Artificial groundwater recharge in forests : soil fauna and microbiology

At present, approximately half of the drinking water for the city of Basel (Switzerland) is obtained by an artificial groundwater recharge system in a former floodplain area called “Lange Erlen”. Generally, the use of groundwater for drinking water production may lower groundwater tables. Artificial groundwater recharge is a practice of directing and simultaneously purifying water into aquifers, thereby raising the groundwater table and guaranteeing sufficient drinking water sources. Water purification through artificial groundwater recharge is widespread. However, it more commonly involves areas without vegetation, i.e. slow sand filters, dunes or channels and is combined with long flooding periods. In contrast, at the “Lange Erlen”, forested areas are periodically flooded (max. 10 days) with water from the river Rhine. This routine is interrupted by longer regeneration periods. To date, water infiltration and purification processes have remained constant and satisfactory since the system has been established almost 100 years ago. However, detailed knowledge on the belowground processes that have been sustaining the water purification capacity of “Lange Erlen” is scarce.
Intensive flooding may detrimentally affect earthworm populations and result in clogging of the topsoil, which is a common problem in groundwater recharge systems. Earthworms are known to influence water infiltration and aeration of soils, but most of the existing knowledge relates to grasslands and little is known about the role of earthworms for water infiltration in forests. To better understand the “Lange Erlen” system in the shallow soil layer, earthworm populations, microbial parameters (substrate induced respiration, SIR) and water infiltration rates were studied at the recharge areas. The findings suggest that earthworms are an important factor of the long-term sustainability of this system (for the past ~100 years). The total earthworm
numbers and biomass in watered sites exceeded those of non-watered sites (+51% and +71%, respectively). Total earthworm numbers, numbers of endogeic (mineral forms) and epigeic (litter dwellers) earthworms, and numbers of two species (Lumbricus rubellus and Allolobophora chlorotica) significantly and positively correlated with water infiltration rates. Microbial biomass and activity was significantly enhanced in the top soil layer of the watered sites. The results imply that the flooding regime at the “Lange Erlen” favors earthworm populations which in turn prevent soil clogging, aerate the top soil layer, and stimulate microbial growth.
Groundwater quality is directly influenced by subsurface microbial, chemical and physical soil processes. However, most studies on microbial communities have been limited to the top soil layer. These studies disregarded deeper soil horizons although subsurface microorganisms are crucial for the degradation of natural organic compounds or contaminants and the maintenance of groundwater quality. Therefore, vertical soil profiles down to approximately 4 m of depth from two watered sites and one non-watered site were investigated for the structural (phospholipid fatty acids, PLFAs) and the functional (extracellular hydrolytic enzymes) microbial community composition. Furthermore, additional microbial (by SIR), physical and chemical soil parameters were obtained from the same soil samples.
The microbial biomass did not differ between watered sites and the non-watered site, however considerable fractions of the microbial biomass (25-42% by PLFA and 42-58% by SIR) were located in 40-340 cm depth at all sites. The microbial activity (CO2 emission) and the specific respiration (qCO2) were highest at the watered sites. The microbial community structure differed significantly between watered and non-watered sites (predominantly below 100 cm depth), whereas the functional structure (based on the relative enzyme pattern) differed significantly between all sites. The latter finding could probably be explained by different soil structures in each soil
profile rather than by flooding. Proportions of the bacterial PLFAs 16:1ω5, 16:1ω7, cy17:0 and 18:1ω9t, and the long chained PLFAs 22:1ω9 and 24:1ω9 were more prominent at the watered sites, whereas branched, saturated PLFAs (iso/anteiso) dominated at the non-watered site. The PLFA community indices indicated stress response and higher nutrient availability due to flooding. The analysis of extracellular soil enzymes revealed that acid phosphatase showed highest absolute activities at all field sites throughout the soil depth transect and was followed by L-leucine aminopeptidase and β-glucosidase. Combining the structural and the functional diversity of the microbial community in one analysis revealed significant correlations between the PLFA pattern and specific enzymes activities in the non-watered site. However, at the watered sites these relationships were not detected and the same factors appeared uncoupled from each other.
Overall, this implies that adding labile nutrients (i.e. DOC or DON by flooding) to a soil where other nutrients are limiting microbial growth (i.e. P as indicated by acid phosphatase) increases microbial activity but not biomass. This in turn results in waste respiration by overflow metabolism. Additionally, slight nutrient leaching (e.g. nitrate) into the groundwater is observed due to P-limiting conditions.
No differences in absolute and specific enzyme activities between watered sites and the non-watered site indicated complex organic matter input at the recharge sites to be impeded by flooding water pretreatment. In conclusion, water recharge processes resulted in a microbial community adapted to resource and environmental conditions, which was predominantly located in the upper (100-220 cm depth) and partly in the lower vadose zone (220-280 cm depth). Given a better understanding, the system may be more widely adopted and used to provide sufficient and reliable drinking water to the city of Basel.