A survey of symbiotic nitrogen fixation by white clover grown on metal contaminated soils

There is conflicting evidence about toxic effects of heavy metals in soil on symbiotic nitrogen fixation. This study was set-up to assess the general occurrence of such effects. Soils with metal concentration gradients were sampled from six established field trials, where sewage sludge or metal salts have been applied, or from a transect in a sludge treated soil. Additional contaminated soils were sampled near metal smelters, in floodplains, in sludge amended arable land and in a metalliferous area. Symbiotic nitrogen fixation was measured with N-15 isotope dilution in white clover (Trifolium repens L.) grown in potted soil that was not re-inoculated, and using ryegrass (Lolium perenne L.) as reference crop. The fraction nitrogen in clover derived from fixation (N-dff) varied from 0 to 88% depending on soil. Pronounced metal toxicity on N-dff was only confirmed in a sludge treated soil where nitrogen fixation was halved from the control value at soil total metal concentration of 737 mg Zn kg(-1), 428 mg Cu kg(-1) and 10 mg Cd kg(-1). The N-dff was significantly reduced by increasing metal concentration in soils from two other sites where N-dff was low throughout and where these effects might be attributed to confounding factors. No significant effects of metals on N-dff were identified in all other gradients even up to elevated total metal concentration (e.g. 55 mg Cd kg(-1)). The variation of N-dff among all soils (n = 48), is mainly explained by the number of rhizobia in the soil (log MPN, log (cells g(-1) soil)), whereas correlations with total or soil solution metal concentrations were weak (R-2 < 0.25). The log MPN is significantly affected by the presence or absence of the host plant at the sampling site. No effects of metals were identified at log MPN > 3, even at total Zn concentrations of about 2000 mg Zn kg(-1), whereas metal toxicity could be identified at lower most probable number (MPN) values. This survey shows that the metal toxicity on symbiotic nitrogen fixation cannot be generalized and that survival of a healthy population of the microsymbiont is probably the critical factor. (C) 2004 Elsevier Ltd. All rights reserved.