Introduction

Pristine peatlands—i.e., peatlands (where peat soil has at least 30% dry organic matter and a peat depth generally exceeding 40 cm1) that have thus far not been directly impacted by anthropogenic activities, such as draining for agriculture or forestry land uses2—are among the most valuable ecosystems on Earth1. These ecosystems are critical for preserving global biodiversity, providing safe drinking water, minimizing flood risk, sequestering carbon (C) and mitigating climate change3,4. Although peatlands cover only ~2.84% of the Earth’s land surface5, they store 500–700 Gt (1 Gt = 103 Mt) of soil C1, which contributes 21–47% of the global soil C storage (1500–2400 Gt C) and is equivalent to more than 50% of the C in the atmosphere (860 Gt C)6. Therefore, peatlands are recognized as critically important and long-term C sinks to support the Climate Action Sustainable Development Goal of the United Nations7,8. However, peatlands worldwide are vulnerable to drainage and climate drying, which can substantially lower their water table (WT) level3,8,9 and consequently stimulate soil respiration (SR) and cause widespread peat subsidence (PS)10,11,12. SR, the second largest C flux between terrestrial ecosystems and the atmosphere after net ecosystem production13, is the major pathway for releasing CO2 through the soil surface due to microbial decomposition of organic matter and litter (heterotrophic respiration: HR) and due to belowground vegetation respiration (autotrophic respiration: AR)14,15. SR plays a vital role in regulating atmospheric CO2 concentrations and climate warming13,15 and is mainly regulated by WT levels in peatlands10,11,16.

Drainage was the most important driver of WT decline and global pristine peatland shrinkage in the past century17,18. Almost 50% of the total European pristine peatland area (0.53 × 106 km2) has been lost since the 20th century due to drainage3,19. Approximately 11–13% of global pristine peatlands were lost18,20 due to drainage for agriculture21,22, forestry23,24, grassland expansion (to support livestock grazing25,26 and herbage production27,28) and peat extraction29,30. Furthermore, future climate drying is predicted to be widespread in the mid-latitudes and subtropics and has already impacted European peatlands8,31, potentially causing further drops in WT levels and impacting the SR and its components, HR and AR, of pristine peatlands32,33.

The drawdown of WT due to drainage and climate drying directly exposes the anoxic peat layer to oxygen and thus causes significantly higher SR from drained peatlands than from pristine peatlands10,11. Previous meta-studies exclusively reported significant decreases in the net ecosystem production of pristine peatlands due to global WT decline3,4,8,20, however, the responses of SR in pristine peatlands to global WT decline remain unknown. Furthermore, in situ measurements showed that large uncertainties existed in the responses of SR to WT decline, with increases28,29, decreases25,34 or no changes35,36 all observed. To date, the global patterns and controls in the responses of SR to WT decline in pristine peatlands remain largely unclear. Additionally, although it is well known that SR comprises HR and AR14,15, another open question is whether the changes in SR due to global WT decline were contributed by the changes in HR or AR or both, which remains elusive. Disentangling the relative contributions of HR and AR to SR is crucial for understanding soil C dynamics because the former decomposes C that may have accumulated over millennia in the soil, whereas the latter represents the respiration of C recently assimilated by plants37,38. Consequently, the lack of clarity on this issue further limits the understanding of peatland soil C biogeochemistry, as HR is used to appropriately evaluate peat C decomposition and the PS rate (Rps) owing to peat C oxidation12,39. WT decline has caused high Rps, which severely threatens the utilization of agriculturally drained peatlands if peat C is depleted due to oxidation11,12. The Rps is triggered by a combination of processes, such as physical compaction by heavy equipment or livestock trampling and shrinkage through the contraction of organic fibers when drying, consolidation by loss of water from pores in the peat, and oxidation due to the breakdown of peat C10,11,12. Therefore, disentangling the Rps from oxidation (i.e., HR) is crucial for the sustainable management of drained peatland, decelerating peat C loss, and mitigating soil CO2 release. However, the patterns of Rps by oxidation and associated total annual soil heterotrophic CO2 emissions from global drained peatlands are far from clear, underscoring the urgent need for assessments of the impacts of WT decline to inform conservation guidelines for pristine peatlands and sustainable land-use policies for drained peatlands, as well as to reduce soil C losses for climate change mitigation.

Here, we report on a global meta-analysis of the impacts of WT decline (due to drainage and climate drying) on SR and its components, HR and AR, and their controls, as well as on Rps due to oxidation from global pristine peatlands (Figs. 1–8, Table 1, Supplementary Figs. 1–14 and Table 1). The in situ SR observations (measured by the chamber method) in our study covered most peatlands globally except for tropical swamps in Central Africa and South America (Fig. 1a and Supplementary Fig. 1). In total, 386 paired (i.e., pristine vs. drained) observations (250 SR, 101 HR, and 35 AR) were collected from 63 different publications of studies conducted globally (Fig. 1b). Most observations targeted only one of the three efflux components. Only 35 paired observations measured SR, HR, and AR simultaneously (Supplementary Fig. 3). In addition, 485 Rps values and 485 proportions of peat oxidation to Rps (Supplementary Fig. 9), estimated using the empirical models developed in our study (Fig. 5), were used to map the global patterns of Rps due to oxidation and associated soil HR rates from global drained peatlands (Fig. 6). Our study aims to 1) reveal the responses of SR and its components, HR and AR, in pristine peatlands to the global WT decline; 2) clarify the influencing factors for the responses of SR and its components in pristine peatlands to the global WT decline; and 3) estimate the total annual SR and HR due to the drainage of previously pristine peatlands worldwide. We hypothesize that the net increases in SR were mainly due to the net increases in HR rather than in AR, as the declining WT in pristine peatlands directly exposes the anoxic peat layer to oxygen and consequently promotes peat C oxidation. We also hypothesize that drained peatlands are global hotspots for SR, as peatlands store immense amounts of soil C in terrestrial ecosystems1. Consistent with the hypotheses, we found that WT decline significantly stimulated SR from global pristine peatlands mainly through HR rather than AR, and resulted in widespread PS across global drained peatlands. Tropical pristine peatlands drained for agriculture and forestry functioned as the largest SR emitters, while boreal pristine peatlands drained for forestry served as the lowest emitters, mainly due to the significantly higher Rps in tropical peatlands drained for agriculture and forestry. The total annual SR from global drained peatlands contributed 4.6% (3.0–7.3%) to the global total annual anthropogenic CO2 emissions. Our study calls for conserving pristine peatlands and restoring degraded peatlands to reduce peat C loss and mitigate climate change while lowering PS to achieve multiple Sustainable Development Goals (e.g., combat climate change, halt and reverse land degradation and halt biodiversity loss) of the United Nations.

Table 1 Estimated peat subsidence rate (Rps) due to oxidation and associated soil C release from global drained peatlands.
Full size table
Fig. 1: Relative changes in soil respiration (SR) and its components, heterotrophic respiration (HR) and autotrophic respiration (AR), due to water table (WT) decline in global pristine peatlands.
figure 1

a Distribution of the in situ observations. b Frequency distributions of the relative changes in SR and HR due to WT decline (AR was not shown due to the scarcity of data points (n = 35)). c Relative changes in SR, HR, and AR due to WT decline. d Pearson correlation of the growing season HR and SR rates in pristine and drained boreal peatlands. e Pearson correlation of the annual HR and SR rates in pristine and drained tropical peatlands. No data were observed in pristine and drained temperate peatlands. Each data point for WT decline impacts on in situ SR and its components were classified on the basis of the driver of WT decline and duration of WT decline in a. The gray–blue regions in a indicate the global distributions of peatlands derived from PEATMAT5. The larger solid circles and horizontal bars in c denote the weighted means of relative changes and their 95% confidence intervals. The smaller circles and numbers in c indicate individual relative changes due to WT decline and the number of data points, respectively. The open circles and bars in d, e show the means and standard deviations for individual data points. The asterisks *, **, *** indicate significance at the levels of p < 0.05, 0.01 and 0.001, respectively, and n.s. indicates no significance (for details, see Meta-analysis in Methods).

Full size image

Results

Changes in SR and its components, HR and AR, due to water table decline

The meta-data were slightly right-skewed for the impacts of WT decline on SR and HR (Fig. 1b). Overall, WT decline significantly increased SR by 35% (n = 250) and increased HR by 44% (n = 101) but had no significant effect on AR (30%, n = 35) (Fig. 1c). This result was also supported by a separate meta-analysis of incubation observations that measured the SR from peat soil samples under WT decline (n = 21, Supplementary Fig. 2). The simultaneous measurements from 35 paired SR, HR and AR observations were further analyzed, and a similar trend of a significant decline in WT corresponding to significant increases in SR and HR while having no significant effect on AR was observed (Supplementary Fig. 3). Additionally, significant positive linear correlations (Pearson correlation coefficients of 0.76 and 0.75, respectively) were observed for the growing season or annual SR and HR rates in boreal and tropical pristine and drained peatlands (Fig. 1d, e), reflecting the similar dynamics of these two effluxes in response to WT decline. Recent annual measurements also revealed strong linear correlation between SR and HR from temperate to boreal pristine and drained peatlands40. These results collectively indicated that increases in SR from global pristine peatlands under WT decline were mainly due to HR rather than AR.

Subgroup analysis for SR and HR (which had larger sample sizes than AR) (Fig. 2a, b and Supplementary Fig. 4) also supported the finding that WT decline significantly stimulated SR mainly through HR rather than AR (Fig. 1c). Among the drivers of WT decline, significant increases in SR and HR were observed in all land-use types except for grassland and peat extraction, including agriculture (36%, n = 18; 78%, n = 15), agriculture-N (agricultural land use with nitrogen application, 32%, n = 34; 82%, n = 10), forestry (39%, n = 107; 24%, n = 42) and climate-induced drying of pristine peatlands (49%, n = 644; 27%, n = 30). With regard to the impact of the magnitude of the WT decline, a stronger decline (>30 cm) resulted in significantly higher SR (46%, n = 75) and HR (78%, n = 41) than a weaker decline (≤30 cm; 22%, n = 147 for SR; 26%, n = 31 for HR), showing that a much lower WT (Supplementary Fig. 5) stimulates more peat oxidation41,42. Considering the duration of WT decline, significant increases in SR and HR were observed when the period since WT decline implementation was longer than 10 years; otherwise, no significant stimulation effects were observed, particularly when the duration of WT decline was less than 1 year.

In terms of climate zone, the effect of WT decline on increasing SR was significantly higher in boreal areas (54%, n = 93) than in temperate (32%, n = 97) or tropical (31%, n = 60) areas. A similar trend was observed for HR, but this effect was much larger, particularly for boreal (123%, n = 36) and tropical areas (77%, n = 22), highlighting that climate regime functioned as a key mediator of the WT decline effects on SR and HR. We also observed that the impacts of WT decline on SR and HR from pristine peatlands relied upon peatland type and the SOC concentration of the pristine peatlands (Fig. 2a, b). Specifically, WT decline significantly increased SR and HR from bogs, fens and swamps, while it did not have a significant influence on SR from marshes. Significant stimulations of SR and HR were found in pristine peatlands with SOC contents ≥300 g kg–1.

Fig. 2: Relative changes in soil respiration (SR) and heterotrophic respiration (HR) under different conditions due to water table (WT) decline in pristine peatlands.
figure 2

a Relative changes in SR due to WT decline. b Relative changes in HR due to WT decline (AR was not shown due to the scarcity of data points (n = 35)). The larger solid circles and horizontal bars denote the weighted means of relative changes and their 95% confidence intervals. The smaller circles and the numbers indicate individual relative changes due to WT decline and numbers of data points, respectively. The asterisks *, **, *** indicate significance at the levels of p < 0.05, 0.01 and 0.001, respectively, and n.s. indicates no significance (for details, see Meta-analysis in Methods). Net water table declines greater and less than 30 cm are defined as deep and shallow declines, respectively, according to ref. 42.

Full size image

Changes in environmental variables due to water table decline

Overall, WT decline significantly decreased SWC (–29%, n = 56) and SOC concentration (0–30 cm, –10%, n = 88) while increasing Eh (242 mV, n = 15), Ts (4%, n = 88), BD (79%, n = 101), and concentrations of soil \({{{{\rm{NO}}}}}_{3}^{-}\) (325%, n = 44), MBN (17%, n = 4) and DTP (311%, n = 3) (Fig. 3). The decreases in SWC were significantly related to decreases in WTD (Supplementary Fig. 6). WT decline did not exert significant effects on the concentrations of SOC in deeper soil layers (30–100 cm: 11%, n = 22), TN (0–30: 7%, n = 125; 30–100 cm: –17%, n = 19), SP (18%, n = 23), soil \({{{{\rm{NH}}}}}_{4}^{+}\) (48%, n = 44), MBC (–58%, n = 9) and DOC (–31%, n = 11) (Fig. 3). In addition, WT decline did not significantly affect the coefficient of temperature sensitivity (Q10, 42%, n = 9), soil pH (–3%, n = 125) or BGB (2%, n = 26) (Fig. 3).

Fig. 3: Relative changes in environmental variables due to water table (WT) decline in pristine peatlands.
figure 3

Eh (soil redox potential), SWC (soil water content), Ts (soil temperature), O2 (oxygen), Q10 (temperature sensitivity coefficient), SOC (soil organic carbon concentration), TN (total nitrogen concentration), SP (soil phosphorus concentration), BD (soil bulk density), pH, clay content, \({{{{\rm{NH}}}}}_{4}^{+}\) (soil ammonium concentration), soil \({{{{\rm{NO}}}}}_{3}^{-}\) (soil nitrate concentration), MBC (microbial biomass carbon), MBN (microbial biomass nitrogen), DOC (dissolved organic carbon), DON (dissolved organic nitrogen), DTP (dissolved organic phosphorus) and BGB (belowground root biomass). The larger solid circles and horizontal bars denote the weighted means of relative changes and their 95% confidence intervals. The smaller circles and numbers indicate individual relative changes due to WT decline and the number of data points, respectively. The asterisks *, **, *** indicate significance at the levels of p < 0.05, 0.01 and 0.001, respectively, and n.s. indicates no significance (for details, see Meta-analysis in Methods). Eh was denoted as the net change (treatment – control, in mV) due to WT decline (for details, see Meta-analysis in Methods).

Full size image

Controls of the changes in SR and HR due to water table decline

Across the study sites, the increases in SR significantly linearly increased with the peat depth (p = 0.03) and SOC content of pristine peatlands (p < 0.001) due to WT decline (Fig. 4a, b). In addition, greater increases in SR were significantly correlated with greater decreases in WTD (p < 0.001), greater decreases in SWC (p = 0.02) and greater decreases in DOC contents (p = 0.02) (Fig. 4d, e). The changes in SR were not correlated with the drainage durations (Fig. 4c) and were not controlled by meteorological conditions such as the mean annual air temperature and precipitation (Supplementary Fig. 7).

Fig. 4: Relationships between the relative changes in soil respiration (SR) and environmental variables due to WT decline.
figure 4

Relationships between the relative changes in SR and the a peat depth of pristine peatlands, b soil organic C concentration of pristine peatlands, c WT decline duration, d net changes in water table depth, e relative changes in soil water content and f relative changes in dissolved organic carbon concentration due to WT decline. The gray area indicates the 95% confidence interval for the fitted regression curve. We note that the SR observations were classified based on the climate, peatland type, and driver of WT decline (i.e., land use).

Full size image

As there were fewer observations for HR, we only observed that the increases in HR due to WT decline were controlled by the SOC content of pristine peatlands (p = 0.08), changes in WTD (p = 0.03) and changes in MBC (p = 0.03) (Supplementary Fig. 8).

Peat subsidence and associated soil C emissions due to drainage

Drainage of pristine peatlands worldwide has led to widespread PS10,11,12. Our synthetic results showed that the Rps varied greatly, from 0.04 to 20.10 cm yr–1, across different drainage drivers (i.e., land uses) and climate zones, with significant and negative exponential trends associated with the number of years after drainage (Fig. 5a, b). It is well known that Rps is triggered by a combination of processes such as physical densification (e.g., consolidation by dehydration of the peat, compaction associated with trampling due to livestock grazing or farm machinery, and shrinkage through the contraction of organic fibers when drying) and oxidation due to the breakdown of peat organic matter10. Therefore, the proportion of peat oxidation should be extracted from the total Rps to accurately estimate the actual HR from drained peatlands12. Our synthetic results showed that the proportion of oxidation to Rps (Po) ranged from 2% to 95% across different climate zones and land uses, with significant and positive logarithmic trends associated with the number of years since drainage (Fig. 5c, d), consistent with previous observations across drained temperate and boreal peatlands12,41.

Fig. 5: Drainage duration controls peat subsidence rate (Rps) and the proportion of peat oxidation to Rps (Po).
figure 5

a, b Relationships of Rps with drainage duration (i.e., years since drainage) for drained peatlands across different climate zones and land uses. c, d Proportion of subsidence attributed to peat oxidation (Po) with drainage duration for drained peatlands across different climate zones and land uses. The fitted regression curves and their 95% confidence intervals are shown. The same Po was used for drained boreal and temperate peatlands under different land use due to a lack of sufficient observations. Similarly, the same Po was used for drained tropical peatlands under different land uses.

Full size image

Using the established equations between the Rps and drainage duration and Po and drainage duration (Fig. 5a–d) as well as the globally synthesized drainage duration measurements (n = 485) categorized by climate zone and land use (Supplementary Fig. 9a, see also Eqs. (6–11) in Methods), we estimated the spatial patterns of the Rps due to oxidation across global drained peatlands (Fig. 6a). On average, the Rps due to oxidation of drained peatlands varied greatly across climate zones and land uses, where the Rps due to oxidation was significantly higher in the drained tropical peatlands under different land uses (1.89–1.95 cm yr–1) than in temperate or boreal peatlands (0.14–0.82 cm yr–1, Table 1 and Supplementary Fig. 10c). The Rps due to oxidation exhibited significant differences among different land uses following the decreasing order of agriculture>grassland>forestry in both drained boreal and temperate peatlands. Two other important variables, SOC concentration and soil bulk density (BD), in pristine peatlands, together with the Rps due to oxidation, were used for calculating soil HR rates11 and showed large spatial variations worldwide (Fig. 6b, c). Overall, the SOC concentrations of pristine peatlands were significantly higher for tropical areas (mean with 95% CI, 483 (437–524) g kg–1, the same below) and boreal areas (473 (452–495) g kg–1) than for temperate areas (315 (269–362) g kg–1) (Table 1 and Supplementary Fig. 10a). Soil BD was significantly higher in the temperate (0.17 (0.13–0.22) g  cm–3) and tropical pristine peatlands (0.16 (0.13–0.22) g cm–3) than in the boreal pristine peatlands (0.11 (0.08–0.15) g cm–3) (Table 1 and Supplementary Fig. 10b). According to Eqs. (12, 13) (see Methods), on an areal basis, soil HR rates were lowest in forestry-drained boreal peatlands (729 (484–1054) kg C ha–1 yr–1), followed by forestry-drained temperate peatlands (991 (565–1593) kg C ha–1 yr–1), grassland-drained temperate (3149 (1841–4927) kg C ha–1 yr–1), boreal peatlands (3515 (2428–4905) kg C ha–1 yr–1), and agriculture-drained boreal (3992 (2626–5812) kg C ha–1 yr–1) and temperate peatlands (4319 (2400–7082) kg C ha–1 yr–1), and highest in drained tropical peatlands under different land uses (15,408 (10,865–22,264) to 15,975 (10,967–23,685) kg C ha–1 yr–1) (Table 1 and Fig. 6d).

Fig. 6: Peat subsidence rate (Rps) (cm yr–1) due to oxidation and associated soil heterotrophic CO2 emissions.
figure 6

Spatial variations in the a estimated Rps due to oxidation of drained peatlands, b soil organic carbon (SOC) concentration (g kg–1) of pristine peatlands, c soil bulk density (BD, g cm–3) of pristine peatlands, and d estimated soil heterotrophic CO2 emissions due to peat oxidation (kg C ha–1 yr–1) among global drained peatlands. For the calculation processes of Rps and HR, see the Methods. The gray-blue regions indicate the global distributions of peatlands derived from PEATMAT5.

Full size image

To test the reliability of our estimated HR rates, we collected 36 annual HR rates from global drained peatlands. We extracted drainage duration, climate zone, and land use information from these studies and then estimated the annual HR rates by using previously established Eqs. (6–13) (see Methods). We observed a significant linear trend between the estimated and in situ measured annual HR rates across different climate zones and land uses (Fig. 7). This finding suggested that the estimation methods used in our study could be applied for estimating HR rates from regional to global drained peatlands.

Fig. 7: Linear correlation between the estimated and measured annual soil heterotrophic respiration (HR) rates across climate zones and land uses for drained peatlands.
figure 7

The measured annual HR was synthesized from previous publications. The predicted annual HR was obtained using Eqs. (6–13) in the Methods. The climate zone, land use, and drainage duration information were directly extracted from these publications, which simultaneously reported the measured annual HR. For details related to the estimation of annual HR, see Eqs. (6–13) in the Methods.

Full size image

Taking HR rates due to peat oxidation and drained peatland areas together (categorized by climate zone and land use, see Methods), the estimated total annual mean soil HR values were 22 (95% CI: 10–42), 4 (3–6), 5 (1–11), 32 (9–77), 5 (2–10), 13 (5–29), 152 (103–228), 140 (77–250) and 31 (1–87) Mt C yr–1 for agriculture-, forestry- and grassland-drained peatlands in boreal, temperate and tropical climate zones, respectively, with contributions of approximately 5.4 (4.6–5.7), 1.0 (0.9–1.3), 1.1 (0.5–1.5), 7.8 (4.2–10.4), 1.3 (1.1–1.4), 3.3 (2.3–3.9), 37.7 (30.8–49.0), 34.8 (33.8–36.6), and 7.6 (0.4–11.7)% to global total annual soil HR (404 (210–740) Mt C yr–1, total drained area 499 × 103 km2) due to subsidence of drained peatlands (Table 1).

Using 35 simultaneously measured sets of SR, HR, and AR observations across pristine and drained peatlands, we observed that there were no significant differences in the relative contributions of HR to SR between the drained tropical (62%) and boreal (66%) peatlands (Table 1 and Supplementary Fig. 11b). After combining the area-weighted total HR due to drainage and the bootstrapped relative contribution of HR to SR in the different climate zones and land-use types (see Methods), the estimated total SR emissions from global drained peatlands were 645 (95% CI: 401–1025) Mt C yr–1 (Table 1) or 640 (95% CI: 428–953) Mt C yr–1 (Supplementary Table 1).

Discussion

Underlying mechanisms controlling SR and its components from global pristine peatlands due to water table decline

Our meta-analysis results demonstrated that WT decline due to drainage and climate-induced drying significantly promoted SR from global pristine peatlands, mainly through HR rather than AR (Fig. 8), resulting in positive feedback to climate warming. Recent studies also found that drainage significantly stimulated SR primarily through HR rather than AR from tropical pristine peatlands12,39. The significant net decrease in the WTD (overall weighted mean: –25 cm with 95% CI: –29 to –21 cm) enhanced soil aerobic conditions and increased soil redox potentials (Fig. 8 and Supplementary Fig. 5), consequently stimulating soil microbial activity and accelerating peat organic matter decomposition36,43. Similar improvements in soil aerobic conditions and soil redox potentials and consequently stimulations of HR from pristine peatlands due to WT decline were reported44,45. Our meta-regression analysis also showed significant linear increases in SR and HR with decreases in WTD (Fig. 4d and Supplementary Fig. 8b). This finding was further supported by the significant increases in SR and HR with higher pristine peatland SOC concentrations (Fig. 4b and Supplementary Fig. 10a) and with significant decreases in dissolved organic carbon and microbial biomass carbon contents (Fig. 4f and Supplementary Fig. 8c). Peat organic C functioned as the substrate for microbial decomposition11,33,46, with a consequence that the higher substrates were provided for microbial decomposition given the higher SOC concentrations in peatlands11,47. Meanwhile, easily decomposable peat organic C, such as DOC and MBC, was preferentially lost due to WT decline48,49. Moreover, the lowered WT significantly decreased the soil water content and increased the soil temperature (Figs. 3 and 8), further stimulating the activities of soil enzymes and microbes in decomposing soil organic matter50,51, resulting in significant reductions in the topsoil (0–30 cm) SOC concentrations of pristine peatlands, as observed in our study (Fig. 3). This finding was verified by the observed significant linear increase in SR with decreasing soil water content (Fig. 4e). In addition, the decreased WT did not result in a change in belowground root biomass (Figs. 3 and 8) and thus did not exert a significant influence on AR, as roots are the main source of AR15,33. This explanation was consistent with recent findings that drainage exhibited nonsignificant impacts on AR across boreal to tropical pristine peatlands39,52.

Fig. 8: Conceptual model describing the effects of WT decline on soil carbon emissions and peat subsidence from global pristine peatlands.
figure 8

SR (soil respiration), HR (heterotrophic respiration), AR (autotrophic respiration), WTD (water table depth), Ts (soil temperature), SWC (soil water content), O2 (soil oxygen content), Fe3+ (soil trivalent iron concentration), \({{{{\rm{SO}}}}}_{4}^{2-}\,({{{\rm{soil}}}}\,{{{\rm{sulfate}}}}\,{{{\rm{concentration}}}})\), \({{{{\rm{NO}}}}}_{3}^{-}\) (soil nitrate concentration), Eh (soil redox potential), BGB (belowground biomass), SOC (soil organic carbon concentration), TN (total nitrogen concentration) and SP (soil phosphorus concentration). The asterisks *, **, *** indicate significance at the levels of p < 0.05, 0.01 and 0.001, respectively, and n.s. indicates no significance (for details, see Meta-analysis in Methods). The total SR and HR are presented as the means with 95% confidence intervals (for estimation details, see Methods).

Full size image

Although WT decline significantly promoted SR and HR from pristine peatlands, the magnitudes of the relative changes were also regulated by factors such as the driver and intensity of WT decline, climate zone, and peatland type. Decreases in WT due to agriculture were more extensive than those due to other land uses because the much lower WT could provide more soil aerations for crop growth53, dropping to more than two meters belowground (Supplementary Figs. 5 and 12) and thus promoting SR and in particular, HR (Fig. 2). This promotion trend was much more obvious in boreal and tropical peatlands (such as bogs, fens, swamps) than in temperate peatlands (such as marshes), as boreal and tropical peatlands exhibited significantly higher peat SOC concentrations and peat depths than temperate peatlands (Supplementary Figs. 10a and 13a, b). This result was further verified by the finding that WT declines with larger magnitudes stimulated higher emissions from SR, particularly from HR in pristine peatlands with higher SOC concentrations, than WT declines with smaller magnitudes (Fig. 2 and Supplementary Fig. 13). Moreover, agricultural tillage practices that loosen soils further stimulated peat organic matter decomposition16,28. The WT decreases associated with grasslands showed nonsignificant effects on SR and HR, probably because management practices such as grazing and harvesting, which cause trampling, decreased soil porosity and consequently inhibited SR and HR54,55, thus counteracting the stimulatory effects of WT decline on SR and HR. Additionally, consumption by grazing animals and biomass harvesting reduced litter input28 and inhibited microbial growth and activity56 and further reduced soil substrates for microbes34,53. Only three sites with peat extraction were involved in our meta-analysis; thus, there are large uncertainties in the results associated with this land-use type, and more research is warranted. The significant increases in SR and HR associated with WT declines due to climate-induced drying suggested that climate change has already impacted C cycles in peatlands4,8,31. Our study suggests that climate change mitigation actions, as well as minimizing human disturbances, are crucial for the conservation of pristine peatlands and reductions in soil CO2 release7,11,16.

Net emissions of SR and HR from global pristine peatlands due to drainage

Drainage promoted peat organic matter decomposition, as observed by significant increases in HR in our study, and consequently, peat oxidation was considered the overriding contributor to the widespread PS in the drained peatlands10,11,57. The highest Rps due to oxidation was observed in the agriculture-drained tropical peatlands, followed by forestry- and grassland-drained tropical peatlands, and the lowest Rps due to oxidation was observed in the forestry-drained boreal peatlands (Table 1). The warm and humid conditions in tropical regions make them more favorable to peat oxidation when pristine peatlands are drained for agriculture, grassland, or forestry uses than pristine peatlands in temperate and boreal cold environments12,58. This finding demonstrated that the Rps due to oxidation was largely dependent on climate zones and land uses, and therefore, these factors should be considered simultaneously when using this parameter for estimating soil HR from global drained peatlands.

In this study, the estimated total annual soil HR in global drained peatlands was 404 Mt C yr–1 (95% CI: 210–740 Mt C yr–1) (Table 1). Additionally, after aggregating the Rps due to oxidation in the boreal and temperate drained peatlands under each land use (see Methods, Supplementary Fig. 14), the estimated total annual soil HR in global drained peatlands was 400 Mt C yr–1 (95% CI: 226–688 Mt C yr–1) (Supplementary Table 1). The results of these estimates were very similar, indicating that our estimates were robust. Nonetheless, our results suggested that peat oxidation due to drainage was a large source for soil HR, consistent with previous estimates from global drained peatlands (520 Mt C yr–1, area 509 × 103 km2, ref. 18, 355. Mt C yr–1, area 460 × 103 km2, ref. 19).

Previous studies18,19 estimated only global HR due to drainage, with the consequence that the global SR due to drainage remained unknown. Our study revealed that drained peatlands were global hotspots for SR, with emissions of 645 (95% CI: 401–1025) Mt C yr–1 (i.e., 2.36 (1.47–3.76) Gt CO2 yr–1) (Table 1) or 640 (95% CI: 428–953) Mt C yr–1 (i.e., 2.35 (1.57–3.49) Gt CO2 yr–1) (Supplementary Table 1). These estimates are equivalent to ~14% (8.7–22%) of the total annual CO2 emissions from all land use changes and contributed 4.6% (3.0–7.3%) to global annual anthropogenic CO2 emissions (i.e., total emissions from land use change and fossil fuel burning) from 2010 to 20196. Considering the range of the remaining cumulative CO2 emissions (920–1980 Gt CO2-eq. (including methane and nitrous oxide)) associated with limiting warming to 1.5 °C (compared to the preindustrial era) by the end of this century59, our study calls for conserving pristine peatlands and restoring drained peatlands (by rewetting) to reduce peat carbon losses. The potential for enhancing methane emissions through rewetting does not negate the climate change mitigation potential of rewetting degraded peatlands3,8,20.

Our study also showed that agriculture- and forestry-drained tropical peatlands and agriculture-drained temperate peatlands were the top three emitters of SR, occupying half of global drained peatlands while accounting for ~81% (75–91%) of the total annual SR from global drained peatlands, suggesting that these drained peatlands should be given priority for restoration to reduce peat C loss and mitigate SR effluxes.

Limitations and the way forward

Our meta-analysis included only 35 paired simultaneous measurements of SR, HR, and AR from global pristine and drained peatlands; moreover, these measurements were mainly performed in tropical and boreal peatlands. This limitation probably resulted in some uncertainties in our results showing that decreases in WT due to drainage and climate-induced drying significantly promoted pristine peatland SR, mainly through HR rather than AR, and consequently caused widespread peat subsidence. Performing more simultaneous measurements of soil SR, HR and AR will be beneficial for more reliably assessing the impacts of WT decline on pristine peatland soil carbon effluxes and their feedback to climate change. Additionally, in our study, we were unable to obtain empirical models of Rps or Rps due to oxidation with drainage duration for drained boreal or temperate peatlands, respectively, as we currently lack sufficient measurements for these climate zones. Accordingly, we used two methods (i.e., climate was categorized into boreal, temperate, and tropical or boreal+temperate and tropical types) to constrain the uncertainties in the estimated annual SR and HR from global drained peatlands. Similar results were obtained from these two methods, and these results were in the ranges of the previous two estimates from global drained peatlands. Nevertheless, more observations of these variables should be conducted to support more reliable estimates of the annual SR and HR from drained peatlands. Finally, we did not estimate the annual SR and HR from peatlands affected by peat extraction and climate-induced drying, as the extent of the impact (e.g., the net decreases in WTD, impacted areas and their distributions) remains unknown, underscoring the necessity of considering these potentially important sources of changes in future studies.

Conclusions

Our meta-analysis revealed that water table decline due to drainage and climate-induced drying resulted in significant increases in soil respiration from pristine peatlands, mainly through heterotrophic respiration rather than autotrophic respiration, and consequently intensified positive climate-carbon feedback and induced widespread global peat subsidence. Considerable amounts of peat organic matter were lost in the form of soil CO2 emissions from global drained peatlands, contributing almost 14% of the total annual anthropogenic CO2 emissions from all land use changes or 4.6% of the total annual anthropogenic CO2 emissions from fossil fuel burning combined with land use changes. Therefore, urgent measures, such as conserving global pristine peatlands and restoring degraded peatlands particularly agriculture- and forestry-drained tropical peatlands and agriculture-drained temperate peatlands, are crucial for reducing peat carbon loss and mitigating soil CO2 emissions.

Methods

Systematic review

We searched relevant publications through Web of Science (all databases), Google Scholar, and the China National Knowledge Infrastructure Database between 1945 and March 2021 with the following combinations of keywords: (drain* OR lower* water table OR standing water depth OR ground water table level drawdown OR decline OR drought OR dry*) with (peatland* OR mire* OR fen OR bog OR swamp OR marsh*) with (soil respiration OR heterotrophic respiration OR microbial respiration OR soil CO2 OR soil carbon decompos* OR soil carbon minerali* or peat subsidence). Using these search terms, we initially identified 2120 different publications. To reliably evaluate WT decline impacts on SR and peat subsidence-associated soil CO2 emissions, the following further criteria were applied:

1) Only paired studies with pristine peatland (i.e., undrained, near-natural peatland without direct drainage history) as a control and pristine peatland with direct WT decline (due to drainage and land use or climate-induced drying) as a treatment were included by carefully checking the descriptions of field conditions from the publications. For the pristine peatlands, we included the peatland only if the peat soil had at least 30% dry organic matter, a peat depth of >40 cm1, and did not have any direct drainage history2. We acknowledge that few, if any, untouched and completely pristine peatlands currently exist, particularly in Europe.

2) WT decline in peatlands referred to only the WT depth lowered by drainage or climate-induced drying and/or additional management practices related to C or N input (e.g., manure/N fertilizers); treatments in which WT decline was combined with manipulated warming, elevated CO2, N deposition, etc., were excluded, while individual treatments (i.e., peatlands affected by WT decline without additional warming, elevated CO2, N deposition treatments, etc.) were included, as the primary objective of this study was to evaluate the responses of peatland C decomposition to WT decline.

3) Each individual study included SR or at least one of its components (HR and AR), and the measurement intervals were at least monthly. The in situ measurements of SR or its components (HR and AR) covered at least the growing or nongrowing season in temperate/boreal climate zones and the whole wet or dry season in (sub)tropical climate zones.

4) Both in situ and soil core/microcosm/mesocosm measurements of SR or its components (HR and AR) were included. SR and its components were exclusively measured using the chamber method. The results of the latter group were used to test the results of the former.

Finally, 386 paired in situ and 21 paired soil core incubation measurements of SR or its components (HR and AR) were extracted from 63 in situ studies and 9 soil core studies, respectively (see Supplementary Data A). Furthermore, to estimate HR emissions from global drained peatlands, the in situ measured paired peat subsidence rate (Rps, cm yr–1) and drainage duration (i.e., years since first drainage) and the proportion of peat subsidence rate attributed to oxidation (Po, %) and drainage duration, as well as the soil (0–30 cm) organic C and bulk density in pristine peatlands, were extracted from peer-reviewed publications. In drained boreal and temperate peatlands, most studies measured the total subsidence (in meter) during a certain drainage period, therefore the average Rps was calculated as the ratio of total subsidence and drainage years. It was assumed that the Rps was faster at the beginning and lower at the end of drainage duration, so the average subsidence rate is the rate for the middle year of the drainage duration41. The remaining studies directly showed the in situ measured Rps at the ith year of drainage. A similar procedure was applied for the Po in the ith year of drainage. In sum, 230 paired Rps–drainage duration observations and 49 paired Po–drainage duration observations, as well as 76 SOC and 63 BD in pristine peatlands, were taken from 80, 25, 58, and 44 studies, respectively (see Supplementary Data B).

Data compilation

To systematically evaluate the impacts of WT decline on SR in pristine peatlands and clarify the underlying mechanisms, we obtained data related to SR and its components (HR and AR) together with environmental variables such as the mean annual temperature [MAT], mean annual precipitation [MAP], peat depth [PD], WT depth [WTD], soil water content [SWC], soil temperature [Ts], soil redox potential [Eh], soil air oxygen level [O2], soil bulk density [BD], soil pH [pH], soil organic carbon [SOC], soil total nitrogen [TN], soil total phosphorus [TP], soil ammonium [\({{{{\rm{NH}}}}}_{4}^{+}\)], soil nitrate [\({{{{\rm{NO}}}}}_{3}^{-}\)], soil dissolved organic carbon [DOC], microbial biomass carbon [MBC], microbial biomass nitrogen [MBN], dissolved total phosphorus [DTP], belowground biomass [BGB], iron [Fe3+, Fe2+] and sulfate [\({{{{\rm{SO}}}}}_{4}^{2-}\)] when possible. If available, other important information, such as geographic location (latitude, longitude), climate and WT decline driver and duration, intensity, peatland type, Rps, Po, nutrient type, inundated condition, microtopography, and plant functional types, was recorded. For WT decline intensity, net WT declines greater and less than 30 cm were defined as deep and shallow declines, respectively, according to the IPCC wetland report42. The abovementioned information about pristine peatlands and peatlands affected by WT decline is compiled in Supplementary Data A and B.

We subsequently extracted the mean (\(\bar{X}\)), standard deviation (SD) and replicates (n) from different publications. If studies reported standard error (SE) rather than SD, then SD was calculated by SE \(\sqrt{n}\). If studies reported only the median, maximum, minimum, and 25th and 75th percentiles, then the mean and SD were estimated following the mathematical equations recommended by ref. 60. If neither SD nor SE was reported, then the missing SD was estimated by multiplying the reported mean by the average coefficient of variation (CV) obtained from the remaining observations, resulting in both the mean and SD being reported61. The data were either obtained directly from tables and texts or extracted by digitizing graphs using Getdata Graph Digitizer software (version 2.26, Russia).

The final database consisted of 250 paired SR, 101 paired HR and 35 paired AR in situ observations. Only 35 paired observations simultaneously reported SR, HR, and AR. Twenty-one paired SR soil core incubation measurements were also collected to test the results of the in situ measurements. The dataset mainly originated from Europe, North America, and Southeast Asia, and most studies (>70%) were conducted in temperate and boreal peatlands in the Northern Hemisphere (Fig. 1a). Moreover, 230 paired Rps–drainage duration observations and 49 paired Po–drainage duration observations (Fig. 5a, b) and an additional 485 drainage year (Supplementary Fig. 9) observations classified by climate zone (i.e., boreal, temperate and tropical) and land use (i.e., agriculture, forestry, and grassland) were collected. A total of 76 SOC and 63 BD measurements from pristine peatlands categorized by climate zone (i.e., boreal, temperate, and tropical) were extracted to estimate Rps by oxidation and associated soil HR from global pristine peatlands due to drainage activities (Supplementary Fig. 10 and Supplementary Data B). In this study, we were unable to estimate climate drying-induced net CO2 emissions through soil HR, as the areas of pristine peatlands affected by climate drying currently remain unknown.

Meta-analysis

To assess the relative changes in SR and its components (HR and AR), as well as environmental variables (e.g., SOC, BD, Ts, etc.) due to WT decline, the log-transformed response ratio (RR) was used:62

$${{{\mathrm{ln}}}}({{{\rm{RR}}}})=\,{{{\mathrm{ln}}}}({X}_{{{{\rm{t}}}}}/{X}_{{{{\rm{c}}}}})$$
(1)

The results are presented as the percent change ((elnRR  – 1) × 100). The variance (v) of RR was estimated using the following equation:

$$v=\frac{{{{{\rm{SD}}}}}_{{{{\rm{t}}}}}^{2}}{{n}_{{{{\rm{t}}}}}\,{X}_{{{{\rm{t}}}}}^{2}}+\frac{{{{{\rm{SD}}}}}_{{{{\rm{c}}}}}^{2}}{{n}_{{{{\rm{c}}}}}\,{X}_{{{{\rm{c}}}}}^{2}}$$
(2)

where Xt and Xc indicate the means of the treatment and control, SDt and SDc indicate the SDs of the treatment and control and nt and nc indicate the numbers of replicates in the treatment and control, respectively.

However, in our study, approximately 60% of the WTD and Eh observations for the peatlands in pristine condition (control) and affected by WT decline (treatment) showed opposite signs; e.g., the pristine peatlands generally exhibited positive WTDs (higher than the peat surface) and negative Eh values, while those affected by WT decline exhibited negative WTDs (lower than the peat surface) and positive Eh values. Since it is impossible to calculate the logarithm of negative values, we introduced a new study index (net changes) for these two variables in our meta-analysis according to ref. 63:

$$D={X}_{{{{\rm{t}}}}}-{X}_{{{{\rm{c}}}}}$$
(3)

where Xt and Xc indicate the paired annual mean WTD and Eh for the treatment and control, respectively, and D indicates the difference between the treatment and control.

The SD and variance (v) of D were estimated using the following equation:

$${{{\rm{SD}}}}=\sqrt{\frac{({n}_{{{{\rm{c}}}}}-1)\;{{{{\rm{SD}}}}}_{{{{\rm{c}}}}}^{2}+({n}_{{{{\rm{t}}}}}-1)\;{{{{\rm{SD}}}}}_{{{{\rm{t}}}}}^{2}}{{n}_{{{{\rm{c}}}}}+{n}_{{{{\rm{t}}}}}-2}}$$
(4)
$$v=\frac{{{{{\rm{SD}}}}}_{{{{\rm{t}}}}}^{2}}{{n}_{{{{\rm{t}}}}}}+\frac{{{{{\rm{SD}}}}}_{{{{\rm{c}}}}}^{2}}{{n}_{{{{\rm{c}}}}}}$$
(5)

where SDt and SDc indicate the SD of the treatment and control and nt and nc indicate the number of replicates for the treatment and control, respectively.

The weighted mean RR or D was calculated by individual RR or D with bias-corrected 95% confidence intervals (CIs) using the rma.mv function in the metafor package in R software (R core team, 2019)64, in which the variable “study” was regarded as a random effect to account for the dependence of observations derived from the same study. The impact of WT decline on a response variable was considered significant if the 95% CI did not overlap 065. Differences between subgroups (e.g., WT decline driver, climate zone, drainage duration) were considered significant if the 95% CIs did not overlap each other65. The frequency distribution of RR was calculated to test variability among individual studies using the Gaussian function (i.e., normal distribution)66.

Estimation of peat subsidence rate by oxidation and associated HR rate

Drainage has induced widespread peat subsidence and associated large CO2 release through soil HR and consequently reduced the sustainable utilization of drained peatlands and contributed to global warming11,12. In this study, we estimated the spatial patterns of Rps by oxidation and associated soil HR from global drained peatlands. Using the 230 paired Rps and drainage duration observations synthesized in this study, we first constructed empirical models between Rps and drainage duration for drained peatlands categorized by climate zone (boreal, temperate and tropical climate) and land use (i.e., agriculture, forestry and grassland) (Fig. 5a, b). The values of Rps for certain groups classified by climate zone and land use could be estimated by using the corresponding empirical models established in this study and reported drainage durations that were extracted from the literature. The empirical models categorized by climate zone and land use are listed below (Fig. 5a, b):

$${R}_{{{{\rm{ps}}}}}{\mbox{-}}{{{\rm{Bor}}}}{\mbox{-}}{{{\rm{Tem}}}}{\mbox{-}}{{{\rm{Agr}}}}=13.95\,{{{{\rm{Dur}}}}}^{-0.58},\,n=48,\,{R}_{{{{\rm{adj}}}}.}^{2}=0.85,\,p\; < \; 0.0001$$
(6)
$${R}_{{{{\rm{ps}}}}}{\mbox{-}}{{{\rm{Bor}}}}{\mbox{-}}{{{\rm{Tem}}}}{\mbox{-}}{{{\rm{For}}}}=5.36\,{{{{\rm{Dur}}}}}^{-0.83},\,n=21,\,{R}_{{{{\rm{adj}}}}.}^{2}=0.92,\,p\; < \; 0.0001$$
(7)
$${R}_{{{{\rm{ps}}}}}{\mbox{-}}{{{\rm{Bor}}}}{\mbox{-}}{{{\rm{Tem}}}}{\mbox{-}}{{{\rm{Gra}}}}=5.55\,{{{{\rm{Dur}}}}}^{-0.36},\,n=40,\,{R}_{{{{\rm{adj}}}}.}^{2}=0.61,\,p\; < \; 0.0001$$
(8)
$${R}_{{{{\rm{ps}}}}}{\mbox{-}}{{{\rm{Tro}}}}{\mbox{-}}{{{\rm{Agr}}}}{\mbox{-}}{{{\rm{For}}}}{\mbox{-}}{{{\rm{Gra}}}}=6.63\,{{{{\rm{Dur}}}}}^{-0.37},\,n=121,\,{R}_{{{{\rm{adj}}}}.}^{2}=0.55,\,p\; < \; 0.0001$$
(9)

where Rps indicates the peat subsidence rate (cm yr–1), Dur is the drainage duration, and the numbers indicate coefficients for the established empirical models. Bor, Tem, and Tro indicate boreal, temperate, and tropical climate zones, respectively. Agr, For, and Gra represent agriculture, forestry, and grassland land uses, respectively. We note that it was not possible to further distinguish these models between boreal and temperate climate zones and among agriculture, forestry, or grassland land use in tropical climates, as there is currently a lack of sufficient measurements, which warrants more research.

However, the Rps is triggered by a combination of processes such as physical compaction by heavy equipment or livestock trampling and shrinkage through contraction of organic fibers when drying, consolidation by loss of water from pores in the peat and oxidation owing to the breakdown of peat organic matter10,11,12. Therefore, to reliably estimate the soil HR rate from Rps due to oxidation, the proportion of Rps attributed to oxidation (Po, in %) should be considered12. Using the 49 paired Po and drainage duration observations synthesized in this study, we then constructed empirical models between Po and drainage duration for drained peatlands that were also categorized by climate zone (boreal, temperate, and tropical climate) and land use (agriculture, forestry, and grassland) (Fig. 5c, d). Similarly, the Po values of certain groups classified by climate zone and land use could be estimated by using the corresponding empirical models established in this study and reported drainage durations that were extracted from the literature. The empirical models categorized by climate zone and land use are shown below (Fig. 5c, d):

$$ {P}_{{{{\rm{o}}}}}{\mbox{-}}{{{\rm{Tem}}}}{\mbox{-}}{{{\rm{Bor}}}}{\mbox{-}}{{{\rm{Agr}}}}{\mbox{-}}{{{\rm{For}}}}{\mbox{-}}{{{\rm{Gra}}}}=12.05\,{{{\mathrm{Ln}}}}({{{\rm{Dur}}}})+2.15,\,n=30,\\ {R}_{{{{\rm{adj}}}}.}^{2}=0.89,\,p\; < \;0.0001$$
(10)
$$ {P}_{{{{\rm{o}}}}}{\mbox{-}}{{{\rm{Tro}}}}{\mbox{-}}{{{\rm{Agr}}}}{\mbox{-}}{{{\rm{For}}}}{\mbox{-}}{{{\rm{Gra}}}}=14.36\,{{{\mathrm{Ln}}}}({{{\rm{Dur}}}})+37.05,\,n=19,\\ {R}_{{{{\rm{adj}}}}.}^{2}=0.81,\,p\; < \;0.0001$$
(11)

where Po indicates the proportion of Rps attributable to oxidation, Dur is the drainage duration, and the numbers indicate coefficients for the established empirical models. The abbreviations Bor, Tem, Tro, Agr, For, and Gra have been described previously. We note that the different land uses shared the same models across temperate and boreal climates and tropical climate due to a lack of sufficient global observations. This will also induce some uncertainties in our analysis.

Furthermore, the soil HR (FHR, Mt C yr−1) due to peat oxidation induced by drainage was estimated using the following equation according to ref. 11:

$${F}_{{{{\rm{HR}}}}}=\sum {R}_{{{{\rm{ps}}}},i,j}\times {P}_{{{{\rm{o}}}},i,j}\times {{{{\rm{SOC}}}}}_{i}\times {{{{\rm{BD}}}}}_{i}\times {A}_{i,j}$$
(12)

where SOC (g kg–1) and BD (g cm–3) indicate the soil (0–30 cm) organic C concentration and bulk density of pristine peatlands, respectively; A (×103 km2) indicates the drained peatland area; i indicates the climate zone (boreal, temperate or tropical); j indicates the land use (agriculture, forestry or grassland); and Rps (cm yr–1) and Po (%) are described in Eqs. (6–11). Datasets of the SOC concentration and BD and Rps due to oxidation were systematically reviewed and bootstrapped and categorized by climate zones and land uses (see Supplementary Fig. 10 and Supplementary Data B). Regarding the large uncertainties for areas of drained peatlands, we combined two previously published datasets (72, 61, 22, 37, 43, 26, 94, 109, and 39 × 103 km2 by ref. 18, and 37, 55, 4, 109, 63, 58, 96, 72, and 1 × 103 km2 by ref. 20. for agriculture-, forestry- and grassland-drained peatlands in boreal, temperate and tropical climate zones, respectively) and obtained their mean values with 95% CIs (for details, see bootstrapping procedure in Data analysis). Uncertainties (i.e., 95% CI) in total HR (δFHR) were propagated according to the Gaussian random error propagation principle as follows:

$${{{\rm{\delta }}}}{F}_{{{{\rm{HR}}}}}=\sqrt{\sum \sqrt{\begin{array}{c}{(\delta {R}_{{{{\rm{ps}}}},i,j})}^{2}\times {({P}_{{{{\rm{o}}}},i,j}\times {{{{\rm{SOC}}}}}_{i}\times {{{{\rm{BD}}}}}_{i}\times {A}_{i,j})}^{2}+\\ {(\delta {P}_{{{{\rm{o}}}},i,j})}^{2}\times {({R}_{{{{\rm{ps}}}},i,j}\times {{{{\rm{SOC}}}}}_{i}\times {{{{\rm{BD}}}}}_{i}\times {A}_{i,j})}^{2}+\\ {(\delta {{{{\rm{SOC}}}}}_{i})}^{2}\times {({R}_{{{{\rm{ps}}}},i,j}\times {P}_{{{{\rm{o}}}},i,j}\times {{{{\rm{BD}}}}}_{i}\times {A}_{i,j})}^{2}+\\ {(\delta {{{{\rm{BD}}}}}_{i})}^{2}\times {({R}_{{{{\rm{ps}}}},i,j}\times {P}_{{{{\rm{o}}}},i,j}\times {{{{\rm{SOC}}}}}_{i}\times {A}_{i,j})}^{2}+\\ {(\delta {A}_{i,j})}^{2}\times {({R}_{{{{\rm{ps}}}},i,j}\times {P}_{{{{\rm{o}}}},i,j}\times {{{{\rm{SOC}}}}}_{i}\times {{{{\rm{BD}}}}}_{i})}^{2}\end{array}}}$$
(13)

where δFHR, δRps, δPo, δSOC, δBD, and δA indicate the 95% CIs of total soil HR, Rps, Po, SOC, and BD and drained peatland area, respectively, and i and j indicate the climate zone (boreal, temperate, tropical) and land use (agriculture, forestry, or grassland), respectively.

To further estimate the total SR (FSR, Mt C yr−1) and its uncertainty (δFSR) from global drained peatlands, the following equations were used:

$${F}_{{{{\rm{SR}}}}}=\sum \frac{{F}_{{{{\rm{HR}}}},i,j}}{{C}_{{{{\rm{HR}}}},i,j}}$$
(14)
$$\delta {F}_{{{{\rm{SR}}}}}=\sqrt{\sum \sqrt{{(\frac{1}{{C}_{{{{\rm{HR}}}},i,j}})}^{2}\times \delta {F}_{{{{\rm{HR}}}},i,j}^{2}+{(-\frac{{F}_{{{{\rm{HR}}}},i,j}}{{C}_{{{{\rm{HR}}}},i,j}^{2}})}^{2}\times \delta {C}_{{{{\rm{HR}}}},i,j}^{2}}}$$
(15)

where CHR (%) indicates the mean relative contribution of HR to SR from simultaneously measured SR, HR, and AR from our meta-analysis (see Supplementary Fig. 11 and Supplementary Data A) and i and j indicate the climate zone (boreal, temperate, tropical) and land use (agriculture, forestry, or grassland), respectively. FHR and δFHR are given in Eqs. (12, 13). We note that the CHR could be classified only by climate zone, as there is a lack of sufficient measurements of land use; that is, the different land uses under the same climate shared the same CHR value, which may induce uncertainties in estimating the total SR from global drained peatlands.

Regarding the abovementioned lack of sufficient measurements for distinguishing between boreal and temperate drained peatlands, we also used another method to estimate the annual total HR and SR from global drained peatlands. Specifically, we obtained the mean values of Rps by oxidation across boreal and temperate drained peatlands for each land use (i.e., climate zones were classified as boreal+temperate or tropical) (Supplementary Fig. 14 and Supplementary Table 1). The estimation process was the same as that previously described. The different estimation methods were likely to provide results with greater convergence.

Data analysis

Significant differences in observed variables were tested by performing nonparametric analysis. Specifically, tests with two independent samples (i.e., Mann–Whitney U test) were used for only two variables (i.e., to compare the contribution of HR to SR between pristine and drained peatlands), and tests with two or more independent samples (i.e., Kruskal–Wallis test and pairwise comparisons) were used if there were three or more variables (i.e., SOC, BD and Rps due to oxidation in the boreal, temperate and tropical climate zones or different land uses). Linear or nonlinear regression analysis was performed to examine the relationships between the responses of SR and its components with environmental variables or the peat subsidence rate with drainage duration.

To reliably estimate the uncertainties in Rps by oxidation, SOC, BD, drained peatland area, and relative contribution of HR to SR, bootstrap resampling with 10000 iterations was conducted using the boot package, and 95% CIs were calculated using the “basic” type. The ggplot 2 package in R software (R core team, 2019) was used for statistical analysis. Data were expressed as the means with their 95% CIs, and significance of the regression analyses was indicated at the level of p < 0.05 across the study.