COS 57-3
N2O emissions from grain cropping systems: Meta-analysis of management impacts, and cross-scale perspectives

Wednesday, August 12, 2015: 8:40 AM
301, Baltimore Convention Center
Zhen Han, Department of Ecology and Evolutionary Biology, Cornell University, Ithaca, NY
M. Todd Walter, Cornell University, Ithaca
Laurie E. Drinkwater, Horticulture, Cornell University, Ithaca, NY
Background/Question/Methods

There are a growing number of scientific studies seeking improved management practices to reduce N2O emissions from agricultural systems. A comprehensive analysis is needed to systematically evaluate current research proceedings, identify knowledge gaps and indicate promising new research directions. A meta-analysis on 524 pairwise comparisons (from 93 papers) was conducted to assess the efficacy of a wide range of management strategies and quantify the tradeoff between N2O mitigation and yield outcomes. Quantitative synthesis was conducted to evaluate whether or not predictions of N2O production could be improved using data characterizing the (de)nitrifier communities. A conceptual framework for understanding N2O emissions in agroecosystems from cross-scale perspectives was provided.

Results/Conclusions

Past studies have disproportionally focused on adjusting the precision of commercial nitrogen (N) fertilizer inputs, while there were limited numbers of studies on ecologically-based strategies such as the use of organic N sources and cover crops. Comparisons aimed at improving fertilizer uptake efficiency dominated the dataset (38.3%), followed by reduced tillage (30.8%) and manure (20.0%) studies, while studies of more complex, ecologically-based approaches such as diversified rotations only accounted for 10.8% of the dataset.

Reducing fertilizer rates had the most significant N2O reduction effect (by 57.0 % on average). The effect of manure compared to inorganic N fertilizer varied based on the amount of N input and soil texture, but not manure form.  The tradeoff between N2O mitigation and yield improvement was examined. Only 12.6% of data pairs had reduced N2O without yield loss.  49.2% data pairs decreased N2O with less than 10% yield loss, which indicated great potential for achieving environmental-economic co-benefits.

We also provided quantitative evidence that functional gene abundances (FGA) can be used together with environmental variables to improve the prediction of nitrification and denitrification processes. Forty five of the 56 studies found a significant correlation between FGA and potential rates of (de)nitrification.  Twelve of the 39 studies that linked FGA to in situ rates of (de)nitrification or N2O fluxes found significant correlations. Only 18 studies could be used to address the question of whether or not the added information from FGA improved the prediction of (de)nitrification potential and in situ rates in addition to environment variables, and nine studies found improvement. Our synthesis indicates that cross-scale and multi-disciplinary collaboration will be necessary in order to improve predictions of environmental N2O emissions.