively in the form of natural foci until the development of human society violated the environmental barriers that restrained the pathogen within the boundaries of the wild. The socio-economic development of human society and demographic shifts created conditions for the introduction of the plague microbe into human society on a large scale. This occurred through direct contact between hunters and wild rodents, the breeding of domestic animals that encountered wild animals in the territories of natural foci and the ubiquitous synanthropisation of rodents (rats, mice). Thus, the development of human society at a certain stage led to the emergence of an important social problem—the health of society, the solution of which is currently being implemented within the framework of the One Health paradigm.

The map (Figure 3) traces anthropogenic (human-driven) spread of plague, particularly the movement of a non-glycerine-fermenting strain (gly−) associated with human populations and trade routes originating in the Indian subcontinent. This anthropogenic spread of Y. pestis and the plague is also illustrated in Figure 4, which focuses on the development of Y. pestis adaptations to different primary host populations that resulted in different gene variants, as 1.ORI that spread from Yunnan to Hong Kong via Rattus flavipectus and the flea vector, Xenopsylla cheopis [30]. A fundamental transition in the pathogen’s ecology occurred when anthropogenic activities—such as agriculture, trade, and urbanization—disrupted the environmental barriers that had previously contained natural plague foci [19]. This breach initiated a new phase of evolution, in which the pathogen’s success became tied to its ability to adapt to human-altered landscapes and their commensal host species [30].

History knows of three plague pandemics with ascending amount of detail: the Plague of Justinian, which began in 541 in the city of Pelusium in the eastern Mediterranean, the “Black Death” (1346), which swept through medieval Europe, and the 3rd pandemic, which began in 1894 in Hong Kong and spanning many countries of the world [16]. Information about the causes of the first pandemic is limited. The alleged source of the second pandemic is the marmots of Central Asia: hunting for marmots, their fur trade, and military operations in Central Asia, the Lower Volga region, and the northern Black Sea region led to human infection with the plague and its spread to Constantinople and Genoa, and further throughout Europe [16]. However, the most complete, detailed information is available about the 3rd pandemic. It was a pandemic of the “rat plague”: synanthropic rats received the pathogen from wild rodents in natural foci and “transferred” it to human society. In the 20th century, against the background of the third pandemic of the “rat plague” in 1910–1911 and 1947–1949, epidemics of the “marmot plague” broke out in Manchuria, which were associated with the hunting of marmots [16]. But these epidemics had a regional scale and did not spread widely in the world.

3.1.2. Environmental Niche Specialization and the Genesis of the 1.ORI Lineage

The evolutionary pathway of the plague bacterium genovariant 1.ORI (biovar Orientalis), responsible for the 3rd Pandemic, can be traced to its parasitic system. We can trace this evidence of the pathogen’s adaption to 1.ORI primary hosts and the geography tied to these hosts formed in the Indian subcontinent and spread into Yunnan, China [16]. The 1.ORI lineage likely emerged from a Y. pestis enzootic system adapted to the Indian gerbil (Tatera indica) and its flea vector Xenopsylla astia in hot climates, making it distinct from older Y. pestis strains (e.g., Antiqua, Medievalis). Consequently, 1.ORI did not need the gly+ trait of other Asian gerbils, which live in the more northern regions of Asia with temperate climate and cold winters. Therefore, the Y. pestis strain lost its glycerine-metabolizing ability (gly–), a trait tied to its adaptation to T. indica’s physiology populating the hot Indian subcontinent climates [30].

This gly− biological trait has important implications: rodents that are hosts of plague microbes and live in colder countries that store fat metabolize it into glycerol, a substance that older strains of Y. pestis (such as Antiqua and Medievalis) could use, marked by the gly+ metabolic trait [30].

This trait is found in strains emerging from natural plague reservoirs in northern India and possibly parts of the Middle East. Importantly, T. indica is the only primary host of Y. pestis known to carry the X. astia flea, which may have served as a bridge to marmot-associated plague of the 1.IN gene variant typical for populations of the Himalayan marmot (M. himalayana) [30] (Figure 5).

3.1.3. 1.ORI Geographic Dispersal Through Rattus Hosts & the X. cheopis Flea Vector

The 1.ORI Y. pestis strain spread around the world [30,31]. But why this strain? It likely has to do with the efficiency of its transmission through its adaptable and quickly populating host Rattus rattus (RrC) and its flea vector, X. cheopis that has efficiently spread the pathogen. The spread of the 1.ORI lineage in Yunnan occurred through Rattus flavipectus (syn. R. tanezumi), which Aplin et al. (2011) argue belong to the same species complex as India’s R. rattus (RrC) [32]. These closely related rodents show high ecological overlap as forest dwellers but maintain separate nesting sites when coexisting. Their ability to interbreed enhances the complex’s adaptability and population growth, facilitating Y. pestis transmission. While no specialized fleas existed in their natural forest habitats, the introduction of X. cheopis, an African flea that was introduced to South/Southeast Asia, adapted to human environments and created an efficient transmission system. This flea colonized both RrC members equally due to their ecological similarity. By the mid-19th century, Y. pestis had established in Yunnan’s R. flavipectus populations, with Shi et al. (2018) documenting local evolution through unique CRISPR signatures [15]. Although 1.ORI first was introduced in India’s R. rattus populations, the ecological interchangeability within the RrC complex enabled seamless pathogen transfer to Yunnan’s R. flavipectus [15,32]. This host system’s biological plasticity, combined with X. cheopis vector efficiency, advanced conditions for the Third Pandemic’s emergence. Yet the question remains: what triggered its pandemic emergence?

3.1.4. Integrating Molecular Genetic and Ecological Approaches: A One Health Framework for Reconstructing Pathogen Evolution

The proposed evolutionary pathway of the 1.ORI lineage—particularly the role of the gly− phenotype and the distinction between the Tatera indica–Xenopsylla astia (gerbil-flea) and Rattus–Xenopsylla cheopis (rat-flea) systems—is derived largely from the synthesis of previously published genetic studies [e.g., 29–31]. While this evidence is persuasive, it is important to acknowledge that competing interpretations exist. Some studies, for example, have suggested alternative geographic origins for 1.ORI, including emergence within commensal rat populations on the eastern Tibetan Plateau rather than in Indian gerbil reservoirs [31,32]. The One Health framework adopted in this study does not dismiss these alternatives; rather, it evaluates them by triangulating genetic data with ecological, historical, and environmental evidence. In doing so, it privileges not merely molecular similarity, but adaptive and ecological plausibility.

This points to a broader methodological principle. The relatively recent origin of Y. pestis—within the last 30,000 years—offers a rare opportunity to reconstruct not only the phylogeny of a pathogen but also the environmental contexts that shaped its evolution [17]. For most microbes, such ancient ecological reconstruction is impossible, because their evolution occurred too far in the past to allow us to determine the environmental realities to which they were adapting. As a result, a strict Molecular Genetic (MG) approach has become habitual and mainstream, even for Y. pestis [18]. Foundational studies, such as Cui et al. [33] (2013) and Mas Fiol et al. (2024) [31], have elegantly resolved the phylogeny and gene variants of Y. pestis using molecular methods alone. Yet the MG approach, grounded primarily in the theory of neutral evolution, is limited in what it can reveal about adaptive history. Molecular similarity does not always reflect true kinship; homoplasies—parallelisms in evolution where similar traits arise independently in different lineages—can obscure phylogenetic relationships when reconstructed from molecular data alone [19].

The history of a “young” microbe like Y. pestis is more clearly legible through its adaptive features. Its direct ancestor is known (Y. pseudotuberculosis O:1b), the approximate timing of divergence is established, and the distinct ecological niches of ancestral and derived forms are geographically identifiable. This allows us to trace how transitional forms adapted to new hosts, vectors, and environments. Such adaptations are often observed in the subject of Ecology. Thus, a complete evolutionary reconstruction requires the synthesis of Molecular Genetic and Ecological approaches (MG + ECO) [18]—a synthesis that aligns directly with the integrative ethos of One Health. By situating genetic data within ecological and historical context, we move beyond phylogenetic description toward a mechanistic understanding of how Y. pestis evolved from a silent soil bacterium into a pandemic pathogen.

3.1.5. Zoonotic Barrier Erosion Diminishes Control of Pathogen Transmission

Plowright et al. (2024) [12] argue that a pandemic triggering event requires the sequential erosion of ZB (Figure 6). The initial barriers concern the pathogen’s survival and its ability to transmit to new animal hosts, facilitating its spread into new environmental habitats. This process culminates in the breach of the animal-human barrier. Subsequent stages depend on human factors, first the capacity of the immune system to respond to the novel threat, and finally, the convergence of conditions that enable efficient human-to-human transmission.

3.2. Three Interrelated Processes Driving the Emergence of the 3rd Plague Pandemic in Yunnan

The emergence of the Third Plague Pandemic in Yunnan was not a spontaneous biological event but the endpoint of a socio-ecological cascade. This analysis identifies a sequence of interrelated processes that systematically dismantled the multi-layered barriers to zoonotic transmission:

The results of this study found three interrelated processes driving the emergence of the 3rd Plague Pandemic in Yunnan, China. These three interrelated processes were:

(1)

the socio-economic shifts, including monetary depreciation and labour exploitation, which eroded human population resilience and drove landscape change,

(2)

the environmental and biological breakdown of zoonotic barriers and the synanthropic rat amplification, instigating host switching to human hosts, which led to:

(3)

the pandemic triggering event, the Panthay Rebellion, which exacerbated and accelerated the erosion of the final protective barriers, leading to widespread spillover and dissemination.

3.2.1. Socio-Economic Shifts That Compromised Host Resilience and Altered Landscapes

Real Wage Decline & the Devaluation of Coin Currencies: A Monetary Driver of Poverty and Ecological Disruption

The mid-19th century witnessed a decoupling at the very foundation of the Qing economy: its dual-currency system. The rupture between the value of silver—the state’s fiscal pillar—and copper—the people’s daily currency—triggered a major decline in real wages that, according to historical records, threatened subsistence of mine and agricultural labourers (Figure 7). Thus, available evidence indicates that impoverishment became an important force in the reconfiguration of Yunnan’s ecology, driving deforestation, agricultural shifts, and habitat encroachment. This section argues that this monetary and ensuing ecological crisis was a primary non-biological prerequisite, destabilizing environmental and human health barriers and creating the precise conditions that facilitated the spillover and amplification of Yersinia pestis.

• The Bimetallic System and Its Breakdown:

In Qing China, silver served large payments while copper cash (wen) was the medium of exchange for daily wages and local markets. The often-cited equivalence of “one tael to one cash-string (≈1000 wen)” was an accounting convention rather than a fixed legal peg. In practice, the liang–qian exchange ratio varied by time and region. From the late eighteenth century into the mid-nineteenth century, repeated attempts to stabilize the rate failed; after the “last copper century” (c. 1705–1808) [35] ended, shortages and quality deterioration of cash contributed to a persistent rise in the silver–copper ratio in many markets—from roughly 1:1000 to between 1:1500 and 1:2000+ on average nationwide, implying a c.25–50% depreciation of copper cash relative to silver [35,36,37,38]. As shown in Figure 8, the market exchange rate broke decisively above 1000 wen per tael after 1808, soaring to approximately 2250 wen per tael by 1850. This reflected a chronic shortage and debasement of copper coinage, leading to its severe depreciation against silver.

• The Mechanism of Impoverishment

This monetary shift had a catastrophic effect on real wages paid in wen, pushing the purchase of subsistence foods beyond reach. Allen et al. (2011, 23) [34] found that the minimum for food subsistence in Beijing was estimated between approximately 150 to 250 g of silver per day through most of the 19th century [34] (p. 23), until the 1890s quickly surpassing 300 g by 1900. As illustrated in Figure 7, which compares the silver value of daily wages for unskilled labourers across major cities, the estimated wage in Beijing fell very substantially from the early to mid-19th century. The average wage was especially low during the mid-19th century, falling to an estimated 1.21 silver taels (from approximately 110 wen) per day for unskilled labour in 1856 [34]. Crucially, by the 1850s, this wage often fell below the estimated minimum food subsistence level of 150–250 g of silver per day. In comparison, on the eve of the decoupling in 1807, the average wage was approximately 81 wen, equal to 3.32 silver taels, which would have left a small surplus after subsistence purchases [34].

Yunnan: an acute case

In Yunnan, the disparity appears even more acute. Hu Yuefeng’s (2021) analysis shows the province’s silver-copper exchange ratios were among the highest in the empire, frequently reaching 1:3000–4000 [39] (See Figure 9). Additionally, wages in Yunnan were typically lower than in Beijing; a historical survey (1769–1795) indicated a daily wage of 0.077 taels for unskilled public construction work in Beijing versus 0.048 taels in Yunnan [34] (p. 23). While workers may have managed small surpluses in the late 18th or early 19th century, the economic feasibility of supporting habits like opium consumption diminished as the mid-19th century approached. Nevertheless, thousands of kilograms of opium were being confiscated in the province, with the Kunming District accounting for half of all trafficking offenses and five-sixths of consumption offenses for the entire province [40]. Thus, labourers faced mounting economic pressures. A contemporary source in the General Gazetteer of Yunnan lamented: “Silver prices soared, copper coins were difficult to use; workers’ wages could purchase only half a dou of rice, and popular resentment surged [26].” Consequently, conflict and tension between government officials and the people rose, leading to increased litigation. This also led to a growth in counterfeiting and, possibly, more illegal (untaxed) mines [35] (pp. 126–127), [41] (pp. 19–20).

• Global Silver Devaluation

After the Panthay Rebellion, another monetary crisis developed as silver depreciated globally from the 1870s following the establishment of the gold standard in Western nations. As shown in Figure 10 (Gold to Silver Ratio), the consistent historical ratio of c. 16:1 broke after 1870 as Western countries sold off silver. This shift in the global monetary metals marketplace is illustrated by the fact that by 1913, eleven Western nations and Japan held 350% more gold than in 1885, while their silver stocks increased by only 33% [42].

Meanwhile, China’s GDP peaked around 1882, stagnated around 1898, and entered negative growth by 1905 [44]. Allen et al. (2011) observed that while subsistence food prices hit a low around 1850, they did not rise above the equivalent of 250 g of silver until the mid-1890s, after which they escalated rapidly to around 400 g two decades later [34], (pp. 20–26). This coincided with an increase in state welfare assistance from 1856 to its peak between approximately 1890 to 1905 [34], (pp. 26–30), even as the Qing administration was becoming increasingly insolvent. During this period, the silver-tael value of Chinese unskilled labourers’ wages remained relatively steady, gradually increasing until a peak around 1898–1901 [34], (pp. 20, 35).

The global selloff of silver caused it to lose half its value relative to gold by the time the pandemic reached Hong Kong in 1894. This final monetary shock eviscerated the fiscal capacity of the Qing state precisely when resources were most critically needed to combat the emerging Third Pandemic.

Rapid Population Growth and Deforestation in Late 18th and 19th Century

The dramatic population increase in Yunnan of over an estimated 350% growth in a hundred years, from 1750 to 1850 (Figure 11) reflects the intensity the Qing government was putting into its copper and silver mining, which was centred in Yunnan. However, to do so entailed the difficulty of providing the food for these labourers in a mountainous region. Consequently, these rapid demographic changes of higher population densities in rural mountainous regions of Yunnan incentivised rapid deforestation, vegetative clearance, and intensive and expansive agriculture to help meet the demand (Figure 12). Lee (1982) [21] observed this Qing dynasty phenomenon of increased population density in areas of diminished land availability, which was especially the case in Yunnan:

“In contrast to Ming population growth, the expansion of population during the Qing dynasty was inversely correlated to the availability of land. Indeed, in Yunnan the population increased fastest where land was least available. In 1775 Kunming and Chengjiang, the two inner core prefectures, had 863,000 people and over 2 million registered mu of cultivated land, that is, approximately one-quarter of the population and one-quarter the provincial acreage. By 1825 their share of the provincial population had increased to well over 2 million, almost one-third of the registered population. Their proportion of the cultivated acreage, however, had shrunk to 1.6 million mu, less than one-sixth of the provincial acreage. By the early nineteenth century, in other words, each acre of cultivated land in the core on the average supported twice as many people as an acre of cultivated land in the periphery”

[21] (p. 40).

The above population and deforestation graphs (Figure 11 and Figure 12) indicate a connection between a rapid increase in population growth and the intensification of deforestation from around 1775–1780 to meet increased mining and agricultural demand. The need for miners fostered most of the migration to Yunnan during this period, while this population influx led to agricultural intensification of these valley regions of mining zones. Yunnan, as characteristic of many mountainous regions, lacked a lot of agricultural lands. This with the intensity of cash crops, opium and later tobacco [21] led to more food being shipped to mining areas Li et al. (2022) [24] noted how these changes affected the landscape of the larger region, explaining,

“In the mountainous areas of China, the cropland cover expanded with the increase of people and immigration when the land use policy changed after the mid-18th century. According to agricultural historians, for example, the cropland area from 1724 to 1812 had increased by 32.6% in Sichuan (including Chongqing), by 250.6% in Guizhou, by 33.1% in Yunnan, by 63.8% in Hunan, and by 36.4% in Guangxi”

[24] (p. 12).

More specifically, within the large copper mining and production region centred in Kunming (Yunnan) and extending into the neighbouring provinces of Guizhou and Sichuan, Braun et al. (2015) [23] has reconstructed the impact of the mining and agriculture on deforestation (Figure 12). They found, “that the contribution of agriculture is about 80% at the beginning of the mining period, continually decreasing in the following years [23] (p. 50). In 1778 deforestation of primary forests was almost completely caused by mining, during the time of the height in copper production. Due to the following decrease in mining rates and the rising population, the share of agriculture for the destruction of the evergreen broad-leaved forests increased to a constant 100% since 1800. This indicates that agriculture is the main driving factor for deforestation in the 19th century” [23] (p.50). However, they also found that the emergence of agricultural impact was closely related to dramatic increase in copper mining in the region, with the mines and smelters dominating in the early decades of the copper century during the Qing Dynasty, with “ecological succession turned out to have a massive impact by reducing overall deforestation by up to 75% and transiently increasing environmental diversity in mining areas.” [23], (p. 50).

Kim (2018) [45] (pp. 100–120) found similar findings in their modelling vegetation change in a mining area around Dongchuan, about 160km north of Kunming (Figure 13) where they tied 18th century deforestation to mining whose increase in labour may have instigated a minor increase in agricultural lands, particularly south of Dashuitang. This reconstruction of deforestation illustrates the expansion of deforestation is especially concentrated around known mines, not just a general intensification of deforestation in the region, and according to the model, the increase of migrant miners may have instigated a minor increase in agricultural lands, particularly south of Dashuitang.

Figure 13. A & B. Model of Deforested Regions around Dongchuan, Yunnan (1700) (A) and (1800) (B). (Kim 2018) [45] (pp. 108–120) These are models of vegetation change in a mining area around Dongchuan, about 160km north of Kunming. These models were based upon historical records on copper mining outputs and “existing research on outputs, smelting technologies and the organization of mining (esp. Yang Yuda’s research on fuel consumption and deforestation)” to estimate copper production. Then the study mapped population centres, land use, and transport networks using GIS to estimate relative impact of different players and activities based upon historical evidence to reconstruct the regional mine activity and the corresponding demands for charcoal, as well as that which was required for the labourers of the mines [45]. The first model (A) is in 1700, and the second model [45] (p. 112) (B) is in 1800, which illustrates the widespread deforestation and vegetation clearance that has with erosion implications [45] (p. 117). Barren valley landscapes, as pictured in Figure 14, may, at least in part, be the result of such practices.

Figure 13. A & B. Model of Deforested Regions around Dongchuan, Yunnan (1700) (A) and (1800) (B). (Kim 2018) [45] (pp. 108–120) These are models of vegetation change in a mining area around Dongchuan, about 160km north of Kunming. These models were based upon historical records on copper mining outputs and “existing research on outputs, smelting technologies and the organization of mining (esp. Yang Yuda’s research on fuel consumption and deforestation)” to estimate copper production. Then the study mapped population centres, land use, and transport networks using GIS to estimate relative impact of different players and activities based upon historical evidence to reconstruct the regional mine activity and the corresponding demands for charcoal, as well as that which was required for the labourers of the mines [45]. The first model (A) is in 1700, and the second model [45] (p. 112) (B) is in 1800, which illustrates the widespread deforestation and vegetation clearance that has with erosion implications [45] (p. 117). Barren valley landscapes, as pictured in Figure 14, may, at least in part, be the result of such practices.

Figure 14. Minimal Vegetation Landscape from the Xiaojiang Valley to the Huize Plateau. (Kim 2018) is a photo (2007) illustrating the lack of vegetation on the road from the Xiaojiang Valley to the Huize Plateau [45] (p. 101).

Figure 14. Minimal Vegetation Landscape from the Xiaojiang Valley to the Huize Plateau. (Kim 2018) is a photo (2007) illustrating the lack of vegetation on the road from the Xiaojiang Valley to the Huize Plateau [45] (p. 101).

The harsh conditions of those that worked the land also likely played a role in this deforestation and clearance of vegetation. In June 1904, a British businessman, Archibald Little, reached the valley of the Dongchuan prefecture, observed its impoverished state:

“On reaching the bottom we found ourselves upon the edge of paddy fields, the rice being grown right up to the limestone rock; across these our way led to the city, where we were to repose a couple days before going further…Tung-chuan is a poor mountain city with not half of the population of Chao-tung [Zhaotong] and, notwithstanding the rich valley in which it stands, the population has a poverty-stricken aspect, especially in the surrounding villages, while in the city itself we did not notice any good shops, and were told there was not for the sale of silk, whereas in Szechuan silk is an article of dress common to all but the very poorest. Our missionary friends informed us that all the good land was owned by a few rich gentry, ex-officials, who reside within the city walls and extort half the crop from the wretched farmers for rent. There were one very productive copper mines in the neighbourhood, but these, being under official management, were no longer flourishing…”

[46]

Erosion from Mining Areas Leading to Heavy Metal Contamination in Lakes and Rivers

According to the Safe Drinking Water Foundation (SDWF), Water is often the primary casualty of heavy metal pollution [47]. During the Qing Dynasty, mining operations were unregulated, allowing waste products—including tailings, acid mine drainage, and residues from smelting and ore processing—to flow directly into river systems [48] (p. 163). For instance, in copper processing, ores were often submerged in or flushed with water (sometimes water initially used for cleaning rice) for cooling [48] (p. 163). J. Coggin Brown (1923) [49] observed this practice at Pao-p’ing-ch’ang (Debaopingchang, 得宝坪厂, In Chinese) in the broader Dongchuan copper-mining region, where discarded low-grade ores and gangue (containing disseminated sulphides and carbonates) were piled into dumps. He noted, “The water percolating through the dumps was of a bright blue colour and appeared to bear an appreciable quantity of copper salts in solution” [49] (p. 115)—a clear sign of severe copper contamination [50].

This water pollution was not limited to waste dumps; the smelting process itself was a major source. Figure 15 illustrates a Qing Dynasty smelting procedure (a Type B smelter for ‘thorough’ copper ores), showing both the use of water in the crab shell furnace and the cooling of the final copper product. The diagram also highlights the substantial demand for wood fuel (firewood and charcoal), which necessitated the felling of trees around the mine. This deforestation reduced vegetative cover, minimizing barriers to overland flow and allowing water contaminated during smelting to flow more directly into local rivers, carrying heavy metals with it.

Smelting of ‘thorough’ copper ores, Yunnan, 18th–19th century

This legacy of contamination is recorded in lake sediments across Yunnan. Studies indicate that heavy metal pollution increased during the Ming Dynasty and peaked in many areas during the Qing Dynasty. For example:

Dian Lake (south of Kunming): Sediment cores show the highest historical levels of heavy metals (Cu, Cr, Pb, Zn) on the Yunnan Plateau during the Ming-Qing period. Researchers link this directly to mining and metallurgy, exacerbated by deforestation-induced erosion that transported waste into the lake [51] (p. 8).

Xing Yun Lake (further south): Concentrations of Pb and Hg reached their highest levels in approximately 3500 years during the Qing Dynasty, tied to intensified land use and metal extraction [52] (p. 30).

Erhai Lake (near Dali): Similar contamination patterns are evident. Agricultural expansion and subsequent severe gully erosion during the Qing Dynasty increased sediment loads, which carried historical mining pollutants. Notably, a spike in heavy metals in the late 19th century may point to unreported or illegal mining activity [53] (p. 67), [54] (p. 24–25).

• The Role of Erosion and Hydrology

The karst topography prevalent in these regions amplified the problem. Steep, long slopes and intense rainfall lead to rapid runoff and rill formation, which efficiently transports soil and contaminants. Recent studies confirm that topography and hydrology are key factors in distributing heavy metals from mining waste into surrounding farmland and waterways [55] (p. 401), [56]. This relationship is further illustrated by the case of the Dabaoshan mine (Guangdong), where water safety standards in the receiving river were met during the dry season, but surpassed only during the rainy season, highlighting how surface runoff dictates the mobilization of pollutants (Figure 16).

• Long-Term and Contemporary Impacts

Heavy metal contamination from mining is not a historical artifact but a persistent and escalating environmental legacy. Modern research in Yunnan confirms that soils, water, and the food chain in mining regions remain severely impacted, refs. [58,59,60] demonstrating the enduring nature of pollution and pollution pathways that were often established centuries ago.

The widespread contemporary impact is clear. Studies document high contamination rates in soil, cultivated food, milk, and human populations in Yunnan’s mining areas, with direct links to health issues [61,62,63]. Provincial-scale assessments, rank Yunnan among China’s most severely polluted provinces, noting that “the overall pollution levels of heavy metals in the soil… are particularly severe,” a fact reflected in contaminated forage samples [63] (p. 6838).

Crucially, localized studies in the historic mining heartland provide direct, mechanistic evidence for how past operations would have polluted the environment. Research focusing specifically on the Dongchuan copper mining area found that soils within the mining zone contain significantly higher levels of Arsenic (As), Copper (Cu), and Zinc (Zn) compared to surrounding areas, frequently exceeding national safety standards [64]. A geostatistical map of copper presence in local agricultural soils (Figure 17) visually underscores this intense, localized pollution emanating from the mining area.

The process behind this lasting contamination is critical to understanding the historical scenario. Cheng et al.’s (2018) [64] study of the region hypothesizes, the modern extraction process itself activates and releases residual metals:

“These findings suggest that after Cu was extracted from the primary ore via flotation, residual elements accumulated on the surface along with Cu tailings. During ore cracking, these elements are activated and released in ionic or molecular forms into the soil, leading to complex metal pollution. Cu is expected to be the dominant pollutant in Dongchuan, accompanied by Cd, Pb, and Zn contamination”

[64] (p. 13).

This modern assessment provides a reference for reconstructing the historical impact. If inefficient Qing-era mining, roasting, and smelting techniques—which lacked any pollution controls—similarly mobilized metals from ore and tailings, then the contamination burden during periods of peak production would have been profound. The first half of the 19th century, when copper, silver, lead, and zinc production in Yunnan reached historical highs, likely represented a zenith for this type of unmitigated heavy metal release. The severe contamination observed in modern soils around abandoned sites is a direct result of processes that were operating, unchecked, at an industrial scale during the Qing Dynasty. Thus, contemporary evidence does not merely illustrate a similar problem; it validates the physical and chemical mechanisms at play. This would appear to shift the analytical focus from whether water contamination occurred during the 19th century to assessing what scale and geographic extent is.

3.2.2. Environmental and Biological Breakdown of Zoonotic Barriers and the Synanthropic Rat Amplification

Mining, Environmental Toxins, and Host-Pathogen Dysregulation: A Context for Yersinia pestis Emergence

The role of environmental toxins, particularly heavy metals, in human and ecosystem health is a subject of intense contemporary study in China, where an estimated 19.4% of agricultural land is currently contaminated [65]. As Li et al. (2025) note, this is a particularly acute and complex challenge in western Yunnan, a historical mining centre, due to its “unique geologic background… and the overlapping impacts of mining, agriculture, and urbanization” [66] (p. 32213).

These modern concerns find a powerful historical parallel. The intensive mining that reshaped 19th-century Yunnan’s economy and demography also fundamentally degraded its environmental health. We propose that this degradation likely acted as a significant compounding factor, intensifying the impact of Yersinia pestis by helping to erode both the ecological and biological barriers to zoonotic transmission.

• The Broad Ecological Impact: Habitat Degradation and Pollution

At a macro level, the environmental consequences of mining are well-established and form a plausible pathway to increased zoonotic risk. As Ellwanger, Ziliotto, and Chies (2025) describe the metal demand that drives mining triggers a destructive cascade: biodiversity loss, land-use change, and habitat homogenization [67] (p. 105). In Yunnan, this resulted in deforestation and landscape fragmentation near mountain valleys of mining areas, which increased contact between wildlife, rodents, and human settlements.

Concurrently, mining is a primary source of heavy metal pollution. Palaeoecological evidence from Yunnan supports this: sediment cores from lakes adjacent to historical mining operations show that levels of copper (Cu), chromium (Cr), lead (Pb), and zinc (Zn) reached historical peaks during the late Qing Dynasty—the period immediately preceding the pandemic’s emergence. For instance, in the northwest near Dali, Yang et al.’s (2022) analysis of Lake Jian sediments indicates an especially high peak in copper and iron concentrations in the early-mid 19th century, linked to regional mining activity [68]. Similarly, cores from Dian Lake, south of Kunming, reveal the highest historical levels of sediment and these metals found on the Yunnan Plateau [51], p. 8.

This widespread soil and water contamination represents a broad-spectrum environmental stressor. The metals enter the food web primarily through plants [69,70]. As Alengebawy et al. (2021) explain, heavy metals interfere with fundamental plant processes in the rhizosphere, disrupting “mineral transportation, nutrition uptake, and modification of photosynthesis” [69] (p. 42) (Figure 18). This physiological damage reduces crop vigour and yield, resulting in harvests that are both diminished in nutritional quality and contaminated with metallic toxins [69].

The potential human health consequences of ingesting such contaminated food are significant, with chronic exposure linked to organ damage [71] and increased infection susceptibility [72]. This ingestion pathway is key; as Tong et al. (2020) emphasize in their risk assessment, “ingestion was the dominant exposure pathway for having adverse effects on human health” [73] (p. 399). Thus, the mining landscape established a direct route for toxins into local populations. Metals entered the body either through the consumption of contaminated water or via the food chain—from contaminated soil into physiologically stressed crops. Both pathways ensured that rodent and human populations were chronically exposed.

• From Environmental Stress to Host Vulnerability: The Heavy Metal Hypothesis

The potential link between a polluted environment and disease susceptibility is illustrated by a persistent epidemiological observation: the focal nature of some plague outbreaks. Research spanning decades, from Rothchild’s work [74] to recent studies by Wei et al. (2021) [75] in Qingnan, has observed a correlation between soils with specific heavy metal contamination (e.g., As, Cd, Pb, Hg) and localised rodent plague mortality. The proposed mechanism is that rodents—and by extension, humans—consuming plants or water from these areas experience chronic exposure. This could lead to malnutrition (from plants grown in nutrient-leached soils and diminished plant yields) and direct immunotoxicity, as metals like lead are known to diminish immune function [72]. Therefore, the mining landscape may have fostered a reservoir host population that was not only ecologically concentrated but also physiologically stressed, with potentially weakened defences.

• The Microbiological Interface: A Potential “Metal Tug-of-War”

The chronic environmental exposure to heavy metals, as established above, did not merely pose a toxicological threat; it also had the potential to alter the very biochemical terrain within a host’s body where infection is fought. The ingested metals—whether from water or food—entered the bloodstream and tissues, creating a dysregulated internal “micronutrient landscape.” It is at this cellular and molecular level that a second, critical mechanism likely operated: the disruption of the essential competition for metals between host and pathogen.

Yersinia pestis and its mammalian host are engaged in a constant biochemical competition for essential metals—a “metal tug-of-war” (see Figure 19).

The Host’s Need (Nutritional Immunity): The host’s immune system requires a tightly balanced bioavailability of metals like iron (Fe), zinc (Zn), and manganese (Mn) to function optimally [76]. Key immune cells and antimicrobial proteins depend on these nutrients. The body actively sequesters these metals to starve invading pathogens, a defence strategy termed “nutritional immunity” [76].

“Nutritional immunity is a process by which a host organism sequesters trace minerals to limit pathogenicity during infection. Circulating concentrations of minerals, such as iron and zinc, decline rapidly and dramatically with the inflammation associated with infection. The decline in iron and zinc is thought to starve invading pathogens of these essential elements, limiting disease progression and severity…”

[77].

The Pathogen’s Strategy (Metal Theft): Y. pestis is an adept “metal thief.” To circumvent nutritional immunity, it deploys a suite of specialized high-affinity transport systems (e.g., Yfe, Ybt, Feo) to scavenge these same metals, particularly iron, which is absolutely required for its growth, metabolism, and spread within the host [6] p. 205–206.

The historical mining pollution in Yunnan could have catastrophically disrupted this delicate balance, potentially tipping it in the pathogen’s favour in two key ways:

Impairing Host Defences: Chronic exposure to immunotoxic metals like lead and cadmium—identified as key contributors to persistent infection risk (Zhang et al., 2024)—can directly compromise immune cell function. This would logically dampen the host’s overall immune response and impair the efficiency of its metal-sequestering ‘nutritional immunity,’ a defence strategy critical for starving Y. pestis of essential metals like iron [72].

Inadvertently Aiding the Pathogen: Paradoxically, an excess of certain bioavailable metals in the host’s system might aid the bacterium. High circulating iron can facilitate Yersinia colonization by providing a resource it is evolutionarily optimized to capture. Furthermore, elevated zinc and manganese have been shown in related bacterial models to alter host microbiomes and enhance bacterial virulence mechanisms [76].

• Synthesis: A Confluence of Stresses in a High-Risk Landscape

It is critical to acknowledge that many factors—burrow microclimates, specific plant metal uptake, and precise exposure thresholds—remain areas of active research. However, the different fields of evidence of 19th-century Yunnan converge to illustrate intensive mining did not merely alter the landscape; it saturated the local biosphere with immune-disrupting pollutants at historical concentrations. This created a zoonotic bottleneck: a dense, stressed population of rodent hosts living in degraded habitats and proximity to humans, while the very soil and water were primed to undermine their health and potentiate pathogen virulence. This multi-layered assault—ecological, toxicological, and microbiological—likely represented the final, critical erosion of the ZB, transforming Yunnan’s mining districts into a potent incubator for the coming pandemic.

Opium, Immunosuppression, and the Fuelling of a Pandemic in 19th-Century Yunnan

While environmental degradation, heavy metal exposure, and food insecurity undoubtedly eroded population health in Yunnan’s mining districts, these were often inescapable conditions of life and labour. Opium consumption presents a more complex and deliberate factor. Despite widespread poverty and malnutrition, labourers appear to have often prioritized purchasing opium above reasons of addiction, often at the direct expense of food. This may have indeed been a tragic convergence of economic distress, entrenched medical beliefs, and a biological trap that may have critically amplified susceptibility to Yersinia pestis.

• The Socio-Economic Trap: Opium over Sustenance

During the mid-to-late 19th century, as real wages of low and unskilled labour collapsed and the Panthay Rebellion (1856–1873) destabilized the region, opium paradoxically became more accessible. The drug’s legalisation, driven by outside pressure from Western countries and the Qing dynasty’s desperation for increased tax revenue, certainly played an important role. Following its de facto legalization in the 1850s, Yunnan rapidly transformed into China’s leading cultivator of opium poppies for the world’s largest domestic market [40] (pp. 1111–1112), [78]. While precise acreage of opium cultivated in Yunnan is difficult to ascertain, since it was an illegal activity until the 1850s, contemporary government officials considered it a widespread issue. Censor Lu Yinggu explained in 1839:

“Officials had identified locations where poppies were cultivated and opium was processed and sold, as well as key entry points through which foreign merchants trafficked opium into the interior and the main routes by which opium was transported from Yunnan into Sichuan (据该御史指出栽种罂粟熬烟售卖处所，并由外夷贩烟入内要口，及由滇省贩烟入川要路, In Chinese)”

[79]

This agricultural shift diverted land from food crops to more profitable—but soil-depleting—opium and tobacco [25,80], exacerbating local food insecurity. Within this context, opium consumption became a devastating budget priority for the labouring poor. Customs physicians in 1880s Kunming observed that some railway and mining workers spent about one-third of their daily income on opium, a choice that “undermined nutritional intake and basic livelihood security” [79]. Local gazetteers vividly echo this. The Gazetteer of Gejiu Subprefecture laments that for miners, “a day’s labour could not match half a day’s smoke” (一日之工，不敌半日之烟, In Chinese), explicitly noting opium’s precedence over food [81] (p. 105). This created a vicious cycle: wages depleted by opium reduced caloric and protein intake, compounding physical decline. As the Gazetteer of Guantong Department noted during epidemics, “the poor had no access to medicine, while the rich relied on opium to resist illness … leading to countless deaths” (贫者无医，富者吸烟抵病… 致死无数, In Chinese) [82] (p. 59). Opium was not just a recreational vice; for many, it was a perceived medicinal necessity, however misplaced, in the absence of other healthcare.

• The Medical and Cultural Rationale: Opium as Panacea

To understand this choice, one must consider opium’s deep integration into Chinese medical and cultural practice. Long before the 19th century, opium was enshrined in the Qing therapeutic remedies. Manuals like the Jiyan liangfang (《集验良方》, 1724, In Chinese) recommended opium preparations for a wide spectrum of ailments: cholera (huolan, 霍乱, In Chinese), fever, diarrhoea, stomach pains, and—crucially—zhangqi (miasmatic disorders, often associated with plague; 瘴气, In Chinese) [83]. It was seen as a substance that could control bodily fluids, preserve “vital energy” (yuanqi, 元气, In Chinese), warm the organs, and provide comfort against hunger and misery [83].

This perceived prophylactic and therapeutic role became especially salient during epidemics. Historian Lars P. Laamann notes that reports of “rat disease” (shuyi, 鼠疫, In Chinese)—a common descriptor for plague—often correlated with increased opium use [83]. The association went beyond ingestion; the practice of using fire and smoke to kill rats, observed by French explorer de Lagrée in 1867–1868 and linked to folk traditions of Marshal Wen, mirrored the act of opium smoking itself, potentially reinforcing a belief in smoke as a purifying barrier against miasma [83]. Thus, when plague threatened, the cultural script directed people toward opium not only as relief from symptoms but as a purported defensive agent.

• The Biological Trap: Opium as an Immunosuppressant

The tragic irony is that the very substance consumed for protection likely rendered users more vulnerable. The raw opium smoked in Yunnan, containing 10–12% morphine [84], is a potent immunomodulator [84,85]. Modern neuroimmunology explains that opioids like morphine activate receptors in the brain, triggering the hypothalamic-pituitary-adrenal (HPA) axis and sympathetic nervous system. This leads to the release of immunosuppressive hormones (e.g., cortisol) and catecholamines, which in turn suppress key immune defences [86,87] (Figure 20).

As summarized by Vallejo et al., acute and chronic opioid use inhibits humoral and cellular immune responses, including:

Antibody production

Natural killer (NK) cell activity

Pro-inflammatory cytokine expression

Phagocytic activity of macrophages [85]

The clinical consequence, well-documented in modern studies of heroin addicts, is a significantly increased incidence of severe infections [86]. This link was already being clinically noted in the late 19th and early 20th centuries. Physicians like Kee (1908) observed that opium addicts in China suffered disproportionately from tuberculosis, pneumonia, bronchitis, and diarrheal diseases [84]—a pattern consistent with opioid-induced immunosuppression.

• Synthesis: Opium interwoven within the Context for the 3rd Pandemic

The 19th-century Yunnan opium consumption likely acted as a major population-level immuno-compromising co-factor. The escalating, interconnected chain of events leads to eroded ecological and biological ZBs, which intensified the zoonotic risk:

Economic Distress & Agricultural Shift: Poverty and cash-crop cultivation created food insecurity.

Cultural-Medical Demand: Opium was sought as a cure-all and prophylactic, especially during plague outbreaks, diverting scarce resources from nutrition.

Direct Biological Impact: Chronic consumption of raw opium suppressed innate and adaptive immune functions via neuroendocrine pathways.

Synergistic Vulnerability: The resulting state of malnutrition-stress-immunosuppression created a biologically primed host population. When Yersinia pestis spilled over from enzootic reservoirs—a process facilitated by the mining-driven erosion of the rat-human barrier—individuals with opium-weakened immune systems would have been less able to mount an effective defence.

In summary, opium in Yunnan was more than a commodity or a habit; but was rather an integrated part of the societal response to health issues the people were dealing with. Its consumption sat at the intersection of economic pressure, cultural belief, and biological effect, creating a feedback loop that may have significantly amplified the deadliness of the developing Third Plague Pandemic.

3.2.3. The Incubation Phase (1840s–1850s): Localized Barrier Breaches and the Genesis of Systemic Crisis

Historical records from the 1840s and 1850s reveal the first documented erosion of ZBs in specific, ecologically pressured regions of Yunnan, concurrent with escalating ethnic strife that would culminate in the Panthay Rebellion. As summarized in

Table 2. Provincial Phasing of Zoonotic Barrier Disruption to Plague Outbreaks in 19th-Century Yunnan.

Region of Yunnan	Phase 1 (1840s–1850s)	Phase 2 (1860s–1870s)	Phase 3 (1880s–1900s)
Northeast (Dongchuan, Huize)	Local Breach: Mine collapses, valley floods, rats in homes, human disease epidemic	Compounded Shock: Warfare effects add to existing pressures	Integrated Crisis: Continued local triggers within continual high-risk ecology
Central/East (Kunming)	NA	Systemic Shock: Warfare & Major Drought (1877–1878) cause first widespread agricultural/famine-driven crises	Dominant Pattern Emerges: Drought/Famine leading to rats in Granary, then granary Spillover to local people becomes a common, repeated narrative
South (Gejiu)	NA	Warfare disruption	Industrial Epidemic Peak: Intensive mining leads to major attributed outbreaks (c. 1890)
Northwest (Dali, Dayao Co.)/West (Tengchong)	report of environmental destruction, rats enter homes	Report of famine, granaries re-opened & rats poured out, human disease followed	Widespread Narrative: Rampant rat commensalism and recurrent human epidemics.
Summary of Provincial trend	Localized, sporadic events in most ecologically pressured zones	Converging Crises: Warfare & flooding/droughts synchronize and amplify across regions	Pandemic Ecology Established: All regional narratives merge into a consistent provincial picture of rampant rat commensalism and spillover

, this period was characterized by localized, sporadic events, with the most severe breaches occurring in the northwest and northeast.

Table 2—The 1840s–1850s historical accounts indicating ZB disruption appears to start in Tengyue Subprefecture, continues east to Dayao County, and then to the Dongchuan-Huize County region, which was where most of the accounts were from in the period. The Dongchuan-Huize County region was also the only one where a human plague epidemic was recorded in the 1840s–1850s. The accounts in all regions of Yunnan progressively became more prevalent in the 1860s–1870s and 1880s–1900s periods, with human plague accounts being more common outside the Northeast of Yunnan.

Note that one historical account (No. 113: Jiaqing 25 [c. 1820; 嘉庆二十五年, in Chinese]; Zhou Xiang (Zhou Qiong, 周琼, in Chinese), 2002 (p. 46)) (See Supplementary Table S1: Historical Quotes Inventory (Yunnan, 1849–1908)), was written around 1820, but unlike the accounts from the 1840s–1850s, it did not meet convergent validation with the environmental and economic data. Thus, this earlier account was treated with low confidence.

The Tengyue-Baoshan Corridor: A Conduit for Strain 1.ORI?

The Tengyue (Tengchong)-Baoshan region, a historical hub for jade and gem trading via mule caravans to Burma [88], presents a plausible early introduction point for Y. pestis strain 1.ORI into Yunnan. As early 20th-century epidemiologist Wu Liang-te argued, plague was not autochthonous to Yunnan but was likely carried overland from Burma, where a natural plague focus borders Yunnan [89]. The robust caravan trade [90,91], which intensified after the First (1824–1826) and Second (1852) Anglo-Burmese Wars, opened Burma to British Indian commerce [1] (p. 148), created a persistent biological corridor. Caravans returning to Yunnan were laden primarily with cotton [91] (p.149)—an ideal material for transporting fleas and rodents [92]. While definitively tracing the 1.ORI strain’s passage is beyond this paper’s scope, the mid-19th-century expansion of this trade network provided a viable mechanism for the introduction and eastward spread of the novel strain from Burmese ports into Yunnan’s interior, coinciding with the first observations of abnormal rodent behaviour.

Social Fracture as an Epidemiological Determinant: The Path to the Panthay Rebellion

Concurrently, the Tengyue-Baoshan region was an epicentre of the ethnic tensions that catalysed the Panthay Rebellion (1856–1873). Han immigration surged into Yunnan, competing with Hui communities for mining and agricultural resources [93] (p. 1084). This competition turned violently systemic: a Han militia killed 1700 Hui in Mianning (1839); a three-day massacre in Baoshan (1845) killed over 8000 Muslims; and by early 1856, provincial officials sanctioned a widespread extermination campaign. On 19 May 1856, in Kunming, a provincial proclamation authorized the indiscriminate killing of Hui, resulting in a three-day massacre that claimed thousands of lives and destroyed the city’s mosques [93] (pp. 1085–1086).

The geography of this early social violence overlaps tellingly with the geography of early zoonotic events. Chuxiong and Dali—sites of pivotal early massacres and battles [93] (p. 1086)—are also among the locations where 1840s–1850s gazetteers first noted rodent invasions or human disease (See Supplementary Table S1: Historical Quotes Inventory (Yunnan, 1849–1908), Table S2: Codebook for Categories and Subcategories, Table S3: Master List of Sources). Kunming, the political heart of the attempted genocide and the rebellion, ref. [93] (p. 1086) would become a persistent hotspot of plague. This suggests that social destabilization—through massacre, flight, and siege—may have increased plague risk, degrading human “immune systems” (biological resilience via nutrition and sanitation, and community cohesion) [94,95,96] in tandem with the environmental shocks degrading ecological barriers.

Converging Pathways in the Northeast

While the northwest corridor illustrates potential pathogen introduction and the social trigger, it was in Dongchuan and Huize County in the northeast where all catalytic factors first coalesced most acutely. Here, intensive mining and environmental degradation had already fundamentally altered the local ecology. When compounded by the floods and mine collapses of the 1840s–1850s, the ZB was breached outright, independent of the widespread warfare soon to follow. This region presented the complete pre-adaptation: a landscape and rodent ecology primed for enzootic transition, awaiting only the final trigger. The subsequent province-wide rebellion would then replicate and amplify this crisis model across Yunnan, synchronizing local breaches into a systemic pandemic ecology.

Ecological Epicentre: The Copper Mining Valleys of Dongchuan and Huize County

The novel Y. pestis strain 1.ORI followed its Rattus rattus (RrC) host, a chain efficiently mediated by the flea vector Xenopsylla cheopis. This represented the first known establishment of Y. pestis in Rattus flavipectus (the dominant rat in Yunnan, a RrC host) as a primary reservoir in South/Southeast Asia [96] (p. 8200). R. flavipectus is a burrow-dwelling rodent typically inhabiting forests but is highly opportunistic and capable of migrating to human settlements.

Historical records pinpoint the copper mining valleys of northeast Yunnan as the initial flashpoint for this migration. The Gazetteer of Dongchuan Prefecture (c. 1862–1874) provides a seminal description: “All the valleys were mined for copper; trees were long felled, grasses withered and earth cracked, water sources dried up, rats bred in abundance, and at night the sound of their rushing was like tides” (谷谷皆采铜, 林木久伐, 草枯土裂, 水源涸竭, 鼠孳甚蕃, 夜闻奔走之声如潮, In Chinese) [97]. These activities to support the mines created what the Imperial Maritime Customs Medical Report, Yunnan Station (June 1888) would term “vermin-infested valleys” [98]. Deforestation for mining and charcoal production eradicated the rodents’ natural habitat and food sources, driving them into the densely populated valley settlements where they survived on human grain stores. These valleys became ideal amplification sites: concentrated rodent populations living in intimate proximity to dense human communities.

Crucially, the presence of R. flavipectus alone was insufficient to trigger plague epidemics. This is exemplified by the green, lowland hills of south-southwest Yunnan (see Figure 21, Figure 22 and Figure 23, red-circled region), where the species resided but no mid-to-late-19th-century plague outbreaks were documented. This region lacked the intensive mining and associated deforestation that characterized the epidemic zones.

The maps (Figure 21, Figure 22 and Figure 23) suggest that recorded mid-19th-century outbreaks cluster (Figure 23) in mid-elevation basins and karst/red-plateau valleys, while extremely high elevations (where Rattus flavipectus is seldom present) and the lowest, densely vegetated tropical belts show fewer records. However, this evidence alone may just reflect a pattern in the available documentation—potentially shaped by settlement and reporting. Field evidence indicates that when R. flavipectus can feed outside of buildings, indoor trap success declines: Yin et al. report that vegetables around houses and maize grown within villages reduced captures of R. flavipectus by ~45% [100] (p. 463). In contrast, the karst and red-plateau regions hosted concentrated mining and agricultural expansion driven by labour migration. As Huang Fei describes for the Southwest, the bazi valley floors—about 6% of land area yet hubs of wet-rice agriculture and market life—became the focal points where groups competed for space and resources [104] (pp. 78–80). In Yunnan gazetteers, these valleys are repeatedly associated with intensive land-use change (deforestation for charcoal, vegetation clearance, expanded grain/opium cultivation) and surges of mobile labour (See Supplementary Table S1: Historical Quotes Inventory (Yunnan, 1849–1908), Table S2: Codebook for Categories and Subcategories, Table S3: Master List of Sources).

A Kunming Prefecture gazetteer (Guangxu years) notes that “poppies were widely cultivated; consecutive droughts and floods hardened the soil; wild animals had no shelter, and rat hordes invaded villages” (光绪年间坝区多种罂粟, 连年旱涝相继, 土地板结, 野物无所栖, 鼠群入寨为患, In Chinese) [105] (“Wuchan zhi”, 物产志, In Chinese; juan 4 [卷四·物产志, In Chinese], p. 23). Taken together, these processes eroded habitat barriers (food available outdoors diminished; refuge altered) and concentrated R. flavipectus around human stores and dwellings, increasing opportunities for human–rat contact and potential spillover (see Supplementary Tables S1 and S2 for coded instances).

The synthesis of historical maps (Figure 21, Figure 22 and Figure 23) and field data suggests a barrier-erosion pathway: habitat destruction reduced food availability outdoors and eliminated refuge, forcibly concentrating R. flavipectus around human grain stores and dwellings, thereby dramatically increasing the probability of flea-borne spillover (see Supplementary Tables S1 and S2 for coded instances).

3.3. The Rodent-Host Health Threshold: From Density to Disease

An inventory of 19th-century Yunnan documentation (Figure 24) systematically categorizes the erosion of environmental barriers: (A) increased rat population density, (B) the collapse of human-rodent separation, and (C) subsequent spillover events. This historical evidence aligns with Begon et al. (2019) ecological understanding: while low rodent density reliably predicts no outbreak, high density alone is not a sufficient predictor [106]. Interestingly, however, people at the time also frequently linked rat health—or perceived threats to rat health—to the anticipation of human plague outbreaks, as a contemporary observer noted: “In the 17th year of Guangxu (1891), rats were seen fleeing the fields south of the city both morning and evening. Ten days later, a plague erupted. Elderly farmers said: ‘Rats are spiritual—they flee poison before it strikes’” (光绪十七年, 邑南鼠出田野, 晨夕皆见奔走. 旬日疫起, 老农曰: ‘鼠有灵性, 先避其毒.’, In Chinese) [107]. This observation of the rats fleeing in large numbers being a signal of upcoming plague outbreaks is commonly seen in the historic record, but what led so many from their burrows? Rat Miasma (odour) has also frequently been cited as toxic in the historic record, even the cause of plague deaths. Strong odour from rat excrement has often been cited as an indicator for high density burrows or nests [108,109]. However, Arakawa et al. (2010) observed particular odours from the excrement can signal to other rats that one or more of their cohabitants are suffering from a serious infection leading them to leave [109], or maybe a rat died and the odour from its de-composition instigated the migration. There appears to be an order of events that starts with a flood, storm, fire, drought and/or famine taking place that appears to affect the food security that leads people and rats going after the same food supply, as observed in shipping during the 1880s, “During a famine, grain was scarce. At night, people competed with rats for food at the hearth. One child died after fighting a rat over a rice ball. The plague followed soon after” [110]. This chronology of food insecurity is often followed by an erosion of the human-rat barrier, as well by rat miasma (which people must be in relatively proximity to smell), and then plague outbreaks. This is demonstrated by this example from Zhaotong (1878) that illustrates an example of drought igniting a large-scale rat migration to people’s environs, rat miasma and then a plague outbreak, “In the summer of the 4th year of Guangxu (1878), prolonged drought cracked the fields. When the granaries were opened for famine relief, warehouse rats surged out, invading markets and shops. People fell ill, calling it ‘rat-poison miasma’” (光绪四年夏, 久旱田裂, 开仓赈济, 仓鼠奔突, 入市肆, 民病疫, 谓之“鼠毒瘴”. In Chinese) [111].

Accordingly, the critical factor for plague risk from animal reservoirs does not appear to be reservoir density alone, but health and immune status of the rat population also seem crucial. These two factors may indeed be connected, but a time gap is often seen in the historical accounts that takes place between a population boom and the health/immune system consequences (that may be the result of the increased competition for food). Huise (1888) exemplified this phenomenon by observing, “In the 14th year of Guangxu (1888), great forest fires destroyed rodent nests, forcing them into rural settlements. Local people said, ‘When rats come in April, the dead cry in June.’ Three thousand perished in the outbreak” [112]. The historical record makes many references to an abundance of rats that overrun grain storage, agricultural, and home food storage areas, which may have held a sufficient supply in the beginning, but often are seen as migrating, searching for new food sources.

The historical record from 19th-century Yunnan suggests a consistent socio-ecological cascade leading to plague emergence:

Anthropogenic Pressure and Environmental Shock: Historical and gazetteer accounts document that state-promoted mining and agricultural expansion drove widespread deforestation and landscape alteration. This anthropogenic environmental change is widely recognized to degrade ecosystem resilience, which in turn would have increased the region’s vulnerability to climatic shocks. The sources report that floods, droughts, and famines—events that destabilize natural food webs—became more frequent and severe during this period.

Habitat Loss and Zoonotic Barrier Collapse: Multiple sources explicitly link this environmental degradation to wildlife displacement. Gazetteers note that the loss of forest habitat forced wildlife, including rodents, to migrate into human settlements and agricultural lands. This erosion of the environmental barrier is documented to have placed humans and wildlife in unprecedented proximity, forcing direct competition for dwindling food stores in granaries and homes.

Rodent Distress and Consequent Susceptibility: The historical evidence describes consequential signs of a rodent population in crisis: observations of environmental change and unusual rat mass migration The Gazetteer of Tengyue Subprefecture from the late Daoguang reign (Daoguang, 道光, In Chinese; c. 1850) illustrates this vividly: “During the Daoguang era, rebel troops entered the region; villages were destroyed, fields abandoned, vegetation vanished, and rats descended from the mountains into houses” (道光年间, 贼兵入境, 村落毁坏, 田畴荒芜, 草木尽失, 鼠自山中下入民舍, In Chinese) [113]. And intense infestation of human food stores. While we lack direct biological data on rodent health from the period, these behaviours are consistent with a population under severe stress from overcrowding and food insecurity. Modern plague ecology indicates that such conditions typically lead to malnutrition and compromised immune function in rodent hosts, making them more susceptible to fulminant Y. pestis infection and less capable of limiting epizootic spread [114,115]. In contrast, a stable, healthy rodent population can often resist low-dose exposures, with survivors conferring some degree of herd immunity [116,117,118].

Pathogen Spillover to Human Populations: The final, well-documented step occurred when this distressed rodent population died in close contact with human dwellings. Accounts consistently describe rats nesting in roofs, under stoves, and within walls immediately preceding human outbreaks, facilitating the transfer of infected fleas to human hosts [100,119,120,121].

This reconstructed chronology implies that epidemic risk escalated not merely from observed high rodent density, but from the likely co-occurrence of that density with poor host population health. The conditions described in Yunnan’s valleys—intense crowding, food competition, and visible distress—are precisely those that, according to modern understanding, would undermine a rodent population’s resilience, facilitating explosive epizootics and increasing the probability of human spillover.

3.3.1. The Pandemic “Critical Mass”: Amplification by the Panthay Rebellion

The mid-19th century Panthay Rebellion—a devastating conflict between the Hui Muslim community, other non-Han minorities, and the Qing state—appears to have served as a critical trigger for the Third Plague Pandemic. The rebellion did not just cause death through combat; it systematically collapsed the region’s socio-ecological defences, creating ideal conditions for the plague bacillus (Yersinia pestis) to explode from a localized zoonosis into a widespread human pandemic. French diplomat Émile Rocher, traveling through Yunnan during the conflict, witnessed a terror that surpassed the fear of combatants. He reported villagers abandoning their homes to camp on high ground, fleeing “an adversary even more inhumane than the insurgents: the plague epidemics” [121]. From the late Qianlong into the Tongzhi/Guangxu reigns, plague activity intensified province-wide, with the Xianfeng–Tongzhi period (c. 1861–1874) marking a sharp spatial expansion of outbreaks driven by refugee flows and troop movements. Contemporary Chinese sources and later epidemiological surveys consistently attribute much of the excess mortality in these years to plague and famine rather than battle itself [122,123,124,125].

This collapse of “zoonotic barriers” followed a predictable cascade of disruption. Warfare devastated agricultural production, leading to famine and a desperate reliance on centralized granaries, as the Gazetteer explained towards the end of the 19th century, “In spring of the 20th year of Guangxu (1894), as military provisions ran out, the granaries were opened. Rats poured out like tides. People in the capital fell ill with chills and fever and died within days. Pestilence spread”(光绪二十年春, 兵饷既匮, 开仓出粟. 鼠涌出如潮, 都中人病寒热, 数日卽毙, 疫气流衍, In Chinese) [126]. This intensified pressure on already eroded zoonotic barriers, as a record from Xuanwei in 1868 describes the same cascade: “In the 7th year of the Tongzhi reign (1868), after military chaos, the granaries within the city collapsed, and rat burrows spread beneath them. When the rainy season arrived, houses collapsed and flooded. An epidemic broke out beside the granaries, widely attributed to ‘rat miasma’” (同治七年, 军乱过后, 城中仓廪倾圮, 鼠穴蔓延其下. 及雨季至, 屋塌水浸, 疫起仓侧, 众以为鼠瘴所致, In Chinese) [127]. Food insecurity forced dependency upon the granaries that collapsed, fostering grave nutritional risks for both humans and commensal rats—which had long adapted to human settlements—competed for the same dwindling food stores within these granaries. Meanwhile, the broader economy collapsed, copper currency devalued, unemployment soared, the risk of injuries grew in often increasingly unsanitary conditions, while banditry and the opium economy spread, spurring large-scale migration of impoverished, malnourished, and often diseased populations.

Thus, waves of vulnerable people, intimately exposed to plague-infected rat fleas in their homes and granaries, and often further weakened by opium, heavy metal contamination, and malnutrition, carried the disease with them as they fled (See Supplementary Table S1: Historical Quotes Inventory (Yunnan, 1849–1908), Table S2: Codebook for Categories and Subcategories, Table S3: Master List of Sources [102,128,129]. This pattern repeated in areas such as Yongshan in 1885, where post-war resettlement occurred amid lingering rat colonies and sparse vegetation, leading to outbreaks of illness that local physicians attributed to the “rat-burrow poisonous wind” (shuxue dufeng, 鼠穴毒风, In Chinese) [130]. Accordingly, the Panthay Rebellion, as a large-scale human conflict, created a socio-economic crisis that triggered a domino effect through amplifying the erosion of ZBs during an already present zoonosis and lost containment with pandemic-scale consequences.

3.3.2. National Context: Systemic Unsustainability in the Late Qing

The Panthay Rebellion was not an isolated event but part of a devastating wave of mid-19th century rebellions that crippled China. These included, most catastrophically, the Taiping Rebellion (Figure 25), which originated in Guangxi province bordering Yunnan and resulted in tens of millions of deaths [131]. These rebellions may be seen as regional networks of varying degrees of connection that were often tied to banditry and opium trade, which appears to often be the main funding source, while often attacking mines, a major funding source for the Qing Dynasty [132]. These opium trade networks largely worked in the mountain regions, which is also where the rebellions often would begin [132]. The extreme poverty most labourers found themselves in may have assisted the level of support the rebellions received.

This period of unprecedented violence caused a massive decline in a population that had quadrupled over the previous 150 years, leading to the abandonment of farmland and a collapse in agricultural cultivation. It appears the rapid population growth triggered an increase in large-scale deforestation and land clearance for agriculture [129]. If this is indeed the case, this would have two critical consequences: first, it significantly increased the frequency and severity of floods and droughts by removing natural vegetation that stabilized soils and watersheds. Scholars such as Lee and Zhang (2013) [132] have demonstrated a strong correlation between flooding and drought and epidemic outbreaks during the Qing Dynasty. Second, this habitat destruction forced animal reservoir hosts, like the plague-carrying rodent Rattus flavipectus, to migrate into human settlements in search of food and shelter, drastically increasing human-wildlife contact.

The Third Plague Pandemic emerged within an economic context that forced populations into human activity that helped create unsustainable environmental practices. According to Laybourn-Langton and Hill (2019), such “environmental breakdown” is a consequence of socioeconomic systems driven by unsustainable resource use [133] p. 4. This pattern, which intensifies in the Anthropocene, erodes the barriers between humans and animal diseases, increasing the risk of zoonotic epidemics [134]. In mid-19th century China, this systemic unsustainability was starkly visible. Although China’s GDP represented the world’s largest at 33% of global output in 1820, a century of internal strife, including the Panthay and Taiping rebellions, cut its relative economic size to half that of Western Europe by 1870 [34,135,136]. This catastrophic decline shattered the state’s capacity to manage disasters and maintain public health. As historian Carol Benedict (1988) has argued, the resulting vacuum—filled by poverty, mass migration, and a destabilizing opium trade—created the perfect pathways for the plague to spread from its epicentre in the war-torn and environmentally degraded province of Yunnan [137].

Figure 25. Map of Qing China and the Taiping Heavenly Kingdom captured territories (M. Bitton 2022) [138]. Tan shaded areas had various occurrences of the rebellion, which started in Guangxi, the orange areas indicate early lands of the rebellion, and the brown areas are late territories of the rebellion. Note. This map, created by M. Bitton (2022), is a derivative work based on Peng (2021), Reilly [Year of The Taiping Heavenly Kingdom…], and the base map “China 1820 de.svg.” Reprinted from Wikimedia Commons, by M. Bitton, 2022 [138]. (https://upload.wikimedia.org/wikipedia/commons/2/23/Taiping_Heavenly_Kingdom_map.svg) (accessed on 26 February 2026).

Figure 25. Map of Qing China and the Taiping Heavenly Kingdom captured territories (M. Bitton 2022) [138]. Tan shaded areas had various occurrences of the rebellion, which started in Guangxi, the orange areas indicate early lands of the rebellion, and the brown areas are late territories of the rebellion. Note. This map, created by M. Bitton (2022), is a derivative work based on Peng (2021), Reilly [Year of The Taiping Heavenly Kingdom…], and the base map “China 1820 de.svg.” Reprinted from Wikimedia Commons, by M. Bitton, 2022 [138]. (https://upload.wikimedia.org/wikipedia/commons/2/23/Taiping_Heavenly_Kingdom_map.svg) (accessed on 26 February 2026).

4. Discussion: Triangulating Pandemic Origins—A Dynamic One Health Synthesis

4.1. The Interdisciplinary Imperative: Building a Case from Convergent Evidence

Understanding pandemic emergence is an exercise in historical reconstruction, akin to building a case from circumstantial evidence. No single field—be it genomics, ecology, or history—holds the definitive proof. The strength of the argument instead relies on convergent validation, where independent lines of evidence from disparate disciplines align to support a coherent narrative. This study employs such an interdisciplinary, spatiotemporal One Health framework to analyse the Third Plague Pandemic’s genesis in Yunnan. We move beyond a static model of health to trace how the dynamic co-evolution of the environment, animal hosts, human society, and pathogen biology over decades created the necessary conditions for spillover.

4.2. The Epicentre Revealed: Dongchuan-Huize as the Probable Crucible of Emergence

Our analysis points to the copper mining districts of Dongchuan and Huize County as the probable initial epicentre for the pandemic. This inference is supported not by a single source, but by the convergence of multiple, independent lines of evidence that collectively indicate a uniquely high-risk socio-ecological configuration:

Historical Documentation: Contemporary gazetteers from this region provide the earliest and most coherent accounts of a full breach sequence, explicitly correlating mining-driven deforestation with rodent population irruptions into settlements and subsequent human epidemic mortality.

Ecological Preconditioning: The extractive economy fundamentally altered the local ecology. Historical gazetteers explicitly report that mining-driven deforestation eroded the environmental barrier, displacing Rattus flavipectus from degraded forests into densely populated valley settlements. This created conditions conducive to epizootic amplification, as the historical record suggests high-density, nutritionally stressed rodent colonies in intimate contact with human habitations. The resultant landscape concentrated both rodent and human populations into a confined, resource-stressed environment, significantly increasing the frequency of contact and the potential for cross-species transmission.

Compromised Human Host Resilience: Critically, this landscape also fostered conditions that likely degraded population-level host defences. The economic structure of mining, particularly a declining copper-to-silver wage ratio, led to a significant reduction in real wages for labourers, directly constraining their access to adequate nutrition. This created a populace vulnerable to malnutrition, a condition well-documented to compromise immune function. Concurrently, the historic record indicates the widespread cultivation and use of opium—driven both by its economic role and by its contemporary perception as a prophylactic and treatment for disease, including plague symptoms—introduced a substance with documented immunosuppressive effects. Furthermore, the enduring environmental legacy of unregulated 19th-century smelting—evidenced by persistent heavy metal (as Copper) contamination in modern soils—suggests a plausible historical pathway for chronic immunotoxic exposure via local water and food sources. These factors—economic hardship, medicinal-opium use, and environmental toxicity—would have synergistically eroded the biological barriers to infection.

Pathogen Introduction and Fit: The eventual introduction of the highly transmissible Y. pestis 1.ORI strain into this preconditioned system likely served as the means for an outbreak or a local epidemic to turn to a pandemic. The region’s integration into provincial and transnational trade networks via the northeastern mining corridor presents a plausible route for such pathogen introduction or heightened connectivity.

This convergent evidence suggests the pandemic’s emergence was endogenous to processes within Yunnan’s transforming landscape. Rather than merely receiving an introduced pathogen, the region appears to have generated the synergistic conditions-a perturbed reservoir host ecology and a human population with diminished biological resilience-that enabled a zoonotic agent to transition to epidemic spread.

4.3. The Provincial Amplifier: The Panthay Rebellion’s Systemic Shock

The Panthay Rebellion (1856–1873) did not create the initial epicentre but acted as a critical province-wide amplifier and synchronizer. By collapsing agriculture, destroying granaries, and triggering mass migrations of malnourished refugees, the rebellion systematically replicated the high-risk Dongchuan model across Yunnan. It transformed a series of localized outbreaks into a generalized pandemic by simultaneously degrading ZBs (through habitat destruction and famine) and biological barriers (through human impoverishment and immune compromise) on a provincial scale.

4.4. Linkages, Limitations, and Future Directions

We observed the pattern of:

Demographic density increases in valley settlement areas

Regional economic streamlining (overdependence) on mining

Anthropogenic environmental change

Host habitat loss & population stress

Compromised host immunity

Pathogen spillover and amplification

This pattern also resonates with modern frameworks for zoonotic risk. It emphasizes that spillover is not a simple function of pathogen presence but requires the breakdown of multiple interdependent barriers separating reservoir hosts from susceptible human populations.

We explicitly acknowledge the limitations inherent in this historical reconstruction. Retrospective diagnosis relies on descriptive historical sources. While our representation of distressed contemporary Yunnan residents is supported in primary historical documentation, rodent immunocompromise is supported by documented historical observations, ecological principles and consequential evidence (e.g., distress migrations), direct biological proof is lacking, as it typically is in centuries old historical investigations. Additionally, the precise entry route of the 1.ORI strain is just a hypothesis and requires further investigation.

These limitations define a clear agenda for future research:

Epicenter Validation: Multi-proxy studies in Dongchuan-Huize, combining palaeoecology, bioarchaeology (e.g., Y. pestis aDNA from remains), and granular historical assessment of cultural, economic, and educational developments, would increase the confidence of our hypothesis as Dongchuan-Huize County as the first location of the 3rd Plague Pandemic.

Pathway Tracing: Research in Myanmar should investigate the caravan trade hypothesis, seeking historical and archaeological evidence linking colonial-era land use (e.g., cotton, tea) to rodent ecology and potential pathogen movement.

Origin Refinement: Phylogeographic and historical ecological work in South Asia is crucial to pinpoint the 1.ORI host-shift from Indian gerbil to Rattus and its connection to colonial commodity production. This agricultural production would also need to be reached to link Rattus host transmission through trade to Burma and continuing to Yunnan.

Expansion Dynamics: Studies beyond Yunnan should analyse how transmission dynamics evolved during global spread, integrating genomics and historical epidemiology to understand shifts between rural zoonotic and urban pneumonic spread.

5. Conclusions: Towards a Model for Pandemic Reconstruction

This study demonstrates that a dynamic, interdisciplinary One Health approach is not merely beneficial but essential for what we term “pandemic reconstruction,” which builds upon Bendrey et al. [4] interdisciplinary One Health assessment through triangulation. By triangulating genetic, historical, and ecological evidence, we move beyond isolated speculation to identify the specific socio-ecological conditions that transformed a zoonotic pathogen into a global catastrophe. The Dongchuan-Huize epicentre exemplifies how unsustainable resource extraction—by eroding multiple ZBs simultaneously—can create a receptacle for disease emergence.

This historical insight underscores an enduring principle: pandemic prevention lies not only in pathogen surveillance, but in sustaining the ecological and social integrity of the landscapes we share with animal reservoirs. As Rayfield et al. [5] has highlighted using the One Health approach in assessing the past, the anthropogenic impact often is instrumental and transformative in human zoonotic risk. However, little One Health assessment focus has been put in what leads to these transformative anthropogenic environmental changes. This appears to need more emphasis, as the One Health Centre at the University of Alaska Fairbanks, which emphasizes resilience and adaptation and individual and community well-being (https://www.uaf.edu/onehealth/) (24 March 2026) [20]. If environmental change is disproportionately driven by human activity, then community health and environmental health are not separate concerns but mutually constituting. A community’s health—including the knowledge systems, values, and institutions that shape its decisions—directly influences the socio-economic structures that, in turn, affect environmental, plant, and animal health. Diagnosing what leads to unhealthy outcomes must therefore include an examination of these underlying drivers.

Crucially, this approach must be applied dynamically across time and space, recognising that the relevant temporal scales for biological, environmental, and socio-economic change are not uniform. The evolution of a pathogen, the slow erosion of soils, and the rapid shift of an economic policy each operate on different clocks and across different geographies. A practical One Health framework must therefore embrace this multi-scalar complexity, tracing how interconnected domains of health and environment co-evolve to shape pandemic risk.

By integrating the study of past pandemics with the analysis of present-day zoonotic threats, we can extract lessons that not only deepen our understanding of outbreak dynamics but also inform more resilient and equitable pathways forward. The past, properly interrogated, is not merely prologue—it is evidence.