Meta-analysis of environmental impacts on nitrous oxide release in response to N amendment

MINI REVIEW ARTICLE published: 31 July 2012 doi: 10.3389/fmicb.2012.00272 Meta-analysis of environmental impacts on nitrous oxide release in response to N amendment Emma L. Aronson 1 * and Steven D. Allison 1,2 1 2 Department of Ecology and Evolutionary Biology, University of California Irvine, Irvine, CA, USA Department of Earth System Science, University of California Irvine, Irvine, CA, USA Edited by: Graeme W. Nicol, University of Aberdeen, UK Reviewed by: Hongchen Jiang, China University of Geosciences, China Nick Bouskill, Lawrence Berkeley National Laboratory, USA *Correspondence: Emma L. Aronson, Department of Ecology and Evolutionary Biology, University of California Irvine, 321 Steinhaus, Irvine, CA 92627, USA. e-mail: Atmospheric nitrous oxide (N2 O) accounts for approximately 5% of the global greenhouse effect and destroys stratospheric ozone. Soils are the most important source of N2 O, which is produced during nitrification and denitrification. To assess the impact of environmental variables and ecosystems on N2 O flux, we performed a meta-analysis comparing N2 O flux in N amended and matched control plots in non-agricultural soils. We found that N2 O release increased with N amendment in the short term. Although there were few studies in shrubland, this ecosystem showed the greatest response. The N2 O response to N amendment was greater in year-round studies and in studies with more measurements, but lower in longer studies. The N2 O response was greater at higher latitudes and precipitation rates. We also observed an unexpected 55% decline in the N2 O response to N amendment over the 23 years covered by the studies. This pattern may reflect a suppression of the N2 O response from long-term N deposition accumulation, particularly in temperate regions. Although short term increases in reactive N entering natural systems may cause positive feedbacks to the release of N2 O, this effect may diminish over time in locations with high rates of N deposition. Keywords: denitrification, meta-analysis, N amendment, N deposition, N2 O, nitrification, nitrous oxide INTRODUCTION Human activities have doubled the amount of nitrogen (N) entering soils, primarily through fossil fuel combustion and the application of N fertilizer to agricultural land (Schlesinger, 2009). Increased N inputs and cycling have led to increased N deposition on natural ecosystems (Galloway et al., 2008), stimulating plant growth, and altering soil microbial responses (Lu et al., 2011). Increased soil N availability can also stimulate losses of trace gases, such as nitrous oxide (N2 O), which accounts for approximately 5% of global greenhouse gas forcing. Oxides of N derived from N2 O can also react with the Earth’s protective stratospheric ozone layer and expose the surface to harmful UV rays from the sun (Intergovernmental Panel on Climate Change, 2007). Therefore, predictions of climate feedbacks involving N deposition must account for N2 O release (Zaehle et al., 2011). The main sources of atmospheric N2 O are the reduction of nitrate through denitrification, and the oxidation of ammonia to nitrite, and further to nitrate during nitrification (Pathak, 1999). Soil type, oxygen status, moisture, temperature, carbon (C), and N status, as well as N amendment to the soil, can influence both of these processes and subsequent N2 O release (Pathak, 1999; Burgin and Groffman, 2012). In agricultural systems, fertilizer type, fertilizer amount, and crop type are important factors determining the rate of N2 O release (Bouwman et al., 2002). In non-agricultural systems, most reactive N enters the system through atmospheric deposition. For example, elevated N deposition can increase N2 O release by up to fivefold in forest soils (Butterbach-Bahl et al., 1998). Most N2 O research has focused on agricultural soils, which generally release more N2 O than non-agricultural soils (Reay et al., www.frontiersin.org 2012). However, non-agricultural soil sources are increasing in importance as perturbation of the N cycle continues (Gruber and Galloway, 2008). In order to address the response of nonagricultural systems to greater N loads, we collected published comparisons of soil N2 O release in fertilized and unfertilized soils in upland ecosystems. This analysis contrasts with previous metaanalyses that have focused on the fertilization response of plant N (Lu et al., 2011), and C pools and fluxes (Treseder, 2008; Liu and Greaver, 2009; Aronson and Helliker, 2010; Janssens et al., 2010). We focus on non-wetland and non-agricultural soils, since far less is known about the response of these soils to N amendments. We hypothesized that reactive N amendment would significantly increase N2 O release across studies (e.g., Barnard et al., 2005). In addition, we hypothesized that the response of N2 O release to N amendment would increase with level of N amendment and daily precipitation. The precipitation hypothesis was based on data showing increased N2 O release from denitrification following precipitation and irrigation (Freney et al., 1978; Duxbury et al., 1982; Pathak, 1999). We therefore hypothesized that more arid systems, such as deserts, savannas, grasslands, and shrublands would show the lowest N2 O response to N amendment, whereas forests, particularly in the humid tropics, would show the greatest response. MATERIALS AND METHODS DATA SOURCES Nitrous oxide production data were extracted from published studies containing matched N amendment and control treatments (Appendix). The studies included in this meta-analysis were performed in non-agricultural, non-wetland ecosystems. All studies July 2012 | Volume 3 | Article 272 | 1 Aronson and Allison Environment regulates N2 O fertilization response performed field flux measurements using chambers, although the chamber type and N2 O quantification methods varied. All the original data are extracted from text, figures, and tables in the published papers, as in Aronson and Helliker (2010). The studies were located using ISI Web of Knowledge with the search terms: “nitrous oxide” and “release” or “flux” and “fertilization” or “nitrogen” and “amendment,” “addition” or “deposition.” The resulting database consisted of 99 entries from 33 papers, each containing a single added N versus control comparison (Supplementary Material). There were multiple comparisons from many studies due to different levels of N treatment or multi-factorial designs. When multiple N amendment levels were used, the average flux from un-amended plots was used for the control in all comparisons. In multi-factorial experiments, comparisons were made between N treatment and N control plots that had received the same set of crossed treatments. Ancillary information from each data source included: latitude, average annual temperature, average daily precipitation, start year, study duration, number of measurements, ecosystem type, season(s) studied, form(s) and amount of N used, and gas flu (...truncated)