How changes in species identification can affect disease prediction. A case study of a Cricetid rat and Hantavirus.
-
1
University of California, Davis, United States
Rats belonging to the Oligoryzomys genus can be found all over Central and South America, they are usually difficult to identify and have a lot of species described (González‐Ittig et al., 2010). They are of particular importance to public health since many have been implicated as natural reservoirs of a series of zoonotic diseases (Jonsson et al., 2010; Meerburg et al., 2009). However, there has been a lot of changes in the taxonomy of this species and not always has the medical field been caught up with the latest species reviews. A recent example is of the recently described O. mattogrossae, formerly known as O. fornesi. This species was identified as the reservoir for Anajatuba hantavirus but under its former classification (O. fornesi) (Weksler et al., 2017). For the different classifications of Oligoryzomys genus I used the specimens described in the article by Weksler et al., 2017. Geolocations were determined according to the catalog number of the published individuals. The MZV and USNM online databases were used to determine specific coordinates or a description of the area that allowed geolocation. Other locations were obtained in the scientific literature that described these specific catalogued individuals (Pereira & Geise 2007, Oliveira 2008, Bonvicino et al 2014, Rocha et al 2001, Paresque et al 2007, Weksler & Bonvicino 2005, Teta et al 2009). This information provided the data for 3 sets of species distributions:The official Oligoryzomys mattogrossae distribution, the old distribution for Oligoryzomys fornesi before this taxonomic review (matching the extent described in the IUCN site for this species) and the official Oligoryzomys fornesi distribution. The final dataset was extracted from the GBIF (Global Biodiversity Information) online database under the search term “Oligoryzomys fornesi”. A search for “Oligoryzomys mattogrossae” was also performed but did not output any results. All geolocations were plotted on Google Earth to check for accuracy. Final plots with species points was plotted using RStudio vs. 1.1463. Climate data from the BioClim database was uploaded and cropped to the extent of South America. A correlation matrix plot was done to evaluate collinearity among variables. A Maxent (Maximum Entropy) analyses was done with the Maxent interface through the dismo package in RStudio vs 1.1463. Once the model was built, variable importance was observed for each dataset as well as a prediction of suitable occurrence areas in the form of a raster file. These rasters were subtracted from each other to view possible errors in prediction according to the datasets. Finally, to evaluate model fit k-folds were done to cross validate the data and the resulting AUC from the train and test data was plotted. All statistical analyses were done in RStudio vs 1.1463. The following variables were kept in the model: annual mean temp., mean diurnal range, isothermality, temp. annual range, annual precipitation, precipitation of driest month, precipitation seasonality, precipitation of warmest quarter, precipitation of coldest quarter. There was a great variability among the distribution of the species according to the datasets. A Maxent model was done of species distribution pairing the points to the bioclimatic variables and the most important variables determined . To discover the error among the predictions all models were compared to the gold standard, in this case the correct distributions for O. fornesi and O. mattogrossae. To do this, one raster was subtracted from the other and the difference was plotted. The datasets vary on different degrees in term of species distribution. The gold standard of O. fornesi distribution greatly resembles that of the GBIF database. This can be better seen in Table 1, where the difference between predicted areas is projected. This small difference is probably due to the fact that the GBIF database only contains records from museum specimens for this species, many, correctly identified by Weksler and collaborators in the paper that served as the gold standard for this article. During my literature search I also came across a seemingly correct map of O. fornesi distribution on the latest edition of Mammals of the World, that is due to the fact that Weskler was responsible for writing that particular chapter and included preliminary findings to the article that he later published and was used here as gold standard (Nowak & Walker, 2015). However, other sources of data, such as the IUCN site did not have the updated distribution. If a researcher relies on that information, they could have a serious bias in their work. All datasets had similar variable importance for prediction, that usually included isothermality, average temperature and precipitation. The AUC plots also indicated good model fit. However, this data must be viewed with caution since for most datasets the sample size is small and only bioclimatic variables were included in the model. Other variables such as elevation, land use, soil and vegetation type should be included to make more robust predictions. However, for the challenge at hand, just the climatic variables were enough to show discrepancies in species prediction according to the use of different data origins. Researchers must be wary of changes in host taxonomy especially when host distribution is used for disease risk analyses. Literature updates should be frequent and contact with a local natural history museum is important to deposit specimens identified as hosts, that way future taxonomic revisions can include this data and provide more reliable information for ecologists and epidemiologists alike.
Acknowledgements
I am very grateful for the comments and suggestions by Prof. Beatriz Martinez Lopez and Prof. Christopher Barker greatly improved this project. Also for Eliezer Guiterrez for teaching me the basics and fundamentals of ENM
References
Bonvicino, CR; Lazar, A; Corrêa, MMO; Weksler, M; Paula, AC; Bezerra, AMR. 2014 Conservation units in the core area of the Cerrado domain: An overview on the small nonvolant mammals (Rodentia and Didelphimorphia). Heringeriana 8(2).
González‐Ittig RE; Salazar‐Bravo J; Barquez RM; Gardenal CN. 2010 Phylogenetic relationships among species of the genus Oligoryzomys (Rodentia, Cricetidae) from Central and South America. Zoologica Scripta. 39(6):511-26.
Jonsson CB; Figueiredo LT; Vapalahti O. 2010 A global perspective on hantavirus ecology, epidemiology, and disease. Clinical microbiology reviews. 23(2):412-41.
Meerburg BG; Singleton GR; Kijlstra A. 2009 Rodent-borne diseases and their risks for public health. Critical reviews in microbiology. 35(3):221-70.
Nowak RM; Walker EP. 2015 Walker's Mammals of the World. JHU Press
Oliveira, FCG. 2008 Avaliação preliminar de impacto ambiental sobre a fauna de pequenos mamíferos e suas taxas de infecção por Trypanosoma cruzi e hantavírus na área de influência da usina hidrelétrica espora, Apore – GO. Master’s Thesis for the Environmental and Health Science Graduate Program, Universidade Católica de Goiás, Brazil.
Paresque, R; Silva, JJ; Yonenaga-Yassuda, Y; Fagundes, V. 2007 Karyological geographic variation of Oligoryzomys nigripes Olfers, 1818 (Rodentia, Cricetidae) from Brazil Genet. Mol. Biol. 30(1)
Pereira, LGI; Geise, L. 2007 Karyotype composition of some rodents and marsupials from Chapada Diamantina (Bahia, Brasil) Braz. J. Biol. 67(3).
Rocha, RG; Ferreira, E; Costa, BMA; Martins, ICM; Leite, YLR; Costa, LP; Fonseca, C. 2011 Small mammals of the mid-Araguaia River in central Brazil, with the description of a new species of climbing rat. Zootaxa 2789:1–34
Teta, P; Pereira, JA; Muschetto, E; Fracassi, N. 2009 Mammalia, Didelphimorphia, Chiroptera, and Rodentia, Parque Nacional Chaco and Capitán Solari, Chaco Province, Argentina. Check List 5(1): 144–150
Weksler, M; Bonvicino, CR. 2005 Taxonomy of pigmy rice rats genus Oligoryzomys Bangs 1900 (Rodentia, Sigmodontinae) of the Brazilian Cerrado with the description of two new species. Arquivos do Museu Nacional, Rio de Janeiro, 63(1):113-130
Weksler M; Lemos EM; D'Andrea OS; Bonvicino CR. 2017 The taxonomic status of Oligoryzomys mattogrossae (Allen 1916)(Rodentia: Cricetidae: Sigmodontinae), reservoir of Anajatuba Hantavirus. American Museum Novitates. (3880):1-33.
Keywords:
niche model,
Hantavirus,
rodent,
Taxonomy,
South America
Conference:
GeoVet 2019. Novel spatio-temporal approaches in the era of Big Data, Davis, United States, 8 Oct - 10 Oct, 2019.
Presentation Type:
Regular oral presentation
Topic:
Spatial methods for environmental & exposure epidemiology and climate change
Citation:
Grossmann
NV
(2019). How changes in species identification can affect disease prediction. A case study of a Cricetid rat and Hantavirus..
Front. Vet. Sci.
Conference Abstract:
GeoVet 2019. Novel spatio-temporal approaches in the era of Big Data.
doi: 10.3389/conf.fvets.2019.05.00078
Copyright:
The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers.
They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.
The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.
Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.
For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.
Received:
21 Jun 2019;
Published Online:
27 Sep 2019.
*
Correspondence:
DVM. Narjara V Grossmann, University of California, Davis, Davis, United States, nvgrossmann@ucdavis.edu