Abstract

The virus family Filoviridae includes 2 genera, the Marburg viruses and the Ebola viruses. The ecology of the filoviruses is poorly known, and indeed their host relationships remain completely unknown. An earlier effort prioritized mammalian taxa as to their possible status as the long-term coevolved reservoir of the filoviruses based on a coarse, regional classification of occurrences; here, we greatly refine the geographic data set for the mammalian taxa based on rich occurrence data sets and range interpolations from ecological niche models for each species involved. This improved detail permits a much more detailed inspection of distributional overlap patterns, and consequently a shorter list of candidate taxa—geographic analysis of 124 mammalian clades led to identification of 55 groups of interest that coincide spatially with known filovirus outbreaks, and fulfill the requirements of several additional assumptions. We discuss implications of our results for the search for the filovirus reservoir, and for research in African mammalogy.

Disease associated with members of the virus family Filoviridae has emerged sporadically since 1967, when Marburg virus caused an outbreak of hemorrhagic disease associated with exposure to primates imported (probably from Uganda) into Germany. Marburg and Ebola viruses have subsequently caused isolated cases or epidemics of hemorrhagic fever in humans or nonhuman primates across Africa (Anonymous 1978a, 1978b, 2001; Formenty et al. 1999; Georges et al. 1999; Gradon 2000; Heymann et al. 1980; Khan et al. 1999; Lahm et al. 2007; Leroy et al. 2002; Locsin 2002; Locsin and Matua 2002; Oyok et al. 2001a, 2001b; Weir 2001; Zeller 2000) and elsewhere when infected primates were moved for commercial purposes (Joffe and Haarhoff 2002; Kilpatrick 2001; Miranda et al. 1999; Rollin et al. 1999).

Given the high mortality associated with disease caused by filoviruses, considerable interest has focused on the identity of their sylvatic reservoir or reservoirs, here defined as the set of populations of animals or plants that sustains the pool of virus from which infections in humans and other primates have derived (Haydon et al. 2002). Several suites of tools have been brought to bear on this question. Field epidemiological studies of filovirus outbreaks (Baiter 2000; Conrad et al. 1978; Miranda et al. 1999; Monath 1999; Morvan et al. 2000; Murphy et al. 1990; Peters et al. 1993; Tucker et al. 2002; Weber and Rutala 2001) have been limited by the early deaths of primary human cases (who might otherwise have been able to recount their activities), but have established that filovirus infections from the natural reservoir occur in rural situations, outside of domiciles, and that single persons have usually been the initiation of outbreaks, which argues against involvement of arthropod vectors in the transmission process.

Further work has included laboratory tests of effects of infection on potential hosts (Chepurnov et al. 2001; Schou and Hansen 2000; Swanepoel et al. 1996; Turell et al. 1996). These studies appear to have established that no trace of successful infection and viral replication exists in inoculated plants and invertebrates (Swanepoel et al. 1996; Turell et al. 1996); that reptiles and birds show limited (if any) ability to be infected (Swanepoel et al. 1996); and that bats at least can become infected without developing frank disease, and can at least in some cases develop viremia (Swanepoel et al. 1996). However, conclusions from inoculation experiments have been limited because few species occurring in central Africa have been tested, no individuals from filovirus-endemic regions of central Africa have been tested, and some major mammal clades (e.g., Soricidae) have never been tested.

Finally, searches have been conducted for natural virus infections among animal species in localities where outbreaks have occurred (Germain 1978; Leirs et al. 1999; Leroy et al. 2005). Although these studies have resulted in thousands of individuals of more than 130 species of mammals being tested for filovirus infection (Peterson et al. 2004b), unambiguously positive results have been elusive (see “Discussion”). These searches have been based on convenience sampling of mammalian taxa, taking species that end up in traps and nets set in areas of interest, rather than focusing on full representation of the mammalian fauna or of a shorter list of priority taxa (Peterson et al. 2004b). As such, many species have been unnecessarily oversampled and many taxa of interest have not been sampled adequately or at all.

This paper constitutes a 3rd step in an effort to marshal a new suite of tools and approaches for improving the probability of identifying the natural reservoir(s) of filoviruses. A 1st step involved detailed characterization of the geography and ecology of filovirus disease occurrences (Peterson et al. 2004a, 2006). A 2nd step was to assess all mammal “clades” in the African mammal fauna for crude regional distributional coincidence with the 4 African filovirus species, distributed in western (Ebola Ivory Coast), central (Ebola Zaire), and eastern (Ebola Sudan and Marburg) Africa. This latter analysis resulted in identification of a set of mammalian taxa as potential filovirus reservoirs (Peterson et al. 2004b), but suffered from imprecise definition of distributional areas for mammalian taxa.

As such, this study aims to improve over the previous analysis in several ways. First, because of the lack of precise, compiled distributional data on central African mammals, the geographic filter used was quite coarse, asking, for example, if the clade has a representative in West Africa, rather than specifically whether the clade is represented along the Ivory Coast-Liberia border, where Ebola Ivory Coast occurs. Review of mammal collections and specimen records from many of the major African mammal collections of the world has allowed us to define much more precisely the distributions of many African mammalian taxa. Second, our assumptions about the age of the filovirus radiation may have been incorrect. The dramatic differentiation of Ebola and Marburg viruses in nucleotide and protein sequences had led to a focus at the level of mammalian genus to subfamily (Peterson et al. 2004b). However, the only study to date of evolutionary rates and processes in the group (Suzuki and Gojobori 1997) indicates a much younger origin, given relatively fast evolutionary rates in the filoviruses. Finally, we have been criticized for elimination of medium-sized mammals from our candidate reservoir lists. Hence, here, we use results from an exhaustive study of detailed occurrence data for African mammals and ecological niche modeling to assess species- and genus-level distributions of African mammal clades as regards fine-scale coincidence with known filovirus distributional areas.

Materials and Methods

Assumptions.—In this study, we largely follow the assumptions outlined in Peterson et al. (2004b), but with several important adjustments. Briefly, we assumed the following: First, the reservoir is a mammal (Leroy et al. 2005; Murphy et al. 1990; Murphy and Peters 1998; Peters et al. 1993; Swanepoel et al. 1996=). Second, the reservoir supports persistent, largely asymptomatic filovirus infections, which eliminates most primates and ungulates from consideration (Lahm et al. 2007). Third, the geographic distribution of the virus is restricted to the geographic distribution of the reservoir. Fourth, the reservoir has small body size, on the grounds that a larger animal would likely be more memorable, and thus have been noted by index cases who had some sort of contact. Fifth, a revolutionary link exists between the virus and its reservoir(s). Sixth, the reservoir is not commensal with humans. In the current study, we modified the assumptions of the previous study to allow inclusion of widespread species with ranges that include the ranges of all known African filovirus species, considering that many of these single “species” may actually be highly structured genetically, and represent complexes of related species. We also allowed inclusion of larger-sized mammals, up to about 10–15 kg. Finally, and most importantly, we used a considerably more detailed geographic filter in defining geographic coincidence, described in more detail below.

Occurrence data.—Occurrence data for each species listed in the original, coarse-filter analysis (Peterson et al. 2004b), plus additions based on consideration of widespread species and species of larger body size, were accumulated from natural history museum specimens (Table 1) by direct inspection of specimens and capture of data from specimen labels, by review of digital catalog data provided by collections personnel, or by consulting the Mammal Networked Information System (MaNIS) distributed biodiversity information database system (Stein and Wieczorek 2004).

Table 1

Natural history museum collections from which locality data were obtained for the distributional maps for mammal species, and how data were obtained.

Institution How data obtained 
American Muséum of Natural History Electronica 
Natural History Muséum (London) Manualb 
Los Angeles County Muséum of Natural History Electronic 
United States Natiónal Muséum of Natural History Electronic 
University of Michigan Muséum of Zoology Electronic 
Yale Peabody Muséum Electronic 
Muséum Natiónal d'Histoire Naturelle, Paris Manual, electronic 
Musée Royal de l'Afrique Centrale, Tervuren Manual, electronic 
Muséum fur Naturkunde, Berlin Manual 
Data sets served in MaNISc: University of Washington Burke Muse;um, Field Museum, Royal Ontario Museum, University of New Mexico Mueum of Southwestern Biology, California Academy of Sciences, University of Kansas Natural History Museum, Museum of Vertebrate Zoology Electronic 
Texas Tech University Electronic 
Louisiana State University Museum of Natural Science Electronic 
Institution How data obtained 
American Muséum of Natural History Electronica 
Natural History Muséum (London) Manualb 
Los Angeles County Muséum of Natural History Electronic 
United States Natiónal Muséum of Natural History Electronic 
University of Michigan Muséum of Zoology Electronic 
Yale Peabody Muséum Electronic 
Muséum Natiónal d'Histoire Naturelle, Paris Manual, electronic 
Musée Royal de l'Afrique Centrale, Tervuren Manual, electronic 
Muséum fur Naturkunde, Berlin Manual 
Data sets served in MaNISc: University of Washington Burke Muse;um, Field Museum, Royal Ontario Museum, University of New Mexico Mueum of Southwestern Biology, California Academy of Sciences, University of Kansas Natural History Museum, Museum of Vertebrate Zoology Electronic 
Texas Tech University Electronic 
Louisiana State University Museum of Natural Science Electronic 
a

Specimen data provided in electronic format by collections personnel.

b

Specimens examined and data recorded manually.

c

Specimen data accessed via query of the Mammal Networked Information System (Stein and Wieczorek 2004).

Data extracted from these sources (museum, catalog number, species, country, state, locality) were assigned geographic coordinates via consultation of a variety of electronic (Natiónal Geospatial-Intelligence Agency 2004; University of California, Santa Barbara 2003) and print (Davis and Misonne 1964) gazetteers. All georeferencing was checked for accuracy and consistency with taxonomic concepts by plotting occurrences on maps of Africa in Arc View version 3.2 (Environmental Systems Research Institute, Inc., Redlands, California); although some misidentifications and other errors may persist, the picture of species' distributional areas is consistent with authoritative references (Kingdon 1997; Rosevear 1969; Wilson and Reeder 1993).

Ecological niche modeling.—To provide further detail and an educated interpolation of likely distributional patterns for each species, we used ecological niche modeling. This general class of procedures is based on known occurrences of species, as they relate to digital raster geographic information system coverages that summarize potentially relevant ecological parameters. The idea is to identify a suite of ecological conditions within which the species in question can likely maintain populations without immigrational input (Grinnell 1917, 1924). The result is a picture of the species' potential geographic distribution, defined as the area meeting the species' ecological niche requirements that characterize known distributional areas (Soberón and Peterson 2005).

We used the Genetic Algorithm for Rule-set Prediction, or GARP (Stockwell and Noble 1992; Stockwell and Peters 1999), which has seen extensive testing and application to such questions (Peterson and Cohoon 1999; Stockwell and Peterson 2002a, 2002b, 2003). GARP is an evolutionary-computing approach that relates known occurrences of species to raster data layers summarizing relevant environmental parameters to create a model of the ecological niche of the species, which can in turn be used to identify a potential geographic distribution (Soberón and Peterson 2005). We used the 19 bioclimatic variables from the WorldClim global 0.17° data set (Hijmans et al. 2005), plus data on topography and land form, including elevation, slope, aspect, and compound topographic index (United States Geological Survey 2001) to characterize ecological landscapes. All data were resampled to 10 × 10-km spatial resolution for analysis.

Because our use of GARP is for visualization and interpolation purposes only, and because the technique has been documented in detail elsewhere (Anderson et al. 2002, 2003; Illoldi et al. 2004; Martx00ED;nez-Meyer et al. 2004; Ortega-Huerta and Peterson 2004; Peterson 2005; Soberón and Peterson 2004), we do not provide a full, detailed description herein. We used half of the available occurrence data for training models, and half to provide test data sets that characterized model success in predicting independent occurrence points. Using the GARP algorithm, we developed 100 replicate models for each species, and followed recent recommended protocols (Anderson et al. 2003) in using independent measures of omission and commission (type I and type II) prediction error to identify an optimal 10% of models from the 100 replicate models originally produced; the sum of these 10 models was taken as the best hypothesis of the species' distribution.

Procedures for applying filters.—Known and predicted distributions of each mammal species or set of mammal species making up broader clades were reviewed by ATP, MP, DSC, HL, and James N. Mills (see “Acknowledgments”) as a group, and compared with the known distribution of filovirus disease outbreaks (Peterson et al. 2004a; Fig. 1) across Africa. We considered 3 features of distributions of mammal clades: whether the geographic distribution of the mammal species or clade overlaps all of the known virus outbreak sites, a criterion that was considered necessary for further consideration; whether the mammal species or clade is distributed more broadly than the virus, and in which directions; and whether single mammal species (using taxonomy from Wilson and Reeder [1993]) cover the range of multiple virus species or whether multiple mammal species were represented within the distributional area of a particular filovirus species. The latter 2 features of mammalian distributions were not considered to be grounds on which mammal clades would be eliminated from consideration, but they are certainly relevant to the issue, because broad distributions can complicate bio-geographic interpretations (Page 2003). Mammal clades were identified based on traditional, hierarchical taxonomy (Wilson and Reeder 1993).

Fig. 1

Distributional summary of Ebola and Marburg viruses, with predicted distributions based on ecological niche models of outbreak coordinates (Peterson et al. 2004a). Ebola virus distribution shown as highest probability in black and intermediate probability in dark gray; Marburg virus distribution shown as highest probability in white and intermediate probability in light gray. Square, Ebola virus outbreak; triangle, Marburg virus outbreak.

Fig. 1

Distributional summary of Ebola and Marburg viruses, with predicted distributions based on ecological niche models of outbreak coordinates (Peterson et al. 2004a). Ebola virus distribution shown as highest probability in black and intermediate probability in dark gray; Marburg virus distribution shown as highest probability in white and intermediate probability in light gray. Square, Ebola virus outbreak; triangle, Marburg virus outbreak.

Results

In all, 124 mammalian clades were analyzed in this study, representing all genera identified in our previous study; 3 supergeneric clades that were discernable and that appeared to be of particular interest (Chalinolobus + Glauconycteris, Potamogale + Micropotamogale, and Heterohyrax + Proca-via + Dendrohyrax); clades holding species of larger body size, reflecting our modified maximum body size criterion; and 85 widespread single species with distributions that cover the entire known African distribution of the Filoviridae. These 124 clades are listed in Appendix I, along with their characteristics as they pertain to the present study.

Mammal clades showed variable coincidence or noncoincidence with filovirus occurrences (Fig. 2). In 2 of the examples shown (Hypsignathus monstrosus and Funisciurus), the geographic distribution of the mammal clade covers the entire range of Ebola virus, but does not overlap the entire known distribution of Marburg virus, particularly regarding its southernmost occurrence in Zimbabwe. The other 2 examples (Eidolon helvum and Heliosciurus) overlap the entire geographic distribution known for the filoviruses. As such, the former 2 examples would be discarded using the spatial filters in this analysis and the latter 2 would be retained.

Fig. 2

Four examples of distributional summaries for mammal taxa analyzed in this study. Shown are Hypsignathus monstrosus, Eidolon helvum, Funisciurus, and Heliosciurus. Predicted potential distributional areas are shown in gray, and known filovirus outbreaks in black outlines; occurrence points are shown as dotted circles for the 2 individual species.

Fig. 2

Four examples of distributional summaries for mammal taxa analyzed in this study. Shown are Hypsignathus monstrosus, Eidolon helvum, Funisciurus, and Heliosciurus. Predicted potential distributional areas are shown in gray, and known filovirus outbreaks in black outlines; occurrence points are shown as dotted circles for the 2 individual species.

Considering geographic coincidence with the 4 African filovirus species known to date, the distributions of 55 clades covered all virus distributional areas. Of 70 clades eliminated from consideration on grounds of noncoverage of virus distribution, 62–65 were invalidated based on noncoverage of Marburg virus distribution, 7 or 8 based on Ebola Zaire, 5 based on Ebola Ivory Coast, and 3 based on Ebola Sudan (ranges of numbers are provided because some uncertainty remains in the knowledge of distributions of some of the mammalian groups). The recognized range of Marburg virus is extended far to the south (to Zimbabwe) on the basis of a single occurrence, for which the site of exposure is not well understood (Conrad et al. 1978; Peterson et al. 2006). This southernmost Marburg outbreak alone was responsible for elimination of >50 of the clades. Thus, the 1975 Zimbabwe case becomes key in narrowing down the suite of possible reservoir clades. Conversely, if the site of this case were somehow to be in error, then these clades would need to see more consideration; however, the recent outbreak in Angola lends considerable credence to the Zimbabwean origin of the 1975 outbreak (Peterson et al. 2006).

The considerations of overextension, widespread mammal species, and multiple mammal species per viral distributional area are summarized in Appendix I. Only 9 clades showed no overextension; 110–112 clades were overextensive to the west, often to Senegal, where humid rain forest habitats are not present (some uncertainty was present here as to what constitutes overextension, hence the range of numbers), whereas overextension in other directions was less frequent. In 33 cases, mammal clades were overextensive in all 4 cardinal directions, clearly constituting species broadly distributed in the African tropics. Only in 4 cases were mammal species not widespread with respect to the distributions of virus species. Finally, multiple mammal species per viral species' distributional area were also common, occurring in 37 of the 124 mammal clades.

Discussion

We provide a short list of candidate mammal clades for the reservoir(s) of the filoviruses. Although not guaranteed to include the reservoir taxa, we believe that this list at least represents a high-probability set of candidates that can help to simplify, prioritize, and increase the likelihood of success of future reservoir searches. Our inferences regarding distributional areas of mammal species are based on interpretations of ecological niche models derived from locality data associated with natural history museum specimens. Many previous analyses have indicated that this analytical framework offers significant predictive ability regarding geographic distributions of species (Elith et al. 2006; Guisan et al. 2007, in press; Illoldi et al. 2004).

Assumptions and their implications.—Our earlier, coarse-scale analysis of distributions of African mammals and filoviruses (Peterson et al. 2004b) incorporated a series of assumptions based on knowledge of filovirus biology and ecology (Murphy 1978; Murphy et al. 1990; Murphy and Peters 1998; Peters et al. 1993). The present analysis differs from the original in 3 ways. First, given the likely younger age of the filovirus clade (Suzuki and Gojobori 1997), we relaxed the assumption regarding cospeciation and close phylogenetic and geographic correspondence with currently recognized mammal species. As a consequence, we included single, widespread mammal species with distributions that overlap the distributions of all of the filovirus species. Second, given the possibility of transmission without direct contact, such as via contact with feces, we relaxed the assumption of small body size to an assumption of small-to-medium body size. Finally, as described below, we imposed a much finer, more detailed geographic filter. Instead of simply requiring representation in western, central, and eastern Africa, we developed fine-scale distributional maps for each component species in each clade under consideration using distributional data associated with natural history museum specimens and ecological niche modeling. This last filter was much more restrictive than our original, regional filter, and produced a significantly shorter set of candidate clades for being the filovirus reservoir.

Admitting species-level taxa in our list brought into the study 85 widespread species with distributions covering all of the African tropics. Although we have relaxed the assumption that the reservoir of the combined Ebola viruses would be a supraspecific taxon, we still suspect that the hosts and viruses have undergone a period of coevolution similar to that shown for other small mammal-borne hemorrhagic fever viruses such as the hantaviruses and arenaviruses. Given a perfect understanding of host and virus taxonomy, such a relationship might predict a genus-level host taxon that would encompass the distribution of all filovirus reservoirs with a single host species within that genus coinciding with the geographic range of each virus species (Ebola Sudan, Ebola Ivory Coast, Ebola Zaire, and Marburg). However, because the radiation of the filoviruses appears to be relatively young and understanding of the taxonomy of African small mammals is incomplete, it is possible that the clade that includes the reservoirs for all filoviruses or some subset of them could be currently recognized as a single widespread species. Upon closer examination, this “species” may be shown to consist of several genetically distinct cryptic species or geographic races (Corti et al. 2004; Lavrenchenko et al. 2004), each associated with a separate filovirus. By admitting medium-sized species as well as those of smaller size, numerous groups were added to consideration, including small carnivores (e.g., Galerella sanguinea), pangolins (Manis), and even some strepsirrhine primates (e.g., Gala go).

This 2nd exercise in identification of high-priority taxa for further study leaves only 55 clades on the list, considering species, genera, and a few higher taxa. A significant portion of these clades, nonetheless, has distributions broader than that of the filovirus species, are widespread mammalian taxa with respect to virus species, or present multiple mammal species present per virus species' distributional area. Although we do not consider these points to invalidate a candidate reservoir taxon, they are certainly relevant to the issue of which mammalian taxa are possible candidate reservoir taxa.

The final list for study is quite focused on Chiroptera. Indeed, of the 55 clades, 36 are bats, 6 are small-bodied carnivores, and 6 are rodents; the remainder includes strepsirrhine primates, hyraxes, and pangolins. Shrews are difficult to treat; for example, both Sylvisorex and Crocidura satisfy all of the criteria for inclusion, but the species-level taxonomy is poorly resolved, especially in the latter genus, and several candidate subclades may be included.

Implications for mammalogy in Africa.—The existence of filoviruses and arenaviruses that cause hemorrhagic fevers in African mammals should be a sobering thought for mammalogists working in Africa. Given the close contact involved in preparation activities, transmission of filoviruses to mammalogists could occur easily as they prepare mammal specimens. Although such has not yet to our knowledge occurred, appropriate precautions should be taken, following accepted guidelines (Mills et al. 1995a, 1995b).

However, mammalogists working in Africa have much to offer and contribute to the ongoing efforts to identify the filovirus reservoir. As discussed in our previous paper (Peterson et al. 2004b) and above, researchers in on-site investigations and searches for filovirus-infected mammals have used opportunistic methods, and have not conducted systematic efforts to locate and sample all mammal species present at the sites. Prioritizations based on range and other assumptions, such as that presented herein, can identify taxa on which to focus these efforts. As such, involvement of professional mammalogists, who could direct efforts to obtain series of each important taxon, would be critical.

Conclusions.—A recent paper reported finding evidence of Ebola virus infection in 3 species of fruit bats (H. monstrosus, Epomops franqueti, and Myonycteris torquata) in Gabon and the Republic of the Congo (Leroy et al. 2005). Although no virus could be isolated, detection of specific antibodies for Ebola virus and small quantities of viral RNA suggests that these bat species may be involved in hosting the virus. All 3 species were also retained in our list of taxa that represent potential reservoirs for filovirus (see Appendix I). All 3 or their congeners are widespread and have distributions that include all areas where Ebola virus has been documented.

Epomops franqueti and H. monstrosus even have, in our terminology, overextensive distributions westward. None extends broadly to the areas where Marburg virus occurs. This result is interesting because 1 of our assumptions was that Ebola and Marburg virus are carried by 2 different, but closely related, species, or at least 2 distinct clades within a single species. For example, the related species Epomophorus labiatus has a distribution that does not cover the area where Ebola is expected, but that does include most of the modeled Marburg area, except in Zambia and Zimbabwe.

Considerable attention has been paid to pteropodid phylog-eny in the literature on mammalian systematics (Giannini and Simmons 2003; Hollar and Springer 1997; Juste-B. et al. 1999; Romagnoli and Springer 2000). However, taxon representation is incomplete in all of these studies, and inconsistent between them, because each omits at least 1 of the 3 genera. What is more, even among the 2 most recent studies (Giannini and Simmons 2003; Romagnoli and Springer 2000), topologies differed significantly, making interpretation difficult. More generally, the questions we pose require a species-level resolution to the phylogenetic tree, so we must await publication of a more finely resolved phylogenetic picture for this family. The phylogenetic relationships among the 3 species that tested positive in Leroy et al. (2005) will be very informative.

Unless these species are very closely related, at least 2 of them were probably positive via spillover to other species from the real reservoir species, which could explain the lack of detection of live virus. Detection of viral material in wild bat species also lends further support to the results of an earlier study, in which experimental inoculation with Ebola virus resulted in successful infection and replication in 3 other bat species, Epomophorus wahlbergi, Mops condylurus, Chaerephon pumilus (Swanepoel et al. 1996). However, it should be pointed out that the latter 2 species are not fruit bats and belong to a different suborder. This situation again shows the potential for spillover between species, but if spillover happens relatively easy, it raises the question of why “spillover positives” were never detected in earlier studies.

Clearly, the ideal next step in identifying the filovirus reservoir will be a multidisciplinary challenge. Molecular genetic studies may allow more accurate estimates of the age of the virus radiation, which may allow a further narrowing of the field of candidate taxa. Most importantly, however, focused and directed searches using nucleic acid probes to examine mammalian tissue samples have an ever-higher probability of success.

Acknowledgments

We thank our valued colleague J. N. Mills for his fundamental contributions to this effort—from molding the concept to interpreting the results of the analysis. Thanks also to S. Goodman, J. Peterhans Kerbis, P. Gaubert, W. Stanley, R. Hutterer, and W. Sechrest for their generous provision of occurrence data associated with their ongoing research programs on mammal faunas of Africa, and to W. Sechrest and J. Peterhans Kerbis for helpful comments on a draft of the manuscript. J. Montgomery provided helpful insights from his experience with these viruses. We gratefully acknowledge the generous welcome that we were afforded at many natural history museums (Table 1), and the kind support and assistance offered by the staff of each institution. Data contributed by the IUCN Global Mammal Assessment, which in collaboration with the upcoming book project The Mammals of Africa (edited by J. Kingdon, T. Butynski, and D. Happold), provided distributional data on some of the African small mammals analyzed here. E. Brown and M. Howard-Heretakis contributed long hours georeferencing occurrence localities. Much of the work in this effort was supported by a contract from the United States Department of Defense. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention or the United States Government.

Appendix I

Summary of mammalian taxa that represent potential reservoirs of filoviruses in Africa. Clade indicates the species or set of species under consideration, for which the higher taxonomy and component species are then provided. The 4 aspects of virus-potential host codistribution described in the “Materials and Methods” are detailed: virus species missing, areas of overprediction, mammal species widespread, and multiple mammal species per virus distributional area. Occasional notes on habitats are provided in the last column. M, I, Z, and S indicate Marburg, Ebola Ivory Coast, Ebola Zaire, and Ebola Sudan, respectively. In areas of overprediction, N, S, E, W, and combinations thereof indicate cardinal directions. —, no; +, yes.a

Clade Higher taxonomy Component species Virus species missing Areas of overprediction Mammal species widespread? Multiple mammal species per virus distributional area? Habitats 
Afrosoricida: Tenrecidae; Micropotamogale lamottei— — — Aquatic habits 
 Potamogalinae M. ruwenzorii, Potamogale      
  velox      
Afrosoricida: Soricidae; Crocidura — — —  
 Crocidurinae       
 Sylvisorex — NE   
Chiroptera: Pteropodidae; Eidolon helvum — W, E, NE —  
 Pteropodinae       
 Epomophorus labiatus I, Z, M W, E, NE —  
 Epomops ∼M  
 Epomops franqueti ∼M  Forest—savanna 
       ecotone 
 Hypsignathus monstrosus   
 Micropteropus W, NE Forest—savanna 
       ecotone 
10  Myonycteris torquata — —  
11  Rousettus aegyptiacus — W, S, E, NE —  
12  R. angolensis — W, NE —  
13 Chiroptera: Pteropodidae; Megaloglossus woermanni — —  
 Macroglossinae       
14 Chiroptera: Emballonuridae Coleura afra — W, E — Caves 
15  Saccolaimus peli —  
16  Taphozous — W, S, E, NE  
17 Chiroptera: Nycteridae Nycteris — W, S, E, NE  
18  N. arge —  
19  N. grandis — W, E —  
20  N. hispida — W, S, E, NE —  
21  N. intermedia —  
22  N. macrotis — W, S, E, NE —  
23  N. major M, S ∼W —  
24  N. nana —  
25 Chiroptera: Megadermatidae Lavia frons W, E, NE —  
26 Chiroptera: Rhinolophidae; Rhinolophus — W, S, E, NE  
 Rhinolophinae       
27  R. alcyone — —  
28  R. landed — W, E, NE —  
29  R. simulator —  
30 Chiroptera: Rhinolophidae; Hipposideros — W, S, E, NE  
 Hipposiderinae       
31  H. commersoni — W, S, E, NE —  
32  H. cyclops M, ∼Z —  
33  H. fuliginosus —  
34  H. ruber — W, E, NE —  
35 Chiroptera: Vespertilionidae; Kerivoula — W, S, E, NE  
 Kerivoulinae       
36  K. lanosa — W, S, E, NE —  
37 Chiroptera: Vespertilionidae; Chalinolobus including — W, E, NE  
 Vespertilioninae C. poensis      
38  Chalinolobus excluding W, E  
  C. poensis      
39  Chalinolobus poensis W, ∼E —  
40  Chalinolobus variegatus — W, E — Woodlands 
41  Eptesicus — W, S, E, NE  
42  E. capensis — W, S, E, ∼NE —  
43  E. guineensis —  
44  E. rendalli W, E, NE — Woodlands and savanna 
45  E. somalicus — W, NE — Woodlands and savanna 
46  E. tenuipinnis —  
47  Mimetillus moloneyi W, E —  
48  Myotis bocagei — W, E, NE —  
49  Pipistrellus — W, S, E, NE  
50  P. eisentrauti W, E —  
51  P. nanulus W, E —  
52  P. nanus — W, S, E, NE —  
53  P. rustic us — W, S, E, NE —  
54  Scotoecus albofuscus — Woodlands 
55  S. hirundo — W, NE — Woodlands 
56  Scotophilus dinganii — W, S, NE —  
57  S. nigrita — W, NE —  
58  S. nux — —  
59 Chiroptera: Vespertilionidae; Miniopterinae Miniopterus — W, S, E, NE  
60  M. schreibersii — W, S, E, NE —  
61 Chiroptera: Molossidae Chaerephon — W, S, E, NE  
62  C. aloysiisabaudiae —  
63  C. bemmeleni —  
64  C. nigeriae — W, NE — Woodlands 
65  C. pumilus — W, S, E, NE —  
66  Mops — W, S, E, NE  
67  M. brachypterus —  
68  M. condylurus — W, S, E, NE —  
69  M. congicus M, Z ∼W —  
70  M. midas W, E, NE — Woodlands 
71  M. nanulus W, NE —  
72  M. thersites —  
73  Myopterus  
74  M. whitleyi I, M — — —  
75  Tadarida I, Z W, S, E, NE  
76  T. aegyptiaca I, Z W, S, E, NE —  
77 Primates: Lorisidae Perodicticus potto —  
78 Primates: Galagidae Galago — W, E  
79  G. senegalensis — W, E —  
80  Galagoides demidojf —  
81 Carnivora: Felidae; Felinae Profelis aurata — —  
82 Carnivora: Herpestidae; Herpestinae Atilax paludinosus    Aquatic habits 
83  Crossarchus  
84  Galerella sanguinea — W, S, E, NE —  
85 Carnivora: Viverridae; Nandiniinae Nandinia binotata  W, E   
86 Carnivora: Viverridae; Civettictis civetta  W, S, E, NE —  
 Viverrinae       
87  Genetta — W, S, E, NE  
88  G. maculata — W, S, E, NE —  
89 Hyracoidea: Procaviidae Dendrohyrax, Heterohyrax brucei, Procavia capensis — W, S, E, NE  
90  Dendrohyrax W, E —  
91  D. dor sails —  
92 Pholidota: Manidae Manis — W, S, E  
93  M. gigantea W, E —  
94  M. tetradactyla —  
95  M. trlcuspls —  
96 Rodentia: Sciuridae; Sciurinae Funlsclurus  
97  F. pyrropus —  
98  Heliosciurus — W, E, NE Woodlands and 2nd growth 
99  H. gambianus  W, E, NE — Woodlands and 2nd growth 
100  H. rufobrachium W, E  Woodlands and 2nd growth 
101  Protoxerus  
102  P. stangeri —  
103  Xerus ∼M W, S, E, NE Forest and savanna 
104  X. erythropus M, Z W, NE — Forest and savanna 
105 Rodentia: Nesomyidae; Cricetomyinae Cricetomys emini W, E —  
106  C. gambianus — W, E —  
107 Rodentia: Muridae; Murinae Grammomys — W, S, E, NE  
108  G. dolichurus — W, S, E, NE   
109  G. poensis (formerly rutilansW, NE   
110  Hybomys M, S —  
111  Hylomyscus W, E  
112  Lemniscomys — W, S, E, NE  
113  L. striatus W, NE —  
114  Lophuromys W, E, NE  
115  L. sikapusi W, E —  
116  Malacomys  
117  M. longipes —  
118  Praomys W, E  
119  P. jacksoni I, M W, E —  
120  P. tullbergi —  
121 Rodentia: Anomaluridae; Anomalurinae Anomalurus  
122  A. beecrofti M, S —  
123 Rodentia: Anomaluridae; Zenkerellinae Idiurus  
124 Rodentia: Gliridae; Graphiurinae Graphiurusb  W, S, E, NE  
Clade Higher taxonomy Component species Virus species missing Areas of overprediction Mammal species widespread? Multiple mammal species per virus distributional area? Habitats 
Afrosoricida: Tenrecidae; Micropotamogale lamottei— — — Aquatic habits 
 Potamogalinae M. ruwenzorii, Potamogale      
  velox      
Afrosoricida: Soricidae; Crocidura — — —  
 Crocidurinae       
 Sylvisorex — NE   
Chiroptera: Pteropodidae; Eidolon helvum — W, E, NE —  
 Pteropodinae       
 Epomophorus labiatus I, Z, M W, E, NE —  
 Epomops ∼M  
 Epomops franqueti ∼M  Forest—savanna 
       ecotone 
 Hypsignathus monstrosus   
 Micropteropus W, NE Forest—savanna 
       ecotone 
10  Myonycteris torquata — —  
11  Rousettus aegyptiacus — W, S, E, NE —  
12  R. angolensis — W, NE —  
13 Chiroptera: Pteropodidae; Megaloglossus woermanni — —  
 Macroglossinae       
14 Chiroptera: Emballonuridae Coleura afra — W, E — Caves 
15  Saccolaimus peli —  
16  Taphozous — W, S, E, NE  
17 Chiroptera: Nycteridae Nycteris — W, S, E, NE  
18  N. arge —  
19  N. grandis — W, E —  
20  N. hispida — W, S, E, NE —  
21  N. intermedia —  
22  N. macrotis — W, S, E, NE —  
23  N. major M, S ∼W —  
24  N. nana —  
25 Chiroptera: Megadermatidae Lavia frons W, E, NE —  
26 Chiroptera: Rhinolophidae; Rhinolophus — W, S, E, NE  
 Rhinolophinae       
27  R. alcyone — —  
28  R. landed — W, E, NE —  
29  R. simulator —  
30 Chiroptera: Rhinolophidae; Hipposideros — W, S, E, NE  
 Hipposiderinae       
31  H. commersoni — W, S, E, NE —  
32  H. cyclops M, ∼Z —  
33  H. fuliginosus —  
34  H. ruber — W, E, NE —  
35 Chiroptera: Vespertilionidae; Kerivoula — W, S, E, NE  
 Kerivoulinae       
36  K. lanosa — W, S, E, NE —  
37 Chiroptera: Vespertilionidae; Chalinolobus including — W, E, NE  
 Vespertilioninae C. poensis      
38  Chalinolobus excluding W, E  
  C. poensis      
39  Chalinolobus poensis W, ∼E —  
40  Chalinolobus variegatus — W, E — Woodlands 
41  Eptesicus — W, S, E, NE  
42  E. capensis — W, S, E, ∼NE —  
43  E. guineensis —  
44  E. rendalli W, E, NE — Woodlands and savanna 
45  E. somalicus — W, NE — Woodlands and savanna 
46  E. tenuipinnis —  
47  Mimetillus moloneyi W, E —  
48  Myotis bocagei — W, E, NE —  
49  Pipistrellus — W, S, E, NE  
50  P. eisentrauti W, E —  
51  P. nanulus W, E —  
52  P. nanus — W, S, E, NE —  
53  P. rustic us — W, S, E, NE —  
54  Scotoecus albofuscus — Woodlands 
55  S. hirundo — W, NE — Woodlands 
56  Scotophilus dinganii — W, S, NE —  
57  S. nigrita — W, NE —  
58  S. nux — —  
59 Chiroptera: Vespertilionidae; Miniopterinae Miniopterus — W, S, E, NE  
60  M. schreibersii — W, S, E, NE —  
61 Chiroptera: Molossidae Chaerephon — W, S, E, NE  
62  C. aloysiisabaudiae —  
63  C. bemmeleni —  
64  C. nigeriae — W, NE — Woodlands 
65  C. pumilus — W, S, E, NE —  
66  Mops — W, S, E, NE  
67  M. brachypterus —  
68  M. condylurus — W, S, E, NE —  
69  M. congicus M, Z ∼W —  
70  M. midas W, E, NE — Woodlands 
71  M. nanulus W, NE —  
72  M. thersites —  
73  Myopterus  
74  M. whitleyi I, M — — —  
75  Tadarida I, Z W, S, E, NE  
76  T. aegyptiaca I, Z W, S, E, NE —  
77 Primates: Lorisidae Perodicticus potto —  
78 Primates: Galagidae Galago — W, E  
79  G. senegalensis — W, E —  
80  Galagoides demidojf —  
81 Carnivora: Felidae; Felinae Profelis aurata — —  
82 Carnivora: Herpestidae; Herpestinae Atilax paludinosus    Aquatic habits 
83  Crossarchus  
84  Galerella sanguinea — W, S, E, NE —  
85 Carnivora: Viverridae; Nandiniinae Nandinia binotata  W, E   
86 Carnivora: Viverridae; Civettictis civetta  W, S, E, NE —  
 Viverrinae       
87  Genetta — W, S, E, NE  
88  G. maculata — W, S, E, NE —  
89 Hyracoidea: Procaviidae Dendrohyrax, Heterohyrax brucei, Procavia capensis — W, S, E, NE  
90  Dendrohyrax W, E —  
91  D. dor sails —  
92 Pholidota: Manidae Manis — W, S, E  
93  M. gigantea W, E —  
94  M. tetradactyla —  
95  M. trlcuspls —  
96 Rodentia: Sciuridae; Sciurinae Funlsclurus  
97  F. pyrropus —  
98  Heliosciurus — W, E, NE Woodlands and 2nd growth 
99  H. gambianus  W, E, NE — Woodlands and 2nd growth 
100  H. rufobrachium W, E  Woodlands and 2nd growth 
101  Protoxerus  
102  P. stangeri —  
103  Xerus ∼M W, S, E, NE Forest and savanna 
104  X. erythropus M, Z W, NE — Forest and savanna 
105 Rodentia: Nesomyidae; Cricetomyinae Cricetomys emini W, E —  
106  C. gambianus — W, E —  
107 Rodentia: Muridae; Murinae Grammomys — W, S, E, NE  
108  G. dolichurus — W, S, E, NE   
109  G. poensis (formerly rutilansW, NE   
110  Hybomys M, S —  
111  Hylomyscus W, E  
112  Lemniscomys — W, S, E, NE  
113  L. striatus W, NE —  
114  Lophuromys W, E, NE  
115  L. sikapusi W, E —  
116  Malacomys  
117  M. longipes —  
118  Praomys W, E  
119  P. jacksoni I, M W, E —  
120  P. tullbergi —  
121 Rodentia: Anomaluridae; Anomalurinae Anomalurus  
122  A. beecrofti M, S —  
123 Rodentia: Anomaluridae; Zenkerellinae Idiurus  
124 Rodentia: Gliridae; Graphiurinae Graphiurusb  W, S, E, NE  
a

Species eliminated from consideration on the basis of being primarily open-habitat species include the following: Taphozous mauritianus, Rhinolophus fumigatus, Hipposideros coffer, Leptailurus serval, Helogale, Herpestes, Ichneumia albicauda, Mungos, Aonyx, Mellivora capensis, Ictonyx striatus, Dendromus, and Mastomys.

b

Graphiurus lorraineus could be treated as an additional candidate clade, but information available on its distribution and variation is incomplete.

Literature Cited

Anderson
R.P.
Gómez-Laverde
Peterson A.T.
.
2002
.
Geographical distributions of spiny pocket mice in South America: insights from predictive models
.
Global Ecology and Biogeography
 
11
:
131
141
.
Anderson
R. P.
Lew
D.
Peterson
A. T.
.
2003
.
Evaluating predictive models of species' distributions: criteria for selecting optimal models
.
Ecological Modelling
 
162
:
211
232
.
Anonymous.
1978a
.
Ebola haemorrhagic fever in Sudan, 1976
.
Bulletin of the World Health Organization
 
56
:
247
270
.
Anonymous.
1978b
.
Ebola haemorrhagic fever in Zaire, 1976
.
Bulletin of the World Health Organization
 
56
:
271
293
.
Anonymous.
2001
.
Outbreak of Ebola haemorrhagic fever, Uganda, August 2000-January 2001
.
Weekly Epidemiological Record
 
76
:
41
48
.
Balter
M.
2000
.
Emerging diseases—on the trail of Ebola and Marburg viruses
.
Science
 
290
:
923
925
.
Chepurnov
A. A.
Dadaeva
A. A.
Kolesnikov
S. I.
.
2001
.
Study of the pathogenesis of Ebola fever in laboratory animals with different sensitivity to this virus
.
Bulletin of Experimental Biology and Medicine
 
132
:
1182
1186
.
Conrad
J. L.
et al
.
1978
.
Epidemiologic investigation of Marburg virus disease, southern Africa, 1975
.
American Journal of Tropical Medicine and Hygiene
 
27
:
1210
1215
.
Corti
M.
Castiglia
R.
Annesi
F.
Verheyen
W.
.
2004
.
Mitochondrial sequences and karyotypes reveal hidden diversity in African pouched mice (subfamily Cricetomyinae, genus Saccostomus)
.
Journal of Zoology (London)
 
262
:
413
424
.
Davis
D. H. S.
Misonne
X.
.
1964
.
Gazetteer of collecting localities of African rodents
.
Documentation Zoologique 7
 .
Musee Royal de l'Afrique Centrale
,
Tervuren, Belgium
.
Elith
J.
et al
.
2006
.
Novel methods improve prediction of species' distributions from occurrence data
.
Ecography
 
29
:
129
151
.
Formenty
P.
et al
.
1999
.
Ebola virus outbreak among wild chimpanzees living in a rain forest of Côte dTvoire
.
Journal of Infectious Diseases
 
179
:
S120
S126
.
Georges
A. J.
et al
.
1999
.
Ebola hemorrhagic fever outbreaks in Gabon, 1994–1997: epidemiologic and health control issues
.
Journal of Infectious Diseases
 
179
:
S65
S75
.
Germain
M.
1978
.
Collection of mammals and arthropods during the epidemic of haemorrhagic fever in Zaire
. Pp.
185
189
in
Ebola virus haemorrhagic fever
  (
Pattyn
S. R.
, ed.).
Elsevier
,
Amsterdam, The Netherlands
.
Giannini
N. P.
Simmons
N. B.
.
2003
.
A phylogeny of megachiropteran bats (Mammalia: Chiroptera: Pteropodidae) based on direct optimization analysis of one nuclear and four mitochondrial genes
.
Cladistics
 
19
:
496
511
.
Gradon
J.
2000
.
An outbreak of Ebola virus: lessons for everyday activities in the intensive care unit
.
Critical Care Medicine
 
28
:
284
285
.
Grinnell
J.
1917
.
Field tests of theories concerning distributional control
.
American Naturalist
 
51
:
115
128
.
Grinnell
J.
1924
.
Geography and evolution
.
Ecology
 
5
:
225
229
.
Guisan
A.
Elith
J.
Graham
C. H.
Phillips
S. J.
Peterson
A. T.
Zimmermann
N. E.
.
In press
.
What matters for predicting spatial distributions of trees: techniques, data, or species' characteristics?
Ecological Monographs
 .
Guisan
A.
Graham
C.H.
Elith
J.
Huettman
F.
Species Distribution Modelling GroupNCEAS
.
2007
.
Sensitivity of predictive species distribution models to change in grain size
.
Diversity and Distributions
 
13
:
332
340
.
Haydon
D. T.
Cleaveland
S.
Taylor
L. H.
Laurenson
M. K.
.
2002
.
Identifying reservoirs of infection: a conceptual and practical challenge
.
Emerging Infectious Diseases
 
8
:
1468
1473
.
Heymann
D. L.
Weisfeld
J. S.
Webb
P. A.
Johnson
K. M.
Cairns
T.
Berquist
H.
.
1980
.
Ebola hemorrhagic fever: Tandala, Zaire, 1977–1978
.
Journal of Infectious Diseases
 
142
:
372
376
.
Hijmans
R. J.
Cameron
S.
Parra
J.
.
2005
.
WorldClim, version 1.3
 . http://biogeo.berkeley.edu/worldclim/worldclim.htm.
Accessed 1 January 2006
.
Hollar
L. J.
Springer
M. S.
.
1997
.
Old World fruitbat phylogeny: evidence for convergent evolution and an endemic African clade
.
Proceedings of the Natiónal Academy of Sciences USA
 
94
:
5716
5721
.
Illoldi
P.
SAnchez-Cordero
V.
Peterson
A. T.
.
2004
.
Predicting distributions of Mexican mammals using ecological niche modeling
.
Journal of Mammalogy
 
85
:
658
662
.
Joffe
H.
Haarhoff
G.
.
2002
.
Representations of far-flung illnesses: the case of Ebola in Britain
.
Social Science & Medicine
 
54
:
955
969
.
Juste
B. J.
Álvarez
Y.
Tabarés
E.
Garrido-Pertierra
A.
Ibanez
C.
Bautista
J. M.
.
1999
.
Phylogeography of African fruitbats (Megachiroptera)
.
Molecular Phylogenetics and Evolution
 
13
:
596
604
.
Khan
A. S.
et al
.
1999
.
The reemergence of Ebola hemorrhagic fever, Democratic Republic of the Congo, 1995
.
Journal of Infectious Diseases
 
179
:
S76
S86
.
Kilpatrick
K.
2001
.
Canada's Ebola scare over but questions just beginning
.
Canadian Medical Association Journal
 
164
:
1031
-
1032
.
Kingdon
J.
1997
.
The Kingdon field guide to African mammals
 .
Academic Press
,
London, United Kingdom
.
Lahm
S. A.
Kombila
M.
Swanepoel
R.
Barnes
R. F. W.
.
2007
.
Morbidity and mortality of wild animals in relation to outbreaks of Ebola hemorrhagic fever in Gabon, 1994–2003
.
Transactions of the Royal Society of Tropical Medicine and Hygiene
 
101
:
64
78
.
Lavrenchenko
L. A.
et al
.
2004
.
Divergent and reticulate processes in evolution of Ethiopian Lophuromys flavopunctatus species complex: evidence from mitochondrial and nuclear DNA differentiation patterns
.
Biological Journal of the Linnaean Society
 
83
:
301
316
.
Leirs
H.
et al
.
1999
.
Search for the Ebola virus reservoir in Kikwit, Democratic Republic of the Congo: reflections on a vertebrate collection
.
Journal of Infectious Diseases
 
179
:
S155
S163
.
Leroy
E.
Souquiere
S.
Rouquet
P.
Drevet
D.
.
2002
.
Reemergence of Ebola haemorrhagic fever in Gabon
.
Lancet
 
359
:
712
.
Leroy
E. M.
et al
.
2005
.
Fruit bats as reservoirs of Ebola virus
.
Nature
 
438
:
575
576
.
Locsin
R. C.
2002
.
Ebola at Mbarara, Uganda: aesthetic expressions of the lived worlds of people waiting to know
.
Nursing Science Quarterly
 
15
:
123
130
.
Locsin
R. C
Matua
A. G.
.
2002
.
The lived experience of waiting-to-know: Ebola at Mbarara, Uganda—hoping for life, anticipating death
.
Journal of Advanced Nursing
 
37
:
173
181
.
Martínez-Meyer
E.
Peterson
A. T.
Hargrove
W. W.
.
2004
.
Ecological niches as stable distributional constraints on mammal species, with implications for Pleistocene extinctions and climate change projections for biodiversity
.
Global Ecology and Bio-geography
 
13
:
305
314
.
Mills
J. N.
Childs
J. E.
Ksiazek
T. G.
Peters
C. J.
Velleca
W. M.
.
1995a
.
Methods for trapping and sampling small mammals for virologic testing
 .
Department of Health and Human Services
,
Atlanta, Georgia
.
Mills
J. N.
et al
.
1995b
.
Guidelines for working with rodents potentially infected with hantavirus
.
Journal of Mammalogy
 
76
:
716
722
.
Miranda
M. E.
et al
.
1999
.
Epidemiology of Ebola (subtype Reston) virus in the Philippines, 1996
.
Journal of Infectious Diseases
 
179
:
S115
S119
.
Monath
T. P.
1999
.
Ecology of Marburg and Ebola viruses: speculations and directions for future research
.
Journal of Infectious Diseases
 
179
:
S127
S138
.
Morvan
J. A.
Nakoune
E.
Deubel
V.
Colyn
M.
.
2000
.
Ebola virus and forest ecosystem
.
Bulletin de la Société de Pathologie Exotique
 
93
:
172
175
.
Murphy
F. A.
1978
.
Pathology of Ebola virus infection
. Pp.
43
49
in
Ebola virus haemorrhagic fever
  (
Pattyn
S. R.
, ed.).
Elsevier
,
Amsterdam, The Netherlands
.
Murphy
F. A.
Kiley
M. P.
Fisher-Hoch
S. P.
.
1990
.
Filoviridae: Marburg and Ebola viruses
. Pp.
933
942
in
Virology
  (
Fields
B. N.
Knipe
D. M.
, eds.).
Raven Press, Ltd.
,
New York
.
Murphy
F. A.
Peters
C. J.
.
1998
.
Ebola virus: where does it come from and where is it going?
Pp.
375
410
in
Emerging infections: biomedical research reports
  (
Krause
R. M.
, ed.).
Academic Press
,
San Diego, California
.
Natiónal Geospatial-Intelligence Agency
.
2004
.
GEOnet Names Server
 , http://earth-info.nga.mil/gns/html/
Accessed 1 November 2004
.
Ortega-Huerta
M. A.
Peterson
A. T.
.
2004
.
Modelling spatial patterns of biodiversity in northeastern Mexico
.
Diversity and Distributions
 
10
:
39
54
.
Oyok
T.
et al
.
2001a
.
Outbreak of Ebola hemorrhagic fever— Uganda, August 2000-January 2001
.
Archives of Dermatology
 
137
:
839
840
.
Oyok
T.
et al
.
2001b
.
Outbreak of Ebola hemorrhagic fever— Uganda, August 2000-January 2001 (reprinted from MMWR, vol 50, 73–77, 2000)
.
Jama-Journal of the American Medical Association
 
285
:
1010
1012
.
Page
R. D. M.
2003
.
Tangled trees: phylogeny, cospeciation, and coevolution
 .
University of Chicago Press
,
Chicago, Illinois
.
Peters
C. J.
et al
.
1993
.
Filoviruses
. Pp.
159
175
in
Emerging viruses
  (
Morse
S. S.
, ed.).
Oxford University Press
,
Oxford, United Kingdom
.
Peterson
A. T.
2005
.
Kansas Gap Analysis: the importance of validating distributional models before using them
.
Southwestern Naturalist
 
50
:
230
236
.
Peterson
A. T.
Bauer
J. T.
Mills
J. N.
.
2004a
.
Ecological and geographic distribution of filovirus disease
.
Emerging Infectious Diseases
 
10
:
40
47
.
Peterson
A. T.
Carroll
D.
Mills
J. N.
.
2004b
.
Potential mammalian filovirus reservoirs
.
Emerging Infectious Diseases
 
10
:
2073
2081
.
Peterson
A. T.
Cohoon
K. C.
.
1999
.
Sensitivity of distributional prediction algorithms to geographic data completeness
.
Ecological Modelling
 
117
:
159
164
.
Peterson
A. T.
Lash
R. R.
Carroll
D. S.
Johnson
K. M.
.
2006
.
Geographic potential for outbreaks of Marburg hemorrhagic fever
.
American Journal of Tropical Medicine and Hygiene
 
75
:
9
15
.
Rollin
P. E.
et al
.
1999
.
Ebola (subtype Reston) virus among quarantined nonhuman primates recently imported from the Philippines to the United States
.
Journal of Infectious Diseases
 
179
:
S108
S114
.
Romagnoli
M. L.
Springer
M. S.
.
2000
.
Evolutionary relationships among Old World fruitbats (Megachiroptera: Pteropodidae) based on 12S rRNA, tRNA valine, and 16s rRNA gene sequences
.
Journal of Mammalian Evolution
 
7
:
259
284
.
Rosevear
D. R.
1969
.
The rodents of West Africa
 .
British Muséum of Natural History
,
London, United Kingdom
.
Schou
S.
Hansen
A. K.
.
2000
.
Marburg and Ebola virus infections in laboratory non-human primates: a literature review
.
Comparative Medicine
 
50
:
108
123
.
Soberón
J.
Peterson
A. T.
.
2004
.
Biodiversity informatics: managing and applying primary biodiversity data
.
Philosophical Transactions of the Royal Society of London, B. Biological Sciences
 
359
:
689
698
.
Soberón
J.
Peterson
A. T.
.
2005
.
Interpretation of models of fundamental ecological niches and species' distributional areas
.
Biodiversity Informatics
 
2
:
1
10
.
Stein
B. R.
Wieczorek
J.
.
2004
.
Mammals of the world: MaNIS as an example of data integration in a distributed network environment
.
Biodiversity Informatics
 
1
:
14
22
.
Stockwell
D. R. B.
Noble
I. R.
.
1992
.
Induction of sets of rules from animal distribution data: a robust and informative method of analysis
.
Mathematics and Computers in Simulation
 
33
:
385
390
.
Stockwell
D. R. B.
Peters
D. P.
.
1999
.
The GARP modelling system: problems and solutions to automated spatial prediction
.
International Journal of Geographical Information Science
 
13
:
143
158
.
Stockwell
D. R. B.
Peterson
A. T.
.
2002a
.
Controlling bias in biodiversity data
. Pp.
537
546
in
Predicting species occurrences: issues of scale and accuracy
  (
Scott
J. M.
Heglund
P. J.
Morrison
M. L.
, eds.).
Island Press
,
Washington, D.C.
Stockwell
D. R. B.
Peterson
A. T.
.
2002b
.
Effects of sample size on accuracy of species distribution models
.
Ecological Modelling
 
148
:
1
13
.
Stockwell
D. R. B.
Peterson
A. T.
.
2003
.
Comparison of resolution of methods used in mapping biodiversity patterns from point occurrence data
.
Ecological Indicators
 
3
:
213
221
.
Suzuki
Y.
Gojobori
T.
.
1997
.
The origin and evolution of Ebola and Marburg viruses
.
Molecular Biology and Evolution
 
14
:
800
806
.
Swanepoel
R.
et al
.
1996
.
Experimental inoculation of plants and animals with Ebola virus
.
Emerging Infectious Diseases
 
2
:
321
325
.
Tucker
C. J.
Wilson
J. M.
Mahoney
R.
Anyamba
A.
Linthicum
K.
Myers
M. F.
.
2002
.
Climatic and ecological context of the 1994–1996 Ebola outbreaks
.
Photogrammetric Engineering and Remote Sensing
 
68
:
147
152
.
Turell
M. J.
Bressler
D. S.
Rossi
C. A.
.
1996
.
Lack of virus replication in arthropods after intrathoracic inoculation of Ebola Reston virus
.
American Journal of Tropical Medicine and Hygeine
 
55
:
89
90
.
United States Geological Survey
.
2001
.
HYDRO lk elevation derivative database
 , http://edcdaac.usgs.gov/gtopo30/hydro/.
Accessed 1 January 2004
.
University of California, Santa Barbara
.
2003
.
Alexandria Digital Library Gazetteer
 , http://www.alexandria.ucsb.edu/.
Accessed 15 November 2004
.
Weber
D. J.
Rutala
W. A.
.
2001
.
Risks and prevention of nosocomial transmission of rare zoonotic diseases
.
Clinical Infectious Diseases
 
32
:
446
456
.
Weir
E.
2001
.
Ebola erupts again
.
Canadian Medical Association Journal
 
164
:
685
685
.
Wilson
D. E.
Reeder
D. M.
(eds.).
1993
.
Mammal species of the world: a taxonomic and geographic reference
 .
2
nd ed.
Smithsonian Institution Press
,
Washington, D.C.
Zeller
H.
2000
. [
Lessons from the Marburg virus epidemic in Durba, Democratic Republic of the Congo (1998–2000).
]
Medecine Tropicale
 
60
:
23S
26S
(in French)
.

Author notes

Associate Editor was John A. Yunger.