*2.5. Data Analysis*

Dry-bulb (Tdb), dew point temperatures (Tdp) and relative humidity data were downloaded from the data loggers after each monitoring campaign. Ambient T and humidity corresponding to the monitored duration of each trip were obtained from weather data downloaded from nearby weather stations on the days that monitored trips took place.

*Animals* **2019**, *9*, 31

Apparent equivalent temperature (AET) were calculated from the dry bulb temperature and the relative humidity [17]. Specifically,

$$\text{AET} = \text{T} + \frac{10^{(30.5905 - 8.2 \times \log\_{10}(\text{K}) - \frac{342.31}{\text{K}})} \times \left(\frac{\text{RH}}{100}\right)}{0.93 \times (0.0006363601 \times \text{K} + 0.472)}\tag{1}$$

where T = recorded air temperature, ◦C, K = T corrected to Kelvin (◦C + 273.15), RH = recorded relative humidity, %.

Except for mean comparison of recorded temperature, elevated temperature above ambient values (i.e., Δt) were calculated by subtracting mean ambient temperature of each transit from the recorded temperature at each measurement location. For each managemen<sup>t</sup> configuration, elevated temperature at three longitudinal planes (width, i.e., driver, midline and curb), at three horizontal planes (height, i.e., top, middle and bottom), and at five cross-section planes (Figure 3) were analyzed by ANOVA with means separated by Tukey's range test [18].

Mean comparison of elevated temperature of different transit segments at three longitudinal and three horizontal planes of plastic wrap and double board were made [18]. The first segmen<sup>t</sup> was chosen as 15-min, due to an observed fast change immediately after the trailers departed from the farms, especially in summer months. Other segments were 30-min long. Only data from four segments were retained for this analysis. Differences of segment-average temperature were analyzed within groupings for each plane along the width and height axes [18], and considered significant if *p* < 0.05. Additionally, mean comparisons of elevated temperature above ambient ( Δt) at different planes of the two paired trailers with either plastic wrapped or double board were made using paired t-test to determine the effect of plastic wrap on the double board winter trips.

Relative humidity values represent how close air is to saturation at the measured temperature. Due to its temperature dependence, it is invalid to compute averages of recorded relative humidity over several hours directly. A "representative" relative humidity variable (RH\*) was derived from a time-averaged humidity ratio (also called absolute humidity) and a corresponding time-averaged temperature from the same logging interval [19]. It serves as an "averaging" variable of relative humidity using appropriate psychometric manipulation.

Specifically, for each data logger, a humidity ratio (W) was computed at a specific time using its corresponding T and RH values [20]. After calculating the humidity ratio for a x time interval, a time-averaged humidity ratio ( W) and a time-average temperature ( T) were calculated for this duration. Before the representative RH\* can be calculated, the partial pressure of water vapor in moist air (p w) was calculated using:

$$\mathbf{p}\_{\rm w} = \frac{\overline{\mathbf{W}} \times \mathbf{p}}{\left(\overline{\mathbf{W}} + 0.62198\right)}\tag{2}$$

where p w = partial pressure of water vapor in moist air (Pa), W = time-averaged humidity ratio, p = total pressure of moist air assumed equal to atmospheric pressure (Pa).

The representative relative humidity was determined by:

$$\text{RH}^\* = \frac{\text{P}\_{\text{W}}}{\text{P}\_{\text{WS}}} \times 100 \tag{3}$$

where RH\* = representative relative humidity (%), pws = saturation pressure of moist air (Pa).

## **3. Results and Discussion**

Figure 5 illustrates examples of the temperature, RH and air speed profiles of interior and exterior of a summer live-haul trip. The ambient temperature at the start of this transport was 30.6 ◦C (Figure 5a). The first arrow indicates the beginning of loading, the second the beginning of transport, and the third the arrival at the processing plant for holding period. Cooling, including convective fans

and water sprays, was applied to the trailer during loading, resulting in a sharp temperature drop and relative humidity increase. Air speeds during the 40-min transport were variable (Figure 5b), likely in part determined by the speed of the vehicle. Air speeds at interior and exterior locations averaged 0.5 and 1.9 m s<sup>−</sup>1, respectively. Webster et al. [5] reported mean air movement of open trucks in motion of 3.3 m s<sup>−</sup><sup>1</sup> (range 0.0 to 8.9 m s<sup>−</sup>1) of commercial broiler transporters in England. Weeks et al. [10] calculated that air speeds in moving vehicles varied from 0.9 to 2.4 m s<sup>−</sup><sup>1</sup> with maxima of 6.0 m s<sup>−</sup>1.

**Figure 5.** Temperature, relative humidity (**a**) and air velocity (**b**) profiles from interior and exterior logger positions of a summer trip. The first arrow on each graph indicates the beginning of loading, the second the beginning of transport, and the third the beginning of holding period.

The representative relative humidity of 28 trips and their trailer average temperatures are plotted in Figure 6. Using a physiological stress response model, Mitchell et al. [17] identified "safe", "alert" and "danger" thermal zones, defined by AET values based on temperature-humidity combinations (Figure 1). AETs of 65 ◦C or greater were deemed dangerous due to potential severe physiological stress [17]. AETs of the monitored summer trips averaged 80.5 ◦C, indicating possible dangerous thermal conditions. Note that the laboratory experiments used to collect physiological parameters for derivation of the AET index were three hours in length with no air speed reported [17]. Majority of transport trips in this study area were less than two hours, with a median of one hour. Air speed on the moving trailers (Figure 5b) may have allowed convective cooling, although this was not uniformly experienced by all chickens on board. Additionally, partial surface wetting of broilers by hand sprayers may have alleviated or delayed the onset of heat stress of cooling-assisted transport trips based on literature reports [21,22].

The severity of physiological stress in the summer is unknown due to unavailable mortality data, which could have allowed correlation analysis with the thermal conditions. Future research should focus on improved research protocols, such as mortality data collection, measurements of core body temperature of broilers under various micro-environments, and behavior monitoring with video cameras. Nevertheless, it is important for commercial companies in South Central USA to improve the efficiency of the catching, loading and transporting process, and to minimize the duration of exposure of live chickens to uncontrolled environments in the summer. Additionally, better measures, such as stocking density adjustments, route optimization to avoid heavy traffic, and adding on-board sprinkler systems to the modules, should be considered for trips with longer distances to mitigate heat stress conditions.

**Figure 6.** Average representative relative humidity vs. average trailer temperature of 28 monitored trips moving broilers to processing plants. The AET values corresponding to 40 and 65 ◦C [17] were overlaid.

Tables 3 and 4 show the mean temperature and representative relative humidity that birds were exposed to for every 15 or 30-min transit duration of each trailer configuration. Under winter conditions, double boards and plastic wraps allowed heat produced by the broilers to be retained within the transporters, resulting in a mean trailer temperature of 5.5 ◦C (Table 3). While recorded trailer temperatures were higher than ambient T in winter (Table 3), they remained below the thermoneutral range [10], in spite of the boarding procedures to ameliorate the cold temperatures. Knezacek et al. [6] reported a 1 ◦C rectal temperature reduction of broilers that were exposed to 3.9 ◦C crate temperature during a 178-min winter trip of −28.2 ◦C ambient temperature. In a simulated 3-h wind tunnel experiment with chamber temperatures ranging between −4 and 12 ◦C, 1 ◦C rectal temperature reduction from broilers were reported under all dry chamber environments [23]. However, moderate to severe hypothermia (3 to 14 ◦C rectal temperature reduction) were observed when wetting was imposed in an increasingly colder chamber environment [23].

**Table 3.** Mean recorded temperature (◦C) of 15 or 30-min segments during transit under five trailer configurations found in commercial broiler transporters.


**Table 4.** Mean representative relative humidity (RH\*, %) every 15 or 30-min during transit under five trailer configurations found in commercial broiler transporters.


During the warm season with cooling assistance, overall trailer temperatures were within a narrow range of ambient conditions (up to 36.1 ◦C) for the duration of the journey (Table 3), with mean representative relative humidity less than 80% (Table 4). Both the air speed on trailers (Figure 5b) and the water retained by birds and trailer modules during on-farm loading could have prevented trailer air temperatures from rising far beyond ambient temperatures. Panting was observed from birds in transit based on camera footages from one selected summer trip, indicating that the efficacy of cooling assistance was limited due to a simultaneous increase of humidity level.

Temperature throughout transporters with open and single side boards were mostly within the thermoneutral zone for broiler chickens (Table 3). Open transporters operating in the mild seasons provided a reasonably comfortable thermal environment. It is important to note that the greatest number of broilers are transported using the open sided configuration.

Humidity plays an important role in heat and mass exchanges in the livestock and poultry environment [4,15]. For example, moist air can compromise feather insulation properties, placing broilers at risk of cold stress [8,23]. Hunter et al. [23] concluded that broiler chickens could be safely transported at crate temperatures as low as −4 ◦C, if they are dry, or experience moderate hypothermia at temperatures as high as 8 ◦C when wet.

Representative relative humidity was selected to express the extent of moisture saturation in the modules. When moist air comes into contact with cooler surfaces (i.e., the modules and interior surface of wrapping plastics), condensation forms. Burlinguette et al. [8] used a threshold of 80% relative humidity value to determine susceptibility to condensation. In our study, mean representative relative humidity of four winter boarded transport was around 80% (Figure 6), indicating that a small amount of air exchange existed. Movement-induced ventilation prevented the excessive accumulation of moisture produced by the birds within the trailers (Table 4).

#### *3.1. Spatial Uniformity of Air Temperature on Trailer*

The industry practice of installing fiberglass boards on the modules is intended to reduce ventilation and conserve heat produced by the broilers in cool seasons. The practice seemed to be effective, elevating mean air temperature above their corresponding ambient temperature (with ranges of −15.8 to 2.8 ◦C and −16.4 to 8.9 ◦C, Table 2) by 10.7 ◦C and 9.3 ◦C for the plastic wrap and double board, respectively (Table 5). In comparison, three levels of curtains, and closure of roof vents used by Canadian transporters, resulted in an average temperature elevation of 14.4, 12.7 and 11.2 ◦C above ambient, as reported by Burlinguette et al. [8].

Differences in elevated temperature above ambient between locations were analyzed for each configuration. In winter, elevated temperatures at three longitudinal planes along the Width axis (Table 5) were different (*p* < 0.05) for all three boarding configurations. Mean elevated T at Midline were warmer than those at the outward-facing planes of the trailers (*p* < 0.05) when side boards were used, likely a result of lower airflow in the central locations. Top modules recorded mean T elevations of 8 ◦C from the ambient, which were several degrees lower than those gained by the middle or bottom modules, likely due to lack of protection from motion-induced ventilation. The lowest observed T was −1.1 ◦C at top, curb-side module, while the highest T, 18 ◦C, occurred on the midline, bottom module on double board trailers. This was similar to earlier report of highly variable and extreme thermal conditions when side curtains and most roof vents were closed on a Canadian transporter [8]. Large temperature gradients with up to 20 ◦C difference of crate temperatures (i.e., 3 to 26 ◦C) in an ambient temperature of −28 ◦C were reported when only the fourth roof vent was opened in a 178-min trip on a Saskatchewan transport trailer [6]. Kettlewell et al. [1] reported airflow movement from the back to the front of the trailers based on the temperature trends observed throughout trailers in the UK. However, the airflow distribution of the trailers in this study is unknown due to many undefined small openings at the back of each module (opposite to the door) (Figure 3b) and around the fiberglass boards.


**Table 5.** Spatial variation of air temperatures elevation above-ambient (Δt, ◦C) across the trailer during transport in different seasons.

a,b,c Superscripts denote differences (*p* < 0.05) within each column and axis, 1 Code 1, 2, 3, 4 and 5 denote instrumented cross sections from the front to the back of trailers.

Transporters using cooling assistance displayed slightly different temperature profiles than those during cold or mild seasons. Temperatures tend to be higher (*p* < 0.05) at the midline (Table 5), although the difference was small (1.0 ◦C). The top tier displayed higher temperature elevations (*p* < 0.05), likely due to exposure of sheet metal roof to direct sunlight in summer.

#### *3.2. Effect of Journey Length on Thermal micro-Environment*

 8.0 c

75–105 min

Although journey lengths in this study were shorter (less than 2 h) compared to those reported elsewhere [1,6,10,19], elevated temperatures still differed from the beginning to the end (*p* < 0.05) (Table 6) for trips using plastic wrap. Trailer T elevation decreased significantly during transit at various locations in plastic wrap (up to 4.1 ◦C). A similar decline of elevated temperature above the ambient were observed in double board trailers without wrap (Table 7).


**Table 6.** The effect of trip duration on elevated temperature above ambient (Δt, ◦C) at measured locations across width and height of the trailer with plastic wrap (*n* = 5).

8.6 a,b,c,d Superscripts denote differences (*p* < 0.05) within each column and axis.

d 6.6 c 10.9 c 8.7 b

9.7 c

**Table 7.** Effect of trip duration on elevated temperature above ambient (Δt, ◦C) at measured locations across the width and height of the trailer with double boards (*n* = 4).


a,b,c,d Superscripts denote differences (*p* < 0.05) within each column and axis.

When cooling assistance was used during loading in summer, temperatures increased from the first 15 min to the following 30 min across the trailer (*p* < 0.05) (Table 8). Air temperatures inside the trailers were lower than the corresponding outdoor conditions during the first 15 min immediately after trailers departed from farms. This was clearly the residual effect of liquid water retained on transporters from loading on farms. Fans and various types of water treatments, including misters and hand-held sprayers, were used in all eight trips monitored and reported in this category. Water retained by modules and birds' feathers continued evaporating as transporters traveled on the roads. Ritz et al. [13] also reported that the use of multiple high-velocity fans positioned parallel to the live-haul trailers during loading was effective at cooling birds prior to transport. However, water likely diminished around 15 min after transporters' departure, allowing temperature rises of 2 to 3 ◦C at various locations after one hour (*p* < 0.05, Table 8). For hot weather conditions, even with 1 to 2 ◦C temperature rises within the trailer, thermal load could shift to a more dangerous level.

**Table 8.** Effect of trip duration on elevated temperature above ambient (Δt, ◦C) at measured locations across width and height of the trailer when cooling assistance was used (*n* = 8). Cooling assistance consisted of propeller fan(s) blowing air and misters or hand-held pressure washers applying water toward trailers being loaded.


a,b,c Superscripts denote differences (*p* < 0.05) within each column of width and height axis.

#### *3.3. Effect of Plastic Wrappping on the Micro-Environment*

Plastic wrap, in addition to the double boarded trailers, raised the mean air temperature by around 3.2 ◦C compared to double boarded trailers on winter nights with average ambient temperatures of −5 and −17 ◦C (Table 9). Average representative relative humidities of double board and plastic wrapped trailers were 72% and 79%, respectively (Figure 6). Plastic wrapping was only used to further reduce wind penetration through modules in order to protect birds from extremely cold weather conditions when birds with incomplete feather coverage (with 1.7 kg live weight) were transported. This practice seemed to moderately retain heat and water vapor inside the modules without creating moisture saturation. Better protection, such as more insulation, might be needed in order to alleviate cold stress without any risk of creating saturated air conditions.

**Table 9.** Means and standard errors of elevated temperatures above ambient (Δt, ◦C and the paired sample t-test at various locations of paired trailers with either plastic wrap or double board on two winter nights with average ambient temperatures of −5 and −17 ◦C, respectively.


a,b Superscripts denote differences (*p* < 0.05) within each column.
