Emissions of Volatile Organic Compounds from a Dairy Cattle Shed in Japan

To clarify the nature and characteristics of volatile organic compounds (VOCs) emitted from dairy cattle within a cattle shed located in Chiba, Japan, air samples were collected and analyzed for the four seasons in 2017–2018. Thirty-four VOCs were determined by gas chromatography-mass spectrometry and high performance liquid chromatography. In addition, air temperature and relative humidity inside and outside of the shed were monitored during each sampling campaign to estimate the ventilation rate of the shed. The average concentrations of total VOCs (μg m−3) in the shed in each season were 50.5 (spring), 128.4 (summer), 168.8 (autumn), and 199.5 (winter). Ketones were always the most dominant components followed by alcohols and volatile fatty acids (VFAs). The sum of ketones, alcohols, and VFAs accounted for more than 80% of the total VOCs in all seasons. Acetone, 3-pentanone, 1-butanol, and acetic acid were the major components regardless of the season, accounting for more than 60% of the total VOCs. The average emission rates of total VOCs from the shed (μg h−1 kg−1) were calculated to be 623 (spring), 1520 (summer), 585 (autumn) and 469 (winter). The emission rates of almost all the VOCs except alcohols increased exponentially with increase of air temperature in the shed. The ranges of the emission rates for each class of chemical (μg h−1 kg−1) were 39–170 (VFAs), 247–913 (ketones), 65–134 (alcohols), 40–122 (phenols), 10–122 (aldehydes), 4.17–22.3 (sulfur compounds), and 0.0067–0.74 (indoles). Furthermore, the annual emissions of VOCs for a single dairy cattle and for the cattle shed were estimated to be 5.5 kg and 44 kg, respectively.


INTRODUCTION
Volatile organic compounds (VOCs) are one of the most prominent class of chemicals in the atmosphere.VOCs in the atmosphere impact on climate, the atmospheric environment, and human health.Some of the ambient VOCs are known to cause odor pollution near the emission source (Shusterman, 2013;Parker et al., 2012), and have the potential to cause cancer or other health impacts such as headache, respiratory disease, and neurological disorders (e.g., Akdeniz et al., 2013).In addition, gas-phase oxidation of VOCs leads to the formation of secondary organic aerosols (SOAs) (e.g., Camredon et al., 2007).Such SOAs may influence cloud formation and, especially for polar organic compounds such as those with carboxyl groups, have the potential to act as cloud condensation nuclei, which can form cloud droplets (Nga et al., 2005).In addition, the presence of atmospheric VOCs causes a rise in the concentration of atmospheric ozone (World Health Organization, 2000).Ozone in the troposphere is a prevalent photochemical oxidant, is known to be highly phytotoxic (Krupa and Manning, 1988), and has impacts on human health (e.g., Finlayson-Pitts and Pitts Jr., 1997).In addition, tropospheric ozone contributes to the formation of airborne toxic chemicals (e.g., Finlayson-Pitts and Pitts Jr., 1997).Therefore, VOCs are of great concern from the perspective of human health and environmental management.
Recently, the livestock industry has been reported to be one of the major sources of atmospheric VOCs in the USA (Hafner et al., 2010;Howard et al., 2008;Shaw et al., 2007), Europe (Sintermann et al., 2014), China (Qi et al., 2017;Qiu et al., 2014;Fu et al., 2013), and India (Varshney and Padhy, 1998).Livestock feed-yard operations are major stationary source of odorous VOCs emissions from agriculture and cause public odor contamination in the USA (Lu et al., 2008).Borhan et al. (2012) studied seasonal emissions of phenol and p-cresol, the primary odorous VOCs, in a dairy operation in central Texas, and reported that emission rates (ERs) for phenol and p-cresol amounted to several hundred to several thousand mg head -1 day -1 , respectively.Chung et al. (2010) evaluated non-methane VOC (NMVOC) emissions from dairy operations in California, and indicated that silage and total mixed rations were the dominant sources of VOCs.
The ER, flux, and emission inventory of VOCs from livestock have been estimated in various studies (Qi et al., 2017;Qiu et al., 2014;Fu et al., 2013;Borhan et al., 2012;Hu et al., 2012;Hales et al., 2012;Parker et al., 2010;CDPR, 2006;Varshney and Padhy, 1998).For example, in California, VOCs emitted from the stockbreeding industry have been estimated to account for 9.6% of the total VOC emissions (CDPR, 2006).Hence, VOCs from livestock have the potential to adversely impact the environment.In fact, Hu et al. (2012) estimated that the O 3 formation potential of VOCs due to animal feed emissions was almost at the same level as that of mobile-source VOC emissions.
In contrast to the situation in the USA, VOCs emitted from livestock in Japan have rarely been reported.In 2018 for example, approximately 1.3 million head of dairy cattle, 2.5 million head of beef cattle, 9.2 million head of swine, 139 million head of hens, and 139 million head of broiler chicken were reared in Japan ( Japanese Ministry of Agriculture, Forestry and Fisheries, 2018).Therefore, the VOC emissions from these livestock may adversely impact on human health and the environment in Japan.Previously, we reported the concentrations, compositions, and seasonal variations of VOCs emitted from swine in Japan and showed that the ER of VOCs from a swine shed was about 1-2 × 10 3 μg (kg-swine) -1 (Osaka et al., 2018).On this basis, the total annual emissions of VOCs from one swine shed was estimated to be on the order 10 3 g year -1 .As indicated above, about 9.2 million head of swine are reared on an annual basis.Hence it is considered that the nation's annual emissions of VOCs due to swine rearing is not negligible.
In this study, emissions of VOCs from dairy cattle in Japan have been studied with the aim of clarifying the concentrations, compositions, seasonal variations, and ERs for the VOCs.

2 Sampling Site
Air samples from a cattle shed were collected at Asahi Agricultural High School (latitude 35°43 00 N, longitude 140°39 36 E) located in the northeast of Chiba Prefecture, Japan.A schematic of the shed is given in Fig. 1.The volume of the shed was 522.7 m 3 (11.0m width × 14.4 m depth × 3.3 m height).Most of the doors and wind ows of the shed were kept open throughout the day re ga r d less of season, except when there were strong win ds due to typhoons, so the air within the shed was regarded as being naturally ventilated.The floor was made of concrete and some of that was covered with straw.Waste from the cattle was removed daily at 7:00, 11:00, and 16:00.
Holstein and Japanese Black cattle, both major species for dairy cattle in Japan, were raised in the shed.All the cattle stayed in the shed in stalls throughout the day.The number and total weight of the cattle are shown in Table 1.The cattle were fed about 2 kg day -1 head -1 of hay (Sudan grass and Bermuda grass), 1 kg day -1 head -1 of hay (Phleum pretense and alfalfa), and 6-12 kg day -1 head -1 of formula feed (Miracle U 70, Marubeni Nisshin Feed) every day, the feedstuffs being split into three portions and provided at 7:00, 11:00, and 16:00.The cattle were able to drink water freely at any time from the watering stations within the facility and they could also take mineral salt.

3 Sampling Procedure
Air sampling in the cattle shed was conducted in spring (April 2017), summer ( July and August 2017), fall (October 2017), and winter (January 2018) for three or four days in each season (Table 1).Sample collection was performed according to a previous study (Osaka et al., 2018).The VFAs, phenols, sulfur compounds, indoles, some ketones (2-butanone, 2-pentanone, and 3-pentanone), and some alcohols (2-butanol, 1-propanol, and 1-butanol) were collected with stainless steel or glass sorbent tubes filled with Tenax TA ® sorbent (3.5 in.× 0.25 in.OD, 60/80 mesh, COMSCO).Prior to sample collection, all tubes were conditioned by a stream of pure nitrogen gas at a flow rate of 50 mL min -1 at 300°C.Air samples were collected at the center of the shed, as shown in Fig. 1.Air in the shed was continuously collected every hour in the sorbent tubes using a tube sampler (MTS-32, MARKES) at a flow rate of 0.1 L min -1 throughout every sampling period.After sampling, the tubes were closed and stored in a cool, dark place.In addition, trip blank and field blank samples were also collected for each sampling campaign.
For measurement of aldehydes and acetone, samples were collected with two 2,4-dinitrophenylhydrazine (DNPH) cartridges containing DNPH derivatizing agents (InertSep mini AERO DNPH-LG, GL Sciences) that connected to an ozone scrubber cartridge (InertSep mini AERO Ozone Scrubber, GL Sciences) upstream.Sample air was collected at the center of the shed at a flow rate of 0.1 L min -1 from 8:00 to 16:00 and from 16:00 to 8:00.After sampling, the DNPH cartridges were closed at both ends and placed in a cool, dark place.Field blanks were also processed along with the sorbent tube samples.
Air samples for methanol and ethanol were collected from the center of the shed using a Florisil ® cartridge (Presep-C ® Florisil, Wako Pure Chemical Industries) at a flow rate of 0.1 L min -1 for 30 min at 0:00, 8:00, 12:00, and 16:00 in each season.After sampling, the Florisil cartridges were closed at both ends and placed in a cool, dark place.Field blanks were also processed along with the sorbent tubes.
Number of samples collected in each sampling campaign were shown in Table 1.

4 Monitoring of Air Temperature and Relative
Humidity The air temperature and relative humidity (RH) inside and outside of the shed were monitored using hydro-thermographs (RTR-503, T&D, ±0.3°C, ±5% RH).Monitoring points inside the shed were the same as for the air sampling.The hydrothermograph for outside measurement was located at 10 m distance from the shed (Fig. 1).The external hydrothermograph was covered with tin foil to give protection from the effects of direct sunlight and rain.The air temperature and RH were monitored every 10 min throughout the sampling campaigns.The differences in temperature and RH between the two hydrothermographs were 0.07±0.08°Cand 0.36±0.55%RH, respectively.Based on these data, the air temperature and RH were judged to be calibrated.

5 Analytical Procedure
Analytical procedures, except for the determinations of methanol and ethanol, were conducted according to a previous study (Osaka et al., 2018).The VFAs, phenols, and indoles were determined by gas chromatographymass spectrometry (GC/MS; (GCMS-QP2020, Shimadzu Corporation) equipped with a thermal desorption injector (TD-GC/MS; TDTS-2020, Shimadzu Corporation).An InertCap WAX capillary column (30 m × 0.25 mm × 0.25 μm, GL Sciences) was used for separation.Chemical compounds collected on the adsorbent were desorbed for 3 min at 230°C with a purge flow of 50 mL min -1 with trapping at -20°C.The cold trap was rapidly heated to 230°C and the trapped chemical substances were injected into the GC/MS.The GC oven temperature program was as follows: 40°C (hold 3 min) → (ramp 8°C min -1 ) → 230°C (hold 5 min).The temperatures of the injection port and the ion source were 200 and 210°C, respectively.The samples were analyzed using the selected ion monitoring (SIM) mode.For signal quantitation, standard solutions of the analytes at 1, 10, and 100 ng μL -1 were measured by TD-GC/MS.
Methanol and ethanol samples were processed before analysis.Three mL of purified water were added onto a Florisil cartridge for extraction of methanol and ethanol.The eluate was analyzed by GC/MS (GCMS-QP2020, Shimadzu Corporation) equipped with an InertCap WAX capillary column (30 m × 0.25 mm × 0.25 μm, GL Sciences).The GC oven temperature program was as follows: 40°C (hold 1 min) → (ramp 5°C min -1 ) → 75 °C →(ramp 15°C min -1 ) → 120°C (hold 1 min).The temperatures of the injection port and the ion source were 200 and 210°C, respectively.The samples were analyzed using the SIM mode.
The trip blanks for all of the target compounds were below or near the detection limit.The field blanks were detected at most 10 ng, but they were sufficiently low compared with the measured values of the samples.The measured values of the samples were corrected by subtracting the blanks.

6 Estimation of ERs of VOCs from the Shed
The ERs of VOCs from the shed were estimated according to a previous study (Osaka et al., 2018).ERs of VOCs from the shed were evaluated using Eq. ( 1), where ) is the ventilation rate of the shed, and W (kg) is the total weight of cattle in the shed.The concentrations of the VOCs in the air in the shed were measured by the air sampling method.The ventilation rate was estimated using the water balance method (Urano and Katayama, 1985), which established the vapor equilibrium based on the water balance of the shed.In the shed, where air was exchanged by ventilation, the mass balance formula (Eq.( 2)) can be written as follows: where G* (kg) is the weight of air in the shed, t (s) is the time, G in (kg s -1 ) and G out (kg s -1 ) are the weights of intake air and exhaust air, respectively, and G g (kg s -1 ) is the weight of waste material, such as feces and urine, generated in the shed.On the assumption that humid air in the shed consists of vapor and dry air that are well mixed by ventilation, the two mass balance formulae (Eqs.(3-a) and (3-b)) for vapor and dry air, respectively, may be derived from Eq. (2): where V c (m 3 ) is the volume of the shed, V in (m 3 s -1 ) is the volume of intake air of the shed, V out (m 3 s -1 ) is the volume of exhaust air of the shed, x i (kg kg -1 ) is the indoor absolute humidity, x 0 (kg kg -1 ) is the outdoor absolute humidity, v i (m 3 kg -1 ) is the specific volume in the shed, v 0 (m 3 kg -1 ) is the specific volume exiting out of the shed, and W g (kg s -1 ) is the amount of moisture emission.The absolute humidity and the specific volume were determined from the air temperature and the RH inside and outside of the shed using a psychrometric chart.Moisture emission, that is, the amount of moisture to be removed from the room, was referred to the regression equation reported by Yeck et al. (1959).Yeck et al. (1959) described the relationship between the air temperature and the moisture ER for dairy cattle (water emission (lb.) per body weight of dairy cattle (lb.)) in the shed.Moisture emission may be converted to the latent heat of the room by multiplying the moisture released from the cattle by the latent heat of vaporization of the body temperature of the cattle, and by multiplying the moisture picked up from sources such as urine and feces by the latent heat of vaporization for the air temperature of the room (Yeck et al., 1959).Thus, moisture emission was estimated from the regression curves for the weight of one dairy cattle and the temperature in the shed.V out may be determined from Eqs. (3-a) and (3-b) as follows: where Δt is the time interval, x i * is the indoor absolute humidity after t min, and v i * is the specific volume in the shed after t min.On the assumption that the air in the shed was at a steady state, Eq. ( 4) may be re-written as Eq. ( 5).
The ventilation rate of the shed was estimated according to Eq. ( 5) on the assumption that the air exhaust is equal to the air intake.As described in the next section, in cases where the differences of temperature and RH between the inside and outside of the shed were below the measurement errors of the hydrothermographs, such data were removed from the calculation of the ventilation rate of the shed.The total annual emissions of VOCs from the shed was estimated according to Eq. ( 6): where N (g/year) is the amount of total annual emissions of VOCs from the shed.

1 Temperature and RH Inside and Outside of the Shed
The time variation of the temperature and RH differences between the inside and outside of the cattle shed during the spring sampling campaign is shown in Fig. 2. It was found that the differences in temperature and RH between the two locations were often smaller than the measurement errors (±0.3°C, ±5% RH) of the hydrothermographs used for measurement.Similar trends were also observed in summer, autumn, and winter.During the sampling campaigns, the windows and the doors of the cattle shed were kept open throughout the year except when typhoons occurred, so the shed was considered to be well-ventilated.Hence, relatively small differences in temperature and RH between the inside and outside of the shed occurred regardless of the season.The data whereby the temperature and RH differences were below the measurement errors were not considered reliable for calculating the absolute humidity and the specific volume (see former section).Therefore, these data were not used for estimating the ER of VOCs from the cattle.

2 Concentration and Composition of VOCs in the Cattle Shed
The average concentrations for each VOC in the air in the shed for the four seasons are shown in Table 2.The average concentrations of total VOCs (μg m -3 ) were 50.5 (spring), 128.4 (summer), 168.8 (autumn), and 199.5 (winter).In general, emissions of VOCs from the shed tended to increase with increase in the air temperature.Nevertheless, the concentrations of VOCs in the air in the shed were not particularly high in summer relative to the other seasons.This indicated that the shed was wellventilated by the outside air.Especially in summer, a large fan was always operated in addition to having the doors and windows of the shed open to control the phys-   SD: standard deviation 2) BDL: below detection limit ical conditions inside the shed.Thus, the concentrations of VOCs in the shed were not high in the summer.
The chemical composition of the VOCs for each season are indicated in Fig. 3.It was found that the ketones were always the most dominant component followed by alcohols and VFAs.The sum of ketones, alcohols, and VFAs accounted for more than 80% of the total VOCs in all seasons.Acetone, 3-pentanone, 1-butanol, and acetic acid were the major chemicals regardless of the season, accounting for more than 60% of the total VOCs.Filipy et al. (2006) and Chung et al. (2010) showed that acetone was one of the dominant VOCs emitted from dairy cattle.Alanis et al. (2010) estimated the VFA flux emitted from dairies and clarified that more 90% of the VFAs emitted was in the form of acetic acid.The present data are generally consistent with previous reports concerning this latter point.
The relationships between the temperature inside the shed and the ER of total VOCs, each VOC class and each individual VOC (except for the compounds which were not detected in all four seasons) are shown in Figs.5-7.The ERs of total VOCs and each VOC class except alcohols increased exponentially with increase of temperature inside the shed (Figs. 5 and 6).In addition, similar trends were also observed regarding the ER of each VOC (Fig. 7a-7g).The Claperyon equation, which descri bes the relationship of vapor pressure and temperature, is as follows:

) RT
where P is the vapor pressure, ΔvapHm is the molar enthalpy of vaporization, R is the molar gas constant, and T is the temperature.According to the equation, vapor pressure increases rapidly with rise in temperature.Hence VOCs contained in feces, urine, feed, and bedding in the shed would undergo increased emission with rise of Emission rate [μg h -1 temperature.Also, the amount of transpiration from the body surface of the livestock would generally increase as air temperature rises.Moreover, Shibata and Mukai (1981) reported that the respiratory quotient of dairy cattle increased with rise of temperature, suggesting that the amounts of VOCs in expired air increase with increase in temperature.Thus, in general, the ER of VOCs is considered to increase with rise of air temperature within the shed.Previous studies also support the trends observed in this work.For instance, Hafner et al. (2012Hafner et al. ( , 2010) ) reported that the ER of ethanol from loose corn silage increased exponentially with rise of temperature.In addition, Rumsey et al. (2012) showed that ethanol emissions weakly correlated with the air temperature of the swine barn.Osaka et al. (2018) estimated that the mean ERs of VOCs from a swine shed in each season were 1-2 × 10 3 μg (h kg-swine) -1 .In addition, VFAs were found to be the dominant emission component from the swine, accounting for 47-75% of the total VOCs (Osaka et al., 2018).Compared with those from swine, the ERs of VOCs from dairy cattle were almost at the same level as those of swine, although the compositions of VOCs emitted from the swine and cattle were different.

4 Annual Emissions of VOCs from Dairy Cattle
The annual VOC emissions from one dairy cow and

]
Emission rate ] Fig. 7a.Relationship between the temperature in the shed and ER of VFAs.The error bars show the standard deviation.Emission rate [μg h -1 Emission rate from the cattle shed were estimated using the ERs of VOCs shown above.The results are given in Assuming that the annual emissions of VOCs from dairy cattle for the present study is taken as the average for Japan, the annual VOC emissions from dairy cattle in Japan are estimated to be 7.3 Gg.Kannari et al. (2007) reported NMVOCs emitted from several sources in Japan and estimated that the annual emissions from industrial combustion and navigation were 45 Gg and 14 Gg, respectively.The annual emissions of VOCs from dairy cattle estimated in this study is of a similar level as that of industrial combustion and navigation.This finding confirms that emissions associated with the livestock industry represent a substantial fraction of the VOC emissions inventory in Japan.Clearly, there will be differences in husbandry practices and environmental condi-tions for dairy cattle operations in Japan; hence, emissions and compositions of VOCs from each cattle shed are expected to cover wide ranges.Accordingly, the value of annual emissions of VOCs from dairy cattle in Japan estimated by using only the data obtained in this study will be subject to a degree of uncertainty and potential error.Given that there is currently no other data available for VOC emissions from dairy cattle in Japan, the annual emissions of VOCs from dairy cattle estimated in the present study is considered to have great significance despite being subject to considerable uncertainty.

CONCLUSION
In this study, VOCs in a cattle shed located in Asahi Agricultural High School, Chiba Prefecture, Japan, were collected over four seasons in an attempt to clarify the emission characteristics and composition of VOCs from the livestock industry in Japan.As part of the study, the temperature and humidity inside and outside of the cattle shed were also monitored.The average VOC concentrations in the cattle shed for the four seasons were in the range 50.5-199.5μg m -3 .Ketones were always the most dominant VOC class followed by alcohols and VFAs irrespective of season.Acetone, 3-pentanone, 1-butanol, and acetic acid were the major chemicals regardless of the season, accounting for more than 60% of the total VOC.The average ERs of total VOCs from the cattle shed (μg h -1 kg -1 ) were calculated to be 469 (winter), 623 (spring), 1520 (summer), and 585 (autumn).The ERs for almost all the VOCs increased exponentially with increase in the air temperature of the shed.Furthermore, the annual emissions of VOCs for a single dairy cattle and for the cattle shed were estimated to be 5.5 kg and 44 kg, respectively.

Fig. 2 .
Fig. 2. Differences of temperature (upper) and RH (lower) between the inside and outside of the shed during the spring sampling campaign.Dotted lines represent the range of measurement errors.

Fig. 6 .
Fig. 6.Relationship between the temperature in the shed and ER of each class of VOC.The error bars show the standard deviation.
cattle shed [°C] Temperature in the cattle shed [°C] Temperature in the cattle shed [°C] Temperature in the cattle shed [°C] Temperature in the cattle shed [°C] Temperature in the cattle shed [°C] Temperature in the cattle shed [°C] Temperature in the cattle shed [°C]

Fig. 7b .
Fig. 7b.Relationship between the temperature in the shed and ER of alcohols.The error bars show the standard deviation.

[Fig. 7f .
Fig. 7c.Relationship between the temperature in the shed and ER of phenols.The error bars show the standard deviation.

Table 1 .
Outline of sample collection.

Table 2 .
Concentrations of VOCs in the cattle shed for each season (in μg m -3

Table 3 .
The annual emissions of VOCs per single dairy cattle and the dairy cattle shed were estimated to be 5.5 kg and 44 kg, respectively.The number of dairy cattle raised in

Table 3 .
Annual emissions of VOCs from one cow and from the cattle shed.