- Email: [email protected]
Environmental assessment of three egg production systems – Part III: Airborne bacteria concentrations and emissions
Environmental assessment of three egg production systems – Part III: Airborne bacteria concentrations and emissions Y. Zhao,∗ D. Zhao,∗,† H. Ma,∗,‡ K. Liu,∗ A. Atilgan,∗,§ and H. Xin∗,1 ∗
Department of Agricultural and Biosystems Engineering, Iowa State University, Ames, IA, USA; † Department of Mechanical Engineering, Nanjing Institute of Industry Technology, Nanjing, China; ‡ Key Laboratory of the Ministry of Agriculture for Agricultural Engineering in Structure and Environment, China Agricultural University, Beijing, China; and § Department of Agricultural Structures and Irrigation, Suleyman Demirel University, Isparta, Turkey trations in the afternoon (with litter access) than in the morning (without litter access). The overall means and standard deviation of airborne total bacteria emission rates, in log CFU/[h-hen] (or log CFU/[h-AU], AU = animal unit or 500 kg live weight) were 4.8 ± 0.4 (or 7.3 ± 0.4) for CC, 6.1 ± 0.7 (or 8.6 ± 0.7) for AV, and 4.8 ± 0.5 (or 7.3 ± 0.5) for EC. Both concentration and emission rate of airborne total bacteria were positively related to PM10 . Gram− bacteria were present at low concentrations in all houses; and only 2 samples (6%) in CC, 7 (22%) samples in AV, and 2 (6%) samples in EC out of 32 air samples collected in each house were found positive with Gram− bacteria. The concentration of airborne Gram− bacteria was estimated to be <2% of the total bacteria. Total bacteria counts in manure on belt (in all houses) and floor litter (only in AV) were similar; however, the manure had much more Gram− bacteria than the litter. The results point out the need to mitigate airborne total bacteria in laying-hen houses, especially in AV houses.
Key words: indoor air quality, total bacteria, Gram-negative bacteria, airborne, alternative hen housing 2016 Poultry Science :1–9 http://dx.doi.org/10.3382/ps/pew053
INTRODUCTION
In addition to their local hazards, the microorganisms can be exhausted and transported in an ambient atmospheric environment, thus posing regional health risks (Seedorf et al., 1998). Therefore, indoor concentrations and facility emissions of airborne microorganisms are critical indicators of occupational health, animal wellbeing, food safety, and environmental impact. Efforts dedicated to quantify concentration and emission of airborne microorganisms have been made to obtain baseline values for poultry houses. The reported microbial concentrations in hen houses vary largely, ranging from 4 to 9 log colony forming unit (CFU)/m3 (Muller and Wieser, 1987; Seedorf et al., 1998; Oppliger et al., 2008; Matkovi´e et al., 2013; Wang-Li et al., 2013). The wide range of values is believed to result from the differences in production systems (litter-based vs. cage-based systems), thermal environment, and house ventilation (Zhao et al., 2014). Compared to indoor
Airborne microorganisms, with a large portion being bacterial species (Muller and Wieser, 1987), are usually found at high concentrations in laying-hen houses (Seedorf et al., 1998). These microorganisms are typically associated with dust particles, and can cause infection or trigger respiratory diseases to animals and caretakers after inhalation (Thelin et al., 1984; Rylander and Carvalheiro, 2006; Cambra-Lopez et al., 2010). It is also a concern that the microorganisms may deposit to the eggshell and result in contamination of table eggs (de Reu et al., 2008; Vuˇcemilo et al., 2010). C 2016 Poultry Science Association Inc. Received July 13, 2015. Accepted January 3, 2016. 1 Corresponding author: [email protected]
1
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
ABSTRACT Airborne microorganism level is an important indoor air quality indicator, yet it has not been well documented for laying-hen houses in the United States. As a part of the Coalition for Sustainable Egg Supply (CSES) environmental monitoring project, this study comparatively monitored the concentrations and emissions of airborne total and Gram-negative (Gram− ) bacteria in three types of commercial laying-hen houses, i.e., conventional cage (CC), aviary (AV), and enriched colony (EC) houses, over a period of eight months covering the mid and late stages of the flock cycle. It also delineated the relationship between airborne total bacteria and particulate matter smaller than 10 μm in aerodynamic diameter (PM10 ). The results showed airborne total bacteria concentrations (log CFU/m3 ) of 4.7 ± 0.3 in CC, 6.0 ± 0.8 in AV, and 4.8 ± 0.3 in EC, all being higher than the level recommended for human environment (3.0 log CFU/m3 ). The much higher concentrations in AV arose from the presence of floor litter and hen activities on it, as evidenced by the higher concen-
2
ZHAO ET AL.
MATERIALS AND METHODS Hen Houses The CSES environmental monitoring was carried out in three hen housing systems (CC, AV, and EC) located at the same farm in the Midwest United States. The CC house had a nominal capacity of 200,000 hens and was equipped with manure belts that conveyed the accumulated manure out of the house twice a week. The AV house had a nominal capacity of 50,000 hens and was provided with colonies and litter area accessible by the hens part of a day to perform foraging and dustbathing behaviors. Manure belts were installed in all hen colonies to remove manure out of the house twice a week, while the manure deposited/accumulated on the litter floor was removed at the end of each flock. The EC house also had a nominal capacity of 50,000 hens, and all manure was deposited onto the manure belts and was removed out of the house twice a week.
Detailed description of the housing systems and management practices was provided by Zhao et al. (2015a).
Samplers and Measurements All glass impingers (AGI-30, Ace Glass Inc., Vineland, NJ) were used for sampling airborne bacteria. The AGI-30 has a collection cut-off diameter of 0.3 μm, which means that collection efficiency is 50% for particles with aerodynamic diameter of 0.3 μm and is greater for larger particles. The sampler collects microorganisms in the air going through its sampling line into 20 mL liquid medium at an orifice-regulated air flow rate of 12.5 L/min. In this study, the medium used to collect the bacteria was sterile physiological saline (0.9% sodium chloride solution). The sampling medium was stored in individual 50 mL centrifuge tubes (20 mL liquid medium per tube), and kept in an ice chest at approximately 4◦ C before it was used. After collection, the microbe-laden liquid was transferred to its original centrifuge tube, put back in the ice chest, and transported to the Microbiology Analytical Lab at ISU campus for further analysis. Airborne bacteria were measured one day every month from January to August 2013, covering mid to near the end of a production cycle (45 to 78 weeks of age). To cover spatial and temporal variations, on each sampling day the bacterial concentrations were measured at the air exhausts (near stage-1 fan) and in the middle of the three houses in both the morning and the afternoon. This experimental design is important, especially in the AV house where higher airborne bacteria levels were expected in the afternoon when the hens had access to the litter. Figure 1 shows the sampling locations for airborne bacteria and PM10 . Gas concentrations and other parameters (e.g., temperature, relative humidity or ‘RH’, building static pressure or ‘SP’) were also measured in the environmental monitoring, but are not marked in the figure as their locations and results have been published in other companion papers (Zhao et al., 2015a; Zhao et al., 2015b). The sampling duration with AGI-30 is typically less than 30 min to avoid excessive medium evaporation that would cause a drop in collection efficiency. In this study, each sampling lasted for 15 min, which led to about 15% of the liquid medium being evaporated (about 17 of 20 mL was left after sampling). Ambient airborne bacteria level was not measured because it is negligible as compared to indoor concentrations (Lonc and Plewa, 2011; Wang-Li et al., 2013; Hu et al., 2015). On each sampling day from March to August, fresh manure on belts was collected at six different locations in each hen house, and made into a composite sample by mixing and well stirring it in a centrifuge tube. The bacteria concentration and moisture content were determined in these manure samples. Environmental conditions (temperature and RH), ventilation conditions (fan operation status and SP),
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
concentration, emission rates of microorganisms have been less investigated and are reported with different units, making it difficult to compare results across studies. Seedorf et al. (1998) reported total bacteria emission rate of 7.2 log CFU/[h-AU] (1 AU or animal unit = 500 kg live body weight) for cage laying-hen houses, and 9.5 log CFU/[h-AU] for broiler houses. Agranovski et al. (2007) reported a maximum emission rate of 9 log CFU/[h-house] for broiler houses; however bird numbers and body weight were not provided to allow estimation of emission rate on per AU basis. Most previous work for microbial measurement in hen houses was performed in European countries (especially for alternative housing systems) and focused on one specific housing system at a time. These baseline data would not necessarily reflect the situation in the United States where different management practices and housing systems are used. The Coalition for Sustainable Egg Supply (CSES) Project was launched in 2011. The goal of the project was to holistically evaluate the impacts of two alternative egg production systems, i.e., aviary (AV) and enriched colony (EC) housing systems, as compared to the conventional cage (CC) production system, with regards to hen well-being and health, egg quality and safety, environmental impact, food affordability, and worker safety. As a part of the CSES research team, our group at Iowa State University (ISU) examined the environmental impact of the three hen housing systems, and parts of the results have been published in a series of companion papers (Shepherd et al., 2015; Zhao et al., 2015a,b). This paper reports the concentrations and emission rates of airborne total and Gram-negative (Gram− ) bacteria in the CC, AV, and EC houses, and their relationship with particulate matter of 10 μm or smaller in aerodynamic diameter (PM10 ).
3
AIRBORNE BACTERIA OF THREE LAYING-HEN HOUSING TYPES
and PM10 concentrations were continuously monitored. Air temperature and RH were measured, respectively, with type-T thermocouples (Cole-Parmer, Vernon Hills, IL) and capacitance-type humidity sensors (HMP 61U, Vaisala Inc., Woburn, MA) at 1 s intervals. PM10 concentrations were measured with real-time Tapered Element Oscillating Microbalances (TEOM, Model 1400a, Thermo Fisher Scientific Inc., Waltham, MA) that were set to a 300 s integration time.
Sample Analysis The outside of all centrifuge sample tubes was cleaned and disinfected using alcohol before being transported to the biosafety cabinet for analysis. The samples were homogenized by vortexing for 5 s, and then viable counts of total bacteria and Gram− bacteria in each sample were determined by plating 0.2 mL portions directly and after serially diluted (1:10) in physiological saline onto trypticase soy agar (TSA, for total bacteria, Catalog No. R455002, Fisher Scientific, Hanover Park, IL) and Macconkey No. 3 agar (for Gram− bacteria, Catalog No. OXCM0115B, Fisher Scientific) plates. The plates were aerobically incubated at 37◦ C for 24 h (total bacteria) or 48 h (Gram− bacteria). The colonies formed on plates (30 to 300 colonies) were counted and used for calculating concentration of airborne bacteria (Equation 1, ‘log’ in this study is base-10 logarithm). The principle detection limit of the plating culture method is 2.7 log CFU/m3 (1 colony on the agar plate with undiluted sample after incubation). Emission rate was calculated using concentration data and ventilation rate (VR) simultaneously measured at the corresponding 15 min bacteria sampling time (Equation 2). The VR was corrected to standard temperature and
pressure conditions.
C = log10
1 N × 10n × Vs × Vp Va
(1)
where C is the airborne bacteria concentration, log CFU/m3 ; N is the number of colonies on a countable plate (30 to 300 colonies); n is serial dilution factor (n = 0 for undiluted sample, n = 1 for 10-fold diluted sample, etc.); Vp is the sample volume plated, mL (Vp = 0.2 mL in this study); Vs is total volume of original liquid sample, mL; and Va is the total air volume sampled using AGI-30, m3 (Va = 0.1875 m3 in this study).
E = log10
Tstd Pa 10 × V R × × Th Pstd C
(2)
where E is airborne bacteria emission rate, log CFU/[hhen] or log CFU/[h-AU]; VR is building ventilation rate, m3 /[h-hen] or m3 /[h-AU]; Tstd is standard temperature (273.15 K); Th is house temperature, K; Pstd is standard atmospheric pressure (101.325 kPa); and Pa is actual atmospheric pressure, kPa. Enrichment was further conducted to boost levels of bacteria that were potentially too low for detection using direct plating. For total bacteria, 2 mL of the undiluted samples was transferred to a sterile tube containing 50 mL of trypticase soy broth (TSB, Catalog No. R455052, Fisher Scientific), and incubated at 37◦ C for 48 h. Following enrichment, 0.5 mL culture suspension was spread on TSA plates, and the plates were incubated at 37◦ C for 48 h before being checked. The same procedure was applied to Gram− bacteria enrichment and culturing, except that Macconkey broth (7185, Neogen Corp., Lansing, MI) and Macconkey No. 3 agar were used. The presence of bacteria in the enriched
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Figure 1. A schematic representation of the house layout and sampling locations for airborne bacteria in the conventional cage (CC), enriched colony (EC), and aviary (AV) laying-hen houses. Particulate matter (PM) concentrations and thermal environment conditions were measured at the same locations.
4
ZHAO ET AL. Table 1. Daily mean air temperature, relative humidity, ventilation rate, and PM10 concentration in the conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses on the sampling days for airborne bacteria. Air temperature (◦ C)
Relative humidity (%)
Ventilation rate (m3 /[h-hen])
PM10 concentration (μ g/m3 )
Date (d/m/y)
Climate
Amb
CC
AV
EC
Amb
CC
AV
EC
CC
AV
EC
CC
AV
EC
1/9/13 2/16/13 3/16/13 4/11/13 5/10/13 6/7/13 7/4/13 8/9/13 Avg±SD
Cold Cold Cold Cold Mild Mild Warm Warm
−2.1 −7.4 1.8 0.9 12.6 14.9 22.2 21.4 8.0±11.2
23.2 22.7 23.6 23.1 24.4 24.7 26.0 26.3 24.3±1.4
26.4 26.1 26.7 26.6 26.8 27.1 26.7 27.0 26.7±0.3
25.7 25.5 25.3 25.1 25.4 25.6 25.0 25.5 25.4±0.3
74 76 82 96 56 74 64 69 74±12
66 65 63 67 47 53 53 55 59±8
60 60 57 59 45 50 50 53 54±5
62 66 63 60 50 53 54 57 58±6
0.5 0.5 0.7 0.6 1.5 1.6 3.4 3.5 1.5±1.2
0.5 0.6 0.7 0.6 1.1 1.5 4.4 3.5 1.6±1.5
0.5 0.6 0.8 0.6 1.2 1.6 4.8 4.6 1.9±1.8
695 908 701 879 699 762 526 555 716±135
8221 9151 9026 9815 7171 7535 2225 2406 6944±2984
645 843 680 768 563 584 390 404 610±160
Note: ‘Avg’ stands for average, ‘SD’ stands for standard deviation.
Data Analysis Statistical analysis was performed to investigate the airborne bacteria concentration and emission rate as affected by house type, sampling location, sampling moment, and climatic condition using Statistical Analysis System version 9.3 (SAS 9.3, SAS Institute Inc., Cary, NC). Based on sampling dates, average daily ambient temperature and VR, three climatic conditions were categorized, i.e., cold (January to April), mild (May and June), and warm (July and August). The actual ambient temperatures and VR of these three climatic conditions are presented in Table 1. For consistency with analysis in companion papers in this series, climate condition was selected as model factor instead of VR, although they are inherently interchangeable. A model fitting process was performed using PHREG procedure with all concerned factors included. The sampling location was confirmed as the only insignificant factor (P = 0.537) and was excluded from subsequent analysis. The effects of remaining factors on bacteria concentration and emission rate (or ‘Y’ in equation 3) were then tested using GLIMMIX procedure (Equation 3). A random term of house × location was included to account for dependency of measurements taken from the same house in the same location. The effects were considered significant at a probability level of 0.050. Since Gram− bacteria was non-detectable in most of the air samples, the presence (yes/no) of Gram− bacteria as
affected by the factors was tested assuming a binomial distribution.
Y = house + climate + house × climate + moment (3) Regression analyses was performed to delineate the relationships between concentration and emission rate of airborne bacteria vs. PM10 .
RESULTS AND DISCUSSION Temperature, Relative Humidity (RH), and Ventilation Rate (VR) The daily mean ambient temperature on the sampling days ranged from −7.4 to 22.2◦ C, with overall mean and standard deviation or ‘SD’ of 8.0 ± 11.2◦ C. The indoor temperatures and RH were well maintained at the thermal comfortable zone for laying hens in all three houses, 25.4 ± 1.3◦ C and 57 ± 6%. Daily mean VR in m3 /[h-hen] and SD were 1.5 ± 1.2 for CC, 1.6 ± 1.5 for AV, and 1.9 ± 1.8 for EC. Daily mean PM10 concentrations in μg/m3 and SD were 716 ± 135 for CC, 6944 ± 2984 for AV, and 610 ± 160 for EC. Table 1 shows the details of the results.
Concentration of Airborne Total Bacteria The overall mean of airborne total bacteria concentration in log CFU/m3 was 4.7 ± 0.3 for CC and 4.8 ± 0.3 for EC, and both were significantly lower than 6.0 ± 0.8 for AV (P = 0.002 and 0.003, respectively). These concentration levels were within the range reported in previous studies, but at the lower end of the range. It is known that a variety of factors affect airborne bacteria concentration in poultry houses, including animal activity, thermal environment, ventilation, and manure/litter management. Controlling indoor gas and dust concentrations may also lead to reduced airborne bacteria concentration. For instance, to reduce ammonia volatilization from the manure and make the manure easier to handle, manure on belts was continuously blow dried. Minimum ventilation rates are
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
samples was examined; however, the bacterial concentration was not determined as the bacterial growth rate in the enrichment medium was unknown. The principle detection limit of the plating culture method is 1.7 log CFU/m3 (assuming 1 bacteria in 2 mL undiluted sample is enrichable). One gram of the composite manure sample was transferred into 10 mL physiological saline, followed by vortexing at 3,000 rpm for 1 min. The mixture was left standing for 5 min, and bacterial concentration in the supernatant was determined using the same methods as for the liquid air samples. The enrichment test was not performed to the manure samples as the fecal bacteria were abundant enough for plate culturing.
5
AIRBORNE BACTERIA OF THREE LAYING-HEN HOUSING TYPES
also provided in wintertime to ensure that indoor RH does not become excessive. Both practices would help maintaining a relative dry indoor environment that is less favorable for bacterial growth and multiplication. Furthermore, hens in AV were allowed to access litter only part of the day, as opposed to full-day access that is typically practiced in Europe. This management strategy may have reduced the overall generation of airborne bacteria and particles from litter because of less hen activities. Although the negative effects of airborne bacteria on animal and human health are well recognized, no threshold limit value (TLV) or permissible exposure limit (PEL) exists in the working environment. The bacteria are usually not a single species but rather exist in complex mixtures. The inhalation of airborne bacteria, their deposition in respiratory tract, and final infection are complicated processes and affected by bacterial particle size, respiration pattern (oral vs. nasal), use of personal protection equipment or not, recipient health status, etc. Also, the bacteria concentrations measured with some samplers might underestimate the real biological contamination levels because of inherent sampling stress that inactivates bacterial viability during sampling. For these reasons, it is difficult or even impossible to establish TLV and PEL at the present time. Instead of a compulsory threshold, Occupational Safety & Health Administration (OSHA, 1999) recommended a contamination-indicating level of 3 log CFU/m3 . Concentrations in excess of this level does not necessarily
Table 2. Summary of statistical analysis results for airborne total bacteria concentration and factors. Effect House Climate House × climate Moment
Num DF
Den DF
F value
P-value
2 2 4 1
3 83 83 83
67.36 50.57 12.11 19.02
0.0032 < 0.0001 < 0.0001 < 0.0001
Note: ‘Num DF’ is the degrees of freedom of the numerator; ‘Den DF’ is the degrees of freedom of the denominator.
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Figure 2. Airborne total bacteria concentrations in the conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses at two sampling moments (morning or ‘AM’, and afternoon or ‘PM’) and three climatic conditions (cold, mild, and warm). The error bars are 95% confidence interval (CI95); means with non-overlapping error bars are significantly different (P < 0.050).
imply an unsafe environment, but might require further surveillance to assess the health hazard to human. Figure 2 shows the airborne total bacteria concentrations in the three hen houses measured in the morning and afternoon under the cold, mild, and warm climatic conditions. The AV house tended to have lower bacteria concentrations under mild (P = 0.025) and warm (P < 0.001) climatic conditions than under cold, likely owing to the dilution effect of higher ventilation rate. A similar phenomenon was reported in swine barns (Duchaine et al., 2000). This dilution effect was, however, not obvious in CC or EC houses. The probable reason for these different results in AV vs. CC/EC might be that the AV house has much higher bacteria generation rates from the bird activities on the litter which would continue to benefit from increasing VR; in comparison, the CC and EC houses have much lower (limited) generation rate and further increasing VR would not provide as much of dilution effect. Due to the presence of litter and hen activities on it, the AV house consistently had a much higher bacteria concentration than the CC and EC houses for all climatic conditions. At cold condition, total bacteria in AV was 1.5 to 1.9 log higher than (or 30 to 80 times) those of CC and EC. As ambient temperature increased, the concentration differences between AV and CC/EC became smaller. In the AV house, bacteria concentrations measured in the early morning were numerically lower than in the afternoon. This is because the measurements were taken before litter access in the early morning and after litter access in the afternoon. The CC (P = 0.995) and EC (P = 1.000) houses had similar bacteria concentrations in the morning and afternoon. Table 2 shows results of statistical analysis for airborne total bacteria concentration and factors using GLMMIX procedure in SAS. It can be seen that all factors and interaction of concern affected the indoor airborne total bacterial concentrations. In the past few decades, there has been increased attention to small particulate matters (e.g., PM10 ) for their adverse occupational health and environment impact caused by their microbiological and chemical components (Cambra-Lopez et al., 2010). The continuous measurement of PM10 concentrations in the CSES environmental monitoring allowed us to delineate the relationship between airborne bacteria and this PM fraction in the three laying-hen houses. It is shown that higher airborne total bacteria concentrations generally coincided with higher PM10 concentrations in all three
6
ZHAO ET AL.
houses (Figure 3), indicating that PM is the carrier of airborne microorganisms. The difference in regression equations among the houses shown in Figure 3 is a result of heterogeneity in dust size distribution and culturable bacteria count per unit weight of dust particles attached. While Figure 3 reveals the positive relationship between airborne total bacteria and PM10 concentrations, cautions should be taken when interpreting the results as bacteria measured in this study are those carried by total dust (all particles >0.3 μm) but the PM10 fraction contains particles equal to and smaller than 10 μm. Use of logarithmic regression in Figure 3 is based on the assumptions of constant PM10 mass fraction to total dust and uniform bacterial distribution in the dust, i.e., total bacteria count in dust is assumed to be proportional to the dust mass/volume.
Emission Rate of Airborne Total Bacteria The overall means and SD of airborne total bacteria emission rates in log CFU/[h-hen] (or log CFU/[h-AU]) were 4.8 ± 0.4 (or 7.3 ± 0.4) for CC, 6.1 ± 0.7 (or 8.6 ± 0.7) for AV, and 4.8 ± 0.5 (or 7.3 ± 0.5) for EC. The emission rates found in CC and EC were in line with the data (7.2 log CFU/[h-AU]) reported by Seedorf et al. (1998) for laying-hen house in Europe, but
Table 3. Summary of statistical analysis results for airborne total bacteria emission and factors. Effect House Climate House × climate Moment
Num DF
Den DF
F value
P-value
2 2 4 1
3 38 38 38
37.17 0.51 6.61 20.16
< 0.0001 0.6043 0.0004 < 0.0001
Note: ‘Num DF’ is the degrees of freedom of the numerator; ‘Den DF’ is the degrees of freedom of the denominator.
they were much lower than that for AV (P < 0.001). The house × climate interaction and sampling moment affected emission rates as well (Table 3). Figure 4 shows airborne total bacteria emission rates in the three hen houses measured in the morning and afternoon at the cold, mild, and warm climatic conditions. At mild (P = 0.012) and warm (P = 0.047) climatic conditions, the emission rates of AV were higher in the afternoon than those measured in the morning, which is again due to dust generation by hen activities on the litter floor. Emission rates in AM and PM were not different in EC under any climatic conditions (P = 0.996 for cold, P = 0.985 for mild, and P = 1.000 for warm), nor in CC (P = 1.000 for cold, P = 0.280 for mild, and P = 1.000 for warm). Emission rates of airborne total bacteria were positively related with the PM10 emission rates in all
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Figure 3. Relationship between total bacteria and PM10 concentrations (both measured at air exhaust) in conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses, and for pooled data (data from all three houses).
AIRBORNE BACTERIA OF THREE LAYING-HEN HOUSING TYPES
7
Concentration of Airborne Gram− Bacteria
houses, as shown in Figure 5. It should be emphasized that emission rates obtained in this study were based on measurements taken in daytime only. The emission rates were not monitored at night when bacterial concentration and VR were expected to differ from those during daytime as a result of much lower bird activities and lower ambient temperature at night.
Figure 5. Relationship between total bacteria and PM10 emission rates in conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses, and for pooled data (data from all three houses).
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Figure 4. Airborne total bacteria emission rates of the conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses at two sampling moments (morning or ‘AM’, and afternoon or ‘PM’) and three climatic conditions (cold, mild, and warm). The error bars are 95% confidence interval (CI95); means with non-overlapped error bars are significantly different (P < 0.050).
Results of airborne Gram− bacteria measurements are shown in Table 4. Out of 32 total air samples collected in each house, 2 samples (6%) in CC, 7 (22%) samples in AV, and 2 (6%) samples in EC were found to be positive for Gram− bacteria (Table 4). The statistical analysis shows that presence of Gram− bacteria was not affected by house (P = 0.312), climate (P = 0.968), or sampling moment (P = 0.804). Most of these positive samples were confirmed using enrichment method, indicating that the airborne Gram− bacteria concentrations were in the range of 1.7 (detection limit of enrichment test) to 2.7 (detection limit of plating culture) log CFU/m3 . The only positive sample confirmed using plating culture was found on 5/10/13 in the AV house at a concentration of 3.5 log CFU/m3 . European and Australian studies reported a wide range of Gram− bacteria concentration from 1 to 4.7 log CFU/m3 in laying-hen houses. Those studies also found that Gram− bacteria accounted for 0.6 to 15% of the total number of culturable bacteria (Seedorf et al., 1998; Zucker et al., 2000; Agranovski et al., 2007). The result of our study showed the Gram− to total bacteria ratio of <2%. Considering the possible low presence of Gram− bacteria, increasing sampling duration and/or using a larger aliquot for
8
ZHAO ET AL. Table 4. Airborne Gram− bacteria concentrations in the exhaust and middle areas of the conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses. Gram− bacteria concentration (log CFU/m3 ) AV
CC Exhaust
Middle
Exhaust
EC
Middle
Exhaust
Middle
Date (m/d/y)
AM
PM
AM
PM
AM
PM
AM
PM
AM
PM
AM
PM
1/9/13 2/16/13 3/16/13 4/11/13 5/10/13 6/7/13 7/4/13 8/9/13
ND ND ND ND ND ND ND ND
ND + ND ND ND ND ND ND
ND ND ND ND ND ND ND +
ND ND ND ND ND ND ND ND
ND ND ND ND ND ND ND +
ND ND + + ND ND ND ND
ND ND + ND ND ND ND ND
ND ND + + 3.5 ND ND ND
ND ND ND ND ND ND ND ND
ND ND ND ND ND ND ND ND
ND ND ND ND ND ND ND +
ND ND ND ND ND ND + ND
Table 5. Bacteria concentrations in the manure and litter of the conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses. Total bacteria
Bacteria concentration (mean ± SD, log CFU/g) Gram− bacteria
Manure on belt
As-is Dry-base
Litter
Manure on belt
CC
AV
EC
AV
CC
AV
9.5 ± 0.9 10.0 ± 0.9
9.2 ± 0.6 9.7 ± 0.6
8.8 ± 0.7 9.3 ± 0.6
9.2 ± 0.8 9.3 ± 0.8
7.9 ± 0.6 8.4 ± 0.6a a
Litter EC
7.3 ± 0.2 7.7 ± 0.2a a
AV
7.4 ± 0.8 7.9 ± 0.8a a
4.5 ± 0.6b 4.5 ± 0.6b
Note: ‘As-is’ concentration is the concentration calculated based on weight of as-is manure; ‘Dry base’ concentration is the concentration calculated based on weight of manure dry matter. Gram− bacteria concentrations in the same type of manure (as-is or dry-base) with different lower-case superscripts are significantly different (P < 0.001). Each datum is a mean of six values measured once every month from March to August 2013.
plating culture to get a countable bacteria number are recommended.
Bacteria Concentrations of Manure and Floor Litter As an important source of airborne microorganisms, manure and litter were analyzed for the fecal bacteria concentrations in this study. The total bacteria concentrations in the hen houses were 8.8 to 9.2 log CFU/g for as-is manure and litter, and 9.3 to 10.0 log CFU/g for dry-base manure and litter (Table 5). Total bacteria concentrations of manure and floor litter were similar. The Gram− bacteria in manure (7 to 9 log CFU/g) were more abundant than in the floor litter (4.5 ± 0.6 log CFU/g) (P < 0.001). This result can be explained by the fact that Gram− bacteria are more vulnerable to dehydration stress than Gram+ bacteria. The moisture contents of manure on the belts were 69.4 ± 5.6% in CC, 66.2 ± 5.3% in AV, 67.3 ± 2.3% in EC; whereas moisture content of the litter in AV was only 15.7 ± 2.6%. The fraction of Gram− bacteria (of total bacteria) in manure was much greater than that in airborne particles (P < 0.001). The difference in these fractions indicates that (wet) manure on belt is not the major source of bacterial particulates in laying-hen houses. The high moisture content of manure prevents fecal particles from becoming airborne. In contrast, the much
drier floor litter could be easily agitated by animal activity and eventually contribute to the higher airborne bacteria concentrations.
CONCLUSIONS This study investigated the concentrations and emission rates of airborne bacteria of three types of layinghen houses, i.e., conventional cage, enriched colony, and aviary. The monitoring was performed for an 8-month period covering mid to end of the production flock. The relationship between airborne total bacteria and PM10 is delineated. The following observations and conclusions were made. r Concentrations and emission rates of airborne to-
tal bacteria in the laying-hen houses were affected by housing type and climatic condition, and were variable within the day. r The AV house had much higher airborne total bacteria concentrations and emission rates as compared to the CC and EC houses. r Higher ventilation rate in warmer weather helped diluting the airborne bacteria concentration in the AV houses, although the dilution effect was not as obvious in CC and EC houses. r In the AV house, the total bacteria concentration and emission rates tended to be higher in the afternoon than in the morning due to more bacterial
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Note: ‘ND’ is not detectable; ‘+’ means a sample is positive in the enrichment test, but negative in plating culture.
AIRBORNE BACTERIA OF THREE LAYING-HEN HOUSING TYPES
particles generated by hen activities on the floor litter. r Total bacteria were positively correlated with PM 10 in terms of concentration and emission rate. r Airborne Gram− bacteria were present in the laying-hen houses at low concentrations. The fraction of Gram− bacteria (of the total bacteria) was estimated to be less than 2%. r Manure on the belt and floor litter had similar total bacteria concentrations; however, litter contained much lower levels of Gram− bacteria.
ACKNOWLEDGMENTS
REFERENCES Agranovski, V., T. Reponen, and Z. D. Ristovski. 2007. Survey of bioaerosol emissions from Australian poultry buildings. Proc. European Aerosol Conference 2007, Salzburg, Austria. Cambra-Lopez, M., A. J. A. Aarnink, Y. Zhao, S. Calvet, and A. G. Torres. 2010. Airborne particulate matter from livestock production systems: A review of an air pollution problem. Environ. Pollut. 158:1–17. de Reu, K., W. Messens, M. Heyndrickx, T. B. Rodenburg, M. Uyttendaele, and L. Herman. 2008. Bacterial contamination of table eggs and the influence of housing systems. World’s Poult. Sci. J. 64:5–19. Duchaine, C., Y. Grimard, and Y. Cormier. 2000. Influence of building maintenance, environmental factors, and seasons on airborne contaminants of swine confinement buildings. Am. Ind. Hyg. Assoc. J. 61:56–63. Hu, D., L. Wang-Li, O. D. Simmons, III, J. J. Classen, and J. A. Osborne. 2015. Spatiotemporal Variations of Bioaerosols in the Vicinity of an Animal Feeding Operation Facility in the US. J. Environ. Prot. 6:614–627. Lonc, E., and K. Plewa. 2011. Comparison of Indoor and Outdoor Bioaerosols in Poultry Farming in Advanced Topics in Environ-
mental Health and Air Pollution Case Studies. A. Moldoveanu, ed. InTech. ˇ ˇ c, D. Maruˇsi´c, B. Matkovi´c, K., M. Vuˇcemilo, I. Stokovi´ c, R. Simi´ Vinkovi´c, and S. Matkovi´c. 2013. Concentrations of airborne bacteria and fungi in a livestock building with caged laying hens. Veterinarski Arhiv. 83:413–424. Muller, W., and P. Wieser. 1987. Dust and microbial emissions from animal production. Pages 47–89 in Animal production and environmental health. D. Strauch, ed. Elsevier. Oppliger, A., N. Charri`ere, P. Droz, and T. Rinsoz. 2008. Exposure to bioaerosols in poultry houses at different stages of fattening; use of real-time PCR for airborne bacterial quantification. Ann. Occup. Hyg. 52:405–412. OSHA. 1999. OSHA technical manual (OTM) section III: Chapter 2 Indoor Air Quality Investigation (Directive Number TED 01-00-015). Occupational Safety & Health Administration. Rylander, R., and M. F. Carvalheiro. 2006. Airways inflammation among workers in poultry houses. Int. Arch. Occ. Env. Hea. 79:487. Seedorf, J., J. Hartung, M. Schroder, K. H. Linkert, V. R. Phillips, M. R. Holden, R. W. Sneath, J. L. Short, R. P. White, S. Pedersen, H. Takai, J. O. Johnsen, J. H. M. Metz, P. W. G. Groot Koerkamp, G. H. Uenk, and C. M. Wathes. 1998. Concentrations and emissions of airborne endotoxins and microorganisms in livestock buildings in Northern Europe. J. Agric. Eng. Res. 70:97–109. Shepherd, T. A., Y. Zhao, H. Li, J. P. Stinn, M. D. Hayes, and H. Xin. 2015. Environmental assessment of three egg production systems—Part II. Ammonia, greenhouse gas, and particulate matter emissions. Poult. Sci. 94:534–543. Thelin, A., O. Tegler, and R. Rylander. 1984. Lung reactions during poultry handling related to dust and bacterial endotoxin levels. Eur. J. Respir. Dis. 65:266–271. ˇ Vuˇcemilo, M., B. Vinkovi´c, K. Matkovi´c, I. Stokovi´ c, S. Jakˇsi´c, - . Stubiˇcan. 2010. The influS. Radovi´c, K. Grani´c, and D ence of housing systems on the air quality and bacterial eggshell contamination of table eggs. Czech J. Anim. Sci. 55:243–249. Wang-Li, L., Q. Li, and G. E. Byfield. 2013. Identification of bioaerosols released from an egg production facility in the southeast United States. Environ. Eng. Sci. 30:2–10. Zhao, Y., A. J. A. Aarnink, M. C. M. de Jong, and P. W. G. Groot Koerkamp. 2014. Airborne microorganisms from livestock production systems and their relation to dust. Crit. Rev. Env. Sci. Tec. 44:1071–1128. Zhao, Y., T. A. Shepherd, J. Swanson, J. A. Mench, D. M. Karcher, and H. Xin. 2015a. Comparative evaluation of three egg production systems: Housing characteristics and management practices. Poult. Sci. 94:475–484. Zhao, Y., T. A. Shepherd, H. Li, and H. Xin. 2015b. Environmental assessment of three egg production systems–Part I: Monitoring system and indoor air quality. Poult. Sci. 94:518–533. Zucker, B. A., S. Trojan, and W. Muller. 2000. Airborne Gramnegative bacterial flora in animal houses. J. Vet. Med. B. 47:37–46.
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Cash funding for the study was provided by the Coalition for Sustainable Egg Supply (CSES). In-kind contributions by Iowa State University and the Egg Industry Center (located at ISU) were provided by availing the state-of-the-art environmental monitoring equipment to the project. We sincerely appreciate the cooperation and assistance of the egg producer in the implementation of this field study. We also thank Lendie Follett of Statistics Department at ISU for her assistance in the statistical analysis.
9
Department of Agricultural and Biosystems Engineering, Iowa State University, Ames, IA, USA; † Department of Mechanical Engineering, Nanjing Institute of Industry Technology, Nanjing, China; ‡ Key Laboratory of the Ministry of Agriculture for Agricultural Engineering in Structure and Environment, China Agricultural University, Beijing, China; and § Department of Agricultural Structures and Irrigation, Suleyman Demirel University, Isparta, Turkey trations in the afternoon (with litter access) than in the morning (without litter access). The overall means and standard deviation of airborne total bacteria emission rates, in log CFU/[h-hen] (or log CFU/[h-AU], AU = animal unit or 500 kg live weight) were 4.8 ± 0.4 (or 7.3 ± 0.4) for CC, 6.1 ± 0.7 (or 8.6 ± 0.7) for AV, and 4.8 ± 0.5 (or 7.3 ± 0.5) for EC. Both concentration and emission rate of airborne total bacteria were positively related to PM10 . Gram− bacteria were present at low concentrations in all houses; and only 2 samples (6%) in CC, 7 (22%) samples in AV, and 2 (6%) samples in EC out of 32 air samples collected in each house were found positive with Gram− bacteria. The concentration of airborne Gram− bacteria was estimated to be <2% of the total bacteria. Total bacteria counts in manure on belt (in all houses) and floor litter (only in AV) were similar; however, the manure had much more Gram− bacteria than the litter. The results point out the need to mitigate airborne total bacteria in laying-hen houses, especially in AV houses.
Key words: indoor air quality, total bacteria, Gram-negative bacteria, airborne, alternative hen housing 2016 Poultry Science :1–9 http://dx.doi.org/10.3382/ps/pew053
INTRODUCTION
In addition to their local hazards, the microorganisms can be exhausted and transported in an ambient atmospheric environment, thus posing regional health risks (Seedorf et al., 1998). Therefore, indoor concentrations and facility emissions of airborne microorganisms are critical indicators of occupational health, animal wellbeing, food safety, and environmental impact. Efforts dedicated to quantify concentration and emission of airborne microorganisms have been made to obtain baseline values for poultry houses. The reported microbial concentrations in hen houses vary largely, ranging from 4 to 9 log colony forming unit (CFU)/m3 (Muller and Wieser, 1987; Seedorf et al., 1998; Oppliger et al., 2008; Matkovi´e et al., 2013; Wang-Li et al., 2013). The wide range of values is believed to result from the differences in production systems (litter-based vs. cage-based systems), thermal environment, and house ventilation (Zhao et al., 2014). Compared to indoor
Airborne microorganisms, with a large portion being bacterial species (Muller and Wieser, 1987), are usually found at high concentrations in laying-hen houses (Seedorf et al., 1998). These microorganisms are typically associated with dust particles, and can cause infection or trigger respiratory diseases to animals and caretakers after inhalation (Thelin et al., 1984; Rylander and Carvalheiro, 2006; Cambra-Lopez et al., 2010). It is also a concern that the microorganisms may deposit to the eggshell and result in contamination of table eggs (de Reu et al., 2008; Vuˇcemilo et al., 2010). C 2016 Poultry Science Association Inc. Received July 13, 2015. Accepted January 3, 2016. 1 Corresponding author: [email protected]
1
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
ABSTRACT Airborne microorganism level is an important indoor air quality indicator, yet it has not been well documented for laying-hen houses in the United States. As a part of the Coalition for Sustainable Egg Supply (CSES) environmental monitoring project, this study comparatively monitored the concentrations and emissions of airborne total and Gram-negative (Gram− ) bacteria in three types of commercial laying-hen houses, i.e., conventional cage (CC), aviary (AV), and enriched colony (EC) houses, over a period of eight months covering the mid and late stages of the flock cycle. It also delineated the relationship between airborne total bacteria and particulate matter smaller than 10 μm in aerodynamic diameter (PM10 ). The results showed airborne total bacteria concentrations (log CFU/m3 ) of 4.7 ± 0.3 in CC, 6.0 ± 0.8 in AV, and 4.8 ± 0.3 in EC, all being higher than the level recommended for human environment (3.0 log CFU/m3 ). The much higher concentrations in AV arose from the presence of floor litter and hen activities on it, as evidenced by the higher concen-
2
ZHAO ET AL.
MATERIALS AND METHODS Hen Houses The CSES environmental monitoring was carried out in three hen housing systems (CC, AV, and EC) located at the same farm in the Midwest United States. The CC house had a nominal capacity of 200,000 hens and was equipped with manure belts that conveyed the accumulated manure out of the house twice a week. The AV house had a nominal capacity of 50,000 hens and was provided with colonies and litter area accessible by the hens part of a day to perform foraging and dustbathing behaviors. Manure belts were installed in all hen colonies to remove manure out of the house twice a week, while the manure deposited/accumulated on the litter floor was removed at the end of each flock. The EC house also had a nominal capacity of 50,000 hens, and all manure was deposited onto the manure belts and was removed out of the house twice a week.
Detailed description of the housing systems and management practices was provided by Zhao et al. (2015a).
Samplers and Measurements All glass impingers (AGI-30, Ace Glass Inc., Vineland, NJ) were used for sampling airborne bacteria. The AGI-30 has a collection cut-off diameter of 0.3 μm, which means that collection efficiency is 50% for particles with aerodynamic diameter of 0.3 μm and is greater for larger particles. The sampler collects microorganisms in the air going through its sampling line into 20 mL liquid medium at an orifice-regulated air flow rate of 12.5 L/min. In this study, the medium used to collect the bacteria was sterile physiological saline (0.9% sodium chloride solution). The sampling medium was stored in individual 50 mL centrifuge tubes (20 mL liquid medium per tube), and kept in an ice chest at approximately 4◦ C before it was used. After collection, the microbe-laden liquid was transferred to its original centrifuge tube, put back in the ice chest, and transported to the Microbiology Analytical Lab at ISU campus for further analysis. Airborne bacteria were measured one day every month from January to August 2013, covering mid to near the end of a production cycle (45 to 78 weeks of age). To cover spatial and temporal variations, on each sampling day the bacterial concentrations were measured at the air exhausts (near stage-1 fan) and in the middle of the three houses in both the morning and the afternoon. This experimental design is important, especially in the AV house where higher airborne bacteria levels were expected in the afternoon when the hens had access to the litter. Figure 1 shows the sampling locations for airborne bacteria and PM10 . Gas concentrations and other parameters (e.g., temperature, relative humidity or ‘RH’, building static pressure or ‘SP’) were also measured in the environmental monitoring, but are not marked in the figure as their locations and results have been published in other companion papers (Zhao et al., 2015a; Zhao et al., 2015b). The sampling duration with AGI-30 is typically less than 30 min to avoid excessive medium evaporation that would cause a drop in collection efficiency. In this study, each sampling lasted for 15 min, which led to about 15% of the liquid medium being evaporated (about 17 of 20 mL was left after sampling). Ambient airborne bacteria level was not measured because it is negligible as compared to indoor concentrations (Lonc and Plewa, 2011; Wang-Li et al., 2013; Hu et al., 2015). On each sampling day from March to August, fresh manure on belts was collected at six different locations in each hen house, and made into a composite sample by mixing and well stirring it in a centrifuge tube. The bacteria concentration and moisture content were determined in these manure samples. Environmental conditions (temperature and RH), ventilation conditions (fan operation status and SP),
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
concentration, emission rates of microorganisms have been less investigated and are reported with different units, making it difficult to compare results across studies. Seedorf et al. (1998) reported total bacteria emission rate of 7.2 log CFU/[h-AU] (1 AU or animal unit = 500 kg live body weight) for cage laying-hen houses, and 9.5 log CFU/[h-AU] for broiler houses. Agranovski et al. (2007) reported a maximum emission rate of 9 log CFU/[h-house] for broiler houses; however bird numbers and body weight were not provided to allow estimation of emission rate on per AU basis. Most previous work for microbial measurement in hen houses was performed in European countries (especially for alternative housing systems) and focused on one specific housing system at a time. These baseline data would not necessarily reflect the situation in the United States where different management practices and housing systems are used. The Coalition for Sustainable Egg Supply (CSES) Project was launched in 2011. The goal of the project was to holistically evaluate the impacts of two alternative egg production systems, i.e., aviary (AV) and enriched colony (EC) housing systems, as compared to the conventional cage (CC) production system, with regards to hen well-being and health, egg quality and safety, environmental impact, food affordability, and worker safety. As a part of the CSES research team, our group at Iowa State University (ISU) examined the environmental impact of the three hen housing systems, and parts of the results have been published in a series of companion papers (Shepherd et al., 2015; Zhao et al., 2015a,b). This paper reports the concentrations and emission rates of airborne total and Gram-negative (Gram− ) bacteria in the CC, AV, and EC houses, and their relationship with particulate matter of 10 μm or smaller in aerodynamic diameter (PM10 ).
3
AIRBORNE BACTERIA OF THREE LAYING-HEN HOUSING TYPES
and PM10 concentrations were continuously monitored. Air temperature and RH were measured, respectively, with type-T thermocouples (Cole-Parmer, Vernon Hills, IL) and capacitance-type humidity sensors (HMP 61U, Vaisala Inc., Woburn, MA) at 1 s intervals. PM10 concentrations were measured with real-time Tapered Element Oscillating Microbalances (TEOM, Model 1400a, Thermo Fisher Scientific Inc., Waltham, MA) that were set to a 300 s integration time.
Sample Analysis The outside of all centrifuge sample tubes was cleaned and disinfected using alcohol before being transported to the biosafety cabinet for analysis. The samples were homogenized by vortexing for 5 s, and then viable counts of total bacteria and Gram− bacteria in each sample were determined by plating 0.2 mL portions directly and after serially diluted (1:10) in physiological saline onto trypticase soy agar (TSA, for total bacteria, Catalog No. R455002, Fisher Scientific, Hanover Park, IL) and Macconkey No. 3 agar (for Gram− bacteria, Catalog No. OXCM0115B, Fisher Scientific) plates. The plates were aerobically incubated at 37◦ C for 24 h (total bacteria) or 48 h (Gram− bacteria). The colonies formed on plates (30 to 300 colonies) were counted and used for calculating concentration of airborne bacteria (Equation 1, ‘log’ in this study is base-10 logarithm). The principle detection limit of the plating culture method is 2.7 log CFU/m3 (1 colony on the agar plate with undiluted sample after incubation). Emission rate was calculated using concentration data and ventilation rate (VR) simultaneously measured at the corresponding 15 min bacteria sampling time (Equation 2). The VR was corrected to standard temperature and
pressure conditions.
C = log10
1 N × 10n × Vs × Vp Va
(1)
where C is the airborne bacteria concentration, log CFU/m3 ; N is the number of colonies on a countable plate (30 to 300 colonies); n is serial dilution factor (n = 0 for undiluted sample, n = 1 for 10-fold diluted sample, etc.); Vp is the sample volume plated, mL (Vp = 0.2 mL in this study); Vs is total volume of original liquid sample, mL; and Va is the total air volume sampled using AGI-30, m3 (Va = 0.1875 m3 in this study).
E = log10
Tstd Pa 10 × V R × × Th Pstd C
(2)
where E is airborne bacteria emission rate, log CFU/[hhen] or log CFU/[h-AU]; VR is building ventilation rate, m3 /[h-hen] or m3 /[h-AU]; Tstd is standard temperature (273.15 K); Th is house temperature, K; Pstd is standard atmospheric pressure (101.325 kPa); and Pa is actual atmospheric pressure, kPa. Enrichment was further conducted to boost levels of bacteria that were potentially too low for detection using direct plating. For total bacteria, 2 mL of the undiluted samples was transferred to a sterile tube containing 50 mL of trypticase soy broth (TSB, Catalog No. R455052, Fisher Scientific), and incubated at 37◦ C for 48 h. Following enrichment, 0.5 mL culture suspension was spread on TSA plates, and the plates were incubated at 37◦ C for 48 h before being checked. The same procedure was applied to Gram− bacteria enrichment and culturing, except that Macconkey broth (7185, Neogen Corp., Lansing, MI) and Macconkey No. 3 agar were used. The presence of bacteria in the enriched
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Figure 1. A schematic representation of the house layout and sampling locations for airborne bacteria in the conventional cage (CC), enriched colony (EC), and aviary (AV) laying-hen houses. Particulate matter (PM) concentrations and thermal environment conditions were measured at the same locations.
4
ZHAO ET AL. Table 1. Daily mean air temperature, relative humidity, ventilation rate, and PM10 concentration in the conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses on the sampling days for airborne bacteria. Air temperature (◦ C)
Relative humidity (%)
Ventilation rate (m3 /[h-hen])
PM10 concentration (μ g/m3 )
Date (d/m/y)
Climate
Amb
CC
AV
EC
Amb
CC
AV
EC
CC
AV
EC
CC
AV
EC
1/9/13 2/16/13 3/16/13 4/11/13 5/10/13 6/7/13 7/4/13 8/9/13 Avg±SD
Cold Cold Cold Cold Mild Mild Warm Warm
−2.1 −7.4 1.8 0.9 12.6 14.9 22.2 21.4 8.0±11.2
23.2 22.7 23.6 23.1 24.4 24.7 26.0 26.3 24.3±1.4
26.4 26.1 26.7 26.6 26.8 27.1 26.7 27.0 26.7±0.3
25.7 25.5 25.3 25.1 25.4 25.6 25.0 25.5 25.4±0.3
74 76 82 96 56 74 64 69 74±12
66 65 63 67 47 53 53 55 59±8
60 60 57 59 45 50 50 53 54±5
62 66 63 60 50 53 54 57 58±6
0.5 0.5 0.7 0.6 1.5 1.6 3.4 3.5 1.5±1.2
0.5 0.6 0.7 0.6 1.1 1.5 4.4 3.5 1.6±1.5
0.5 0.6 0.8 0.6 1.2 1.6 4.8 4.6 1.9±1.8
695 908 701 879 699 762 526 555 716±135
8221 9151 9026 9815 7171 7535 2225 2406 6944±2984
645 843 680 768 563 584 390 404 610±160
Note: ‘Avg’ stands for average, ‘SD’ stands for standard deviation.
Data Analysis Statistical analysis was performed to investigate the airborne bacteria concentration and emission rate as affected by house type, sampling location, sampling moment, and climatic condition using Statistical Analysis System version 9.3 (SAS 9.3, SAS Institute Inc., Cary, NC). Based on sampling dates, average daily ambient temperature and VR, three climatic conditions were categorized, i.e., cold (January to April), mild (May and June), and warm (July and August). The actual ambient temperatures and VR of these three climatic conditions are presented in Table 1. For consistency with analysis in companion papers in this series, climate condition was selected as model factor instead of VR, although they are inherently interchangeable. A model fitting process was performed using PHREG procedure with all concerned factors included. The sampling location was confirmed as the only insignificant factor (P = 0.537) and was excluded from subsequent analysis. The effects of remaining factors on bacteria concentration and emission rate (or ‘Y’ in equation 3) were then tested using GLIMMIX procedure (Equation 3). A random term of house × location was included to account for dependency of measurements taken from the same house in the same location. The effects were considered significant at a probability level of 0.050. Since Gram− bacteria was non-detectable in most of the air samples, the presence (yes/no) of Gram− bacteria as
affected by the factors was tested assuming a binomial distribution.
Y = house + climate + house × climate + moment (3) Regression analyses was performed to delineate the relationships between concentration and emission rate of airborne bacteria vs. PM10 .
RESULTS AND DISCUSSION Temperature, Relative Humidity (RH), and Ventilation Rate (VR) The daily mean ambient temperature on the sampling days ranged from −7.4 to 22.2◦ C, with overall mean and standard deviation or ‘SD’ of 8.0 ± 11.2◦ C. The indoor temperatures and RH were well maintained at the thermal comfortable zone for laying hens in all three houses, 25.4 ± 1.3◦ C and 57 ± 6%. Daily mean VR in m3 /[h-hen] and SD were 1.5 ± 1.2 for CC, 1.6 ± 1.5 for AV, and 1.9 ± 1.8 for EC. Daily mean PM10 concentrations in μg/m3 and SD were 716 ± 135 for CC, 6944 ± 2984 for AV, and 610 ± 160 for EC. Table 1 shows the details of the results.
Concentration of Airborne Total Bacteria The overall mean of airborne total bacteria concentration in log CFU/m3 was 4.7 ± 0.3 for CC and 4.8 ± 0.3 for EC, and both were significantly lower than 6.0 ± 0.8 for AV (P = 0.002 and 0.003, respectively). These concentration levels were within the range reported in previous studies, but at the lower end of the range. It is known that a variety of factors affect airborne bacteria concentration in poultry houses, including animal activity, thermal environment, ventilation, and manure/litter management. Controlling indoor gas and dust concentrations may also lead to reduced airborne bacteria concentration. For instance, to reduce ammonia volatilization from the manure and make the manure easier to handle, manure on belts was continuously blow dried. Minimum ventilation rates are
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
samples was examined; however, the bacterial concentration was not determined as the bacterial growth rate in the enrichment medium was unknown. The principle detection limit of the plating culture method is 1.7 log CFU/m3 (assuming 1 bacteria in 2 mL undiluted sample is enrichable). One gram of the composite manure sample was transferred into 10 mL physiological saline, followed by vortexing at 3,000 rpm for 1 min. The mixture was left standing for 5 min, and bacterial concentration in the supernatant was determined using the same methods as for the liquid air samples. The enrichment test was not performed to the manure samples as the fecal bacteria were abundant enough for plate culturing.
5
AIRBORNE BACTERIA OF THREE LAYING-HEN HOUSING TYPES
also provided in wintertime to ensure that indoor RH does not become excessive. Both practices would help maintaining a relative dry indoor environment that is less favorable for bacterial growth and multiplication. Furthermore, hens in AV were allowed to access litter only part of the day, as opposed to full-day access that is typically practiced in Europe. This management strategy may have reduced the overall generation of airborne bacteria and particles from litter because of less hen activities. Although the negative effects of airborne bacteria on animal and human health are well recognized, no threshold limit value (TLV) or permissible exposure limit (PEL) exists in the working environment. The bacteria are usually not a single species but rather exist in complex mixtures. The inhalation of airborne bacteria, their deposition in respiratory tract, and final infection are complicated processes and affected by bacterial particle size, respiration pattern (oral vs. nasal), use of personal protection equipment or not, recipient health status, etc. Also, the bacteria concentrations measured with some samplers might underestimate the real biological contamination levels because of inherent sampling stress that inactivates bacterial viability during sampling. For these reasons, it is difficult or even impossible to establish TLV and PEL at the present time. Instead of a compulsory threshold, Occupational Safety & Health Administration (OSHA, 1999) recommended a contamination-indicating level of 3 log CFU/m3 . Concentrations in excess of this level does not necessarily
Table 2. Summary of statistical analysis results for airborne total bacteria concentration and factors. Effect House Climate House × climate Moment
Num DF
Den DF
F value
P-value
2 2 4 1
3 83 83 83
67.36 50.57 12.11 19.02
0.0032 < 0.0001 < 0.0001 < 0.0001
Note: ‘Num DF’ is the degrees of freedom of the numerator; ‘Den DF’ is the degrees of freedom of the denominator.
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Figure 2. Airborne total bacteria concentrations in the conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses at two sampling moments (morning or ‘AM’, and afternoon or ‘PM’) and three climatic conditions (cold, mild, and warm). The error bars are 95% confidence interval (CI95); means with non-overlapping error bars are significantly different (P < 0.050).
imply an unsafe environment, but might require further surveillance to assess the health hazard to human. Figure 2 shows the airborne total bacteria concentrations in the three hen houses measured in the morning and afternoon under the cold, mild, and warm climatic conditions. The AV house tended to have lower bacteria concentrations under mild (P = 0.025) and warm (P < 0.001) climatic conditions than under cold, likely owing to the dilution effect of higher ventilation rate. A similar phenomenon was reported in swine barns (Duchaine et al., 2000). This dilution effect was, however, not obvious in CC or EC houses. The probable reason for these different results in AV vs. CC/EC might be that the AV house has much higher bacteria generation rates from the bird activities on the litter which would continue to benefit from increasing VR; in comparison, the CC and EC houses have much lower (limited) generation rate and further increasing VR would not provide as much of dilution effect. Due to the presence of litter and hen activities on it, the AV house consistently had a much higher bacteria concentration than the CC and EC houses for all climatic conditions. At cold condition, total bacteria in AV was 1.5 to 1.9 log higher than (or 30 to 80 times) those of CC and EC. As ambient temperature increased, the concentration differences between AV and CC/EC became smaller. In the AV house, bacteria concentrations measured in the early morning were numerically lower than in the afternoon. This is because the measurements were taken before litter access in the early morning and after litter access in the afternoon. The CC (P = 0.995) and EC (P = 1.000) houses had similar bacteria concentrations in the morning and afternoon. Table 2 shows results of statistical analysis for airborne total bacteria concentration and factors using GLMMIX procedure in SAS. It can be seen that all factors and interaction of concern affected the indoor airborne total bacterial concentrations. In the past few decades, there has been increased attention to small particulate matters (e.g., PM10 ) for their adverse occupational health and environment impact caused by their microbiological and chemical components (Cambra-Lopez et al., 2010). The continuous measurement of PM10 concentrations in the CSES environmental monitoring allowed us to delineate the relationship between airborne bacteria and this PM fraction in the three laying-hen houses. It is shown that higher airborne total bacteria concentrations generally coincided with higher PM10 concentrations in all three
6
ZHAO ET AL.
houses (Figure 3), indicating that PM is the carrier of airborne microorganisms. The difference in regression equations among the houses shown in Figure 3 is a result of heterogeneity in dust size distribution and culturable bacteria count per unit weight of dust particles attached. While Figure 3 reveals the positive relationship between airborne total bacteria and PM10 concentrations, cautions should be taken when interpreting the results as bacteria measured in this study are those carried by total dust (all particles >0.3 μm) but the PM10 fraction contains particles equal to and smaller than 10 μm. Use of logarithmic regression in Figure 3 is based on the assumptions of constant PM10 mass fraction to total dust and uniform bacterial distribution in the dust, i.e., total bacteria count in dust is assumed to be proportional to the dust mass/volume.
Emission Rate of Airborne Total Bacteria The overall means and SD of airborne total bacteria emission rates in log CFU/[h-hen] (or log CFU/[h-AU]) were 4.8 ± 0.4 (or 7.3 ± 0.4) for CC, 6.1 ± 0.7 (or 8.6 ± 0.7) for AV, and 4.8 ± 0.5 (or 7.3 ± 0.5) for EC. The emission rates found in CC and EC were in line with the data (7.2 log CFU/[h-AU]) reported by Seedorf et al. (1998) for laying-hen house in Europe, but
Table 3. Summary of statistical analysis results for airborne total bacteria emission and factors. Effect House Climate House × climate Moment
Num DF
Den DF
F value
P-value
2 2 4 1
3 38 38 38
37.17 0.51 6.61 20.16
< 0.0001 0.6043 0.0004 < 0.0001
Note: ‘Num DF’ is the degrees of freedom of the numerator; ‘Den DF’ is the degrees of freedom of the denominator.
they were much lower than that for AV (P < 0.001). The house × climate interaction and sampling moment affected emission rates as well (Table 3). Figure 4 shows airborne total bacteria emission rates in the three hen houses measured in the morning and afternoon at the cold, mild, and warm climatic conditions. At mild (P = 0.012) and warm (P = 0.047) climatic conditions, the emission rates of AV were higher in the afternoon than those measured in the morning, which is again due to dust generation by hen activities on the litter floor. Emission rates in AM and PM were not different in EC under any climatic conditions (P = 0.996 for cold, P = 0.985 for mild, and P = 1.000 for warm), nor in CC (P = 1.000 for cold, P = 0.280 for mild, and P = 1.000 for warm). Emission rates of airborne total bacteria were positively related with the PM10 emission rates in all
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Figure 3. Relationship between total bacteria and PM10 concentrations (both measured at air exhaust) in conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses, and for pooled data (data from all three houses).
AIRBORNE BACTERIA OF THREE LAYING-HEN HOUSING TYPES
7
Concentration of Airborne Gram− Bacteria
houses, as shown in Figure 5. It should be emphasized that emission rates obtained in this study were based on measurements taken in daytime only. The emission rates were not monitored at night when bacterial concentration and VR were expected to differ from those during daytime as a result of much lower bird activities and lower ambient temperature at night.
Figure 5. Relationship between total bacteria and PM10 emission rates in conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses, and for pooled data (data from all three houses).
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Figure 4. Airborne total bacteria emission rates of the conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses at two sampling moments (morning or ‘AM’, and afternoon or ‘PM’) and three climatic conditions (cold, mild, and warm). The error bars are 95% confidence interval (CI95); means with non-overlapped error bars are significantly different (P < 0.050).
Results of airborne Gram− bacteria measurements are shown in Table 4. Out of 32 total air samples collected in each house, 2 samples (6%) in CC, 7 (22%) samples in AV, and 2 (6%) samples in EC were found to be positive for Gram− bacteria (Table 4). The statistical analysis shows that presence of Gram− bacteria was not affected by house (P = 0.312), climate (P = 0.968), or sampling moment (P = 0.804). Most of these positive samples were confirmed using enrichment method, indicating that the airborne Gram− bacteria concentrations were in the range of 1.7 (detection limit of enrichment test) to 2.7 (detection limit of plating culture) log CFU/m3 . The only positive sample confirmed using plating culture was found on 5/10/13 in the AV house at a concentration of 3.5 log CFU/m3 . European and Australian studies reported a wide range of Gram− bacteria concentration from 1 to 4.7 log CFU/m3 in laying-hen houses. Those studies also found that Gram− bacteria accounted for 0.6 to 15% of the total number of culturable bacteria (Seedorf et al., 1998; Zucker et al., 2000; Agranovski et al., 2007). The result of our study showed the Gram− to total bacteria ratio of <2%. Considering the possible low presence of Gram− bacteria, increasing sampling duration and/or using a larger aliquot for
8
ZHAO ET AL. Table 4. Airborne Gram− bacteria concentrations in the exhaust and middle areas of the conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses. Gram− bacteria concentration (log CFU/m3 ) AV
CC Exhaust
Middle
Exhaust
EC
Middle
Exhaust
Middle
Date (m/d/y)
AM
PM
AM
PM
AM
PM
AM
PM
AM
PM
AM
PM
1/9/13 2/16/13 3/16/13 4/11/13 5/10/13 6/7/13 7/4/13 8/9/13
ND ND ND ND ND ND ND ND
ND + ND ND ND ND ND ND
ND ND ND ND ND ND ND +
ND ND ND ND ND ND ND ND
ND ND ND ND ND ND ND +
ND ND + + ND ND ND ND
ND ND + ND ND ND ND ND
ND ND + + 3.5 ND ND ND
ND ND ND ND ND ND ND ND
ND ND ND ND ND ND ND ND
ND ND ND ND ND ND ND +
ND ND ND ND ND ND + ND
Table 5. Bacteria concentrations in the manure and litter of the conventional cage (CC), aviary (AV), and enriched colony (EC) laying-hen houses. Total bacteria
Bacteria concentration (mean ± SD, log CFU/g) Gram− bacteria
Manure on belt
As-is Dry-base
Litter
Manure on belt
CC
AV
EC
AV
CC
AV
9.5 ± 0.9 10.0 ± 0.9
9.2 ± 0.6 9.7 ± 0.6
8.8 ± 0.7 9.3 ± 0.6
9.2 ± 0.8 9.3 ± 0.8
7.9 ± 0.6 8.4 ± 0.6a a
Litter EC
7.3 ± 0.2 7.7 ± 0.2a a
AV
7.4 ± 0.8 7.9 ± 0.8a a
4.5 ± 0.6b 4.5 ± 0.6b
Note: ‘As-is’ concentration is the concentration calculated based on weight of as-is manure; ‘Dry base’ concentration is the concentration calculated based on weight of manure dry matter. Gram− bacteria concentrations in the same type of manure (as-is or dry-base) with different lower-case superscripts are significantly different (P < 0.001). Each datum is a mean of six values measured once every month from March to August 2013.
plating culture to get a countable bacteria number are recommended.
Bacteria Concentrations of Manure and Floor Litter As an important source of airborne microorganisms, manure and litter were analyzed for the fecal bacteria concentrations in this study. The total bacteria concentrations in the hen houses were 8.8 to 9.2 log CFU/g for as-is manure and litter, and 9.3 to 10.0 log CFU/g for dry-base manure and litter (Table 5). Total bacteria concentrations of manure and floor litter were similar. The Gram− bacteria in manure (7 to 9 log CFU/g) were more abundant than in the floor litter (4.5 ± 0.6 log CFU/g) (P < 0.001). This result can be explained by the fact that Gram− bacteria are more vulnerable to dehydration stress than Gram+ bacteria. The moisture contents of manure on the belts were 69.4 ± 5.6% in CC, 66.2 ± 5.3% in AV, 67.3 ± 2.3% in EC; whereas moisture content of the litter in AV was only 15.7 ± 2.6%. The fraction of Gram− bacteria (of total bacteria) in manure was much greater than that in airborne particles (P < 0.001). The difference in these fractions indicates that (wet) manure on belt is not the major source of bacterial particulates in laying-hen houses. The high moisture content of manure prevents fecal particles from becoming airborne. In contrast, the much
drier floor litter could be easily agitated by animal activity and eventually contribute to the higher airborne bacteria concentrations.
CONCLUSIONS This study investigated the concentrations and emission rates of airborne bacteria of three types of layinghen houses, i.e., conventional cage, enriched colony, and aviary. The monitoring was performed for an 8-month period covering mid to end of the production flock. The relationship between airborne total bacteria and PM10 is delineated. The following observations and conclusions were made. r Concentrations and emission rates of airborne to-
tal bacteria in the laying-hen houses were affected by housing type and climatic condition, and were variable within the day. r The AV house had much higher airborne total bacteria concentrations and emission rates as compared to the CC and EC houses. r Higher ventilation rate in warmer weather helped diluting the airborne bacteria concentration in the AV houses, although the dilution effect was not as obvious in CC and EC houses. r In the AV house, the total bacteria concentration and emission rates tended to be higher in the afternoon than in the morning due to more bacterial
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Note: ‘ND’ is not detectable; ‘+’ means a sample is positive in the enrichment test, but negative in plating culture.
AIRBORNE BACTERIA OF THREE LAYING-HEN HOUSING TYPES
particles generated by hen activities on the floor litter. r Total bacteria were positively correlated with PM 10 in terms of concentration and emission rate. r Airborne Gram− bacteria were present in the laying-hen houses at low concentrations. The fraction of Gram− bacteria (of the total bacteria) was estimated to be less than 2%. r Manure on the belt and floor litter had similar total bacteria concentrations; however, litter contained much lower levels of Gram− bacteria.
ACKNOWLEDGMENTS
REFERENCES Agranovski, V., T. Reponen, and Z. D. Ristovski. 2007. Survey of bioaerosol emissions from Australian poultry buildings. Proc. European Aerosol Conference 2007, Salzburg, Austria. Cambra-Lopez, M., A. J. A. Aarnink, Y. Zhao, S. Calvet, and A. G. Torres. 2010. Airborne particulate matter from livestock production systems: A review of an air pollution problem. Environ. Pollut. 158:1–17. de Reu, K., W. Messens, M. Heyndrickx, T. B. Rodenburg, M. Uyttendaele, and L. Herman. 2008. Bacterial contamination of table eggs and the influence of housing systems. World’s Poult. Sci. J. 64:5–19. Duchaine, C., Y. Grimard, and Y. Cormier. 2000. Influence of building maintenance, environmental factors, and seasons on airborne contaminants of swine confinement buildings. Am. Ind. Hyg. Assoc. J. 61:56–63. Hu, D., L. Wang-Li, O. D. Simmons, III, J. J. Classen, and J. A. Osborne. 2015. Spatiotemporal Variations of Bioaerosols in the Vicinity of an Animal Feeding Operation Facility in the US. J. Environ. Prot. 6:614–627. Lonc, E., and K. Plewa. 2011. Comparison of Indoor and Outdoor Bioaerosols in Poultry Farming in Advanced Topics in Environ-
mental Health and Air Pollution Case Studies. A. Moldoveanu, ed. InTech. ˇ ˇ c, D. Maruˇsi´c, B. Matkovi´c, K., M. Vuˇcemilo, I. Stokovi´ c, R. Simi´ Vinkovi´c, and S. Matkovi´c. 2013. Concentrations of airborne bacteria and fungi in a livestock building with caged laying hens. Veterinarski Arhiv. 83:413–424. Muller, W., and P. Wieser. 1987. Dust and microbial emissions from animal production. Pages 47–89 in Animal production and environmental health. D. Strauch, ed. Elsevier. Oppliger, A., N. Charri`ere, P. Droz, and T. Rinsoz. 2008. Exposure to bioaerosols in poultry houses at different stages of fattening; use of real-time PCR for airborne bacterial quantification. Ann. Occup. Hyg. 52:405–412. OSHA. 1999. OSHA technical manual (OTM) section III: Chapter 2 Indoor Air Quality Investigation (Directive Number TED 01-00-015). Occupational Safety & Health Administration. Rylander, R., and M. F. Carvalheiro. 2006. Airways inflammation among workers in poultry houses. Int. Arch. Occ. Env. Hea. 79:487. Seedorf, J., J. Hartung, M. Schroder, K. H. Linkert, V. R. Phillips, M. R. Holden, R. W. Sneath, J. L. Short, R. P. White, S. Pedersen, H. Takai, J. O. Johnsen, J. H. M. Metz, P. W. G. Groot Koerkamp, G. H. Uenk, and C. M. Wathes. 1998. Concentrations and emissions of airborne endotoxins and microorganisms in livestock buildings in Northern Europe. J. Agric. Eng. Res. 70:97–109. Shepherd, T. A., Y. Zhao, H. Li, J. P. Stinn, M. D. Hayes, and H. Xin. 2015. Environmental assessment of three egg production systems—Part II. Ammonia, greenhouse gas, and particulate matter emissions. Poult. Sci. 94:534–543. Thelin, A., O. Tegler, and R. Rylander. 1984. Lung reactions during poultry handling related to dust and bacterial endotoxin levels. Eur. J. Respir. Dis. 65:266–271. ˇ Vuˇcemilo, M., B. Vinkovi´c, K. Matkovi´c, I. Stokovi´ c, S. Jakˇsi´c, - . Stubiˇcan. 2010. The influS. Radovi´c, K. Grani´c, and D ence of housing systems on the air quality and bacterial eggshell contamination of table eggs. Czech J. Anim. Sci. 55:243–249. Wang-Li, L., Q. Li, and G. E. Byfield. 2013. Identification of bioaerosols released from an egg production facility in the southeast United States. Environ. Eng. Sci. 30:2–10. Zhao, Y., A. J. A. Aarnink, M. C. M. de Jong, and P. W. G. Groot Koerkamp. 2014. Airborne microorganisms from livestock production systems and their relation to dust. Crit. Rev. Env. Sci. Tec. 44:1071–1128. Zhao, Y., T. A. Shepherd, J. Swanson, J. A. Mench, D. M. Karcher, and H. Xin. 2015a. Comparative evaluation of three egg production systems: Housing characteristics and management practices. Poult. Sci. 94:475–484. Zhao, Y., T. A. Shepherd, H. Li, and H. Xin. 2015b. Environmental assessment of three egg production systems–Part I: Monitoring system and indoor air quality. Poult. Sci. 94:518–533. Zucker, B. A., S. Trojan, and W. Muller. 2000. Airborne Gramnegative bacterial flora in animal houses. J. Vet. Med. B. 47:37–46.
Downloaded from http://ps.oxfordjournals.org/ at University of California, Santa Barbara on May 15, 2016
Cash funding for the study was provided by the Coalition for Sustainable Egg Supply (CSES). In-kind contributions by Iowa State University and the Egg Industry Center (located at ISU) were provided by availing the state-of-the-art environmental monitoring equipment to the project. We sincerely appreciate the cooperation and assistance of the egg producer in the implementation of this field study. We also thank Lendie Follett of Statistics Department at ISU for her assistance in the statistical analysis.
9