[This corrects the article on p. 1638 in vol. 7, PMID: 27807431.].

Savannas cover at least 13% of the global terrestrial surface and are often nutrient limited, especially by nitrogen. To gain a better understanding of their microbial diversity and the microbial nitrogen cycling in savanna soils, soil samples were collected along a granitic and a basaltic catena in Kruger National Park (South Africa) to characterize their bacterial and archaeal composition and the genetic potential for nitrification. Although the basaltic soils were on average 5 times more nutrient rich than the granitic soils, all investigated savanna soil samples showed typically low nutrient availabilities, i.e., up to 38 times lower soil N or C contents than temperate grasslands. Illumina MiSeq amplicon sequencing revealed a unique soil bacterial community dominated by Actinobacteria (20-66%), Chloroflexi (9-29%), and Firmicutes (7-42%) and an increase in the relative abundance of Actinobacteria with increasing soil nutrient content. The archaeal community reached up to 14% of the total soil microbial community and was dominated by the thaumarchaeal Soil Crenarchaeotic Group (43-99.8%), with a high fraction of sequences related to the ammonia-oxidizing genus Nitrosopshaera sp. Quantitative PCR targeting amoA genes encoding the alpha subunit of ammonia monooxygenase also revealed a high genetic potential for ammonia oxidation dominated by archaea (~5 × 10⁷ archaeal amoA gene copies g^-1 soil vs. mostly < 7 × 10⁴ bacterial amoA gene copies g^-1 soil). Abundances of archaeal 16S rRNA and amoA genes were positively correlated with soil nitrate, N and C contents. Nitrospira sp. was detected as the most abundant group of nitrite oxidizing bacteria. The specific geochemical conditions and particle transport dynamics at the granitic catena were found to affect soil microbial communities through clay and nutrient relocation along the hill slope, causing a shift to different, less diverse bacterial and archaeal communities at the footslope. Overall, our results suggest a strong effect of the savanna soils' nutrient scarcity on all microbial communities, resulting in a distinct community structure that differs markedly from nutrient-rich, temperate grasslands, along with a high relevance of archaeal ammonia oxidation in savanna soils.