Maize weed communities composition in a highly anthropized landscape: whichvegetation response variables and landscape descriptors?

Abstract: Few scientific guidelines are available for the definition of parameters to be taken intoaccount in studies aiming at determining the spatial interactions between weeds and theirenvironment. The choice of both landscape descriptors and weed response variables and the scale atwhich they should be measured is important. The objective of this study was to define amethodology to support the choice of landscape descriptors and weed response variables and thespatial scale at which their interactions occur. We assumed that weed communities respond tolandscape parameters based on the various ecological and biological characteristics of thecomponent species. A case study in a highly anthropized landscape along the Tuscan and Liguriancoastline is presented, with the main aim to determine how land abandonment and urbanisationaffect in-field weed communities. We selected a variety of landscape descriptors for (a) fieldmargins, (b) landscape structure and (c) landscape composition. Landscape structure andcomposition were considered in circles of various radii (100, 250, 500, 1000 and 1500m around thecentroid of each field). The sensitivity of each weed response variable depended on the scale atwhich the landscape descriptors were measured and on the characteristics of the field margins. Thisstudy shows that the selection of weed response variables, landscape descriptors and the scale atwhich data are collected are extremely important and may considerably affect results.