GROUNDWATER SAMPLES DO NOT REFLECT BACTERIAL DENSITIES AND ACTIVITY IN SUBSURFACE SYSTEMS

Total cell numbers, abundance of respiring bacteria and [(3)]thymidine and [C-14]leucine incorporation rates were investigated in four groun dwater wells of low nutrient content. Total cell numbers in the pumped groundwater were low (14 x 10(3) to 279 x 10(3) cells mL(-1)), and [H -3]thymidine and [C-14]leucine incorporation rates were, with one exce ption, below the detection limit. Therefore we exposed sediments in si tu for 2 months which allowed us to determine bacterial numbers and in corporation rates of labeled substrates by bacteria attached to sedime nt particles. The two habitats differed considerably in all bacterial parameters both in magnitude and seasonal trends. Total bacterial numb ers of sandy sediments (52.1 +/- 21.3 x 10(6) cells cm(-3)) correspond ed in average to 663 cm(3) of pumped groundwater (78.5 +/- 61.5 x 10(6 ) cells L(-1)). For the fraction of respiring bacteria this ratio was on average 3032 cm(3) (sediments: 10.3 +/- 5.4 x 10(6) respiring cells cm(-3); groundwater: 3.39 +/- 6.01 x 10(6) respiring cells L(-1)). Th e percentage of respiring bacteria in sandy sediments was between 6.0 and 41.4% (average 19.8) compared to 1.0 to 24.9% (average 5.23) in th e pumped groundwater. Our results stress the importance of studying th e microbial communities attached to sediment, as pumped groundwater sa mples may not be representative for the real structure and dynamics of microbial assemblages in subsurface environments. (C) 1997 Elsevier S cience Ltd.