Methods
Hydrology
Landscape Genetics
Genetic Analysis
Geographic Information Systems (GIS) and Remote Sensing (RS)
GPS/Tracking
Field Epidemiology
Molecular detection / Cercariometry
Snail Sampling
Stochastic Individual-Based Modeling
________________________________________________________
Hydrology
We use lumped
parameter rainfall-runoff and Geographic Infrormation Systems (GIS)
-based hydrological modeling techniques to characterize inter-village
hydrological transport between sites in our study area. We use remotely
sensed digital elevation models to map and characterize hydrological
pathways. Combining these
data with direct measures of parasite diffusion and snail dispersal, we
aim to predict:
(a) Dispersal
among sites upstream and downstream within a watershed,
(b) Dispersal
among distant sites within a watershed.
(c) Dispersal
among sites straddling
watershed boundaries.
We have good
information on the transport of cercariae via hydrological connectivity
from previous studies in the Xichang area of
China.
We have an elevation model from SRTM for the entire study region, which
may be refined to look at more detailed watersheds than simple
inter-village slopes, as used in Xichang, for the purposes of defining
inter-village hydrological connectivity.
Publications:
Hydrology
(Top)
Landscape Genetics
Landscape genetics
is an emerging field integrating landscape
ecology, population genetics, and spatial statistics to better
understand how physical, biological, and chemical variation in the
landscape shape genetic diversity and structure. We are currently
applying landscape genetic tools to model the influence of landscape
variables such as riparian habitat quality, topography, and hydrology on
snail migration and parasite diffusion. We are employing a
geographic information systems hydro-model to delineate potential
migration corridors, accounting for topographic barriers and landscape
variation that may facilitate or impede gene flow between study sites.
(Top)
Genetic Analysis
To estimate gene flow in both snail
intermediate host and different parasite life-stage populations, we are
using established panels of molecular markers (AFLP and msats,
respectively). We apply genotype assignment methods to analyze recent
gene flow; these techniques derive information about migration within
the last few generations from transient disequilibrium at individual
multilocus genotypes of migrants or their recent offspring. We apply
frequency methods to estimate long-term gene flow; these methods test
hypotheses about an event based on the expected frequency of that event
happening over a large number of trials, and are based in simplified
models of population structure. These two classes of techniques provide
information about gene flow on different timescales, and are thus
complementary.
Publications:
Genetic Analysis
(Top)
Geographic Information Systems
(GIS)
and Remote Sensing (RS)
Remote
sensing and GIS
technology promise to aid the identification of new areas of potential
snail habitat and sites with high potential for disease transmission,
based on image analysis and specification of land-cover features
associated with agriculture and human habitation.
GIS
analyses will also aid in understanding hydrological connectivity and
land use changes that may be associated with social connectivity.
Remote sensing and GIS
have been used effectively in the study of several infectious diseases
including malaria, Lyme disease, and rift valley fever (Wood, Washino et
al. 1991; Dister, Beck et al. 1993; Beck, Rodriguez et al. 1994; Washino
and Wood 1994; Glass, Schwartz et al. 1995; Linthicum, Bailey et al.
1997).
Our work includes the establishment of a
GIS
to manage field epidemiological data
(Spear et al, 1998), and the creation of new
GPS-based
protocols for mapping snail density (Seto et al, 2001).
An early remote sensing study in our group found that it is
possible to use Landsat TM data to identify the habitat of the
Oncomelania snail in the Anning River Valley of Sichuan (Seto et al,
2002), and in Poyang Lake in the Lower Yangtze River Valley (Seto et al,
2007, Davis, et al, 2003, Wu et al, 2002, Seto et al, 2002).
Recent studies in our group have quantified snail densities via remote
sensing (Xu, et al, 2004, Xu, et al, 2006).
Most recently GIS
methods are being used to map social and hydrological pathways to
understand parasite diffusion and disease re-emergence.
Publications:
GIS/RS
(Top)
GPS/Tracking
Water contact questionnaires, self-reported
contact diaries, and direct observation are all typical methods for
assessing water contact behaviors that may be associated with
schistosomiasis transmission.
However, such methods are problematic due to poor recall for
surveys and diaries, and altered behavior for observational studies.
Ultimately, we would like to know where, when and the intensity
of water contact behaviors and their relationship to parasites in the
environment. Global
positioning system (GPS) receivers may be used for personal
time-activity monitoring to assess these relationships.
By monitoring study subjects with GPS vests, it was possible to
create hourly time-activity maps, which were subsequently used in
interviews to ascertain the timing and location of water-contacts. We
found that individuals averaged more than one water contact per day, and
were surprisingly mobile, with a large number of study participants
spending time outside of their village. Such mobility suggests the need
for further research into social patterns that may facilitate the spread
of parasites, and contribute to sustained transmission (Seto, et al.,
2007).
Publications:
GIS/RS
(Top)
Field Epidemiology
Field epidemiologic studies have played an important role in all our
work, whether it be parameterization and calibration of our mathematical
model, understanding the spatial and temporal determinants of infection
risk and re-emergence, identifying effective strategies for sustainable
local control, and elucidation of the roles that parasite diffusion and
social and hydrological factors play in transmission.
Much of our work has environmental epidemiology, emphasizing the
role that environmental factors play in allowing for and mediating
transmission. Studies from
our first field site in
Xichang,
Sichuan
established the importance of risk factors that operate at the village
level (Spear et al, 2004).
Kinship analysis further reinforced this notion, showing that after
adjusting for village of residence, additive genetic factors at the
household and individual-level were not responsible for intensity of
infection (Seto et al, 2005).
Most recently, our field epidemiologic data has led to a better
parameterization of individual-level exposure that has shown to be
correlated with both infection status and intensity of infection (Seto
et al, 2007, Lee and Seto, in review).
This has allowed for a better understanding of within-village
risk. Other work has
explored the linkage between schistosomiasis and cancer (Qiu et al,
2005), and how variation in egg counts may be used to estimate the
distribution of work burdens for different at-risk populations (Hubbard
et al, 2002).
Publications:
Field Epidemiology
(Top)
Molecular
Detection / Cercariometry
The
current method for estimating the spatial variation in cercarial
concentration in village water bodies is to expose mice to potentially
infective surface water at different points along a water course for a
total exposure period of 10 hours. Because the
exposures are integrated over multiple days, reasonable information is
gained on the relative hazard of each location if the timing of the
assays is reasonably coincident , yet the method has extremely limited
temporal and spatial resolutions (Spear et al, 2004).
In response, we have pioneered a highly sensitive and specific
PCR assay for the detection of genomic cercarial DNA in water samples
(Driscoll, 2005).
We have also developed methods
for determining the decay rates for cercariae infectivity with distance
by conducting cercarial release experiments (Lowe et al, 2005).
These decay functions, along with a remote sensing-derived digital
elevation model were used in a spatial dynamic model that assessed
intervillage hydrologic connections for approximately 200 villages in
Xichang (Xu et al, 2006).
Publications:
Cercariometry ,
Genetic analysis
(Top)
Snail
Sampling
We have created a
protocol for performing geographically randomized snail surveys for
schistosomiasis research using the global positioning system (GPS) (Seto et al, 2001). This protocol differs from traditional surveys in
its ability to accurately map and measure the spatial distribution of
snail habitat. The protocol was used to map irrigation ditches, the
primary habitat for Oncomelania
hupensis, in two residence areas in
Sichuan Province,
China.
We have also explored the habitat of the snail in both the Anning
River Valley and Poyang Lake (Seto et al, 2002, Seto et al, 2002)
We have also developed a novel, longitudinal mark-recapture for
for Oncomelania hupensis, the intermediate host for
Schistosoma japonicum, to
better understand the population dynamics of the intermediate host (Remais
et al, 2007). Snail density, recruitment and death rates were
estimated monthly and environmental variables recorded continuously, at
three sites in a mountainous region of southwestern
China.
We have incorporated these snail population parameters from field
surveys into a mathematical population dynamics model, with the
objective of informing control decisions
Publications:
Snail ecology
(Top)
Stochastic Individual-Based Modeling
We
use this model to explore the efficacy
of various surveillance strategies.
A
stochastic model enables us to simulate all events in the population
relevant to the dynamics of the transmission of the disease: the
acquisition of a new worm by an individual (infection events), the death
of a worm (death events), the introduction in the village of an infected
individual or animal, and the introduction of snails or larval stages of
the parasite through the irrigation system (introduction events). The
ability to account for the stochasticity linked to the discrete nature
of the worm population, often called demographic stochasticity is
especially important in the context of an early epidemic when that
parasite population is still very small.
Publications:
Transmission modeling
(Top)