Résumés des présentations - Webinaire Juillet 2021

Panayiota Touloupou (Univ. Birmingham ; oratrice invitée) - Statistical methods for linking geostatistical maps and transmission models

Geostatistical modelling is increasingly used in epidemiology to combine surveys from multiple locations into a detailed model of local prevalence or incidence. Spatial heterogeneity is recognised as an important epidemiological factor in many diseases, however predictions of future cases are frequently performed on aggregated data, risking the epidemiological fallacy.

When mathematical modelling is used to evaluate potential intervention strategies, spatial heterogeneity is also frequently ignored. Motivated by this fact, we developed a novel methodology to combine geospatial models of disease prevalence with transmission models of epidemic dynamics in order to make maps of future projections of disease. An important feature of the approach is the ability to capture uncertainty at every stage through a Bayesian framework.

The value of using the developed methodology is demonstrated on lymphatic filariasis in Africa, where by exploring the effect of a variety of intervention strategies on the future predictions, we will able to give advice to the national control programmes on the best routes towards elimination with their appropriate uncertainty.

 

Laura Poitou (INRAE) - Modeling the relationship between temperature and development of the pine processionary moth to predict its phenology 

Insect development is heavily impacted by temperature and the development time could vary across life-stages. In the literature, two types of model are generally used to simulate insect development: 1) linear models that are easy to use but not reliable near developmental limits and 2) non linear models that are more robust but need substantial datasets to be correctly parameterized. The good practice is to test both types of models and compare their predictions to find the best model.
The pine processionary moth (PPM), Thaumetopoea pityocampa, is an important forest pest in Europe currently expanding its range with climate warming. Due to its univoltine development, this species is also suspected to respond to climate change with shifts of its life cycle depending on climatic features. To simulate PPM phenology, we built three phenology models: 1) a linear model based on degree-days needed to achieve the development of each life stage, 2) a non-linear model based on the thermal performance curve (TPC) of each stage considering hourly mean temperatures, and 3) a non-linear model also based on the TPC of each stage but considering daily mean temperatures. To compare the accuracy of each model, their predictions were compared to phenological observation data. The non-linear model with daily mean temperatures was found to best fit the observations. This model was used to explore the effects of climate warming on PPM phenology

 

Pauline Clin (Institut Agro Rennes) - Host mixtures for plant disease control: from induced pathogen competition to the benefits of immune priming

Host mixtures are a promising method for agroecological plant disease control. First, we developed an epidemiological model in n dimensions with a large number of resistant host genotypes to explore the effect of host mixtures on the equilibrium prevalence of the disease. A significant amount of resistance genes must be deployed to achieve low disease prevalence, especially if the cost of resistance
breaking is low. Also, plant immunity is key to the success of host mixtures against polymorphic pathogen populations. This results from priming-induced cross-protection, whereby plants able to resist infection by specific pathogen genotypes become more resistant to other pathogen genotypes. Strikingly, this phenomenon was thus far absent from mathematical models aiming at designing host mixtures. We previously showed in a simple model of gene-for-gene interaction that priming reduces the prevalence of the disease by the cross-protection process (Clin et al, 2020). We here extend the previous n-dimensional model to take priming into account and show that it could help reduce the number of plant genotypes needed to drop the prevalence below an acceptable level. Given the limited availability of resistance genes in cultivars, this mechanism of plant immunity would make the use of host mixtures more realistic.

P. Clin, F. Grognard, L. Mailleret and F. Hamelin.

 

Pauline Ezanno (INRAE) - Movement rewiring among relevant herd statuses to control paratuberculosis at a regional scale

Paratuberculosis is a worldwide disease mainly introduced through trade. Protecting herds from purchasing infected animals is difficult due to the low sensitivity of diagnostic tests. Our objective was to assess if trade movement rewiring to promote risk-based movements could reduce the spread of Mycobacterium avium subsp. paratuberculosis (MAP) between dairy cattle herds at a regional scale. Two levels of control strategies were assessed. At the between-herd scale, trade rewiring aimed to prevent animals from high risk herds moving into low risk herds. At the within-herd scale, complementary additional measures were considered according to the herd infection status, aiming at limiting the within-herd spread through reducing calf exposure to adult faeces and shortening the delay before culling animals detected as highly test-positive. We used a stochastic individual-based and between-herd mechanistic epidemiological model, adapted to the 12,857 dairy cattle herds located in Brittany, western France. We compared MAP regional spread when using the observed trade movements versus a rewiring algorithm rendering trade movements risk-based. Herds were distributed annually into three statuses based on the test results of all their females older than two years of age, accounting for test sensitivity: A if the estimated true prevalence was below 7%, B if it ranged from 7 to 21%, C otherwise. We also identified herds having a high probability of being MAP-free (AAA herds having obtained an A status on three consecutive years) to assess the effect on MAP regional spread of decreasing their risk of purchasing infected animals. We showed that movement rewiring to prevent selling animals from high- to low-prevalence herds reduces MAP regional spread. Targeting AAA herds enabled minimizing the control effort to decrease regional MAP spread. However, animals purchased by AAA herds should be guaranteed as MAP-free with a moderate to high level, especially if the risk to purchases animals from herds of unknown status cannot be managed. Hygiene improvement and early culling of positive animals were relevant complementary on-farm control options to further decrease MAP spread. Future studies should identify how best to define herd statuses to optimally target combinations of control measures to most effectively reduce the spread of MAP on a regional scale.