Habitat loss and fragmentation are considered to be one of the main causes of terrestrial biodiversity decline worldwide. The effect of fragmentation at a given level of habitat amount is however debated, and experimental tests remain scarce. To discriminate the effects of both quantity and fragmentation per se, we designed a landscape-scale experiment in which a wide range of habitat amount (9 levels from 1% of habitat to 100%) is crossed with three levels of fragmentation (from clumped to fragmented). Here, we investigate the short-term response of ground-dwelling arthropod communities to our experimental design with a multi-taxa approach. At the year scale, only the number of captures declined significantly with the habitat amount. The coming years will tell us more on a potential delayed effect of fragmentation on ground-dwelling arthropods.