Biological invasions facilitate zoonotic disease emergences

The rapid increase in zoonotic diseases, caused by pathogens that spill over from animals to humans, threatens global economy, public health, and social stability. Alien animal host invasions have long been suspected to increase zoonosis risk by boosting existing pathogens and introducing novel ones, especially amid recent rises in both alien introductions and zoonoses such as avian influenza, SARS, and COVID-19. Human activities (e.g., pet trade, aquaculture) often accompany alien introductions, elevating human–host contact opportunities. Numerous case reports link alien hosts to zoonoses (e.g., invasive rats with plague and leptospirosis; raccoons with West Nile virus; alien mosquitoes with chikungunya and dengue; alien lice and fleas with typhus and plague), but no prior global synthesis has quantified established alien hosts’ roles across broad taxa. Because alien invasions and zoonosis emergences share cofactors—propagule pressure, climate change, biodiversity loss, land-use change, human population density, and surveillance/research biases—few studies have isolated alien hosts’ unique contributions. This study compiles 10,473 zoonosis events since 1348 and quantifies the effect of 795 established alien hosts on zoonosis events worldwide, controlling for confounders including non-zoonotic alien host richness (as a propagule pressure control), climate and land-use change, biodiversity loss, native biodiversity, surveillance capacity, reporting bias, spatial autocorrelation, and lack of independence via random effects. Additional analyses assess which alien taxonomic groups contribute most, examine relationships through time to address possible temporal mismatches, and map locations where historical zoonoses were most influenced by alien host introductions. After controls, zoonosis events increased with alien zoonotic host richness across mammalian (notably Artiodactyla, Carnivora, Rodentia), avian (waterfowl, Galliformes, Passeriformes), and Dipteran invertebrate hosts, both spatially and temporally, whereas alien non-zoonotic host richness showed no significant effect.

The paper synthesizes prior evidence linking alien species to zoonoses, citing examples across taxa and regions: invasive rats associated with plague, murine typhus, scrub typhus, leptospirosis, and hantavirus hemorrhagic fever; introduced raccoons in Europe with West Nile virus and human roundworm infections; alien mosquitoes with chikungunya and dengue in Europe; and alien lice and fleas linked to typhus and plague. It highlights the longstanding recognition that invasions and zoonoses correlate with propagule pressure, climate change, biodiversity loss, land-use change, human density, and surveillance biases. Despite these insights and growing introductions and disease emergences, a comprehensive, multi-taxa, global-scale assessment controlling for these cofactors had been lacking. The study builds on frameworks quantifying zoonotic risk from host traits and environmental drivers and adapts global datasets (GIDEON, CLOVER) previously used in zoonosis research.

Study design and scope: Global analysis at the administrative-unit level (201 jurisdictions) using GIDEON to compile 10,473 events for 161 zoonoses (nonhuman zoonotic and multihost diseases) from 1348–2020. Zoonosis events include both emerging (first-time in unit) and re-emerging events. Events were categorized by host taxonomic groups (mammals, birds, invertebrates, reptiles, amphibians) and pathogen types (bacteria, viruses, parasitic animals, fungi). Counts were converted to densities by dividing by jurisdictional area to control for area effects. To adjust for surveillance intensity, the dependent variable was the residuals from a regression of zoonosis event density against all disease event density. Alien host data: Compiled 4,522 establishment events for 795 established alien animals across mammals, birds, herpetofauna, and invertebrates from multiple sources (Kraus compendium and updates for reptiles/amphibians; GAVIA for birds; Introduced Mammals of the World and updates for mammals; GISD for invertebrates). For each alien species, zoonotic status (zoonotic vs non-zoonotic host) was determined via an intensive literature review and the CLOVER dataset integrating GMPD2, EID2, HP3, and Shaw databases. Alien host richness (zoonotic and non-zoonotic) was computed per jurisdiction and by taxonomic order/group. Taxa analyzed in depth included mammalian orders Carnivora, Cetartiodactyla (Artiodactyla), Lagomorpha, Rodentia; avian groups (waterfowl encompassing Anseriformes, Gruiformes, Pelecaniformes, Phoenicopteriformes, Suliformes; Columbiformes; Galliformes; Passeriformes; Psittaciformes); Diptera among invertebrates; and herpetofauna combined—each with alien populations in ≥50 administrative units. Covariates and controls: Environmental and global change covariates included Global Environmental Stratification (GEnS) as a composite bioclimate variable; climate change quantified as slopes of temperature and precipitation from 1901–2009 (CRU 0.5° grids) averaged per jurisdiction; land-use modification quantified as the proportion of grids transitioning to more anthropogenic classes between 1900 and 2000 (Anthromes v2, 15 transition scenarios); biodiversity loss measured as counts of threatened species (NT, VU, EN from IUCN Red List) per taxon per unit; native vertebrate richness (amphibians, birds, mammals, reptiles) from global diversity maps; human population density (World Bank) as a proxy for propagule pressure and human-assisted pathogen movements. Sampling effort and bias were controlled using the Infectious Disease Vulnerability Index (IDVI), PubMed citations per disease per jurisdiction (automated via PubCrawler), and spatial autocorrelation via centroid latitude/longitude. Statistical analyses: Generalized additive mixed models (GAMMs) were used with the surveillance-corrected zoonosis event density residuals as the response. The full model included 13 smoothed fixed effects: GEnS, human population density, native species richness density, biodiversity loss, alien zoonotic host richness density, alien non-zoonotic host richness density, temperature change, precipitation change, land-use change, IDVI, PubMed citations, longitude, latitude. Random intercepts controlled for lack of independence: continent (human settlement history), pathogen identity, and host order. Thin-plate spline smoothers with 6, 8, and 10 knots were explored (primary results shown for 10 knots). Model selection used AIC across all combinations of predictors (2^13 models), with ΔAIC ≤ 2 indicating highly supported models. Significance was inferred when 95% CIs of standardized coefficients did not overlap zero. The proportion of deviance explained by each variable was estimated by comparing models with and without the variable. Interactions between alien zoonotic host richness and host order were included to identify taxonomic groups with stronger associations. Spatial mapping was performed by predicting zoonosis events with and without alien zoonotic host variables and mapping the difference (average and 95% CI bounds) per administrative unit to infer contribution of alien hosts to historical zoonoses. Temporal analyses: Compiled introduction years for alien zoonotic and non-zoonotic hosts (birds: GAVIA and literature; mammals: Introduced Mammals; invertebrates: GISD). Annual counts of new zoonosis events were extracted from GIDEON. Multiple regression/GAM treated year as replicate, with counts of alien zoonotic and non-zoonotic introductions as predictors of annual zoonosis events. Breakpoint regression (segmented package) identified years of sharp increases (left-horizontal and two-slope models; AIC-based breakpoint selection) for zoonosis events and alien host introductions. Data were pooled across host taxa for temporal analyses due to limited timing data in some orders. Data and code: Data sources are publicly described (GIDEON, GADM, CLOVER, CRU, Anthromes v2, World Bank, Biodiversity Mapping, IUCN Red List, GISD; details and access links provided). R code provided in Supplementary Notes.

• At least 35.6% (283/795) of established alien animal species are zoonotic hosts. There is an average of approximately 5.9 (±0.58 SE) zoonoses per alien zoonotic host.
• Zoonosis events (10,473 events for 161 zoonoses, 1348–2020) are most associated with mammalian hosts, followed by avian, invertebrate, and herpetofaunal hosts; geographically concentrated in Europe (3,495), Asia (2,180), North America (2,120), Africa (1,301), South America (897), and Oceania (480); and primarily caused by bacteria (4,622) and viruses (4,002), then parasitic animals (1,621) and fungi (225).
• Spatial GAMM results: Across the top AIC-supported models, six predictors consistently had positive, significant effects with 95% CIs not overlapping zero: alien zoonotic host richness, human population density, biodiversity loss, temperature change, land-use modification, and PubMed citations. Latitude was also significant in some top models, with higher zoonosis event density at higher latitudes.
• Alien non-zoonotic host richness (propagule pressure control) did not have a significant effect, indicating the association with alien zoonotic hosts is not merely due to general propagule pressure.
• Taxonomic contributions: Stronger positive associations between alien zoonotic host richness and zoonosis events for mammalian orders Carnivora (0.0150 ± 0.0028, P<0.001), Artiodactyla (0.011 ± 0.0023, P<0.001), and Rodentia (0.0197 ± 0.0021, P<0.001); avian groups waterfowl (0.0149 ± 0.0022, P<0.001), Galliformes (0.0059 ± 0.0024, P=0.0134), and Passeriformes (0.0082 ± 0.0019, P<0.001); invertebrate order Diptera (0.0151 ± 0.0053, P<0.01). No significant effects detected for Columbiformes, Lagomorpha, Psittaciformes, amphibians, or reptiles.
• Temporal analyses: Annual zoonosis events positively correlated with annual introductions of alien zoonotic hosts (Estimate = 3.04, 95% CI 2.23–3.85, P<0.001), but not with alien non-zoonotic host introductions (Estimate = 0.62, 95% CI −0.69–1.94, P=0.345). Breakpoints: alien zoonotic host introductions increased sharply in 1960; zoonosis events increased sharply in 1962; non-zoonotic host introductions breakpoint was 1948, not coincident with zoonosis breakpoint.
• Mapping: Predicted contributions of alien zoonotic hosts to historical zoonoses were highest in Europe, Oceania, and the Caribbean islands, aligning with global hotspots of alien animal establishment.

The study provides global, spatial, temporal, and multi-taxa evidence that established alien animal species facilitate zoonosis emergences. By controlling for propagule pressure via non-zoonotic alien host richness and numerous environmental, demographic, and sampling biases, the observed positive association between alien zoonotic host richness and zoonosis events likely reflects a causal contribution of alien hosts to zoonotic emergence. The strong effects in specific mammalian orders (Carnivora, Artiodactyla, Rodentia) and avian groups (waterfowl, Galliformes, Passeriformes), and in Diptera, align with known biology: close human association, phylogenetic relatedness to humans, use of human-dominated habitats, and efficient vector capacities all increase spillover opportunities. The findings that human population density, land-use change, climate warming, and biodiversity loss are positively associated with zoonosis events underscore how global change processes amplify spillover risk by increasing contact rates, expanding vector and pathogen ranges, and facilitating establishment of alien and native reservoirs. The significance of PubMed citations confirms the necessity of accounting for research effort to interpret spatial patterns. Temporal concordance between rises in alien zoonotic host introductions and zoonosis events further supports alien invasions’ role in recent zoonotic increases. The global map of alien-host-related zoonosis contributions highlights regions where surveillance and biosecurity could be prioritized to mitigate future risks.

This study delivers the first comprehensive global assessment linking established alien animal invasions to zoonotic disease emergences across taxa, space, and time. It shows that zoonosis events increase with alien zoonotic host richness independent of propagule pressure and that specific mammalian, avian, and Dipteran groups are key contributors. Zoonosis events are also positively associated with human population density, land-use modification, climate warming, and biodiversity loss. These insights inform policy by emphasizing the health benefits of preventing and managing introductions of zoonotic host species, strengthening biosecurity and surveillance—particularly in identified hotspots—and addressing global change drivers that elevate zoonotic risk. Future research should expand host-pathogen association data (especially for under-studied taxa and regions), incorporate native invertebrate diversity, refine temporal linking of introduction and outbreak timelines at finer scales, and evaluate mechanistic pathways by which alien hosts promote spillover.

• Underestimation of zoonotic alien hosts is likely due to conservative classification criteria and incomplete host-pathogen knowledge; taxonomic and geographic sampling biases persist despite controls.
• Temporal mismatch cannot be entirely excluded, though analyses using introduction timing and breakpoint tests mitigate this concern.
• Native invertebrate richness was not included due to data limitations, potentially omitting relevant biodiversity effects.
• Some host orders (e.g., Chiroptera, Primates) were excluded from order-level analyses because too few alien species exist, limiting taxonomic generality.
• Surveillance and reporting biases remain, though proxies (IDVI, PubMed citations, overall disease event adjustment) were incorporated.
• Administrative-unit aggregation and area-standardization may mask subnational heterogeneity and local processes.