Assessment of heavy metals associated with bacteria in soil

Abstract

The heavy metals associated with bacteria in soil are dicult to quantify due to the problem of extracting sucient soil microbial biomass for analysis. The e€ects of soil minerals, organic matter and dissolved ions on the solubility and the reactivity of metals would all complicate the quanti®cation of heavy metals associated with bacteria. One approach to overcome this problem would be to separate the bacterial cells from soil in adequate quantities followed by analysis of heavy metals in the bacterial fraction. The essential steps are dispersal of the soil aggregates, followed by centrifugation in which cells and soil particles are separated according to size, density or both (Bakken and Lindhal, 1994). We describe a new method to quantify bacterial heavy metals that involve cell extraction and analysis of heavy metals in and absorbed onto cells. The method was tested in soil with or without manure addition. Samples were collected from two soil treatments (with or without addition of cow manure at the rate of 4% w/w) of a cultivated soil (Typic Cryoboroll). The soil samples (two replicates) were moistened with a tracer solution (a mixed solution of Cd and Zn) to bring the moisture content to 35% (w/w) and the total activity of the two tracers to 2000 Bq gy1 soil. After 1 week of soil incubation at 228C, the soils were used for cell dispersion and cell separation as follows. The soils were suspended in distilled water (soil:water=1:10) and homogenized in a Waring blender (3 1 min homogenization at full speed). The homogenates were placed in 250 ml centrifuge tubes and a cushion of Nycodenz (Nycomed Pharma AS, Torshov, Oslo, Norway) with a density of 1.3 g mly1 (24 g Nycodenz to 30 ml ®ltered distilled water) was placed underneath each homogenate. The cushions were placed with care to reduce mixing of the soil homogenate with the Nycodenz. The cushion of Nycodenz was large enough to extend above the rounded end of the centrifuge tubes. In a 250 ml centrifugation tube, we used a 50 ml Nycodenz cushion underneath 200 ml of soil homogenate. The samples were placed in a swing-out rotor to centrifuge at 10,000g for 60 min. After centrifugation, the bacterial cells had oated to the top of the Nycodenz cushion and soil material formed a soil pellet at the bottom of the tube. The supernatant above the bacterial layer was sampled separately (20 ml aliquots removed consecutively) to investigate the distribution of heavy metal in the washed portion. The bacterial layer (with or without a cushion) was harvested with a pipette. This bacterial fraction was diluted (1:20) and was subjected to prolonged centrifugation (2 h). This is because in the ®rst centrifugation (60 min at 10,000g), a true isopycnic equilibrium was often not reached. This is required for higher purity, since the cells lost are those with adhering soil colloids. After the prolonged centrifugation, the supernatents were discarded. The pure bacterial pellet was diluted (1:50) in distilled water and one portion analyzed for a bacterial count and the other for the activity of Cd and Zn. The soil pellet was also sampled for bacterial counts and for activity of the Cd and Zn. The percentage distribution of Cd and Zn between the fractions (supernatant, bacterial fractions and soil pellets) were estimated after counting the gamma emissions by Cd and Zn in the ®nal extracts and residues, by a NaI-detector. The counts were corrected for background emission and the energy spectra overlapping between the two nuclides, whereas the energy maxima are 88 and 1115 keV for Cd and Zn, respectively. The extraction eciency and the total bacterial count was investigated by acridine orange direct counting (AODC). The activity of Cd and Zn associated with bacteria (Table 1) are estimated by taking account of the bacterial count in both the bacterial fraction and soil pellet (Table 2). The Cd activity associated with bacteria was less than 1.3% of the total activity. Similarly, the Zn activities in the bacterial fraction were less than 3.6% of the total activities (Table 1). This shows that soil bacSoil Biology and Biochemistry 31 (1999) 315±316