No abstract

Muddy floods, i.e. water flowing from agricultural fields and carrying large quantities of soil, affect routinely numerous municipalities of central Belgium, northern France and southern England. A comparison of flood frequency between different European regions is difficult, because of the lack of a uniform and official database as well as the landscape heterogeneity of administrative entities. Agri-environmental measures [AEMs; e.g. grass buffer strips (GBS)] can contribute to the control of muddy floods but their installation is voluntary and depends therefore on farmers' willingness. Actions to increase awareness and to inform the farmers proved to increase drastically their participation rate in AEM programmes. In all the studied regions, flood prone areas are increasingly taken into account to define land approved for development. Moreover, several schemes for the control of muddy floods have also been proposed at the regional scale. However, there is a spatial mismatch between the scale at which muddy floods are triggered (small catchment scale) and the scale at which public authorities can operate (municipality, grouping of municipalities, delineated flood prone areas, river basin). In future, beside curative measures (e.g. retention ponds and dams), farming techniques preventing runoff and erosion in the field (e.g. conservation tillage) should be encouraged. This could be achieved by the creation of a new AEM. Moreover, guidelines for the location of AEMs could usefully be introduced. Existing flood control schemes should also be systematically carried out by catchment agencies including legal, environmental and financial expertise. These agencies should be set up for local groupings of municipalities and provide them technical assistance to equip the flood prone areas and carry out maintenance of the implemented control measures.

No abstract

Humans have colonized and adapted to extremely diverse environments, and the genetic basis of some such adaptations, for example to high altitude, is understood. In some cases, local or regional variation in selection pressure could also cause behavioural adaptations. Numerous genes influence behaviour, such as alleles at the dopamine receptor locus D4 (DRD4), which are associated with attitude toward risk in experimental settings. We demonstrate genetic differentiation for this gene, but not for five unlinked microsatellite loci, between high- and low risk environments around Mount Merapi, an active volcano in Java, Indonesia. Using a behavioural experiment, we further show that people inhabiting the high risk environment are significantly more risk averse. We provide evidence of a genetic basis for this difference, showing that heterozygotes at the DRD4 locus are more risk averse than either homozygotes. In the high risk environment, allele frequencies are equilibrated, generating a high frequency of heterozygotes. Thus it appears that overdominance (i.e. selective advantage of heterozygotes) generates negative frequency dependent selection, favouring the rarer allele at this locus. Our results therefore provide evidence for adaptation to a marginal habitat through the selection of a neurocognitive trait with a genetic basis.