Although it is widely accepted that global nitrogen enrichment leads to biodiversity loss and reduced ecosystem resilience, there is debate about whether or not N addition experiments, characterized by infrequent large additions of N, reflect ecosystem responses to actual patterns of N deposition (high frequency, smaller individual inputs of N). Recent studies indicate that plant species richness decreases more rapidly with a low frequency of high N inputs, suggesting that such fertilization studies over-estimate the effects of N deposition on plant species loss. However, our understanding of how belowground net primary productivity (BNPP), NPP and biomass allocation may respond to N input patterns is lacking. We manipulated the amount (0-50 g N m-2 year-1) of NH4NO3 inputs and the frequency (twice vs. monthly additions per year) for six consecutive years in a temperate grassland in northern China. We measured aboveground net primary production (ANPP) and BNPP annually from 2012 to 2014 to test two hypotheses: (i) ANPP and BNPP responses to increasing amounts of N addition would vary inversely, leading to reduced belowground allocation; and (ii) BNPP and biomass allocation would respond differently to frequent small N additions (12 times yr-1) vs. infrequent large N additions (2 times yr-1).
Results/Conclusions
We found that increasing N, regardless of 2 or 12 additions yr-1, significantly increased ANPP by 36%, and suppressed BNPP by 29%, resulting in 7% NPP increase. Further, the response of BNPP, ANPP and NPP to N saturated above 30 g N m-2 yr-1. However, both BNPP and biomass allocation responded differently to the frequency of N addition. BNPP and the fraction of BNPP was higher when N addition was frequent and N amounts were low. BNPP was also more sensitive to N than ANPP, especially at low amounts of N addition. This suggests that past N fertilization experiments may underestimate root response to N deposition. Our findings provide new insight into how plants regulate carbon allocation to different organs with increasing N amounts and changing N frequencies. These root response patterns should be incorporated into earth system models to improve the predictive power of C dynamics of dryland ecosystems in the face of global atmospheric N deposition.