Effects of vegetation management by mowing on ground-dwelling arthropods

Species-rich grasslands are rare in the Netherlands and need consistent vegetation management to retain their characteristic biodiversity. Roadside verges are important refuges for grassland plants since the mowing management no longer aims at traffic safety only but also strives for botanical diversity. Although arthropods are highly abundant in roadside verges, the effect of different mowing practices on this group is largely unknown. During 4 years, we studied ground beetles, weevils, ants and ground-dwelling spiders with pitfall traps in experimental plots in roadside verges with five different mowing treatments: (i) no management, (ii) and (iii) mowing once a year with and without hay removal, (iv) and (v) mowing twice a year with and without hay removal. This was done in a plant productivity gradient; the experiment was repeated in low-, medium- and high-productive verges. In the low-productive site, the effect of management on the arthropods only existed in a higher abundance in plots mown twice per year with hay removal. In the medium- and high-productive sites, mowing twice a year with hay removal resulted not only in highest abundances but also in highest arthropod species richness. Mowing twice without hay removal and mowing once with removal showed intermediate values, while mowing once per year without removal and particularly the absence of management resulted in low diversity and low abundance. To promote ground-dwelling arthropods in medium-to-high-productive grassland verges, we recommend a management of mowing twice a year with the removal of hay. It is reasoned that some form of rotational management, aiming at leaving some vegetation refuges intact after mowing events, may further promote arthropod survival. However, caution should be taken that these refuges are not too large, as overall suitability for ground-dwelling arthropod decreases rapidly in such patches. Out of several studied vegetation characteristics, the number of flowering plant species (medium-productive verge) and total flower abundance (high-productive verge) appeared to represent suitable, and easily monitored, proxies that significantly mirror arthropod diversity.