Physiological response of weaned piglets to two transport durations observed in a Canadian commercial setting

Abstract

Observational studies describing the impact of transport duration on weaned piglet welfare are limited. Current Canadian transport regulations are heavily informed by studies involving market hogs. Due to physiological differences between weaned piglets and market hogs, additional data on their response to transport are needed for age-specific evidence-based recommendations. A cohort study was conducted to describe and compare mortality, injury, weight change, hematological or biochemical changes in hydration, muscle injury and stress response observed in weaned piglets undergoing short duration (SD, <3 h), or long duration (LD, >30 h) commercial summertime transport events. Data collection on 440 of 11,434 transported piglets occurred the morning of the day before transport (T0), at arrival (T1) and approximately 3 to 4 d (78 to 93 h) after arrival at the nursery barn (T2). Low mortality occurred over all transport events (0.06%) with no association observed between transport duration and odds of death during transport (P = 0.62). The incidence of lameness between T0 and T1 was low (1.84% of the 435 focal piglets scored) with all lameness cases identified as mild in severity. Lesions on ears and skin were more prevalent than other injury types after transport (T1) and may have been related to mixing aggression associated with weaning rather than transport alone. LD piglets weighed 0.39 kg less than SD piglets at T1 (P < 0.01), but no difference in group weight was observed at T2 (P = 0.17). Hematological and biochemical differences were present between groups at T1. LD piglets had increased hematocrit levels compared with SD piglets (P = 0.01), suggesting increased body water losses. SD piglets showed greater levels of muscle injury compared with LD piglets including elevated aspartate aminotransferase (P < 0.01) and creatine kinase (P < 0.01). However, these parameters were within normal reference ranges for piglets of this age group. Indicators of physiological stress response including cortisol and neutrophil to lymphocyte ratios were elevated in SD piglets compared with LD piglets (P = 0.02 and P < 0.01, respectively). The results of this study demonstrate that both short and long transport durations can result in detectable physiological changes in weaned piglets. The overall impact of these durations on piglet welfare should be further explored by analyzing behavioral time budgets during and after transport.

Introduction

Multi-site pig production was developed and implemented to prevent the transmission of pig diseases between age groups (Harris, 2000). This production strategy typically involves transporting pigs to different farm sites multiple times throughout their lives. The first opportunity for transporting pigs is at weaning. Weaned piglet transport occurs commonly within the North American swine industry: studies characterizing shipment events within large pork producing regions in Canada have reported the transport of weaned piglets from sow farms to nursery farms as the second most logged category of pig movement (Melmer et al., 2018; Thakur et al., 2016). Although transport for the purpose of relocation to a different site can be beneficial for piglet welfare by reducing the risk of transmission of infectious agents to susceptible pigs, many aspects of commercial transport have been reported as stressful for piglets (Roldan-Santiago et al., 2013). Transport duration is recognized as one of these aspects and has been identified as a priority swine welfare issue by the Canadian National Farm Animal Care Council Transportation Code of Practice Scientific Committee (Rioja-Lang et al., 2019). Canadian federal transport regulations mandate the maximum duration that pigs under transport can be without access to appropriate feed, water, or a rest period (Health of Animals Regulations, 2020). Currently, this maximum duration is the same for all ages of swine and the majority of regulation-informing data have been collected from market hogs (Lewis, 2008; Nielsen et al., 2011; Sutherland et al., 2014; Rioja-Lang et al., 2019).

The difference between these life stages (market hog vs. weaned piglet) with regard to body size and proximity to weaning may result in weaned piglets having different responses to transport conditions or durations compared with market hogs. For example, the commercial weaning process and accompanying management changes may result in piglet fatigue or injury due to aggression after mixing (Mei et al., 2016). Piglets also have significantly decreased feed intake after weaning (McCracken et al., 1999; Bruininix et al., 2001; Tao et al., 2016). Although fasting piglets for 12 h before transport has been observed to lower in-transit mortality rate (Averós et al., 2010), the period of postweaning anorexia could overlap with long transport durations without feed accessible in-transit, therefore resulting in durations without significant energy intake exceeding the maximum time periods. This may also have implications for thermoregulation; cold stress experienced by piglets during transit may be exacerbated by prolonged fasting as a lower energy intake reduces piglet heat production (Collin et al., 2001). Furthermore, young weaned piglets (7 to 15 kg) require higher ambient temperatures compared with finishing hogs; suggested ambient temperature zones in the pen environment reduce from 24 to 30 °C to 16 to 20 °C between these life stages (Kyriazakis and Whittemore, 2006). Even during the summertime, optimal environmental temperatures for weaned piglets may be challenging to maintain during long transport events in North American climates where temperatures may be cool during early morning loading or overnight periods but rise rapidly during the day. These factors underscore why the effect of transport duration on the welfare of weaned piglets is an area requiring additional research (Rioja-Lang et al., 2019).

Previous studies investigating the effects of transport duration on young pigs have been conducted using commercial and experimental transport conditions with a range of studied response variables. However, observational research involving a comprehensive physiological assessment of weaned piglets’ response to continuous transport durations >24 h in length with a short duration (SD) comparator group is limited. Averós et al. (2009) used commercial transport conditions to compare the physiological effects of short (<1 h) and medium (~8 h) duration transport events and observed some evidence of increased fatigue for short-transported piglets, and no change in cortisol levels after transport for either group. The interactive effects of transport and temperature on early-weaned piglet performance after durations up to 24 h have been investigated using simulated transport (Berry and Lewis, 2001). This study demonstrated the negative physiological impact of exposure to high temperatures for medium and long periods of time; piglet weight loss lasted for up to 7 d after piglets were “transported” for both 6 and 24 h at 35 °C. Garcia et al. (2015) conducted comprehensive physiological and behavioral assessments on weaned piglets transported for 32 h under experimental conditions (small group size and interrupted transport) compared with nontransported weaned piglets and nonweaned control piglets. Authors reported no difference in neutrophil to lymphocyte ratios (N:L) among transported, weaned, and nonweaned piglets, though significant weight loss occurred for piglets transported for 32 h without access to feed and water compared to nontransported piglets with feed and water access. Finally, weaned piglet predicted mortality rates have been calculated from American commercial swine transportation records (Zhao et al., 2016). While rates of >0.4% were reported, predicted mortality changed markedly with environmental conditions and trip duration times. Studies reporting predicted piglet mortality rates or probabilities have examined transport events less than 24 h in length (Harmon et al., 2017), distance categories up to >1,500 km (Zhao et al., 2016), and durations >24 h with time taken for rest stops included (Averós et al., 2010).

As durations >24 h in length are possible in North America due to geographical span and a widely dispersed but connected swine industry, research addressing the effects of transport durations >24 h on piglet welfare would be useful to support evidence-based recommendations for optimizing piglet well-being during and after transport. The overarching objectives of this research project were to generate data on the physiological and behavioral response of weaned piglets to commercial transport through an observational study design. The specific objective of this study was to describe and compare mortality, injury, weight change, and hematological or biochemical changes in hydration, muscle injury, and stress response observed in weaned piglets after short (<3 h) or long (>30 h) summertime commercial transport events. Due to the marked difference in transport duration between groups, the accompanying time without access to feed and water, and anticipated changes in environmental exposures, it was hypothesized that there would be significant differences among the evaluated physiological parameters collected between duration groups.

Materials and Methods

Animals and sampling

The data collection procedures for this project were approved by the University of Guelph (Animal Use Protocol (AUP) #4165, Regional Ethics Board (REB): 19-02-034) and University of Saskatchewan (AUP-REB # 20180072) Animal Care Committees and Animal Research Ethics Boards. A prospective cohort study was completed between June and August of 2019 using scheduled transport events from a convenience sample of two Canadian commercial farm systems. The farms were selected based on the duration of transport between their sow and nursery barn locations allowing categorization of the transport events into long (>30 h) or short (<3 h). Descriptions of the observed transport events are presented in Table 1. The long duration (LD) system transported piglets from a sow farm in Saskatchewan to nursery barns in Ontario. The SD system transported piglets between a sow farm and nursery barns within Ontario. Both sow farms shipped piglets (commercial crosses of Landrace, Duroc, and Large White breeds) to affiliate nursery farms after weaning at ~17 to 24 d old. Data collection occurred at three time points; the morning of the day before transport (T0), immediately after unloading at the destination (T1), and approximately 3 to 4 d (78 to 93 h) after arrival at the nursery barn (T2). A description of the data collection timepoints and measures collected for each treatment (duration) group is presented in Figure 1a (LD) and Figure 1b (SD).

Table 1.

Description of transport events observed

Treatment group	Short (SD)	Long (LD)
Duration and piglet characteristics
Transport events observed	4	4
Piglets sampled1	200	240
Livestock trailer and route descriptions
Trailer type	Flat deck	Potbelly
Average load size (min, max)	508.5(454, 567)	2350(2200, 2500)
Main drivers per trip2	1	2
Decks	1	4
Total compartments	1	9
Monitored compartments	1	3
Total piglets sampled per compartment (male/female)	Flat deck: 200(100/100)	Upper-back:60(33/27) Belly:59(29/30) Bottom-front:121(58/63)
Ventilation type	Passive	Passive
Bedding availability	Yes	Yes
Water availability	No	No
Destination nurseries3	3	3

¹Number of focal piglets sampled per transport event differed to accommodate nursery pen sizes: 50 per SD and 60 per LD transport event.

²Primary drivers within each treatment group were consistent for each transport event.

³Each transport event delivered to a commercial nursery facility. In each treatment group, one nursery received piglets twice.

Figure 1.

Data collection timeline for transport events. (a) Long duration transport events. (b) Short duration transport events.

Long distance transport events

The large distance between the LD sow farm and the nursery barns required the formation of two research teams (one team at the sow barn and the other team at the nursery barn) to complete data collection at the three timepoints (Figure 1a). One team (Team 1) collected piglet data at T0 and were responsible for mounting monitoring equipment on the LD trailer before loading. At T0, Team 1 purposively enrolled 60 focal piglets, balanced for sex, from temporary pens of piglets weaned for transport. The sample size of focal piglets was calculated based on previous research evaluating piglet performance outcomes after transport (Wamnes et al., 2006), and changes in multiple market hog physiological parameters after transport (Weschenfelder et al., 2013) to assess changes in physiology with adequate power (>80%) while balancing for the reduction of sampled animals and feasibility of sampling. The number of transport events followed per duration group was determined based on the number of focal piglets required and the number of transport events possible to measure during the single summer season within the existing farm transport schedule. All piglets enrolled were weaned between one and 6 d before transport and were selected based on being of average size out of the available group (~5 kg). At T0, focal piglets were individually identified with ear tags and livestock spray to allow for identification during loading and unloading. Focal piglets were weighed using a small shipping scale (Dymo Heavy Duty Digital Mailing Scale 100lb capacity, Dymo Corp). The same scale model was used at each subsequent data collection location. Injury assessment of focal piglets was conducted by recording the presence and severity of injury (lameness and lesions on the ear, tail, and body regions) by trained team members using the scoring scheme presented in Table 2.

Table 2.

Injury severity score chart for lesion and lameness categorization at each data collection timepoint

	Absent (0)	Mild (1)	Moderate (2)	Severe (3)
Skin1	Absence of any wounds or scratches	1–2 superficial wounds, scabs, or scratches (skin not broken) involving face and trunk	3–4 superficial wounds or scratches or one small would (skin broken, <2 cm in length)	>1 large wound (skin broken, >2 cm in length) or ≥5 superficial wounds or scratches
Ear2	No injury or necrosis present; the ears blemish free	1–2 superficial wounds, scabs, or scratches	3–4 superficial wounds or scratches or 1 small would	>1 large wound or ≥5 superficial wounds or scratches
Tail3	No evidence of lesions or tail biting observed	Presence of a scab or indication of superficial biting along the length of the tail. No evidence of fresh blood or swelling	Fresh blood visible on the tail; there is evidence of some swelling and possibly infection. Tail remains intact	Severe wounds and a significant part of the tail missing. A crust may have formed where there is missing tail
Gait4	Piglet easily and comfortably moves on all limbs	Piglet moves relatively easily but with reluctance to bear weight on one leg	Piglet is nonweight bearing on one or two limbs and may exhibit compensatory behavior such as head dipping	Piglet is reluctant to move at all or bear weight on any limb (nonambulatory)

¹Skin on the face, legs, and trunk were assessed for lesions.

²The tips and skin of the external pinnae were assessed for lesions and/or necrosis.

³The length and tip of the tail were assessed for lesions and/or necrosis.

⁴Piglet gait was observed and scored when animal was set down on an even floor surface (solid or slatted) after weighing.

From each group of focal piglets per transport event, 20 had blood samples collected by suborbital sinus sampling using a 16 to 18 gauge, 1- or 1.5-in. needle (detectable aluminum hub, SyrVet, QC, Canada). A total of 15 mL of blood was collected from each piglet for hematological and biochemical testing: 5 ml of whole blood using BD Vacutainer spray-coated K2EDTA tubes which were inverted 10 to 15 times after collection, and 10 mL for serum separation using BD Vacutainer silicone-coated serum tubes (BD, Franklin Lakes, NJ). One drop of whole blood was used for pen-side lactate testing using precalibrated portable analyzers (Lactate Scout/Lactate Plus, Sports Resource Group, Inc., MN). The remaining whole blood samples were submitted for complete blood count (CBC) analysis to Prairie Diagnostic Services, University of Saskatchewan. Serum samples were left to clot at room temperature for approximately 1 h, then all samples were transported on ice. Serum samples were centrifuged between 3 and 5 h after collection using a Centra CL3R Thermo IEC centrifuge at 1,400 × g for 20 min at 4 °C. Samples were stored at −80 °C before shipment to the Animal Health Laboratory (AHL) at the University of Guelph for serum biochemistry and cortisol analysis by chemiluminescence.

The morning of transport, Team 1 mounted environmental monitoring equipment inside and outside of the LD trailer before the piglets were loaded. The LD system transported piglets using a four-deck potbelly trailer, with three compartments selected to house the focal piglets during transport and for environmental monitoring (Figure 2a). The selected compartment locations were the upper-back (C-UB), bottom-front (C-BF), and bottom-middle (belly) (C-B), which have previously been observed to have differences in environmental conditions during transport (Brown et al., 2011). A temperature (°C) and relative humidity (%RH) logger (Hygrochron (DS1923-F5), Maxim Integrated) was secured above pig level at the front and back of each focal compartment and recorded measurements every five minutes during transport (Figure 2). Loggers were also attached to each of the tractor side mirrors to record external temperature and humidity at the same intervals. Neither farm withdrew feed nor water from piglets before loading and equipment used to assist with loading and unloading was limited to methods approved by the Canadian Code of Practice for the Care and Handling of Pigs, including shaker cans and pig boards (National Farm Animal Care Council, 2014). All LD transport events included brief stops; an additional loading stop at an affiliated farm and breaks for refueling and driver changes. A second team (Team 2) collected data on LD piglets at T1 and T2 (Figure 1a). Team 2 was present at the nursery farms when the transported piglets were delivered and recorded the time taken to unload each trailer and the incidence of in-transit mortality. Focal piglets were placed together in separate pens in the nursery, which corresponded to their transport compartment. At this time (T1), sampling was repeated on the focal piglets as described for T0, except whole blood samples were submitted for CBC analysis to the AHL. At T2, the research team returned to the nursery farm to re-weigh and score all focal piglets for injury presence and severity.

Figure 2.

Schematic of the commercial trailers and environmental monitoring equipment used for each treatment group, stars represent temperature and relative humidity logger (Hygrochron DS1923-F5, Maxim Integrated) locations, suspended from the compartment ceiling. (a) Potbelly trailer with nonhydraulic decks and ramps used for long duration transport events. Three of the compartments were monitored and had focal piglets loaded for the study (C-UB, C-BF, and C-B). (b) Flat deck trailer with a single deck and shortened compartment (F) to approximate the dimensions of C-BF on the potbelly trailer housed all focal and nonfocal piglets for the short duration transport events.

Short distance transport events

Data collection for the SD transport events followed the same timeline and procedures as the LD data collection described above, with some exceptions. Team 2 collected data from the SD piglets at all timepoints (Figure 1b). The SD sow farm weaned smaller batches of piglets into a corridor the morning of transport immediately before loading, allowing Team 2 to conduct T0 enrollment and sampling while piglets were still with their sow in farrowing crates. The SD focal piglets were also purposively selected for enrollment; however, SD sow barn farrowing crate numbers were randomized in advance of T0 sampling using a random number generator (www.random.org) with two to four average size piglets enrolled from each selected litter depending on the number of litters available. For each SD transport event, 50 piglets were selected, balanced for sex. This number of focal piglets was enrolled for the SD transport events due to different nursery pen sizes after transport. Whole blood samples collected at T0 and T1 were submitted for CBC analysis to the AHL. All focal piglets from the SD farm were loaded onto the bottom floor of a flat deck trailer, which was penned off to be of comparable size and stocking density to C-BF of the LD potbelly trailer (Figure 2b). The SD transport events included a stop to weigh the loaded trailer to measure piglet batch weight, which was standard operating procedure for the system. At arrival, SD focal piglets were divided into one or two pens, with 50 or 25 pigs per pen, respectively, depending on the nursery barn pen sizes. The duration between T1 and T2 sampling was longer for SD than for LD transport events due to farm access limitations (~92 to 93 h compared with 78 to 80 h).

Analysis

All statistical analyses, including descriptive statistics, outcome modeling and model diagnostics were completed in SAS (SAS Studio University Edition, SAS Inst. Inc., Cary, NC). A probability level of 5% was used for significance interpretation of all results. Temperature (T) in °C and RH as a percentage (%) were downloaded from the hygrochrons as individual.csv files. Average values and standard errors of the mean reported were recorded during transport only; data collected during loading and unloading periods were excluded. Fisher’s exact tests and odds ratios were calculated using PROC FREQ to assess the association between transport duration and in-transit mortality for all transported piglets. Injury scores at each timepoint were dichotomized into absent/mild (scores zero and one) or moderate/severe (scores 2 and 3) and associations between transport duration and injury severity were evaluated using Fisher’s exact tests and odds ratios using PROC FREQ. The proportion of sampled piglets which had a decrease, no change, or increase in injury score severity between T0 and T1 was calculated, with differences among duration groups evaluated using a Fisher’s exact test.

For body weight and blood parameter variables, descriptive statistics were used to identify missing observations or lack of variability. Before regression model building, variance component analysis was completed in an empty model to determine the contributions of random effects to error explanation. The random effects tested included transport event, trailer compartment, nursery pen, and piglet ID. To evaluate if differences in piglet body weight existed between the treatment groups at each time point (T1 and T2), a mixed multivariable linear regression model was produced using PROC MIXED (referred to as BW1). Forward step-wise model building was used to test fixed effects including treatment group (long, short), sex (male, female), time of sampling (T1, T2), and the interaction between treatment group and time of sampling, with only significant effects (P < 0.05) or confounding variables remaining in the final model. Confounding variables were identified a priori with causal diagrams or identified if their inclusion in a model resulted in greater than a 20% change in another model coefficient on the natural log scale. Piglet body weight at T0 was included as a covariate regardless of significance. Random intercepts included were transport event number, to account for environmental and batch variation between transport events; and piglet ID, to adjust for lack of independence between repeated observations. The Satterthwaite adjustment for denominator degrees of freedom was used. To evaluate if differences in piglet body weight existed between the compartments of the LD potbelly trailer at each time point (T1 and T2), a mixed multivariable linear regression model was produced using PROC MIXED (referred to as BW2). This second body weight model (BW2) was constructed in the same fashion as BW1, except with trailer compartment included as the independent variable of interest in place of treatment group. For both models, PROC UNIVARIATE and plotted residuals were used to assess normality, homoscedasticity, and identify outliers.

The average values for all hematological and biochemical parameters were calculated and compared to normal reference ranges previously published for a similar piglet population to allow for improved interpretation of clinical significance (Perri et al., 2017), as reference ranges reported by the laboratory were calculated from adult animals. Mixed multivariable linear regression models were built using PROC MIXED to assess the effect of treatment group on dependent variables measuring hydration, muscle injury, and stress response, listed in Table 3. Similar to the body weight models, forward step-wise model building was used with the T0 parameter value included as a covariate, and transport event included as a random effect. PROC UNIVARIATE and plotted residuals were used to assess normality, homoscedasticity, and identify outliers. A log₁₀ transformation of the data was conducted if residuals were not normally distributed or homoscedastic, and back-transformed estimates (geometric means) reported. A second set of models for the same dependent variables (Table 3) were constructed to compare differences between piglets transported in different compartments of the LD potbelly trailer. These models were built in a similar fashion as the first set, except with trailer compartment included as the independent variable of interest in place of treatment group.

Table 3.

Hematological and biochemical parameters selected to assess hydration, muscle injury, and physiological stress response

Parameter	Test
Hydration
Hematocrit	Whole blood complete blood count1
Total protein	Serum biochemistry2
Albumin	Serum biochemistry
Muscle injury
Creatine kinase	Serum biochemistry
Aspartate aminotransferase	Serum biochemistry
Lactate	Whole blood portable analyzer3
Physiological stress
Cortisol	Serum chemiluminescence test
White blood cell count	Whole blood complete blood count
Neutrophil:lymphocyte ratio	Whole blood complete blood count

¹Complete blood count on whole blood sample (EDTA) collected from the suborbital sinus.

²Comprehensive serum biochemistry profile from suborbital sinus samples.

³Lactate Scout and Lactate Plus portable monitors, using EDTA whole blood samples.

Results

Over the 3-mo period, data were collected on four transport events from each treatment group including a total of 11,434 piglets with 440 of these directly sampled. The length of LD transport events were at most, 4.5% different from each other (111 min), whereas the length of SD transport events were at most, 21.1% different from each other (19 min). All hygrochron loggers from SD transport events launched and recorded environmental data for extraction after transport. Data from a mixed subset of loggers from the LD transport events were lost due to technical problems; data were not available from three of the 24 mounted compartment loggers and three of the eight mounted exterior loggers with no exterior environmental data available from the first LD transport event. The exterior T and RH values recorded over all SD transport events (average ± SEM) were 22.3 °C ± 0.11 and 70.7% ± 1.03 compared with 19.1 °C ± 0.10 and 67.4% ± 0.35 for LD transport events. Exterior SD T and RH recordings ranged from 19.1 to 25.6 °C and 49.7% to 97.4%, respectively. Long duration transport event exterior T and RH recordings ranged from 7.1 to 36.1 °C and 24.5% to 100.6%, respectively. Interior compartment conditions for both treatment groups were, on average, warmer and less humid than exterior conditions; the average interior values were 24.3 °C ± 0.10 and 62.7% ± 0.82 for SD transport events and 23.0 °C ± 0.053 and 55.1% ± 0.15 for LD transport events. Interior SD compartment T and RH recordings ranged from 20.6 to 27.1 °C and 47.2% to 82.7%, respectively. Long duration compartment T and RH recordings ranged from 6.1 to 41.6 °C and 18.6% to 91.2%, respectively.

All piglets found dead on arrival were nonfocal piglets from LD transport events. From all of the piglets transported during the eight observed transport events, seven piglets (0.06%) died between loading and unloading. No association was observed between transport duration and odds of death during transport (P = 0.62). The distribution of injury scores for each piglet at each timepoint is presented in Figure 3, showing changes in the prevalence of score severity over time, both within and between treatment groups. Results of the Fisher’s exact test and odds ratio calculations for injury scores are presented in Table 4. Odds ratios could not be computed in all circumstances as all SD piglets scored absent/mild for the skin region at T0 and the tail region at all three timepoints. Measures of association were also not possible for gait scores when dichotomized, as no piglets had a gait score of greater than one (mild lameness). The proportion of piglets in each treatment group showing an increase or decrease in injury or gait score, or no change, over the transport period are presented in Table 5. In both treatment groups, the ear location had the greatest proportion of piglets showing an increase in lesion severity, followed by the skin location. After transport, SD piglets had significantly greater proportions of gait and skin injury score severity increases compared to LD piglets.

Figure 3.

Distribution of injury scores for the sampled population by treatment group at each sampling timepoint. See Table 2 for corresponding score descriptions. Score severity is represented by legend colors with each row displaying a score for an individual piglet. Timepoints included T0 (before transport), T1 (arrival), and T2 (~3 to 4 d [78 to 93 h] after arrival). Five piglet scores were not recorded for the fourth SD transport event at T1 and are represented by white bars.

Table 4.

Odds of short duration focal piglets having a moderate or severe lesion or gait score¹ compared to long duration focal piglets

	Number of moderate/severe injury scores		Odds ratio (95% CI)2	P-value3
	Long	Short
T0 (n = 440)
Skin	3/1	0/0	NC4	0.130
Tail	2/0	0/0	NC4	0.503
Ear	53/11	19/0	0.29 (0.16, 0.51)	<0.001
Gait	0/0	0/0	NC4	NC5
T1 (n = 435)6
Skin	4/0	9/0	2.86 (0.78, 12.86)	0.091
Tail	0/0	0/0	NC4	NC5
Ear	79/15	28/0	0.26 (0.16, 0.43)	<0.001
Gait	0/0	0/0	NC4	NC5
T2 (n = 440)
Skin	10/0	1/0	0.12 (0.0027, 0.83)	0.014
Tail	1/0	0/0	NC4	1.00
Ear	66/1	109/12	3.96 (2.60, 6.02)	<0.001
Gait	0/0	0/0	NC4	NC5

¹All outcomes were scored on a scale from 0 to 3 (Table 2) but were dichotomized to ≤1 (absent, mild) or ≥2 (moderate, severe).

²Odds ratios with exact 95% confidence limits reported, with long duration (LD) as the referent.

³Two-sided P-value reported.

⁴Odds ratios not possible to calculate (NC).

⁵No lesion or gait scores ≥2 reported making evaluation of the association not possible to calculate (NC).

⁶Five piglets were not scored for the fourth short duration transport event at T1.

Table 5.

Proportion of sampled piglets which had a decrease, no change, or increase in injury score severity¹ between the before and after transport sampling timepoints (T1–T0)

	Short duration (SD) (n = 195)2			Long duration (LD) (n = 240)
Location	Decrease	No change	Increase	Decrease	No change	Increase
Gait	1.0% (2/195)	95.4%a (186/195)	3.6%c (7/195)	0.0% (0/240)	99.6%b (239/240)	0.4%d (1/240)
Skin	2.6%a (5/195)	79.0%c (154/195)	18.5%e (36/195)	24.2%b (58/240)	65.8%d (158/240)	10.0%f (24/240)
Ear	22.6% (44/195)	53.9% (105/195)	23.6% (46/195)	22.5% (54/240)	50.8% (122/240)	26.7% (64/240)
Tail	7.7%a (15/195)	92.3%c (180/195)	0.0% (0/195)	22.1%b (53/240)	76.3%d (183/240)	1.7% (4/240)

¹Score severity descriptions presented in Table 2.

²Five piglet scores were not recorded for the fourth SD transport event at arrival.

Values within change categories (decrease, no change, and increase) between duration groups with different letters differ significantly at P < 0.05, as determined through a Fisher’s exact test.

Average focal piglet body weights at T0 were 6.2 kg (range: 4.5 kg) for the SD group and 6.0 kg (range: 6.1 kg) for the LD group. Weights for SD and LD piglets at T1 after transport were 6.2 kg (range: 4.8 kg) and 5.6 kg (range: 5.4 kg), respectively. By T2, the group average for SD piglets was 6.6 kg (range: 5.1 kg) and 6.3 kg (range: 7.4 kg) for LD piglets. Variance component analysis showed that the majority of body weight variance was explained at the piglet level (57.8%), and by transport events (10.2%), with far less explained at the nursery pen (1.3%) and trailer compartment (0.29%) levels. The fixed effects of sex (P = 0.027), T0 body weight (P < 0.0001), time (P < 0.0001), duration (P = 0.007), and the interaction between duration and time (P < 0.0001) were all significant in the BW1 Model (Table 6). Piglets from the LD group weighed significantly less than piglets from the SD group at T1, but not at T2 (Table 6). No significant differences in group weight were observed among the LD compartments at either timepoint after transport (data not presented).

Table 6.

Differences of least squares means¹ of piglet weights (kg) observed

Effect	Variables		Difference	95% CI for difference		SE	P-value
				Lower bound	Upper bound
Sex	Female	Male	−0.068	−0.13	−0.0078	0.030	0.027
Duration	Long	Short	−0.25	−0.40	−0.098	0.061	0.007
Time2	T1	T2	−0.55	−0.61	−0.49	0.030	<0.0001
Duration × time	Long, T1	Short, T1	−0.39	−0.55	−0.24	0.068	0.0002
	Long, T2	Short, T2	−0.099	−0.25	0.053	0.068	0.174

¹Results of the mixed multivariable linear regression model BW1.

²Timepoints included T1 (arrival), and T2 (~3 to 4 d [78 to 93 h] after arrival), as T0 (before transport) weight was included as a covariate. Five piglet weights were not recorded for the fourth short duration transport event at T1.

All hematological values measured in this study were within the 95% reference intervals published for Ontario piglets around the time of weaning (Perri et al., 2017). In some instances, serum cortisol values were reported by the lab as <28 nmol/L (below the minimum detection threshold) and were assigned a value of 14 nmol/L to facilitate statistical analysis. There was one piglet with markedly elevated aspartate aminotransferase (AST) and creatine kinase (CK) values (146 and 6,049 U/L, respectively) at T0. The sample was from a female piglet from the first LD transport event, with a T0 weight of 6.1 kg, mild ear lesions present, and no recording errors or other abnormalities found. The observation was removed from analysis due to the influence observed on the models.

All selected parameters used to evaluate hydration, muscle injury, and physiological stress response (Table 3) required log₁₀ transformation except for white blood cell count (WBC), hematocrit (HCT), total protein (TP), and albumin. Despite transformation, data outliers and observations with high influence on the stress, injury, and hydration model coefficients remained. The number of observations with leverage values greater than the specified cut-off ranged from five to 13 observations for each of these models. Multiple observations with high leverage indicate limitations of the models to fit the data, possibly due to the sample size not adequately capturing the full range of biological variation observed in these parameters, and therefore should be interpreted with caution.

When the T0 value was controlled for, duration of transport had a significant effect on HCT, CK, AST, cortisol, WBC, and N:L levels measured from the sampled piglets (Table 7). After transport, LD piglets had greater levels of HCT (P = 0.014; Table 7), but lower CK and AST levels (P = 0.006 and 0.0004; Table 7). The SD piglets also had increased cortisol (P = 0.024), WBC (P = 0.043), and N:L (P < 0.0001) levels after transport relative to LD piglets (Table 7).

Table 7.

Effect of duration of transport on indicators of hydration, muscle injury, and physiological stress response

Variable	n	Least squares means estimate (95% CI)1		Fixed effects
				T0 value		Duration
		Short duration	Long duration	F-value	P-value	F-value	P-value
Hematocrit, L/L	1582	0.39 (0.38, 0.41)	0.42 (0.41, 0.43)	871.86	<0.0001	10.55	0.014
Total protein, g/L	160	50.41(48.76, 52.07)	49.30 (47.65, 50.95)	271.01	<0.0001	1.30	0.294
Albumin, g/L	160	36.47 (34.57, 38.36)	36.64 (34.74, 38.53)	594.74	<0.0001	0.02	0.883
Creatine kinase, U/L3	1594	492.85 (405.38, 599.19)	312.39 (256.94, 379.80)	11.62	0.0008	15.62	0.006
Aspartate aminotransferase, U/L2	1594	48.89 (44.20, 54.08)	32.95 (29.79, 36.45)	107.89	<0.0001	44.74	0.0004
Lactate, nmol/L3	1395	2.28 (1.98, 2.62)	2.46 (2.18, 2.78)	3.76	0.055	0.99	0.354
Cortisol, nmol/L3	160	71.47 (51.19.99.78)	40.36 (28.91, 56.35)	3.66	0.058	7.83	0.024
White blood cell count, ×109/L	1592	11.55 (8.59, 14.51)	7.18 (4.23, 10.14)	50.26	<0.0001	6.42	0.043
Neutrophil:lymphocyte ratio3	1592	2.03 (1.69, 2.43)	0.75 (0.63, 0.90)	24.83	<0.0001	79.05	<0.0001

¹Mixed multivariable linear regression models, including T0 (pretransport) value as a covariate and transport event as a random effect.

²Sample results not returned from laboratory analysis.

³Backtransformed estimates and 95% confidence intervals presented.

⁴Values from one piglet excluded from analysis.

⁵Missing samples due to portable analyzer failure.

Few differences among compartments monitored in the LD transport events were detected, with no single compartment consistently demonstrating adverse effects on piglets. Piglets transported in the C-BF had greater HCT levels compared with those transported in the C-UB (C-BF: 0.42 L/L, 95% CI: 0.41 to 0.43 and C-UB: 0.41 L/L, 95% CI: 0.40 to 0.42, P = 0.046). Piglets transported in the C-BF also had greater TP levels compared with those transported in the C-UB (C-BF: 49.64 g/L, 95% CI: 47.97 to 51.31 and C-UB: 48.38 g/L, 95% CI: 46.62 to 50.13, P = 0.031). Piglets from C-B had elevated lactate levels at T1 compared with those in the C-UB (C-B: 2.94 nmol/L, 95% CI: 2.47 to 3.48 and C-UB: 2.13 nmol/L, 95% CI: 1.81 to 2.51, P = 0.003). Piglets transported in C-B also had greater serum cortisol levels at T1 compared with those transported in C-UB (C-B: 54. 24 nmol/L, 95% CI: 36.17 to 81.34 and C-UB: 34.78 nmol/L, 95% CI: 23.31 to 51.90, P = 0.048), with pigs in C-BF being intermediate (37.82 nmol/L, 95% CI: 26.57 to 53.86).

Discussion

The results of this study contribute to the existing knowledge gap on the differences in the physiological response for weaned piglets undergoing short (<3 h), or long (>30 h) transport events under Canadian commercial conditions. This research utilized an observational study design to assess piglets of comparable age and management undergoing SD transport (the same day as weaning) or LD transport (after weaning). Notably, the timeline between weaning and transport was not manipulated or made equivalent between systems. Long duration transport events often require larger loads of piglets to remain economically feasible and it is common to load trailers in the early morning hours to maximize traffic-free travel time and favorable environmental conditions. These factors result in the practices observed in the LD system in this study; with the sow farm weaning piglets in the days leading up to transport and/or stopping at affiliate farm(s) to pick up additional piglets to fill the trailer. While the management differences between duration groups may be seen as a limitation in achieving a counterfactual state, the data collected in this study may be more applicable to the target population (commercial piglets) than examination of the effects of transport duration in isolation under experimental or simulated conditions; a review of the physiological and behavioral response of piglets to the commercial weaning process is described by Sutherland et al. (2014), who emphasize that evaluation of piglet welfare around transport should not be studied in isolation from weaning.

Additionally, while the trailer types included in this study differed between groups, both the LD potbelly trailer and the SD flat deck trailer were equipped with passive ventilation, had bedding amounts provided according to expected weather conditions, and did not have the ability to provide feed or water to piglets during transport, therefore limiting the potential of the trailer design to impact the results of this study. If extrapolating these results to the broader industry, it should be considered that farms were enrolled on a volunteer basis and piglets were purposively chosen during enrolment to represent the average piglet transported. Furthermore, the potential for varying herd health status among transport events to impact the physiological outcomes in this study should be recognized.

In this study, no association was observed between the number of piglets found dead on arrival and the duration of transport. The in-transit mortality level observed (0.06%) was comparable to previously published mortality rates reported for weaned piglets (6.5 kg ± 0.7 kg) from 3,174 US transport records (Zhao et al., 2016). Zhao et al. reported 0.016% ± 0.007% predicted mortality rate for piglets transported in mild ambient conditions (15 to 25 °C), and 0.091% ± 0.04% and 0.026% ± 0.01% for piglets transported in warm/hot (>25 °C) and cold (<15 °C) ambient conditions, respectively.

In addition to the low mortality observed, injury resulting in moderate or severe lameness was not observed in any piglets after transport and is acknowledged as a positive finding supporting welfare. The high prevalence of non-zero lesion scores, particularly for the ear region in LD piglets at T1, may be attributed to the delay between weaning and transport that occurred for the LD piglets rather than the increased transport duration, though it is not possible to verify this based on the study design. The hypothesis that the delay between LD weaning and transport provided an opportunity for postmixing aggression to occur is supported by the increased odds of moderate or severe ear lesions being present for LD piglets compared to SD piglets at T0 as well. Additionally, SD piglet ear lesion severity appeared to be greatest at T2, which may also be explained by post-weaning and mixing-associated aggression after their arrival at the nursery (Mei et al., 2016), as this timepoint followed their first occurrence of mixing in the pen setting after weaning. A similar pattern of results was present for non-zero tail lesion scores observed, though with reduced frequency.

The change in piglet body weight observed in the duration groups in this study is comparable with observations of Wamnes et al. (2006) who reported that weaned piglets lost 7.14% of weaning weight after 24 h of transport. Percent body weight losses for piglets transported for 32 h immediately after weaning have been reported as approaching 9% (Garcia et al., 2015). The smaller losses reported here may be explained by the fact that pretransport weight for LD piglets was not a direct reflection of weaning weight as piglets were weighed one day before transport and up to 5 d postweaning. Piglet feed intake has been observed to increase by 2 d after weaning (McCracken et al., 1995; Bruininix et al., 2001); therefore, LD piglets may have been off feed before the T0 sampling time resulting in relatively smaller losses between T0 and T1 compared with if piglets were weighed after uninterrupted feed consumption immediately at weaning.

While body water losses may account for a component of weight loss over transport for the LD piglets, marked effects of LD transport on hydration was not observed in the results of the physiological parameters included in this study. The HCT levels in LD piglets were significantly greater than SD piglets after transport; however, they were not outside of the normal range (Perri et al., 2017). Therefore, the impact of LD transport on piglet hydration and implication for piglet welfare appears to be low under the observed conditions but would best be supported by future research examining the effects of providing water to piglets during LD transport with the addition of behavioral assessments. Including behavioral assessments is important to further understand how behavioral signs of thirst relate to changes in blood parameters reflecting dehydration, and which is more sensitive for evaluating piglet welfare.

The SD piglets showed evidence of increased injury after transport with greater CK and AST levels compared with the LD group, though not outside of normal reference ranges reported for these parameters (Perri et al., 2017). However, treatment group did not explain variation in lactate levels after transport, although these results should be interpreted with caution as the portable lactate analyzer failed to function at the T1 sampling of the first SD transport event resulting in no values obtained. These results support investigation as to whether or not a recovery period from weaning before short transport would lessen physical exertion or exacerbate the physical demands of mixing, loading, transport and unloading due to decreased energy intake after weaning (Bruininix et al., 2001).

In this study, SD piglets also appeared to be undergoing a greater physiological stress response after transport relative to LD piglets. Serum cortisol and N:L ratios have been used as indicators of stress response in previous piglet transport research (Sutherland et al., 2009; Garcia et al., 2015) and were greater in SD piglets at T1. While it is not possible to separate the contributions of weaning, loading, transport, or unloading on this stress response given the study design, these components are all common in SD transport and therefore it is practical to interpret them together. Although not a direct comparison, in research populations not affected by weaning stress such as the transport of 98 kg fasted market hogs to slaughter, increased cortisol values at exsanguination after short transport (~1 h) compared with medium duration transport (~13 h) has been attributed to a lack of opportunity to recover from stress associated with loading before unloading and mixing (Averos et al., 2007). Similarly, decreased WBC values for pigs (previously weaned and raised to ~19 kg) transported for ~8 h compared with less than 1 h under commercial conditions have been observed (Averós et al., 2009). Garcia et al. (2015) reported no difference in N:L ratios for piglets transported for 32 h without feed or water compared with control piglets that remained with the sows and offered habituation to transport as an explanation which could also explain observations from the present study.

In conclusion, differential changes in weight, injury, and indicators of hydration, muscle injury, and stress were observed for piglets sampled from these two duration groups. Marked physiological changes were not observed for either group nor among the compartments of the LD trailer evaluated. Investigation into the effect of transport duration under environmental conditions not captured in this research would be valuable.

Glossary

Abbreviations

AST

aspartate aminotransferase

CK

creatine kinase

HCT

hematocrit

N:L

neutrophil-to-lymphocyte ratio

RH

relative humidity

TP

total protein

WBC

white blood cell count

Conflict of interest statement

The authors declare no real or perceived conflicts of interest.