Volume 47, Issue 19 e2020GL089373
Research Letter
Free Access

Global Methyl Halide Emissions From Rapeseed (Brassica napus) Using Life Cycle Measurements

Yi Jiao

Corresponding Author

Yi Jiao

Department of Geography, University of California Berkeley, Berkeley, CA, USA

Correspondence to:

Y. Jiao and R. C. Rhew,

[email protected];

[email protected]

Search for more papers by this author
Jerrold Acdan

Jerrold Acdan

Department of Geography, University of California Berkeley, Berkeley, CA, USA

Search for more papers by this author
Rong Xu

Rong Xu

Department of Environmental Science, Policy and Management, University of California Berkeley, Berkeley, CA, USA

Search for more papers by this author
Malte Julian Deventer

Malte Julian Deventer

Department of Geography, University of California Berkeley, Berkeley, CA, USA

Now at Bioclimatology, University of Göttingen, Göttingen, Germany

Search for more papers by this author
Wanying Zhang

Wanying Zhang

Anhui Province Key Laboratory of Polar Environment and Global Change, School of Earth and Space Sciences, University of Science and Technology of China, Hefei, China

Search for more papers by this author
Robert C. Rhew

Corresponding Author

Robert C. Rhew

Department of Geography, University of California Berkeley, Berkeley, CA, USA

Department of Environmental Science, Policy and Management, University of California Berkeley, Berkeley, CA, USA

Correspondence to:

Y. Jiao and R. C. Rhew,

[email protected];

[email protected]

Search for more papers by this author
First published: 21 September 2020
Citations: 5

Abstract

Global budgets of methyl halides are not balanced between currently identified sources and sinks. Among biological sources, rapeseed is regarded as the second largest terrestrial source of CH3Br, extrapolated from laboratory-based incubations and limited field measurements. This study analyzes the CH3Br budget from rapeseed (Brassica napus “Empire”), using field-based life cycle measurements, yielding a globally scaled emission rate of 2.8 ± 0.7 Gg year−1. Though this verifies that rapeseed is a significant global source, it is just half of the previous estimation, even after accounting for the doubling of global annual rapeseed production since then. The ozone-depleting potential of rapeseed is further sustained through CH3Cl and CH3I emissions, which were measured for the first time and scaled to 5.3 ± 1.3 and 4.0 ± 0.8 Gg year−1 globally.

Key Points

  • Brassica napus (rapeseed) is a smaller global source of atmospheric CH3Br than previously believed, estimated at 2.8 ± 0.7 Gg year−1 in 2018
  • Rapeseed also emits an estimated 5.3 ± 1.3 Gg year−1 of CH3Cl and 4.0 ± 0.8 Gg year−1 CH3I to the atmosphere over the same time period
  • Emissions of ozone-depleting substances from B. napus croplands continue to increase, becoming an ever-larger global source

Plain Language Summary

Stratospheric ozone absorbs incoming solar UV radiation, attenuating the harmful radiation exposure for life on Earth's surface. Halogen atoms transported via halocarbons, including methyl halides, can catalyze ozone destruction efficiently in the stratosphere. Anthropogenic sources of halocarbons have been decreasing consistently since the implementation of the 1987 Montreal Protocol and its amendments. However, some natural sources, especially those influenced by anthropogenic activities, may offset some of the achievement of reduced halocarbon emissions. This study quantifies methyl halide emissions from cultivated rapeseed (Brassica napus, cultivar: Empire), based on life cycle measurements and normalized to seed production. This yields a global crop contribution of 2.8 ± 0.7 Gg of methyl bromide (CH3Br) annually, which is smaller than previous estimates (5.1–6.6 Gg), supporting the conventional view that there must be other unidentified or underestimated sources for CH3Br. This study also quantifies for the first time that rapeseed emits 5.3 ± 1.3 Gg of methyl chloride (CH3Cl) and 4.0 ± 0.8 Gg of methyl iodide (CH3I) each year. Due to the increasing demand on rapeseed products such as canola oil, its global methyl halide emissions are expected to grow in the future.

1 Introduction

As a result of the Montreal Protocol, atmospheric levels of most anthropogenic ozone-depleting halocarbons (e.g., CFCs, HCFCs, carbon tetrachloride, and methyl chloroform) have decreased (Montzka et al., 1996, 2003), leaving halocarbons from natural sources relatively more pronounced, especially those influenced by human activities and climate change (Liang et al., 2017). Among the halocarbons with natural sources, methyl chloride (CH3Cl) and methyl bromide (CH3Br) are presently the largest carriers of chlorine and bromine, respectively, to the stratosphere. Methyl iodide (CH3I) also contributes to ozone destruction, though mostly in the lower troposphere due to its relatively shorter lifetime of 12–15 days (Zhang et al., 2020).

Known methyl halides sources include biomass burning (Andreae et al., 1996), tropical forests (Bahlmann et al., 2019; Xiao et al., 2010; Yokouchi et al., 2007), salt marshes (Drewer et al., 2006; Manley et al., 2006; Rhew et al., 2000), wetlands (Lee-Taylor et al., 2001; Varner et al., 1999), oceans (Carpenter et al., 2014; Hu et al., 2013), fungus (Lee-Taylor et al., 2001; Lee-Taylor & Holland, 2000), agricultural fields (such as rapeseed, Gan et al., 1998; Mead et al., 2008; rice paddies, Lee-Taylor & Redeker, 2005; Redeker et al., 2000), and anthropogenic sources (such as coal combustion, McCulloch et al., 1999, and the use of CH3Br as a fumigant for quarantine and preshipment uses, Carpenter et al., 2014). Estimated magnitudes of these sources do not balance the known sinks (Carpenter et al., 2014), including loss to the stratosphere, reaction with OH radicals, and degradation in oceans and soils (bidirectional interfaces), which suggests that either missing or underestimated sources exist. The magnitude of the “missing” sources for CH3Cl and CH3Br are 748 and 39 Gg year−1, equivalent to 20% and 46% of all known sources, respectively; global CH3I emissions are also not well constrained (Tables 1-4 and 1-8 in Carpenter et al., 2014).

The second largest estimated biological terrestrial source of CH3Br is the global crop of rapeseed (Brassica napus), with an estimated global emission of 5.2 Gg year−1 (Carpenter et al., 2014). Plants of the Brassicaceae family were first reported to have large emission potentials for methyl iodide (CH3I) based on a survey of floating leaf discs of 118 herbaceous species in KI solutions (Saini et al., 1995). Among several species of Brassica plants, B. napus emitted the largest amount of CH3Br per unit biomass, and the global crop flux scaled to 6.6 Gg year−1 (Gan et al., 1998). However, these studies were limited to juvenile plants in laboratory-based incubations or snapshot measurements in the field. Rapeseed has a usual lifetime of 130–160 days (Buntin et al., 2007; Diepenbrock, 2000) with physiological factors at different life stages varying significantly, such as leaf area index, flower number, pocket number, biomass, plant height, and so forth (Bouttier & Morgan, 1992; Morrison et al., 1992). It is expected that rapeseed, similar to other biological emitters, would reveal varied methyl halide emission levels over the life cycle (Deventer et al., 2018; Khan et al., 2013; Manley et al., 2006; Redeker et al., 2000). Extrapolating a global CH3Br budget after incorporating the life cycle variations in emissions may improve estimates.

Using the laboratory-based measurements of juvenile plants and field measurements from Gan et al. (1998), Mead et al. (2008) applied the global crop harvest index and life length to extrapolate the rapeseed flux to yield 5.1 Gg year−1. Rapeseed is the third largest source for vegetable oil (Liu et al., 2016), and its production (20-fold increase) and growing area (sixfold increase) have been consistently increasing over the last six decades (FAOSTAT Database, supporting information Figure S1). Therefore, methyl halide emissions from rapeseed have likely increased with time and since the most recent quantification was conducted (Mead et al., 2008).

Biological methyl halide production can occur when S-adenosyl-l-methionine (SAM) donates methyl groups to bond with halogen ions to form methyl halides and with bisulfide to produce methane thiol, reactions catalyzed by SAM-halide/thiol methyltransferases (Attieh et al., 2000; Bayer et al., 2009; Rhew et al., 2003). Thus, rapeseed should have the potential to produce the other ozone destroying compounds, CH3Cl and CH3I. Improved estimates of all of these methyl halide emissions from rapeseed will help predict the impact of this crop expansion on the future of stratospheric ozone.

In this study, methyl halide emissions were measured from the whole plants of B. napus “Empire” over their entire life cycle, as well as from soil samples collected from the field. These results are used to provide an updated global CH3Br budget and the first estimates of CH3Cl and CH3I emissions with respect to rapeseed.

2 Methods

2.1 Site Description and Rapeseed Cultivation

The cultivation of rapeseed (Brassica napus “Empire”) was conducted both in the greenhouse and in the field at Oxford Tract (37°52′31″N, 122°16′01″W), a University of California, Berkeley facility located approximately 5 km east of San Francisco Bay. The greenhouse was vented and was not temperature controlled during the experiment. The field plot (6 m × 30 m) was located at the southeastern corner of the tract, adjacent to cultivated corn crops. The tract was plowed regularly each year and was irrigated roughly twice a week with treated tap water from the East Bay Municipal Utility District during the dry season (July to September). Air temperature was recorded for every 10 min via sensors (HOBO® U23 Pro v2 Data Logger, Onset Computer Corporation, Bourne, MA, USA).

On 3 July 2015, uncoated seeds (i.e., no neonicotinoids) of B. napus (Empire, non-GMO), from the Brassica Breeding and Research group, University of Idaho, were sowed into six different trays (L × W × D: 56 cm × 56 cm × 30 cm) of soil. On 5 August 2016, B. napus seeds were sowed in the field garden (Figure S2). Prior to both experiments, the soil above 20 cm depth was removed and autoclaved at 145°C for 6 hr to destroy the viability of weed seeds and pests and returned either to the field or translocated to the trays. Open-ended aluminum chamber bases (L × W × H: 52 cm × 52 cm × 22.5 cm) were inserted 10 cm into the soil and remained there for the duration of the growing season for static-chamber enclosure measurements. For the field measurements only, leaf area, plant height, flower numbers, and fruit pocket numbers, if available, were recorded each week over the life cycle (Figure S3), and the above-ground biomass was harvested and dried after the seeds were collected at the end of the experimental cultivation. The harvest data for the greenhouse measurements were lost; thus, only the diurnal studies (see below) were used for the derivation of normalized flux responses to temperature (Figure S4), while the weekly fluxes were used to plan for the field study (Table S1).

2.2 Gas Collection and Analysis

Starting from the 5th day after sowing, static flux chamber measurements were conducted over two rapeseed plots/trays (n = 2) each week until the end of the life cycle (Table S1). For headspace air sampling, an aluminum chamber lid (L × W × H: 53.5 cm × 53.5 cm × 44.5 cm) was placed on top of an aluminum base, which was rimmed with aluminum channels (width: 3 cm) filled with deionized water, making an airtight seal. Gas samples were collected into previously evacuated canisters (1-L electro-polished stainless steel [Lab Commerce, San Jose, CA] or 1-L or 3-L silica-lined stainless steel [Restek Corporation, Bellefonte, PA, USA]) through stainless-steel tubing at 2, 17, and 32 min after the chamber closure. The entire chamber was covered by reflective insulation to maintain a stable internal temperature. Air inside the chamber was well mixed with two electric fans (0.012 m3 min−1 each) on opposite ends, angled in different directions to enhance circulation. The temperature within the chamber was recorded with stainless-steel thermocouple dataloggers (iButtons, Maxim Inc., Sunnyvale, CA, USA). When the rapeseed plants grew higher than the chamber lid, an intermediate connection piece (L × W × H: 54.5 cm × 54.5 cm × 45 cm) was placed in between the base and the lid to increase the overall height to 1.2 m; at the maximum plant heights (1.3 m), stems were bent gently to avoid breakage.

Methyl halide concentrations were analyzed by gas chromatography coupled with mass spectrometry (GC/MS; Agilent 6890N/5973, Agilent Technologies, Santa Clara, CA, USA) with a custom preconcentration system (Jiao et al., 2018). Each gas sample was analyzed at least twice consecutively, and the concentration was reported as the average ± standard deviation in units of ppt (parts per trillion). Before and after the measurements of each batch of gas samples, calibration curves and daily instrumental drift corrections were constructed using a whole-air standard collected at Niwot, Colorado, and calibrated at the Global Monitoring Division Laboratory of the National Oceanographic and Atmospheric Administration (B. Hall, pers. comm., 13 June 2014). The averaged instrumental precisions for CH3Cl, CH3Br, and CH3I after applying drift corrections were 1.0%, 1.9%, and 8.7%, respectively.

2.3 Flux Calculation and Approximation

Fluxes were calculated (Equation 1) by linear regression between methyl halide concentrations and the enclosure time (dc/dt), converted to a net rate of change of moles of trace gas in the chamber (Na) and then normalized to the basal area of the chamber (A) and expressed in the units of micromoles per square meter per day (μmol m−2 day−1).
urn:x-wiley:00948276:media:grl61261:grl61261-math-0001(1)

Methyl halide emissions have been shown to vary dramatically with stage of plant growth (Deventer et al., 2018; Khan et al., 2013; Redeker & Cicerone, 2004). Therefore, the life cycle of rapeseed was segmented into four growing stages in this study: leafing period (from sowing to the maximum leaf area), flowering period (to the maximum flower number), fruit-bearing period (to when the seed pods started to dry out), and the senescence period (to harvest).

For the flux chamber measurements in the greenhouse, a 24-hr study was conducted at 3- or 4-hr intervals starting at 0:00 on a randomly selected day within each of the four growing stages to capture the diurnal relationships between fluxes and air temperature. Assuming the biogenic methyl halide fluxes have an exponential relationship with the air temperature (Deventer et al., 2018; Guenther et al., 2012), an equation can be proposed as,
urn:x-wiley:00948276:media:grl61261:grl61261-math-0002(2)
where Fi represents the flux of the gas species i, Temp is the air temperature within the chamber when the flux measurement was carried out, and p and q are unknown coefficients, which can be derived from fitting with exponential regression (R2 > 0.90).
Furthermore, assuming a similar methyl halide flux response to the temperature both in the greenhouse and in the field, the daily averaged methyl halide fluxes can be calculated by the exponential model with the following equation:
urn:x-wiley:00948276:media:grl61261:grl61261-math-0003(3)
where Fday represents the average flux over the course of a day, Ti represents air temperature measured at the ith hour of the day, and TMeas. and FMeas. represent the temperature within the chamber when measurement was carried and the corresponding trace gas flux, respectively.

After the autoclaved soil was returned back to the field, intact soil cores (diameter: 4.5 cm; depth: 10 cm) from adjacent sites undergoing similar treatment were collected with a slide hammer (AMS, Inc., American Falls, ID, USA) during each growing stage. The cores were subsequently incubated in the laboratory in glass jars (V = 1.9 L) to derive the background methyl halide fluxes from the soil by analyzing the changing methyl halide concentrations in the headspace over enclosure time.

2.4 Emission Extrapolation

Studies have shown that the leafing area, floral initiation, floral bud number, and yield of rapeseed are proportional to each other and to the accumulated air temperatures (Luo et al., 2018; Wright et al., 1988). Therefore, to correct for the local temperature bias, the weekly methyl halide fluxes were integrated over the plant life span and then normalized to the weight of the seeds harvested for enclosed plants at the end of the experiment to derive a total “flux per gram of seed.” This is multiplied by the global seed harvest, assuming that the harvest-normalized methyl halide production rates in this study are representative of crops globally. An unquantified uncertainty is the variation of methyl halide emissions between different plant cultivars, as only one cultivar was analyzed in this study. Cultivar has been shown to be a significant variable in rice plants (Redeker & Cicerone, 2004). The annual data of global/regional rapeseed production, yield, and cultivation area since 1961 are accessible to the public from the online Food and Agriculture Organization Corporate Statistical Database.

3 Results and Discussion

3.1 Methyl Halide Fluxes From Plant-Soil System

The average soil fluxes of CH3Cl, CH3Br, and CH3I were 66 ± 33, 2.1 ± 0.3, and 17 ± 24 nmol m−2 day−1 during the leafing period, 98 ± 163, 2.8 ± 3.9, and 20 ± 4 nmol m−2 day−1 within the flowering period, −172 ± 37, 1.6 ± 4.3, and 5.6 ± 0.9 nmol m−2 day−1 over the fruiting period, and −211 ± 97, 2.8 ± 0.8, and 12.0 ± 0.1 nmol m−2 day−1 at the senescence period, respectively. Hence, CH3Br and CH3I fluxes from soil were relatively consistent, while CH3Cl gradually switched from net emissions to net absorption over the experiment. Soil is usually a bidirectional interface for methyl halides with abiotic production and biotic degradation happening concurrently (Keppler et al., 2020). The switch from CH3Cl soil sink to source may be a result of the initial autoclave process killing methyl halide consuming microbes, which slowly recolonized the soils over the season.

Flux measurements from cultivated rapeseed (B. napus) plots were much larger than incubated soil fluxes, roughly 10, 100, and 20 times larger in magnitude for CH3Cl, CH3Br, and CH3I, respectively. CH3Cl, CH3Br, and CH3I emissions all had similar seasonal patterns over the life cycle of the crop, with the peak emissions observed during the flowering period (Figure 1). The average fluxes (note different units) of CH3Cl, CH3Br, and CH3I during the leafing period were 0.29 ± 0.33, 0.06 ± 0.09, and 0.13 ± 0.18 μmol m−2 day−1; then climbed up to 2.04 ± 0.46, 0.65 ± 0.19, 0.57 ± 0.18 μmol m−2 day−1 within the flowering period; decreased to 0.77 ± 0.40, 0.19 ± 0.06, and 0.17 ± 0.07 μmol m−2 day−1 over the fruiting period; and dropped further to 0.12 ± 0.03, 0.02 ± 0.01, and 0.03 ± 0.03 μmol m−2 day−1 at the senescence period, respectively.

Details are in the caption following the image
Life cycle emissions of methyl chloride (CH3Cl), methyl bromide (CH3Br), and methyl iodide (CH3I) from rapeseed (B. napus) normalized to per unit area. Error bars represent the standard deviations of the replicates (n = 2). The four different background colors represent the four life stages as indicated by the drawings: leafing, flowering, fruiting, and senescence. The open diamond symbols represent the background soil fluxes based on the controlled laboratory incubations.

Fluxes of methyl chloride (CH3Cl) and methyl bromide (CH3Br) both showed strong correlations with air temperature (Figure S4, Table S3) within each plant life stages (R2 > 0.90). Similar results were found for perennial pepperweed (Lepidium latifolium), a related species within the Brassicaceae family, which increased with temperature toward an optimum temperature at 33°C, consistent with enzymatically mediated production by the plants (Deventer et al., 2018). This is also consistent with measurements from rice plants where biogenic production of methyl halides correlated with temperature over a large range, with an optimum temperature around 36°C (Redeker & Cicerone, 2004). It is proposed that SAM acts as a methyl donor to halogen ions to form the respective methyl halides catalyzed by the methyltransferase enzyme (Rhew et al., 2003; Saini et al., 1995). Studies reported that some methyltransferases, such as nicotinic acid-N-methyltransferase from soybeans (Glycine max) and (S)-tetrahydroprotoberberinecis-N-methyltransferase from Eschscholtzia and Corydalis, have an optimum temperature at 40°C and remain stable up to 45°C (Chen & Wood, 2004; Rueffer et al., 1990). However, the temperature during the experimental period of this study was almost entirely under 33°C (Figure S5). Therefore, impacts of high temperature could either enhance or reduce enzymatic activity, but these are not accounted for in this study. In addition, the temperature sensitivity of CH3Cl and CH3Br emissions varies across the life stages: leafing, flowering, fruiting, and senescence (Table S3), which suggest that life cycle scale methyl halide emissions from rapeseed were predominantly influenced by physiological factors.

3.2 Global Methyl Halide Emissions From Rapeseed

CH3Br emissions from several Brassica plants were found to be linearly correlated to soil halide content (Gan et al., 1998). In the present study, soil bromide content was 1.1 ± 0.3 mg kg−1, similar to the typical bromide content (1.0 mg kg−1) in natural soils globally (Flury & Papritz, 1993). Allometric factors of rapeseed may be influenced by several factors, including local climate, agrarian techniques and chemical inputs, such as fertilizer, pesticide, and so forth. As a consequence, the yield and density of rapeseed may vary significantly across the producing regions (e.g., Canada, China, Europe, and India) with different local climates. The single cultivar (Empire, non-GMO) used in this experiment may show different methyl halide emission patterns compared to other genotypes. However, studies have shown that leafing area, floral bud numbers, and yield of rapeseed are proportional to each other and to the accumulated air temperatures (Luo et al., 2018; Wright et al., 1988). Rapeseed in this experiment showed a harvest index of 24.4 ± 2.2, which was close to those (21.3 ± 1.5) of hundreds of other genotypes of rapeseed (Lu et al., 2016). Therefore, methyl halide emissions per gram of seed harvested (corrected for soil fluxes) were calculated as a means to partially offset the locality bias in plant vitality. Daily fluxes were calculated based on the weekly flux measurements and temperature relationships (Equation 3) and summed over the plant life span. The integrated life cycle emissions were then divided by the fresh weight of the seeds to determine total emissions per gram of seed produced. This method provides both a means for intersite comparison as well as for global extrapolation using agricultural commodity data.

In this experiment, methyl halide production rates by seed biomass were 71.0 ± 17.1, 37.8 ± 9.8, and 53.2 ± 11.4 μg g−1 for CH3Cl, CH3Br, and CH3I, respectively. Assuming these parameters are representative globally, 2.8 ± 0.7 Gg CH3Br were emitted from global cultivated rapeseed in 2018. This is smaller than the prior estimations, which concluded rapeseed emitted 5.1–6.6 Gg of CH3Br each year (Gan et al., 1998; Mead et al., 2008). The lower value is determined despite the fact that the global production of rapeseed has more than doubled as well as the growing area has increased over 1.5-fold from 1998–2003 to 2018.

The major reason for the discrepancy involves the method of extrapolation. Prior extrapolations (Gan et al., 1998; Mead et al., 2008) were based on CH3Br fluxes in low bromide soils (25 ng g−1 dry wt above ground biomass day−1), which scaled up 20-fold to 515 ng g−1 day−1 assuming a global average soil bromide content of 1 mg kg−1. The end result was roughly 9 times the life cycle average flux reported here despite similar soil bromide contents (Table 1). On the other hand, numerous other factors remain to be accounted for, including variations in cultivar, agricultural practices, climate, and soils. The low number of replicates (n = 2) also poses substantial uncertainties in representing the true mean flux for the field site. Based on studies of chamber variance in rice paddies, we estimate that these measurements are within 30% of the field means (Redeker et al., 2002).

Table 1. The Comparison of Measured or Extrapolated Methyl Bromide (CH3Br) Fluxes Per Unit Above-Ground Biomass From Rapeseed at Different Life Stages
Comparison of CH3Br fluxes Prior studiesa This studyb
Period of measurements Leafing Leafing Flowering Fruiting Senescence Life cycle average
Unit (ng g−1 day−1 [dry weight]) 25 82.0 ± 32.4 89.2 ± 50.5 16.8 ± 6.0 2.8 ± 1.8 55.9 ± 48.6
515
  • a The previous studies include Gan et al. (1998) and Mead et al. (2008).
  • b The averaged soil bromide content of this study was 1.1 ± 0.3 mg kg−1. CH3Br fluxes of this study listed in the table were normalized to per unit above-ground biomass in order to provide comparison to prior studies. However, the global extrapolation of this study was calculated from fluxes per unit of seeds harvested.

Interestingly, this study yielded a much larger average emission rate for CH3Br per unit above-ground biomass than the prior study, which was based on the leafing phase only. The life cycle average was twice that of the prior study. Considering the leafing phase only, the emission rate in this study (82 ng g−1 dry wt day−1) was over threefold the prior study. Thus, while this study supports the assumption of higher emissions with higher soil bromide content, it does not concur with the magnitude of scaling up.

Even with smaller CH3Br annual fluxes than previously estimated, rapeseed remains one of the largest nonindustrial terrestrial sources of CH3Br, next to biomass burning, 23 Gg year−1 (Andreae et al., 1996; Blake et al., 1996), salt marshes, 8–29 Gg year−1 (Deventer et al., 2018; Manley et al., 2006; Rhew et al., 2000, 2014), and wetlands, 4.6 Gg year−1 (Varner et al., 1999). However, CH3Br budgets with respect to rapeseed still cannot balance the current identified and quantified sources and sinks, with the missing sources estimated to be about half the magnitude of all the known sources put together.

The average chloride content in soils globally is about 100 mg kg−1 (Geilfus, 2019), which is slightly lower than that of the soil (148 ± 71 mg kg−1) in this experiment. Using the same extrapolation method and assumptions yields a global CH3Cl source from rapeseed of 5.3 ± 1.3 Gg in 2018, which is comparable to that of rice paddies, 2.4–5.8 Gg year−1 (Lee-Taylor & Redeker, 2005; Redeker et al., 2000), mangroves, 12 Gg year−1 (Manley et al., 2007), and at the lower limit of that of salt marshes, 10 Gg year−1 (Deventer et al., 2018). However, rapeseed itself cannot bridge the considerably large CH3Cl gap between currently identified sources and sinks.

As for CH3I, rapeseed is estimated to account for an annual flux of 4.0 ± 0.8 Gg year−1, which is a minor source. The ocean is believed to be the largest CH3I source, with an average flux of 371.4 Gg year−1, predominantly through abiotic photochemical production (Butler et al., 2007; Stemmler et al., 2014) and/or biogenic emissions from phytoplankton and algae (Manley & de la Cuesta, 1997; Smythe-Wright et al., 2006), depending on the region or season. The largest identified terrestrial source for CH3I is rice paddies, with a quantified flux of 16 to 71 Gg year−1 (Lee-Taylor & Redeker, 2005; Redeker et al., 2000). However, a CH3I profile measurement study at North Carolina (Sive et al., 2007) suggested the existence of midlatitude terrestrial sources of 33 Gg year−1, in which rapeseed (B. napus) would be considered as one of them. Other ubiquitous species among the Brassicaceae family, such as Brassica oleracea (e.g., broccoli, brussels sprout, cabbage, cauliflower, collards, and kale), Brassica rapa (e.g., bok choy, Chinese cabbage, and turnip), Brassica juncea (mustard), Lepidium latifolium (e.g., perennial pepperweed), and so forth, have been reported as significant emitters for CH3Cl and CH3Br (Deventer et al., 2018; Gan et al., 1998; Khan et al., 2013). Thus, it is rational to speculate that they may account for part of the terrestrial sources of CH3I as well. However, measurements of CH3I from these widespread species have not been reported, and their global budget extrapolations have not been established as a consequence. In contrast to CH3Cl and CH3Br, currently identified and quantified sources of CH3I exceed the sum of the sinks, with a positive imbalance of ~146 Gg year−1, suggesting that either the oceanic CH3I production may be overestimated and/or the photolysis sinks of CH3I may be underestimated.

Atmospheric CH3Br concentrations have an interhemispheric ratio of 1.130 (NH/SH, 2017) (HATS Data Archive, Global Monitoring Laboratory, & National Oceanic & Atmospheric Administration, 2019), suggesting the unevenly distributed regional sources with the major sources located in the Northern Hemisphere, especially in the mid northern latitudes. This is in agreement with a hypothesis of the existence of significant terrestrial sources or biogenic sources. Given the geographical distribution of rapeseed, methyl halides from this crop are mainly emitted in the Northern Hemisphere (93.6%, 2018), which will contribute to this interhemispheric gradient. However, the updated CH3Br budget of rapeseed itself cannot account for the interhemispheric difference due to its size. The dominant emission of CH3Cl from rapeseed from the Northern Hemisphere also has a negligible impact on the interhemispheric gradient of CH3Cl. Atmospheric CH3Cl displayed a relatively even distribution across the hemispheres with the peak concentrations observed at the intertropical zone (HATS Data Archive, Global Monitoring Laboratory, & National Oceanic & Atmospheric Administration, 2019), suggesting the predominant sources are located near the equator.

Global estimates of rapeseed emissions will be improved after additional measurements are conducted at different agricultural sites and with different cultivars. This study demonstrates the importance of life cycle measurements to determine overall emissions per seed crop biomass, a key metric to extrapolate results over time. Emission rates by biomass are highest during the leafing and flowering stage and are highest by surface area at the flowering stage. This study also demonstrates the relative importance of rapeseed on the global CH3Cl and CH3I budgets.

The increase of rapeseed harvest area and production over the past 60 years has resulted in increased emissions of CH3Cl, CH3Br, and CH3I by 21-fold between 1961 and 2018 (Figure 2). The demand for rapeseed oil in the world is projected to grow even larger in the future (Jat et al., 2019). Using the extrapolations developed here and assuming the future global rapeseed production follows the same rate of increase as the past 60 years, the atmospheric sources of CH3Cl, CH3Br, and CH3I may increase to 13.1, 7.0, and 9.8 Gg year−1, respectively, by 2050, highlighting the increasing contribution from natural sources of ozone depleting substances influenced by human activities. Noncontrolled anthropogenic halocarbon emissions (Dhomse et al., 2019; Li et al., 2017; Montzka et al., 2018), together with the projected increasing of natural sources associated with global warming (Smythe-Wright et al., 2006; Yokouchi et al., 2001), land use change (Deventer et al., 2018; Mead et al., 2008), biomass burning (Blake et al., 1996; Westerling et al., 2006), and sea level rise (Jiao et al., 2018; Wang et al., 2016), may offset some of the halocarbon emission reductions achieved by the Montreal Protocol (Liang et al., 2017).

Details are in the caption following the image
Extrapolated global methyl chloride (CH3Cl), methyl bromide (CH3Br), and methyl iodide (CH3I) budgets with respect to rapeseed (B. napus) from 1961 to 2018. Estimated emissions from the Northern Hemisphere (shaded bars) greatly exceed those from the Southern Hemisphere (white bars).

Acknowledgments

This project was funded by the National Science Foundation (ATM-1258365). The authors would like to thank the Brassica breeding and research group at the University of Idaho for providing the seeds; Christina Wistrom for coordinating field work at the Oxford Tract; Wenchen Liu, Yujia Tao, and Connor Shingai for assistance on sampling; Bernard Koh, Anna Mikheicheva, and Ross Ward for GC/MS measurements; and Dennis Zellmann for producing rapeseed drawings. This paper was substantially improved by incorporating the constructive comments from two anonymous reviewers.

    Data Availability Statement

    Rapeseed global annual production data can be accessed through the Food and Agriculture Organization Database (http://www.fao.org/faostat/en/#data/QC). The data set of weekly methyl halide fluxes, rapeseed physiological variables, soil halide contents, and temperature is available online (https://doi.org/10.6078/D1WD8H).