Despite a considerable knowledge of the significant role of termites in the global methane budget, very little is known about their contribution to the global nitrous oxide (N2O) budget. Release of N2O from termite (Cubitermes fungifaber) mounds was measured at a natural savanna site in the southwest of Burkina Faso from May to September 2006. Termite N2O emissions were around 20 μg N2O-N m–2 h–1 at the end of the dry season, and up to two orders of magnitude higher than N2O emissions from the surrounding termite-free soil after the onset of the rainy season. The average N2O emission rate from termite mounds during the observation period was 204 μg N2O-N m–2 h–1, and termite mounds contributed 3.0% to total N2O emissions from this savanna ecosystem. However, in other tropical terrestrial ecosystems with other termite species and/or higher termite density this share might be significantly higher.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 Tropical savannas cover huge parts of the Earth's land surface (11.5% or 16.1 × 106 km2 [Scholes and Hall, 1996]) between the equatorial rainforests and mid-latitude deserts. Although they contribute significantly to global C and N cycling and global greenhouse gas budgets due to their vast extension, savannas have received much less attention by research than other ecosystems, like forests, up to now.
 N2O has a global warming potential of 298, i.e. a 298 times higher warming effect than the same mass unit of CO2 over a 100 year period [Intergovernmental Panel on Climate Change (IPCC), 2007], thus, it is of high relevance for the Earth's greenhouse effect. Several studies indicate that microbiological processes in soils are the main source of atmospheric N2O [e.g., Donoso et al., 1993], amplified by anthropogenic activities, like, e.g., application of fertilizer, irrigation and ploughing, burning and clear-cutting. It is assumed that tropical savannas contribute around 17% to the global emission of N2O from non-agricultural soils, but this figure is associated with a great uncertainty [IPCC, 2001]. On the other hand data from several field studies suggest that soils can also consume N2O, albeit the strength of this sink has not yet been evaluated [e.g., Donoso et al., 1993; IPCC, 2001; Chapuis-Lardy et al., 2007].
 It is known that termites have a significant impact on carbon cycling, ecological functioning and trace gas emissions in tropical savannas by processing large quantities of soil or plant material [Zimmerman et al., 1982; Khalil et al., 1990; Rondón et al., 1993; Konaté et al., 2003], and it is assumed that they are one of the major biogenic CH4 sources with an estimated global emission of approx. 20 Tg CH4 yr−1 [IPCC, 2001]. However, the role of termites in N cycling in tropical ecosystems is much less understood. In an early study dealing with N2O emissions from termite nests, Khalil et al.  did not find significant differences between N2O fluxes from mounds of six different Australian termite species and the surrounding soil, indicating no significant acceleration or alteration of soil nitrogen cycling by termites. In contrast, Rondón et al.  have found four times higher nitrate concentrations in and approx. ten times higher NO emissions from termite mounds as compared to the surrounding soil in tropical South America as a result of accumulation of organic matter in the termite mounds. A similar result was reported by Le Roux et al.  and Serça et al.  for West African savanna, who found significantly higher net nitrate accumulation in and NO fluxes from termite mounds as compared to the surrounding soil. However, the latter three studies did not report any data on termite-related N2O fluxes.
 The objective of the present study was to quantify the N2O release from termite mounds at a nature reserve site in the South-Sudanian savanna of Burkina Faso, an almost unexplored region in the savanna belt of West Africa, with non-destructive manual static chamber measurements over a longer time period, and to assess its relevance for total ecosystem N2O exchange in this savanna ecosystem.
2. Material and Methods
 Release of N2O from mounds of the soil-feeding termite Cubitermes fungifaber Sjostedt was measured at a natural savanna site within a nature reserve in the southwest of Burkina Faso, approx. 280 km southwest of the capital Ouagadougou, close to a small village named Bontioli. Mean annual air temperature in this region of Burkina Faso is 29.5°C. Annual mean precipitation is 926 mm with a maximum from July to September. In contrast, during the very dry months from November to February the total amount of rainfall is only about 20 mm. The soil texture of the site is a sandy loam, with a pH of 4.9 ± 0.2, a C:N ratio of 11.08 ± 0.15, and a soil organic carbon (SOC) content of 0.56 ± 0.09%. The soil formed one homogeneous layer of 35–50 cm thickness above a water-impermeable laterite rock layer and showed no development of soil horizons. The natural vegetation during the rainy season from May to October is mainly dominated by the grass species Andropogon gayanus and Loudetiopsis kerstingii. Further details on site characteristics can be found in work by Brümmer et al. . A termite mound density of approx. 36 ha–1 was found at the Bontioli reserve site, estimated by counting the number of mounds in a larger area (approx. 10 ha) around the sampling plots.
 Soil moisture and soil temperature were measured manually with a handheld TDR probe (ML-2x, Delta-T, Cambridge, UK) and a cut-in thermometer (Novodirect, Kehl, Germany) at 5 cm depth in the soil in close vicinity to the termite mounds. Field data of volumetric water content (VWC) were converted into water-filled pore space (WFPS) using the following equation:
where 1.43 is the bulk density of the soil and 2.65 is the density of quartz.
 N2O flux measurements were conducted non-destructively from May to September 2006 on five termite mounds with the static chamber technique. During sampling, the termite mounds were completely covered by opaque plastic chambers (length × width × height = 0.8 m × 0.6 m × 0.4 m) that were mounted gas-tight on plastic frames driven approx. 0.1 m into the soil. For N2O analysis, four chamber air samples were taken within 30 min from each chamber with pressure-lock syringes. N2O was analyzed usually the next day, but no longer than 3 d after each sampling, in a laboratory 40 km away from the experimental site with a gas chromatograph (GC-14A, Shimadzu, Duisburg, Germany) equipped with an electron capture detector and a manual injection port. Reference gas (Air Liquide, Munich, Germany), containing 396 nL L–1 N2O in synthetic air, was injected once every four chamber air injections to calculate drift of the GC detector. Flux rates were determined via linear regression of the four sampling points for each chamber and by applying a temperature and pressure correction. The r2 value was usually higher than 0.95, in many cases higher than 0.99. The headspace volume of the chambers was calculated as the difference between chamber volume and mound volume. All mounds were mushroom-shaped and of approximately the same size with an average cap diameter of 0.5 m, and an average shaft diameter and height of 0.3 m. A constant mound volume of 21 L was assumed and subtracted from the gross chamber volume. In total, 104 N2O flux rates from all investigated termite mounds were calculated. The contribution of termite mounds to total N2O soil emissions (data from Brümmer et al. ) was calculated based on a mound density of 36 ha–1 and the area of the chambers enclosing each mound (0.48 m2).
 Although some rainfall events had already occurred in the second half of April prior to the first measurements, WFPS was still almost zero during the first two sampling dates around mid-May 2006 (Figure 1). After rainfall in the second half of May 2006, WFPS rose to values around 50%. Except for one sampling date in June 2006, when soil moisture had decreased again after some days of drought, WFPS rose gradually to almost 100% at the end of the observation period around mid-September 2006.
 N2O fluxes between 6 and 36 μg N2O-N m–2 h–1 were determined under dry soil conditions and high soil temperatures at the beginning of the measurements in May 2006, whereas high termite N2O emissions occurred at high soil water content and relatively low soil temperatures (Figure 1). Following the first major rain events and the first significant increase in soil water content, N2O emissions rose to 80–680 μg N2O-N m−2 h−1. Later in the rainy season, at constantly high soil water content, values ranged around 200 μg N2O-N m−2 h−1 (Figure 1). N2O fluxes showed a weak positive correlation with WFPS, and a weak negative correlation with soil temperature (data not shown).
 The mean value of termite N2O emissions throughout the observation period was 204 μg N2O-N m−2 h−1. Taking the area covered by termite mounds and the mean soil N2O fluxes during the same period (11.5 μg N2O-N m−2 h−1) [Brümmer et al., 2008] as a basis, termite-related N2O emissions amounted to 3.0% of total N2O emissions at this particular savanna site during the observation period.
 N2O emissions from termite mounds have been virtually uninvestigated in tropical ecosystems. In this paper we present for the first time a longer-term dataset of termite N2O emissions. To the best of our knowledge, the only paper on termite N2O field emissions published up to now is by Khalil et al. , who studied the influence of six Australian termite species on soil–atmosphere trace gas exchange and found no significant differences between N2O emissions from termite mounds and the surrounding soil. In contrast, the N2O emissions from termite mounds in this study (May–September 2006) were on average by a factor of 17.7 higher than N2O release from the surrounding soil during the same period (11.5 μg N2O-N m–2 h–1) [Brümmer et al., 2008]. Similarly, Rondón et al.  found, independent of the nest size, ten times higher NO emissions from termite (Velocitermes paucipilus and Nasutitermes spp.) mounds than from the surrounding soil in the Orinoco Basin in Venezuela. Also Le Roux et al.  and Serça et al.  found elevated NO fluxes from termite mounds up to 4.9 times higher than from termite-free soil in tropical savanna of Western Africa, coincident with elevated nitrate concentrations within the termite mounds as compared to the surrounding soil. The elevated termite NO and N2O fluxes as compared to the fluxes of the surrounding soil emphasizes the role of termites in mediating and accelerating soil N cycling. The reason for the fact that Khalil et al.  did not observe any significant enhancement of N2O emissions by termites remains unclear but might be related to differences in termite species behaviour and impact on soil nitrogen cycling.
 Termites can change soil properties and ameliorate soil fertility significantly, on the one hand—in the case of underground fungus-growing termites—by translocation of large amounts of organic material into the termite nests that leads to a concentration of nutrients in a relatively small volume and to hot spots of soil respiration [Konaté et al., 2003; Ohashi et al., 2007], and on the other hand—in the case of soil-feeding termites like Cubitermes fungifaber—by direct alteration of soil properties, such as pH, soil organic carbon and water content, during the passage of the soil through the termite gut [Donovan et al., 2001, 2004; Ndiaye et al., 2004]. The soil-ameliorating activity of termites creates suitable conditions for microbial degradation and mineralization of the collected organic material in combination with denitrification of mineral nitrogen. For example, Ndiaye et al.  found 100 times higher ammonium and 50 times higher nitrate contents in mounds of the soil-feeding termite Cubitermes niokoloensis in Southern Senegal as compared to the surrounding savanna soil, associated with a higher density of denitrifying bacteria and higher denitrification potential than in termite-free savanna soil, and creating the potential of elevated NO and N2O emissions from termite mounds.
 Like in our study for termite N2O emissions, Rondón et al.  found an immediate but short increase in termite NO emissions with the onset of the first rains as well as a moderate decrease later in the rainy season, indicating that oxygen [Ndiaye et al., 2004] and water availability play a key role in N transformation and N trace gas production in termite mounds. This initial large stimulation of N2O emissions at the onset of the rainy season with subsequent decrease to a lower level was also observed by Brümmer et al.  for termite-free soil at the same location, albeit less pronounced. The likely reason for this pattern is on the one hand an accumulation of easily available mineral nitrogen during the dry season in the soil and especially within the termite mounds, and on the other hand an improvement of soil physical conditions with higher soil moisture, leading to optimal conditions for denitrifiers as soon as the soil water content is high enough. As denitrification rates are higher at high water content than nitrification rates, the nitrate content will quickly decrease and lead to decreasing N2O emission rates.
 Even though the soil was very dry before the first rainfall of the up-coming wet season, N2O emissions from termite mounds were still higher than soil N2O fluxes [Brümmer et al., 2008], suggesting that beside soil-dwelling also termite-associated denitrifying bacteria might have contributed to N2O production in the mounds. A similar observation can be made for CH4 emissions from termites that have methanogenic microbes in their hindguts which are also still active when the surrounding soil is dry. However, a higher contribution (8.8%) of termite-related CH4 emissions to total ecosystem CH4 fluxes was found at the same location [Brümmer et al., 2009]—compared to 3.0% in the case of N2O in this study, as also relatively high CH4 emissions from termite mounds were measured when the surrounding soil was still very dry and served as a CH4 sink. This finding implies that either the contribution of termite-associated microorganisms to total N2O emissions from termite mounds is smaller than in the case of methane, or that N2O consumption in the mound plays a more important role than CH4 consumption. However, this could only be revealed by stable isotope analysis or micro-scale concentration profile measurements within the mounds.
 The present study demonstrates that N2O fluxes from termite mounds can be much higher than from the surrounding soil over a longer time period, although their contribution to the total areal N2O emission budget at the particular savanna site of the present study in Western Africa was relatively small. As the termite N2O emissions were only significant at relatively high soil water content, onset of the rainy season and amount of rainfall will determine the magnitude of termite-related N2O fluxes. In view of the huge range of termite species worldwide, the vast extension of tropical ecosystems, the differences in ecology and abundance between different termite species in different ecosystems, and the role of termites in creating “fertility islands”, intensified research is needed to constrain termite-related N2O fluxes on a global scale.
 Funding for this research work was provided by the Helmholtz Association of German Research Centres (Virtual Institute, VH-VI-001). The authors thank Georg Willibald for technical assistance and Gildas Houénagnon Boko for collecting climate data, Michael Dannenmann and Boris Matejek for scientific advice, and Konrad Vielhauer of the Center for Development Research (ZEF) and Dominik Schmengler of the Dreyer Foundation for generous infrastructural support in Burkina Faso. Special thanks go to Leama Paulin, Sawadogo Oussoumane, and Dramane Barry for their assistance during the field campaigns.