Effect of soil ammonium concentration on N2O release and on the community structure of ammonia oxidizers and denitrifiers

Abstract

The effect of ammonium addition (6.5, 58, and 395 microg of NH4+-N g [dry weight] of soil(-1)) on soil microbial communities was explored. For medium and high ammonium concentrations, increased N2O release rates and a shift toward a higher contribution of nitrification to N2O release occurred after incubation for 5 days at 4 degrees C. Communities of ammonia oxidizers were assayed after 4 weeks of incubation by denaturant gradient gel electrophoresis (DGGE) of the amoA gene coding for the small subunit of ammonia monooxygenase. The DGGE fingerprints were invariably the same whether the soil was untreated or incubated with low, medium, or high ammonium concentrations. Phylogenetic analysis of cloned PCR products from excised DGGE bands detected amoA sequences which probably belonged to Nitrosospira 16S rRNA clusters 3 and 4. Additional clones clustered with Nitrosospira sp. strains Ka3 and Ka4 and within an amoA cluster from unknown species. A Nitrosomonas-like amoA gene was detected in only one clone. In agreement with the amoA results, community profiles of total bacteria analyzed by terminal restriction fragment length polymorphism (T-RFLP) showed only minor differences. However, a community shift occurred for denitrifier populations based on T-RFLP analysis of nirK genes encoding copper-containing nitrite reductase with incubation at medium and high ammonia concentrations. Major terminal restriction fragments observed in environmental samples were further described by correspondence to cloned nirK genes from the same soil. Phylogenetic analysis grouped these clones into clusters of soil nirK genes. However, some clones were also closely related to genes from known denitrifiers. The shift in the denitrifier community was probably the consequence of the increased supply of oxidized nitrogen through nitrification. Nitrification activity increased upon addition of ammonium, but the community structure of ammonium oxidizers did not change.

FIG. 1.

Effect of incubation at LA, MA, and HA concentrations on ammonium transformation. The stacked bars indicate percent contributions of nitrification (open) and denitrification (shaded) to total N₂O emission, and the small squares indicate the rates of total N₂O emission. Means ± standard errors are shown (n = 3).

FIG. 2.

DGGE analysis of amoA fragments from soil samples taken before treatment (ZT) and after incubation with LA, MA, and HA concentrations at 4°C for 4 weeks.

FIG. 3.

Phylogenetic Fitch-Margoliash tree (using global rearrangement and randomized input order [three jumbles]) based on partial amoA sequences (150 amino acids). Clones obtained from this experiment are shown in boldface. The sources of sequences were soil samples taken before treatment (ZT) and after incubation with LA, MA, and HA concentrations at 4°C for 4 weeks. The scale bar indicates 10 mutations per 100 sequence positions.

FIG. 4.

T-RFs of amplified nirK fragments (514 bp) with major differences in relative abundance according to soil treatment. Soil samples were taken before treatment (ZT) and after incubation with LA, MA, and HA concentrations at 4°C for 4 weeks. The key indicates the lengths of the T-RFs.

FIG. 5.

Neighbor-joining tree of partial nirK genes based on 102 amino acids. The consensus tree was reconstructed based on neighbor-joining, parsimony, FITCH, and maximum-likelihood (MOLPHY) analyses. Unresolved nodes are displayed as multifurcations and are indicated by dashed lines. Clones obtained from this experiment are shown in boldface, and the calculated sizes of T-RFs are indicated in parentheses. The scale bar represents 10 mutations per 100 sequence positions.