Microbial activity and biogeochemical cycling in a nutrient-rich meromictic acid pit lake

Microbial activity, abundance and biomass, and biogeochemical cycling of iron, sulphur and carbon were studied in the extremely acidic (pH 2.5–4.3), meromictic, nutrient-rich pit lake Cueva de la Mora in Spain. The goal was to find out (1) if the relatively high nutrient content influenced plankton abundance in the water column and alkalinity-producing microbial processes in the sediments compared to other acid pit lakes, and (2) if sediments in the shallow, mixed and the deep, stagnant parts of the lake exhibited differences in microbial activities and geochemical sediment composition related to meromixis. We hypothesised that redox cycling was more intense in the mixed part and higher amounts of reduced components would accumulate in the stagnant part. Especially phytoplankton biomass, CO2 production, and sulphate reduction were indeed higher than reported from typical acid pit lakes and were rather within the range of neutral or weakly acidic lakes, which can be attributed to the relatively high nutrient contents of Cueva de la Mora. Even in the monimolimnion, anaerobic processes occurred mainly in the sediments. Sediments from the mixed and stagnant parts of the lake differed markedly in their biogeochemistry. Mixolimnetic sediments showed high iron and sulphate reduction rates, and they appeared to undergo substantial recycling, as supported by reactive Fe, relation between gross sulphate reduction rate and net accumulation of reduced sulphur, and viable counts of iron and sulphur bacteria. Monimolimnetic sediments exhibited lower anaerobic microbial activities, and surprisingly they accumulated more Fe(II) than mixolimnetic sediments, but less reduced sulphur and carbon. This might be explained by a strong separation of the two water bodies, resulting in comparably less input of energy (light) and allochthonous matter into the monimolimnion. Regarding the total extent of alkalinity-generating microbial processes, their net effect is not sufficient to neutralis