BG3.19

Flooded savannas are extensive in South America and this study was conducted to assess two digital soil mapping (DSM) approaches to predict the spatial distribution of soil organic carbon (SOC) content and stocks in the Orinoco flooded savannas of Casanare department, located in the eastern plains of Colombia. SOC was estimated using a total of 80 sites sampled at two soil depth intervals (0-10 cm and 10-30 cm). SOC ranged from 0.41% at 0-10 cm and 0.23% at 10-30 cm in drier soils found in continental dunes with sandy textures and low vegetation cover (steppe) to over 14.50% and 7.51% in soils that experienced seasonal flooding located in depressions with loamy textures and flooded savanna vegetation. Predictions of the spatial distribution of SOC were done through Expert Knowledge (EK) and Random Forest (RF) approaches across the study area at 0-10 cm and 10-30 cm soil depth. Both DSM approaches were assessed through root mean square errors, mean absolute errors, and coefficients of determination. Although both DSM approaches performed very well, EK was considered slightly superior to predict SOC in the Casanare flooded savannas. Covariates derived from vegetation coverage, topography, and soil texture properties were identified as key drivers in controlling its distribution at the study area. We found total SOC stocks of 55.07 Mt with a mean density of 83.13 ± 24.32 t ha^-1 stored in the first 30 cm soil depth, with 12.3% of this being located in the flooded parts of the savanna landscape, which represented only 7.9% of the study area (664,752 ha). This study provides the first effort to systematically quantify SOC stocks in the Casanare flooded savannas and shows the importance of conserving this ecosystem with the aim of avoiding SOC losses and consequent increased CO₂ emissions to the atmosphere. We estimate that the department of Casanare would release an average of 2,42 Mton of CO₂ emissions per year over 30 years if there were large scale conversion of the flooded savannas to intensive agriculture, which corresponds to 62% of the current emissions of the department. At regional level, the impact of a large-scale land use conversion of the flooded Llanos del Orinoco ecosystem area (15 Mha) would represent 1/3 of the current net Colombian CO₂ emission (AFOLU), which makes this region a potentially important source of emissions if correct decisions are not taken regarding the land management.

How to cite: Martín-López, J. M., Verchot, L., Martius, C., and da Silva, M.: Modeling the spatial distribution of soil organic carbon and carbon stocks for the Casanare flooded Savannas, Colombia, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-1840, https://doi.org/10.5194/egusphere-egu22-1840, 2022.

Tropical peat swamp forests are major global carbon (C) stores that are particularly vulnerable to human intervention. In the Peruvian Amazonia they have been severely degraded through recurrent cutting of Mauritia flexuosa palms for fruit harvesting, and potentially been transformed from a CO₂ sink into a significant source. To estimate emissions associated with degradation, we combined C stock changes in aboveground biomass with peat C losses along a gradient comprising undegraded (Intact), moderately degraded (mDeg) and heavily degraded (hDeg) palm swamps. Temporal and spatial dynamics of the main components of the peat C budget (heterotrophic soil respiration (Rh) and litterfall) were investigated (bi)monthly over three years, while annual site-specific root C inputs and default dissolved organic C exports were taken from the literature. Variables measured at tree or microtopographic level were site-scaled considering forest structural changes from degradation. Site-scale litterfall (Mg C ha⁻¹ year⁻¹) at the hDeg site (2.3 ± 0.5) was less than half the rate at the Intact and mDeg sites (5.2 ± 0.9 and 6.0 ± 1.6, respectively). Conversely, site-scale Rh (Mg C ha⁻¹ year⁻¹) was higher at the hDeg site (9.6 ± 0.6) than at the Intact and mDeg sites (7.5 ± 1.1 and 6.1 ± 0.5, respectively). The peat carbon budget (Mg C ha⁻¹ year⁻¹) indicated that medium degradation reduced the sink capacity of the soil (from -1.8 ± 1.8 at the Intact site to -0.3 ± 0.7 at the mDeg site) while high degradation turned the soil into a high C source (6.0 ± 0.6 at the hDeg site). The large total C stock loss rates of 23.5 ± 14.3 and 57.7 ± 14.3 Mg CO₂ ha⁻¹ year⁻¹ at the mDeg and hDeg sites, respectively, which originated 94 and 77% from aboveground biomass changes clearly highlight the need for sustainable management of these peatlands.

How to cite: Kristell, H., Jeffrey, V. L., Nelda, D., Louis Vincent, V., Jan Willem, V. G., Mariela, L., and Julio, G.-R.: Major CO2 losses from degradation of Mauritia flexuosa peat swamp forests in western Amazonia, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-13155, https://doi.org/10.5194/egusphere-egu22-13155, 2022.

The megadiverse Andean mountain rain forests in southern Ecuador are threatened by climate and land use change, which are expected to alter biodiversity and thus functional traits impacting ecosystem processes. However, the high biodiversity of tropical mountain forests is still poorly represented in Land Surface Models (LSMs). We developed a biodversity-informed LSM entitled HUMBOL-TD (Hydroatmo Unified Model of Biotic interactions and Local Trait Diversity) to analzye the impact of climate and land-use change on carbon- and water fluxes. HUMBOL-TD consists of three coupled submodels specialized to represent different processes at the land surface. As such, energy- and water fluxes between land surface and atmosphere (LSMatmo) are simulated by the Community Land Model (CLM), vegetation dynamics including C, N and P cycling (LSMbio) are simulated by the Lund-Potsdam-Jena General Ecosystem Simulator (LPJ-GUESS), while the soil hydrology (LSMhydro) is represented by the Catchment Modeling Framework (CMF). A first test towards the simulation of the mountain forests and their replacement systems is conducted for a pasture site at 2000 m elevation. The model is parameterized and validated using a year of local site data. The first runs of the model enable the investigation of the differences in accuracy of modeled changes in the carbon- and water fluxes between coupled, partially coupled (LSMatmo – LSMbio, LSMatmo – LSMhydro, LSMbio – LSMhydro) and the fully coupled model (LSMatmo – LSMbio – LSMhydro).

How to cite: Limberger, O., Dantas De Paula, M., Windhorst, D., Trachte, K., Breuer, L., Hickler, T., and Bendix, J.: The new coupled land surface model HUMBOL-TD - Concept and performance, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12150, https://doi.org/10.5194/egusphere-egu22-12150, 2022.

Deforestation followed by fertilizer intensive agriculture is widely recognized as a significant contributor to anthropogenic greenhouse gas emissions (GHG), particularly carbon dioxide (CO₂), methane (CH₄), and nitrous oxide (N₂O). However, empirical studies focusing on soil GHG flux dynamics from deforestation hotspots in the tropics are still limited creating major uncertainties for constraining global GHG budgets. In this study, we investigated how deforestation for fertilizer intensive sugarcane cultivation in Uganda affects soil-borne GHGs. Therefore, soil GHG fluxes were measured in a primary forest and in a completely randomized experiment premised in the neighboring sugarcane fields with different fertilizer regimes, representing both smallholder and industrial-scale sugarcane farm management. Despite the use of different fertilization rates (low, standard, and high) as treatments for the sugarcane CRD experiment, neither auxiliary controls nor soil GHG fluxes significantly differed among the CRD treatments. Soil respiration was higher in the sugarcane than in the forest, which we attribute to the increased autotrophic respiration from the sugarcane’s fine root biomass and the likely exposure of the sugarcane’s larger soil organic carbon stocks to microbial decomposition through ploughing operations. The forest soils were a stronger net sink of CH₄ than the sugarcane soils despite forest soils having both higher bulk densities and larger water-filled pore space (WFPS), and we suspect that this was due to alteration of the methanotroph abundance upon the conversion. Soil N₂O emissions were smaller in the sugarcane than in the forest, which was surprising, but most likely resulted from the excess N being lost either through leaching or uptake by the sugarcane crop. Only seasonal variability in WFPS, among the auxiliary controls, affected CH₄ uptake at both sites and soil CO₂ effluxes in the sugarcane. Noteworthy, soil N₂O fluxes from both sites were unaltered by the seasonality-mediated changes in auxiliary controls. All the findings put together suggest that forest conversion for sugarcane cultivation alters soil GHG fluxes by increasing soil CO₂ emissions and reducing both soil CH₄ sink strength and soil N₂O emissions.

How to cite: Tamale, J., van Straaten, O., Hüppi, R., Turyagyenda, L. F., Fiener, P., and Doetterl, S.: Soil greenhouse gas fluxes following tropical deforestation for fertilizer-intensive sugarcane cultivation in northwestern Uganda, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-761, https://doi.org/10.5194/egusphere-egu22-761, 2022.

Indonesia is one of the hotspots of land transformation from forest ecosystems toward oil palm and other cash-crop monocultures. Land-cover changes directly impact below-canopy microclimate, which are critical drivers for many ecological functions, such as greenhouse gas exchange and soil microbial activity. However, microclimatic variability below canopies, even within the same land-use type can be quite large due to structural heterogeneity, vegetation age or vitality, and differences in management practices.

In this study, we focused on the assessment of microclimatic differences within the most common land-use types in tropical lowland Jambi province (Sumatra, Indonesia), using mini-meteorological stations. We used a rapid assessment approach in which we monitored below-canopy key meteorological parameters at a total of 120 different locations from June to November 2021, covering lowland tropical rainforest, oil palm monoculture, rubber monoculture and agroforestry systems, and fallow shrublands. We clustered the study region into 16 micro-regions, each with a radius of four kilometres. In each micro-region, an open-land area served as a reference meteorological location. Based on the gradients of meteorological parameters between below-canopy and open-land conditions we derived the site-specific impact of the respective land-use type on below-canopy microclimate. To further explore microclimatic characteristics of the different land-use types, we used airborne laserscanning (ALS) data available at a subset of 90 plots as well as information on age, management intensities and ownerships of plantations, distance between plantations and forests, and overall land cover distribution.

Preliminary results show that forests and fallow shrublands are generally cooler, wetter and receive lower below-canopy radiation compared to agricultural systems and open land. Forests show a strong capacity to buffer high levels of open-land air temperature and atmospheric vapour pressure deficit (VPD) variability by, on average, 1.7°C and 6.4 hPa, respectively, while oil palm showed very little buffering capacities (0.2°C and 2.2 hPa). At a regional scale, mixed land-use systems tend to be slightly warmer (+0.36±0.18°C) and drier (+1.47±0.52 hPa VPD) compared to forest-dominated land-use systems. Within the mixed land-use systems, forests tend to be drier (+1.05±0.41 hPa VPD) while below-canopy temperature remains similar (+0.38±0.34°C) compared to forests in the forest-dominated land-use systems. Interception is an important component in the hydrology of the studied forest locations, with approx. 66% of precipitation being intercepted, while at fallow shrubland, rubber and oil palm locations, only 24, 25 and 17%, respectively, of precipitation was intercepted. Overall, our preliminary results show that there is high variability in meteorological conditions, even within the same micro-region or land-use type.

How to cite: Stiegler, C., Camarretta, N., Ali, A., June, T., Wenzel, A., and Knohl, A.: A rapid assessment of microclimate and meteorological conditions in the tropical lowlands of Jambi province (Sumatra, Indonesia): Land-use intensity gradients and spatial small-scale climate variability across 120 plot locations, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7375, https://doi.org/10.5194/egusphere-egu22-7375, 2022.

Knowledge about microclimatological conditions strongly contributes to our understanding of land surface – atmosphere interactions as drivers of the Earth’s surface energy budget. Particularly the radiative fluxes are major determinants providing energy for vital climate processes and are crucial for climate warming, water availability, primary productivity and ecosystem services. The partitioning into sensible and latent heat fluxes are highly dependent on the land coverage and represent feedback effects affecting the cycling of heat and water in the vegetation-atmosphere continuum. In the Reserva Biologica San Francisco (RBSF) on the eastern escarpment of the South Ecuadorian Andes on 2000m elevation above sea level (a.s.l.) two eddy-covariance measurement stations have been installed over natural rain forest and pasture ecosystem to observe atmospheric water and carbon fluxes. The aim is to assess net-ecosystem exchange (NEE) and evapotranspiration (ET) in order to estimate the impact of deforestation on the carbon sink function and the water availability. Additionally, microclimatological conditions in terms of e.g. radiative fluxes and soil conditions are supposed to further disentangle effects of the respective land surface properties on the environmental conditions. Over the last three years generally higher water fluxes could be observed during daytime over the forest ecosystem compared to pasture. Concerning NEE a clear carbon sink was revealed for both ecosystems indicated by a mean gross primary productivity (GPP) of 12.7 gC/m²day (forest) and 6.5 gC/m²day (pasture), while a mean ecosystem respiration (Reco) of 10.6 gC/m²day (forest) and 5.9 gC/m²day (pasture) was obtained. However, a mean NEE of 2.1 gC/m²day (forest) and 0.6 gC/m²day (pasture) clearly shows the stronger productivity of the forest ecosystem and thus, a higher carbon sink as a contribution to climate change mitigation.

How to cite: Trachte, K., Pucha Cofrep, F., Raffelsbauer, V., Limberger, O., Fries, A., Carillo-Rojas, G., and Bendix, J.: Microclimatological conditions along a land use gradient in the tropical Andes of South Ecuador, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11987, https://doi.org/10.5194/egusphere-egu22-11987, 2022.

How to cite: Ba, S. M., Roupsard, O., Chapuis-Lardy, L., Bouvery, F., Diongue, D. M. L., Agbohessou, Y. F., Guérin, F., Tagesson, H. T., Sambou, B., and Serça, D.: Monitoring soil greenhouse gas (GHG) emissions in a Sahelian agrosilvo-pastoral parkland, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-4337, https://doi.org/10.5194/egusphere-egu22-4337, 2022.

The importance of the land sector in addressing the climate and nature crises has gained worldwide attention. Nature-based solutions were a key topic at the recent United Nations Conference of the Parties (COP26) in Glasgow to limit global warming to well below 2 degrees. The conservation, restoration, and improved management of peatlands play a significant role in Indonesia's nature-based solutions.

The eddy covariance measurements of net ecosystem carbon dioxide and methane exchanges from a coastal peatland in Sumatra, Indonesia indicate that the GHG balance increased from 20.0 ± 4.5 tCO₂e ha⁻¹ yr⁻¹ at the intact site (undrained and undisturbed forest cover) to 43.8 ± 1.5 tCO₂e ha⁻¹ yr⁻¹ at the degraded site (drained with canal system and selectively logged). The significant carbon dioxide emissions from the intact site, during an extreme drought caused by a positive Indian Ocean Dipole phase combined with El Niño event, highlight the potential importance of climate regime in determining the GHG budget of tropical peatlands.

Although the measurements indicate that both intact and degraded peatlands in this study are warming the atmosphere, it remains clear that protection of the remaining intact tropical peatlands offers a viable way to avoid substantial GHG emissions from this globally important ecosystem, which for our study in Sumatra was 24 ± 5 tCO₂e ha⁻¹ yr⁻¹. These results highlight that protecting all remaining intact peat swamp forests in Indonesia (6.2 Mha) from degradation will avoid GHG emissions of around 0.15 GtCO₂e yr^-1, this equates to ~10% of Indonesia’s GHG emissions in 2016.

Additionally, tropical peatland conservation contributes directly to the UN Sustainable Development Goals by fostering unique biodiversity and ecosystem services.

How to cite: Deshmukh, C. S., Susanto, A., Asyhari, A., Desai, A. R., Page, S., Nardi, N., Nurholis, N., Hendrizal, H. M., Kurnianto, S., Suardiwerianto, Y., Agus, F., Astiani, D., Sabiham, S., Gauci, V., and Evans, C.: Tropical peatland conservation in Indonesia as a nature-based solution, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-6896, https://doi.org/10.5194/egusphere-egu22-6896, 2022.

Logged and degraded tropical forests are fast becoming one of the most dominant land-use types throughout the tropics, yet there is limited understanding of the impact of logging on tropical forest function and carbon balance. To date, previous research on the carbon dynamics of logged and degraded forests has mostly focused on carbon stock recovery during forest regrowth and asserted these ecosystems as an important carbon sink due to rapid increase in stem biomass. These estimates of biomass sink function do not, however, serve as an assessment of the ecosystem carbon balance, as they do not include estimates of the carbon losses through ecosystem respiration, particularly from heterotrophic sources. We quantified the complete carbon budget in old-growth, moderately logged, and heavily logged forests within Malaysian Borneo, a region that is a hotspot for deforestation and degradation. We present the first direct measurements of net ecosystem CO₂ exchange from a logged and structurally degraded tropical forest and show how this landscape represents a substantial net carbon source to the atmosphere, using both eddy covariance technique and ground-based biometric estimates. We estimated a net carbon source of 4.66 ± 1.36 Mg C ha^-1 year^-1 across the logged plots sampled (n=5), compared to 0.69 ± 1.06 Mg C ha^-1 year^-1 within old-growth plots (n=6). Our results showed a high level of variability along the logging gradient, ranging from 1.88 ± 4.29 Mg C ha^-1 year^-1 in a moderately logged plot to 8.16 ± 4.16 Mg C ha^-1 year^-1 in a heavily logged plot, highlighting that unsustainably logged areas function as substantial net carbon sources. Eddy covariance measurements over the heavily logged landscape estimated a net carbon source of 12.24 ± 2.06 Mg C ha^-1 year^-1_, similar to that of the heavily logged biometric plot located within its footprint. Consistent with existing literature, our study showed a significantly greater woody biomass gain during regrowth across moderately and heavily logged forests, compared with old-growth forests. This was not due to higher total net primary productivity but due to an allocation shift towards the increased production of woody tissue. Gross and net primary production was largely unaffected by logging, but ecosystem respiration, particularly from heterotrophic sources was significantly higher in logged forests. Despite increased tree growth rates within recovering logged forest compared to old-growth forests, these systems do not necessarily function as a net carbon sink, especially if past disturbances cause persistent carbon losses from soil and necromass. We, therefore, demonstrate critically how focussing on carbon gain from woody biomass accumulation alone does not provide a complete picture of carbon cycling within logged tropical forests, and how heavily degraded forests function as net carbon sources.

How to cite: Mills, M., Malhi, Y., Ewers, R. M., Kho, L. K., Teh, Y. A., Burslem, D. F. R. P., Both, S., Majalap, N., Nilus, R., Huaraca Huasco, W., Turner, E., Reynolds, G., and Riutta, T.: Logged tropical forests are a net carbon source to the atmosphere as investigated by eddy covariance and biometric ground-based estimates, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-2537, https://doi.org/10.5194/egusphere-egu22-2537, 2022.

Conventional intensive management, such as high fertilizer and herbicide applications, are common practice in large-scale oil palm plantations. One of the proposed solutions to lessen its environmental impact is to reduce fertilization and employ mechanical weeding without sacrificing yield and profit. A full factorial experiment with two fertilization rates (260 N, 50 P, 220 K kg ha^-1 yr^-1 as conventional practice, and 136 N, 17 P, 187 K kg ha^-1 yr^-1, equal to harvest export, as reduced management) and two weeding methods (conventional herbicide application, and mechanical weeding as reduced management) was established in 2016 at a large-scale oil palm plantation (planted in 1998-2002) on a sandy clay loam Acrisol soil in Jambi, Indonesia. Soil CO₂, N₂O, and CH₄ fluxes were measured monthly from July 2019 to June 2020, using vented static chambers. At each plot, the measurements were conducted on two randomly selected subplots, and in each subplot, we measured at three management zones (palm circle, inter-row, and frond-stacked area). During 2017-2020, fruit yield did not differ among treatments (fertilization: P=0.35; weeding control: P=0.11). Soil CO₂, N₂O, and CH₄ fluxes also did not differ among treatments (fertilization: P>0.81; weeding control: P>0.28). Area-weighted from the three management zones, soil CO₂ fluxes (mg C m^-2 h^-1) were 61±2 for conventional and 65±4 for reduced fertilization and 64±4 for herbicide and 62±2 for mechanical weeding. Soil N₂O fluxes (µg N m^-2 h^-1) were 46±12 for conventional and 45±16 for reduced fertilization and 57±15 for herbicide and 34±12 for mechanical weeding. Soil CH₄ fluxes (µg C m^-2 h^-1) were -17±2 for conventional and -17±3 for reduced fertilization and -17±3 for herbicide and -17±2 for mechanical weeding. Distinct differences were observed among the three management zones. Frond-stacked area, with high soil organic carbon and low soil bulk density, had the highest soil CO₂ emission and soil CH₄ uptake (P≤0.01). Palm circle, with fertilizer application and high soil bulk density, had the highest soil N₂O emission and lowest soil CH₄ uptake (P≤0.01). Inter-row area, with low soil organic carbon and no direct fertilizer application, had the lowest soil CO₂ and N₂O emission (P≤0.01). Soil CO₂ (rho=0.64, P≤0.05) and N₂O (rho=0.53, P≤0.05) fluxes were positively correlated with total mineral N. Soil CH₄ flux was negatively correlated with total mineral N (rho=-0.30, P≤0.05) and positively correlated with water-filled pore space (rho=0.66, P≤0.05). Although the frond-stacked area only accounted for 15% of the oil palm plantation area, it contributed 30% of soil CO₂ emission and 41% of soil CH₄ uptake. The palm circle accounted for 18% of the oil palm plantation area and contributed 79% of soil N₂O emissions. Our results indicated that the inherent management zones in oil palm plantations should be spatially represented for accurate quantification of soil greenhouse gas fluxes. Our findings showed that reduced management maintained yield whereas soil greenhouse gas fluxes remained high at least during the 3-4 years of this management experiment, which signified the legacy effect of more than a decade-long conventional management in this mature oil palm plantation.

How to cite: Chen, G., Veldkamp, E., Tjoa, A., Damris, M., Irawan, B., and Corre, M. D.: Soil greenhouse gas fluxes from large-scale oil palm plantation under conventional and reduced management systems, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-2664, https://doi.org/10.5194/egusphere-egu22-2664, 2022.

Oil palm (OP) plantations have replaced large areas of forest in the tropical landscape of Southeast Asia and are major emitters of greenhouse gases (GHGs). However, within established plantations there are management options which may reduce these emissions, including altered management practices within plantations and restoring forest within the landscape. Managing the vegetation within and around plantations could potentially minimise environmental damage and maximise co-benefits such as soil protection, pest control and support for biodiversity. Such practices include relaxed management, passive restoration, and active restoration. The impact of these management practices is uncertain, and there is a real need for an evidence-base to guide improvements in the environmental sustainability of OP management.

Here we present GHG fluxes from two long-term experiments as part of ‘The Biodiversity and Ecosystem Function in Tropical Agriculture’ (BEFTA) Project. The first experiment is investigating the impact of three alternative understory management treatments on biodiversity, ecosystem functioning and yield in Sumatra, Indonesia:

• Normal biodiversity complexity: standard industry practice, intermediate level of herbicide use in harvest circles around the palms.
• Reduced biodiversity complexity: spraying/removing all understory vegetation with herbicides.
• Enhanced biodiversity complexity: reduced-input management with no herbicide application and limited understory cutting.

The second experiment focusses on riparian restoration options (‘Riparian Ecosystem Restoration in Tropical Agriculture’ (RERTA) Project). The experimental site began as a mature OP plantation, followed by felling in April 2019 and replanting and riparian restoration in October 2019. Four management strategies were applied on both sides of a river to create 50 m riparian buffers, 400 m in length:

• A control treatment of no restoration, the removal of mature OP and replanting of young OP to the river margin.
• Passive restoration: Little to no agricultural management of mature OP.
• Active restoration A: Clearance of mature OP and enrichment planting with native forest trees.
• Active restoration B: Little or no agricultural management of mature OP and additional enrichment planting with native forest trees.

For both experiments, we measured the GHGs nitrous oxide (N₂O), methane (CH₄) and ecosystem respiration/carbon dioxide (CO₂) from static chambers and analysis by gas chromatography (GC-µECD/FID). Additionally, meteorological and basic soil parameters were measured as potential variables or drivers of measured fluxes that might be greater than any ‘treatment’ or ecological management effect. Measurements were carried out monthly from the understory treatments, taken from 54 static chambers for the duration of one year starting in October 2018. For the riparian restoration project, monthly background measurements were taken between January and April 2019 and then approximately monthly after replanting from 6 chambers in each riparian treatment and 16 in the actual OP plantation resulting in 40 chambers in total.

We investigated whether the observed ecological benefits of alternative management and restoration options such as introducing native tree species in riparian buffers and allowing the natural regrowth of understory in plantations may be associated with an additional or reduced GHG burden; thereby assessing the overall environmental impact.

How to cite: Drewer, J., Sionita, R., White, S., Luke, S., Turner, E., Raine, B., Banin, L., Skiba, U., Advento, A. D., Ketut Aryawan, A. A., Caliman, J.-P., and Pujianto, P.: Diversifying understory vegetation and riparian restoration as ecological management options to regulate greenhouse gas fluxes in oil palm plantations, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-13120, https://doi.org/10.5194/egusphere-egu22-13120, 2022.

Mangrove and other coastal wetlands such as saltmarsh and seagrass are termed ‘blue carbon’ ecosystems due to their substantial capacity for carbon storage and sequestration over a long-term time scale. Policymakers and stakeholders are currently promoting mangroves into national carbon management as part of nature-based climate change mitigation and adaptation strategy. Unfortunately, global mangroves area with particularly in the tropics is decreasing at a rapid rate due to land-use and land-cover change (LULCC). Yet, there has been limited study of carbon emissions impacted by multiple mangrove conversions at the landscape scale. Here we assessed spatio-temporal patterns of soil CO₂ and CH₄ effluxes across six land uses, namely mangroves converted to 15 yrs oil palm, 20 yrs coconut, and 20 yrs aquaculture (pond wall and water surface), as well as newly logged mangrove, 10 yrs planted mangrove, and undisturbed mangrove forests reference in North Sumatra, Indonesia. Direct measurement of soil CO₂ and CH₄ effluxes were performed by using an ultra-portable LGR gas analyser during low tide condition between 08.00 and 16.00, with triplicated PVC 10-inch diameter and 25 cm height opaque static chambers (closed system) were installed at each land use in September-October 2021 -- representing wet season in the study site. The soil CO₂ and CH₄ effluxes were collected three times for each chamber and 3 days of measurement during this field campaign with a total of 193 measurements were performed. We observed that the top three highest soil CO₂ and CH₄ effluxes were among aquaculture pond wall soils (591±104 mgCO₂ m² h^-1 and 0.40±0.17 mgCH₄ m² h^-1), logged mangroves (480±104 mgCO₂ m² h^-1 and 3.21±1.34 mgCH₄ m² h^-1), and natural mangroves (274±71 mgCO₂ m² h^-1 and 0.58±0.28 mgCH₄ m² h^-1). By contrast, relatively low effluxes (< 200 mgCO₂ m² h^-1 and < 0.1 mgCH₄ m² h^-1) were observed across other land-use types. Our preliminary results suggest that the variation of soil CO₂ and CH₄ in our study sites may be controlled by the duration of the disturbances, particularly we observed the highest CO₂ and CH₄ effluxes at newly (occurred at the same year with our measurement) constructed pond wall and logged mangrove locations. On the other hand, low CO₂ and CH₄ effluxes were observed at both oil palm and coconut plantations. These new land uses were constructed more than 10 years ago with the application of drainage and tidal blocking. Our current limited data constraint further essential factors that commonly control CO₂ and CH₄ in the coastal wetlands, such as tidal elevation, bioturbation, seasonal variation, and soil properties. Overall, our dataset will be essential to guide policymakers in related to the improvement of land-based low carbon development and climate change mitigation strategies for Indonesia to meet the targeted 29% of unconditional carbon emissions reduction by 2030 as outlined in the Nationally Determined Contributions (NDCs) as part of the Paris Agreement.

How to cite: Sasmito, S., Arriyadi, D., Bimantara, Y., Amelia, R., Saragi-Sasmito, M., Darusman, T., Basyuni, M., Maher, D., Hutley, L., and Murdiyarso, D.: CO2 and CH4 effluxes across six land uses in coastal wetlands of North Sumatra, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-13486, https://doi.org/10.5194/egusphere-egu22-13486, 2022.

Tropical forest plays a fundamental role in the ecosystem services maintenance. Amazon forests have been subject to intense land use and cover changes (LUCC), mainly in the Southeast portion. Like many tropical countries, more than 70% of Brazilian greenhouse gasses emissions come from LUCC. Under the framework of the CARBAM Project, atmospheric CO₂ measurements in four sites of the Amazon, show that there is a reduction in the Amazon forest capacity to absorb C in the proximities of previous deforested and degraded forest areas, such as the well-known “Deforestation Arc” in the Southeast amazon. There are many LUCC databases now available that allow to assess the deforestation, degradation and second forest dynamics and contribute to a better understanding of the carbon dynamics of nine years of in situ atmospheric CO₂ measurements. Nevertheless, in order to know how much CO₂ is released to the atmosphere due to LUCC, it is necessary to quantify how much carbon is stored in the forest biomass and to assess the biomass variability along the different datasets. Here we compared the forest biomass quantity of three biomass maps: the fourth national communication of Brazil map (official), a global map (Baccini et al. 2012) and a regional map for the Brazilian Amazon (EBA project). We found significant differences for the Brazilian Amazon: between the official biomass map and the regional map 27%, between the global and regional map 25% and the smallest difference was between the official and the global map (3%). Even though the official and the regional maps were obtained using the same data inputs, the official map refers to a potential biomass for 2010 and the regional map reflects the real biomass in 2016, this could explain the difference. The official and global maps represent the potential biomass, and as we used the mean forest area, the biomass content is similar. When comparing these maps at a deforested pixel level the differences could be larger. The spatial and temporal scale of biomass maps make it hard to estimate the CO₂ emissions of degradation and secondary forest loss and growth which are fundamental to understand the Amazon C balance under climate change and LUCC pressures.  

Key words: Amazon, CO₂ emissions, forest biomass, land use and cover change, carbon balance

How to cite: Tejada, G., V. Gatti, L., S. Basso, L., A. V. Mataveli, G., L.G. Cassol, H., and Von Randow, C.: Forest biomass - an uncertainty source of land use and land cover change related carbon emissions in the Amazon, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10410, https://doi.org/10.5194/egusphere-egu22-10410, 2022.

Among the greatest threats to the global climate is the possibility that the Amazon rainforest, Earth’s largest carbon stock, becomes a net carbon source. To estimate the Amazon’s carbon budget, Gatti et al. (2021) performed 590 atmospheric vertical profiling measurements from four sites using aircraft over the course of eight years. They found that intact forests of the southeastern Amazon already act as a carbon source. This is likely related to decreased precipitation levels, stressing the importance of maintaining or enhancing precipitation levels in that region. The level and variability of precipitation partly depends on the land cover at the location where the moisture has evaporated. Forests in the Amazon enhance evapotranspiration, which significantly contributes to regional precipitation levels. This spatial connection between evapotranspiration and precipitation implies a causal link between forest cover at a certain location and the carbon budget at remote locations. To determine these evapotranspiration-precipitation connections, we use a high-resolution Lagrangian atmospheric moisture tracking model forced with ERA5 reanalysis data. We determine the seasonally changing spatial distributions of the moisture sources of different parts of the Amazon that have different carbon dynamics. We obtain land characteristics of these moisture-source areas to explore the potential of forest restoration for maintaining or regaining the carbon sink in the Amazon. We find that, on average, about one-third of the precipitation in the area identified as a carbon source originates as evaporation from land, the majority of which in this region itself. We find seasonality in the amount of moisture that is recycled within this region, peaking in the fourth quarter. The results indicate that deforestation in the southeastern Amazon may accelerate the carbon emissions from remaining intact parts of the Amazon. Further, they show where forest restoration may be prioritized to prevent these emissions.

How to cite: Staal, A., Tejada, G., and Gatti, L.: Moisture sources of the Amazon carbon source, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3828, https://doi.org/10.5194/egusphere-egu22-3828, 2022.

The increasing demand for charcoal in Sub-Saharan Africa (SSA) is a growing threat to tropical ecosystems as more forest areas get cleared to meet the high energy needs. While the region’s current socio-economic trends, such as increasing population, urbanisation and high poverty levels, will likely drive high charcoal demands into the future, current estimates indicate that charcoal production contributes up to 7% of total deforestation in tropical ecosystems every year, with carbon emissions corresponding to 71.2 million tonnes of CO₂ and 1.3 million tonnes of CH₄. Although forest management practices could enable sustainable production by using harvest cycles to allow forest regeneration, emissions from charcoal production may contribute to exacerbate global warming. A transition for other energy carriers in SSA has been called for, which may be a slow process as it depends on investments and cultural changes, thus projected demands for charcoal could severely impact the balance and timing of carbon fluxes and the overall carbon budget of tropical ecosystems. To better understand how charcoal production affects tropical ecosystems carbon dynamics, we parameterised a dynamic global vegetation model, LPJ-GUESS, to determine the magnitude and direction of carbon fluxes following charcoal production. We simulated 300 model years for two forest governance regimes, natural and managed forest, on 782 gridcells at 0.5° x 0.5° resolution covering the tropical rain forest of Africa. We allowed for tree harvesting for charcoal only in managed forests, where we vary the fraction of trees cut (10%, 20%, and 30%) and harvest rotation cycles (10, 20, and 30 years). We find that Net Ecosystem Exchange (NEE) under all charcoal production regimes cause tropical forests to transition from a net carbon sink (NEE natural = -0.024 ± 0.047 kg C/m² yr-1) to a net carbon source. We estimate NEE = 0.005 ± 0.432 kg C/m² yr-1 under the least intense management regime (10% forest cut every 30 years) and a mean NEE of 0.027 ± 0.630 kg C/m² yr-1 for the most intense regime (30% forest cut every 10 years). We further observe an initial and steep drop in vegetation carbon following the start of charcoal production for all management regimes, and this change quickly stabilises as tree harvest keeps vegetation under a new stable state that is lower than that of natural forests. Compared to our modelled natural forest, we find that all charcoal regimes lead to more than a 25% decline in vegetation carbon over time. We further examined carbon partitioning into pools of litter and soil and find consistent patterns of transition from sink to source. These findings suggest that while carbon dynamics vary in tropical systems depending on the intensity and frequency of charcoal production, even a management regime of 10% charcoal production every 30 years can result in forest carbon loss with amplified vegetation carbon losses in the order of 25%. 

How to cite: Sakala, D. and Santos, M. J.: Effects of charcoal production on carbon cycling in African tropical forests, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7091, https://doi.org/10.5194/egusphere-egu22-7091, 2022.

Application of mineral nitrogen (N) fertilizer and water management are two very essential farming practices, used to optimize potential yields in sub-Sahara African rice cultivation. Differences in both practices, however, might affect the patterns of climate relevant gaseous carbon (C) emissions (CO₂ and CH₄) and soil C losses, thus contributing to global climate change. To date, knowledge about the combined effects of different N fertilizer rates together with different water management practices on the gaseous C emissions and soil C losses are very limited. This is even more the case for arable lands in sub-Sahara Africa. Our study aims to identify the best combination of water management and N fertilizer amount to reduce gaseous C emissions and limit soil C losses for an irrigated rice production in Benin. We hypothesize that especially a combination of alternate wetting and drying (AWD) as water management and an optimum amount of N fertilizer reduce gaseous C emissions and might help to enhance C sequestration by reducing soil C losses from irrigated rice production in Benin. To test this hypothesis, a field experiment was established at Koussin lélé, Cove district, southern Benin using a full factorial, split-plot experimental design. Within the experiment the combination of three levels of water management and two levels of N fertilizer amount are tested. The water management technologies include continuous flooding (CF) and two alternate wetting and drying (AWD) methods (AWD15 and AWD25) of irrigation. Nitrogen fertilizer levels is 90 kg/ha (farmer’s practice) and 120 kg/ha (high amount of fertilizer). To measure gaseous C emissions (CO₂ and CH₄) and estimate dynamics in soil C losses, an innovative, customized low cost dynamic NFT-NSS closed chamber system is used. The system consists of CO₂/CH₄ NDIR sensors connected to a microcontroller for data storage and transparent (NEE measurements) polycarbonate chambers (40 cm x 40 cm x 100 cm). To measure R_eco, transparent chambers where covered with an opaque hood. Chamber measurements for diurnal variability in CH₄ and CO₂ fluxes are performed biweekly at all plots. In addition, agronomy and crop growth indices such as the Normalized difference vegetation index (NDVI) are measured weekly. Here we present CO₂ and NECB balances for the first crop growth period.

Key words: Water management, N fertilizer, CO₂ emission, net ecosystem carbon balance (NECB), rice

How to cite: Sossa, L. G., Naab, J., Augustin, J., Sintondji, L., Sanogo, S., and Hoffmann, M.: Effect of N fertilizer amount and water management on CO2 exchange and net ecosystem C balance of rice cultivation in Southern Benin, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-232, https://doi.org/10.5194/egusphere-egu22-232, 2022.

Degradation, conversion and drainage of tropical peatlands generate sizeable emissions of greenhouse gases (GHG). Current IPCC default emission factors (EF) for drained tropical peatlands are based on a very limited number of observations, thereby resulting in large uncertainties in emissions estimates. Impacts of disturbance on peat GHG emissions in undrained tropical peatlands can also be substantial but are not well characterized and not considered by IPCC guidelines. Research is critically needed to support development of more accurate EF for national GHG accounting for both drained and undrained degraded tropical peatlands. To explore the potential of process-based modelling to refine tropical peat EF, we used the DeNitrification DeComposition (DNDC) model to simulate peat GHG emissions and biogeophysical variables in oil palm plantations and undrained primary and secondary peat swamp forests of Central Kalimantan, Indonesia.

The simulated magnitude of C inputs (litterfall and root mortality) and dynamics of annual heterotrophic respiration and peat decomposition N₂O fluxes in oil palm plantations were generally consistent with field observations. The modelled onsite oil palm peat CO₂ EF was lower than the IPCC default (11 Mg CO₂-C ha^-1 yr^-1) and decreased from 7.7 ± 0.4 Mg C ha^-1 yr^-1 in the first decade to 3.0 ± 0.2 and 1.8 ± 0.3 Mg C ha^-1 yr^-1 in the second and third decades of the rotation. The modelled N₂O EF from peat decomposition was higher than the IPCC default (1.2 kg N ha^-1 yr^-1) and increased from 3.5 ± 0.3 kg N ha^-1 yr^-1 in the first decade to 4.6 ± 0.5 kg N ha^-1 yr^-1 in the following ones. Modelled fertilizer-induced N₂O emissions were minimal and much less than 1.6% of N inputs indicated by the IPCC EF in wet climates regardless of soil type. Temporal variations in oil palm EF were strongly linked to soil C:N ratio and mineral N content for CO₂ and fertilizer-induced N₂O emissions, and to precipitation, water table level, and soil NH₄⁺ content for peat decomposition N₂O emissions. These results suggest that current IPCC EF for oil palm on organic soil could over-estimate onsite CO₂ emissions and underestimate peat decomposition N₂O emissions and that decadal-scale temporal variation in emissions should be considered for further improvement of EF. Simulations allowed the generation of oil palm EF disaggregated by plantation age and emission source (decomposition, fertilizer-induced), a practical and useful application for GHG inventories in tropical peatlands.

In unconverted land uses, the GHG budget (Mg CO₂-equivalent ha^-1 yr^-1) was ten times higher in the secondary forest (10.2 ± 4.5) than in the primary forests (0.9 ± 3.9) on the account of a larger peat C budget and N₂O emission rate. Preliminary modelling results suggest increased peat C outputs from heterotrophic respiration and decreased C inputs from litterfall and root mortality in secondary forest compared to primary forest. Our study highlights the disastrous atmospheric impact associated with not only conversion to oil palm but also forest degradation in tropical peatlands and stresses the need to investigate GHG fluxes in disturbed undrained lands.

How to cite: Swails, E., Hergoualc'h, K., Deng, J., and Frolking, S.: How can process-based modelling improve tropical peat greenhouse gas emission factors?, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5145, https://doi.org/10.5194/egusphere-egu22-5145, 2022.

Land use change specially affects greenhouse gases (GHGs) emissions, and it can act as a sink/source of GHGs. Alterations in edaphic properties and microbial attributes induced by land use change can individually/interactively contribute to GHGs emission, but how they predictably affect soil CO₂, CH₄, and N₂O emissions remain unclear. Here, we investigated the direct and indirect controls of edaphic properties [i.e. dissolved organic C (DOC), soil organic C (SOC), total N (TN), C: N ratio, NH₄⁺-N, NO₃^－-N, soil temperature (ST), soil moisture (SM), pH, bulk density (BD)] and microbial attributes [i.e. total PLFAs (Phospholipid fatty acids), 18:1ω7c, nitrifying genes (ammonia-oxidizing archaea (AOA), ammonia-oxidizing bacteria (AOB)), and denitrifying genes (nirS, nirK, and nosZ)] over the annual soil CO₂, CH₄, and N₂O emissions from the woodland, shrubland, and abandoned land in subtropical China. Soil CO₂ and N₂O emissions were higher in the afforested lands (woodland and shrubland) than in the abandoned land, but the annual cumulative CH₄ uptake did not significantly differ among all land use types. The CO₂ emission was positively associated with microbial activities (e.g., total PLFAs), while the CH₄ uptake was tightly correlated with soil environments (i.e. ST, SM) and chemical properties (i.e. DOC, C:N ratio, NH₄⁺-N concentration), but not significantly related to the methanotrophic bacteria (i.e. 18:1ω7c). Whereas, soil N₂O emission was positively associated with nitrifying genes, but negatively correlated with denitrifying genes especially nosZ. Overall, our results suggested that soil CO₂ and N₂O emissions were directly dependent on microbial attributes, and soil CH₄ uptake was more directly related to edaphic properties rather than microbial attributes. Thus, different patterns of soil CO₂, CH₄, and N₂O emissions and associated controls following land use change provided novel insights into predicting the effects of afforestation on climate change mitigation outcomes.

How to cite: Chen, Q. and Cheng, X.: Differential response of soil CO2, CH4, and N2O emissions to edaphic properties and microbial attributes following afforestation in central China, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3098, https://doi.org/10.5194/egusphere-egu22-3098, 2022.