Planting materials, sample collection and preparation
PFAL kale trials were conducted at the Agrowlab plant factory in Bangkok, Thailand. The samples were collected 3 months after planting. A total of 32 buckets of kale were separated into 3 parts; 12:8:12, three varieties of PFAL kale were randomly collected from each part.
N-CF kale trials were conducted at a farm in Pathum Thani province, Thailand. There were 180 kale plants in the experimental plot. The N-CF kale samples were collected at 3, 6, and 12 months after planting. The kale plot was divided into three parts; three varieties of N-CF kale were randomly collected from each part of the plot.
An identical quantity of kale seeds was utilised in each agricultural system. In addition, leaf samples for the assessment of nutrition were collected throughout the same growth phases. The sets of kale samples were washed individually with water to remove dirt and contaminants. After washing, the edible part of each set will be cut into small pieces and homogenized. The homogenized samples are going to be divided into two portions for nutritional compositions and antioxidant contents analysis and stored at − 20°C until analysis.
Proximate Analysis
The proximate analysis of kale samples followed standardized methods according to the AOAC (George W. Latimer, 2019). The Kjeldahl method (AOAC 981.10) was used to determine protein content, and fat content following the Soxhlet method (AOAC 922.06). Moisture content was assessed by drying the samples at 100°C in a hot air oven, according to AOAC 925.45 A. Ash content was determined through incineration at 550°C in a muffle furnace (AOAC 920.153). Total dietary fibre analysis was determined by Enzymatic-Gravimetric Method (AOAC 985.29). The total carbohydrate content of the dry mass was calculated using the following formula:
Total carbohydrates (%) = 100 − (protein + total fat + moisture + ash + total dietary fibre)
Total Polyphenol Content
Total phenolic concentration (TPC) was determined according to the method described by Kongkachuichai et al. (2015), with slight modifications based on Brune et al. (1991). Two grams of samples were extracted with 20 ml of 50% dimethylformamide (Sigma–Aldrich, St. Louis, MO, USA) in 0.1 M acetate buffer for 16 h by constant shaking at room temperature. After extraction, samples were centrifuged (20 min, 25°C, 3000 rpm) and then supernatants were collected. The supernatants were diluted to 25 µl with water. Then the samples were added to a 96-well plate, followed by 125 µl of 10% Folin–Ciocalteau reagent (Merck, Darmstadt, Germany) and 100 µl 0.5 M sodium hydroxide to each well and mixed. Absorbance was measured at 750nm using an automated microplate reader (SunriseTM Tecan, Victory, Australia) after standing for 15 min at 25°C. Gallic acid (Sigma–Aldrich, St. Louis, MO, USA) was used as the standard with concentrations ranging from 0.00 to 80.00 ppm. Results were expressed as in milligram gallic acid (GAE) equivalents per 100 g of fresh weight (mg GAE/100 g). (Brune et al., 1991; Kongkachuichai et al., 2015)
Antioxidant activity
Antioxidant activity was analyzed by two methods, oxygen radical absorbance capacity (ORAC) assay and ferric reducing antioxidant power (FRAP) assay.
The FRAP solution was freshly prepared before analzing a sample. The solution consists of 0.3 M sodium acetate buffer solution (pH 3.6), 10 mM 2,4,6-tripyridyl-s-triazine (TPTZ, Sigma–Aldrich, St. Louis, MO, USA) in 40 mM of concentrated HCl and 20 mM FeCl3 (Sigma–Aldrich, St. Louis, MO, USA) at the ratio 10:1:1 (v/v/v). The sample was added to the FRAP solution and incubated at 37°C for 4 min. The absorbance of each sample was measured at 593 nm using a spectrophotometer (UV-1601 Shimadzu, Kyoto, Japan). The standard curve of Trolox (6-hydroxy-2,5,7,8 tetramethychroman-2-carboxylic acid, Sigma–Aldrich, St. Louis, MO, USA) was used to calculate the FRAP values of the samples. The results were expressed as micromole Trolox equivalents per 100 g of fresh weight (µmol TE/100 g). (Benzie & Strain, 1996; Kongkachuichai et al., 2015).
The ORAC assay was performed using a spectrofluorometer (PerkinElmer LS 55 luminescence spectrofluorometer, Perkin Elmer, Waltham, MA, USA) with excitation and emission wavelengths set to 493 nm and 515 nm, respectively. 6.25 to 100 µM trolox (6-hydroxy-2,5,7,8-tetramethychroman-2-carboxylic acid) was used as the standard. Briefly, a 500 µl sample, a 3.0 ml fluorescein solution (8.16 × 10− 2 µM), and 500 µl of AAPH (153 mM) were mixed in each tube. All dilutions were prepared using a 75 mM potassium phosphate buffer solution at pH 7.2. The ORAC value was calculated from a linear regression equation of net area under the curve (AUC) of the standard Trolox concentrations. Results were expressed as micromole Trolox (TE) equivalents per 100 g fresh weight (µmol TE/100 g fresh weight) (Huang et al., 2002; Kongkachuichai et al., 2015).
Statistical analysis
Statistical analysis was performed using Statistical Package for the Social Science (SPSS) program 18.0 (IBM Software, Chicago, II, USA). The significant differences are represented by P < 0.05.
An independent t-test was used to compare the results for the quantitative data (nutritional compositions, antioxidant contents, environmental impact, and carbon footprints) between the two groups, kales in PFAL and N-CF systems, to determine whether there is a statistically significant difference.
One-way analysis of variance (ANOVA) was used for the quantitative data (nutritional compositions, antioxidant contents, environmental impact, and carbon footprints) in the 4 sets of kale samples, kale grown in PFAL and three samples from N-CF systems with repeated observations at 3, 6, and 12 months.
Carbon footprint calculation
Goal and scope definition
This study analyzed the carbon footprints of kale production in PFAL and N-CF production systems, whereby the seeding process started in January 2022. PFAL kale production occurred from February 2022 to May 2022 (3 months), and N-CF kale production occurred from February 2022 to January 2023 (12 months). The cultivation practices commonly applied to the kale fields in this study differed slightly between the two systems:
-
PFAL: Kale was planted for around 3 months and harvesting occurred 45–55 days after planting. Subsequently, kales were collected around 3 times before being removed from the bucket.
-
N-CF: Kale was typically planted for around 12 months, and harvesting took place 45–55 days after planting. Subsequently, kales were collected weekly until 12 months after planting, after which they were plowed away from the field.
The production systems were analyzed from the gate-to-gate perspective. The functional unit (FU) in this study was a mass-balanced FU defined as 1 kg of edible kale during the planting cycle of PFAL and N-CF. Thus, the life cycle inventory included the material and energy requirements of agricultural activities, as well as the transportation of farm inputs. Figure 1 shows the system boundary as they were underpinned by different farm processes, except for seeding process which was similar in both harvest systems.
(Added Fig. 1)
Life cycle inventory
The inventory analysis quantified the inputs and outputs of the agricultural systems. Inputs and outputs data for both farming systems, including water, fertilisers and manure, pesticides, on-farm electricity use, transportation, and other factors related to kale production, were collected from the experimental farm through questionnaires. All inputs and outputs data were collected through follow-up questionnaires over different periods: 3 months for the PFAL farm and 12 months for the N-CF farm. The agricultural compositions of kale production studied in this work are shown in Table 1.
(Added Table 1)
Description of sites: The study site of PFAL was located at the Agrowlab plant factory in Khlong Kum, Bueng Kum District, Bangkok, Thailand (Fig. 2). The study site of N-CF was located on an experimental non-chemical farm in Khlong Song, Khlong Luang District, Pathum Thani province, Thailand (Fig. 2). Data are collected from a 3 m2 kale planting area. Although these two experimental sites were in different provinces, they have similar environmental conditions, including temperature and humidity. In this study, the following assumptions are summarized:
-
Energy consumption and CO2 emissions from manual labour are not included in this analysis.
-
In FPAL, only direct N2O emission released into the atmosphere was considered, as the mixed-chemical fertiliser was not applied to the soil.
Limitation of this study, N-CF was susceptible to flooding events due to the nature of open-field cultivation, which can damage or destroy the kale crop and impact production yields. Additionally, insect infestations pose a significant challenge in N-CF systems, potentially leading to crop damage and yield losses if not adequately managed, which may impact the overall carbon footprint analysis of the N-CF system.
(Added Fig. 2)
Calculation of Carbon Footprints from PFAL and N-CF
The carbon footprints of kale grown in PFAL and N-CF systems were calculated following the Intergovernmental Panel on Climate Change (IPCC) guidelines (Pörtner et al., 2019). Data pertaining to both systems were collected and validated for accuracy and reliability. Emissions from various sources, including energy consumption, transportation, production processes, and waste management, were calculated in accordance with the IPCC guidelines and the Thai National Life Cycle Inventory (LCI) Database (Thailand Greenhouse Gas Management Organization (Public Organization), 2022).
The Carbon footprint of kale farming systems was calculated using Eq. (1).
Carbon Footprint (kgCO2e) = EF×Q (1)
where EF is the emission factor for the activity (kgCO2e per unit) (Table 2) and Q is the quantity of the raw materials, chemicals, and other emissions (in units).
(Added Table 2)
Application of manure and synthetic fertiliser results in direct and indirect emissions of N2O. N2O emissions from N-CF came from nitrogen in non-chemical fertilisers applied in the field, including direct N2O emissions released into the atmosphere, indirect N2O emissions from atmospheric deposition of volatilized N, and indirect N2O emissions from nitrogen leaching and runoff. N2O emissions from the use of nitrogen fertiliser are represented by Eq. (2).
N2O Emissions (kgN2O) = N2Odirect-Ninputs+N2Odirect-Nos+N2Oindirect−volatilize+N2Oindirect−leaching (2)
where N2Odirect-Ninputs represent direct N2O emissions released into the atmosphere from organic N additions to soils, N2Odirect-Nos represent direct N2O emissions released from managed areas annually, N2Oindirect − volatilize represent indirect N2O emissions from atmospheric deposition of volatilized N, and N2Oindirect − leaching represent indirect N2O emissions form nitrogen leaching and runoff. (Eggleston et al., 2006)