Characterization of the pollen beetle, Brassicogethes aeneus, dispersal from woodlands to winter oilseed rape fields.

Many crop pests rely on resources out of crop fields; understanding how they colonize the fields is an important factor to develop integrated pest management. In particular, the time of crop colonization and damage severity might be determined by pest movements between fields and non-crop areas. Notably, the pollen beetle, Brassicogethes aeneus, previously named Meligethes aeneus, one of the most important pests of winter oilseed rape, overwinters in woodlands. As a result, its abundance increases in oilseed rape fields near wooded areas. Here, we assessed the spatio-temporal patterns of the dispersal from woodlands to oilseed rape fields in diversified landscapes of a same region. We observed on four dates the abundance of pollen beetles in 24 fields spread in the Eure department, France. We modeled the abundance as a result of the dispersal from the neighboring woodlands. We compared the modalities of dispersal corresponding to different hypotheses on the dispersal origin, kernel shape and sources of variability. Within oilseed rape the distance to the edges of woodlands is not the main determinant of pollen beetle abundance. On the contrary, the variability of the abundance between fields is largely explained by the dispersal from neighboring woodlands but there is considerable variability between dates, sites and, to a lesser extent, between fields. The two dispersal kernels received similar support from the data and lead to similar conclusions. The mean dispersal distance is 1.2 km but seems to increase from a few hundred meters the first week to more than two kilometers the fourth, allowing the pollen beetles to reach more distant OSR fields. These results suggest that early varieties away from woodlands and late varieties close to the woodlands may limit attacks at the time when oilseed rape is the most sensitive.

Using biophysical models to manage nitrogen pollution from agricultural sources: Utopic or realistic approach for non-scientist users? Case study of a drinking water catchment area in Lorraine, France.

The objectives of this comparison of two biophysical models of nitrogen losses were to evaluate first whether results were similar and second whether both were equally practical for use by non-scientist users. Results were obtained with the crop model STICS and the environmental model AGRIFLUX based on nitrogen loss simulations across a small groundwater catchment area (<1 km(2)) located in the Lorraine region in France. Both models simulate the influences of leaching and cropping systems on nitrogen losses in a relevant manner. The authors conclude that limiting the simulations to areas where soils with a greater risk of leaching cover a significant spatial extent would likely yield acceptable results because those soils have more predictable leaching of nitrogen. In addition, the choice of an environmental model such as AGRIFLUX which requires fewer parameters and input variables seems more user-friendly for agro-environmental assessment. The authors then discuss additional challenges for non-scientists such as lack of parameter optimization, which is essential to accurately assessing nitrogen fluxes and indirectly not to limit the diversity of uses of simulated results. Despite current restrictions, with some improvement, biophysical models could become useful environmental assessment tools for non-scientists.