The benefits of low-intensity mixed livestock-crop farming systems for Mediterranean farmland biodiversity have been widely documented. However, little is known about the effects of farming specialisation, even though the European Union's Common Agricultural Policy (CAP) has promoted the replacement of mixed systems by specialised livestock and crop systems. We investigated how specialisation affects birds in open farmland of southern Portugal, by analysing spatial associations between farming systems, land uses, and birds’ species richness and occurrence. We found a dominant landscape gradient contrasting specialisation in either sheep or cattle. The sheep system was positively related to woodland/shrubland-related species, Galerida larks and Melanocorypha calandra, while the cattle system was negatively related to these species. The specialised crop system, when associated with olive groves (< 10 % of the area), was negatively related to fallow-related species, M. calandra and Calandrella brachydactyla and positively so to overall species richness and that of woodland/shrubland-related species. Land uses had strong effects on birds, regardless of the farming systems in which they were embedded. Fodders were positively related to woodland/shrubland- and cereal-related species, Emberiza calandra and Cisticola juncidis. Rainfed grain cereals were positively related to cereal-related species, C. juncidis and Coturnix coturnix. Pastures were positively related to fallows- and ploughed-related species, M. calandra, Tetrax tetrax, and Galerida larks. Results suggest that programs for grassland conservation should consider both farming systems and land uses. CAP should promote diversified farming systems with reduced livestock densities along with high-quality habitats for grassland birds like pastures, fallows, rainfed grain cereals and semi-natural features.

The correct SPOM parametrization is one of the most crucial steps of every simulation procedure allowed by the package. The parameters are the numeric translation of the relation of the species with the landscape. Although this is not the main focus of this package, MetaLandSim offers some basic tools that will help to estimate the parameters to run the simulations. However, the user can have the parameters values available from other sources (whether published papers or estimated with other software tools) in each case the simulations can be run using those parameters. For a good overview of the several methods to parametrize the SPOM see Etienne et al. (2004). Most of the methods described can be implemented with MetaLandSim. Also, it is recommended that the user reads the book by Hanski (1999) in order to acquire the basic knowledge about metapopulation ecology and the first estimation procedures. The function parameter.estimate is the package tool to parametrize the spom function. This vignette clarifies the options available and gives a step-by-step guide of the function usage. The current version of the package does not executes all computations. Rather, some of the methods ((’MCsim’,’rescue’and ’norescue’) create the files needed as input to two applications provided by Moilanen (1999) (MCsim) and ter Braak and Etienne (2003) ((’rescue’ and ’norescue’). Future versions of the package should incorporate these procedures in the R code. The objective of this vignette is to describe the parametrization procedures, allowing the user to produce robust estimates of the parameters in order to proceed with the landscape simulation process or with the range expansion simulation. Third vignette is run with made up data based upon real information.