Article pubs.acs.org/est
Nitrous Oxide and Methane Fluxes Following Ammonium Sulfate and Vinasse Application on Sugar Cane Soil Debora da S. Paredes,† Bruno J. R. Alves,*,‡ Marco A. dos Santos,† Denizart Bolonhezi,§ Selenobaldo A. C. Sant’Anna,‡ Segundo Urquiaga,‡ Magda A. Lima,∥ and Robert M. Boddey‡ †
COPPE, Centro de Tecnologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro 21941-450, Rio de Janeiro Brazil Embrapa Agrobiologia, Seropédica 23891-000, Rio de Janeiro Brazil § Instituto Agronômico (IAC), APTA Regional Centro-Leste, Ribeirão Preto 14032-800, São Paulo Brazil ∥ Embrapa Meio-Ambiente, Jaguariúna 13820-000, São Paulo Brazil ‡
ABSTRACT: This study aimed to quantify nitrous oxide (N2O) and methane (CH4) emission/sink response from sugar cane soil treated with fertilizer nitrogen (N) and vinasse applied separately or in sequence, the latter being investigated with regard to the time interval between applications for a possible effect on emissions. The study was carried out in a traditional area of unburned sugar cane in São Paulo state, Brazil. Two levels of N fertilization (0 and 100 kg N ha−1) with no added vinasse and combined with vinasse additions at different times (100 m−3 ha−1 at 3 and 15 days after N fertilization) were evaluated. Methane and N2O fluxes were monitored for 211 days. On average, the soil was a sink for CH4, which was not affected by the treatments. Emissions of N2O were induced by N fertilizer and vinasse applications. For ammonium sulfate, 0.6% of the added N was emitted as N2O, while for vinasse, this ranged from 1.0 to 2.2%. Changes in N2O fluxes were detected the day after application of vinasse on the N fertilized areas, but although the emission factor (EF) was 34% greater, the EF was not significantly different from fertilizer N alone. Nevertheless, we recommend to not apply vinasse after N fertilization to avoid boosting N2O emissions.
■
INTRODUCTION According to the Bulletin of the World Meteorological Organization (WMO),1 the concentration of greenhouse gases (GHG) in the atmosphere was at unprecedented levels in 2013. In consideration of such a scenario, the search for fuels with low GHG emissions has been on the scientific agenda, for which ethanol from sugar cane stands out due to its large potential for GHG mitigation.2 The International Panel on Climate Change (IPCC) guidelines3 have been used in the calculation of the potential mitigation promoted by biofuels, for which global GHG emission factors for N2O and CH4 emissions have been utilized owing to the scarcity of data from cane fields. Brazil is the largest producer of sugar cane in the world, with an area of approximately 9.8 million hectares and a production of 690 million tons of stems in the 2013/2014 harvest.4 This crop is semiperennial with a first crop cycle of 12 or 18 months followed by 4−6 ratoon crops (harvested every 12 months) until renewal, which means ∼80% of the crop is in a ratoon phase every year. In the 2013/2014 harvest, ethanol production was 27.96 billion liters, resulting in the production of approximately 363 billion liters of diluted vinasse,5 also known as stillage, which is the liquid byproduct generated by distilleries when ethanol is separated by distillation. On average, for each liter of ethanol, 13 L of vinasse are produced,6 and this © 2015 American Chemical Society
is spread onto the cropped area as source of K, for emergency irrigation, or both. The high C (8000−22 000 g dm−3) and N (0.10−0.71 g dm−3) contents6 of vinasse make it a potential GHG source in sugar cane cropping areas, which has been demonstrated for N2O in Brazilian soils,7−9 but for CH4, the emissions were reported to be negative.7 Fertilizer N is applied to the crop at rates varying from 60 to 120 kg N ha−1, the highest being for ratoon crops, usually in bands beside the plant row. It is a known source of N2O, but the establishment of emission factors are still required for the crop under Brazilian conditions to avoid the use of average data from other countries, many of them from contrasting climatic and crop management conditions (e.g., Lisboa et al.10). In addition, a common practice in sugar cane production, for practical and economic reasons, is to accomplish all operations for the next ratoon cycle as soon as possible after stem harvesting. Hence, the applications of vinasse and fertilizer N are usually made in sequence but not necessarily in this order, and the time interval between both is variable but usually within Received: Revised: Accepted: Published: 11209
March 25, 2015 July 31, 2015 August 21, 2015 August 21, 2015 DOI: 10.1021/acs.est.5b01504 Environ. Sci. Technol. 2015, 49, 11209−11217
Article
Environmental Science & Technology Table 1. Soil Chemical Characteristicsa soil layers (cm)
pH
Al3+ (cmolc dm−3)
Ca2+ (cmolc dm−3)
Mg2+ (cmolc dm−3)
K+ (mg dm−3)
P (mg dm−3)
N (%)
C (%)
0−10 10−20
5.43 5.38
0.03 0.03
3.20 2.62
1.17 0.89
88 41
6 7
0.15 0.13
1.50 1.41
a Soil pH was determined in water (2:1); Al3+ was determined by alkali titration and Ca2+ and Mg2+ were determined by atomic absorption after KCl 1 M extraction; K+ was determined by flame photometry, and P was determined by spectrometry after Mehlich 1 extraction; and N and C were determined by dry combustion in a CN analyzer.
Table 2. Chemical Characterization of the Vinasse Applied at 3 and at 15 Days after N Fertilizationa vinasse
N (g dm−3)
K (mg dm−3)
organic C (mg dm−3)
dissolved CH4 (mg dm−3)
dissolved N2O (μg dm−3)
applied at 3 days applied at 15 days
0.17 ± 0.09 0.19 ± 0.09
1240 ± 16 1210 ± 25
1757 ± 102 2137 ± 154
0.91 ± 0.06 3.24 ± 0.88
0.23 ± 0.02 0.23 ± 0.02
Nitrogen was determined by Kjeldhal digestion; K was determined by flame photometry; organic C was determined by a total organic carbon analyzer model TOC 5000 (Shimadzu, Tokyo, Japan); and dissolved CH4 and N2O were determined by gas chromatography after the headspace sampling technique. a
a few days, up to a week or two. Paredes et al.9 showed evidence that the sequence of N fertilizer application and vinasse changed the magnitude of N2O emissions. More specifically, the application of fertilizer N followed by vinasse brought about increased emissions, but not when the opposite was done. Apart from confirmation, it is desirable that the impact on these emissions of the time interval between applications should also be studied. Possible effects on CH4 emissions from vinasse should also be evaluated because an increase in ammonium concentration in soil solution is considered an inhibitor of methanotrophic activity in aerobic soils.11 Hence, the present study was aimed to quantify N2O and CH4 emission response from a sugar cane crop treated with fertilizer N and vinasse applied separately or in sequence, the latter being investigated with regard to the time interval between applications for a possible interaction between the two N sources.
■
B followed by application of vinasse after 3 days; (D) application of vinasse 3 days after the beginning of the experiment; (E) the same as in treatment B followed by application of vinasse after 15 days; and (F) vinasse application 15 days after the beginning of the experiment Plots of approximately 4 m2 were delimited in the sugar cane area following the adopted experimental design. On November 29, ammonium sulfate (100 kg N ha−1) was applied in treatments B, C, and E, while the other treatments remained without N addition. On December 2, 3 days after the application of fertilizer, vinasse was applied at the equivalent to 100 m3 ha−1 in treatments C and D and, on December 15, vinasse was applied at the same rate in treatments E and F, and treatment A was the only remaining control. The treatments without vinasse received the same volume of water. Owing to some variation in the composition of vinasse being produced by the distillery over time, the amount of N applied in the form of vinasse was approximately 18.2 kg N ha−1, 17.1 (±0.9) kg N ha−1 for the first application and 19.2 (±1.0) kg N ha−1 for the second application (Table 2). The fertilizer N was side-dressed approximately 10 cm from the sugar cane row, while vinasse was spread over the whole plot according to standard practice. Soil CH4 and N2O fluxes were monitored using vented non flow-through closed chambers similar to that described by Carmo et al.,7 placing one chamber per plot close to plant row. Briefly, it was a base-top chamber both parts made of polyvinyl chloride (PVC), 5 mm thick, with 30 cm diameter by 20 cm height. The base was inserted into the soil to a depth of 7 cm and left in place until the end of measurements. The lid was put in place only during the period of soil gas flux measurement, which was carried out always in the morning between 8:00 and 10:00 h.12 Known amounts of fertilizer N and vinasse were placed into the chamber in the same way they were applied in plots. The amount of ammonium sulfate in each chamber was equivalent to 64 g N m−2 as the chamber area enclosed the fertilizer band, but not the whole inter-row space. The control plots (no fertilizer, no vinasse) were also used to represent the area not fertilized with N. Daily gas sampling was performed from November 28 to December 23 and from December 26 to December 30, 2013. From January 2 to June 27, 2014, sampling frequency was reduced to twice a week. Samples were collected at 0, 15, and 30 min during incubation using 60 mL polypropylene syringes. Approximately,
MATERIALS AND METHODS
The experiment was conducted in an area of the Experimental Station of the Agronomic Institute of Campinas (IAC) Regional East Central Pole (APTA), in Ribeirão Preto, São Paulo State (24°10′40″ S and 47°48′36″ W), altitude 546 m.a.s.l. The climate is tropical, with a rainy summer (December to March) presenting an average temperature of 24 °C. The average annual rainfall is 1423 mm, but rainfall is very much reduced during winter, and the mean temperature drops to 18 °C. The soil class is Rhodic Hapludox and its chemical characteristics are described in Table 1. The experiment was installed on November 28, 2013, in an area of sugar cane established between 1980 and 1985, which has since been harvested without burning, with total trash conservation. The sugar cane variety IAC 95-5000 was planted with 1.5 m row space and was in the second ratoon stage. The amount of straw present on site at the time of experiment setup (2 weeks after harvest) was 17.8 Mg ha−1. The experimental design was a factorial 2 × 3 in randomized blocks with five replications. The two factors were N fertilization (with and without) and vinasse application (no vinasse and vinasse applied at 3 and 15 days after N fertilization). The treatments were (A) control (without fertilizer or vinasse); (B) nitrogen fertilizer (ammonium sulfate) applied at the beginning of the experiment; (C) the same as in treatment 11210
DOI: 10.1021/acs.est.5b01504 Environ. Sci. Technol. 2015, 49, 11209−11217
Article
Environmental Science & Technology
Figure 1. (A) Daily mean air temperature and rainfall and (B) GHG fluxes from the control treatment during the experimental period. Bars represent the standard error of the mean.
40 mL of air were withdrawn from each chamber after flushing the dead volume (inner-chamber access tube, three-way valve and syringe) with 20 mL of air, which was discarded before the final air sampling. From the 40 mL of air sampled, 25 mL were transferred into 20 mL chromatography vials evacuated to