Land use change—for example, the conversion of natural habitats to agricultural or urban ecosystems—is widely recognized to influence the risk and emergence of zoonotic disease in humans1,2. However, whether such changes in risk are underpinned by predictable ecological changes remains unclear. It has been suggested that habitat disturbance might cause predictable changes in the local diversity and taxonomic composition of potential reservoir hosts, owing to systematic, trait-mediated differences in species resilience to human pressures3,4. Here we analyse 6,801 ecological assemblages and 376 host species worldwide, controlling for research effort, and show that land use has global and systematic effects on local zoonotic host communities. Known wildlife hosts of human-shared pathogens and parasites overall comprise a greater proportion of local species richness (18–72% higher) and total abundance (21–144% higher) in sites under substantial human use (secondary, agricultural and urban ecosystems) compared with nearby undisturbed habitats. The magnitude of this effect varies taxonomically and is strongest for rodent, bat and passerine bird zoonotic host species, which may be one factor that underpins the global importance of these taxa as zoonotic reservoirs. We further show that mammal species that harbour more pathogens overall (either human-shared or non-human-shared) are more likely to occur in human-managed ecosystems, suggesting that these trends may be mediated by ecological or life-history traits that influence both host status and tolerance to human disturbance5,6. Our results suggest that global changes in the mode and the intensity of land use are creating expanding hazardous interfaces between people, livestock and wildlife reservoirs of zoonotic disease.
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
All code for this study is available at Figshare https://doi.org/10.6084/m9.figshare.7624289.
Myers, S. S. et al. Human health impacts of ecosystem alteration. Proc. Natl Acad. Sci. USA 110, 18753–18760 (2013).
Gottdenker, N. L., Streicker, D. G., Faust, C. L. & Carroll, C. R. Anthropogenic land use change and infectious diseases: a review of the evidence. EcoHealth 11, 619–632 (2014).
Keesing, F. et al. Impacts of biodiversity on the emergence and transmission of infectious diseases. Nature 468, 647–652 (2010).
Ostfeld, R. S. & LoGiudice, K. Community disassembly, biodiversity loss, and the erosion of an ecosystem service. Ecology 84, 1421–1427 (2003).
Johnson, P. T. J. et al. Living fast and dying of infection: host life history drives interspecific variation in infection and disease risk. Ecol. Lett. 15, 235–242 (2012).
Johnson, P. T. J., Preston, D. L., Hoverman, J. T. & Richgels, K. L. D. Biodiversity decreases disease through predictable changes in host community competence. Nature 494, 230–233 (2013).
Newbold, T. et al. Global effects of land use on local terrestrial biodiversity. Nature 520, 45–50 (2015).
Newbold, T. et al. Widespread winners and narrow-ranged losers: land use homogenizes biodiversity in local assemblages worldwide. PLoS Biol. 16, e2006841 (2018).
Faust, C. L. et al. Pathogen spillover during land conversion. Ecol. Lett. 21, 471–483 (2018).
Becker, D. J., Streicker, D. G. & Altizer, S. Using host species traits to understand the consequences of resource provisioning for host-parasite interactions. J. Anim. Ecol. 87, 511–525 (2018).
Plowright, R. K. et al. Pathways to zoonotic spillover. Nat. Rev. Microbiol. 15, 502–510 (2017).
Shah, H. A., Huxley, P., Elmes, J. & Murray, K. A. Agricultural land-uses consistently exacerbate infectious disease risks in Southeast Asia. Nat. Commun. 10, 4299 (2019).
Gottdenker, N. L., Chaves, L. F., Calzada, J. E., Saldaña, A. & Carroll, C. R. Host life history strategy, species diversity, and habitat influence Trypanosoma cruzi vector infection in changing landscapes. PLoS Negl. Trop. Dis. 6, e1884 (2012).
Fornace, K. M. et al. Association between landscape factors and spatial patterns of Plasmodium knowlesi infections in Sabah, Malaysia. Emerg. Infect. Dis. 22, 201–209 (2016).
Pulliam, J. R. C. et al. Agricultural intensification, priming for persistence and the emergence of Nipah virus: a lethal bat-borne zoonosis. J. R. Soc. Interface 9, 89–101 (2012).
Kilpatrick, A. M. Globalization, land use, and the invasion of West Nile virus. Science 334, 323–327 (2011).
Popp, A. et al. Land-use futures in the shared socio-economic pathways. Glob. Environ. Change 42, 331–345 (2017).
Civitello, D. J. et al. Biodiversity inhibits parasites: broad evidence for the dilution effect. Proc. Natl Acad. Sci. USA 112, 8667–8671 (2015).
LoGiudice, K., Ostfeld, R. S., Schmidt, K. A. & Keesing, F. The ecology of infectious disease: effects of host diversity and community composition on Lyme disease risk. Proc. Natl Acad. Sci. USA 100, 567–571 (2003).
Purvis, A., Gittleman, J. L., Cowlishaw, G. & Mace, G. M. Predicting extinction risk in declining species. Proc. Biol. Soc 267, 1947–1952 (2000).
Johnson, P. T. J., Ostfeld, R. S. & Keesing, F. Frontiers in research on biodiversity and disease. Ecol. Lett. 18, 1119–1133 (2015).
Olival, K. J. et al. Host and viral traits predict zoonotic spillover from mammals. Nature 546, 646–650 (2017).
Young, H., Griffin, R. H., Wood, C. L. & Nunn, C. L. Does habitat disturbance increase infectious disease risk for primates? Ecol. Lett. 16, 656–663 (2013).
Hudson, L. N. et al. The database of the PREDICTS (Projecting Responses of Ecological Diversity In Changing Terrestrial Systems) project. Ecol. Evol. 7, 145–188 (2017).
Lloyd-Smith, J. O. et al. Epidemic dynamics at the human–animal interface. Science 326, 1362–1367 (2009).
Joseph, M. B., Mihaljevic, J. R., Orlofske, S. A. & Paull, S. H. Does life history mediate changing disease risk when communities disassemble? Ecol. Lett. 16, 1405–1412 (2013).
Kamiya, T., O’Dwyer, K., Nakagawa, S. & Poulin, R. What determines species richness of parasitic organisms? A meta-analysis across animal, plant and fungal hosts. Biol. Rev. Camb. Philos. Soc. 89, 123–134 (2014).
Lee, K. A., Wikelski, M., Robinson, W. D., Robinson, T. R. & Klasing, K. C. Constitutive immune defences correlate with life-history variables in tropical birds. J. Anim. Ecol. 77, 356–363 (2008).
Rohr, J. R. et al. Towards common ground in the biodiversity–disease debate. Nat. Ecol. Evol. 4, 24–33 (2020).
Hosseini, P. et al. Does the impact of biodiversity differ between emerging and endemic pathogens? The need to separate the concepts of hazard and risk. Philos. Trans. R. Soc. Lond. B 372, 20160129 (2017).
Brearley, G. et al. Wildlife disease prevalence in human-modified landscapes. Biol. Rev. Camb. Philos. Soc. 88, 427–442 (2013).
Burkett-Cadena, N. D. & Vittor, A. Y. Deforestation and vector-borne disease: Forest conversion favors important mosquito vectors of human pathogens. Basic Appl. Ecol. 26, 101–110 (2018).
Hassell, J. M., Begon, M., Ward, M. J. & Fèvre, E. M. Urbanization and disease emergence: dynamics at the wildlife–livestock-human interface. Trends Ecol. Evol. 32, 55–67 (2017).
Holmes, E. C., Rambaut, A. & Andersen, K. G. Pandemics: spend on surveillance, not prediction. Nature 558, 180–182 (2018).
Wardeh, M. et al. Database of host-pathogen and related species interactions, and their global distribution. Sci. Data 2, 150049 (2015).
Stephens, P. R. et al. Global Mammal Parasite Database version 2.0. Ecology 98, 1476 (2017).
Plourde, B. T. et al. Are disease reservoirs special? Taxonomic and life history characteristics. PLoS ONE 12, e0180716 (2017).
Han, B. A., Schmidt, J. P., Bowden, S. E. & Drake, J. M. Rodent reservoirs of future zoonotic diseases. Proc. Natl Acad. Sci. USA 112, 7039–7044 (2015).
Chamberlain, S. & Szocs, E. Taxize: taxonomic search and retrival in R. F1000Res 18, 191 (2013).
Wertheim, H. F. L., Horby, P. & Woodall, J. P. Atlas of Human Infectious Diseases (Wiley-Blackwell, 2012).
Taylor, L. H., Latham, S. M. & Woolhouse, M. Risk factors for human disease emergence. Philos. Trans. R. Soc. B 356, 983–989 (2001).
Han, B. A., Kramer, A. M. & Drake, J. M. Global patterns of zoonotic disease in mammals. Trends Parasitol. 32, 565–577 (2016).
Røttingen, J. A. et al. New vaccines against epidemic infectious diseases. N. Engl. J. Med. 376, 610–613 (2017).
Harris, N. C. & Dunn, R. R. Using host associations to predict spatial patterns in the species richness of the parasites of North American carnivores. Ecol. Lett. 13, 1411–1418 (2010).
Cooper, N. et al. Phylogenetic host specificity and understanding parasite sharing in primates. Ecol. Lett. 15, 1370–1377 (2012).
Young, H. S. et al. Interacting effects of land use and climate on rodent-borne pathogens in central Kenya. Philos. Trans. R. Soc. B 372, 20160116 (2017).
Newbold, T. et al. Global patterns of terrestrial assemblage turnover within and among land uses. Ecography 39, 1151–1163 (2016).
Cooper, N., Kamilar, J. M. & Nunn, C. L. Host longevity and parasite species richness in mammals. PLoS ONE 7, e42190 (2012).
González-Suárez, M., Lucas, P. M. & Revilla, E. Biases in comparative analyses of extinction risk: mind the gap. J. Anim. Ecol. 81, 1211–1222 (2012).
Ducatez, S. & Lefebvre, L. Patterns of research effort in birds. PLoS ONE 9, e89955 (2014).
Paull, S. H. et al. From superspreaders to disease hotspots: linking transmission across hosts and space. Front. Ecol. Environ. 10, 75–82 (2012).
Blangiardo, M., Cameletti, M., Baio, G. & Rue, H. Spatial and spatio-temporal models with R-INLA. Spat. Spatio-Temporal Epidemiol. 4, 33–49 (2013).
Hooten, M. B. & Hobbs, N. T. A guide to Bayesian model selection for ecologists. Ecol. Monogr. 85, 3–28 (2015).
Zuur, A. F., Ieno, E. N., Walker, N., Saveliev, A. A. & Smith, G. M. Mixed Effects Models and Extensions in Ecology with R (Springer, 2010).
Dallas, T. A. et al. Host traits associated with species roles in parasite sharing networks. Oikos 128, 23–32 (2019).
R: A language and environment for statistical computing. http://www.R-project.org/ (R Foundation for Statistical Computing, 2019).
We thank L. Enright, A. Etard, L. Franklinos, R. Freeman, R. Lowe and R. Pearson for discussion on previous versions of the manuscript. This research was supported by a University College London Graduate Research Scholarship (R.G.); the Ecosystem Services for Poverty Alleviation Programme, Dynamic Drivers of Disease in Africa Consortium, NERC project no. NE-J001570-1 (D.W.R. and K.E.J.); an MRC UKRI/Rutherford Fellowship (MR/R02491X/1) and Wellcome Trust Institutional Strategic Support Fund (204841/Z/16/Z) (both to D.W.R.); and a Royal Society University Research Fellowship (T.N.). C.A.D. thanks the UK MRC and DFID for Centre funding (MR/R015600/1), and the UK National Institute for Health Research Health Protection Research Unit in Modelling Methodology at Imperial College London in partnership with Public Health England for funding (grant HPRU-2012–10080).
The authors declare no competing interests.
Peer review information Nature thanks Noam Ross and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
Extended Data Fig. 1 Conceptual framework for the effects of land use change on zoonotic disease transmission.
Pathogen transmission between potential hosts is shown as black arrows. Land use change (green driver) acts on ecological community composition and human populations (white boxes), and on environmental features that influence contact and transmission both locally (light blue box) and at broader geographical scales (dark blue box). These processes occur within a broader socio-ecological system context also influenced by additional environmental (for example, climatic), socioeconomic and demographic factors. Unpicking the relative influence of these different processes on disease outcomes is challenging in local disease system studies, in which multiple processes may be acting on pathogen prevalence and transmission intensity. The aim of this analysis was therefore to specifically examine, at a global scale, the effects of land use change on the composition of the potential host community (excluding domestic species), denoted by the red box.
Extended Data Fig. 2 Approximating research effort bias for non-host species within the PREDICTS dataset.
For all non-host species, we approximated the likelihood of false classification given research effort (that is, probability of being a host, but not detected), based on the distribution of publication effort across known zoonotic hosts within the same taxonomic order (Supplementary Methods 1). a, b, Line graphs show, for several orders, the cumulative curve of publication counts for known zoonotic hosts (a; shown on log-scale), and approximated false classification probability, which declines and asymptotes with increasing levels of research effort (b) (line colours denote taxonomic order). c–e, Points and box plots show the distribution of PubMed publications for all host and non-host species in PREDICTS (c; total n = 6,921), and false classification probabilities (used as bootstrap transition rates) for all non-host species per taxonomic class in PREDICTS (d; total n = 3,665), and per key mammalian and avian order (e; total n = 2,927) (bracketed numbers denote number of species per group; boxes show median and interquartile range, whiskers show values within 1.5× IQR from quartile). f, The histogram shows the number of non-host species transitioned to host status for each of 1,000 bootstrapped models of the full dataset (median 121, 95% quantile range 102–142).
Extended Data Fig. 3 Random (study-level) and geographical cross-validation of community models (full dataset).
We tested the sensitivity of fixed effects estimates to both random and geographically structured (biome-level) subsampling. a, For random tests we fitted 8 hold-out models, excluding all sites from 12.5% of studies at a time (mean 12.5% of total sites excluded per model, range 4–19%). b, For geographical tests we fitted 14 hold-out models, with each excluding all sites from one biome (mean 7% of sites excluded per model, range 0.07–32%). Points and error bars show posterior marginal parameter distributions for each hold-out model (median and 95% quantile range, with colour denoting hold-out group or biome), calculated across samples from 500 bootstrap iterations per-model to account for variable research effort across species. Directionality and evidence for fixed-effects estimates are robust to both tests, suggesting that our results are not driven by data from any particular subset of studies or regions. Urban parameters are, however, the most sensitive to exclusion of data, probably owing to the relatively sparse representation of urban vertebrate diversity in the PREDICTS database (17 studies in our full dataset).
Extended Data Fig. 4 Effects of land use on site-level mammalian reservoir host species richness and total abundance.
a–d, Points, wide and narrow error bars show differences in diversity metrics from primary minimal use baseline (posterior marginal median, 67% and 95% quantile ranges respectively, across 1,000 bootstrap models). Models are of species richness (a) and total abundance (b) of reservoir host and all other (non-host) species, and of hosts as a proportion of site-level richness (c) and total abundance (d). For managed and urban sites, use intensities were combined to improve evenness of sampling (n = 2,026 sites from 63 studies: primary (589 and 572 for minimal and substantial use respectively), secondary (144, 257), managed (348) and urban (116)). Posterior estimates were calculated across an ensemble of 1,000 bootstrapped models (median 51, range 38–62 non-hosts transitioned to host status, that is, increasing host number by 28–46%) (Methods). Results from urban sites show the same trend as the full dataset (Fig. 2), but are not visualized owing to wide uncertainty: 88.7% (−2.1, 252.3) proportion richness, 307% (78.8, 500.7) proportion abundance (posterior median and 95% quantile range; see Supplementary Table 4). Point shape indicates use intensity (minimal, substantial or both combined) and colour indicates host (brown) or non-host (green). Reservoir species are listed in Supplementary Table 1 (mammal species listed as ‘Detection/reservoir’ in the ‘Evidence of host status’ column).
Extended Data Fig. 5 Effects of land use on occurrence and zero-truncated abundance (abundance given presence) of mammalian and avian hosts and non-hosts of zoonotic agents.
Each row of three plots shows the results of species-level modelling for each of five mammalian and two avian orders, and for mammals overall. Points, wide and narrow error bars show average difference in species occurrence probability (left column) and ZTA (middle column) (posterior median, 67% and 95% quantile ranges across 500 and 750 bootstrap iterations, for each order and all mammals respectively). Differences are shown in secondary (Sec), managed and urban sites relative to a primary land baseline (dashed line), across all host (brown) and non-host (green) species. Histograms show, for each taxonomic group, the distribution of host species counts across all bootstrap models (that is, after reclassifying non-hosts) compared to current number of known hosts (red vertical line), and the total number of species included in models (brackets in plot title). Estimates from occupancy and ZTA models (Supplementary Table 6) were combined, assuming independence of processes, to give the hurdle predictions in Fig. 3. Mammal reservoir status was defined on the basis of strict criteria (pathogen detection or isolation), and the full list of host species included in these estimates is provided in Supplementary Table 1 (scored ‘1’ in the’ zoonotic agent host’ column). Silhouettes obtained from PhyloPic (http://phylopic.org/).
a–c, Distribution of human-shared and non-human-shared pathogen richness (a) and relationship to publication counts (b, c) are shown for mammals in our host–pathogen association dataset (n = 780 species; points represent species shaded by order, associations defined on serological or stronger evidence). d, e, Observed versus fitted plots show where observed deviates from expected pathogen richness given log-publications and taxonomic group (Poisson likelihood with random intercepts and slopes for order and family; slope estimates for log-publications are similar for both human and non-human-shared pathogens, β of 0.298 and 0.248 respectively). f, Fitted models were used to predict expected pathogen richness for mammals in PREDICTS (n = 546) and derive residuals from observed values, which were used in land use models (Extended Data Fig. 7). g, h, Calculating per-species residual quantile ranges across 2,500 posterior parameter samples shows that within-species residual variance is generally small relative to residual size, points and error-bars show posterior median, 67% and 95% intervals, scaled to unit variance), and land use model results are robust to including this uncertainty (Methods, Supplementary Table 7).
Extended Data Fig. 7 Effects of land use on the relationship between mammal species pathogen richness and occurrence probability.
a–d, Points and error bars show intercept (a, b) and slope parameters (c, d) of the relationship between residual pathogen richness (scaled to mean 0 and unit variance) and mammal species occurrence probability (on the log odds scale; median and 95% credible interval). Model was fitted to occurrence data for all mammals in the database (n = 29,569 records of 546 species, 1,950 sites, 66 studies). Intercept parameters represent the average occurrence probability of a species with residual pathogen richness of 0 (that is, with average pathogen richness given research effort and taxonomy), and slope parameters represent the change in occurrence probability for one scaled unit (s.d.) increase in residual pathogen richness (Extended Data Fig. 6g, h). Intercept and slope parameters for primary and secondary land measure the differences relative to managed land (that is, delta-intercept or delta-slope; b, d). e, f, Plotted lines show these relationships on the probability scale, showing the median (black line), 67% (dark shading) and 95% (light shading) quantile range, based on 3,000 samples from the joint posterior distribution. For both human-shared and non-human-shared pathogens, there is a positive relationship between the residual pathogen richness of a species and its probability of occurrence in human-managed land. For human-shared pathogens, the strength of this relationship (slope parameter) is significantly larger in managed sites than in both primary and secondary land, and for non-human-shared pathogens significantly larger in managed than in primary land (d; slopes for primary land not significantly different from 0). Full model summaries and results of sensitivity analyses are in Supplementary Table 7.
Extended Data Fig. 8 Differences in human population density between land use types, for all sites within the full dataset.
Points and boxplots show the distributions of log-transformed human population density by land use type and intensity, across all sites included in community models (n = 6,801). Boxes show median and interquartile range with whiskers showing values within 1.5 × IQR from quartile, and are coloured by land use type, and numbers denote the number of sites in each category. Human population density estimates were extracted from CIESIN Gridded Population of the World 4, for 2005, the median year of studies included in the dataset. Per-site log human density estimates were considered as fixed effects in community models of host diversity, because human-tolerant or synanthropic species might respond to human population change independently of land use (Methods).
Extended Data Fig. 9 Diagnostic plots for all community models (full dataset and mammal reservoirs subset).
Species richness counts were modelled with a Poisson likelihood, and abundance (adjusted counts) were log-transformed and modelled with a Gaussian likelihood (see Methods). Plot titles refer to model response variables: species richness (SR), total abundance (Abundance), for hosts, non-hosts, and for hosts as a proportion of the community (Prop). a, b, Observed data against model-fitted values are shown in a. The red line shows the expectation if observed equals fitted (n = 6,801 for full SR; n = 6,093 for full abundance; n = 2,026 for mammals SR; n = 1,963 for mammals abundance). We also tested for spatial autocorrelation of residuals across all sites within each study, with histograms (b) showing the distribution of per-study Moran’s I P values (indicating significance of spatial autocorrelation among sites within that study) for each model (n = 184 for full SR; n = 164 for full abundance; n = 63 for mammals SR; n = 60 for mammals abundance). Numbers in brackets are the percentage of studies that contained significant spatial autocorrelation (P < 0.05, shown as a red line). Overall, spatial autocorrelation was fairly low across the dataset (statistically significant in 14–34% of studies, with maximum 26% for models with host metrics as response variables). Residuals and statistics were derived from a single fitted model including community mean false classification probability as a linear covariate to account for research effort (with known hosts given a false classification probability of 0), rather than the full bootstrap ensemble.
This file contains Supplementary Methods (x1; approximation of research effort bias for non-host species) and Supplementary Tables (x8; includes tabular summaries of host-pathogen and ecological communities data, and full numeric summaries of model results).
About this article
Cite this article
Gibb, R., Redding, D.W., Chin, K.Q. et al. Zoonotic host diversity increases in human-dominated ecosystems. Nature 584, 398–402 (2020). https://doi.org/10.1038/s41586-020-2562-8