Phosphorus but not nitrogen addition significantly changes diazotroph diversity and community composition in typical karst grassland soil.

• N and P addition affect the dry-cold and wet-hot seasonal dynamics of diazotroph communities. • Relative abundance of Rhizobiales and Rhodospirillales was high in July. • Relative abundance of Nostocales and Burkholderiales was high in December. • N addition decreased diazotroph abundance; P addition increased diazotroph diversity. • P availability affected diazotroph diversity and community composition the most. Diazotrophs have a vital role in nitrogen (N) supply to ecosystems. Yet, response of diazotroph community profiles and N 2 -fixation to N deposition and phosphorus (P) addition remains poorly understood. Two-year nutrient addition experiments (control, N, P, and a combination of N plus P) were conducted in a karst grassland. The relative abundances of diazotroph orders Rhizobiales and Rhodospirillales were higher in July, while those of Nostocales and Burkholderiales were higher in December. Diazotroph abundances and N 2 -fixation activity in July were significantly higher compared with December under control, while an opposite pattern was observed for diazotroph richness and Shannon diversity index. N, P, and N plus P plots showed no significant changes between July and December in terms of diazotroph abundance and diversity. Diazotroph abundances and N 2 -fixation activity in N and NP plots were significantly reduced compared with control plots in July. P addition plots in July showed a significant increase in diazotroph diversity and stronger interaction patterns among diazotroph taxa compared with control plots. Diazotroph diversity and community compositions were primarily affected by P availability. Overall, our findings suggested diazotroph diversity and community compositions were more sensitive to P addition than to N addition, and P availability played an important role in regulating N 2 fixation by increasing diazotroph diversity in karst ecosystems. N and P addition may alter the dry-cold and wet-hot seasonal dynamics of diazotroph communities, while sampling months also affect diazotroph community changes among treatments (no difference between treatments in December). [ABSTRACT FROM AUTHOR]