Behavior, blood stress indicators, skin lesions, and meat quality in pigs transported to slaughter at different loading densities

Abstract

A total of 1,936 pigs were transported through 8 trips (8.4 ± 0.5 hr) from 2 grow-to-finish farms to a commercial slaughter plant, all located in Southern Brazil. On each trip, a sub-sample of each load (36 barrows/load, weighing 118.9 ± 9.8 kg) was randomly allocated into one of the following loading densities during transport: 200 kg/m² (D200), 235 kg/m² (D235), and 270 kg/m² (D270). Behavioral recordings of postures and activities were made during transportation and lairage using video-cameras. At slaughter, blood samples were collected to assess the concentrations of creatine kinase (CK) and lactate. Carcass weights and skin lesion scores were assessed on-line, and meat quality was evaluated in the longissimus thoracis (LT) muscle by assessing pH, color, and drip loss. During transportation, the proportion of animals lying down was higher (P < 0.05) in D200 and D235 groups compared with D270. The proportion of sitting animals during transportation was higher (P < 0.01) in D270 compared with D200. In lairage, D200 and D235 pigs stood more compared with D270 (P = 0.01), while the proportion of lying pigs was higher (P = 0.02) for D270 pigs compared with D200 and D235. The frequency of drinking bouts in lairage was higher (P < 0.05) for D200 group compared with D235 and D270. The levels of CK were lower (P < 0.05) in D200 pigs transported compared with D270. Lesion scores tended to be higher (P = 0.06) in D270 carcasses compared with D200 and D235. A tendency for lower (P = 0.10) pH1 values in the LT muscle of D270 pigs compared with D200 pigs was also found. Based on the results of this study, the application of lower loading densities (≤235 kg/m²) in the truck allows pigs to have sufficient space to rest, travel more comfortably and arrive less fatigued at the slaughter plant.

Introduction

Transportation is a multicomponent phenomenon consisting of effects of multiple factors, such as handling, vehicle design, animal position in the truck compartment, environmental conditions, journey duration, loading density, and driving conditions, which greatly impact the welfare of pigs (Faucitano and Goumon, 2018). Among these factors, loading density can account for the largest portion of the variation in total animal losses, i.e., dead, fatigued, and injured pigs compared with other variables, such as driver, handling crew, ambient quality, wind speed, loading duration, and waiting time before unloading (Fitzgerald et al., 2009).

Recommendations for optimal loading density or space allowance, expressed as kg/m² and m²/100 kg, respectively, ensuring a comfortable travel for pigs, vary between countries and according to climate conditions and market weight (Rioja-Lang et al., 2019). Loading density higher than 235 kg/m², which is the maximum legal allowance in the EU (Council Regulation EC 1/2005), has been associated with lower body temperature, heart rate, blood creatine kinase (CK) levels at slaughter, skin lesions and rectal prolapses, eventually resulting in decreased proportion of dead and nonambulatory pigs at unloading at the plant, and DFD pork meat (Faucitano and Lambooij, 2019). These effects most likely result from physical stress or fatigue (Barton-Gade and Christensen, 1998; Warriss et al., 1998) as under conditions of limited floor space not all pigs can rest as they are either forced to continually change their position in the compartment seeking a place to rest or are not even able to lie down (Lambooy et al., 1985). Based on this evidence, the EU loading density is also recommended for the transport of pigs to slaughter in Brazil (Ludtke et al., 2016). However, results arising from recent studies using planimetric measures of the floor surface covered by pigs in different positions during transport showed that a pig of modern genetics, being larger and heavier, may need a minimal greater space (~0.49 m²) to entirely lie down in fully lateral recumbent position on the truck floor (Arndt et al., 2019), which suggests a revision of the current EU recommendation for space allowance during transport (minimum of 0.425 m²; Council Regulation EC 1/2005).

However, besides providing more floor space to lie down, lower loading density (<235 kg/m²) also provides pigs with greater freedom to move around in the compartment during transport, which can cause muscle fatigue due to pigs’ attempt to maintain their balance in response to unexpected and sudden movements of the truck, or fighting (Barton-Gade and Christensen, 1998). This physical effort may cause excessive muscle glycogen depletion and a greater risk of DFD pork meat production (Guàrdia et al., 2005). Increased risk for skin lesions and pale, soft, and exudative (PSE) pork meat production has also been reported when more floor space (0.50 vs. 0.25 m²/100 kg) is provided in summer (vs. winter) and shorter transports (1 vs. 3 hr), respectively (Guàrdia et al., 2005, 2009).

The conflicting results, especially related to the physiological condition of pigs at slaughter, skin lesions, and pork meat quality, between studies may be explained by the effects of loading density on post-transport behavior during rest in the lairage pen. To our knowledge, such evaluation was never done in any similar study before.

The objective of this study was to evaluate the effects of 3 loading densities (200, 235, and 270 kg/m²) during transport to slaughter on pigs’ behavior in the truck and lairage, physiological conditions at slaughter, and carcass and meat quality.

Materials and Methods

All experimental procedures performed in this study were approved by the institutional animal care committee at the Faculty of Agricultural and Veterinary Sciences of the São Paulo State University (Jaboticabal, Brazil; Protocol n. 08569/19), based on the current Brazilian legislation (Brasil, 2009).

Animals and treatments

A total of 1,936 crossbred pigs of mixed sex and same genetics (synthetic cross) were transported from 2 commercial finishing farms to a commercial slaughterhouse, all located in Southern Brazil, through 8 shipments or replicates (8.4 ± 0.52 hr trips), in terms of 2 shipments per week. Pigs were transported using the same passively ventilated triple-flat-deck trailer (Carrozzeria Pezzaioli, Montichiari, Italy), featuring 2 moving decks (middle and top), 18 compartments (6/deck; 2.6 × 2.4 × 0.9 m each), drinkers (one/compartment) and water misters evenly mounted inside the trailer wall compartments. Shipments were balanced by farm of origin (i.e., 4 loads of 242 pigs per farm, totaling 968 pigs per farm). Farms had the same standard feeding program and handling system and similar loading facilities design, and were at the same distance from the abattoir (477 km).

Within each load, the day of transport 36 sentinel barrows were selected by weight (118.9 ± 9.8 kg) in order to have a uniform weight of pigs in each treatment group and distributed into 3 different loading densities in the trailer: 200 kg/m² (D200; n = 10 pigs), 235 kg/m² (D235; n = 12 pigs) and 270 kg/m² (D270; n = 14). The lowest density was arbitrarily decided to provide pigs with the maximum space to ensure their physical and thermal comfort during transport. The intermediate loading density corresponded to that recommended by the European Union (Council Directive 95/29/EC), while the highest density represented the commercial practice in Brazil.

On the day of transport, at the farm sentinel barrows were sorted-out from the same finishing pen (containing 12 to 16 pigs) and driven to shipping pens near the farm exit, where they were grouped and kept separated by treatment (neither mixing of unfamiliar pigs within treatment nor between treatments) until loading. Pigs were withdrawn of feed for 3 hr before loading, following the standard practice of the producer and loaded onto the trailer in small groups (5 pigs/group) using rattle paddles and boards. Sentinel pigs were allocated into 3 compartments (1 compartment/treatment), each representing a loading density, in the middle deck (Figure 1). In each shipment, loading densities were randomized across compartments in order to avoid the bias of the loading order and position in the truck on pigs’ response to handling and transport reported in previous studies (Correa et al., 2013, 2014; Goumon et al., 2013; Scheeren et al., 2014).

Figure 1.

Location of trailer test compartments and position of video cameras and weather trackers.

Before departure from the farm, pigs were water sprayed in the trailer for ~10 min. The same driver was used throughout the study and the journey was mostly on paved roads. On arrival at the slaughterhouse (between 2:00 and 4:00 a.m.), the truck and its load had to wait for the opening of the abattoir gates before unloading (average waiting time: 2.16 ± 0.84 hr). During this wait, as temperatures were above 20 °C, pigs were cooled-off through water misting as recommended to ensure comfortable thermal conditions for pigs located in a stationary passively ventilated truck (Fox et al., 2014). Groups were unloaded and driven to the lairage pen by compartment using compressed air prods and rattle paddles.

Given the different size between transport compartments and lairage pens, to standardize stocking density (0.6 m²/pig) in the lairage pen, a subsample of 10 animals from each compartment/treatment group (total of 30 pigs/load) was randomly selected at unloading and kept in 3 separate lairage pens (no mixing between treatments). During rest, pigs had free access to water.

After 3 hr of lairage, pigs were driven along the alley feeding the stunning chute using rattle paddles and boards and electric prods while lining-up in the stunning chute. Pigs were electrically stunned with an automatic electrical head-to-chest stunner (250 V on the head and 120 V on the chest), producing cardiac arrest, and exsanguinated in the prone position. At slaughter time, pigs were fasted for an average of 17.1 ± 0.6 hr.

External and within trailer environmental conditions

External ambient temperature and relative humidity (RH) data were collected from a weather station in the area of Toledo (Paraná, Brazil). Throughout the transport trials, the average ambient temperature was 22.5 °C (ranging from 15 to 33 °C), the average of RH was 63.3% (ranging from 20% to 94%) and the average of temperature-humidity index (THI) was 69.5 (ranging from 58.6 to 90.3).

Air temperature (T°) and RH in the trailer were monitored during transport using 2 data loggers (Kestrel Meters 400, Boothwyn, PA) installed on the gates of the test compartments (Figure 1). Both loggers were programmed to record T° and RH at 5-min intervals from loading to unloading. Data from each logger were downloaded using the Kestrel Communicator Interface Software – 4000 (Kestrel Instruments, Boothwyn, PA). From the T° and RH values, an average THI was calculated according to the formula: THI = (1.8 × T + 32) − [(0.55 − 0.0055 × RH) × ((1.8 × T − 26)] (Pereira et al., 2018), where T is in °C and RH in %.

Behavioral observations

During transport, behaviors were continuously recorded using 3 cameras (Intelbras VMH 1010 D HD 720p, Intelbras SA, São José, Brazil), each installed on the ceiling of each test compartment of the trailer (Figure 1). During transport, the number of pigs in each posture (lying, standing, sitting, or any other posture; Table 1) was recorded every 5-min in each selected compartment. As the compartment group was not always entirely visible by the camera, only recordings with at least 7 visible pigs in each group were used for the analysis.

Table 1.

Description of the behaviors evaluated during transport and in lairage

Behavior	Description
Transport
Lying	Pig is in a horizontal position, with the weight of its body supported in one of its shoulder (lateral position) or on the sternum/belly (sternal position).
Standing	Pig is in a vertical position, maintaining all legs straight, and supporting the weight of its body on the four hooves
Sitting	Pig maintain the front legs straight, while the hind legs are bent, keeping its hindquarters in contact with the floor
Other postures	Any other posture that is not the previous ones
Lairage
Drinking bouts	Pig puts its mouth around the drinker for any period of time
Agonistic acts	Pig bites or bangs another pig with the head
Latency to lie down	Period of time, measured in each pen, from the arrival of the pigs in lairage until at least 50% of them lie down (in min)

Behavior in lairage was recorded during the first hour of rest using a camera (Action GoCam Pro Sport Ultra 4k HD Sport 1080, Action GoCam, Guangdong, China) installed above the animals in each pen. The recording started as soon as the pen was filled with pigs and the gate was closed. A scan sampling (every 2 min) was used to record the percentage of pigs lying down, standing, sitting, or showing other postures (Table 1). The latency to lie down was calculated and the frequency of drinking bouts and agonistic behavior were recorded using continuous recording during 4 periods of 8 min (Table 1). A new drinking bout was recorded if the pig’s mouth was off of the drinker for at least 5 s (Fox et al., 2014). Behavioral observations from video recordings were performed by one trained observer.

Physiological measurements

Blood samples from a total subsample of 240 sentinel pigs (80 pigs/treatment) were collected at exsanguination in plastic cups (~150 mL). Blood was transferred into two 4 mL vacuum tubes (FirstLab, São José do Pinhais, Brazil), of which one contained fluoride and EDTA for lactate-level analysis, while the other contained a clot activator for serum extraction for CK level analysis. Tubes were transported in refrigerated containers to the laboratory where they were immediately centrifuged at 3,000 × g during 10 min at 18 °C. Once extracted, serum and plasma samples were frozen at −20 °C until analysis.

Blood lactate and CK concentrations were measured using analytical kits (Siemens Dimensions Flex Reagent Cartridge LA and CKI, Siemens Healthcare Diagnostic Inc., Newark, DE). Quantitative determination of blood lactate and CK concentrations was performed using the Siemens Dimension Xpand Plus Chemistry Analyzer (Siemens Healthcare Diagnostic Inc, Newark, DE) and the values expressed in mmol/L and UI/L for blood lactate and CK concentrations, respectively.

Carcass quality measurements

After slaughter, carcasses were eviscerated, split, and chilled according to standard commercial practices. Hot carcass weight (HCW) was recorded on the slaughter line, while cold carcass weight (CCW) was registered after overnight chilling.

Skin damages were scored on the left carcass side according to 5-point photographic scale (from 1 = none to 5 = severe; MLC, 1985). Bruises were also classified as handling-type bruises (1 = 1 bruise; 2 = 2 bruises; and 3 = more or equal than 3 bruises), fighting-type bruises (1 = less than 10 bruises; 2 = 10 to 20 bruises; and 3 = >20 bruises) or mounting-type bruises (1 = <5 bruises; 2 = 5 to 10 bruises; and 3 = >10 bruises) by visual assessment of lesion shape and size, according to photographic standards (ITP, 1996; Faucitano, 2001).

Meat quality measurements

Meat quality was assessed by measuring pH at 1 hr (pH1) after slaughter (in the cooler) in the left longissimus thoracis (LT) muscle (between the last and the second last rib) using a portable pH-meter fitted with an integrated temperature probe (Testo 205, Testo AG, Lenzkirch, Germany), which was calibrated before the start of measurements and recalibrated every 5 readings. At 24 hr postmortem, LT muscle chops (~5-cm thick) were taken and transported in refrigerated conditions to the laboratory for analysis of pH24, using the same pH-meter, drip loss, and color. Drip loss was assessed using the filter paper wetness (FPW) test (Kauffman et al., 1986). Briefly, a filter paper (# 589 Blue Ribbon, Whatman, International Co., Mont Royal, Canada) was placed on the LT muscle cut surface after 10-min of air exposure and weighed using an analytical scale (M124A, Bel Engineering srl, Monza, Italy) after 3 s of fluid accumulation on the paper. The percentage of drip loss was calculated using the equation: % Drip loss = −0.1 + (0.06 × FPW) (Rocha et al., 2016). Instrumental color (L*, a*, b*) was measured with a Minolta Chromameter (CR-10, Konica Minolta, Inc., Osaka, Japan) equipped with a 8-mm aperture, 10° viewing angle, and D65 illuminant after exposing the muscle surface to 15-min blooming time.

Loins were classified into 5 pork quality categories, namely PSE, PFN (pale, firm, and nonexudative), RSE (red, soft, and exudative), RFN (red, firm, and nonexudative), and DFD (dark, firm, and dry) according to pH24, light reflectance (L*), and drip loss variation (Table 2).

Table 2.

Pork quality classification including pH24, color brightness (L* value) and drip loss (Rocha et al., 2016)

Quality class1	pH24	L*	Drip loss, %
PSE	<6.0	>50	>5
PFN	<6.0	>50	<5
RSE	<6.0	43 to 48	>5
RFN	<6.0	43 to 48	<5
DFD	≥6.0	<42	<2

¹PSE = pale, soft, exudative; PFN = pale, firm, nonexudative; RSE = red, soft, exudative; RFN = red, firm, nonexudative; and DFD = dark, firm, dry.

Statistical analysis

Behavior data were analyzed using generalized linear mixed models for repeated measures with PROC GLIMMIX in SAS (version 9.4, SAS Institute Inc., Cary, NC). Raw and standardized residuals were plotted to determine the distributions of the dependent variables. For pigs’ behavior in the truck, the variables lying and standing showed a log-normal distribution, while sitting showed a normal distribution. All models included treatment (D200, D235, and D270) and farm of origin (1, 2) as fixed effects, shipment (considered as the repeated measure within travel time) as a random effect, and travel time as a covariate. Due to the low presentation of other postures, the statistical model did not converge.

For the behaviors during lairage, the variables lying, standing, and sitting showed a log-normal distribution. For the analysis, the observation time was divided into 6 periods of 10-min. All models included treatment (D200, D235, and D270), farm (1, 2), and observation period (1 to 6) as fixed effects, shipment (considered as the repeated measure within travel time) as a random effect and travel time as a covariate. The variables drinking behavior, agonistic acts, and latency to lie down were analyzed using generalized linear mixed models with PROC GLIMMIX in SAS, and after residuals were plotted, drinking behavior and latency to lie down showed normal distribution and agonistic acts showed a log-normal distribution. Models included treatment (D200, D235, and D270) and farm (1, 2) as fixed effects; and travel time as a covariate.

Blood CK and lactate concentration values were transformed into a logarithmic scale and showed normal distribution. They were analyzed using generalized linear mixed models with PROC GLIMMIX in SAS. Models included farm (1,2), treatment (D200, D235, D270) as fixed effects, and THI and travel time as covariates.

Hot and CCW data showed a normal distribution and were analyzed using generalized linear mixed models with PROC GLIMMIX in SAS. All models included treatment (D1, D2, and D3), farm (1, 2), farm × treatment interaction as fixed effects, and travel time and THI as covariates. The differences in carcass and meat FPW scores between treatments were tested using the Wilcoxon test.

All variables were analyzed using generalized linear mixed models with PROC GLIMMIX in SAS. Raw and standardized residuals were plotted to determine the distributions of the dependent variables. The variables pH1, color a* and b* had a normal distribution, while pH24, drip loss and L* color values showed a log-normal distribution. All models included treatment (D200, D235, and D270) and farm (1,2) as fixed effects, and travel time and THI as covariates.

In the analyses of blood CK and lactate concentration values, and meat quality traits, the effects of treatment × travel time and farm × treatment interactions were tested for all variables, but as no significant effects were observed (P > 0.10), these interactions were excluded from the final model. THI was included in the model only when a significant effect was found (P < 0.10).

Chi-square test was used to verify the relationship between meat quality classes and treatments. For all models, means were compared using post hoc Tukey tests. A probability level of P ≤ 0.05 was chosen as the limit for statistical significance in all tests, whereas probability levels of P between > 0.05 and < 0.10 were considered to be a tendency.

Results and Discussion

Ambient conditions within the compartment

The average air temperature inside the trailer for all trips was 24.9 °C, ranging from 23.1 to 25.9 °C. These temperature values are within the range recommended to ensure the thermal comfort of market weight pigs during transport (15 to 30 °C; Bracke et al., 2020). The average RH was 61.2%, varying from 69.2% to 81.8%. The average THI was 72.5, ranging from 70.6 to 76.6.

Behavioral observations

Transport

As shown in Table 3, the loading densities applied during transport in this study only influenced the lying and sitting postures (P = 0.05 and P = 0.03, respectively). More specifically, the proportion of lying pigs was higher (P < 0.05) in transports at the lowest and intermediate loading densities (D200 and D235, respectively) compared with the highest loading density (D270), while no difference in this posture was found between pigs transported at D200 and D235 (P > 0.10). The higher proportion of lying pigs at D200 and D235 results from the greater space allowed to these pigs during transport (Lambooy and Engel, 1991; Gerritzen et al., 2013). A higher (P < 0.05) proportion of pigs remained in the sitting position during transports at D270 compared with D200, with the proportion of sitting pigs at D235 being intermediate (P > 0.10; Table 3). The dog-like sitting posture is usually interpreted as a certain discomfort of pigs either suffering from heat stress (Ritter et al., 2008) or, at normal temperatures, being provided with limited space availability to settle down and rest in the truck (Dalla Villa et al., 2009). Gerritzen et al. (2013) also reported a higher proportion of pigs sitting when transported at a greater loading density (235 vs. 179 kg/m²).

Table 3.

Arithmetic means (±SD) of the proportion of pigs performing each posture during transport according to the loading density

	Loading density1
Behavior	D200	D235	D270	P
Lying, %	77.3 ± 20.8a	69.6 ± 24.6a	63.0 ± 26.0b	0.05
Standing, %	9.0 ± 14.2	9.0 ± 15.3	7.7 ± 11.9	0.71
Sitting, %	13.5 ± 12.0b	19.4 ± 14.1ab	29.0 ± 19.5a	0.03
OP2, %	0.2 ± 0.6	0.1 ± 0.6	0.2 ± 0.7	—

^a,bWithin a row, means lacking a common superscript letter differ (P < 0.05).

¹D200: 200 kg/m²; D235: 235 kg/m²; D270: 270 kg/m².

²OP: other postures

The interaction between the time (every hour) and treatment in the models had no effect on the variation in on-truck behaviors in this study (P > 0.10).

Lairage

The applied loading densities during transport had an impact on the proportion of animals standing and lying in the lairage pen during rest before slaughter (P = 0.01 and P = 0.02, respectively), while no effect was found on the sitting posture (P > 0.10; Table 4). During rest time in the lairage pen, D200 and D235 pigs stood more than D270 ones (P < 0.05), while no difference was found between D200 and D235 pigs (P > 0.10). As a result, a greater (P < 0.05) proportion of D270 pigs lied down during rest time compared with D200 and D235 pigs, which did not differ between them (P > 0.10). However, the different transport conditions had no effect on the time taken for at least 50% of pigs in the group to lie down (P > 0.10). Loading density during transport influenced the drinking behavior of pigs in the lairage pen (P < 0.01), with D200 pigs drinking water 10 times more frequently during the first hour of lairage compared with D235 and D270 (P < 0.05). The greater proportion of pigs lying down and the lower frequency of drinking behavior in pigs transported at the higher density may be related to their greater fatigue on arrival at the abattoir. This can be associated to lack of rest in the trailer compartment due to insufficient space to lie down and recover from the stress of handling at loading. Several studies showed that pigs arriving fatigued at the abattoir prefer to lie down to recover from the physical stress from transport and handling rather than performing other activities (e.g., fighting, drinking, etc.; Brandt and Aaslyng, 2015; Dalla Costa et al., 2016).

Table 4.

Arithmetic means (±SD) of the proportions of pigs performing each posture, frequencies of drinking bouts and agonistic acts, and latency to lie down in lairage according to the loading density during transport

	Loading density1
Behavior	D200	D235	D270	P
Standing, %	27.0 ± 30.0a	25.4 ± 23.7a	15.9 ± 20.6b	0.01
Lying, %	70.9 ± 30.4b	70.8 ± 24.3b	82.8 ± 21.5a	0.02
Sitting, %	1.9 ± 3.9	3.8 ± 6.3	1.5 ± 3.6	0.24
Drinking bouts2	22.0 ± 9.0a	13.6 ± 6.5b	11.9 ± 4.5b	<0.01
Agonistic acts2	1.5 ± 2.1	0.4 ± 0.5	0.1 ± 0.3	0.64
Latency to lie down, min	16.1 ± 2.0	13.6 ± 4.5	12.1 ± 5.2	0.36

^a,bWithin a row, means lacking a common superscript letter differ (P < 0.05).

¹D200: 200 kg/m²; D235: 235 kg/m²; D270: 270 kg/m².

²Per group of 10 pigs.

Providing pigs with more or less space in the trailer compartment during transport had no effect on pigs’ agonistic behaviors in lairage (P > 0.10). The conditions applied in lairage, i.e., keeping pigs in unmixed and small groups at the recommended stocking density in the lairage pen, might have limited aggressive behaviors, as previously reported in a number of studies (Faucitano, 2018).

The interaction between the time (every hour) and treatment in the models had no effect on the variation in on-truck behaviors this study (P > 0.10).

Physiological conditions at slaughter

In this study, loading density during transport influenced serum CK concentrations at slaughter (P = 0.03; Table 5), with greater (P < 0.05) levels being found in pigs transported at the highest density (D270) compared with those transported at the lowest one (D200). Pigs transported at 235 kg/m² (D235) showed intermediate serum CK concentrations (P > 0.10). The higher CK values in exsanguination blood of D270 pigs may be explained by the effect of long-term stress on animal fatigue and exhaustion (Faucitano and Lambooij, 2019) caused by their physical effort to maintain balance and stability in the sitting position while the vehicle was in motion during an extended lapse of time (8 hr transport). Previous studies also reported higher blood CK levels with increased loading densities in the truck during transport (Barton-Gade and Christensen, 1998; Kim et al., 2004; Gerritzen et al., 2013).

Table 5.

Arithmetic means (±SD) and minimum and maximum values [min – max] of blood CK and lactate concentrations at slaughter according to loading density during transport

	Loading density1
Blood variables	D200	D235	D270	P
CK, UI/L	15,782 ± 15,298b [1,880 to –75,360]	19,301 ± 16,450ab [2,680 to 86,880]	24,974 ± 23,926a [960 to 102,360]	0.03
Lactate, mM/L	11.4 ± 5.3 [3.6 to 27.2]	11.4 ± 4.2 [2.6 to 22.0]	11.9 ± 6.1 [1.50 to 34.20]	0.76

^a,bWithin a row, means lacking a common superscript letter differ (P < 0.05).

¹D200: 200 kg/m²; D235: 235 kg/m²; D270: 270 kg/m².

Blood lactate concentrations at slaughter did not differ between treatments (P > 0.10). This result is not surprising, considering the time provided to pigs in lairage to recover from transport stress and handling (Warriss et al., 1998) and the half-life of lactate concentration in blood after stress (2 hr; Anderson, 2010). Other studies also failed to find effects of loading density during transport on blood lactate levels at slaughter in pigs (Barton-Gade and Christensen, 1998; Gerritzen et al., 2013). However, in another study, Kim et al. (2004) reported higher lactate dehydrogenase concentrations in pigs transported at higher density (0.31 vs. 0.39 m²/100 kg) and rested for 2 hr before slaughter. The discrepancy in the results between studies may be explained by the pigs’ different response to peri-mortem handling which may be influenced by their memory of previous negative handling experience (e.g., loading; Correa et al., 2010; Edwards et al., 2010). Furthermore, based on the short-term peaking of lactate levels in blood (4 min; Anderson, 2010), in this study blood lactate results may have been biased by the acute handling stress (i.e., single line handling and electric prodding; Edwards et al., 2010) experienced by all pigs in the stunning chute.

Carcass quality

Similar to other studies (Nanni Costa et al., 1999; Kim et al., 2004), loading density during transport had no impact on HCW and CCW (P > 0.10; Table 6). The application of recommended total feed withdrawal time (up to 24 hr; Faucitano et al., 2010) and the controlled carcass chilling conditions (Savell et al., 2005) may explain this lack of effects on carcass weight losses in this study.

Table 6.

Arithmetic means (±SD) of HCW and CCW, respectively, and arithmetic means and minimum and maximum values [min–max] of skin damage scores according to the loading density during transport

	Loading density1
Carcass traits	D200	D235	D270	P
HCW, kg	87.6 ± 5.6	87.0 ± 5.8	86.1 ± 5.5	0.50
CCW, kg	85.8 ± 5.4	85.2 ± 5.7	84.2 ± 5.6	0.25
Skin damage score2	2.0b [1 to 3]	2.0b [1 to 3]	2.1a [2 to 4]	0.03
Handling-type bruise score3	1.8b [1 to 3]	1.7b [1 to 3]	2.0a [1 to 3]	0.06
Mounting-type bruise score4	1.0 [1 to 3]	1.0 [1 to 3]	1.0 [1 to 3]	0.72
Fighting-type bruise score5	1.2 [1 to 2]	1.2 [1-1]	1.3 [1 to 3]	0.24

^a,bWithin a row, means lacking a common superscript letter differ (P < 0.10).

¹D200: 200 kg/m²; D235: 235 kg/m²; D270: 270 kg/m².

²1 = none to 5 = severe (MLC, 1985).

³1 = 1 bruises to 3 = greater or equal than 3 bruises (ITP, 1996).

⁴1 = <10 bruises to 3 = greater than 21 bruises (ITP, 1996).

⁵1 = <5 bruises to 3 = greater than 11 bruises (ITP, 1996)

In this study, skin damage scores were higher (P < 0.05) in D270 carcasses compared with D200 and D235 ones (Table 6). The trend for a greater (P = 0.06) number of handling-type bruises on D270 carcasses may have contributed to this difference between treatment groups (P = 0.03). These results may be explained by the effects of the fatigue condition (based on blood CK levels at slaughter) of pigs resulting from their inability to rest during long-term transport that may have made them more difficult to handle during unloading or in the lairage alleys (uncontrolled in this study). There is evidence that fatigued pigs are harder to move and require more handler’s interventions than fit pigs (Rocha et al., 2019).

No differences were found between treatments for mounting- and fighting-type bruises (P > 0.10). While the lack of effect of loading density on fighting-type bruise score can be explained by the similar expression of agonistic behavior between treatment groups in lairage (Table 4), the similar mounting-type bruise scores between treatments are surprising and contrast with the results of other studies reporting increased overlapping and trampling in search of a place to lie down in pigs provided with limited space conditions in the truck (Gispert et al., 2000; Guàrdia et al., 2009).

Meat quality parameters

Except for the pH1 value, loading density during transport had no effect on meat quality parameters in this study (P > 0.10; Table 7). In the LT muscle of D270 pigs the pH1 value tended to be lower (P = 0.10) than that of D200 loins, with pH1 value of D235 being intermediate, suggesting a slight effect of transport conditions on postmortem meat acidification rate. However, the differences between treatments in this meat quality parameter are of little biological and economical importance.

Table 7.

Arithmetic means (±SD) of meat quality traits as assessed in the LT muscle of pigs according to the loading density during transport

	Loading density1
Traits	D200	D235	D270	P
pH1	6.23 ± 0.24a	6.18 ± 0.2ab	6.15 ± 0.19b	0.10
pH24	5.73 ± 0.19	5.73 ± 0.15	5.78 ± 0.20	0.19
L*	51.99 ± 3.50	52.03 ± 3.31	51.27 ± 4.01	0.42
a*	3.18 ± 1.47	2.83 ± 1.26	2.79 ± 1.26	0.15
b*	12.26 ± 1.45	12.23 ± 1.32	11.93 ± 1.66	0.36
Drip loss, %	1.68 ± 1.00	1.77 ± 1.07	1.56 ± 0.98	0.79

^a,bWithin a row, means lacking a common superscript letter differ (P < 0.10).

¹D200: 200 kg/m²; D235: 235 kg/m²; D270: 270 kg/m².

In general, more than 57 % of the loins evaluated in this study were classified as PFN (Table 8). This meat quality defect has been associated to the animal response to some kind of preslaughter stress, such as negotiating trailer internal ramps or poor handling at loading and unloading (Correa et al., 2013; Rocha et al., 2016). Pork meat with DFD characteristics, indicator of antemortem muscle glycogen depletion, has been only found in D270 loins, which confirms the effects of transporting pigs at high loading density on fatigue condition at slaughter (based on blood CK levels at slaughter).

Table 8.

Distribution of pork quality classes according to the pigs’ loading densities during transport

		Loading density1
Quality class2	D200	D235	D270	Total
PSE, %	0	0	0	0
PFN, %	20.4	19.1	17.9	57.4
RSE, %	0	0	0	0
RFN, %	4.6	4.6	2.1	11.2
DFD, %	0	0	0.8	0.8

¹D200: 200 kg/m²; D235: 235 kg/m²; D270: 270 kg/m².

²PSE = pale. soft, exudative; PFN = pale, firm, nonexudative; RSE = red, soft, exudative; RFN = red, firm, nonexudative; DFD = dark, firm, dry.

Overall, the lack of or minimal effects of in-transit loading densities on meat quality traits in this study may be either explained by the sufficient time allowed to pigs for resting and recover from transport stress (Faucitano, 2018) or by the poor handling just before slaughter that levelled the stress conditions in all pigs, regardless of the transport treatment. Another explanation may be the muscle chosen for meat quality assessment, with postural muscles, such as the loin, being less prone to more rapid glycogen exhaustion after physical exercise compared with ham muscles (Correa et al., 2013; Rocha et al., 2015).

Conclusions

Based on the results obtained under the transport conditions of this study, i.e, long transport and wait times before unloading, transporting market weight pigs at a loading density of 270 kg/m², which is a common commercial practice in Brazil, should not be recommended as it results in pigs’ discomfort during the journey, fatigue at slaughter and increased carcass lesion scores. Loading densities lower than or equal to 235 kg/m² appear more suitable to respect the physical needs of pigs during transport without negative consequences on carcass quality. However, aiming at providing a more comprehensive and useful recommendation on loading densities to apply under more variable transport conditions, and limiting animal losses, and carcass and meat quality defects (including ham muscles in the assessment), larger scale studies, including season (winter and summer) and travel time (short vs. long) factors, and more controlled handling conditions in lairage and at slaughter are needed.

Glossary

Abbreviations

CK

creatine kinase

DFD

dark, firm, dry

FPW

filter paper wetness

LT

longissimus thoracis

PFN

pale, firm, nonexudative

PSE

pale, soft, exudative

RFN

red, firm, non-exudative

RH

relative humidity

RSE

red, soft, exudative

T°

air temperature

THI

temperature-humidity index

Funding

This study was supported by Grupo de Pesquisa em Etologia e Ecologia Animal (ETCO) and Grupo de Pesquisa em Análise de Carnes (GPAC). We are grateful to Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for granting the scholarship to V. M. Urrea. The authors thank E. R. Correia, É. R. Dos Santos. D. K. Terto, J. G. Vero, G. A. Ferreira, and A. G. Barro for their valuable contribution in the data collection at the farm and abattoir. Many thanks go to the Clinical Pathology laboratory at the Universidade Estadual de Londrina (Brazil) for performing the blood analysis. We are also grateful to Agrosuinos Serafini trucking company for providing their trucks and staff and to Rainha da Paz Frigorífico for providing pigs, slaughter facilities and manpower.

Conflict of Interest Statement

The authors declare no real or perceived conflicts of interest.