Towards an optimal sampling strategy to assess groundwater biodiversity: comparison across six regions of Europe

1. Reliable assessments of groundwater biodiversity are urgently needed to resolve current issues relating to the protection of aquifers. The assessment of groundwater biodiversity is hampered by the physical complexity and difficult access to the subterranean environment, which is related to the vastness, high degree of fragmentation and environmental heterogeneity of groundwater systems. Knowledge on groundwater biodiversity is also biased towards penetrable karstic habitats (caves), whereas other common habitats such as those found in porous aquifers have been neglected. This situation calls for a standardised and comprehensive strategy to sample an exhaustive and balanced set of groundwater habitats. 2. A standardised sampling protocol aimed at capturing the main sources of environmental heterogeneity within regions was applied in six regions across Europe. Four hierarchical levels were considered: (i) region (c. 400 km2); (ii) basin (c. 100 km2); (iii) aquifer type (karstic or porous) and (iv) habitat (hyporheic and phreatic zones for porous aquifers; saturated and unsaturated zones for karst aquifers). A total of 192 sampling sites equally distributed among habitats were sampled within each region. 3. Stygobiotic species richness significantly varied across regions, probably as a result of important difference in physical and biogeographical characteristics among the regions. Only one species (Graeteriella unisetigera) occurred in all six regions, underlining the narrow geographic range and high degree of endemism of stygobiotic fauna. The low frequency of occurrence of stygobionts also points to the importance of rarity in ground waters and its relevance for drawing up sampling designs. 4. Rarefaction curves were calculated to determine sampling efficiencies within each region. Despite the high sampling effort, the curves did not reach saturation, especially in the Cantabria, Lessinia and Krim regions, which had the greatest numbers of rare species. 5. Species accumulation curves were also calculated by considering the main sources of environmental heterogeneity among basins, aquifer types and habitats captured by the sampling protocol. In two regions (Roussillon and Jura) sampling efficiency was improvedby considering several basins, aquifers and habitats. In the other regions the effect of environmental heterogeneity among basins was low, or absent, but efficiency was higher in the Lessinia and Krim regions when considering both kinds of aquifers and all four habitats. In Cantabria and Wallonia, possible improvement would be expected only when sampling effort is increased or other sources of environmental heterogeneity are integrated. 6. General recommendations to refine the sampling strategy thus include identification of sources of environmental heterogeneity such as latitude, altitude, anthropogenic disturbances and regional history. Optimal hydrologic conditions during sampling are an important prerequisite for reliable assessments of groundwater biodiversity and repeated sampling is necessary if this requirement is not met.