Article Size 16 Μm Diameter (Methods)
Total Page:16
File Type:pdf, Size:1020Kb
Biogeosciences, 18, 169–188, 2021 https://doi.org/10.5194/bg-18-169-2021 © Author(s) 2021. This work is distributed under the Creative Commons Attribution 4.0 License. Increased carbon capture by a silicate-treated forested watershed affected by acid deposition Lyla L. Taylor1, Charles T. Driscoll2, Peter M. Groffman3,4, Greg H. Rau5, Joel D. Blum6, and David J. Beerling1 1Leverhulme Centre for Climate Change Mitigation, Department of Animal and Plant Sciences, University of Sheffield, Sheffield S10 2TN, UK 2Department of Civil and Environmental Engineering, 151 Link Hall, Syracuse University, Syracuse, NY 13244, USA 3City University of New York, Advanced Science Research Center at the Graduate Center, New York, NY 10031, USA 4Cary Institute of Ecosystem Studies, Millbrook, NY 12545, USA 5Institute of Marine Sciences, University of California, Santa Cruz, CA 95064, USA 6Department of Earth and Environmental Sciences, University of Michigan, Ann Arbor, MI 48109, USA Correspondence: Lyla L. Taylor (l.l.taylor@sheffield.ac.uk) Received: 23 July 2020 – Discussion started: 31 July 2020 Revised: 8 November 2020 – Accepted: 13 November 2020 – Published: 12 January 2021 Abstract. Meeting internationally agreed-upon climate tar- which aims to accelerate the natural geological process of gets requires carbon dioxide removal (CDR) strategies cou- carbon sequestration by amending soils with crushed reac- pled with an urgent phase-down of fossil fuel emissions. tive calcium (Ca)- and magnesium (Mg)-bearing rocks such However, the efficacy and wider impacts of CDR are as basalt (The Royal Society and The Royal Academy of poorly understood. Enhanced rock weathering (ERW) is a Engineering, 2018; Hartmann et al., 2013). Forests repre- land-based CDR strategy requiring large-scale field trials. sent potential large-scale deployment opportunities where Here we show that a low 3.44 t ha−1 wollastonite treat- rock amendments may provide a range of benefits, includ- ment in an 11.8 ha acid-rain-impacted forested watershed ing amelioration of soil acidification and provisioning of in- in New Hampshire, USA, led to cumulative carbon cap- organic plant nutrients to cation-depleted soils (Hartmann −1 ture by carbonic acid weathering of 0.025–0.13 t CO2 ha et al., 2013; Beerling et al., 2018). Although ERW has not −1 over 15 years. Despite a 0.8–2.4 t CO2 ha logistical carbon yet been demonstrated as a CDR technique at the catch- penalty from mining, grinding, transportation, and spread- ment scale, a forested watershed experiment at the Hubbard ing, by 2015 weathering together with increased forest pro- Brook Experimental Forest (HBEF, 43◦560N, 71◦450 W) in −1 ductivity led to net CDR of 8.5–11.5 t CO2 ha . Our results the White Mountains of New Hampshire, USA, provides an demonstrate that ERW may be an effective, scalable CDR unusual opportunity for assessing proof of concept in this strategy for acid-impacted forests but at large scales requires priority research area. sustainable sources of silicate rock dust. The HBEF watershed experiment, designed to restore soil calcium following decades of leaching by acid rain, involved application of a finely ground rapidly weathered calcium sil- −1 icate mineral wollastonite (CaSiO3; 3.44 t ha ) on 19 Oc- 1 Introduction tober 1999 to an 11.8 ha forested watershed (Likens et al., 2004; Peters et al., 2004; Shao et al., 2016). Unlike the car- The Intergovernmental Panel on Climate Change (IPCC) bonate minerals (e.g. CaCO3) commonly applied to acidified (Rogelj et al., 2018) Special Report on global warming in- soils (Lundström et al., 2003), wollastonite does not release dicates large-scale deployment of carbon dioxide removal CO2 when weathered (Supplement) so is much better suited (CDR) technologies will be required to avoid warming in for CDR (Hartmann et al., 2013). It also has dissolution ki- ◦ excess of 1.5 C by the end of this century. Land-based netics comparable to or faster than other calcium-rich sili- CDR strategies include enhanced rock weathering (ERW), Published by Copernicus Publications on behalf of the European Geosciences Union. 170 L. L. Taylor et al.: Increased carbon capture by a silicate-treated forested watershed cate minerals such as labradorite found in basalt (Brantley 2.1 Site and treatment et al., 2008). Thus, the HBEF experiment provides a timely and unparalleled opportunity for investigating the long-term 2.1.1 Site description (15 years) effects of ERW on CDR potential via forest and stream-water chemistry responses. The HBEF has a temperate climate with ∼ 1400 mm mean In the case of ERW with wollastonite, CDR follows as Ca annual precipitation of which up to one-third falls as snow cations (Ca2C) liberated by weathering consume atmospheric (Campbell et al., 2007). The mean temperatures in January − − ◦ CO2 through the formation of bicarbonate (HCO3 ) by charge and July are 9 and 18 C respectively, and the period from balance, as described by the following reaction: mid-May to mid-September comprises the growing season (Campbell et al., 2007). There are six small southeast-facing CaSiO C 3H O C 2CO ! Ca2C C H SiO C 2HCO−: (R1) 3 2 2 4 4 3 watersheds in the HBEF (Fig. 1) with 20 %–30 % slopes However, forests in the northeastern USA have experienced (Groffman et al., 2006), including one which received the acid deposition (Likens and Bailey, 2014), changes in ni- silicate treatment (watershed W1, 11.8 ha, 488–747 m a.s.l.) trogen cycling (Goodale and Aber, 2001; McLauchlan et and a biogeochemical reference (watershed W6, 13.2 ha, al., 2007), and increases in dissolved organic carbon (DOC) 545–791 m a.s.l.). Carbonate and evaporite minerals are in fluxes (Cawley et al., 2014) that may affect CO2 removal ef- very low abundance (< 1 % calcite in the crystalline rocks ficiency by ERW processes. In particular, CO2 consumption and glacial deposits) in these silicate-mineral-dominated wa- as measured by bicarbonate production may be diminished if tersheds (Johnson et al., 1981). Well-drained Typic Hap- 2− − sulfate (SO4 ), nitrate (NO3 ), or naturally occurring organic lorthod soils with pH < 4.5 and mean depth of 0.6 m formed − acid (OA) anions (Fakhraei and Driscoll, 2015) (H2A ) in from relatively impermeable glacial till, which restricts wa- DOC intervene to inhibit the following mineral weathering ter flow and protects the underlying schist bedrock from reactions. For example weathering. Overland runoff and flow through bedrock are both thought to be negligible (Likens, 2013). Hydrologically, CaSiO C H O C H SO ! Ca2C C H SiO C SO2−; (R2) 3 2 2 4 4 4 4 the HBEF watersheds are typical of small catchments in C C ! 2C C C − CaSiO3 H2O 2HNO3 Ca H4SiO4 2NO3 ; (R3) northern New England (Sopper and Lull, 1965). Flow rates 2C − CaSiO3 C H2O C 2H3A ! Ca C H4SiO4 C 2H2A : (R4) for W1 and W6 along with stream-water pH are shown in Fig. S1 in the Supplement. Prior to treatment, stream-water These environmental effects on stream-water chemistry are − calcium concentrations were under 30 µmol L 1 while bi- well documented at the HBEF (Cawley et al., 2014; Likens − carbonate concentrations were under 5 µmol L 1, below the and Bailey, 2014; Rosi-Marshall et al., 2016; McLauchlan ranges for typical world rivers (Moon et al., 2014) (60– et al., 2007) and may be exacerbated under future climate C − − − 2293 µmol Ca2 L 1, 179–4926 µmol HCO L 1). change (Sebestyen et al., 2009; Campbell et al., 2009). 3 Fagus grandifolia Britt., Betula alleghaniensis Ehrh., and Here we exploit the experimental design and long-term Acer saccharum Marsh. are the dominant trees in this north- monitoring of stream-water chemistry, trees, and soils for ern hardwood forest, while Betula papyrifera Marsh., Abies two small forested HBEF watersheds to evaluate the effects balsamea (L.) Mill., and Picea rubens Sarq. are common at of the wollastonite treatment in 1999 on catchment CO con- 2 the highest elevations where soils tend to be shallow and sumption via inorganic and organic pathways. Further, we wetter (Cho et al., 2012). A. saccharum and P. rubens are examine how biogeochemical perturbations in S, N, and or- both calcium-sensitive, but soil calcium-bearing minerals are ganic carbon cycling affect catchment inorganic CO con- 2 less available to A. saccharum (Blum et al., 2002), and total sumption. We consider the forest response, the carbon cost bioavailable calcium content decreases with elevation (Cho for ERW deployment (mining, grinding, transportation, and et al., 2012). This silicate-addition experiment was designed application), and the net greenhouse gas balance for the treat- to replace bioavailable calcium which had been stripped from ment. Finally, we provide an initial assessment of the net the soils by decades of acid deposition. CDR potential of silicate treatments deployed over larger ar- eas of acidified forest in the northeastern United States. 2.1.2 Treatment description − 2 Methods On 19 and 21 October 1999, W1 was treated with 344 g m 2 of pelletized wollastonite (CaSiO3) by a GPS-equipped heli- This section describes the site and wollastonite treatment copter with a motorized spreader to ensure even deployment (Sect. 2.1) and our approaches for modelling the inorganic across the catchment, including the 1804 m2 streambed (Pe- carbon fluxes in stream water (Sect. 2.2) and other green- ters et al., 2004). Following treatment, the lignin-sulfonate house gas fluxes associated with the treatment (Sect. 2.3). binder forming the pellets dissolved within several days (Pe- The variables from each of the seven equations in Methods ters et al., 2004), and the ground wollastonite itself dissolved are tabulated in Table 1 along with the section, equation, fig- rapidly in the upper Oie soil horizon, increasing Oie base sat- ure, and table numbers where they appear.