Agriculture Reference
In-Depth Information
Mouth Disease (FMD) (Blackall and Gloster, 1981; Gloster, 1983a) and Newcastle
Disease in poultry (Gloster, 1983b). This model was also used to investigate the
2001 UK FMD outbreak (Mikkelsen
et al.
, 2003).
Several studies of potential long distance aerial transport of plant pathogens have
used air parcel trajectory analysis to establish links between source and receptor
regions (Aylor, 1986; Davis, 1987). Trajectory analysis is a standard tool in the
study of air pollutant movement and it tracks the movement of air parcels using
information on wind fields and atmospheric temperature structures (Davis, 1987;
Stohl, 1998). It is widely used in air pollution studies and computational methods
and applications have been reviewed by Stohl (1998). Back-trajectory analysis of
wind contributed to evidence for long distance dispersal of exotic
Bacillus
bacteria
1800 km from the black sea to Sweden, where the species was isolated from red-
pigmented snow (Bovallius
et al.
, 1978). Web-based trajectory models are available
from the USA National Ocean and Atmosphere Administration (HYSPLIT model
http://www.arl.noaa.gov/) and the British Atmospheric Data Centre (NERC Centre
for Atmospheric Sciences, http://badc.nerc.ac.uk/community/). Trajectory modelling
can account for large scale movement of air parcels due to wind direction changes
and track air movements over large distances, although errors propagated during the
calculations must be carefully considered (Kottmeier and Fay, 1998). Trajectory
models can also be adapted to take into account particle dispersal within the air
parcel (Aylor, 1986; Davis, 1987; Aylor, 1999). In this approach, the dispersing
spores are treated as an expanding 'puff ' travelling along the path of the trajectory.
The expanding 'puff ' can be treated in a similar manner to the Gaussian plume
models, where the vertical (
z
), cross-wind (
y
) and along-wind (
x
) concentrations in
the puff are assumed to follow Gaussian distributions, unless constrained by an
atmospheric boundary, such as an inversion (Aylor, 1986, 1999). The spore
concentration (
C
) in the puff is:
⎡
⎤
⎡
⎤
⎡
⎤
2
Q
(
t
)
−
(
x
−
U
t
)
2
−
y
2
−
z
2
C
(
x
,
y
,
z
,
t
)
=
.
exp
.
exp
.
exp
(6.11)
c
⎢
⎢
⎥
⎥
⎢
⎣
⎥
⎦
⎢
⎣
⎥
⎦
(
2
π
)
3
/
2
σ
σ
σ
2
σ
2
2
σ
2
2
σ
2
⎣
⎦
x
y
z
x
y
z
where
Q
(
t
) is the proportion of released spores that are viable and still air-borne at
time after release
t
;
z
are the standard deviations of the distributions in
C
in the along-wind, cross-wind and vertical directions.
U
c
is a constant transport
speed representative of the atmospheric layer in which the spores are travelling. The
form of
Q
(
t
) can take account of loss of spores by deposition by sedimentation (dry
deposition), rainfall (wet deposition) and loss in viability (Aylor, 1999). Davis and
Main (1986) used a similar approach to investigate the spread of tobacco blue mold
(causal agent
Peronospora tabacina
) in south-eastern USA. This led to the
development of the North American Plant Disease Forecast Centre, North Carolina
State University, Raleigh, NC, that provides an internet-based disease risk
forecasting system for tobacco and cucurbit growers (Main
et al.
, 2001). The
scheme has been operating since 1996 and uses the NOAA HYSPLIT trajectory
model to calculate potential inoculum dispersal from areas where the disease is
σ
x
,
σ
y
and
σ
