Annual Variations of Odor Concentrations and Emissions from Swine Gestation, Farrowing, and Nursery Buildings

ABSTRACT

To obtain annual odor emission profiles from intensive swine operations, odor concentrations and emission rates were measured monthly from swine nursery, farrowing, and gestation rooms for a year. Large annual variations in odor concentrations and emissions were found in all the rooms and the impact of the seasonal factor (month) was significant (P < 0.05). Odor concentration was low in summer when ventilation rate was high but high in winter when ventilation rate was low, ranging from 362 (farrowing room in July) to 8934 (nursery room in December) olfactory unit (OU) m⁻³. This indicates that the air quality regarding odor was significantly better in summer than that in winter. Odor emission rate did not show obvious seasonal pattern as odor concentration did, ranging from 2 (gestation room in November) to 90 (nursery room in April) OU m⁻² sec⁻¹; this explains why the odor complaints for swine barns have occurred all year round. The annual geometric mean odor concentration and emission rate of the nursery room was significantly higher than the other rooms (P < 0.05). In order to obtain the representative annual emission rate, measurements have to be taken at least monthly, and then the geometric mean of the monthly values will represent the annual emission rate. Incorporating odor control technologies in the nursery area will be the most efficient in reducing odor emission from the farm considering its emission rate was 2 to 3 times of the other areas. The swine grower-finisher area was the major odor source contributing 53% of odor emission of the farm and should also be targeted for odor control. Relatively positive correlations between odor concentration and both H₂S and CO₂ concentrations (R ² = 0.58) means that high level of these two gases might likely indicate high odor concentration in swine barns.

IMPLICATIONS

The emissions of air pollutants including odors, greenhouse gases, and toxic gases have become a major environmental issue facing animal farms in the U.S.A. and Canada. To ensure the air quality in the vicinity of intensive livestock farms, air dispersion models have been used to determine setback distances between livestock facilities and neighboring residences based on certain air quality requirement on odor and gases. Due to the limited odor emission data available, none of the existing models can take account of seasonal variations of odor emissions, which may result in great uncertainties in setback distance calculations. Therefore, the obtained seasonal odor and gas emission rates by this study can be used by the government regulatory organizations and researchers in air dispersion modeling to get improved calculation of setback distances.

INTRODUCTION

Odor emission from intensive swine farms has become a concern for the neighboring communities. Researchers have been using air dispersion models to estimate reasonable setback distances between the livestock operations and neighboring residences; however, the results have not been well accepted yet.1 5 One of the main reasons is the limited odor emission data available to be used in the modeling. The odor emissions depend on many factors, including building characteristics, ventilation rates, animal size and density, diet, weather conditions, manure handling systems, etc., and are highly variable with large diurnal and seasonal fluctuations.6 11 Odor emission rates have been measured more or less randomly and the means or geometric means of these limited measured odor emission data were used as representative values in odor dispersion and setback modeling without considering the diurnal and seasonal variations,2 4 5 which may result in great uncertainties of setback distance calculation. Hence, it is necessary to quantify seasonal variations of odor emissions from livestock farms in order to provide accurate emission data for air dispersion modeling and setback determination.

The objective of this study was to measure seasonal variations of odor concentrations and emission rates from three types of swine buildings, i.e., gestation, farrowing, and nursery, for 1 year to obtain the annual profiles. For the swine grower-to-finisher buildings, a previous study conducted by Sun et al.12 13 have measured diurnal and seasonal odor emissions from two types of these buildings; thus, the results from all four types of rooms will also be compared in this study.

MATERIALS AND METHODS

Description of Experimental Swine Rooms

This research was conducted at the Prairie Swine Centre Inc. (PSCI) Elstow Research Farm in Saskatchewan, Canada, from August 2006 to July 2007. This farm was located near Elstow, 50 km east of Saskatoon, Saskatchewan, Canada. It was a 600-sow farrowing to finishing operation. There were a total of eight identical nursery rooms, eight identical farrowing rooms, and two gestation rooms. A nursery room, a farrowing room, and a gestation room with individual stalls were selected for odor emission measurement. All the rooms were ventilated mechanically by wall-mounted exhaust fans. Manure was dropped into the shallow pits beneath the slatted floor and was emptied once every 3 to 8 weeks. The specifics of the experimental rooms were presented in Table 1.

Table 1. Basic information of the swine rooms

Facility	Floor Type	Size (Length × Width × Height^a, Area)	Capacity (Head)	Fan No.	Feeding System	Manure Removal Frequency (Week)
Nursery	Fully slatted	14.9 × 7.2 × 3.7 m, 107 m^2	256	3	Automatic	8
Farrowing	Fully slatted	14.9 × 7.2 × 3.7 m, 107 m^2	14	2	Manual	4
Gestation	Partially slatted	57.5 × 15.2 × 3.7 m, 874 m^2	416	10	Automatic	3
Notes: ^aThe ceiling height of the barn was 3.7 m rather than conventional 3 m to allow for better view of the rooms from a public viewing gallery located along the center of the attic of the barn.

Experiment Design and Statistical Analysis

In order to monitor annual odor concentration and emission rate trends, a repeated measurement method was used that was suitable for the same experimental unit over a period of time. In statistical analysis, a two-factorial experiment design was used: “room type” factor with 3 levels (nursery room, farrowing room, and gestation room) and seasonal “month” factor with 12 levels (January to December), which is treated as a repeat factor. The measurements were conducted around the 20th of each month from August 2006 to July 2007. During the monthly measurement, two identical air samples were collected from each room during the 9:00 a.m. to 12:00 p.m. period when the pigs had high activity level. This period of time was selected because the animal activity was generally the highest of the day so the odor and gas concentrations were the highest. Odor measurement was limited to one pair of samples per room per month mainly due to the high cost of odor of odor measurements (over $250 per sample). For the same reason, identifying possible relationships, if any, between odor and other gases would be highly beneficial in that the gas could be used as an odor indicator. Therefore, concentrations and emissions of a few gases were also measured simultaneously in this study including ammonia (NH₃), hydrogen sulfide (H₂S), carbon dioxide (CO₂), and methane (CH₄). However, gas concentrations were only used in this study for the purpose of identifying any correlation between odor and gases. Detailed gas concentration and emission data will be presented in other publications.

All data were subjected to an analysis of variance using SPSS.14 General linear model (GLM) procedure in SPSS was conducted to evaluate if odor concentration and emission rate were significantly affected by season or room type. The significances of the effects were determined at the 5% level.

Odor and Gas Measurement

During each monthly measurement, two identical samples were collected from each room into two 10-L Tedlar bags (sample bags 232 series; SKC, Inc., Eighty Four, PA) continuously for 3 hr from 09:00 a.m. to 12:00 p.m. through a Teflon FEP tubing (Cole-Parmer, Vernon Hills, IL) and a peristaltic pump (Master flex pump; Cole-Parmer) at a rate of 0.05 L/min, which represented average 3-hr concentration. The pumps were set up at the sampling station outside of each room and the inlet of the tubing was located close to the exhaust fans. There was one sampling point for the nursery and farrowing rooms, respectively, whereas there were two in the gestation room to obtain representative air samples due to its large size. The two sampling lines merged into one line in the middle of the room, which was the mixture of the exhaust airs from the two sampling points representing average exhaust air.

The odor sample bags were shipped to the Olfactometry Laboratory at University of Alberta, Edmonton, Alberta, Canada, for measurement within 30 hours after the samples were collected. Odor concentration (olfactory unit [OU] m⁻³) was assessed by eight trained panelists with a triangular dynamic forced-choice olfactometer based on the CEN (European Committee for Standardization) and American Society for Testing and Materials (ASTM) olfactometry standards protocol.15 16

The odor concentration of the supply air coming through the air inlets was generally very low as compared with the exhaust air from the building. Weak odor samples are hard to be detected using dynamic olfactometry17; therefore, the inlet concentration of odor was assumed to be negligible.

Three gas analyzers were used to measure H₂S, CO₂, and NH₃ in the rooms. A Teflon tubing was installed parallel to the odor sampling tubing to facilitate sampling with the analyzers located at the sampling station outside each room. An air pump continuously pumped the room air near the exhaust fans to the gas analyzers and gas concentrations were recorded once every 15 min by data loggers located at the sampling stations. The gas analyzers used were an infrared NH₃ analyzer (Chillgard RT refrigerant monitor, ±2% accuracy; MSA Instrument Division, Pittsburgh, PA), a H₂S analyzer (JEROME 631-X, with accuracy of ±0.003 ppm at 0.05 ppm, ±0.03 ppm at 0.5 ppm, and ±0.3 ppm at 5 ppm; Arizona Instrument Corporation, Phoenix, AZ), and a CO₂ analyzer (Guardian Plus Infra-Red Gas Monitor, ±2% accuracy; Edinburgh Sensors Limited, Hingham, MA). The 3-hr average values from 09:00 a.m. to 12:00 p.m. were used in this study.

For CH₄ measurement, the gas samples were collected from the odor sample bags by injectors into 10-mL vacuum tubes and measured in the Gas Chromatography Laboratory, Soil Science Department, the University of Saskatchewan. The measured values represented average 3-hr concentrations.

For odor emission calculations, the total odor emission rate of a room (in OU sec⁻¹) was obtained by multiplying the odor concentration (in OU m⁻³) and ventilation rate (in m³ sec⁻¹). The odor emission rate can also be expressed as odor emission per square meter of the floor area of the room (in OU m⁻² sec⁻¹), which was the total odor emission rate of the room (in OU sec⁻¹) divided by the total floor area of the room (in m²). It can also be expressed as odor emission rate per Animal Unit (in OU AU⁻¹ sec⁻¹ [AU: animal unit, 1 AU = 500 kg live animal mass]), which was the total emission rate of a room (in OU sec⁻¹) divided by the total animal units of the room (AU). The incoming air sampling point was located in an air inlet in the nursery room. The incoming air concentrations of NH₃ and H₂S, and CO₂ were measured twice during each monthly measurement, once before 09:00 a.m. and again after 12:00 p.m. The ventilation rate of a room was the sum of air flow rates of all exhaust fans in the room, which was determined using the fan curve method, i.e., by measuring the vacuum pressure in the room and rotating speeds of each operating fan and then obtaining the air flow rate of a fan from the fan testing report. The five types of fans used in the rooms, i.e., TR12F, TR16F, TR20F, TR24F, and TR36F, were either variable-speed fans or single-speed fans (Prairie Pride Enterprises, Winnipeg, MB, Canada). Fan testing reports from Bioenvironmental and Structural System Laboratory, Department of Agricultural Engineering, University of Illinois, were used to estimate the airflow rates of the fans.18

Each room had a pair of pressure lines (6.35 mm internal diameter [ID] Tygon tubing) and a pressure transducer except the gestation room had two separate sets due to its large size. A differential pressure transducer (Model 265, accuracy of ±1%, measuring range 0 to 0.25 inch water; Setra System Inc., Boxborough, MA) was installed at the sampling station outside of each room and it measured the static pressure differential between the outside and inside of the room. The inlet of the inside tubing was placed in the middle of the room 1.5 m above the floor. The inlet of the outside tubing was placed outside of the building near the eave facing ground to prevent wind interference and ice buildup in the tubing. Rotating speeds (rotations per minute [RPM]) of all fans were measured by micro-switch Hall Effect position sensors (SR3F-A1; Honeywell Inc., Freeport, IL) mounted on the motor of the fans. The readings were taking every 5 min with 15-min average recorded by the data loggers.

Other Measurements

Inside and outside temperatures and relative humidity were also measured every 5 min with the average of 15 min recorded by the data acquisition system. The pig numbers in each room were recorded. Two pens of the nursery room (16 pigs in each pen) were chosen and the total weight of the pigs in each pen was measured and the average pig weight in the room was estimated. The pig weight in gestation and farrowing rooms were taken from the record of each sow, and the average piglet weight was estimated based on experience of the barn workers. The total animal units in each room were obtained by dividing the total animal weight by 500-kg animal live weight per animal unit.

RESULTS AND DISCUSSION

The ambient temperature, room temperatures, and ventilation rates, are shown in Figure 1. The odor concentrations and emission rates of the three rooms are illustrated in Figure 2. The annual average room temperatures and geometric means and standard deviations of odor concentrations and emission rates were also summarized in Table 2. Statistical analysis found that there were interaction between the type of room and the measurement month; therefore, Table 3 gives the monthly data and the statistical analysis results for comparison of monthly variations for each room. Odor concentrations and emissions in October were not available because of sample shipment delay caused by adverse weather and bus service cancellation from Saskatoon to Edmonton.

Table 2. Annual geometric means of odor concentrations and emission rates.^a

	Mean	Mean	Odor Concentration (OU m^−3)				Odor Emission Rate (OU m^−2 sec^−1)				Odor Emission Rate (OU AU^−1 sec^−1)
	Room	Animal	Geometric
Room	T(°C)	Unit	Mean	SD	Max	Min	Geometric mean	S.D.	Max	Min	Geometric mean	S.D.	Max	Min
Nursery	26	9.3	3255	2890	8934	927	34	23	90	17	459	364	1410	134
Farrowing	24	7.0	1990	1685	4871	362	16	11	46	3	259	173	750	48
Gestation	21	146.6	1540	1364	4096	400	10	18	64	2	59	118	418	14
Finishing 1	18	35.5	1145	752	2964	221	14	64	29	4	106	64	242	39
Finishing 2	19	37.6	1929	1175	3822	347	22	44	36	11	149	44	219	104
Note: ^aFinishing room 1 was the average of the two rooms with partially slatted floors, whereas finishing room 2 represented the average of the two rooms with fully slatted floors.

Table 3. Monthly odor concentrations and emissions

	Ambient	Odor Concentration (OU m^−3)*			Odor Emission Rate (OU m^−2 sec^−1)*
Date	T (°C)	Nursery	Farrowing	Gestation	Nursery	Farrowing	Gestation
Aug 2006	26.1	1448^ab	2195^bcd	1448^a	38^a	46^c	64^b
Sep 2006	9.8	4466^de	4095^e	2048^b	30^a	26^bc	13^a
Nov 2006	−8.4	3755^bcde	3444^cde	1024^a	19^a	12^abc	2^a
Dec 2006	−8	8934^f	3756^de	2896^c	58^ab	18^abc	8^a
Jan 2007	−4.5	8192^f	4871^e	3756^d	54^ab	21^abc	9^a
Feb 2007	−6.4	6889^e	2435^bcd	4096^d	48^ab	15^ab	6^a
Mar 2007	−0.6	4467^cde	4096^e	2233^bc	29^a	21^abc	7^a
Apr 2007	14.9	4096^abcd	1024^ab	1218^a	90^b	15^abc	12^a
May 2007	20	1024^a	362^a	512^a	17^a	3^a	10^a
Jun 2007	13.2	2048^abc	1722^abc	N/A**	20^a	15^abc	N/A**
Jul 2007	32.2	927^a	538^a	400^a	24^a	15^abc	18^a
Notes: *The means of odor concentrations or odor emission rates in the same column followed by the same letter are not significantly different (P > 0.05).
**Air sample bags leaked.

Figure 1. Room and ambient temperatures and ventilation rates over the year.

Figure 2. Seasonal variations of odor concentrations and emissions.

As shown in Figure 1, during the monthly measurement days, the outside temperature varied from −8.4 °C in November to 32.2 °C in July, whereas the room temperature did not vary as much because supplemental heating was provided but no cooling measure was in place. The ventilation rate followed the similar pattern as the ambient temperature, which was high in the period between April and August and low between September and March.

The annual geometric means of odor concentrations for the nursery, farrowing, and gestation rooms were 3255, 1990, and 1540 OU m⁻³ with corresponding odor emission rates 34, 16, and 10 OU m⁻² sec⁻¹, respectively (Table 2); if the odor emission rates were expressed in terms of animal unit basis, the annual geometric means of odor emission rates were 452, 259, and 59 OU AU⁻¹ sec⁻¹, respectively. Odor concentrations and emissions from the nursery room were significantly higher than the other two rooms (P < 0.05). Higher odor in the nursery room may have been due to less frequent manure removal, diet, higher room temperature, and greater animal activity.

As shown in Table 3, statistical analysis revealed that the seasonal “month” factor had significant effect on odor concentrations for all three rooms (P < 0.05). The peak concentrations occurred in winter when ventilation rates were low, in December for the nursery room, in January for the farrowing room, and in February for the gestation room. The lowest odor concentrations occurred in July for all the rooms when ventilation rates were high. Odor concentration of the nursery room varied from 927 to 8934 OU m⁻³, the farrowing room from 362 to 4871 OU m⁻³, and 400 to 4096 OU m⁻³ for the gestation room. The results clearly indicate that the air quality in terms of odor concentration varied greatly over the year with the extremely high odor concentration occurred in the coldest months.

As shown in Figure 2b, the odor emission rates fluctuated over the year but did not show obvious seasonal pattern as odor concentrations did. Significant monthly differences were also found in all rooms by Duncan multiple comparison (P < 0.05; Table 3). The odor emission variation range was relatively smaller in the farrowing room comparing with the nursery and gestation rooms. Odor emission rates of the nursery room varied from 17 to 90 OU m⁻² sec⁻¹, the farrowing room from 3 to 46 OU m⁻² sec⁻¹, and 2 to 64 OU m⁻² sec⁻¹ for the gestation room. The peak odor emission rates were 6, 15, and 27 times of the lowest values for the nursery, farrowing, and gestation rooms, respectively. The peak emission rate occurred on April 7 as a combined result of high ventilation rate and high odor concentration considering the emission rate is the product of odor concentration and ventilation rate, which made it higher than the winter emission rates (e.g., in December when odor concentration was the highest but ventilation rate was the lowest) and the summer emission rates (e.g., in August when ventilation rate was the highest but odor concentration was the lowest). Studies by Sun et al.13 and Guo et al.⁹ also had similar observations. The reason that the peaks occurred in spring rather than autumn might be that in spring the increasing manure temperature caused higher manure gas production but the ventilation rate was still low during most of the day due to relatively low ambient temperature, thus once the ventilation rate picked up it might be combined with high manure gas production in barn and resulted in high odor emission rate.

The large standard deviations of odor concentrations and emissions summarized in Table 2 reflected high variations of odor concentrations and emissions throughout the year for all three types of rooms. Odor emission peaked in August for the farrowing and gestation rooms when the ventilation rate reached the maximum level and played a dominant role in emission determination. However, the peak of odor emission from the nursery room occurred in April, although its corresponding concentration and ventilation rate were not at their maximums. This shows the important contribution of both odor concentration and ventilation rate on odor emission rates. Because large seasonal variations in odor concentrations and emission rates were found in all the rooms, it is suggested that random measurement of odor emissions for odor dispersion modeling or setback modeling may contribute to great uncertainty. To obtain representative odor emission rate, either measurement has to be taken at the interested time period of the year, or monthly or at least seasonal (the four seasons) measurements should be taken to get the representative seasonal and annual emission rates.

Results obtained by this study were comparable with those reported by Guo et al.9 and Zhang et al.19, both were conducted in the same region as this study on Canadian Prairies. Guo et al.⁹ measured odor emission rates from a commercial swine farm on three separate sites for a year. The measured rooms included two gestation, two farrowing, four nursery, and three finishing rooms. The odor concentration varied from 476 to 8605 OU m⁻³ (geometric mean 1975) for nursery, 354 to 4752 OU m⁻³ (geometric mean 1226) for farrowing, and 536 to 4993 OU m⁻³(geometric mean 1252) for gestation rooms, respectively. The odor emission rate in nursery rooms varied from 8 to 261 OU m⁻² sec⁻¹ with geometric mean of 31 OU m⁻² sec⁻¹, whereas the geometric mean for gestation and farrowing rooms were 17 (varying from 6 to 40) and 25 (varying from 10 to 58) OU m⁻² sec⁻¹, respectively. Zhang et al.¹⁹ measured odor emissions from two swine farrowing farms from June to September, and the geometric means of odor emission rates were 8 and 12 OU m⁻² sec⁻¹ for the gestation barns and 23 OU m⁻² sec⁻¹ for the farrowing barns, which were in the same range as the results obtained by this study. Additionally, the results were also accord with the odor emission rate review reported by Gay et al.,11 i.e., the odor emission rates ranged from 7 to 48, 3 to 48, and 5 to 21 OU m⁻² sec⁻¹ for nursery, farrowing, and gestation rooms, respectively. Wood et al.¹¹ also presented the results measured in Minnesota over a 3-year period, i.e., the geometric means and ranges of odor emissions from nursery, farrowing, and gestation barns 9 (2 to 97), 5 (0.1 to 17), and 13 (1 to 192) OU m⁻² sec⁻¹, respectively. The differences were likely due to the difference between climate, sampling period, ventilation rate measurement and calculation, diet, and manure management, etc.

This study was part of an air emission quantifying project that also measured odor emissions from two types of grower-to-finisher rooms (i.e., finishing rooms) on this farm12 13 and the results are also included in Table 2. The results for the finishing room 1 was the average of the two rooms with partially slatted floors, whereas the finishing room 2 represented the average of the two rooms with fully slatted floors. The nursery room had the highest odor concentration and emission rate. The finishing rooms with fully slatted floor had higher odor concentration and emissions than the finishing rooms with partially slatted floor, which could be caused by large manure exposure area in fully slatted flooring. Odor concentration and emission profiles found in the finishing rooms¹³ were similar to what have found in the other rooms by this study. Hence, significant seasonal variations in odor emission rate existed in all the four room types and influenced greatly by seasonal “month” factor (P < 0.05); however, no specific seasonal variation pattern was observed.

In order to identify the emission contributions of various areas on this swine farm, the odor emissions from the four areas were summarized in Table 4. The total odor emission from the whole barn was 123,174 OU sec⁻¹. The grower-to-finisher rooms had the largest area in the barn and was also the major emission source, contributing 53% of the total emission. Meanwhile, nursery area was the secondary odor source contributing 23.6%. Farrowing and gestation area contributed almost equally with contribution of 11.2% and 12.2%, respectively. These results provided the swine producers with the primary targeting area to apply odor control technologies.

Table 4. Annual average odor emission from the farm and relative room contribution.^a

Variable	Nursery	Farrowing	Gestation	Finishing
Total area (m^2)	858	858	1390	3502
Total emission of each area (OU sec^−1)	29104	13782	14986	65302
Total emission of the barn (OU sec^−1)		123174
Contribution of each area (%)	23.6	11.2	12.2	53.0
Note: ^aThe annual average emission is based on 8 nursery rooms, 8 farrowing rooms, 1 in-stall and 1 group housing gestation room, and 14 grower-finisher rooms.

Figure 3 shows the correlations of odor and gases by pooling all seasonal measurement data together. It is found that there were poor correlations between odor and NH₃ and CH₄ concentrations (R ² = 0.33); however, there were better correlations between odor and both H₂S and CO₂ (R ² = 0.58) than NH₃ and CH₄. Therefore, H₂S and CO₂ might serve as odor indicators for swine barns. This does not mean that these two gases are the main causes of odor, rather, high level of H₂S and CO₂ could possibly indicate high odor level in a barn. In fact, these four gases measured are mostly produced by manure degradation except the main sources of CO₂ are animal respiration and fossil fuel combustion when supplemental heating is required in winter. Although CO₂ and CH₄ are odorless, their concentrations do reflect the general level of manure gas production (there are 331 compounds in swine manure gases identified by Schiffman et al.20 and ventilation in the building. This is the reason that CO₂ is used as an air quality indicator in heating and ventilation system design in animal buildings.21

Figure 3. Correlations between odor concentrations and other gas concentrations.

CONCLUSIONS

Odor concentrations and emissions from three types of swine rooms (nursery, farrowing, and gestation) were measured over a year's period. Large seasonal variations in odor concentrations and emissions were found in all the rooms and the impact of the seasonal factor (month) was significant (P < 0.05). Odor concentration was lowest in summer when ventilation rates were high (362 OU m⁻³, farrowing room in July), and high in winter when ventilation rates were low (8934 OU m⁻³, nursery room in December). This indicates that the in-barn air quality regarding odor was significantly better in summer than that in winter. Odor emission rate did not show obvious seasonal pattern as odor concentration did, ranging from low of 2 OU m⁻² sec⁻¹ (gestation room in November) to high of 90 OU m⁻² sec⁻¹ (nursery room in April), and there were a few emission peaks occurred in April, December, and August. This correlates well with odor complaint trends against swine barns, which occur all year round. To get a representative specific odor emission rate, measurement has to be taken at least monthly so the annual geometric means may be obtained or measurement needs to be taken during the specific period of interest. The annual geometric mean odor concentration and odor emission rate of the nursery room was significantly higher than the other rooms (P < 0.05). Incorporating odor control technologies in the nursery area will be the most efficient in reducing odor emission from the farm considering its emission rate per square meter was 2 to 3 times of the other areas. Due to the large area, the finishing area was the major odor source contributing 53% of total odor emission of the farm and should also be targeted for odor control. Better correlations were found between odor concentration and both H₂S and CO₂ concentrations (R ² = 0.58) compared with NH₃ and CH₄ (R ² = 0.33); therefore, high concentrations of H₂S and CO₂ might indicate high odor concentrations in a barn.

ACKNOWLEDGMENTS

The authors wish to acknowledge the research funding from the Natural Sciences and Engineering Research Council of Canada and the Saskatchewan Agricultural Development Fund. The authors also express appreciation to the Prairie Swine Centre, Elstow Research Farm Inc., Saskatchewan, Canada, for their support.