FUTURE Research directions
Throughout different projects and
parnterships, we have been, or we are yet implicated in geospatial
analyses of HPAI H5N1 in many countries, including Thailand, Vietnam,
Indonesia, Bangladesh, India, China, and Egypt. At some point, we aim
to integrate the outcomes of those different analyses so that we can
better indetify emerging properties of HPAI H5N1 epidemiology
encountered in different environmental, agricultural and social
conditions. Furthermore, much of our work so far focused on the spatial
dimension of epidemiological data, and we aim to better explore how
space-time dynamics of HPAI H5N1 outbreaks can be studied.
Toward a better integration of country-scale studies
Toward improving the qualtity of key epidemiological indicators
Our previous work demonstrated the importance of segregating analysis of host species in order to understand the epidemiology of HPAI. Obtaining data on poultry composition was therefore essential to analyse HPAI risk (in particular separating chickens, ducks and geese). The only available global raster database of poultry pools together all poultry species (FAO Gridded Livestock of the World, GLW), which could be improved by breaking it down by poultry type.The quality of these data ranged from recent species-specific data sets with separate categories of production systems (e.g., Thailand) to country-level data where all categories are pooled together in a simple “poultry” count (e.g., China). A first step toward improving this was thus to compile in collaboration with FAO available species-specific data for chickens, duck and geese. We are now developping revised modelling approaches based on those data to predict the distribution of species-specific poultry following the method presented in Wint and Robinson (2007).
The density of hosts alone is of lower epidemiologically informative value if it can not be divided in production sectors. For example, very high host population density have very different epidemiological implications if they concerns industrialized production (e.g., Netherlands, Denmark, Belgium) or extensive production (e.g., Indonesia, Egypt). To date the FAO Global Livestock Production and Health Atlas, which contains data on different poultry groups for different categories does not differentiate between different poultry sectors and could be improved in that regard. We aim to collaborate with FAO to enhance modelling differrent levels of poultry production intensification.
Develop complementary approaches through collaboratoin: mathematical modelling and evolution
Two complementary aspects of HPAI H5N1 epidemiology will be studied in the future.
First, we are co-investigator of a project submitted by Michael Tildesley (Univ. Edimburg). Should the project be funded and we will explore how mathematical modelling can be used to adress two questions: i) how do the conditions of HPAI H5N1 persistence change as a function of the duck / chicken ration in a given landscape, and ii) compare the outcomes of different control strategies.
Second, HPAI H5N1 continuously evolves in countries where it has been introduced, spread and persists. These countries differ considerably in their agro-ecological conditions, including in their level of poultry production intensification and production structure, the availability of domestic poultry reservoirs, the distribution and abundance of live-bird markets, and the abundance and composition of potential alternate hosts. A reasonable assumption, supported by work on other agriculture-associated pathogens is that variation in these conditions should influence the virus’s evolutionary dynamics as different environments select for different viral features. This work aims to explore how the multidimensional set of agro-ecological factors relate quantitatively to HPAI H5N1 evolutionary dynamics, and will be carried out in collaboration to Rob Wallace (Univ. of Minnesota), Lenny Hogerwerf (Utrecht Univ.), Jan Slingenbergh (FAO) and (Yale Univ.).
Understanding the geography of production intensification
Beyond avian influenza, intensification of animal production has many epidemiological implications (see the background section). Hpowever, to date the spaio-temporal dynamics of the intensification of animal production has not been studied quantitatively. Using a set of represetnative countries, we aim to better characterize the geography of animal production intensification, so that forecasts of future animal production can be translated into future distribution patterns.