Rapid nitrification involving comammox and canonical Nitrospira at extreme pH in saline-alkaline lakes

Abstract

Nitrite-oxidizing bacteria (NOB) catalyse the second nitrification step and are the main biological source of nitrate. The most diverse and widespread NOB genus is Nitrospira, which also contains complete ammonia oxidizers (comammox) that oxidize ammonia to nitrate. To date, little is known about the occurrence and biology of comammox and canonical nitrite oxidizing Nitrospira in extremely alkaline environments. Here, we studied the seasonal distribution and diversity, and the effect of short-term pH changes on comammox and canonical Nitrospira in sediments of two saline, highly alkaline lakes. We identified diverse canonical and comammox Nitrospira clade A-like phylotypes as the only detectable NOB during more than a year, suggesting their major importance for nitrification in these habitats. Gross nitrification rates measured in microcosm incubations were highest at pH 10 and considerably faster than reported for other natural, aquatic environments. Nitrification could be attributed to canonical and comammox Nitrospira and to Nitrososphaerales ammonia-oxidizing archaea. Furthermore, our data suggested that comammox Nitrospira contributed to ammonia oxidation at an extremely alkaline pH of 11. These results identify saline, highly alkaline lake sediments as environments of uniquely strong nitrification with novel comammox Nitrospira as key microbial players.

FIGURE 1

Environmental parameters measured for lake water (A, C, E) and sediment pore water (B, D, F) in monthly samples taken from April 2014 to June 2015 from the two studied saline‐alkaline lakes HS (lake Herrensee) and US (lake Unterer Stinkersee).

FIGURE 2

Phylogenetic maximum likelihood analysis of canonical and comammox Nitrospira in lakes HS and US. The trees show the affiliations of Nitrospira nxrB (A) sequences and Nitrospira AmoA (B) sequences, which were retrieved from monthly samples and from sediment microcosms, to reference sequences. Sequences obtained from in this study are printed in bold. The nxrB gene sequences of cultured cultured Brocadia, Jettenia, Kuenenia, Scalindula, and Nitrospina species and the AmoA gene sequences of cultured Nitrosomonas, Nitrososphaera, and Nitrosocosmicus species were used as outgroups for the Nitrospira nxrB and Nitrospira AmoA tree, respectively. The phylogenetic calculations included model prediction by ModelFinder (Kalyaanamoorthy et al., 2017), which identified the best‐fit models to be GTR + F + I + G4 and LG + F + G4 for the Nitrospira nxrB and Nitrospira AmoA trees, respectively. The Nitrospira AmoA sequences obtained from sediment of lake HS in an earlier study (Pjevac et al., 2017) are marked with an asterisk (B). Sequences of nxrB (A) shown in red or blue are affiliated with comammox Nitrospira clade A or B (as identified by Amo gene‐containing, high‐quality MAGs or genomes from cultured strains), respectively. Black and grey stars indicate isolates and enrichment cultures, respectively. Circles at nodes indicate statistical support of branches (1000 bootstrap iterations). The scale bar indicates 30% (A) and 40% (B) estimated sequence divergence.

FIGURE 3

Nitrification activity in pH‐controlled microcosm incubations. (A) Concentrations of nitrite, nitrate, and ammonium in the microcosms during the incubations. The data are split according to the pH treatments. Data points represent means of replicate microcosms (n = 4). (B) Gross nitrifications rates determined during the last 24 h of the microcosm incubations. The bars represent means of replicate microcosms (n = 4) with standard errors. Lower‐ and upper‐case letters indicate significant differences between microcosms for the same lake HS and US, respectively, while asterisks show significant differences between microcosms for different lakes at a given pH condition. HS, lake Herrensee; US, lake Unterer Stinkersee. Error bars (s.e.m. in both panels) are not visible if smaller than symbols.