Risk factors for Listeria monocytogenes contamination in French laying hens and broiler flocks

The objective of this study was to identify potential risk factors for Listeria monocytogenes contamination in French poultry production. Eighty-four flocks of layer hens kept in cages and 142 broiler flocks were included in this study. For each production type, a questionnaire was submitted to farmers and fecal samples were taken to assess the L. monocytogenes status of the flocks during a single visit to the farm. Two logistic regression models (specific to each production) were used to assess the association between management practices and the risk of L. monocytogenes contamination of the flock. The prevalence of L. monocytogenes-positive flocks was 30.9% (95% CI: 21.0; 40.9) and 31.7% (95% CI: 24.0; 39.4) for cage-layers and broiler flocks, respectively. For layer flocks, the risk of L. monocytogenes contamination was increased when pets were present on the production site. When droppings were evacuated by conveyor belt with deep pit storage, the risk of L. monocytogenes contamination decreased significantly. Feed meal was found to be associated with a higher risk of L. monocytogenes contamination than feed crumb. For broiler flocks, the risk of L. monocytogenes contamination was increased when farmers did not respect the principle of two areas (clean and dirty) at the poultry house entrance. A first disinfection by thermal fogging and the absence of pest control of the poultry house before the arrival of the next flock was found to increase the risk of contamination. When litter was not protected during storage and when farm staff also took care of other broiler chicken houses, the risk of L. monocytogenes contamination increased significantly. In the case of the watering system, nipples with cups were found to decrease the risk of contamination.     

The utility of process-based models for simulating N2O emissions from soils: A case study based on Costa Rican coffee plantations

Soil moisture and gaseous N-flux (N₂O, N₂) dynamics in Costa Rican coffee plantations were successively simulated using a mechanistic model (PASTIS) and two process-based models (NGAS and NOE). Two fertilized (250 kg N ha⁻¹ y⁻¹) coffee plantations were considered, namely a monoculture and a system shaded by the N₂ fixing legume species Inga densiflora. In situ N₂O fluxes were previously measured in these plantations. NGAS and NOE used specific microbial activities for the soils. To parameterize NGAS, we estimated N mineralization via in situ incubations and the contribution of heterotrophic soil respiration to total soil respiration. Potential denitrification rates and the proportion of denitrified N emitted as N₂O were measured in the laboratory to define the values of NOE parameters, as well as nitrification rates and related N₂O production rates for parameterizing both models. Soil moisture and both NGAS and NOE N₂O fluxes were best modelled on an hourly time step. Soil moisture dynamics were satisfactorily simulated by PASTIS. Simulated N₂O fluxes by both NGAS and NOE (3.2 and 2.1 kg N ha⁻¹ y⁻¹ for NGAS; 7.1 and 3.7 kg N ha⁻¹ y⁻¹ for NOE, for the monoculture and shaded plantations respectively) were within a factor of about 2 of the observed annual fluxes (4.3 and 5.8 kg N ha⁻¹ y⁻¹, for the monoculture and shaded plantations respectively). Statistical indicators of association and coincidence between simulated and measured values were satisfactory for both models. Nevertheless, the two models differed greatly in describing the nitrification and denitrification processes. Some of the algorithms in the model NGAS were apparently not applicable to these tropical acidic Andosols. Therefore, more detailed information about microbial processes in different agroecosystems would be needed, notably if process-oriented models were to be used for testing strategies for mitigating N₂O emissions.