Novel Methods In Disease Biogeography: A Case Study With .
Technology Reportpublished: 17 July 2017doi: 10.3389/fvets.2017.00105Novel Methods in DiseaseBiogeography: A Case Study withHeterosporosisLuis E. Escobar1,2,3*, Huijie Qiao4, Christine Lee1 and Nicholas B. D. Phelps1,21Minnesota Aquatic Invasive Species Research Center, University of Minnesota, St. Paul, MN, United States, 2 Department ofFisheries, Wildlife, and Conservation Biology, University of Minnesota, St. Paul, MN, United States, 3 Escuela de Estudios dePostgrado, Facultad de Medicina Veterinaria y Zootecnia, Universidad de San Carlos de Guatemala, Guatemala, Guatemala,4Key Laboratory of Animal Ecology and Conservation Biology, Institute of Zoology, Chinese Academy of Sciences,Beijing, ChinaEdited by:Victoria J. Brookes,University of Sydney, AustraliaReviewed by:Hans-Hermann Thulke,Helmholtz-Zentrum fürUmweltforschung (UFZ), GermanyLina Mur,Kansas State University,United States*Correspondence:Luis E. Escobarecoguate2003@gmail.comSpecialty section:This article was submitted toVeterinary Epidemiology andEconomics,a section of the journalFrontiers in Veterinary ScienceReceived: 31 January 2017Accepted: 19 June 2017Published: 17 July 2017Citation:Escobar LE, Qiao H, Lee C andPhelps NBD (2017) Novel Methods inDisease Biogeography: A Case Studywith Heterosporosis.Front. Vet. Sci. 4:105.doi: 10.3389/fvets.2017.00105Disease biogeography is currently a promising field to complement epidemiology, andecological niche modeling theory and methods are a key component. Therefore, applying the concepts and tools from ecological niche modeling to disease biogeographyand epidemiology will provide biologically sound and analytically robust descriptive andpredictive analyses of disease distributions. As a case study, we explored the ecologicallyimportant fish disease Heterosporosis, a relatively poorly understood disease caused bythe intracellular microsporidian parasite Heterosporis sutherlandae. We explored twonovel ecological niche modeling methods, the minimum-volume ellipsoid (MVE) and theMarble algorithm, which were used to reconstruct the fundamental and the realizedecological niche of H. sutherlandae, respectively. Additionally, we assessed how themanagement of occurrence reports can impact the output of the models. Ecologicalniche models were able to reconstruct a proxy of the fundamental and realized nichefor this aquatic parasite, identifying specific areas suitable for Heterosporosis. We foundthat the conceptual and methodological advances in ecological niche modeling provideaccessible tools to update the current practices of spatial epidemiology. However, carefuldata curation and a detailed understanding of the algorithm employed are critical for aclear definition of the assumptions implicit in the modeling process and to ensure biologically sound forecasts. In this paper, we show how sensitive MVE is to the input data,while Marble algorithm may provide detailed forecasts with a minimum of parameters.We showed that exploring algorithms of different natures such as environmental clusters,climatic envelopes, and logistic regressions (e.g., Marble, MVE, and Maxent) providedifferent scenarios of potential distribution. Thus, no single algorithm should be used fordisease mapping. Instead, different algorithms should be employed for a more informedand complete understanding of the pathogen or parasite in question.Keywords: disease biogeography, risk map, ecological niche modeling, minimum-volume ellipsoid, heterosporosisINTRODUCTIONDisease biogeography is the study of the geographic distribution of infectious diseases (1). It isa powerful approach for mapping disease events, which can inform decision-makers, managers,researchers, and animal and public health specialists (2, 3). Disease biogeography has been proposedas a promising field that can help understand why diseases emerge in one site, but not in anotherFrontiers in Veterinary Science www.frontiersin.org1July 2017 Volume 4 Article 105
Escobar et al.Methods in Disease Biogeography(descriptive analyses), and also provides information to identifysuitable areas where outbreaks could occur in the future (predictive analysis) (1).where data for susceptible individuals, reservoirs, and vectors isscarce (3). Complex ecological niche models can be developedwhen more information is available, such as seasonality, densityof vectors and reservoirs, and immunity of susceptible hosts,allowing to identify with more detail the different levels ofdisease transmission risk across areas, periods, and populations(1).Theoretically, species’ niches can be described as FundamentalNiche (NF) and Realized Niche [NR (5, 6); Figure 1]. The NFwould resemble the abiotic conditions not modifiable by thespecies and that are necessary by the species to survive and,most importantly, to maintain populations in the long termwithout the need for immigration. The NR is represented by theportion of the NF that is actually occupied by the species (2).NF and NR are usually estimated in ecological niche modelingbased on field observations also termed occurrences and theenvironmental conditions in a region, here termed background.In the field of ecological niche modeling, considerable effortshave been made to develop methods and environmental variables to determine the NF and NR of species under the assumption that occurrences NR NF background. Ecological nichemodeling estimations are therefore developed in environmentaldimensions to be later projected to geography in the form ofmaps of areas occupied and potentially occupied by the speciesin question (Figure 1).Conceptual BasesAccording to the assumption of disease biogeography, diseasesare not distributed at random across the landscape, instead occurin non-random tractable and quantifiable landscape or environmental conditions. Disease biogeography incorporates the concept of the ecological niche as a crucial element to understand theenvironmental requirements of a disease transmission systemas well as the geographic distribution of the species involvedin the system (1, 2). Disease biogeographers use the conceptualbases and methods from the field of ecological niche modeling tomake disease biogeography more quantitative (3, 4). Ecologicalniche modeling links field reports with environmental variables,allowing for development of the descriptive and predictive analyses required by disease biogeography. When ecological nichemodeling is used for spatial epidemiology, it varies in complexity, ranging from simple “black-box” approaches (focusing oninfected individuals only to reconstruct the conditions wherethe disease may persist) to more complex hierarchical ecological niche models (including several components of the diseasesystem, e.g., intermediate host, reservoir, vector) (2). Black-boxecological niche models are usually employed for rare diseasesFIGURE 1 The theoretical scenarios of Fundamental (NF) and Realized Niches (NR) of an aquatic parasite in environmental space. Left: all the set of abioticenvironmental conditions suitable for the parasite resembling NF (teal cloud). Right: the sub-set of abiotic environmental conditions suitable for the speciesresembling NR (teal cloud). In this scenario, the species is restricted to a portion of NF due to the effect of biotic interactions (red; e.g., competition with otherparasites or absence of fish hosts in the red region making this portion of the niche unusable). Note the background of abiotic environmental conditions available forthe species (gray lines) composed by water temperature and sunlight.Frontiers in Veterinary Science www.frontiersin.org2July 2017 Volume 4 Article 105
Escobar et al.Methods in Disease BiogeographyApplications in Epidemiologythe muscle tissue. Advanced stages of the disease likely result inindirect parasite-induced mortality due to decreased overall fitness, inability to capture prey or escape predation, and increasedhost stress (Figure 2B). The transmission of H. sutherlandae isthought to be horizontal, through the consumption of infectedprey or contact with mature spores shed into the water column.Consequently, the overland transport of infected fish or water arelikely risk factors for the spread of this pathogen. The possibilitydoes exist for vertical transmission, similar to other microsporidian species infecting fish (16).With Heterosporosis as a case study, we explored the use ofnext generation biogeography tools to evaluate how these tools andapproaches can help (i) understand the ecology of a rare infectiousdisease and (ii) forecast the geographic areas where future investigation is necessary. This contribution aims to use the most state-ofthe-art algorithms and variables available in order to incorporatedisease biogeography in the toolkit of modern epidemiology.While biogeographic methods have gained attention in theepidemiology of terrestrial ecosystems (3), they have been barelyexplored in the epidemiology of aquatic organisms (7). Examplesof biogeographic analyses applied to infectious aquatic diseasesinclude forecasts of Gyrodactylus salaris an ectoparasite of salmon(8), Vibrio cholera in coastal waters (9), and Viral HemorrhagicSepticemia virus in the Great Lakes (10). Descriptive biogeographicanalyses are useful to understand the natural history of novel infectious diseases, poorly known diseases, or diseases barely exploredin the field (11–13). Predictive analyses are useful to anticipate riskin areas where the diseases has not yet been reported, and to guideactive surveillance and research (14). A poorly understood infectious disease of epidemiological importance is Heterosporosiswhich infects fish in the Great Lakes region. Heterosporosis iscaused by the microsporidian parasite Heterosporis sutherlandaeand is known to infect at least eight fish species of economic andecological importance (15). This disease was first confirmed in2000 in Leech Lake and Catfish Lake in Minnesota and Wisconsinand has since been reported in waterbodies in Minnesota (n 26),Wisconsin (n 16), Michigan (n 2) in the USA and Lake Ontario(15). The obligate intracellular parasites proliferate inside skeletalmuscle cells (Figure 2A), eventually leading to liquefaction ofMETHODSOccurrencesWe obtained Heterosporosis-positive occurrence locationsfrom Miller (17) and Phelps et al. (15), who in turn receivedFIGURE 2 Species used in this exploration. (A) Necrotic muscle tissue of the fish Fathead minnows (Pimephales promelas) infected with large aggregations ofspores from the parasite Heterosporis sutherlandae. (B) Fathead minnows experimentally challenged with H. sutherlandae. (C) Heterosporosis-positive occurrences(black points) across the Great Lakes region used for this study. Lines denote administrative boundaries.Frontiers in Veterinary Science www.frontiersin.org3July 2017 Volume 4 Article 105
Escobar et al.Methods in Disease Biogeographythe reports from natural resource management agencies(i.e., Minnesota Department of Natural Resources, WisconsinDepartment of Natural Resources, and U.S. Fish and WildlifeService). Reports were confirmed by gross lesions and histopathology, and in some cases by PCR and sequencing. Anecdotalreports not verified in the laboratory were not included in thisstudy. Lake centroids were used to determine latitude and longitude locations, and duplicate coordinates were removed. Toexplore the effect of data curation in the model’s performance,models were developed using all the final occurrences availableand a subset of resampled occurrences without environmentaloutliers (see below).TABLE 1 Environmental variables used to construct the background.Fundamental Niche (NF)The NF was estimated in a large model calibration regionincluding: all the occurrences and the filtered occurrences.Specifically, we focused on the Laurentian Great Lakes regionof North America (41.4 and 49.3 N and 97.8 and 74.8 W),a bi-national Canadian–American region with portions of theAmerican states of Ohio, Illinoi, Indiana, Minnesota, Wisconsin,Michigan, Pennsylvania, New York, and the Canadian provinceof Ontario (Figure 2C). We used climatic variables from thiscalibration region to construct a background of environmentalconditions in which the NF was estimated (18) resembling thelandscape and terrestrial environmental drivers where parasitesand hosts co-occur. We used climate data from the CliMondrepository (19), selecting the first 35 bioclimatic variableswith original measurable information on annual, weekly, andseasonal temperature, soil moisture, radiation, and precipitation (Table 1), as these variables are a proxy to reconstructecoregions and present-day faunistic distributions (20). Thesevariables are a summary of climatic conditions between 1961and 1990 in the form of rasters at 19 km spatial resolution.A principal component analysis was developed using NicheAsoftware 3.0 (21) to reduce dimensionality and correlationbetween variables, retaining the first three components as theycontained 83.85% of the information from the original set ofvariables. These three components composed the environmentalbackground that summarized the environmental patterns in thearea with reduced spatial and temporal autocorrelation and wereused in posterior analyses. The background developed was thenused by the ecological niche model algorithms to identify therelationship of parasite occurrences with this environmentalbackground. Once this relationship is established, models searchfor this combination of conditions across the entire study area todefine locations suitable and unsuitable for the parasite.To mitigate uncertainty implicit in occurrences, we employeda method modified from Van Aelst and Rousseeuw (22) as filterto remove potential errors in occurrences. This filtering methodis robust for outlier detection: we estimated minimum ellipsoidsaround occurrences displayed in environmental space andremoved 5% [i.e., α 0.05 (3, 23)] of occurrences with the mostmarginal environmental values, as these outlier values could beassociated with occurrence errors [e.g., misidentification; see,Ref. (24)]. The script for occurrences filtering by detection of theoutliers has been included as Supplementary Material S1. Wethen estimated the NF using NicheA with the remaining filteredFrontiers in Veterinary Science www.frontiersin.orgFundamental nicheRealized nicheAnnual mean temperature ( C)Mean value of the monthly MODISenhanced vegetation index (EVI) timeseries data (index)Mean diurnal temperature range[mean(period max-min)] ( C)SD of the monthly MODIS EVI timeseries data (index)Isothermality (Bio02 Bio07)Mean value the 8-day MODISday-time land surface temperature(LST) time series data ( C)Temperature seasonality (C of V)SD of the 8-day MODIS day-time LSTtime series data ( C)Max temperature of warmest week ( C)Minimum value of the 8-day MODISday-time LST time series data ( C)Min temperature of coldest week ( C)Maximum value of the 8-day MODISday-time LST time series data ( C)Temperature annual range(Bio05-Bio06) ( C)Mean value the 8-day MODISnight-time LST time series data ( C)Mean temperature of wettestquarter ( C)SD of the 8-day MODIS night-timeLST time series data ( C)Mean temperature of driestquarter ( C)Minimum value of the 8-day MODISnight-time LST time series data ( C)Mean temperature of warmestquarter ( C)Maximum value of the 8-day MODISnight-time LST time series data ( C)Mean temperature of coldestquarter ( C)Mean value of the 8-day MODISday-time LST time series data forDecember/January ( C)Annual precipitation (mm)Mean value of the 8-day MODISday-time LST time series data forFebruary/March ( C)Precipitation of wettest week (mm)Mean value of the 8-day MODISday-time LST time series data forApril/May ( C)Precipitation of driest week (mm)Mean value of the 8-day MODISday-time LST time series data forJune/July ( C)Precipitation seasonality (C of V)Mean value of the 8-day MODISday-time LST time series data forAugust/September ( C)Precipitation of wettest quarter (mm)Mean value of the 8-day MODISday-time LST time series data forOctober/November ( C)Precipitation of driest quarter (mm)Precipitation of warmest quarter (mm)Precipitation of coldest quarter (mm)Annual mean radiation (W m 2)Highest weekly radiation (W m 2)Lowest weekly radiation (W m 2)Radiation seasonality (C of V)Radiation of wettest quarter (W m 2)Radiation of driest quarter (W m 2)Radiation of warmest quarter (W m 2)Radiation of coldest quarter (W m 2)Annual mean moisture indexHighest weekly moisture indexLowest weekly moisture indexMoisture index seasonality (C of V)Mean moisture index of wettest quarterMean moisture index of driest quarterMean moisture index of warmest quarterMean moisture index of coldest quarterFundamental niche: variables based on climatic data at 19 km spatial resolution.Realized Niche: variables based on MODIS data at 1 km spatial resolution.4July 2017 Volume 4 Article 105
Escobar et al.Methods in Disease Biogeographythe areas predicted by the NF model. The NF and NR were thenprojected to the geographic space to identify areas suitable aspredicted by the models.Finally, to highlight the predictions of MVE and Marble vs.a classic ecological niche modeling method, we developed aseries of models using Maxent algorithm (32). Maxent is a typeof logistic regression (33) and is currently a standard method toestimate species’ ecological niches (34). Maxent models includedthe estimation of the NF based on climate data and NR basedon remote sensing data. The NF and NR were estimated usingthe original occurrences and filtered occurrences as describedbefore. Models were calibrated using default settings in Maxent3.3.3k (34).All models were compared using a cumulative binomialdistribution test using two sets of occurrences, one for modelcalibration and one for model evaluation, as in Peterson et al.(24). The R script used here for automated data split is includedas Supplementary Material S3. Evaluation occurrences were notused during model calibration and instead were used to test theability of the model to predict independent data using evaluationpoints as trials, evaluation points predicted correctly as successes,and the proportion of area predicted suitable as the probabilityof a success (23). The method used to develop this evaluation isincluded as Supplementary Material S4 to facilitate replication.occurrences. The NF was calculated as the minimum-volumeellipsoid (MVE) from the occurrences in a three-dimensionalenvironmental scenario composed by the first three componentsfrom the original environmental variables, described elsewhere(21, 22). Basically, occurrences are displayed and analyzed inthree environmental dimensions instead of two geographicdimensions (i.e., latitude and longitude). NicheA estimates thecentroid point of the occurrences’ cloud, which will be the centerof the ellipsoid. Then, the Euclidean distance is estimated betweenthe center of the ellipsoid and the most external occurrences. Thetwo most external occurrences are the coordinate axes of theellipsoid and in tandem with the Euclidean distances are used asparameters for a standard tri-axial ellipsoid equation (22). Thisellipsoid was then used to simulate Gaussian response curves ofthe species to the environmental data employed to resemble ecological theories of species responses to environmental conditions(5, 25–27). To visualize the impacts of occurrences curation inestimations, a second model was developed as described above,but without occurrences filtered, i.e., using all the reports available to us.Realized Niche (NR)The NR was estimated in a reduced calibration region, including only areas falling inside the NF model (Figure 1). In thesesub-regions, we used 16 remotely sensed variables summarizingland surface temperature (LST) and primary productivity (28).Specifically, we used MODIS data at 1 km spatial resolution,including day and night-time values of LST, and primaryproductivity in t
ecological niche modeling theory and methods are a key component. Therefore, apply-ing the concepts and tools from ecological niche modeling to disease biogeography and epidemiology will provide biologically sound and analytically robust descriptive and predictive analyses of disease distributions. As a case study, we explored the ecologically
dispersal biogeography) Three phases of historical biogeography are usually seen: Historical biogeography is the flipside to ecological biogeography Most of its practitioners are not geographers but systematists specializing on specific groups of organisms Historical Biogeography "What lives where and why? "
Jonathan B. Losos and Robert E. Ricklefs Robert MacArthur and Edward Wilson’s 1967 book, The Theory of Island Biogeography, is the dominant symbol of a transition that took place four de cades ago from descriptive to analytical approaches in ecol-ogy and biogeography. Change was in the air during the dynamic de cade
Glossopteris - Permian “fern” Mesosaurus - Permian freshwater reptile Cynognathus - Triassic land reptile Lystrosaurus - Triassic land reptile Biogeography includes . . . plate tectonics Jurassic Period (208 - 146 mya) “Modern”genera are first seen for Ginkgo Biogeography includ
ing the biogeography of the region, terrestrial bio-geographers have built upon the equilibrium theory of island biogeography (Macarthur et al. 1967). Luzon's biogography has been described by Merill (1923) when he identified the following flo-ral regions or subprovinces (Fig. 5). These are the 1) Eastern Luzon-Bicol peninsula region, 2) Luzon
5/18/22, 1:56 PM Quiz: Hawai'i Physical Geography: Biogeography - Treelines and Plant Succession . and the topic of Primary Plant Succession are classic themes in biogeography. So we combine them into one lab. . GPH 112 Hawaii Physical Geography Question on Treeline There will be a complication -- that lava flows have periodically wiped out .
Atherosclerotic Disease is a Pan Vascular Process Coronary Artery Disease (CAD) Non-coronary Atherosclerosis -Peripheral Artery Disease (PAD) -Lower extremity -Upper extremity (subclavian stenosis) -Carotid artery disease -Renal artery disease -Mesenteric artery disease -Aortic aneurysm -Vasculogenic Erectile Dysfunction Vascular disease is
Download novel the alchemist terjemahan. The alchemist (novel) plot. What books to read after the alchemist. The alchemist graphic novel free download. The alchemist novel in urdu pdf download. The alchemist novel price. The alchemist (novel) summary.
Albert Woodfox is a former Black Panther who spent 45 years unjustly incarcerated in a Louisiana State Penitentiary. He was released in 2016, having served more than 43 years in VROLWDU\ FRQ¿QHPHQW WKH ORQJHVW SHULRG RI VROLWDU\ FRQ¿QHPHQW in American prison history. Kano is a British rapper, songwriter and actor. Kano is one of the pioneers of grime music and culture. In 2004, Kano released .
