Plant diversity in Mediterranean cereal fields: Unraveling the effect of landscape complexity on rare arable plants

• We examine changes in plant diversity depending on the changes on landscape structure.
• The major relevance of landscape structure for total species is at the boundary.
• The major relevance of landscape structure for rare arable plants is at the edge.
• Landscape structure modulates rare arable plants populations.
• Not all rare arable plants respond in the same way to changes in landscape structure.

Landscape complexity is thought to increase plant diversity in Mediterranean dryland cereal fields, although this assumption has not been tested specifically for rare arable plants (RAP). Standardized landscape metrics may help elucidate efforts to enhance RAP conservation. Our paper evaluates the impact of the surrounding landscape on the plant diversity and species composition for both total species and rare arable plants (RAP) at three contrasted field positions (boundary – first metre of non-cultivated habitat surrounding a field-, edge – first metre of cultivated habitat adjacent to the boundary- and centre) in 90 conventional cereal fields in 45 landscapes (2 fields per landscape) over three regions in the NE Iberian Peninsula (15 landscapes per region). Total species richness (S) and Shannon diversity (H′) were partitioned into α- and β-components. β-Components accounted the maximum contribution to the total diversity. Sαlandscape – species richness within-landscape and H′αlandscape – Shannon diversity within-landscape of total species and Sαlandscape of RAP were higher in structurally complex landscapes than in simple ones. The positive effect of landscape complexity on the αlandscape diversity was highest at the boundary for total species but at the edge for RAP. Two subsets of RAP, G1 and G2, were identified according to their response to landscape metrics gradients. G1 and G2 were assembled in complex and simplified landscapes, respectively. Landscape metrics explained the highest variation in species composition at the boundary and the edge for total species and exclusively at the edge for RAP. Moreover, the variation in species composition explained by landscape metrics was higher for RAP than for total species. Thus, RAP assemblages in arable fields are greatly influenced by processes operating at the landscape scale which may filter plant species. Our study provided a formal framework to help policy makers identify landscape configurations that most benefit plant conservation policies. As a depleted species pool may prevent the re-assembly of RAP, agri-environmental schemes targeting landscapes with useful structural elements supporting diversity may increase populations of RAP. Low-input farming practices at the edge of the arable fields in complex landscapes are expected to be the best cost-effective methods for enhancing RAP.