Introduction
As the world enters an era marked by climate change, the role of cities in achieving carbon neutrality is becoming increasingly critical. Although cities occupy only about 2% of the global land area (Nations, 2019), they account for over 70% of greenhouse gas emissions worldwide (Luqman et al., 2023), with more than half of these emissions originating from roughly 15% of all cities (Wei et al., 2021). Furthermore, around 55% of the global population lived in urban areas in 2015, a figure that is projected to rise to 68% by 2050 (Nations, 2019). This urban population increase is expected to drive greater economic activity and energy consumption, implying that both total carbon emissions and per capita carbon emissions are closely linked to urbanization (Yao et al., 2018).
In the context of climate change, cities are recognized as both the most vulnerable to its impacts and as key hubs for the implementation and innovation of climate change-related technologies and policies (Commission et al., 2020). Achieving carbon neutrality at the city level is now a key task in climate change response strategies, and is being discussed intensively in urban planning and energy policies. Research is focusing on the use of carbon sinks through securing and maintaining urban green spaces as a strategy to reduce greenhouse gases. Urban green spaces, including parks, street trees and urban forests, are vital green infrastructures that provide social, economic and environmental benefits. These benefits include improving the urban living environment, enhancing quality of life, and mitigating the urban heat island effect (Boulton et al., 2018; Cheng et al., 2021).
In these green spaces, vegetation plays a vital role in carbon storage and sequestration. Through photosynthesis, plants absorb atmospheric CO2 and convert it into biomass, thus contributing to the reduction of overall carbon emissions (Berry et al., 2010; Warrick, 1986; Y. Zhang et al., 2022). Numerous studies have quantitatively assessed the carbon sequestration capacity of vegetation in urban green spaces, and based on these findings, have examined the potential of cities to offset carbon emissions (W. Y. Chen, 2015; McPherson & Simpson, 2003; Nowak et al., 2013).
Soil also plays a central role in carbon storage and sequestration. Globally, approximately 1,500Pg(1Pg=1Gt= 1015g) of organic carbon is stored in soils, which is about three times the amount stored in plant biomass and roughly double the amount in the atmosphere (Batjes, 1996; Houghton, 2001). Through processes such as plant respiration, photosynthesis, microbial activity, as well as the application of fertilizers and irrigation, soils capture and store atmospheric carbon (Gougoulias et al., 2014; K. Zhang et al., 2024). This function is recognized as a key attribute of urban green spaces. Consequently, urban green spaces, composed of both vegetation and soil, can act as comprehensive carbon sinks and play a pivotal role in urban carbon storage and sequestration (Adetoye et al., 2018; Nowak et al., 2013; McPherson & Simpson, 2003; Nowak & Crane, 2002).
However, in cities that are undergoing rapid urbanization or are already highly urbanized, rising land prices and high development densities create spatial and financial barriers to the establishment and maintenance of new green spaces (Kim et al., 2018; Rupprecht, 2017; Yokohari et al., 2010). In response to these challenges, many cities are exploring alternative green spaces, such as rooftop green spaces and other forms of green infrastructure on artificial surfaces (Mihalakakou et al., 2023; Seyedabadi et al., 2021). When placed on building rooftops, these green spaces help regulate the microclimate by reducing heat transfer, mitigating the urban heat island effect, and decreasing stormwater runoff (Y. Gong et al., 2020; Tabatabaee et al., 2019; Berndtsson et al., 2009). Additionally, rooftop green spaces contribute to various ecosystem services, such as enhancing urban biodiversity (Cristiano et al., 2021; Pauleit et al., 2018). Rooftop green spaces also can fulfill social and aesthetic functions that improve urban well-being and overall city aesthetics, ultimately boosting the quality of urban life (Williams et al., 2019).
Recent studies have begun to examine not only the environmental and ecological benefits of rooftop green spaces but also their potential for carbon storage and sequestration (Shafique et al., 2020). In high-density urban environments, green infrastructure on artificial surfaces—like rooftop green spaces—can serve as significant carbon sinks, and thus may help cities meet carbon-neutral goals. Nevertheless, current research has largely focused on the vegetation layer of rooftop green spaces, leaving relatively few studies that investigate the carbon reduction effects of the soil materials used in such systems.
For rooftop green spaces and similar forms of green infrastructure on artificial surfaces, soil materials must be lightweight to avoid overloading building structures, and must simultaneously provide sufficient nutrients for plant growth. According to Korea’s national construction standard (MOLIT, 2024) natural soil is recommended for green infrastructure on artificial surfaces, although lightweight artificial soil is also permissible when required due to safety and structural considerations. In this context, it is generally recommended to use a lightweight soil mix that maintains an optimal balance between water retention and drainage, preventing excessive drying of the artificial ground while reducing structural load. The most commonly used materials in such contexts include sand, volcanic sand (pumice), perlite, and peat-moss (MOLIT, 2024), in addition to other lightweight options such as cocopeat, vermiculite, and zeolite (Kazemi & Mohorko, 2017). Perlite is the most commonly used soil material for rooftop green spaces and other green infrastructure on artificial surfaces due to its relative affordability and light weight. However, due to its limited capacity for water retention, perlite requires frequent irrigation in arid climates, which can impede root development (Akther et al., 2018; Cascone, 2019; Green, 1968). Additionally, perlite is deficient in inherent nutrients, necessitating the incorporation of chemical fertilizers, and is susceptible to erosion caused by wind or rain, compromising soil stability and impeding root anchorage (Cascone, 2019). Moreover, perlite production involves substantial energy consumption, resulting in considerable environmental impact and limiting its viability as a carbon-reducing or carbon-neutral material (Toboso-Chavero et al., 2021).
For this reason, research into the sustainability and carbon-reducing potential of the soil materials used in rooftop green spaces is an important task. This is particularly relevant at a time when green infrastructure on artificial surfaces, such as rooftop gardens, is attracting attention as a carbon sink as part of climate change response strategies. This study aims to evaluate the potential of greenhouse gas-reducing soil by conducting physicochemical properties analysis and elemental composition analysis, as well as a plant growth experiment according to the mixing ratio of various soil materials used in rooftop green spaces.
Research Methods
Main materials
In this study, four experimental groups were established by combining various artificial soil materials commonly used in urban rooftop greening projects. Each experimental group underwent analyses of physicochemical properties, elemental composition, and plant growth performance. Notably, instead of relying on chemical fertilizers, organic matter was used to supply nutrients. The objective was to explore the potential of greenhouse gas (GHG)-mitigating soils that consider both the physical and chemical properties of the substrate. To this end, the experimental materials selected were bottom ash, biochar, zeolite, peat-moss, carbonized rice hull (CRH), humus, vermiculite, and cocopeat.
Bottom ash is a byproduct generated during the rapid combustion of crushed wood pellets in thermal power plants, which accumulates in the boiler hopper. Generally, bottom ash is discharged using seawater; therefore, repeated washing with fresh water is required to remove salts before it can be used as a soil amendment (J.-H. Park et al., 2023). Research in horticulture and agriculture has noted that bottom ash exhibits structural and chemical properties similar to those of biochar derived from woody biomass (J.-H. Park et al., 2020). Biochar is a porous carbon material produced by the low-temperature pyrolysis of biomass under anaerobic conditions, with feedstocks such as rice straw, livestock manure, and sewage sludge (Tan et al., 2021; Wang et al., 2020; Yue et al., 2017). Recognized as a low-cost, high-efficiency soil conditioner, biochar not only promotes plant growth through its excellent water retention capacity, but also reduces the carbon footprint in rooftop greening (H. Chen et al., 2018; Tan & Wang, 2023). Zeolite is a natural mineral that despite containing impurities such as quartz is commonly used as a stabilizing agent to remediate heavy-metal-contaminated soils, owing to its unique crystalline structure (Budianta et al., 2020; Zheng et al., 2020). When added to soil in appropriate proportions, zeolite reduces the mobility of heavy metals, increases the availability of potassium, calcium, magnesium, sodium, and phosphorus, and enhances the soil’s cation exchange capacity (CEC). Due to its environmentally friendly and cost-effective nature, zeolite has attracted considerable attention in both research and practice (Tasharrofi et al., 2020). Peat-moss is formed from aquatic plants and mosses that partially decompose under low-oxygen conditions. It typically exhibits a mildly acidic pH range of 3.5 to 4.5, features a fine structure with ample pore space, and has an excellent water-retention capacity. These properties promote root development and air exchange (Walmsley, 1977). However, the harvesting process raises concerns about its environmental sustainability, and peat-moss also regenerates slowly. Many countries, including South Korea, depend entirely on imports of peat-moss, a fact which has prompted active research into possible alternatives (Gruda & Schnitzler, 2004; Gupta et al., 2022; Rozas et al., 2023; D. Zhang et al., 2023). CRH is produced by burning rice hulls, and contains essential trace elements and minerals for plant growth, such as K, P, Ca, and Mg. In addition, CRH increases soil’s organic matter and the carbon-to-nitrogen (C:N) ratio, thereby positively affecting nutrient balance (Haefele et al., 2011). Recent studies indicate that CRH can enhance carbon storage in soils (W.K. Park et al., 2016), and shows strong potential for use in organic farming as a sustainable soil material (Nierras, 2019). Humus is an organic layer formed over a considerable period by the partial decomposition of plant materials, such as leaf litter and small branches. Rich in organic matter and with excellent CEC, humus is highly effective in improving soil’s fertility. Consequently, it is widely used in agriculture and for isolating functional microbes, in addition to increasing soil organic carbon and microbial biomass carbon, thereby enhancing overall soil health. Vermiculite is a mineral-based soil conditioner produced by expanding the mineral at temperatures higher than 1,000 °C. Classified as a clay mineral, vermiculite has small particle sizes and high CEC, offering excellent nutrient retention and a tendency toward alkalinity (Malandrino et al., 2011). It also fixes NH4+-N in the soil, thereby reducing nitrogen loss (H. Zhang & Yao, 2017), and enhancing nitrogen fertilizer efficiency to promote crop growth. Cocopeat is an organic soil amendment derived from coconut husks, a byproduct of the coconut processing industry (Arenas et al., 2002). It typically has a near-neutral pH range of 5.2 to 6.0 and exhibits excellent drainage, aeration, and antimicrobial properties (Gbollie et al., 2021). Because coconuts are grown mainly in coastal areas with high salinity, cocopeat shows strong tolerance to sodium chloride and thus tends to have a relatively high electrical conductivity (EC)(Poulter, 2011). Recent research has praised cocopeat as an eco-friendly, renewable resource that could serve as a viable substitute for peat-moss.
In this study, the mixing ratios of soil materials were established to optimize the properties of the artificial soil used in rooftop greening by considering the water retention capacity, drainage, aeration, and carbon storage ability of each material. Specifically, organic materials such as peatmoss and cocopeat enhance water retention, while biochar and CRH promote carbon storage. Bottom ash and zeolite contribute to lightweight properties and improved drainage. Reflecting these characteristics, the ratios of each material were experimentally adjusted, and the optimal mix was determined based on expert consultations. Based on these eight soil materials, four types of artificial soil mixes (ASM) were produced, which have been designated as ASM_A, ASM_B, ASM_C, and ASM_D (Table 1). All ASMs contain 35% bottom ash and 10% zeolite as inorganic components; in addition, ASM_D incorporates 10% vermiculite, resulting in an inorganic-to-organic ratio of 55:45, while ASM_A, ASM_B, and ASM_C maintain a ratio of 45:55.
ASM_A contains 35% bottom ash and 10% zeolite as inorganic components, plus 10% biochar, 35% peat-moss, and 10% humus as organic components. This mix is high in organic matter and offers an excellent water-retention capacity.
ASM_B is derived from ASM_A by removing humus, reducing peat-moss to 10%, and adding 35% cocopeat. This formulation focuses on reducing weight and supplying organic matter.
ASM_C maintains the same composition as ASM_B but replaces 10% biochar with 10% CRH to enhance the carbon storage capacity.
ASM_D has a higher proportion of inorganic components (bottom ash, zeolite, and vermiculite). Although the total organic matter content is lower, it includes 10% biochar and 35% cocopeat to optimize the beneficial water-retention and aeration properties of vermiculite.
Experimental design
To evaluate the sustainability and carbon reduction potential of ASMs as GHG-mitigating soil materials, this study conducted analyses of physicochemical properties, elemental composition, and plant growth performance. To assess the fertility of each ASM and its suitability for plant growth, measurements were taken for pH, cation exchange capacity, electrical conductivity, total organic carbon, P2O5, NH4-N, and NO3-N. Elemental composition analyses were also performed to estimate the relative carbon sequestration capacity of the ASMs.
Furthermore, to empirically demonstrate how the physicochemical properties and elemental composition of each ASM affect plant growth, a plant growth experiment was conducted. Lettuce (Lactuca sativa var. crispa, commonly referred to as "cheong-chima lettuce") was selected due to its short growth cycle and high sensitivity to environmental changes. The control soil (CS) was a commercially available horticultural substrate frequently used in rooftop greening and other artificial ground-based greening projects. The size of the container used for the plant growth experiment was 530 × 260 × 40 mm, which can accommodate 50 seeds. The experimental group (Exp_n) consisted of respectively 5 containers of ASM and CS, each with 250 seeds (Fig. 1). Ultimately, lettuce seeds were sown in 250 cells of each ASM and 250 cells of the CS, and the experiment proceeded for 50 days.
The experiment ran for 50 days in a glass greenhouse, from January 24, 2024 (sowing date) to March 4, 2024. The greenhouse was equipped with an automatic retractable shading system as well as heating and cooling devices. Temperature and humidity were monitored twice daily to maintain daytime temperatures of 23 to 28°C, nighttime temperatures around 20°C, and relative humidity at 55 to 60%. Based on these conditions, watering was performed twice daily using a spray hose. Ten days after sowing, the germination rate was recorded. Over the following 30 days, a damage scale assessment was conducted three times at 10-day intervals. Growth assessments were carried out twice to evaluate plant performance by measuring leaf length (plant height), leaf width, leaf number, and chlorophyll content (SPAD value). Statistical significance for each measurement was determined using Duncan’s multiple range test.
Results and Discussion
Soil characteristics
In this study, four experimental groups were prepared by blending eight types of artificial soil materials commonly used in urban greening projects on artificial ground. These groups were subsequently assessed for their physicochemical properties, elemental composition, and plant growth performance to evaluate their potential as soil media for GHG reduction. The physicochemical characteristics and elemental compositions of each experimental group are summarized in Table 2.
Soil pH
Soil pH not only indicates the acidity or alkalinity of the soil, but is also closely linked to the uptake of nutrients essential for plant growth (Liu et al., 2022; Sultana et al., 2020). An appropriate soil pH promotes the effective absorption of nutrients such as nitrogen and phosphorus, thereby supporting sustainable plant development (Hu et al., 2021). Consequently, soil pH is a critical chemical property that influences nutrient availability, the solubility of toxic substances, and the biochemical reactions within plant roots and microbial communities. Generally, slightly acidic to neutral soils are considered most conducive to plant growth (National Institute of Agricultural Sciences, 2015). According to the Korean Design Standard (MOLIT, 2024), a soil pH range of 5.5 to 7.0 is recommended for planting on artificial ground, whereas the Rural Development Administration of Korea, via the Korean Soil Information System, recommends a pH of 6.0 to 7.0 for protected cultivation. In the present study, the soil pH values of the four ASMs ranged from 6.82 (ASM_A) to 7.09 (ASM_D), a generally neutral range.
Cation Exchange Capacity (CEC)
CEC is an index of the soil’s ability to adsorb and exchange positively charged ions, reflecting its capacity to store nutrients and supply them to plants when needed (Robertson et al., 1999). Soils with a higher density of negative charges can retain more cations, which generally contributes to higher fertility (McKenzie et al., 2004). Soil texture strongly affects CEC values; generally, soils with a CEC of 15 cmol/kg or higher are classified as loam and considered suitable for crop cultivation (Moore, 2001). In Korea, the average CEC of agricultural land (including paddy fields, protected cultivation sites, upland fields, and orchards) is relatively low, at approximately 5.0 to 6.0 cmolc/kg. However, all four ASMs in this study exhibited CEC values characteristic of clay loam, which, owing to its higher clay content, can more effectively retain both water and nutrients. The highest CEC observed in ASM_A is likely attributable to its higher proportion of organic materials, such as peat-moss and humus, relative to the other formulations.
Electrical Conductivity (EC)
EC measures the total concentration of dissolved cations and anions in the soil, and serves as an indicator of soil salinity. High soil salinity implies an abundance of soluble ions, which can hinder seed germination by creating osmotic stress. In major protected cultivation facilities in Korea, the average EC is reported to be approximately 3.10 dS/m (Kong et al., 2016), which is relatively high. According to the Rural Development Administration’s Fertilizer Quality Testing Methods and Sample Collection Standards (ENFORCEMENT DECREE OF THE FERTILIZER CONTROL ACT, 2022) an EC of 1.2 dS/m or lower is considered appropriate. In this study, all four ASMs exhibited EC values below this recommended threshold (≤1.2 dS/m).
Total Organic Carbon (TOC)
TOC is measured in various research fields—including water pollution assessment, waste humification evaluation, estimation of soil carbon content, and studies of carbon fluxes in aquatic systems (Bisutti et al., 2004). In soils, TOC represents the total carbon stored in soil organic matter (SOM), and serves as an important parameter for characterizing both soils and sediments (Nykamp et al., 2024). In agriculture, SOM is a key determinant of soil fertility and quality, which are closely linked to overall productivity. For this reason, higher TOC is considered a major soil quality indicator that influences nutrient supply, improves soil’s physical and biological properties, and can ultimately increase crop yields (Johnston, 1994; Yang et al., 2012). In this study, the TOC contents of the four ASMs were 15.12% (ASM_A), 10.47% (ASM_B), 9.02% (ASM_C), and 8.02% (ASM_D). The relatively higher TOC in ASM_A and ASM_B can be attributed to the greater incorporation of peat-moss, which is rich in organic matter, whereas ASM_D lacked substantial amounts of high-organic materials such as peat-moss or humus, resulting in the lowest TOC.
Phosphorus (P) content
P is an essential nutrient involved in various physiological processes in plants, including root development, cell division, energy production, and gene expression. In soils, P2O5 is one of the primary forms of phosphorus, and serves as a key indicator of the soil’s capacity to support plant growth (Alewell et al., 2020). Because P2O5 strongly binds to soil particles, it exhibits low mobility and a high fixation rate compared with other nutrients. Generally, P is most readily available in soils with a pH of 6 to 7; in acidic soils, it is largely adsorbed by clay minerals and may form complexes that remain accessible to plants (Ara et al., 2018). P is commonly supplied via chemical fertilizers, and its increased application has been shown to significantly influence growth parameters such as plant height, fresh weight, and dry weight (Bae, Eun-Ji et al., 2022). In this study, ASM_A had a relatively higher available P2O5 content than ASM_B, ASM_C, and ASM_D, likely due to the presence of peat-moss and humus, whereas the other formulations—primarily containing cocopeat, CRH, and biochar (substances generally low in P)—exhibited reduced levels of available P2O5.
Ammonium-Nitrogen (NH4-N) and Nitrate-Nitrogen (NO3-N)
NH4-N and NO3-N are critical forms of N absorbed directly by plants for synthesizing proteins and nucleic acids, thereby promoting cell division and chlorophyll formation. According to the Rural Development Administration’s Fertilizer Quality Testing Methods and Sample Collection Standards (Enforcement Decree of the Fertilizer Control Act, 2022), the recommended upper limit for both NH4-N and NO3-N is 500 mg/kg. In this study, NH4-N ranged from 86.7 to 111.2 mg/kg, while NO3-N ranged from 73.07 to 86.63 mg/kg across the four blended soils.
Elemental composition
In soil, total carbon (TC) is generally classified into two main forms: total inorganic carbon (TIC), found in minerals such as limestone or dolomite, and total organic carbon (TOC), which represents the total amount of organic matter. TC has long been recognized as a crucial component of soil quality, and its storage is widely regarded as an effective strategy for mitigating climate change through carbon sequestration (X. Gong et al., 2013). In particular, the storage of soil organic carbon (SOC) has been proposed as a nature-based solution to climate change, garnering considerable attention for its potential in managed land such as parks and agricultural areas (Chien & Krumins, 2022; Tiefenbacher et al., 2021). TOC can be subdivided into labile organic carbon, which is easily decomposed by microorganisms and converted into substances such as CO2 in a short period of time, and stable organic carbon, which contributes to long-term carbon sequestration and soil structure stabilization. Because stable organic carbon is less susceptible to microbial decomposition and is often enriched through high-temperature treatment that volatilizes some oxygen and hydrogen, it forms a chemically complex and rigid structure referred to as black carbon (BC) (Baldock & Smernik, 2002). BC is extremely resistant to microbial degradation, releasing minimal GHG into the atmosphere and thereby helping sequester GHG such as CO2, CH4, and N2O that would otherwise be emitted during natural soil decomposition processes (Fowles, 2007). In this context, this study estimated that soil materials with higher carbon content, alongside organic matter, would likely contain significant amounts of stable BC. ASM_D and ASM_A exhibited comparatively high carbon content (27.60 wt% and 23.81 wt%, respectively), whereas ASM_C and ASM_B showed lower values (16.33 wt% and 16.10 wt%, respectively). Although biochar in ASM_A contributes to an increased overall carbon content, the large proportion of peat-moss and humus also result in relatively high levels of available nitrogen. While humus is rich in nitrogen and beneficial for biological activity, it is less stable for long-term carbon fixation. Thus, although ASM_A may facilitate plant and microbial activity, its relatively low C/N ratio could accelerate the release of CO2 through carbon decomposition. In contrast, ASM_D, with its high carbon content and elevated C/N ratio, achieves a more favorable balance between plant growth and carbon storage, indicating a relatively advantageous carbon reduction effect.
Plant growth
Germination rates were assessed on February 3, 2024—ten days after sowing. Rates were notably higher in ASM_B and ASM_C, presumably because their relatively abundant NH4-N served as an early nutrient source. However, a statistically significant difference emerged only in Exp_B, where ASM_B’s germination rate (98.2%) was significantly higher than that of CS_B (78.4%). In the other experimental groups, germination rates exceeded 80%, and no significant differences were found between ASM and CS (Table 3). Considering that lettuce is highly sensitive to environmental conditions such as soil properties (Hiroki et al., 2014), these findings suggest that the ASM used in this study did not adversely affect seed germination.
To evaluate plant growth after germination, three damage scale assessments (Fig. 2) were taken at ten-day intervals, along with two quantitative growth assessments (Table 4) on days 20 and 30. In the damage scale assessment, ASM_A consistently scored 0 (indicating no inhibitory effects) at each evaluation point, reflecting the most favorable growth status. In contrast, ASM_B and ASM_D exhibited mild growth inhibition (score 1) starting on day 20, whereas ASM_C—despite a relatively high germination rate of 91.2%—displayed more pronounced inhibition (score 2) as early as day 10.
Across the two growth assessments, ASM_B, ASM_C, and ASM_D generally showed significantly lower growth than CS in most parameters, whereas ASM_A did not (Table 4). Specifically, at day 20, ASM_A demonstrated growth values similar to those of CS for all measured parameters, with leaf width reaching 2.77 cm—slightly higher than CS (2.70 cm). By day 30, ASM_A’s performance remained comparable to CS for all variables except leaf length, which showed a statistically significant difference. In contrast, on day 20 ASM_B performed comparably to CS only in leaf number, and showed statistically poorer growth in all other parameters; by day 30, it lagged behind CS in every measured aspect. Similarly, ASM_C exhibited significantly lower growth than CS in both assessments. ASM_D showed growth similar to CS in most parameters on day 20, with its leaf width of 2.7 cm slightly higher than CS (2.67 cm). However, by day 30, ASM_D scored significantly lower than CS in all parameters except SPAD. Over time, ASM_B, ASM_C, and ASM_D increasingly trailed behind CS, particularly in leaf length. The only instance in which growth metrics exceeded those of CS occurred on day 20 in terms of leaf width for ASM_A and ASM_D. SPAD values tended to decline slightly over time in both CS and ASM treatments, with ASM_A and ASM_D remaining close to CS levels.
To compare the relative growth performance of lettuce across the four ASMs, leaf length, leaf width, and leaf number were selected as the primary growth indices. The average of the relative increment rates of these indices was computed to derive the Plant Growth Index (PGI) as follows:
In order to comprehensively compare both the germination rate - which reflects the early growth of lettuce - and the overall growth process among the ASM, each index was subjected to min-max normalization, and the results were visualized using a dumbbell plot (Fig. 3). In ASM_A, the normalized PGI was 1.00 while the normalized germination rate was 0, resulting in the greatest separation between these values. This outcome indicates the although ASM_A exhibits the highest relative growth rate, its germination rate is the lowest. Conversely, ASM_B showed a normalized PGI of 0.70 and a normalized germination rate of 1.00, which demonstrates that despite having the highest germination rate, its PGI is comparatively lower. For ASM_C, a normalized PGI of 0 and a normalized germination rate of 0.84 suggest that while germination was high, the subsequent growth performance was suboptimal. In the case of ASM_D, the normalized PGI was 0.74 and the normalized germination rate was 0.533, indicating that the two indices are relatively close and that both growth and germination rates are moderate. In the dumbbell plot, a longer horizontal line between the two points corresponds to a greater relative difference between the indices. Therefore, ASM_A is identified as the material exhibiting the highest growth performance relative to its low germination rate, whereas a closer proximity between the two points suggests that the PGI and germination rate are similarly ranked. Consequently, ASM_D is interpreted as the material with the most balanced indices.
Plant growth in the ASM generally resembled that in the CS, with only minor differences in certain parameters. Although ASM_A was expected to have an advantage in nutrient retention and supply due to its high CEC and TOC, this effect did not manifest during the early germination stage as clearly as it did in ASM_B and ASM_C. Nevertheless, at later stages, ASM_A exhibited growth that was similar to—or occasionally exceeding—that of CS. In contrast, despite ASM_D having the highest carbon content, and thus a strong potential for carbon reduction, its relatively low N and P2O5 levels appeared to limit its early growth performance.
Conclusion
This study empirically examined the carbon sequestration potential of various soil materials used in green infrastructure on artificial surfaces such as rooftop gardens, and assessed their effects on actual plant growth in the context of rising GHG emissions from urbanization. Four types of artificial soil mixes (ASM_A, ASM_B, ASM_C, and ASM_D) were prepared from eight soil materials, including bottom ash, biochar, peat-moss, cocopeat, humus, zeolite, vermiculite, and CRH. Physicochemical and elemental composition analyses were conducted to characterize each ASM in terms of carbon sequestration and plant growth, and a plant growth test verified each mix’s suitability for cultivation.
During the early stage of lettuce germination and up to 30 days of cultivation, ASM_A performed similarly to or slightly better than the CS. This result is likely attributable to its high organic matter content (TOC 15.12%) and excellent CEC (32.39 cmol/kg), along with its relatively high levels of available P2O5. Although such a nutrient-rich environment promotes rapid early growth, elevated rates of microbial carbon decomposition may limit long-term carbon fixation. In contrast, ASM_B and ASM_C, despite exhibiting high germination rates, showed weaker growth over time compared to the control, possibly due to lower available P and N levels and reduced synergies among the materials; their lower overall carbon contents further suggest a limited capacity for carbon sequestration. ASM_D, on the other hand, exhibited the highest carbon content (27.60 wt%) of the four mixes. The high proportion of recalcitrant carbon in ASM_D suggests notable long-term carbon storage potential; however, its lower N and P levels resulted in poorer germination and early growth performance relative to the CS. Based on the visual damage scale assessment (generally scoring 0 or 1, indicating minimal visible damage), ASM_D is expected to sustain plant growth over the longer term. From a practical perspective, ASM_D appears better suited for long-term carbon reduction and soil stability than for promoting rapid early growth.
These findings indicate that while peat-moss and humus effectively promote germination and early growth, their relatively rapid decomposition makes them less favorable for long-term carbon sequestration. In contrast, biochar and bottom ash support sustained carbon storage but supply fewer nutrients during the initial growth phase. Therefore, in choosing soil materials for green infrastructure on artificial surfaces, it is necessary to balance the objectives of rapid early plant establishment and long-term carbon sequestration. For example, if rapid early growth is the primary objective, a mix like ASM_A, with higher NH4-N, available P, and abundant organic matter may be preferable. Conversely, for long-term carbon reduction and sequestration, a mix such as ASM_D, which is rich in stable carbon (often referred to as BC), would be a strong candidate.
This study contributes to the field by evaluating the carbon reduction potential of ASMs for rooftop green spaces while directly examining plant growth outcomes. In designing green infrastructure—especially in climate change responses at the urban level—it is vital to consider both plant establishment and long-term carbon sequestration. Future research should refine mixing ratios for different urban environments and plant species, investigate more diverse combinations of organic and inorganic components, and extend the monitoring period beyond the initial 30 to 50 days to capture full plant development and long-term carbon storage. Identifying optimal ASM compositions that balance initial nutrient availability with enduring recalcitrant carbon retention will be critical for maximizing both plant growth and carbon sequestration.
PDF Links
PubReader
ePub Link
Full text via DOI
Download Citation
