See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/215475058

Biodiversity loss and emerging infectious disease: An example from the rodent-
borne hemorrhagic fevers

Article  in  Biodiversity · February 2006

DOI: 10.1080/14888386.2006.9712789

CITATIONS READS

93 968

1 author:

James N Mills
Emory University
219 PUBLICATIONS   8,343 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

BIODIS View project

Sin Nombre virus View project

All content following this page was uploaded by James N Mills on 15 May 2014.

The user has requested enhancement of the downloaded file.

 TROPICAL CONSERVANCY

Biodiversity loss and emerging infectious disease:

An example from the rodent-borne hemorrhagic fevers
James N. Mills
Abstract.The relationship between biodiversity loss and emerging infectious diseases is studied in this paper using as James N. Mills
examples the rodent-borne hemorrhagic fevers (those caused by hantaviruses and arenaviruses). Both biodiversity Special Pathogens
loss and infectious disease have taken on importance in the last two decades and both are related to human Branch
disturbance of natural systems. Intuitively, one might expect the most diverse natural habitats to support the greatest Division of Viral and
number of animal diseases that could be transmitted to humans (zoonotic diseases). Yet an examination of the Rickettsial Diseases
circumstances surrounding outbreaks of rodent-borne hemorrhagic fevers suggests that anthropogenically disturbed, National Center for
low-diversity habitats present the greatest risk. Two possible mechanisms are suggested to help explain this increased Infectious Diseases
risk. First, disturbed habitats often favor opportunistic/generalist species which happen to be reservoirs for hemorrhagic Centers for Disease
fever viruses. Second, through a process called the dilution effect, in communities of lowering rodent diversity, more Control and Prevention
and more virus transmission events (aggressive encounters) occur within a single species, usually an opportunistic 1600 Clifton Road
species which is the specific host for a hemorrhagic fever virus. The result is more efficient transmission of the virus (MS G-14)
within the host population, higher prevalence of infection in the host, and greater risk to humans. Results of virus Atlanta, Georgia 30333
monitoring among rodent populations in the southwestern United States are consistent with predictions of the dilution USA
effect hypothesis. Additional monitoring, sampling, and laboratory experimentation are needed to fully understand the [email protected]
relationship between biodiversity and human disease. This monitoring can be accomplished most effectively through
collaboration of scientists from both ecological and public health disciplines. EDITOR’S NOTE:
Dr. Mills is chief of the
INTRODUCTION NATURAL HISTORY OF Medical Ecology Unit
within the Special
The title of this article cites two crises – biodiversity THE RODENT-BORNE HEMORRHAGIC FEVERS Pathogens Branch at the
loss and infectious diseases — that stand out among the The rodent-borne viral hemorrhagic fevers (Mills & U.S. Centers for
Disease Control and
problems facing our world today. Although neither is Childs 2001) consist of those associated with two groups Prevention. His
new, both crises have taken on a particular urgency in of RNA viruses, the arenaviruses (family Arenaviridae) research for the past 19
the last few decades. Both have their roots in human and the hantaviruses (family Bunyaviridae, genus years has involved the
ecology of the reservoir
population growth and human behavior and the Hantavirus). In Asia and Europe, the hantaviruses cause hosts of the viral
interactions between humans and their natural hemorrhagic fever with renal syndrome (HFRS). HFRS hemorrhagic fevers in
environment. As I hope to illustrate, the two crises are is actually a suite of diseases that vary in severity (case Africa, Asia, Europe,
and the Americas.
linked – in ways that seem obvious, and in other ways fatality 1 – 15%) according to the virus that causes the
that seem counterintuitive. The causes and infection. HFRS has been recognized at least since World
consequences of biodiversity loss are widely discussed War I and is a very important disease problem in parts
among the ecological and conservation communities of Asia where there may be as many as 100,000 cases
and the literature specific to the discipline abounds with annually (Peters et al. 1999). In the Americas,
articles
A addressing the problem. The second crisis, the hantaviruses
B cause hantavirus pulmonary syndrome
emergence and increasing incidence of infectious (HPS). HPS was first recognized in 1993 in the
disease, is well recognized among a small group of southwestern United States as an outbreak of severe
epidemiologists, veterinarians, and public health respiratory distress syndrome with a case fatality of
scientists. This problem is discussed at meetings on approximately 50% (Nichol et al. 1993, Childs et al.
tropical medicine, epidemiology, and veterinary public 1994). HPS has since been recognized as a Pan-
health and is a popular subject in the medical literature. American disease with cases occurring throughout most
Nevertheless, only rarely do ecologists and of North, Central, and South America (Schmaljohn &
conservations read the medical literature, and equally Hjelle 1997). The arenaviruses cause the South
rarely do epidemiologists and public health practitioners American hemorrhagic fevers (Argentine, Bolivian, and
read the literature in ecology and conservation. Thus Venezuelan hemorrhagic fevers) and Lassa fever in
the idea that these two problems may be linked is still West Africa. One arenavirus, lymphocytic
poorly recognized by both disciplines. My purpose in choriomeningitis virus, occurs throughout most of the
this article is to illustrate that both of these problems world in association with its cosmopolitan host, the
are of tremendous importance, that an understanding of house mouse (Enria et al. 1999). In general, each
biodiversity loss is critical to an understanding of the hantavirus and each arenavirus is associated with a
emergence of infectious diseases, and that collaboration single species of rodent host (in the family Muridae) in
of ecologists, conservationists, epidemiologists, and which it establishes a chronic, persistent infection
public health scientists can contribute to the goals of involving the shedding of large quantities of infectious
both disciplines. To illustrate my concepts I will use virus into the environment via urine, feces, and saliva
examples from the viral zoonoses, specifically, the for extended periods of time. The viruses are probably
rodent-borne hemorrhagic fever viruses (hantaviruses most frequently transmitted to humans via the inhalation
C D
and arenaviruses). of infectious aerosols of virus shed in urine. Field

B I O D I V E R S I T Y 7 ( 1 ) 2 0 0 6 9
 agents that are being recognized at an increasing rate,
such as AIDS, Nipah virus encephalitis, SARS, and H5N1
avian influenza, and those that represent the re-emergence
of previously described diseases in drug-resistant or more
virulent forms or in expanded ranges, such as tuberculosis,
malaria, the illnesses due to E. coli 0157:H7, monkeypox,
and West Nile virus. The recent awareness of emerging
infectious disease is due to several factors including the
development of antibiotic resistance, improved diagnostics,
physician education, and rapid transportation that quickly
brings victims of remotely acquired diseases into heavily
populated cities. Another important factor is our rapidly
expanding population with the resulting incursion of
humans into remote, natural habitats where previously
unknown diseases have existed for many years in cycles
involving wild-animal hosts.
ZOONOTIC DISEASES
Zoonoses are those diseases that are transmitted between
vertebrate animals and humans. Of the 175 diseases that
were described as “emerging diseases” in 2001, 75%
are zoonotic (Taylor et al. 2001). Zoonotic diseases are
hosted by a wide variety of vertebrates including
Figure 1. research indicates that hantaviruses and at least some amphibians and reptiles (Salmonella infections), birds
New World arenaviruses are transmitted within rodent host (West Nile virus, avian influenza), and mammals.
hantaviruses Discussion of all of these disease organisms would be
recognized as of populations via aggressive encounters and the inflicting
January, 2006. Those of bite wounds (Mills et al. 1992, Mills et al. 1999a). beyond the scope of this work. In order to provide an idea
shown in bold type are of the potential diversity of zoonotic diseases, I will
recognized human EMERGING INFECTIOUS DISEASES provide a few examples associated with the group with
pathogens.
In 1967 the Surgeon General of the United States, which I am most familiar, the rodents.
William H. Stewart, declared that it was time to “close
the book on infectious diseases” and begin to shift our Because of our kinship, mammals are the most likely group
resources toward the study of chronic diseases. In 1998 of animals to share pathogens with humans and rodents
in a speech before the U.S. Congress, another Surgeon are the largest group of mammals, comprising nearly half
General, David Satcher, addressed the continuing threat of all species. This tremendous diversity provides for a
A high diversity of disease agents. Rodents have several
of “emerging infectious diseases” identifying infectious
disease as the number-one killer worldwide. Dr. Satcher additional traits that make them good hosts for zoonotic
went on to say that the death rate due to infectious diseases. Many species are highly opportunistic. Their high
diseases (excluding AIDS) in the United States alone fecundity and rapid reproduction allows them to take
increased by 22% between 1980 and 1992. Within the advantage of temporarily favorable environmental
last two decades, emerging infectious diseases have conditions to reach very high densities that are conducive
been the topic of numerous popular magazine articles, to the transmission of pathogens. Finally, unlike most wild
books, television programs, and films, and the more animals, they are frequently found in our gardens, storage
recent threat of bioterrorism (e.g., the still-unsolved buildings, and even our homes, where we often come into
anthrax mailings) has brought the agents of these contact with them. We have no idea how many potential
diseases increasingly into the public awareness. disease-causing organisms most rodent species are
carrying. No one has looked systematically. However, we
The term “emerging infectious disease” applies to two can make some inferences and extrapolations from what
groups of illnesses: those caused by previously unknown we know in a few cases. I would like to discuss the question
Table 1.
Mouse adenovirus-1 Minute virus of mice
Recognized viruses
associated with Mouse adenovirus-2 Pneumonia virus of mice
Mus musculus
Mouse cytomegalovirus Polyomavirus
*Known or suspected Ectromelia virus Reovirus-3
of causing illness
in humans . Mouse hepatitis virus Mouse rotavirus
Lactic dehydrogenase elevating virus Sendai virus
Murine leukemia viruses (various) Sialodacryoadenitis virus
Lymphocytic choriomeningitis virus* (LCMV) Theiler's murine encephalomyelitis virus
B
Mouse mammary tumor virus* (MMTV) Mouse thymic virus

10 T R O P I C A L C O N S E R V A N C Y
 Table 2.
DISEASE DISTRIBUTION PRIMARY RODENT RESERVOIR A sample of some
I. Viral Diseases significant recog-
nized rodent-borne
Lassa fever Africa M astom ys species diseases (Adapted
from Gratz 1994).
Argentine hemorrhagic fever (HF) Argentina Calom ys m usculinus
Bolivian HF Bolivia Calom ys species
Venezuelan HF Venezuela Zygodontom ys brevicauda
Lymphocytic choriomeningitis virus Worldwide M us m usculus
Hemorrhagic fever with renal syndrome Europe/Asia Muridae species
Hantavirus pulmonary syndrome Americas Sigmodontinae species
Venezuelan equine encephalitis Americas Sigmodontinae species
Tick-borne encephalitis Europe/Asia Muridae species
Monkeypox Africa Unknown terrestrial rodent
II. Rickettsial Diseases
Rocky Mountain spotted fever Americas Many
Boutonneuse fever Europe/Asia/Africa Many
Scrub typhus Asia/India/Australia Rattus species
Murine typhus Global Rattus species
III. Bacterial Diseases
Tick-borne relapsing fever Global Many
Lyme disease Americas/Europe/Asia Perom yscus leucopus and others
Leptospirosis Global Rattus norvegicus and others
Plague Widespread Rattus species/Sciuridae species
Salmonellosis Global Rattus species
IV. Protoz oal Diseases
Toxoplasmosis Global Rattus\M us
Leishmaniasis Asia/Africa/Americas Rattus species and others
Chagas disease Latin America Rattus species and others

of the diversity and importance of rodent-borne zoonoses viral pathogens, the “old world” hantaviruses may cause
from two perspectives. more than 100,000 hospitalized cases of hemorrhagic
What we already know about the diversity fever with renal syndrome each year in Europe and Asia
and importance of pathogens hosted by rodents: (McKee et al. 1991). I will discuss the “new world”
Very few rodent species have been carefully studied to hantaviruses in greater detail below. Among Rickettsial
identify the diseases to which they are susceptible. One diseases, scrub typhus (caused by Orientia tutsugamushi)
species, the common house mouse (Mus musculus) has and murine typhus (caused by Rickettsia typhi), though
been studied because its use in the laboratory has provided vastly under-reported, cause hundreds of thousands of
economic and scientific incentives. Some 40 specific cases annually; up to 50% of some human populations in
agents (viruses, bacteria, rickettsia, and protozoa) have Asia have antibodies to R. typhi (Azad 1990). Lyme
been cited (Committee on Infectious Diseases of Mice disease is a bacterial disease that is found in most of the
and Rats 1991). Only the viruses are listed in Table 1. world. It is the most common vector-borne disease in the
Given that these pathogens are described from animals United States and Europe, causing hundreds to thousands
that may have spent many generations in the relatively of cases per year in the United States (Steere et al. 2004).
sterile environment of the laboratory, it would not be Finally, protozoal Chagas disease affects 16-18 million
unreasonable to expect that the other 2000 or so species people in Latin America (World Health Organization
of rodents in the wild might harbor as many or more 2000).
infectious agents. Not all of these agents, of course, will
What we don’t know about the diversity and
be human pathogens. Of those viruses listed in Table 1,
importance of zoonotic agents hosted by rodents:
to my knowledge, only Lymphocytic choriomeningitis
This is of tremendous importance because the majority
virus is clearly associated with human disease (Jahrling
of rodent-borne pathogens may be still undescribed. I
& Peters 1992) and Mouse mammary tumor virus has a
will use the New World hantaviruses as an example to
hypothesized role in human breast cancer.
illustrate that point. In early 1993, the only recognized
A recent review (Hugh-Jones et al. 1995) described about autochthonous American hantavirus was Prospect Hill
60 diseases affecting humans for which rodents serve as virus. Then an outbreak of HPS in the southwestern
reservoirs. As an illustration that they have a major impact United States in the summer of 1993 led to the
on human health, a few of these diseases are listed (see development of reagents to detect these infections and
Table 2). I will mention only a few examples: Among the the training of physicians to recognize symptoms.

B I O D I V E R S I T Y 7 ( 1 ) 2 0 0 6 11
 Table 3. Disease Virus Location H o st Associated disturbance Reference
Rodent-borne
hemorrhagic fevers Central Calom ys
Argentine HF Junín virus mechanized agriculture (Mills et al. 1992)
that are associated Argentina m usculinus*
with anthropo- subsistence agriculture,
Bolivian HF Machupo virus Beni, Bolivia Calom ys sp.* (Mackenzie et al. 964)
genically disturbed ranching, deforestation
habitats. Venezuela Zygodontom ys agriculture, ranching,
Guanarito virus SW Venezuela (Tesh et al. 1993)
*Opportunistic/ HF brevicauda* deforestation
generalist species villages with open structures
Lassa fever Lassa virus West Africa M astom ys spp.* (Demby et al. 2001)
and abundant food

Korean HF Hantaan virus Korea A podem us agriculture (Lee 1982)

agrarius*
Rattus urban conditions associated
mild HFRS Seoul virus Korea (Lee et al. 1980)
norvegicus* with rats
ranching, construction of
Sin Nombre Perom yscus
HPS SW USA highly permeable housing in (Zeitz et al. 1995)
virus m aniculatus* rural areas
Oligoryzom ys deforestation, rural construc- (Murua & Padula
HPS Andes virus Patagonia
longicaudatus* tion,monocultures of bamboo 2004)
Laguna Negra Paraguay,
HPS Calom ys laucha* deforestation, agriculture (Yahnke et al. 2001)
virus Bolivia
Laguna Negra Santa Cruz, C a l o m ys deforestation, sugar cane
HPS (Carroll et al. 2005)
virus Bolivia callosus* monoculture

Arraraquara Distrito Federal cutting and burning of native Unpublished

HPS and Goias, Bolom ys
virus lasiurus* cerrado, planting of forage observations
Brazil
deforestation, mechanized
Mato Grosso, Unpublished
HPS Unidentified Unidentified agriculture, sugar cane
Brazil observations
monocultures

HPS Choclo virus Panama Oligoryzom ys deforestation, agriculture (Ruedas et al. 2004)
fulvescens*

Twelve years later at least 33 hantavirus genotypes are THE RELATIONSHIP BETWEEN BIODIVERSITY
recognized in association with about as many different AND ZOONOTIC DISEASE: THE CASE OF
rodent reservoir species (Figure 1) and new viruses are THE RODENT-BORNE HEMORRHAGIC FEVERS
added to this list every year. All of the hantaviruses that
The concept of a relationship between biodiversity loss
are known to cause HPS (bold type in Figure 1) are
and disease is one that is frequently cited in the popular
hosted by rodents of the subfamily Sigmodontinae.
press. It is much less frequently cited in the scientific
One theory suggests that we are not even close to literature (Grifo & Rosenthal 1997), and many scientific
recognizing the full extent of hantavirus infection in murid treatments have been largely theoretical (van Buskirk &
rodents. Because of the highly specific association between Ostfeld 1995, van Buskirk & Ostfeld 1998, Dobson 2004)
hantaviruses and their hosts and their association with just because few data are available for testing hypotheses.
three closely related subfamilies of murid rodents, it has
It has been also been stated that high biodiversity is
been hypothesized that an ancestral hantavirus was
causally related to greater levels of infectious disease
associated with an ancestral murid rodent before the
(Sattenspiel 2000) and hypothesized that biodiversity
subfamilial lineages diverged approximately thirty million
loss might decrease the emergence of zoonotic diseases
years ago, and that the numerous existing hantaviruses
(Wolfe et al. 2005). Although it is intuitive to imagine
have been co-speciating and co-evolving with their hosts
that higher overall biodiversity leads to greater diversity
since that time. This theory, which has been highly
of pathogens and thus greater incidence of disease in
corroborated by both molecular and biogeographical
humans, an examination of the circumstances
evidence (Schmaljohn & Hjelle 1997), has some important
surrounding recent outbreaks of rodent-borne
implications. Firstly, hantaviruses are not new; they have
hemorrhagic fever suggests otherwise. In Table 3, I
probably been around for about 30 million years. Secondly,
have listed the etiologic agents, rodent hosts, and general
there may be many more hantavirus-host associations
circumstances surrounding outbreaks of hantavirus and
awaiting discovery – the maximum potential would be the
arenavirus disease in many parts of the world. Two
529 species of Murinae, 423 species of Sigmodontinae,
items stand out. First, all outbreaks occurred in
and 143 species of Arvicolinae that are recognized in the
anthropogenically highly disturbed habitats that were
world (Wilson & Reeder 1993). Whether we approach the
demonstrated to be of low biodiversity (Figures 1-2).
question of the diversity of rodent-borne zoonotic diseases
Second, in each case, the rodent host was a generalist/
by extrapolating from what we know, or by speculating
opportunistic species (see discussion below)
about what we don’t know, these diseases represent a
significant human health problem that is likely to become Thus at least for the case of the rodent-borne
even more important. hemorrhagic fevers, the occurrence of human disease

12 T R O P I C A L C O N S E R V A N C Y
 A B

C D

G H

Figure 1 (A-H). Sites of important outbreaks of viral hemorrhagic fever: A-B, The pampa of southern Santa Fe Province, disease endemic area of Argentine Hemorrhagic fever. Native grassland is largely
destroyed and is now dedicated to extensive mechanized agriculture; C, Southern Portuguesa State, Venezuela, endemic area for Venezuelan hemorrhagic fever. The llanos of Venezuela was a
grassland similar to the Argentine pampa. Native vegetation has now been replaced by a variety of subsistence and mechanized agriculture and ranching practices; D, Department of the Beni, Bolivia,
endemic area for Bolivian hemorrhagic fever. The Beni consists of seasonally inundated grasslands and gallery forests. Humans use the limited high ground for villages and agriculture and share the
high ground with rodents during the wet season; E-F, Hantavirus pulmonary syndrome (HPS) endemic area in Santa Fe Province, Bolivia. Native tropical forest is being cut and burned and converted
to monocultures of sugar cane which support only a few opportunistic rodent species; G-H, Azuero Peninsula of Panama, endemic area for HPS. The native forest has been replaced by a mosaic
of pastures, agricultural fields, and human dwellings.

B I O D I V E R S I T Y 7 ( 1 ) 2 0 0 6 13
seems to be associated not with conditions of high relation to deforestation, agriculture, and ranching
biodiversity, but with situations where natural (Dahlberg 1992, Peters 1997, Lacher et al. 1999, Ruedas
biodiversity has been dramatically diminished. What et al. 2004, Carroll et al. 2005) and time and again, these
might be the explanation or the mechanisms behind this opportunistic species have been associated with viruses
counterintuitive situation? causing human disease (see Table 3). Thus simplification
of ecosystems and decreasing biodiversity may contribute
Life history strategies: to increases in risk of some zoonotic diseases by favoring
Ecologists often assign mammalian life history patterns
the opportunistic hosts of rodent-borne zoonotic pathogens.
to two categories. “Specialist” species are those that
A second mechanism, which has been proposed to operate
are highly adapted to a narrowly defined habitat and
in the case of vector-borne diseases, may also operate for
require one or a few specific food resources. These
directly transmitted zoonotic agents such as the
species also may have a longer lifespan, slower
hantaviruses and arenaviruses.
development to sexual maturity, and lower fecundity.
Specialist species are generally the first to become The dilution-effect hypothesis:
locally extinct when natural habitats are disturbed. Ostfeld and others (Ostfeld & Keesing 2000a, Ostfeld &
“Generalist” or “opportunistic” species are those that Keesing 2000b, Schmidt & Ostfeld 2001) have proposed
have a high adaptability to a wide range of habitats and a general theoretical model for vector-borne diseases
can subsist on a variety of food sources. These species describing a mechanism by which increasing diversity
are often shorter lived, highly fecund, and have rapid of potential vertebrate host species would result in a lower
development. Generalist species may often survive prevalence of infection of a pathogen in the vector and
habitat disturbance and may even benefit if the lower risk of infection to humans. The model requires
disturbance involves the introduction of some food that several assumptions be met, including a generalist
resource. Rodents that have peridomestic affinities vector species, oral acquisition of the pathogen by the
(enter homes or are otherwise associated with humans) vector, and differences in reservoir competency among
are most often generalist/opportunistic species. potential vertebrate hosts. In the specific case of Lyme
disease in the eastern United States, the white-footed
Most murid rodent species are probably habitat and dietary mouse (Peromyscus leucopus), a numerically dominant
specialists, occupying a relatively narrowly defined niche member of the vertebrate community, is the primary and
in any given ecosystem. However, many of the murid most competent reservoir host; the generalist tick, Ixodes
rodents that are recognized hosts of hemorrhagic fever scapularis, is the vector. Several other species of small
viruses are opportunistic species that seem to thrive in mammals (and perhaps birds and reptiles) serve as
disturbed environments in close association with humans. reservoirs of lower competency. Ostfeld and Keesing
Several important hemorrhagic fever virus host species (2000a) argued that, as the relative proportion of
are associated with agriculture (either subsistence, or large- individuals of less competent reservoir species increases,
scale mechanized), land conversion to pasture, or the proportion of tick meals that are taken from these
peridomestic environments (sheds, barns, homes; Table less competent hosts increases. This results in a lower
3). There are two possible explanations for this observation. prevalence of infection in the tick population, and a
Firstly, it may be that there are actually just as many decreased risk of exposure to Lyme disease in humans.
hemorrhagic fever viruses associated with specialist The authors went on to demonstrate a negative correlation
species as with opportunistic species, but we are more between species richness of small mammals and the
likely to encounter those viruses associated with incidence of Lyme disease in humans among 10 regions
opportunistic species because they are the species we are in the Eastern United States. Although the example used
most likely to encounter and are therefore most likely to by Ostfeld and Keesing (2000a) was a bacterial disease,
transmit pathogens to humans. A second possible the theory is equally applicable to vector-borne viral
explanation is that there is something inherent about diseases. A great deal of testing is still required to assess
opportunistic species that makes them more likely to evolve the accuracy and generality of the dilution effect
and maintain associations with hemorrhagic fever viruses hypothesis for vector-borne diseases throughout the world.
(e.g., their greater motility, or their high fecundity and
tendencies to achieve high population densities that are Through a different mechanism, the dilution effect
conducive to more efficient transmission of disease agents). hypothesis might be adapted to some directly transmitted
zoonotic diseases (those that are transmitted by direct
Regardless of the explanation, however, anthropogenic or indirect contact with an infected host, but do not
disturbance to natural ecosystems frequently results in require the participation of a vector in the disease cycle).
dramatic changes to the environment and consequent These diseases include the rodent-borne hemorrhagic
changes in the composition of rodent assemblages. Many fevers caused by the hantaviruses and arenaviruses.
specialist species become locally extinct, and the Hantaviruses, and at least some arenaviruses, are
assemblage becomes species depauperate, often restricted transmitted within host populations by aggressive
to a few opportunistic species whose population densities encounters between individual hosts. As described
may increase dramatically under release from competitive above, host relationships of hantaviruses and
pressures. This pattern has been observed many times in arenaviruses are highly specific and other rodent

14 T R O P I C A L C O N S E R V A N C Y
species are thought to be dead-end hosts (Peters et al. detract from the total encounters with host species and
1999, Enria et al. 1999). Thus it can be expected that decrease the rate of transmission among hosts. Thus
these viruses might be transmitted most efficiently in the prevalence of infection within the host population
host communities of low diversity. In the most extreme would be lower, and the risk of transmission to humans
case, the host “community” would consist of a single lower, in rodent populations of high species diversity.
species, that species being the specific reservoir for
the virus. Every aggressive encounter (potential An alternative mechanism that would create an inverse
transmission event) that an infected host experiences relationship between host diversity and zoonotic
would be with other hosts and the virus would be disease risk to humans involves the combination of
transmitted and further amplified as the newly infected two ecological principles, resource competition and
host encounters other individuals. As additional (non- mass action. Theoretically, higher densities of non-
host) species are added to the community, aggressive host species should result in lower densities of hosts
encounters are more frequently with non-hosts and result through resource competition – reproduction and
in no further amplification of the virus. Assuming a survival of hosts would be limited by the availability
limited amount of energy dedicated to aggressive of resources and the carrying capacity of a given
encounters, each encounter with a non-host would environment for one species would be reduced when
Figure 2 (A-F).
Sites of important
outbreaks of viral
hemorrhagic fever:
A-B, Azuero Peninsula of
Panama, endemic area for
HPS. The native forest
has been replaced by a
mosaic of pastures,
agricultural fields, and
human dwellings;
C-D, Distrito Federal
and Goias State, Brazil,
HPS endemic area.
Native cerrado has been
cut and burned and
A B planted in exotic grasses
(primarily Brachiaria
decumbens) that provide
ideal habitat for a few
opportunistic rodents;
E-F, Mato Grosso State,
Brazil, HPS endemic
area. This transitional
area between cerrado and
tropical forest has been
almost completely
cleared in recent years.
Although farmers are
legally obliged to leave
10% of their land
undisturbed, tiny islands
of native vegetation are
dwarfed by huge
plantations of soybeans,
C sunflower, and sugar
D cane that stretch from
horizon to horizon.

E F

B I O D I V E R S I T Y 7 ( 1 ) 2 0 0 6 15
 the increased density of populations of some
opportunistic species which are hosts for
hemorrhagic fever viruses. Introduced food
supplies associated with agriculture, the planting
of exotic range grasses, or the expansion of
opportunistic plant species (such as bamboos in
Patagonia (Jaksic & Lima 2003); see Table 3)
may contribute to the reproductive success of
opportunistic host species, which increase to
unusually high densities in the absence of
specialist competitors. These high-density
populations are conducive to increased frequency
of intraspecific interactions among host
individuals and higher rates of virus transmission,
leading to a higher prevalence of infection with
the pathogen. In this low-diversity situation,
interactions with non-host species that might
decrease the number of intraspecific encounters
(the dilution effect) are rare. These conditions
may contribute to disease outbreaks in human
Figure 9. resources must be shared with other species. populations as rodents emigrate from high density areas
Linear regression of According to the mass action principle (Dobson & into nearby peridomestic areas, or humans contact
hantavirus antibody infected rodents by entering the agricultural or otherwise
prevalence in the Hudson 1995), the transmission of a pathogen in a
dominant host species population is proportional to the population density. disturbed areas where the host populations are
on cumulative small Thus the lower density host population in a community concentrated.
mammal diversity
(Simpson’s index) at containing many non-host species should support a From the preceding discussion, we are led to two
10 trapping sites in lower prevalence of infection with the pathogen and seemingly paradoxical conclusions. Firstly, as
the southwestern represent a lower risk of infection to humans.
United States, 1994 – exemplified by an examination of the hantaviruses, the
2001. Reproduced The inverse relationship between host community diversity of rodent-borne viral zoonotic diseases and
from Mills 2005 with
permission of the diversity and prevalence of infection with a directly their hosts, is tremendous; secondly, it may be the
publisher. transmitted zoonotic pathogen was tested using data from destruction of this very diversity that leads to increased
longitudinal studies of hantavirus reservoir populations risk of zoonotic disease in human populations.
in the southwestern United States. These studies have In this article, I have discussed the rodent-borne
been conducted since 1994 on 21 trapping arrays at 10 hemorrhagic fevers as an example. Despite their
sites in Arizona, Colorado, and New Mexico (Mills et al. tremendous potential diversity, the hemorrhagic fever
1999b). When we examined the relationship between the viruses are likely only a small proportion of the total (and
prevalence of infection with a hantavirus and the diversity yet unrecognized) zoonotic disease burden carried by
of small mammal communities among the 10 trapping rodents. It is not unreasonable to expect that conditions
sites, we found a highly statistically significant and that lead to increases in risk of human infection with
negative linear relationship (see Figure 9, (Mills 2005)). hantaviruses and arenaviruses might also lead to
As suggested by the previous discussion, the conditions in increased risk of many other rodent-borne diseases.
disturbed ecosystems (abundant food supplies associated
As a precautionary note, I do not intend to imply that
with agriculture, absence of competition, presence of
maintaining high levels of biodiversity will decrease the
artificial shelter) may lead to very different characteristics
incidence of all zoonotic diseases. The majority of rodent
of populations of opportunistic species in disturbed
species (indeed most animal species) are relatively
conditions. Investigators compared characteristics of sylvan
specialist species that occur in the least disturbed
deer mice (those living in relatively undisturbed habitats)
environments. The mechanisms discussed above will apply
to those of peridomestic populations (those living in and
only to those pathogens whose hosts are opportunistic
around human structures). They found that the breeding
species. Although I have suggested that it may be the case,
season of deer mice in peridomestic environments was
it has not been established that opportunistic species are
two months longer than in natural habitats, and the
more likely to harbor zoonotic pathogens than are more
prevalence of infection with Sin Nombre hantavirus was
specialist species. A clear and complete picture of the
significantly higher in peridomestic than in sylvan
overall risk of zoonotic disease can only be determined
populations (Kuenzi et al. 2001).
through achieving a greater understanding of the distribution
CONCLUSIONS of pathogens in host populations worldwide, and especially
Anthropogenic disturbance causing simplification of through increased sampling of populations of specialist
ecosystems results in diminished biodiversity, but favors species. This can best be achieved through multidisciplinary

16 T R O P I C A L C O N S E R V A N C Y
 collaboration among ecologists, conservations, and public 615-632 in R. B. Belshe (Ed.). Textbook of human virology. Mosby
health scientists. Year Book, St. Louis, MO.
Mills, J. N. 2005. Regulation of rodent-borne viruses in the natural
ACKNOWLEDGMENTS host: implications for human disease. Archives of Virology 19: 45-
57.
I am indebted to the many collaborators who made Mills, J. N. & J. E. Childs. 2001. Rodent-Borne Hemorrhagic Fever
these studies possible in Argentina, Bolivia, Brazil, Viruses. Pages 254-270 in E. S. Williams, & I. K. Barker (Eds.).
Chile, Panama, Paraguay, Venezuela, and the United Infectious Diseases of Wild Mammals. Iowa State University, Ames,
IA.
States. My work at the Centers for Disease Control and Mills, J. N., B. A. Ellis, K. T. McKee, G. E. Calderón et al. 1992. A
Prevention has been supported enthusiastically by longitudinal study of Junín virus activity in the rodent reservoir of
James Childs, C.J. Peters, Thomas Ksiazek, and Pierre Argentine hemorrhagic fever. American Journal of Tropical Medicine
and Hygiene 47: 749-763.
Rollin. Thanks to Barbara Ellis, Darin Carroll, Brian Mills, J. N., T. G. Ksiazek, C. J. Peters, & J. E. Childs. 1999a. Long-term
Amman, Joel Montgomery, Pierre Rollin, and Thomas studies of hantavirus reservoir populations in the southwestern United
Ksiazek for helpful comments on the manuscript. States: a synthesis. Emerging Infectious Diseases 5: 135-142.
Mills, J. N., T. L. Yates, T. G. Ksiazek, C. J. Peters, & J. E. Childs.
1999b. Long-term studies of hantavirus reservoir populations in the
REFERENCES southwestern United States: rationale, potential, and methods.
Azad, A. F. 1990. Epidemiology of murine typhus. Annu.Rev.Entomol Emerging Infectious Diseases 5: 95-101.
35: 553-569. Murua, R. & P. Padula. 2004. [Ecology and evolution of hantavirus in
Carroll, D. S., J. N. Mills, J. M. Montgomery et al. 2005. Hantavirus the Southern Cone of America]. Arch.Med.Vet. 36: 1-20.
pulmonary syndrome in central Bolivia: relationships between Nichol, S. T., C. F. Spiropoulou, S. Morzunov et al. 1993. Genetic
reservoir hosts, habitats, and viral genotypes. American Journal of identification of a hantavirus associated with an outbreak of acute
Tropical Medicine and Hygiene 72: 54-58. respiratory illness. Science 262: 914-917.
Childs, J. E., T. G. Ksiazek, C. F. Spiropoulou et al. 1994. Serologic Ostfeld, R. S. & F. Keesing. 2000a. Biodiversity and disease risk: the
and genetic identification of Peromyscus maniculatus as the primary case of Lyme disease. Conservation Biology 14: 722-7287.
rodent reservoir for a new hantavirus in the southwestern United Ostfeld, R. S. & F. Keesing. 2000b. The function of biodiversity in the
States. Journal of Infectious Diseases 169: 1271-1280. ecology of vector-borne zoonotic diseases. Canadian Journal of
Committee on Infectious Diseases of Mice and Rats. 1991. Infectious Zoology 78: 2078.
Diseases of Mice and Rats. National Academy Press, Washington, D.C. Peters, C. M. 1997. Sustainable use of biodiversity: myths, realities,
Dahlberg, K. A. 1992. The conservation of biological diversity and and potential. Pages 312-333 in F. Grifo, & J. Rosenthal (Eds.).
U.S. agriculture: goals, institutions, and policies. Agriculture, Biodiversity and human health. Island Press, Washington, D.C.
Ecosystems and Environment 42: 177-193. Peters, C. J., J. N. Mills, C. F. Spiropoulou, S. R. Zaki, & P. E. Rollin.
Demby, A. H., A. Inapogui, K. Kargbo et al. 2001. Lassa fever in 1999. Hantavirus Infections. in R. L. Guerrant, D. H. Walker, & P. F.
Guinea: II. Distribution and prevalence of lassa virus infection in Weller (Eds.). Tropical Infectious Diseases: Principles, Pathogens
small mammals. Vector Borne and Zoonotic Diseases 1: 283-297. and Practice. W.B. Saunders, New York.
Dobson, A. 2004. Population dynamics of pathogens with multiple Ruedas, L. A., J. Salazar-Bravo, D. S. Tinnin et al. 2004. Community
host species. American Naturalist 164: S64-S78. ecology of small mammal populations in Panama following an
Dobson, A. P. & P. J. Hudson. 1995. Microparasites: observed patterns outbreak of Hantavirus pulmonary syndrome. Journal of Vector
in wild animal populations. Pages 52-89 in B. T. Grenfell, and A. P. Ecology 29: 177-191.
Dobson (Eds). Ecology of Infectious Diseases in Natural Sattenspiel, L. 2000. Tropical environments, human activities, and the
Populations. Cambridge University Press, Cambridge. transmission of infectious diseases. Yearbook of Physical Anthropology
Enria, D. A., M. D. Bowen, J. N. Mills, G. J. Shieh, D. Bausch, & C. J. 43: 3-31.
Peters. 1999. Arenaviruses. Pages 1191-1212 in R. L. Guerrant, D. Schmaljohn, C. S. & B. Hjelle. 1997. Hantaviruses: A global disease
H. Walker, & P. F. Weller (Eds). Tropical infectious diseases, problem. Emerging Infectious Diseases 3: 95-104.
principles, pathogens, and practice. W.B. Saunders, New York. Schmidt, K. A. & R. S. Ostfeld. 2001. Biodiversity and the dilution
Gratz, N. G. 1994. Rodents as carriers of disease. Pages 85-108 in A. effect in disease ecology. Ecology 82: 609-619.
P. Buckle, & R. H. Smith (Eds.). Rodent pests and their control. Steere, A. C., J. Coburn, & L. Glickstein. 2004. The emergence of Lyme
Cab International, Wallingford, UK. disease. Journal of Clinical Investigation 113: 1093-1101.
Grifo, F. D. & J. Rosenthal. 1997. Biodiversity and Human Health. Taylor, L. H., S. M. Latham, & M. E. J. Woolhouse. 2001. Risk factors for
Island Press, Washington, D.C. human disease emergence. Philosophical Transactions of the Royal
Hugh-Jones, M. E., W. T. Hubbert, & H. V. Hagstad. 1995. Zoonoses: Society of London-Series B: Biological Sciences 356: 983-989.
Recognition, Control, and Prevention. Iowa State University Press, Ames. Tesh, R. B., M. L. Wilson, R. Salas, N. M. De Manzione, D. Tovar, T. G.
Jahrling, P. B. & C. J. Peters. 1992. Lymphocytic choriomeningitis virus: Ksiazek, & C. J. Peters. 1993. Field studies on the epidemiology of
A neglected pathogen of man. Arch.Pathol.Lab.Med. 116: 486-488. Venezuelan hemorrhagic fever: implication of the cotton rat
Jaksic, F. M. & M. Lima. 2003. Myths and facts on ratadas: Bamboo Sigmodon alstoni as the probable rodent reservoir. American Journal
blooms, rainfall peaks and rodent outbreaks in South America. of Tropical Medicine and Hygiene 49: 227-235.
Austral Ecology 28: 237-251. van Buskirk, J. & R. S. Ostfeld. 1995. Controlling lyme disease by
Kuenzi, A. J., R. J. Douglass, D. White, C. W. Bond, & J. N. Mills. modifying the density and species composition of tick hosts.
2001. Antibody to Sin Nombre virus in rodents associated with Ecological Applications 5: 1133-1140.
peridomestic habitats in west central Montana. American Journal van Buskirk, J. & R. S. Ostfeld. 1998. Habitat heterogeneity, dispersal,
of Tropical Medicine and Hygiene 64: 137-146. and local risk of exposure to lyme disease. Ecological Applications
Lacher, T. E., R. D. Slack, L. M. Coburn, & M. I. Goldstein. 1999. The 8: 365-378.
role of agroecosystems in wildlife biodiversity. Pages 147-165 in Wilson D. E. & D. M. Reeder. 1993. Mammal species of the world., 2
W. W. Collins, & C. O. Qualset (Eds). Biodiversity in edition. Smithsonian Institution Press, Washington, DC.
agroecosystems. CRC Press, Boca Raton. Wolfe, N. D., P. Daszak, A. M. Kilpatrick, & D. S. Burke. 2005.
Lee, H. W., D. H. Park, L. J. Baek, K. S. Choi, Y. N. Whang, & M. S. Bushmeat hunting, deforestation, and prediction of zoonoses
Woo. 1980. Korean hemorrhagic fever patients in urban areas of emergence. Emerging Infectious Diseases 11: 1822-1827.
Seoul. Korean Journal of Virology 10: 1-6. World Health Organization. 2000. Chagas disease, Chile. Weekly
Lee, H. W. 1982. Korean hemorrhagic fever. Progress in Medical Epidemiologic Record 75: 10-12.
Virology 28: 96-113. Yahnke, C. J., P. L. Meserve, T. G. Ksiasek, & J. N. Mills. 2001. Patterns
Mackenzie, R. B., H.K.Beye, L. Valverde, & H. Garrón. 1964. of Infection with Laguna Negra virus in Wild Populations of Calomys
Epidemic hemorrhagic fever in Bolivia. 1. A preliminary report of laucha in the Central Paraguayan Chaco. American Journal of Tropical
the epidemiologic and clinical findings in a new epidemic area in Medicine and Hygiene 65: 768-776.
South America. American Journal of Tropical Medicine and Hygiene Zeitz, P. S., J. C. Butler, J. E. Cheek, M. C. Samuel et al. 1995. A case-control
13: 620-625. study of hantavirus pulmonary syndrome during an outbreak in the
McKee, K. T., J. W. LeDuc, & C. J. Peters. 1991. Hantaviruses. Pages southwestern United States. Journal of Infectious Diseases 171: 864-870.

B I O D I V E R S I T Y 7 ( 1 ) 2 0 0 6 17

View publication stats