ABSTRACT
Urban design is currently promoting the inclusion of plants in buildings. However, plants emit biogenic volatile organic compounds (BVOCs), which alone or in combination with other airborne molecules such as CO2, may result in a general increase in tropospheric pollution. Many studies have documented the effects of biotic and abiotic factors on plant BVOC responses, but few have assessed the contribution of typical CO2 levels found in indoor work and meeting spaces. To answer this question, we monitored CO2 and constitutive (MT-limonene) and induced (LOX-cis-3-hexenal) BVOC emissions of a fully developed tomato crop grown hydroponically inside an integrated rooftop greenhouse (i-RTG) in a Mediterranean climate. Two distinctive CO2 assays were performed at the level of the i-RTG by supplying or not CO2. The impact of CO2 on plant physiological emittance was then assessed, and the resulting BVOC rates were compared with reference to EU-LCI values. MT-limonene was ubiquitous among the assays and the most abundant, while LOX-cis-3-hexenal was detected only under controlled CO2 management. The highest levels detected were below the indicated LCIs and were approximately tenfold lower than the corresponding LCI for MT-limonene (50.88 vs. 5000 µg m-3) and eightfold (6.63 µg m-3) higher than the constitutive emission level for LOX-cis-3-hexenal. Over extended sampling (10 min) findings revealed a general emission decrease and significantly different CO2 concentration between the assays. Despite similar decreasing rates of predicted net photosynthesis (Pn) and stomatal conductance (gs) their correlation with decreasing CO2 under uncontrolled condition indirectly suggested a negative CO2 impact on plant emission activity. Conversely, increasing CO2 under the controlled assay showed a positive correlation with induced emissions but not with constitutive ones. Because of significantly higher levels of relative humidity registered under the uncontrolled condition, this factor was considered to affect more than CO2 the emission response and even its collection. This hypothesis was supported by literature findings and attributed to a common issue related with the sampling in static enclosure. Hence, we suggested a careful monitoring of the sampling conditions or further improvements to avoid bias and underestimation of actual emissions. Based on the main outcomes, we observed no evidence of a hazardous effect of registered CO2 rates on the BVOC emissions of tomato plant. Furthermore, because of the low BVOC levels measured in the i-RTG, we assumed as safe the recirculation of this air along building's indoor environments.
ABSTRACT
Urban green installations are extensively promoted to increase sustainable and accessible food production and simultaneously improve the environmental performance and liveability of city buildings. In addition to the multiple benefits of plant retrofitting, these installations may lead to a consistent increase in biogenic volatile organic compounds (BVOCs) in the urban environment, especially indoors. Accordingly, health concerns could limit the implementation of building-integrated agriculture. In a building-integrated rooftop greenhouse (i-RTG), throughout the whole hydroponic cycle, green bean emissions were dynamically collected in a static enclosure. Four representative BVOCs, α-pinene (monoterpene), ß-caryophyllene (sesquiterpene), linalool (oxygenated monoterpene) and cis-3-hexenol (LOX derivate), were investigated in the samples collected from two equivalent sections of a static enclosure, one empty and one occupied by the i-RTG plants, to estimate the volatile emission factor (EF). Throughout the season, extremely variable BVOC levels between 0.04 and 5.36 ppb were found with occasional but not significant (P > 0.05) variations between the two sections. The highest emission rates were observed during plant vegetative development, with EFs equivalent to 78.97, 75.85 and 51.34 ng g-1 h-1 for cis-3-hexenol, α-pinene, and linalool, respectively; at plant maturity, all volatiles were either close to the LLOQ (lowest limit of quantitation) or not detected. Consistent with previous studies significant relationships (r ≥ 0.92; P < 0.05) were individuated within volatiles and temperature and relative humidity of the sections. However, correlations were all negative and were mainly attributed to the relevant effect of the enclosure on the final sampling conditions. Overall, levels found were at least 15 folds lower than the given Risk and LCI values of the EU-LCI protocol for indoor environments, suggesting low BVOC exposure in the i-RTG. Statistical outcomes demonstrated the applicability of the static enclosure technique for fast BVOC emissions survey inside green retrofitted spaces. However, providing high sampling performance over entire BVOCs collection is recommended to reduce sampling error and incorrect estimation of the emissions.
