Interdisciplinary research in the ecology of vector-borne diseases: opportunities and needs.
ABSTRACT In addition to their importance to human and animal health, vector-borne diseases are fascinating systems to study. The involvement of multiple species whose biologies and life cycles cover differing space and time scales makes it extremely difficult to predict epidemics. A single environmental factor may have opposite impacts on the system at different points in time. Patchiness at different geographical scales may have very different causes, so it is important to identify the proper scale for a particular study. New developments in remote sensing, GIS, and spatial analysis make it easier to tease out causes of observed patchiness. Interdisciplinary collaboration is essential for many of the projects we carry out, but this requires awareness of the differences between disciplines and the ability to effectively communicate with each other. It is only by forming multi-disciplinary groups to focus on specific vector-host-pathogen systems that we will be able to answer the most interesting (and pressing) problems in our field.
- SourceAvailable from: ca.gov[show abstract] [hide abstract]
ABSTRACT: Present strategies for surveillance, prevention, and control of arbovirus diseases in western North America have been developed from more than 4 decades of epidemiological research and development of mosquito control technology. Methods of prediction of outbreaks remain imprecise, although our understanding of sources of variation associated with indicators used for prediction is improving. Well organized and funded systematic mosquito abatement remains the most effective method of prevention of human cases of mosquito-borne virus disease, although emergency methods must be employed when outbreaks are imminent. The development of information management systems technology, use of recent developments of sampling theory, and research on vector competency and related areas should permit much better precision in estimates of impending outbreaks.The American journal of tropical medicine and hygiene 12/1987; 37(3 Suppl):77S-86S. · 2.53 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: In order to understand adquately the dynamics of vector-borne disease, one must understand how and why vector populations change over time. We describe a long-term, cooperative study of seasonal fluctuation in populations of the Aedes aegypti mosquito in Puerto Rico. During each month of the first 3 years of the project, A. aegypti was found breeding in all five communities studied. Mosquito density was positively correlated with rainfall, the relationship being more marked in the dry, south-coastal part of the island. Discarded tires and animal watering pans were the two most common larval breeding sites. In general, houses in Puerto Rico harbor more potential A. aegypti breeding sites than those in other tropical locations, probably because Puerto Rico is relatively more affluent.The American journal of tropical medicine and hygiene 12/1978; 27(6):1225-31. · 2.53 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Aedes triseriatus (Say) population density patterns and La Crosse encephalitis virus infection rates were evaluated in relation to a variety of habitat parameters over a 14-wk period. Ovitraps and landing collections were used in a La Crosse virus-enzootic area in Nicholas County, WV. Study sites were divided into categories by habitat type and by proximity to the residences of known La Crosse encephalitis cases. Results demonstrated that Ae. triseriatus population densities were higher in sugar maple/red maple habitats than in hemlock/mixed hardwood habitats or in a site characterized by a large number of small red maple trees. Sites containing artificial containers had higher population densities than those without. La Crosse virus minimum infection rates in mosquitoes collected as eggs ranged from 0.4/1,000 to 7.5/1,000 in the 12 study sites, but did not differ significantly among sites regardless of habitat type or proximity to human case residences. La Crosse virus infection rates in landing Ae. triseriatus mosquitoes ranged from 0.0/1,000 to 27.0/1,000. La Crosse virus was also isolated from host-seeking Ae. canadensis (Theobald) in two study sites, at rates similar to those found in the Ae. triseriatus populations. The Ae. triseriatus oviposition patterns and La Crosse virus infection rates suggest that this mosquito species disperses readily in the large woodlands of central West Virginia. The La Crosse enzootic habitats in Nicholas County, WV, are contrasted with those studied in other geographic regions where La Crosse virus is found.Journal of Medical Entomology 08/2000; 37(4):559-70. · 1.86 Impact Factor
218 Journal of Vector Ecology December 2008
Adapted from the Society for Vector Ecology 2008 Distinguished Achievement Award Presentation
at the 40th Annual Meeting
Interdisciplinary research in the ecology of vector-borne diseases: Opportunities
Chester G. Moore
Department of Microbiology, Immunology and Pathology, Colorado State University,
Fort Collins, CO 80523-1692, U.S.A.
ABSTRACT: In addition to their importance to human and animal health, vector-borne diseases are fascinating systems
to study. The involvement of multiple species whose biologies and life cycles cover differing space and time scales makes it
extremely difficult to predict epidemics. A single environmental factor may have opposite impacts on the system at different
points in time. Patchiness at different geographical scales may have very different causes, so it is important to identify the
proper scale for a particular study. New developments in remote sensing, GIS, and spatial analysis make it easier to tease
out causes of observed patchiness. Interdisciplinary collaboration is essential for many of the projects we carry out, but this
requires awareness of the differences between disciplines and the ability to effectively communicate with each other. It is only
by forming multi-disciplinary groups to focus on specific vector-host-pathogen systems that we will be able to answer the
most interesting (and pressing) problems in our field. Journal of Vector Ecology 33 (2): 218-224. 2008.
Keyword Index: Vector ecology, temporal pattern, spatial pattern, scale, communication, multi-disciplinary collaboration.
In this presentation, I want to briefly review several
issues in vector ecology that are of concern to me. To some
extent this will be preaching to the choir, but I think it is
important to remember the basics as we become ever more
deeply focused on the minutiae of our particular area
of study (more about that shortly). The issues I wish to
discuss are: 1) Why is it so difficult to predict, prevent and
control vector-borne diseases? 2) What controls the release
of vector-borne pathogens in time and space, resulting in
epizootics or epidemics? 3) How can we better draw upon
the resources and insights of the various disciplines of
science to improve prediction, prevention, and control of
Prediction, prevention, and control—working with
One of the major reasons vector-borne diseases are so
difficult to predict is the complex interaction of multiple
organisms—vector, vertebrate host, pathogen—in space
and time (Figure 1). Often there are multiple vector and
vertebrate host species involved in the system, each with
slightly different ecologies.
Added to the list of potential hosts and vectors, we must
account for all of the other system components, including
plant assemblages, predators and parasites, topography
and geology, and weather and climate. Short-term weather
patterns as well as longer term climate change can impact
the biological components in different ways, producing
complex transmission patterns. The complex interaction
of all of these components is what makes our field so
complicated, but it also makes it exciting and challenging
Activity of vector-borne pathogens in time and space
Vector-borne pathogen activity is not distributed
randomly over the landscape, either in space or in time. The
concept of spatial focality has been known and described
for many years (e.g., Pavlovsky 1966). Similarly, particularly
in temperate regions of the world, activity is not constant
from year to year. Rather, epidemic or epizootic years may
be separated by extended periods of little or no activity
(Monath 1980, Tsai et al. 1988).
It is clear that environmental factors such as rainfall,
temperature, and humidity, operating at lags of days to weeks
or even months, can impact current transmission patterns
(Moore et al. 1978, Monath 1980). It is easy to see the
relation between recent rainfall and the abundance of larval
habitats (Figure 2). Every mosquito control district biologist
knows the speed of larval development is related to current
temperature and factors that knowledge into the inspection
and treatment schedule. It is less obvious that there may
be important connections between the current situation
and conditions farther back in time. For example, winter
temperature impacts the number and timing of emerging
females in the spring or eggs surviving to hatch with first
flooding (Bennington et al. 1958, Mail and McHugh 1961,
Shemanchuk 1965). Winter snowpack and spring runoff are
known to impact mosquito-transmitted arbovirus cycles,
and are used as predictors in California (Eldridge 1987).
Even farther back in time, conditions during the preceding
year may impact the food sources of vertebrate hosts, leading
to increases or decreases in young for the year (Ostfeld et
al. 2006). Host plant strategies such as masting (Ostfeld et
al. 1996) introduce another complication into our attempts
at modeling and predicting these systems. Thus, we cannot
Vol. 33, no. 2 Journal of Vector Ecology 219
Figure 1. Typical transmission cycle of a vector-borne agent, in this case a mosquito-transmitted arbovirus such as West Nile
virus (Adapted from Moore et al. 1993).
Figure 2. The impact of environmental factors such as temperature, humidity, and rainfall, can significantly impact the
disease transmission cycle at a number of points in time.
220 Journal of Vector Ecology December 2008
simply rely on the current week’s temperature and rainfall to
predict the dynamics of vector-borne diseases. Each weather
variable may have different impacts at different points in
time relative to current events. The current situation might
best be thought of as the integral of the action of daily
weather over the past one to two years.
As noted above, vector-borne disease systems tend to be
focal in nature, at least until conditions permit widespread
activity. With the exception of pathogens that have multiple
transmission pathways (e.g., vertical transmission), all
components of the system, pathogen, vector, and vertebrate
host, must occur together in time and space for epizootics
or epidemics to occur. Variations in landscape structure
create a patchwork of suitable and unsuitable habitats,
leading to focal disease activity (Figure 3). Human actions
tend to expand the amount of suitable habitat within the
larger matrix, creating new and ever expanding areas of
Barriers, such as water bodies, deserts, or mountain
ranges, may prevent the occurrence of a pathogen in an
otherwise suitable location. These barriers can be broken
down by natural or by human activity. In particular, human
travel and commerce have created a massive interchange
of organisms across barriers so that we are now faced with
invasive/exotic vectors (Aedes aegypti, Aedes albopictus,
Aedes japonicus, and numerous tick spp.) and newly
-emerging diseases (West Nile virus, perhaps others).
These introductions produce both health and economic
consequences well beyond what anyone has expected
(Moore and Freier 2005).
Vectors and vector-borne pathogens are not evenly
distributed at any scale I can think of. They are basically
patchy in their distribution. Why is that? Often there are
different causes operating at different scales. We can use
LaCrosse (LAC) encephalitis cases in the United States as
At the national scale (Figure 4A), LAC cases are
basically confined to the eastern half of the country. At
this scale, LAC distribution is defined primarily by the
distribution of eastern hardwood forests, the habitat of
the major LAC vector, Aedes triseriatus (eastern treehole
mosquito). Historically, LAC activity has been highest
in the upper Midwestern states. More recently, however,
increasing numbers of cases are reported from the Mid-
Atlantic states, possibly due to the arrival of Ae. albopictus,
another efficient LAC vector (however, there are alternate
potential explanations for this change).
At the level of individual states (Figure 4B), we again
see a patchy distribution in LAC cases. The example shown
here is the state of West Virginia, where there has been a
major increase in reported LAC over the last 15-20 years.
The reasons for this pattern are not known, although I
suspect that it may largely be a case detection and reporting
When we focus on a finer scale, in this case the
county (Figure 4C), we again see that cases are not evenly
Figure 3. Schematic representation of the requirements for vector-borne zoonotic systems. All components are needed for
maintenance of the transmission cycle. Major barriers to dispersal can prevent the system from becoming established.
Vol. 33, no. 2 Journal of Vector Ecology 221
Figure 4. Distribution of reported LaCrosse (LAC) encephalitis cases in the United States at three scales. A- national scale, B
state scale, C- county scale (A- and B- from CDC unpublished data, C- data from Nasci et al. 2000).
Figure 5. Testing for non-randomness of LaCrosse case distribution in Nicholas County, WV using D-Map software (Rushton
and Armstrong 1997). A- interpolated case distribution surface; B- Probability distribution surface.
222 Journal of Vector Ecology December 2008
distributed. Here we see cases by census block group in a
single West Virginia county, Nicholas County (Nasci et al.
2000). The obvious first explanation for the difference in
cases within a county would be that, since cases are almost
exclusively reported in children under 15 years old, it is
likely there is an uneven distribution of young children
within the county. The sub-county boundaries (Figure 4C)
are census block groups (smallest practical subdivision with
useable data). The map shows cases per 1000 population age
15 and under. Obviously, this pattern is not related to the
distribution of susceptible children. It might also occur to
us that the observed “pattern” is not indicative of anything
at all—that it could have occurred purely by chance. We
now have methods to test that hypothesis.
Using the Disease Mapping and Analysis Program
(DMAP) developed by Rushton and colleagues (Rushton
and Armstrong 1997), I tested the hypothesis that the
observed pattern was due to chance. Figure 5A shows an
interpolated response surface created with the LAC data
from the raw data used to produce Figure 4C (point data
are not shown to protect the privacy of individuals). We can
now see pronounced areas of high and low LAC activity. I
tested the response surface for significance, using the Monte
Carlo routine provided in DMAP, and the results indicate
it is extremely unlikely that the high LAC areas are due to
chance. Similarly, there are areas of lower than expected
incidence—representing more urban communities with
less Ae. triseriatus habitat.
Figure 6. The compartmentalization of science has contributed to the inability of different disciplines to effectively interact
and collaborate on critical research issues.
In reviewing these data with collaborators at USGS in
Reston, VA, it appears that there is a subtle change in tree
species composition—associated with elevation—running
from southeast to northwest. It is possible that this small shift
in the plant community is sufficient to create the observed
LAC pattern, but that hypothesis remains untested.
What are the barriers to communication and cooperation
in vector ecology and related areas?
I suggest that there is a broad lack of communication
and coordination between the various disciplines and sub-
disciplines in biology and related sciences.
On several occasions, I have found it difficult to
communicate the utility of the tools and concepts of
community and landscape ecology to molecular biologists.
Similarly, I think, I have difficulty grasping and integrating
some of the concepts of molecular biology. I have, on several
occasions, attempted to explain the utility of ecological-scale
concepts and techniques to laboratory-based scientists,
with less than stellar results. Apparently I was not a good
communicator, as my ideas were dismissed as not applicable
to current program focus (read “molecular diagnostics”) or
too complicated to produce useful knowledge. Clearly, I was
not making the bridge from genes and gels to the level of
whole disease ecosystems.
This really irritated me. One evening, as I was pondering
the problem, I was watching my son manipulate a Rubik’s
cube—the plastic puzzle with many small cubes that is to be
Vol. 33, no. 2 Journal of Vector Ecology 223
Figure 7. Merging cube and cycle; the methods and perspectives of many disciplines can be brought to focus on vector-borne
to get integrated information flowing across and through
multiple levels to solve specific questions.
How might such an approach work in a vector-borne
disease setting? Figure 7 shows one possible scenario, where
a few of the possible puzzle pieces might fit in. I have taken
Figure 1 and suggested some of the different disciplines and
technologies as they might contribute to an understanding
of the total vector-borne disease system, again thinking
about the Rubik’s cube issue of fragmentation of disciplines.
What I want to suggest here is that no one of us has the
required level of expertise in all of the disciplines identified
in this figure (and there are undoubtedly many more that
I have left out). What is needed are scientists who, while
they are specialists at some level (piece of the puzzle), are
sufficiently “multilingual” to be able to communicate across
disciplines and work effectively with those occupying very
different puzzle pieces.
In addition to their importance to human and animal
health, vector-borne diseases are simply fascinating systems
to study. The involvement of multiple species over differing
space and time scales makes it extremely difficult to predict
epidemics. A single environmental factor may have opposite
adjusted so that all colors on a side are the same. It struck
me that the Rubik’s cube might form a visual aid to assist in
communicating some of the issues of communication and
coordination that we face.
Figure 6 shows a somewhat expanded Rubik’s cube,
labeled so as to convey some of the issues we face in
interdisciplinary communication and coordination. The
cube has 3 axes, which I have labeled Application, Scale, and
Organism. Each piece of the cube represents a particular
group of organisms, studied at a particular scale, with a
specific application area in mind. Each piece tends to have
its own language and acronyms (e.g., RAPDs vs NDVIs),
its own dogmas, journals, its university departments, and
so on. When we look at the whole of biology in this light,
it’s a wonder there is any collaborative research at all!
Now consider interactions between biologists and other
disciplines such as climatology or the social sciences. For
climatologists and meteorologists, for example, the term
vector has a very different meaning than for vector biologists.
Academic departments, government agencies, and other
groups are organized in such a way as to perpetuate this
Balkanization of science, but it seems to me that the only
way to move toward true multidisciplinary research is to
begin to break down the walls between the puzzle pieces—
224 Journal of Vector Ecology December 2008
impacts on the system at different points in time. Patchiness
at different geographical scales may have very different
causes, so it is important to identify the proper scale for
a particular study. New developments in remote sensing,
GIS, and spatial analysis make it easier to tease out causes of
Interdisciplinary collaboration is essential for many of
the projects we carry out, but this requires awareness of the
differences between disciplines and the ability to effectively
communicate with each other. It is only by forming inter-/
cross-/multi-disciplinary groups to focus on specific vector-
host-pathogen systems that we will be able to answer the
most interesting (and pressing) problems in our field. We
who study the ecology of vectors and vector-borne diseases
are concerned with only a tiny slice of the overall range of
possible infectious disease systems. But what a fascinating
slice it is!
I thank the members of SOVE and the awards committee
for this honor. There is not sufficient space here to name
all of the friends and colleagues who have contributed to
my education and thinking, but I am particularly thankful
to F. Louis Blanc who got me “hooked” on entomology,
and G.A.H. ”Andy” McClelland and George B. Craig for
their guidance in medical entomology and vector ecology.
Funded in part by NIH contract #N01-AI-25489.
Bennington, E.E., J.S. Blackmore, and C.A. Sooter. 1958.
Soil temperature and the emergence of Culex tarsalis
from hibernation. Mosq. News 18: 297-298.
Eldridge, B.F. 1987. Strategies for surveillance, prevention,
and control of arbovirus diseases in western North
America. Am. J. Trop. Med. Hyg. 37: 77S-86S.
Mail, G.A. and R.A. McHugh. 1961. Relation of temperature
and humidity to winter survival of Culex pipiens and
Culex tarsalis. Mosq. News 21: 252-254.
Monath, T.P. 1980. Epidemiology, pp. 239-312. In: T. P.
Monath (ed.), St. Louis Encephalitis. American Public
Health Association, Washington, DC.
Moore, C.G. and J. Freier. 2005. Invasive species and
emerging vector-borne infections. In: W. C. Marquardt
(ed.), Biology of Disease Vectors, 2nd ed. Elsevier-
Academic Press, Amsterdam. pp. 219-225.
Moore, C.G., B.L. Cline, E. Ruiz-Tiben, D. Lee, H. Romney-
Joseph, and E. Rivera-Correa. 1978. Aedes aegypti in
Puerto Rico: environmental determinants of larval
abundance and relation to dengue virus transmission.
Am. J. Trop. Med. Hyg. 27: 1225-1231.
Moore, C.G., R.G. McLean, C.J. Mitchell, R.S. Nasci, T. S.
Tsai, C.H. Calisher, A.A. Marfin, P.S. Moore, and D.J.
Gubler. 1993. Guidelines for Arbovirus Surveillance
Programs in the United States. Centers for Disease
Control and Prevention, U.S. Public Health Service,
Fort Collins, CO.
Nasci, R.S., C.G. Moore, B.J. Biggerstaff, N.A. Panella, H.
Q. Liu, N. Karabatsos, B. S. Davis, and E. S. Brannon.
2000. La Crosse encephalitis virus habitat associations
in Nicholas County, West Virginia. J. Med. Entomol.
Ostfeld, R., C. Jones, and J. Wolff. 1996. Of mice and mast.
BioScience 46: 323-330.
Ostfeld, R., C. Canham, K. Oggenfuss, R. Winchcombe, and
F. Keesing. 2006. Climate, deer, rodents, and acorns as
determinants of variation in Lyme-disease risk. PLoS
Biol. 4: 1058-1068.
Pavlovsky, E.N. 1966. Natural Nidality of Transmissible
Diseases With Special Reference to the Landscape
Epidemiology of Zooanthroponoses. University of
Illinois Press, Urbana, IL.
Rushton, G. and M.P. Armstrong. 1997. Improving public
health through geographical information systems:
An instructional guide to major concepts and their
implementation. University of Iowa, Iowa City, IA.
Available at http://www.uiowa.edu/~geog/health/
Shemanchuk, J.A. 1965. On the hibernation of Culex tarsalis
Coquillett, Culiseta inornata Williston, and Anopheles
earlei Vargas, in Alberta. Mosq. News 25: 456-462.
Tsai, T.F., C.J. Mitchell, and T.P. Monath. 1988. St. Louis
encephalitis, pp. 113-143. In: T.P. Monath (ed.), The
Arboviruses. CRC Press, Boca Raton, FL.