Discussions
This study found that the CO2 emission from the mounds of non-fungus growing termites were different between species and their surrounding soils during both dry and wet seasons in a tropical seasonal forest. The mean annual mound CO2 emission was 2.5 times higher than surrounding soils. In particular, CO2emissions from the mounds of G. sulphureus and M. crassuswere much higher than other termite species and greater than their surrounding soils in both seasons. The high CO2emissions were from the mounds from these species because their populations were found ubiquitously in the nest or mound with a thin wall, whereas the other species were found in some part of the nest. Also, most of the nest materials are mixed from the macerated wood and their faeces (Bignell and Eggleton, 2000). Moreover, G. sulphureus and M. crassus are a group of wood/litter feeders that these species may be consumed directly to organic carbon much more than the soil feeders. Although respiratory rate varied inversely with body weight, the substrates for all feeding groups have an effect on the variations in metabolic rate (Jeeva et al., 1999). In the same forest, these two species were found in high abundance and biomass, especiallyM. crassus was 46% and 36% of the termite abundance and biomass, respectively (Yamada et al., 2003). Therefore, the mounds ofG. sulphureus and M. crassus could potentially drive the high spatial variation in soil CO2 emission in the tropical seasonal forests.
As a result, the mean CO2 emission rate from the mound of G. sulphureus was higher than the mound of other species in this study. This emission rate was also higher than the mound from the other studies in both non – fungus grower (Brümmer et al., 2009; Jamali et al., 2013; De Gerenyu et al., 2015; Konemann and Kard, 2016) and fungus – grower termites (Konate´ et al., 2003; Nyamadzawo et al., 2012; Sawadogo et al., 2012; Ohashi et al., 2017) from the different topography, except for some rates of CO2 emissions from the subterranean nest of termites (e.g. Hospitalitermes hospitalis ) in a tropical rainforest (Ohashi et al., 2017) (Table 3).
In addition, the other soil organism was proposed as the CO2 point source such as ants and earthworm (Khalil et al., 1990; Stoyan et al., 2000; Risch et al., 2005; Ohashi et al., 2007). According to Ohashi et al. (2017) reported that the CO2 emission rate from the subterranean nest of antDinomyrmex gigas (45.5 µmol CO2m-2 s-1) was higher than the rate in this study. However, the CO2 emission rate from G. sulphureus mound in this study that was higher than soil CO2 emission which affected by the other ants (Sousa-Souto et al., 2012; Hasin et al., 2014; Fernandez-Bou et al., 2019) and earthworms (Šimek and Pižl, 2010) (Table 3). In term of earthworms, although there were no effect on the emission of CO2 from soil (Guo et al., 2019), earthworms have increased the carbon content and other nutrients (Lavelle et al., 2001) for enhancing soil respiration by microbial activities.
Meanwhile, the CO2 emissions from T. propinquus ,D. makhamensis , and T. comis mounds were both much lower than G. sulphureus and M. crassus and were significantly lower than surrounding soils in both seasons (except for T. comiswhich did not differ between surrounding soils in the wet season). However, these three species are categorized as soil-feeder, whileG. sulphureus and M. crassus are wood/litter-feeder (Wood, 1976; Collins, 1989; Bignell et al., 1997; Bignell and Eggleton, 2000). This was probably due to the respiration rate from termite populations of soil-feeders was lower than wood/litter-feeders. According to Yamada et al. (2005) Termite individual contributions to carbon mineralization by the respiration from wood/litter-feeders and soil-feeders were 2.8% and 1.2% of the annual aboveground litterfall in the DEF at SERS, respectively.
Mound structure differences between species may also contribute to variation in mound CO2 emissions. Nest construction material as the wood carton (a mixture of faeces and macerated wood) is used for the most wood/litter-feeding group, while many soil-feeders use faeces mixed with topsoil (Bignell and Eggleton, 2000). In addition, the soil feeder epigeal mounds may not contain the entire colony population above the mound because most individuals move below the soil for foraging and to avoid adverse environmental conditions (Fig 8). Moreover, the mound CO2 emission rates of T. propinquus , and D. makhamensis were similar to those reported by Song et al. (2013) who reported that the range of CO2emission from termite mounds (species: not given) was from 1.63 to 3.71 µmol CO2 m-2 s-1 and these species did not affect soil CO2 emission in the tropical rainforest. As the result of Song et al. (2013), mounds were either typical soil-feeding group mounds or similar to the fungus growing termites that build a dome-shaped mound with thick walls (20-40 cm thickness) and several branching underground passages (Inoue et al., 2001).
Previous studies have reported that the mound-builder and subterranean nest of termites emitted significantly higher CO2 than typical soils (e.g. Konate et al., 2003; Brümmer et al., 2009; Risch et al., 2012; De Gerenyu et al., 2015; Ohashi et al., 2017). Not only CO2 emission from termite individuals, but their nest also contributed to soil respiration. According to Hu et al. (2017) reported that CO2 emission from termite nest soil was higher than typical soil, due to termite nest soils had higher dissolved organic carbon concentration and C/N ratio, accelerating for microbial activities. In a tropical savanna, the fungus-growing termite was emitted 10-19 µmol CO2 m-2s-1 from the mound which compared to 5-10 µmol CO2 m-2 s-1 from surrounding soils (Konate et al., 2003). In a tropical rain forest, the mean of CO2 emission from termite mound was reached up to 27.9 µmol CO2 m-2s-1 that much higher than the surrounding soil (3.96 µmol CO2 m-2 s-1) (Ohashi et al., 2017). An earlier study (Ohashi et al., 2007) indicated that the rate of soil respiration was sometimes extremely high (>10 µmol CO2 m-2s-1) contributing 10% to the total soil respiration in the tropical rainforest. De Gerenyu et al. (2015) showed that the mound CO2 emission of Odontotermes termites (fungus grower) and G. sulphureus (non-fungus grower) was 2.0-2.5 times higher than the background soil during both dry and wet seasons. Also, those termite mounds contributed up to 10% of the total soil respiration in a tropical monsoon forest of southern Vietnam, showing the maximum rates of the mound CO2 emissions were 17 and 20 µmol CO2 m-2 s-1from Odontotermes and G. sulphureus , respectively. On the other hand, mound CO2 emission rate was 2.5 times lower than the surrounding soil that was affected by termite respiration through their underground passages in the surrounding soil (Boonriam et al., 2021b).
The relationship between soil CO2 emission and its environmental factors was also observed from only soil around each mound, while soil temperature and soil moisture content were not measured from the mounds to avoid disturbing termite activity at the nest. In this study, soil CO2 emission rate was 3.4 times higher in the wet season than the dry season. Soil CO2 emissions showed a consistent relationship with soil moisture content. In addition, soil CO2 emission rate was started to decrease when soil moisture content peaked more than 18% (Adachi et al., 2009) and 21% (Hasin et al., 2014) in diurnal variation of Thai tropical forest. A previous study reported that the rate started to drop at 27 °C soil temperature and 21% soil moisture content in the same topographical forest (Boonriam et al., 2021a). Although precipitation influences soil microbial activities accelerating soil respiration, high precipitation could inhibit CO2emission by a barrier of high moisture content in the soil (Sotta et al., 2004; Wood et al., 2013). This inconstancy within the timing and magnitude of precipitation occasions can influence soil respiration.