Carbon sink services for tropical coastal seagrass are far lower than anticipated when accounting for black carbon

Valuing the sedimentary ‘blue carbon’ stocks of seagrass meadows in mitigating greenhouse gas emissions requires the exclusion of allochthonous recalcitrant forms, such as black carbon (BC) from the stock assessment. Regression models constructed across a tropical estuary predicted that carbon sinks within the more abundant sandy meadows of coastal bays likely support a significant but modest BC fraction. We tested the prediction by measuring BC fractions of total organic carbon (TOC) across three coastal meadows of the same region. One patchy meadow was located close to a major urban centre while the remaining two continuous meadows where contained in separate open embayments of a rural marine park, differing in fetch and species. In all cases, the BC/TOC fractions were significantly greater than predicted constituting a major component of the organic carbon content, 28% ± 1.6, and 26% ± 4.9 to 36% ± 1.5 (±95% confidence intervals) for urban and marine park meadows respectively. The higher BC/TOC fractions were explained by site-specific variability in BC atmospheric supply, patchy coverage, and a presumed increase in the loss of seagrass litter, as determined by the canopy height and proximity to the meadows exposed edge.


Introduction 28
The realisation that anthropogenic emissions of CO 2 is effecting climate change has 29 highlighted the importance of quantifying and managing existing sedimentary organic carbon 30 stocks, buried and protected from remineralisation. Lately, there has been a focus on 'blue 31 carbon' reservoirs. These ecosystems are the seagrass, saltmarsh, mangrove, and macro-algae 32 (1, 2). Of the four, seagrass ecosystems remain better placed to augment their organic carbon 33 stocks. Coastal seagrass meadows filter out organic detritus washed out across intertidal and 34 terrestrial landscapes and sequester them within their sediments (3), which would otherwise 35 be mineralised across the continental shelf (4). 36 Chew and Gallagher (5), however, challenged the traditional biogeochemistry mass 37 balance concept of carbon storage. They argued that because recalcitrant organic carbon 38 produced outside an ecosystem does not require protection from remineralisation, then their 39 presence within sediments cannot be counted as a burial service in the mitigation of 40 greenhouse gas emissions. Black carbon (BC) is an example of an 'allochthonous 41 recalcitrant'. It is formed during the incomplete combustion of biomass and fossil fuels, of 42 which SE Asia is a global hotspot (6). Its supply to seagrass meadow sediments can be both 43 through atmospheric deposition and with soil washout (5). However, the black carbon content 44 of coastal seagrass sediments is unknown. Nonetheless, estimates of its importance to the 45 total organic carbon content (TOC) have been made from a tropical estuarine system, of 46 around 18 ± 3% (±95% confidence interval) within tropical regions located around Sabah 47 (Malaysia) (5). While the BC/TOC fraction was significant, if not moderate, the equivalence 48 may have be confounded. Within the confines of a tide-dominated estuary, the BC fraction 49 not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was this version posted December 13, 2018. . https://doi.org/10.1101/493262 doi: bioRxiv preprint could be reduced by a sizeable fraction, due to the lost seagrass litter returned on the flood 50 tide. Contrast this with coastal seagrass meadows, which lose a large fraction of its litter to 51 the coastal shelf across the whole meadow or close to their exposed boundary (7-9). It is also 52 plausible that the narrow entrance could restrict BC supply from coastal bare sediments by The BB meadow, contained a mix of smaller canopy pioneer species, Cymodocea rotundata 65 and Halodule pinifolia, consistent with a relatively more energetic environ, over that of the 66 LL meadow, dominated by a large leafed climax species, Enhalus acoroides (10). In contrast, 67 the Sepanggar Bay meadow near ODEC beach (OD) was patchy and reminiscent of a 68 degraded system. The meadow supported a mixed bed species of mainly pioneer forms, 69 Cymodocea serrulata, Halodule uninervis, and Halophila ovalis, with only isolated small 70 stands of E. acroides. 71 Across each meadow, the first 1 cm of sediment was taken haphazardly within a 25 72 cm 2 quadrat for BC and TOC contents. Quadrats were laid down every 5 m along two parallel 73 50 m transects perpendicular to the exposed boundary of a prevailing Monsoon 74 not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was this version posted December 13, 2018. . https://doi.org/10.1101/493262 doi: bioRxiv preprint (Supplementary Figure S1). Along with BC and TOC, a range of sedimentological and 75 seagrass biological variables were also measured within each quadrat (table 1) (table 1). Furthermore, BB had on average 2.4 times higher 95 sedimentary TOC contents than LL, which was not reflected in a higher silt/clay fraction 96 (table 1). In contrast, there was no differences for both the average sedimentary TOC and BC 97 between the inner and exposed transects of the patchy meadow at OD. Instead, most of the 98 variance was contained across the transect's TOC resulting in a relatively invariant response 99 not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was this version posted December 13, 2018. . https://doi.org/10.1101/493262 doi: bioRxiv preprint with BC content (figure 2b). It should also be noted that the mean particle sizes within the 100 patch stands at OD were much greater and not smaller than found in adjacent bare patches of 101 the meadow of 8.1µm ± 2.4 and 107.5µm ± 19.0 (±95% CI) respectively (analysed from 102 electronic supplementary material, table S1). 103 104 Discussion 105 As expected, the sedimentary TOC content was less along the outer more exposed 106 transects for both LL and BB meadows. We believe that falling rates of resuspension and 107 subsequent oxidisation closer to the exposed meadows edge is not a likely explanation (4). 108 There was a near doubling of seagrass coverage, and presumably their root's ability to bind 109 sediments (11), across the inner transect of BB, and not for LL (table 1) deposition of these sediments is likely to be greater across the inner transects, as turbulence is 114 increasingly attenuated (12). This should reduce and not increase the inner transects TOC 115 contents. Secondly, a model that describes increases in BC with TOC, which converges 116 towards a positive TOC intercept close to its origin, is more consistent with additional 117 organic carbon not associated with BC. A more likely alternative for the inter and intra 118 meadow differences in TOC is a greater reduction in turbulence and loss of litter across the 119 meadow, by the larger LL canopy species (12) (table 1). 120 The OD meadow variance in both TOC, its relationship to BC and sedimentology 121 appears to illustrate a separate circumstance over the more continuous and rural meadows. 122 The larger mean sediment particle sizes within the patch stands over the bare regions of the 123 OD meadow (see Results section), would imply a seemingly inconsistent greater and not 124 not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was this version posted December 13, 2018. with TOC. Significant differences between transect TOC and BC means could be found for 167 LL (P < 0.05, P < 0.05) and BB (P < 0.07, P < 0.05) respectively. The probablity of the 168 regression intercepts and slopes between BB and LL are being the same was statistically 169 significant (ANCOVA P > 0.38). 170 171 172 not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.