Skip to main page content
U.S. flag

An official website of the United States government

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Jan 22;11(1):2098.
doi: 10.1038/s41598-021-81329-x.

Spatiotemporal clustering and Random Forest models to identify risk factors of African swine fever outbreak in Romania in 2018-2019

Mathieu Andraud  1 , Stéphanie Bougeard  2 , Theodora Chesnoiu  3 , Nicolas Rose  2
Affiliations

Spatiotemporal clustering and Random Forest models to identify risk factors of African swine fever outbreak in Romania in 2018-2019

Mathieu Andraud et al. Sci Rep. .

Abstract

African swine fever (ASF) has affected Romania since July 2017, with considerable economic and social consequences, despite the implementation of control measures mainly based on stamping out of infected pig populations. On the basis of the 2973 cumulative recorded cases up to September 2019 among wild boars and domestic pigs, analysis of the epidemiological characteristics could help to identify the factors favoring the persistence and spread of ASF. A statistical framework, based on a random forest methodology, was therefore developed to assess the spatiotemporal features of the epidemics and their relationships with environmental, human, and agricultural factors. The landscape of Romania was associated with the infection dynamics, particularly concerning forested and wetland areas. Waterways were also identified as a pivotal factor, raising questions about possible waterborne transmission since these waterways are often used as a water supply for backyard holdings. However, human activity was clearly identified as the main risk factor for the spread of ASF. Although the situation in Romania cannot be directly transposed to intensive pig farming countries, the findings of this study highlight the need for strict biosecurity measures on farms, and during transportation, to avoid ASF transmission at large geographic and temporal scales.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1
Figure 1
Cartography of the geographic variables. All data are presented as densities at the commune level.
Figure 2
Figure 2
Demographic and agricultural data. Data are presented at the commune level (human population, backyard pigs) or at the county level (pigs, holdings).
Figure 3
Figure 3
Temporal characteristics of the ASF epidemics in Romania (epidemiologic data): weekly incidence in wild boars, domestic pigs and total population (a), and cumulative incidence throughout the period (b).
Figure 4
Figure 4
Spatial expansion of the ASF epidemics in Romania from June 10, 2018 to August 29, 2019; (a) case locations (map created using R (version 4.0.2), packages tmap and rgdal), (b) number of domestic pig outbreaks per commune.
Figure 5
Figure 5
Geographic distribution of the 14 spatiotemporal clusters from June 10, 2018 to August 29, 2019 (N = 2973 cases; 2371 pig holding outbreaks and 602 wild boars cases). (a) 2-dimensional representation; (b) 3-dimensional plot, time being represented along the third (vertical) axis (interactive view available on request).
Figure 6
Figure 6
Random forest classification for the 14 spatiotemporal clusters of ASF in Romania. (a) Explanatory variable importance plot to differentiate clusters (N = 2895 cases, including 2322 pig holding outbreaks and 573 wild boar cases from June 10, 2018 to August 29, 2019). (b) Most frequent interactions between the eight significant explanatory variables explaining the 14 spatiotemporal clusters of ASF in Romania from June 10, 2018 to August 29, 2019 by means of a random forest classification. N = 2895 cases, including 2322 pig holding outbreaks and 573 wild boar cases.
Figure 7
Figure 7
Random forest regression results for 2018 ASF outbreaks in Romania at the commune scale (N = 2939 communes). (a) Observed versus predicted values of the number ASF outbreaks (pig holdings) per commune. (b) Explanatory variable importance plot considering node purity increase and mean minimal depth. (c) Mean minimal depth for 30 most frequent interactions between the 10 significant explanatory variables.
Figure 8
Figure 8
Random forest regression results for 2019 ASF outbreaks in Romania at the commune scale (N = 2939 communes). (a) Observed versus predicted values of the number ASF outbreaks (pig holdings) per commune. (b) Explanatory variable importance plot considering node purity increase and mean minimal depth. (c) Mean minimal depth for 30 most frequent interactions between the 10 significant explanatory variables.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Galindo I, Alonso C. African swine fever virus: A review. Viruses. 2017;9:103. doi: 10.3390/v9050103. - DOI - PMC - PubMed
    1. Gallardo C, et al. Experimental infection of domestic pigs with African swine fever virus Lithuania 2014 genotype II field isolate. Transbound Emerg. Dis. 2017;64:300–304. doi: 10.1111/tbed.12346. - DOI - PubMed
    1. Guinat C, et al. Inferring within-herd transmission parameters for African swine fever virus using mortality data from outbreaks in the Russian Federation. Transbound Emerg. Dis. 2018;65:e264–e271. doi: 10.1111/tbed.12748. - DOI - PMC - PubMed
    1. Arias M, Jurado C, Gallardo C, Fernandez-Pinero J, Sanchez-Vizcaino JM. Gaps in African swine fever: Analysis and priorities. Transbound Emerg. Dis. 2018;65(Suppl 1):235–247. doi: 10.1111/tbed.12695. - DOI - PubMed
    1. Rowlands RJ, et al. African swine fever virus isolate, Georgia, 2007. Emerg. Infect. Dis. 2008;14:1870–1874. doi: 10.3201/eid1412.080591. - DOI - PMC - PubMed

LinkOut - more resources