Limits...
A spatial simulation model for the dispersal of the bluetongue vector Culicoides brevitarsis in Australia.

Kelso JK, Milne GJ - PLoS ONE (2014)

Bottom Line: Data from midge trapping programmes were used to qualitatively validate the resulting simulation model.This extended model could then be used as a platform for addressing the effectiveness of spatially targeted vaccination strategies or animal movement bans as BTV spread mitigation measures, or the impact of climate change on the risk and extent of outbreaks.These questions involving incursive Culicoides spread cannot be simply addressed with non-spatial models.

View Article: PubMed Central - PubMed

Affiliation: School of Computer Science and Software Engineering, University of Western Australia, Crawley, Western Australia, Australia.

ABSTRACT

Background: The spread of Bluetongue virus (BTV) among ruminants is caused by movement of infected host animals or by movement of infected Culicoides midges, the vector of BTV. Biologically plausible models of Culicoides dispersal are necessary for predicting the spread of BTV and are important for planning control and eradication strategies.

Methods: A spatially-explicit simulation model which captures the two underlying population mechanisms, population dynamics and movement, was developed using extensive data from a trapping program for C. brevitarsis on the east coast of Australia. A realistic midge flight sub-model was developed and the annual incursion and population establishment of C. brevitarsis was simulated. Data from the literature was used to parameterise the model.

Results: The model was shown to reproduce the spread of C. brevitarsis southwards along the east Australian coastline in spring, from an endemic population to the north. Such incursions were shown to be reliant on wind-dispersal; Culicoides midge active flight on its own was not capable of achieving known rates of southern spread, nor was re-emergence of southern populations due to overwintering larvae. Data from midge trapping programmes were used to qualitatively validate the resulting simulation model.

Conclusions: The model described in this paper is intended to form the vector component of an extended model that will also include BTV transmission. A model of midge movement and population dynamics has been developed in sufficient detail such that the extended model may be used to evaluate the timing and extent of BTV outbreaks. This extended model could then be used as a platform for addressing the effectiveness of spatially targeted vaccination strategies or animal movement bans as BTV spread mitigation measures, or the impact of climate change on the risk and extent of outbreaks. These questions involving incursive Culicoides spread cannot be simply addressed with non-spatial models.

Show MeSH

Related in: MedlinePlus

Simulated arrival times following midge dispersal and population establishment.A: Midge arrival times shown by colour for the months following 1st October. White (ocean) and grey indicates areas in which no midge population became established. B: Contours showing monthly arrival time Zones: Zone 1 November, Zone 2 December, Zone 3 January and Zone 4 February.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4126746&req=5

pone-0104646-g007: Simulated arrival times following midge dispersal and population establishment.A: Midge arrival times shown by colour for the months following 1st October. White (ocean) and grey indicates areas in which no midge population became established. B: Contours showing monthly arrival time Zones: Zone 1 November, Zone 2 December, Zone 3 January and Zone 4 February.

Mentions: The combined Culicoides population dynamics and dispersal sub-models described in the Methods section was used in a simulation of C. brevitarsis dispersal and shown to replicate this southerly spread. The model used actual, spatially-explicit weather data for the 12 months from 1st October 1991. The simulated arrival times when the first C. brevitarsis appear are pictured in Figure 7A and by the monthly arrival “zones” overlaid onto the map in Figure 7B. Figure 7B allows comparison with Figure 5, which maps similar monthly arrival time isochrones but which are based on an extensive trapping program [5] (the zone boundaries in the right frame of Figure 7B were drawn by hand to visually separate the sites that fall into each zone). Table 1 presents monthly simulated arrival times together with the corresponding observed trapping months, allowing simulated and actual arrival times to be compared for all locations where trapping occurred, along with a quantitative measure of agreement (Cohen's kappa – see Section 2.4 “Quantifying agreement between simulated and observed midge spread”). Table 1 also includes observed versus simulated midge arrival time comparisons for the seasons immediately prior to 1991/92, namely 1990/91 and 1992/93. Data from these additional years was not used in the calibration of the model, and thus serves as a proof-of-concept validation of the model.


A spatial simulation model for the dispersal of the bluetongue vector Culicoides brevitarsis in Australia.

Kelso JK, Milne GJ - PLoS ONE (2014)

Simulated arrival times following midge dispersal and population establishment.A: Midge arrival times shown by colour for the months following 1st October. White (ocean) and grey indicates areas in which no midge population became established. B: Contours showing monthly arrival time Zones: Zone 1 November, Zone 2 December, Zone 3 January and Zone 4 February.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4126746&req=5

pone-0104646-g007: Simulated arrival times following midge dispersal and population establishment.A: Midge arrival times shown by colour for the months following 1st October. White (ocean) and grey indicates areas in which no midge population became established. B: Contours showing monthly arrival time Zones: Zone 1 November, Zone 2 December, Zone 3 January and Zone 4 February.
Mentions: The combined Culicoides population dynamics and dispersal sub-models described in the Methods section was used in a simulation of C. brevitarsis dispersal and shown to replicate this southerly spread. The model used actual, spatially-explicit weather data for the 12 months from 1st October 1991. The simulated arrival times when the first C. brevitarsis appear are pictured in Figure 7A and by the monthly arrival “zones” overlaid onto the map in Figure 7B. Figure 7B allows comparison with Figure 5, which maps similar monthly arrival time isochrones but which are based on an extensive trapping program [5] (the zone boundaries in the right frame of Figure 7B were drawn by hand to visually separate the sites that fall into each zone). Table 1 presents monthly simulated arrival times together with the corresponding observed trapping months, allowing simulated and actual arrival times to be compared for all locations where trapping occurred, along with a quantitative measure of agreement (Cohen's kappa – see Section 2.4 “Quantifying agreement between simulated and observed midge spread”). Table 1 also includes observed versus simulated midge arrival time comparisons for the seasons immediately prior to 1991/92, namely 1990/91 and 1992/93. Data from these additional years was not used in the calibration of the model, and thus serves as a proof-of-concept validation of the model.

Bottom Line: Data from midge trapping programmes were used to qualitatively validate the resulting simulation model.This extended model could then be used as a platform for addressing the effectiveness of spatially targeted vaccination strategies or animal movement bans as BTV spread mitigation measures, or the impact of climate change on the risk and extent of outbreaks.These questions involving incursive Culicoides spread cannot be simply addressed with non-spatial models.

View Article: PubMed Central - PubMed

Affiliation: School of Computer Science and Software Engineering, University of Western Australia, Crawley, Western Australia, Australia.

ABSTRACT

Background: The spread of Bluetongue virus (BTV) among ruminants is caused by movement of infected host animals or by movement of infected Culicoides midges, the vector of BTV. Biologically plausible models of Culicoides dispersal are necessary for predicting the spread of BTV and are important for planning control and eradication strategies.

Methods: A spatially-explicit simulation model which captures the two underlying population mechanisms, population dynamics and movement, was developed using extensive data from a trapping program for C. brevitarsis on the east coast of Australia. A realistic midge flight sub-model was developed and the annual incursion and population establishment of C. brevitarsis was simulated. Data from the literature was used to parameterise the model.

Results: The model was shown to reproduce the spread of C. brevitarsis southwards along the east Australian coastline in spring, from an endemic population to the north. Such incursions were shown to be reliant on wind-dispersal; Culicoides midge active flight on its own was not capable of achieving known rates of southern spread, nor was re-emergence of southern populations due to overwintering larvae. Data from midge trapping programmes were used to qualitatively validate the resulting simulation model.

Conclusions: The model described in this paper is intended to form the vector component of an extended model that will also include BTV transmission. A model of midge movement and population dynamics has been developed in sufficient detail such that the extended model may be used to evaluate the timing and extent of BTV outbreaks. This extended model could then be used as a platform for addressing the effectiveness of spatially targeted vaccination strategies or animal movement bans as BTV spread mitigation measures, or the impact of climate change on the risk and extent of outbreaks. These questions involving incursive Culicoides spread cannot be simply addressed with non-spatial models.

Show MeSH
Related in: MedlinePlus