Mangrove selective logging sustains biomass carbon recovery, soil carbon, and sediment

Our analysis suggests that, over 465 ha of mangrove area, almost 83% of aboveground tree biomass were harvested annually for commercial timber purposes, using a keyhole harvest pattern (Fig. S3b). Yet after 25 years of natural and human-induced regeneration, both field- and satellite-based assessments reveal that biomass carbon stocks and canopy cover had fully recovered. Our approach using space-for-time substitution indicates that manual selective logging did not significantly affect soil carbon stocks and rates of annual carbon burial. While the differences in soil carbon stock between sites may be due to the diverse hydro-geomorphic settings^8,14 the mangrove root mass in the top 1-m were not disturbed by manual logging activities. Similar situation was found in Tampa Bay, Florida where peat formation from root mass has enhance carbon sequestration¹⁵. These findings reduce uncertainty around the effects of mangrove forest management on the long-term functional capacity of blue carbon storage and provide evidence that managed mangrove ecosystems may deliver nature-based climate solutions.

Recovery of forest structure, canopy cover and species diversity

Along carbon stocks, forest structure and species diversity also demonstrated recovery (Fig. 2, Table S1). Seedling densities were significantly higher in 5 year-old mangrove plots than in plots at any other stage (F_(5,13) = 28.321, p < 0.001). There were no pairwise differences when comparisons were made among other stages of regeneration post-harvest. The younger stands also had a significant number of large trees with an average diameter of 15.9 cm, larger than those found in stands 25 years post-harvest (12.2 cm). This is due to the presence of the seed trees in the harvest area (Fig. 2a). Seedling densities were highest in the younger stands, which suggests that seed trees and the surrounding greenbelt function well in providing propagules. However, it was observed that individual seed trees seem to die within 10 years after harvesting⁴. This does not appear to significantly disturb the regeneration process, the dieback likely to be due to lack of structural support and extreme winds. A study to compare the contribution of seed tree-derived propagules versus that of greenbelt-derived propagules, would help further understand regeneration dynamics. A similar situation of high propagule production was observed in the Hawaiian archipelago, with Rhizophora mangle producing propagules at a rate of 11 ton ha⁻¹ yr⁻¹, and becoming invasive¹⁶.

Mean tree diameters significantly differed across stand ages; the largest mean diameter at breast height (DBH) seen in protected mangroves (17.30 ± 12.00 cm), followed by 5-year-old stands (15.9 ± 0.4 cm), 25-year-old stands (12.15 ± 6.11 cm), 15-year-old stands (9.9 ± 0.5 cm), and 10-year-old stands (7.90 ± 4.71 cm). The fact that 5-year-old stands had a larger mean DBH is due to higher numbers of seed trees with larger diameters, compared to older stands.

The 25-year-old stands had the largest basal areas (BA) (37.12 ± 9.21 m² ha⁻¹), followed by protected mangrove, 15-year-old and 5-year-old stands. The 10-year-old stand had the lowest BA (10.58 ± 7.93 m² ha⁻¹), even lower than that of the 5-year-old stand (16.8 ± 3.3 m² ha⁻¹). These differences indicate that growth in the protected forests has been levelling off or even declining after tree maturity is reached, following a cycle of senescence and regrowth (Fig. 2b). In recently logged forests, the landscape was dominated by seedlings, yet fewer were naturally recruited in later years. Existing seedlings showed relatively low survival rates (around 25%) after 10 years. Overall, the mean BA significantly differed across stand ages.

Rhizophora apiculata dominated logged-over mangroves across different stand ages, with relative frequencies over 70%, but tended to decrease with increasing stand age. The species abundance gradually increased over time (Fig. 2d). R. apiculata reached its lowest frequency (32.7%) in protected forest, close to Ceriops decandra (27.2%). Likewise, Bruguiera parviflora and B. gymnorrhiza were more frequently found in older forest plots, as the established forest structure creates more favorable conditions for these genera to thrive. Species diversity in Bintuni Bay is highly dependent on the stand’s age and tidal zone, some species being present in the fringe mangrove (e.g., Avicennia marina) and others in the interior mangrove (e.g., Ceriops tagal).

Based on time-series Landsat imagery, an estimated 8,361 ha of mangrove area were harvested during 2001–2018, at an average of 465 ha harvested every year. Annual harvesting trends are presented in Fig. S2. This average (465 ha yr⁻¹) is lower than the annual cutting projection for 2010–2020, of 800 ha yr⁻¹ (ref.¹⁷), indicating that selective logging practiced in Bintuni Bay took just half the allowable cut, with extracted timber amounting to 237 m³ ha⁻¹ (see Table S2). Such low logging intensity has significant potential to conserve biodiversity.

Using the same Landsat images, rapid recovery of canopy cover was also demonstrated qualitatively. As shown in Fig. 3, all stands show regreening, including after 5 years post-logging. This indicates that improvement of vegetation cover secures plots from possible erosion, hence, depletion of soil carbon stocks. This impact could be particularly significant for estuarine hydro-geomorphic settings or the fringe mangroves found in the 15-year-old stands.

Reduced impact logging and carbon stocks

Within the Bintuni mangroves undergoing selective logging, we demonstrate that 25 years post-harvest, biomass carbon stocks are close to that of intact or protected mangroves. Secondary forest stands show a recovery in aboveground biomass carbon pools, from 5-year-old stands (16.43 ± 27.16) Mg C ha⁻¹ to 25-year-old stands (114.52 ± 33.99) Mg C ha⁻¹ (Fig. 4, Table S3).

Biomass carbon of 25-year-old manually logged forest stands are similar to that of intact protected undisturbed ones (p = 0.814). Although recovery was gradual, it can be observed as an independent and unbiased measure of sustainably managed mangrove forests. The recovery of canopy cover across differently aged stands is also a measurable indicator that rotational and selective logging secures aboveground biomass gains.

The annual mean of extracted biomass between 2015 and 2018 was 190 Mg ha⁻¹ or approximately 95 Mg C ha⁻¹ (Table S2), which is around 83% of fully recovered biomass carbon (Fig. 4). This amount is equivalent to 192 Mg (dry matter) ha⁻¹, which is the same proxy for emission factors from aboveground biomass removal, to be used in assessing GHG emissions from managed mangroves when using IPCC’s Stock-Difference approach¹¹. It was also observed that dead wood carbon stocks decreased as the stands got older. This suggests that logging residues may have decomposed or shifted elsewhere. The amount of dead wood carbon was relatively large compared with carbon sequestered in the biomass, 39.73 ± 14.54 Mg C ha⁻¹ soon after logging and 10.76 ± 9.17 Mg C ha⁻¹, 25 years after logging. This means that wood harvest waste was at least 40%.

Top 100 cm soil carbon stocks did not differ significantly across post-harvest stands, only varying between 354 ± 71 Mg C ha⁻¹ in 10-year-old stands and 442 ± 57 Mg C ha⁻¹ in 15-year-old stands. Even in newly logged sites, top 100 cm soil carbon stocks were relatively high at 381 ± 68 Mg C ha⁻¹. Here, surface sediments were not significantly disturbed, and skidding trails were properly used for transporting the timber using local system called “Ongkak”. Skidding of logs on the wooden trails using “Ongkak” and loading them on the barges from the log yard at the end of “keyhole” shape logging plot has reduced soil surface disturbances. This suggests that reduced-impact logging preserves soil carbon stocks throughout the cutting cycle.

Deeper soil carbon stocks (i.e., at a depth of 100–300 cm) varied across all sites (Fig. 4, Table S3), confirming variation in carbon density reductions in deeper layers across all sites. This variation corresponds to different degrees of hydro-dynamic processes, under varied spatial and vertical geomorphology gradients^8,14. Looking at a standardized depth of 300 cm to give a full soil carbon pool profile, TECS measurements were significantly different across all sites (p < 0.001, Fig. 4, Table S3); the lowest stocks (665 ± 201 Mg C ha⁻¹) located in 10-year-old stands, and the largest stocks (1389 ± 199 Mg C ha⁻¹) found in protected forests. The fact that post-harvest forests had lower TECS measurements compared to protected forests, was mainly attributed to lower soil carbon density, driven by hydro-geomorphology variation across sampling sites⁸. In addition, wood harvesting practice by coppicing the trees and not followed by excavating soil like in ponds development allows the conservation of soil carbon in generally estuarine landscape. Similar situation was found across Cambodian degraded and deforested mangroves when compared with 25 years old protected stands¹⁸. For example, examining the same complete profiles, we observed larger carbon densities in protected forests than in harvested sites (Table S3). Our findings are consistent with the 2013 IPCC Wetlands supplement¹¹, suggesting that forest management in mangroves substantially affects biomass carbon pool, while soil pool remained unchanged.

The use of top-meter soil carbon stock pools as a standard soil profile from which to calculate coastal wetlands’ carbon storage, and emissions related to LULCC, is widely applied¹⁹ and suggested by the 2013 IPCC Wetlands Supplement for national-scale GHG reporting¹¹. While we acknowledge the substantial roles that hydro-geomorphic variations play in controlling soil carbon stocks, a direct adoption of the Tier 1 global average, without incorporating hydro-geomorphic information, may lead to significant uncertainty. In Bintuni’s protected mangroves, we observed the mean of top-meter soil carbon stocks to be 474 Mg C ha⁻¹, close to the IPCC Tier 1 value of 471 Mg C ha⁻¹ (ref.¹¹). Deeper soil sampling, however, revealed additional stocks of 776 Mg C ha⁻¹ located at a soil depth of 100–300 cm, increasing total soil carbon stock to 1250 Mg C ha⁻¹. This implies that soil sampling below 100 cm is necessary in mangroves, and particularly useful to calculate carbon losses from land conversion that excavates soil to depths of over 100 cm (e.g., aquaculture pond development). Overall, TECS measurements in Bintuni Bay’s protected mangroves are 60% higher than the global average (856 ± 32 Mg C ha⁻¹)¹⁰. Such high TECS measurements should attract the development of project-based activities for climate change mitigation.

Continued sediment accretion and carbon burial

There is a consistent increase pattern of historical and contemporary cumulative sediment accretion across hydro-geomorphic settings and stand ages (Fig. 5 and Table 1). Nevertheless, the rates of contemporary sediment accretion are higher compared to the historical ones, except in fringe protected mangrove site (Table 1). Historical sediment reconstruction results in net sediment accumulation, due to both biological (e.g., litterfall, benthic algae mat growth, root production and decomposition) and physical (e.g., erosion, compaction and groundwater shrink/swell) processes over the course of a decade²⁰. Under shorter observation period, high contemporary accretion in the surface layer suggest a larger net sediment accumulation with less sediment losses from decomposition, erosion, and compaction.

Direct measurements of contemporary sediment vertical accretion show that the younger the stands, the more sediment was accreted vertically (Fig. 5b). This may be associated with the large amount of deadwood materials left on the surface after logging event (see Table S3). These fresh woody debris could potentially trap more sediments and enhance accretion before they are decomposed and contribute to accretion as well. While bioturbation has been known as one of main limitations in coastal sedimentation studies, particularly those who used radionuclides sediment dating approach^21,22,23, here accretion may be additionally affected by logging event. We underline that such logging event may have caused surface sediment dynamics at substantial degree despite manual logging approach was implemented.

Cumulative sediment measured over 4.2 years reached up to 62 mm in 5-year-old stands, 38 mm in 15-year-old stands, and 8 mm in protected mangroves. Accretion was also greater in fringing mangroves than in interior mangroves, consistent with the historical sediment accretion rates. This indicates that the majority of suspended sediment from incoming high tides is initially deposited in fringing areas, where roots and trunks slow down the current’s velocity, and thus the energy required to settle sediments.

Table 1 shows that historical vertical accretion rates derived using the ²¹⁰Pb radionuclide technique ranged between 3 and 13 mm yr⁻¹, while contemporary vertical accretion rates derived from marker horizons (MH), ranged between 1 and 16 mm yr⁻¹. These rates were higher than the rate of global sea level rise, which was 3.2 (2.8–3.6) mm yr⁻¹ between 1993 and 2010²⁴. However, fringe mangroves are also reportedly prone to erosion²⁰. This implies that the hydro-geomorphic setting should be considered when conserving pristine mangroves or restoring degraded mangroves in areas subject to forestry practices, so as to support adaptation to climate-driven sea-level rise.

Based on carbon density estimates, using the soil cores, carbon burial rates were 0.1–13 Mg ha⁻¹ yr⁻¹ for MH, similar to the ²¹⁰Pb radionuclide technique results of 1–3 Mg ha⁻¹ yr⁻¹. This historical carbon burial rates are similar to that found in estuarine mangroves in Vietnam and Palau²⁵, and within the range of global numbers previously reported, 1.3–2.0 Mg ha⁻¹ yr⁻¹ (ref.²⁶). The wide range of sediment age in Table 1 suggests that even when logging operations have been underway for over 25 years, soil carbon remains secure, including in the most recently logged plots. If the ecosystem is well managed and disturbance is minimized, soil carbon can be stored for long periods. In protected mangroves, soil carbon is key to secure permanence.

Leaving seed trees and greenbelts supported rapid canopy recovery and sedimentation. The observed sediment accretion rates by MH suggests that there is a lot of sediment entering these systems supporting a greater net sediment accumulation, while the ²¹⁰Pb radioactive tracer suggests that this dynamic has been occurring for the last 100 years and results in a lower net sediment accumulation. Through efforts to refine and indicate the constraints of global estimates of organic carbon burial rates, it has been confirmed that site-specific measurements are key to reduce uncertainties^21,27. However, measurements may be extended to include mudflat hydrogeomorphic settings and identify organic matter sources, to capture the full fate of sedimentary carbon.

Source: Ecology - nature.com