Comment on ‘Geoengineering with seagrasses: is credit due where credit is given?’

Peter I Macreadie1,4 , Carolyn J Ewers-Lewis1, Ashley A Whitt1, Quinn Ollivier1, Stacey M Trevathan-Tackett1, Paul Carnell1 and Oscar Serrano3 1 Deakin University, School of Life and Environmental Sciences, Centre for Integrative Ecology, 221 Burwood Highway, Burwood, VIC 3125, Australia 2 Climate Change Cluster, University of Technology Sydney, Sydney 2007, Australia 3 School of Science and Centre for Marine Ecosystems Research, Edith Cowan University, Joondalup, WA 6027, Australia 4 Author to whom any correspondence should be addressed. OPEN ACCESS

Over the past decade scientists around the world have sought to estimate the capacity of seagrass meadows to sequester carbon, and thereby understand their role in climate change mitigation. The number of studies reporting on seagrass carbon accumulation rates is still limited, but growing scientific evidence supports the hypothesis that seagrasses have been efficiently locking away CO 2 for decades to millennia (e.g. Macreadie et al 2014, Mateo et al 1997, Serrano et al 2012. Johannessen and Macdonald (2016), however, challenge the role of seagrasses as carbon traps, claiming that gains in carbon storage by seagrasses may be 'illusionary' and that 'their contribution to the global burial of carbon has not yet been established'. The authors warn that misunderstandings of how sediments receive, process and store carbon have led to an overestimation of carbon burial by seagrasses. Here we would like to clarify some of the questions raised by Johannessen and Macdonald (2016), with the aim to promote discussion within the scientific community about the evidence for carbon sequestration by seagrasses with a view to awarding carbon credits. Johannessen and Macdonald (Johannessen and Macdonald 2016) reported that estimates of global carbon burial by seagrasses have been overestimated by 11-to 3100 fold (  (2010) reported estimates of mean seagrass net community production (120 g C m −2 yr −1 ; Duarte et al 2010), accumulation of seagrass autochthonous organic carbon (41-66 g C m −2 yr −1 ), and allochthonous organic carbon (42-67 g C m −2 yr −1 ). Assuming that there is net export of seagrass organic carbon from the meadow, Kennedy et al (2010) concluded that carbon sequestration by seagrass meadows may be better approximated by the sum of their net community production and the allochthonous carbon trapped in their sediments, which results in an estimated sequestration rates of 160-186 g C m −2 yr −1 . This approach taken by Kennedy et al (2010) does not account for post-depositional processes in marine sediments (e.g. biomixing and remineralization), as raised by Johannessen and Macdonald (2016), however, it is important to note that Kennedy et al (2010) aimed to provide estimates of seagrass carbon burial, as opposed to estimating carbon crediting opportunities (i.e. autochthonous carbon only). In order to estimate crediting opportunities from seagrass carbon burial, the data presented by Kennedy et al (2010) should be interpreted in a different way: based only on the accumulation of autochthonous organic carbon (41-66 g C m −2 yr −1 ; 12-40 Tg C yr −1 globally). Using the updated minimum seagrass area reported by Pendleton et al (2012) (170 000 km 2 ), as used by Johannessen and Macdonald (2016), the seagrass global carbon burial estimates (i.e. autochthonous carbon only) that may have implication for crediting would range from 7-40 Tg C yr −1 .

Reliability of global estimates of seagrass carbon sequestration
Second, we were unable to reproduce the 11-to 3100 fold overestimation reported by Johannessen and Macdonald (2016), their calculation for this overestimation is not provided and the units are missing, plus the calculation likely involves some misconceptions. For example, Johannessen and Macdonald (2016) used sediment accumulation rates from general coastal areas to estimate global carbon burial by seagrasses (Alvisi 2009, Boudreau 1994 to determine the % additional organic carbon due to seagrass. However, it seems they did not account for the % of autochthonous and allochthonous carbon in their calculations presented in table 1, despite the fact that they clearly stated in section 4.5 that allochthonous carbon capture does not necessarily represent additional burial. Additionally, the previous global estimates by Kennedy et al (2010) seem to be wrongly reproduced in table 1 from Johannessen and Macdonald (2016): ranging from 4.8 × 10 10 and 1.12 × 10 11 . Assuming that the units reported by Johannessen and Macdonald (2016) are grams (units not shown), then it should be 48 × 10 12 and 112 × 10 12 . As previously indicated, the calculation for this overestimation is not provided, the units of most variables used are missing, literature data is misreported, and the rationale behind the assumptions is not provided; thereby we were not able to reproduce their computations and we believe that Johannessen and Macdonald (2016) have incorrectly estimated global carbon burial by seagrasses.
Third, despite early estimates by Kennedy et al (2010) being based on limited available data and an indirect approach (accounting for plant productivity rather than sediment carbon accumulation), the range they provided is reasonable (12-40 Tg autochthonous C yr −1 globally, or 48-112 Tg total C yr −1 ). Here we show that previous global estimates are within the range of estimates based on seagrass carbon burial data published in peer-reviewed literature. Using the lowest seagrass carbon burial rate (2 g C m −2 yr −1 in Posidonia meadows, burial estimated in a 1 m-thick sedimentary deposit based on 14 C geochronology) (Serrano et al 2014b, Serrano et al 2016b and the highest seagrass carbon burial rate (249 g C m −2 yr −1 in a Posidonia oceanica meadow, burial estimated in a 2 m-thick sedimentary deposit based on 14 C geochronology) (Serrano et al 2016b) reported to date, one could estimate the range of global seagrass carbon burial. Following the approach taken by Johannessen and Macdonald (2016) (global area of seagrass ranging from 177 000-600 000 km 2 ) but assuming that 43%-94% of sediment carbon is due to seagrass presence (based on direct measurements in seagrass cores; Serrano et al 2016a), we estimate that global seagrass carbon burial range 0.26-140 Tg C yr −1 . Therefore, despite the limitations of the early estimates of global seagrass burial provided by Kennedy et al (2010) (48-112 Tg C yr −1 for total carbon, or 7-40 Tg C yr −1 for autochthonous carbon), here we demonstrate that these were not necessarily overestimates, but rather that the variability of seagrass carbon sequestration is larger than initially thought.
Overall, we agree with Johannessen and Macdonald (2016) that the methods used by previous authors were indirect and therefore relied on large assumptions, and that further studies are required to understand differences in carbon burial among seagrass ecosystems, including biological and habitat characteristics, to further refine estimates of global seagrass carbon sequestration capacity. We also agree with Johannessen and Macdonald (2016) that carbon stock estimates in combination with 210 Pb age dating is one of the best approaches to accurately calculate carbon accumulation rates in seagrass meadows. We disagree, however, that only one previous study (Marba et al 2015) has used 210 Pb dating to create seagrass sediment chronologies; Johannessen and Macdonald (2016)

The motivation for seagrass carbon offsetting
Johannessen and Macdonald (Johannessen and Macdonald 2016) stated that 'For climate change mitigation, it is the change in the long-term sequestration rate that ultimately matters'. Here we would like to clarify that the real potential of seagrass ecosystems to mitigate greenhouse gas emissions is towards the preservation of existing meadows and restoration of lost meadows, which can result in avoided emissions from disturbed sediments after canopy loss. The vast majority of carbon stores in seagrass meadows are found in their sediments (Fourqurean et al 2012), and recent literature shows that disturbance of sediments after meadow loss can result in carbon dioxide emissions (Marba et al 2015, Macreadie et al 2015b, Serrano et al 2016c. Indeed the carbon burial capacity of seagrass meadows (ranging from 2-249 g C m −2 yr −1 ; Serrano et al 2016a) is small in terms of potential for crediting: the restoration of 1 ha of seagrass could result in then enhanced sequestration of 0.02-2.5 ton C yr −1 (valued at $0.88-$110, assuming a price of $12 per ton CO 2 ). However, avoided emissions through the preservation of seagrass meadows and the carbon stocks underneath could result in a much larger crediting benefit: the preservation of 1 ha of seagrass could result in avoided emissions of 19-220 ton C (assuming, conservatively, that 25% of stocks in 1 m-thick deposits are remineralized after meadow loss, data from (Serrano et al 2016b, Marba et al 2015, Macreadie et al 2015b); 7.5-88 kg C in 1 m-thick sediments), valued at $826-$9689 (assuming a price of $12 ton CO 2 ). Therefore, further initiatives aiming to determine the potential of seagrass meadows to mitigate climate change emissions should primarily focus on the understanding of the loss and fate of carbon stores after meadow loss.

Moving forward
In conclusion, we argue that global carbon sequestration by seagrasses has not been properly established, but current estimates are within the range reported by growing scientific evidence. Increasing research on carbon sequestration rates by seagrasses showed that their capacity to sequester carbon can be highly variable due to biological, physical and chemical factors. Perhaps the largest current cause of high variance in estimates of global seagrass carbon sequestration is from the high uncertainty in global seagrass area (Macreadie et al 2014). In addition, we need to better understand the fate of allochthonous carbon if it weren't trapped and buried by seagrass meadows. Further research is needed to constrain the range of estimates of seagrass carbon burial rates at local, regional and global scales.