Peatlands are globally important ecosystems covering an area of 463.2 Mha (Leifeld and Menichetti, 2018). However, tropical peatlands, which make up to 11% of the global peatland area (Leifeld and Menichetti, 2018), are the most efficient terrestrial ecosystems for carbon (C) sequestration (Yu et al., 2010). As a long-term global soil organic C stock, they sequester up to 88.6 Gt of C (Page et al., 2011; Gumbricht et al., 2017). Although peatlands are a significant source of atmospheric methane (CH₄), vastly altered by anthropogenic pressure (Petrescu et al., 2015), peatlands are the key ecosystems in the climate change mitigation strategy.

The Peruvian Amazon is the home to one of the largest areas of tropical peatlands globally (Page et al., 2011; Draper et al., 2014; Gumbricht et al., 2017). The region is characterized as aguaje palm (Mauritia flexuosa) dominant swamp forest peatland (Draper et al., 2014; Bhomia et al., 2019). The number of studies on the C pools of the Peruvian Amazon peatlands has increased in recent years (Lähteenoja et al., 2009a,b; Draper et al., 2014).

Peatlands influence climate through the uptake and release of greenhouse gasses (GHG) (van Lent et al., 2019). In addition to carbon dioxide (CO₂) and CH₄, they may also emit or uptake nitrous oxide (N₂O). All these gasses are claimed to be the main contributors to global warming.

Carbon dioxide emission from the peatland depends on the balance between plant C fixation through photosynthesis and ecosystem respiration. While CH₄ is produced in the anaerobic soil of peatlands by methanogens, CH₄ is oxidized in the aerobic zones by methanotrophs, and fixed carbon is released back to the atmosphere as CO₂ (Le Mer and Roger, 2001). Meanwhile, N₂O production and uptake is a microbial-mediated enzyme-driven process via nitrification (the aerobic oxidation of ammonia to nitrate) and denitrification (the reduction of nitrate to N₂O and ultimately to inert N₂ under anaerobic conditions) (Syakila and Kroeze, 2011). In tropical peatlands, other pathways can also be significant sources of N₂O. In nitrate-limited conditions, these processes can be DNRA (dissimilatory nitrate reduction to ammonium) (Espenberg et al., 2018) or nitrifier denitrification (nitrification in which ammonia is oxidized to nitrite followed by the reduction of nitrite to nitric oxide, N₂O and N₂) (Hergoualc’h et al., 2020).

A gas pathway from the soil through herbaceous plants, adapted to high soil water content, to the atmosphere is well-studied (Scholander et al., 1955; Henneberg et al., 2012). Plants exposed to permanent or episodic flooding cope with soil anoxia through morphological adaptations such as the formation of adventitious water roots, hypertrophied lenticels and enlarged aerenchyma (Shimamura et al., 2010; Pangala et al., 2013). Trees can take up the soil-produced GHG by their roots and release it by their aboveground tree surfaces into the atmosphere; however, trees’ role in the production and consumption of these has not been thoroughly studied (Machacova et al., 2016; Maier et al., 2018; Barba et al., 2019). Plant stems are particularly efficient conduits of CH₄ from the soil because the pathway bypasses highly active methanotrophic bacteria populations at the oxic-anoxic interface under the soil surface (Brune et al., 2000). Palms have adventitious root systems spread broadly throughout the shallow soil to anchor the plant (Steffens and Rasmussen, 2016). The root systems enhance gas exchange between the atmosphere and the soil (Megonigal and Day, 1992), in particular, entry of oxygen to the root zone and exit of soil-generated GHG through stems and aerial roots (pneumatophores) into the atmosphere (van Lent et al., 2019). Several studies have demonstrated that temperate trees adapted to wet soil also facilitate egress of soil-produced CH₄ (Gauci et al., 2010; Rice et al., 2010; Machacova et al., 2013; Pangala et al., 2013) via gas transport through aerenchymous tissue, which allows the exchange of gasses between the shoot and the root, and emission to the atmosphere through stem lenticels (Pangala et al., 2014).

Few studies have considered stem CH₄ fluxes in tropical trees (Pangala et al., 2017; Welch et al., 2019; Sjögersten et al., 2020; van Haren et al., 2021). For instance, in the Central Amazon basin, Pangala et al. (2017) have found significantly higher CH₄ fluxes from mature tree stem than from the soil. However, very little is currently known about the magnitudes and patterns of stem CO₂ and N₂O fluxes from different plant species, especially the role of Peruvian Amazon dominant palms, which have remained largely unknown. Therefore, this study aimed to estimate GHG fluxes from palm and tree stems and adjacent soil, and roughly upscale the fluxes to the whole aguaje palm (M. flexuosa) covered area in the region.

Based on the results of previous studies, we hypothesize that:

Monthly precipitation during the study period differed from the long-term pattern (1999–2019), being slightly lower than typical for the wetter season (January to March; Supplementary Figure 3); however, we follow the traditional climatic division in the hereafter and throughout the text. Soil temperature values at 10–40 cm showed low variation, from 24.2 to 27.8°C.

The soil CH₄ fluxes (average 3,617 μg C m^–2 h^–1) in the studied Quistococha palm swamp forest were within the same range of fluxes reported from similar studies in Brazilian flooded swamp forests in the Amazon river basin (Pangala et al., 2017) but lower than fluxes from palm swamps in the Southern Peruvian Amazon (Winton et al., 2017). Palm stem CH₄ emissions were an order of magnitude lower than the stem CH₄ fluxes of various tree and palm species reported by Pangala et al. (2017) and van Haren et al. (2021). The main differences between current and previous studies could be due to the regional soil water regime. The palm swamp forest soil was relatively dry during our study period, with significantly lower precipitation than the long-term average (Supplementary Figure 3). In our study period, the water table varied from −0.125 to +0.08 m (Griffis et al., 2020). In contrast, the water table in the Brazilian floodplain during the flood was up to 8 m higher than during the non-flooding period (Pangala et al., 2017).

At the end of dryer season (September), at a water level above −0.005 m, CH₄ fluxes from the palm stems were relatively high compared to the soil fluxes. That was likely caused by the oxidation of CH₄ in the aerated upper soil layer and, thus, by egress of CH₄ from deeper anoxic soil layers through the aerenchymatous palm roots and stems into the atmosphere (Figures 1, 2). These structures have intercellular (lacunose) spaces, which can serve as transport conduits for soil-produced CH₄ into the palm stems and finally into the atmosphere (Supplementary Figure 2; van Lent et al., 2019).

Average N₂O soil fluxes measured in our study were 2.5–7 times higher than model-predicted emissions of 21 μg N m^–2 h^–1 for the Amazon Basin (Guilhen et al., 2020) but still lower than N₂O emissions measured by Figueiredo et al. (2019) from floodplain soils in the Brazilian Amazon. In Amazonian palm swamps, no previous studies on N₂O stem fluxes are known. Seasonal variations in N₂O emissions are most likely correlated with seasonal changes in the water table and soil nitrogen mineralization. In our study, the higher N₂O production during the dryer period just after the heavy rainfall could indicate the presence of nitrate or other denitrification product in these soils and the effect of pulsing water level (Pärn et al., 2018).

As CH₄ is produced in the anaerobic soil by methanogens, the main factor influencing the soil’s methane production is the groundwater level (Lai, 2009). During our study, the soil CH₄ emission significantly increased when the soil surface was flooded in the wetter season (January–March). However, N₂O emission showed an opposite trend, being significantly lower during the wetter season than the dryer.

Emissions of CH₄ from stems differed by several orders of magnitude among the two studied plant species. This finding supports the results from other studies in temperate and tropical peatland ecosystems, where species-specific differences have been measured (Sjögersten et al., 2020; van Haren et al., 2021). Stem CH₄ emissions were significantly (p < 0.05) larger from palms than boarwood. This finding agrees with the study by van Haren et al. (2021). Higher CH₄ emissions from the palm stems might be explained by porosity of adventitious roots and stems (Jackson and Armstrong, 1999), tree specific wood density (Pangala et al., 2013) and stem water content (Wang et al., 2017), which significantly affect the gas diffusivity through stem wood.

The mean stem CO₂ fluxes from palm and boarwood stems (40.7 and 80.9 mg C m^–2 h^–1, respectively) in this study were in the same range as presented in overview study by Yang et al. (2016) for tropical forest ecosystems (58 mg C m^–2 h^–1). Emitted stem CO₂ could originate from different sources. A portion of CO₂ produced in the soil can be transported from below-ground through the xylem and during the transport diffused to the atmosphere (Teskey et al., 2008). Another part of stem CO₂ flux originates from a plant’s growth and maintenance respiration and is released to the atmosphere as autotrophic respiration (Lavigne and Ryan, 1997).

While studying the tree flux profile, we found the peak of GHG fluxes at the 80 cm height of palm stems (Supplementary Figure 4) which is contradictory to other previous studies in temperate and other tropical trees, where stem CH₄ flux rates decreased either linearly or exponentially with increasing stem height (Sjögersten et al., 2020; van Haren et al., 2021). In the case of palm trees, it might be explained by an effect of the lowest part of pseudobark. Outside of the central cylinder of the palm stem is a region of sclerified tissue known as the cortex and thin epidermis, which is referred to as the pseudobark (Broschat, 2013). The lowest part of the pseudobark is detached from the cortex and functional stem tissues due to the growth of adventitious roots from the root initialization zone (functional stem tissues) below the bark (Tomlinson, 1990). The GHG flux from the bottommost part of the pseudobark, which could be disconnected from the cortex as the new roots of the palm may force the cortex and pseudobark to split (Broschat, 2013), might be influenced.

The contribution of tree stem CH₄ fluxes constituted 4% of the total fluxes of soil and stems, significantly lower than reported in previous mesocosm studies with Alnus glutinosa saplings by Pangala et al. (2014). The share of N₂O stem fluxes was negligible in our study.

While roughly upscaling, our CH₄ measurements to the 27,732 ± 1,101 km² of palm swamp in the Pastaza-Marañon foreland basin (Draper et al., 2014) yield 897 Gg C y^–1. For N₂O, the upscaled estimate is 24 Gg N y^–1. Using global warming potentials (100 years, with feedbacks) of 31 for CH₄ and 298 for N₂O (Myhre et al., 2013), the annual warming effect of these gasses for the Peruvian palm peat swamps is 27.8 and 7.1 Tg CO₂ eq y^–1, respectively. Nevertheless, this should be taken as an initial and rough estimate. A more spatial and temporal cover is needed for a reliable account. The ongoing increase in droughts will likely increase both N₂O and CH₄’s role in the radiative balance of the whole Amazon Basin (Abalos et al., 2017; Zemp et al., 2017).

Our study shows that stems of aguaje palm (M. flexuosa), a dominant plant species in West Amazonian peatlands, can emit large amounts of CH₄, which is significantly higher than from boarwood (S. globulifera) stems. Thus, our first hypothesis was confirmed. Stem N₂O fluxes were low and we found no differences between the species. In contrast to our second hypothesis, the average soil CH₄ flux was significantly higher than flux from the stems: 3,617 μg C m^–2 h^–1 and 193.8 μg C m^–2 h^–1, respectively. The stem CH₄ emission was about a magnitude lower than reported previously for trees in the Amazon floodplain. The CH₄ fluxes were higher during the wetter season, while soil N₂O showed higher values during the dryer season. We suggest that future studies focus on plant species composition, stem anatomy, stand density, and other key factors that control emissions of CH₄ across Amazonian peatland plant species.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Field studies on plants were performed in accordance with relevant institutional, national, and international guidelines and legislation.

ÜM, KS, KM, JP, and RN-J: conceptualization. KS, KM, ÜM, JP, TS, and LF-M: methodology. TS, JP, KM, KS, ÜM, JJ-A, JR-M, LF-M, RZ-G, DG-D, and AA-O: investigation (field studies). KS, KM, ÜM, JP, and TS: data validation, quality check, formal analysis, data curation, and writing – original draft preparation. ÜM, KM, KS, LF-M, and WA-M: resources. KS, JP, KM, ÜM, TS, LF-M, RZ-G, and RN-J: writing – review and editing. ÜM, LF-M, RT-E, and WA-M: supervision. KS, JP, KM, ÜM, TS, LF-M, RN-J, TP-G, RT-E, and WA-M: project administration. ÜM, KM, JP, LF-M, WA-M, and TP-G: funding acquisition. All authors have read and agreed to the published version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.