TY - JOUR
AU - Brown, F.
AB - Abstract The objective of the present study was to document the relationships between selected welfare outcomes and transport conditions during commercial long haul transport of cattle (≥400 km; 6,152 journeys; 290,866 animals). Surveys were delivered to transport carriers to collect information related to welfare outcomes including the number of dead, non-ambulatory (downer) and lame animals during each journey. Transport conditions surveyed included the length of time animals spent on truck, ambient temperature, animal density, shrinkage, loading time, cattle origin, season, experience of truck drivers, and vehicle characteristics. Overall 0.012% of assessed animals became lame, 0.022% non-ambulatory and 0.011% died onboard. Calves and cull cattle were more likely to die and become non-ambulatory during the journey, feeders intermediate, and fat cattle appeared to be the most able to cope with the stress of transport (P ≤ 0.01). The likelihood of cattle becoming non-ambulatory, lame, or dead increased sharply after animals spent over 30 h on truck (P < 0.001). The likelihood of animal death increased sharply when the midpoint ambient temperature fell below −15°C (P = 0.01) while the likelihood of becoming non-ambulatory increased when temperatures rose above 30°C (P = 0.03). Animals that lost 10% of their BW during transport had a greater (P < 0.001) likelihood of dying and becoming non-ambulatory or lame. Animals were more likely to die at smaller space allowances (P < 0.05), particularly at allometric coefficients below 0.015 (P = 0.10), which occurred more frequently in the belly and deck compartments of the trailers, and also at high space allowances in the deck (allometric coefficients > 0.035). The proportion of total compromised animals decreased with more years of truck driving experience (P < 0.001). Mortality was greater in cattle loaded at auction markets compared with feed yards and ranches (P < 0.01). Cull cattle, calves and feeders appear to be more affected by transport based on the likelihood of becoming non-ambulatory and dying within a journey. Most important welfare concerns during long distance transport include the total journey duration, too low or high space allowances, too high or too low ambient temperature, and the experience of the truck drivers. INTRODUCTION Recently, the transport of farm animals has become a topic of much public debate due to its association with animal welfare, food safety, biosecurity, marketing, trade barriers, and product traceability. Concerns regarding animal welfare are related to the effects of transport on stress due to handling, deprivation of feed and water, and extreme environmental conditions such as animal density, ventilation, temperature and humidity. These factors are known to increase BW loss, as well as the incidence of dead, non-ambulatory, and lame animals during and after transport (Grandin, 2001; González et al., 2012c). Inappropriate transport conditions such as either too low or high animal density as well as poor driving may result in poorer welfare and downgrading of carcasses because of injuries, bruising, and dark cutting (Warriss, 1990; Tarrant et al., 1992). The physical and physiological condition of the animals determines their fitness for transport because it impacts their ability to cope with the stress, which may also be influenced by management before and during transport. Therefore, the type of cattle (i.e., calves, feeders, cull, and fat cattle) being shipped may be affected by transport differently. Previous studies from our group have characterized the most relevant conditions during commercial transport of cattle in North America showing great variability and inter-relationships among factors (González et al., 2012a,b,c). For instance, feeders experienced longer delays within a journey, longer transport time and greater BW loss compared with fat cattle. However, there is a lack of science-based information about the relationship between transport conditions and welfare outcomes in cattle, likely as a result of the large datasets required to study events occurring at low frequencies. The objective of the present study was to assess and quantify the relationships between transport conditions and welfare outcomes measured by the incidence of dead, non-ambulatory, and lame cattle associated with commercial long haul transport. MATERIALS AND METHODS The present study collected information about practices employed during the commercial transport of cattle. Care and handling of animals during transportation was therefore not supervised or controlled by the research team. Description of the Survey Field surveys were designed with the participation of the stakeholders in the industry including beef producers, commercial livestock transport companies, government organizations and researchers throughout the country to collect data regarding the characteristics of cattle transport during long hauls departing from, and arriving to the province of Alberta, Canada. A long haul survey required a transport distance equal to or longer than 400 km between the place of origin where cattle were loaded and the place of final destination where cattle were unloaded from the trailer. Surveys were delivered to commercial transport companies throughout the province, face-to-face training sessions were organized with truck drivers to explain the survey and contact information of the research team was provided to the participants so that questions arising regarding their completion of the survey could be addressed. Truck drivers and companies were given a Humane Handling of Beef Cattle booklet from Alberta Farm Animal Care (AFAC, 2005) during training, which had pictures of cattle and recommendations for transport. Further explanation of survey design, data processing and quality control is provided by González et al. (2012a). Truck drivers completed the surveys during their journeys and the research team collected them from the companies or they were mailed to the Lethbridge Research Centre. Surveys consisted of a set of questions separated into 5 sections which were designed to gather information regarding the type of livestock, driver and equipment, animal loading, conditions during transport, and unloading, respectively. Detailed description of the surveys and calculations are presented elsewhere (González et al., 2012a,b,c). Briefly, the first section requested information regarding the characteristics of the animals being transported such as number and category of animals loaded, and average BW. Cattle category was defined according to BW and premise of origin and destination into fat, feeder, calves, and cull cattle as described in González et al. (2012a). Animals were also classified as male or female if the whole load was comprised of the same sex. However, loads with unknown sex or mixed-sex loads were considered as missing information for analysis of differences among sexes because they occurred in low frequency [1 calf dead, 3 non-ambulatory (1 calf and 2 feeders), and 0 lame at unloading out of 26,567 animals with ‘unknown sex’]. Each transport vehicle was composed of a trailer and tractor combination and information was gathered regarding the dimensions of each compartment within the trailer, the location of the gates and decks, as well as the year, make, configuration, and number of axles of the trailer (González et al., 2012a). The number of animals placed in each compartment was also requested while traveling through both Canada and United States, along with their average BW. Maximum and minimum ambient temperatures within the journey were recorded via temperature data loggers located in the side mirrors of the cabin or in the bumper of the truck. Midpoint temperature during the journey was calculated as the average between maximum and minimum temperatures. Start and end time and date of loading and unloading of the animals was also requested in the surveys. Time on truck was calculated as the length of time cattle spent on the trailer from the start of loading at the place of origin until the end of unloading at the place of destination. The reason and length of delays during transport was documented as either border crossing, waiting to unload animals at destination, driver rest stops, mechanical breakdown, traffic, poor road conditions due to inclement weather or traffic problems, and ‘others’. Loaded (scale) weight of the animals before and after transport was also recorded from ground scales at the feedlot, farm or auction market before loading them onto, or unloading them off the trailer. Alternately, scale weights were determined through the difference between loaded (animals present) minus unloaded (no animals present) weight of the vehicle. Average BW of cattle was calculated as scale weight before transport divided by the number of animals loaded. Body weight loss (shrink), was calculated as [1 – (scale weight after transport / scale weight before transport)] × 100. The last section of the survey collected information filled by truck drivers about the number of animals lame at the time of loading and unloading, as well as the number of non-ambulatory and dead animals, as well as the compartments affected animals were transported in. Non-ambulatory cattle were defined as any animal that could not stand up and walk off the trailer without assistance. The number of lame animals at unloading was calculated as the number of lame animals at unloading minus the number of lame animals upon loading. A new variable (Total Compromised) was created to account for the sum of all dead, non-ambulatory, and lame animals at unloading for each trailer and compartment. Animal density within each compartment was calculated for each trailer using the allometric coefficient (k = space allowance / BW2/3; González et al., 2012b). Statistical Analysis All data were nonparametric and therefore analyzed using mixed-effects logistic regression models in the GLIMMIX procedure (SAS Inst. Inc., Cary, NC). The likelihood of dead, non-ambulatory, lame, and total compromised cattle during transport was analyzed as the number of animals affected (events) over the total number of animals (trials) in the trailer as the dependent variable (e.g., dead animals per trailer/total number of animals per trailer). Therefore, each journey was considered the experimental unit (which is the same as survey or cattle load). Zero values (no affected animals) were assumed if the drivers did not fill out these cells within the survey because that was a common decision made by truck drivers to save time filling up cells that did not apply for the actual journey (e.g., drivers left an empty cell rather than writing down ‘zero’ if there were no dead animals). The best model was chosen based on the generalized χ2 per degree of freedom that yielded the value closest to 1. The trucking company was used as a random effect whereas the fixed effects included either categorical or continuous variables (covariates), or both. Continuous variables were divided into ranges to create ‘dummy’ categorical variables as follows: time on truck (less than 10, 10 to 20, 20 to 30, and more than 30 h), midpoint ambient temperature (less than 0, 0 to 20, and above 20°C), temperature range (0 to 10, 10 to 20, 20 to 30, and above 30°C), shrinkage (0 to 4, 4 to 8, 8 to 12, and above 12% of BW), and the allometric coefficient (0.01 to 0.02, 0.02 to 0.03, and above 0.03). All models containing covariates (continuous variables) were tested for both the linear and quadratic effects. To analyze differences in the likelihood of poor welfare outcomes by trailer compartment, the number of affected animals in a compartment (events) and the total number of animals within that same compartment (trials) were used as the dependent variables. This model was later expanded to include space allowance as an independent variable (calculated for each compartment) in the form of a category (dummy variable) or covariate (continuous), and the interaction with compartment of the trailer. The reason for doing this was that space allowance, as measured by the allometric coefficient k, was different among compartments of the trailers and on either side of the Canada/United States border because of redistribution of animals that is typically carried out at the border to meet the axle weight restrictions of each country (González et al., 2012a,b). This analysis was therefore done with space allowance calculated for both Canada and United States in separate analyses although it was unknown in which country animals became non-ambulatory, dead or lame. Significance was declared at P ≤ 0.05 and tendencies discussed at 0.05 ≤ P ≤ 0.10 unless otherwise noted. RESULTS Approximately 15,000 surveys were distributed with a response rate of approximately 45%. After data quality control and elimination of inaccurate surveys, the final dataset for analysis contained a total of 6,152 surveys from 16 livestock transport companies and a total of 327 tractor-trailer units which relocated 290,866 animals. A total of 32 animals were reported dead on board (0.011%), 65 became non-ambulatory (0.022%) and 34 lame (0.012%), for a sum of 130 total compromised (0.045%) out of 290,866 animals accounted for. In addition, 14 animals were reported as being lame upon loading (0.005%). Type of Cattle The likelihood of dying, becoming lame, or non-ambulatory during transport differed amongst cattle categories (P ≤ 0.01). Cull cattle were at greatest risk of poor welfare during long haul transport because they had the greatest probability of becoming lame at the time of loading and unloading, becoming non-ambulatory or dying during the journey compared with the rest of categories (P ≤ 0.05; Table 1). Calves were more likely to become non-ambulatory and dead compared with fat and feeder cattle (P < 0.05). The only difference between the most frequent cattle categories transported (fats and feeders) was that feeders were twice as likely to die during transport as fat cattle (P < 0.05; Table 1). In addition, male cattle were more likely to be lame and non-ambulatory at unloading compared with females (P = 0.04; Table 2). The origin of the cattle was also found to affect welfare outcomes. For example, cattle loaded at the auction market were more likely to become non-ambulatory (P = 0.06) or die during transport (P < 0.01) compared with those loaded at feed yards (Table 3). Table 1. Likelihood of becoming lame, non-ambulatory, or dead (% of animals transported in trailer) in fat, feeder, calves, and cull cattle during long haul transport (>400 km) Item  Fat  Feeder  Calves  Cull  P-value  Lame at loading  0.0028 ± 0.0012x  0.0015 ± 0.0015x  0.0 ± 0.01xy  0.4108 ± 0.1549y  <0.001  Lame at unloading  0.0132 ±0.0025x  0.0091±0.0037x  0.0093 ±0.0093x  0.1174 ± 0.0829y  0.015  Non-ambulatory  0.0203 ± 0.0031x  0.0091 ± 0.0037x  0.1026 ± 0.0309y  0.2934 ± 0.1310y  <0.001  Dead  0.0066 ± 0.0018x  0.0151 ±0.0048y  0.0653±0.0247z  0.0587 ± 0.0587yz  <0.001  Total compromised 2  0.0396 ± 0.0043x  0.0333 ± 0.0071x  0.1772 ± 0.0406y  0.4695 ± 0.1656z  <0.001  No. of animals  212,308  66,131  10,723  1,704  –  Item  Fat  Feeder  Calves  Cull  P-value  Lame at loading  0.0028 ± 0.0012x  0.0015 ± 0.0015x  0.0 ± 0.01xy  0.4108 ± 0.1549y  <0.001  Lame at unloading  0.0132 ±0.0025x  0.0091±0.0037x  0.0093 ±0.0093x  0.1174 ± 0.0829y  0.015  Non-ambulatory  0.0203 ± 0.0031x  0.0091 ± 0.0037x  0.1026 ± 0.0309y  0.2934 ± 0.1310y  <0.001  Dead  0.0066 ± 0.0018x  0.0151 ±0.0048y  0.0653±0.0247z  0.0587 ± 0.0587yz  <0.001  Total compromised 2  0.0396 ± 0.0043x  0.0333 ± 0.0071x  0.1772 ± 0.0406y  0.4695 ± 0.1656z  <0.001  No. of animals  212,308  66,131  10,723  1,704  –  1No animals were reported for this category. 2Sum of lame at unloading, non-ambulatory and dead animals. x-z Within a row, means without a common superscript differ (P ≤ 0.05). View Large Table 1. Likelihood of becoming lame, non-ambulatory, or dead (% of animals transported in trailer) in fat, feeder, calves, and cull cattle during long haul transport (>400 km) Item  Fat  Feeder  Calves  Cull  P-value  Lame at loading  0.0028 ± 0.0012x  0.0015 ± 0.0015x  0.0 ± 0.01xy  0.4108 ± 0.1549y  <0.001  Lame at unloading  0.0132 ±0.0025x  0.0091±0.0037x  0.0093 ±0.0093x  0.1174 ± 0.0829y  0.015  Non-ambulatory  0.0203 ± 0.0031x  0.0091 ± 0.0037x  0.1026 ± 0.0309y  0.2934 ± 0.1310y  <0.001  Dead  0.0066 ± 0.0018x  0.0151 ±0.0048y  0.0653±0.0247z  0.0587 ± 0.0587yz  <0.001  Total compromised 2  0.0396 ± 0.0043x  0.0333 ± 0.0071x  0.1772 ± 0.0406y  0.4695 ± 0.1656z  <0.001  No. of animals  212,308  66,131  10,723  1,704  –  Item  Fat  Feeder  Calves  Cull  P-value  Lame at loading  0.0028 ± 0.0012x  0.0015 ± 0.0015x  0.0 ± 0.01xy  0.4108 ± 0.1549y  <0.001  Lame at unloading  0.0132 ±0.0025x  0.0091±0.0037x  0.0093 ±0.0093x  0.1174 ± 0.0829y  0.015  Non-ambulatory  0.0203 ± 0.0031x  0.0091 ± 0.0037x  0.1026 ± 0.0309y  0.2934 ± 0.1310y  <0.001  Dead  0.0066 ± 0.0018x  0.0151 ±0.0048y  0.0653±0.0247z  0.0587 ± 0.0587yz  <0.001  Total compromised 2  0.0396 ± 0.0043x  0.0333 ± 0.0071x  0.1772 ± 0.0406y  0.4695 ± 0.1656z  <0.001  No. of animals  212,308  66,131  10,723  1,704  –  1No animals were reported for this category. 2Sum of lame at unloading, non-ambulatory and dead animals. x-z Within a row, means without a common superscript differ (P ≤ 0.05). View Large Table 2. Likelihood of animals becoming lame, non-ambulatory, and dead (% of all animals per trailer) according to sex during long haul transport of cattle (>400 km) Item  Female  Male  P-value  Lame at loading  0.0072 ± 0.0023  0.0032 ± 0.0016  0.16  Lame at unloading  0.0085 ± 0.0039  0.0182 ± 0.0074  0.04  Non-ambulatory  0.0165 ± 0.0076  0.0297 ± 0.0130  0.04  Dead  0.0087 ± 0.0025  0.0150 ± 0.0034  0.14  Total compromised1  0.0357 ± 0.0120  0.0707 ± 0.0224  0.001  No. of animals  138,019  126,762  –  Item  Female  Male  P-value  Lame at loading  0.0072 ± 0.0023  0.0032 ± 0.0016  0.16  Lame at unloading  0.0085 ± 0.0039  0.0182 ± 0.0074  0.04  Non-ambulatory  0.0165 ± 0.0076  0.0297 ± 0.0130  0.04  Dead  0.0087 ± 0.0025  0.0150 ± 0.0034  0.14  Total compromised1  0.0357 ± 0.0120  0.0707 ± 0.0224  0.001  No. of animals  138,019  126,762  –  1Sum of lame at unloading, non-ambulatory and dead animals. View Large Table 2. Likelihood of animals becoming lame, non-ambulatory, and dead (% of all animals per trailer) according to sex during long haul transport of cattle (>400 km) Item  Female  Male  P-value  Lame at loading  0.0072 ± 0.0023  0.0032 ± 0.0016  0.16  Lame at unloading  0.0085 ± 0.0039  0.0182 ± 0.0074  0.04  Non-ambulatory  0.0165 ± 0.0076  0.0297 ± 0.0130  0.04  Dead  0.0087 ± 0.0025  0.0150 ± 0.0034  0.14  Total compromised1  0.0357 ± 0.0120  0.0707 ± 0.0224  0.001  No. of animals  138,019  126,762  –  Item  Female  Male  P-value  Lame at loading  0.0072 ± 0.0023  0.0032 ± 0.0016  0.16  Lame at unloading  0.0085 ± 0.0039  0.0182 ± 0.0074  0.04  Non-ambulatory  0.0165 ± 0.0076  0.0297 ± 0.0130  0.04  Dead  0.0087 ± 0.0025  0.0150 ± 0.0034  0.14  Total compromised1  0.0357 ± 0.0120  0.0707 ± 0.0224  0.001  No. of animals  138,019  126,762  –  1Sum of lame at unloading, non-ambulatory and dead animals. View Large Table 3. Likelihood of becoming lame, non-ambulatory, and dead (% of animals per trailer) during long haul transport (>400 km) according to the origin where loading occurred Item  Feed yard  Farm/Ranch  Auction Market  P-value  Lame at loading  0.0031 ± 0.0023  0.0070 ± 0.0063  0.0097 ± 0.0084  0.36  Lame at unloading  0.0134 ± 0.0054  0.0039 ± 0.0041  0.0073 ± 0.0058  0.37  Non-ambulatory  0.0206 ± 0.0029 x  0.0101 ± 0.0071xy  0.0456 ± 0.0161y  0.06  Dead  0.0101 ± 0.0020 x  0.0 ± 0.01y  0.0399 ± 0.0151z  0.006  Total compromised2  0.0464 ± 0.0143 x  0.0124 ± 0.0080y  0.0693 ± 0.0256x  0.03  No. of animals  247,023  19,827  17,544  –  Item  Feed yard  Farm/Ranch  Auction Market  P-value  Lame at loading  0.0031 ± 0.0023  0.0070 ± 0.0063  0.0097 ± 0.0084  0.36  Lame at unloading  0.0134 ± 0.0054  0.0039 ± 0.0041  0.0073 ± 0.0058  0.37  Non-ambulatory  0.0206 ± 0.0029 x  0.0101 ± 0.0071xy  0.0456 ± 0.0161y  0.06  Dead  0.0101 ± 0.0020 x  0.0 ± 0.01y  0.0399 ± 0.0151z  0.006  Total compromised2  0.0464 ± 0.0143 x  0.0124 ± 0.0080y  0.0693 ± 0.0256x  0.03  No. of animals  247,023  19,827  17,544  –  1No animals were reported for this category. 2Sum of lame at unloading, non-ambulatory and dead animals. x-z Within a row, means without a common superscript differ (P ≤ 0.05). View Large Table 3. Likelihood of becoming lame, non-ambulatory, and dead (% of animals per trailer) during long haul transport (>400 km) according to the origin where loading occurred Item  Feed yard  Farm/Ranch  Auction Market  P-value  Lame at loading  0.0031 ± 0.0023  0.0070 ± 0.0063  0.0097 ± 0.0084  0.36  Lame at unloading  0.0134 ± 0.0054  0.0039 ± 0.0041  0.0073 ± 0.0058  0.37  Non-ambulatory  0.0206 ± 0.0029 x  0.0101 ± 0.0071xy  0.0456 ± 0.0161y  0.06  Dead  0.0101 ± 0.0020 x  0.0 ± 0.01y  0.0399 ± 0.0151z  0.006  Total compromised2  0.0464 ± 0.0143 x  0.0124 ± 0.0080y  0.0693 ± 0.0256x  0.03  No. of animals  247,023  19,827  17,544  –  Item  Feed yard  Farm/Ranch  Auction Market  P-value  Lame at loading  0.0031 ± 0.0023  0.0070 ± 0.0063  0.0097 ± 0.0084  0.36  Lame at unloading  0.0134 ± 0.0054  0.0039 ± 0.0041  0.0073 ± 0.0058  0.37  Non-ambulatory  0.0206 ± 0.0029 x  0.0101 ± 0.0071xy  0.0456 ± 0.0161y  0.06  Dead  0.0101 ± 0.0020 x  0.0 ± 0.01y  0.0399 ± 0.0151z  0.006  Total compromised2  0.0464 ± 0.0143 x  0.0124 ± 0.0080y  0.0693 ± 0.0256x  0.03  No. of animals  247,023  19,827  17,544  –  1No animals were reported for this category. 2Sum of lame at unloading, non-ambulatory and dead animals. x-z Within a row, means without a common superscript differ (P ≤ 0.05). View Large Transport Duration The longer the time animals spent on the truck the greater the likelihood of becoming lame, non-ambulatory or dead on board (P < 0.001; Table 4). Interestingly, the likelihood of animals affected increased sharply after 30 h spent on the truck (Figure 1A). Table 4. Proportion of animals lame at unloading, non-ambulatory and dead observed during commercial long haul transport of cattle (>400 km) Item  Lame at unloading  Non-ambulatory  Dead  Total compromised1  Time on truck, h      < 10  0.0029 ± 0.003 x  0.0124 ± 0.0078x  0.0082 ± 0.0058x  0.0228 ± 0.0107x      10 to 20  0.0102 ± 0.0047x  0.0242 ± 0.0111x  0.0035 ± 0.0016x  0.0438 ± 0.0143x      20 to 30  0.0274 ± 0.0147y  0.0027 ± 0.0029y  0.0179 ± 0.0073y  0.0485 ± 0.0198x      > 30  0.1236 ± 0.0622z  0.0891 ± 0.0511z  0.0554 ± 0.0226z  0.3042 ± 0.1129y      P2  < 0.001  < 0.001  < 0.001  < 0.001      Lin2  < 0.001  0.04  < 0.001  < 0.001      Q2  0.67  0.78  0.59  0.81  Midpoint temperature, °C      < 0  0.0149 ± 0.0081xy  0.0132 ± 0.0074x  0.0160 ± 0.0060x  0.0463 ± 0.0166x      0 to 20  0.0067 ± 0.0033x  0.0120 ± 0.0051x  0.0083 ± 0.0020xy  0.0269 ± 0.0087y      > 20  0.0289 ± 0.0132y  0.0372 ± 0.0155y  0.0020 ± 0.0020y  0.0704 ± 0.0228x      P2  0.001  0.001  0.10  < 0.001      Lin2  0.09  0.03  0.009  0.26      Q2  0.13  0.13  0.83  0.01  Temperature range, °C      0 to 10  0.0154 ± 0.0049x  0.0141 ± 0.0067y  0.0093 ± 0.0038  0.0388 ± 0.0129xy      10 to 20  0.0126 ± 0.0031x  0.0128 ± 0.0058y  0.0071 ± 0.0024  0.0329 ± 0.0107y      20 to 30  0.0185 ± 0.0066x  0.0416 ± 0.0190x  0.0093 ± 0.0046  0.0668 ± 0.0233x      > 30  0.0 ± 0.0y3  0.0346 ± 0.0370xy  0.0299 ± 0.0299  0.0681 ± 0.0520xy      P2  < 0.001  0.009  0.58  0.04      Lin2  0.37  0.04  0.10  0.04      Q2  0.95  0.93  0.04  0.35  BW loss, %      0 to 4  0.0061 ± 0.0043  0.0 ± 0.03  0.0 ± 0.03x  0.0064 ± 0.0052x      4 to 8  0.0114 ± 0.0028  0.0248 ± 0.004  0.0040 ± 0.0016y  0.0365 ± 0.0151y      8 to 12  0.0206 ± 0.0103  0.0358 ± 0.0135  0.0307 ± 0.0125z  0.0808 ± 0.0393z      + 12  0.0 ± 0.03  0.0 ± 0.03  0.0508 ± 0.0508z  0.0504 ±0.0543xyz      P2  0.55  0.84  0.001  0.003      Lin2  0.09  <0.001  <0.001  <0.001      Q2  0.11  –4  0.051  –4  Item  Lame at unloading  Non-ambulatory  Dead  Total compromised1  Time on truck, h      < 10  0.0029 ± 0.003 x  0.0124 ± 0.0078x  0.0082 ± 0.0058x  0.0228 ± 0.0107x      10 to 20  0.0102 ± 0.0047x  0.0242 ± 0.0111x  0.0035 ± 0.0016x  0.0438 ± 0.0143x      20 to 30  0.0274 ± 0.0147y  0.0027 ± 0.0029y  0.0179 ± 0.0073y  0.0485 ± 0.0198x      > 30  0.1236 ± 0.0622z  0.0891 ± 0.0511z  0.0554 ± 0.0226z  0.3042 ± 0.1129y      P2  < 0.001  < 0.001  < 0.001  < 0.001      Lin2  < 0.001  0.04  < 0.001  < 0.001      Q2  0.67  0.78  0.59  0.81  Midpoint temperature, °C      < 0  0.0149 ± 0.0081xy  0.0132 ± 0.0074x  0.0160 ± 0.0060x  0.0463 ± 0.0166x      0 to 20  0.0067 ± 0.0033x  0.0120 ± 0.0051x  0.0083 ± 0.0020xy  0.0269 ± 0.0087y      > 20  0.0289 ± 0.0132y  0.0372 ± 0.0155y  0.0020 ± 0.0020y  0.0704 ± 0.0228x      P2  0.001  0.001  0.10  < 0.001      Lin2  0.09  0.03  0.009  0.26      Q2  0.13  0.13  0.83  0.01  Temperature range, °C      0 to 10  0.0154 ± 0.0049x  0.0141 ± 0.0067y  0.0093 ± 0.0038  0.0388 ± 0.0129xy      10 to 20  0.0126 ± 0.0031x  0.0128 ± 0.0058y  0.0071 ± 0.0024  0.0329 ± 0.0107y      20 to 30  0.0185 ± 0.0066x  0.0416 ± 0.0190x  0.0093 ± 0.0046  0.0668 ± 0.0233x      > 30  0.0 ± 0.0y3  0.0346 ± 0.0370xy  0.0299 ± 0.0299  0.0681 ± 0.0520xy      P2  < 0.001  0.009  0.58  0.04      Lin2  0.37  0.04  0.10  0.04      Q2  0.95  0.93  0.04  0.35  BW loss, %      0 to 4  0.0061 ± 0.0043  0.0 ± 0.03  0.0 ± 0.03x  0.0064 ± 0.0052x      4 to 8  0.0114 ± 0.0028  0.0248 ± 0.004  0.0040 ± 0.0016y  0.0365 ± 0.0151y      8 to 12  0.0206 ± 0.0103  0.0358 ± 0.0135  0.0307 ± 0.0125z  0.0808 ± 0.0393z      + 12  0.0 ± 0.03  0.0 ± 0.03  0.0508 ± 0.0508z  0.0504 ±0.0543xyz      P2  0.55  0.84  0.001  0.003      Lin2  0.09  <0.001  <0.001  <0.001      Q2  0.11  –4  0.051  –4  1Sum of dead, non-ambulatory, and lame at unloading. 2P: P-value of the main effect in the categorical model; Lin: P-value of the linear effect in the continuous model; Q: P-value of the quadratic effect in the continuous model 3No animals were reported for this category. 4Procedure failed to converge. x-zWithin a column and item, means without a common superscript differ (P < 0.05). View Large Table 4. Proportion of animals lame at unloading, non-ambulatory and dead observed during commercial long haul transport of cattle (>400 km) Item  Lame at unloading  Non-ambulatory  Dead  Total compromised1  Time on truck, h      < 10  0.0029 ± 0.003 x  0.0124 ± 0.0078x  0.0082 ± 0.0058x  0.0228 ± 0.0107x      10 to 20  0.0102 ± 0.0047x  0.0242 ± 0.0111x  0.0035 ± 0.0016x  0.0438 ± 0.0143x      20 to 30  0.0274 ± 0.0147y  0.0027 ± 0.0029y  0.0179 ± 0.0073y  0.0485 ± 0.0198x      > 30  0.1236 ± 0.0622z  0.0891 ± 0.0511z  0.0554 ± 0.0226z  0.3042 ± 0.1129y      P2  < 0.001  < 0.001  < 0.001  < 0.001      Lin2  < 0.001  0.04  < 0.001  < 0.001      Q2  0.67  0.78  0.59  0.81  Midpoint temperature, °C      < 0  0.0149 ± 0.0081xy  0.0132 ± 0.0074x  0.0160 ± 0.0060x  0.0463 ± 0.0166x      0 to 20  0.0067 ± 0.0033x  0.0120 ± 0.0051x  0.0083 ± 0.0020xy  0.0269 ± 0.0087y      > 20  0.0289 ± 0.0132y  0.0372 ± 0.0155y  0.0020 ± 0.0020y  0.0704 ± 0.0228x      P2  0.001  0.001  0.10  < 0.001      Lin2  0.09  0.03  0.009  0.26      Q2  0.13  0.13  0.83  0.01  Temperature range, °C      0 to 10  0.0154 ± 0.0049x  0.0141 ± 0.0067y  0.0093 ± 0.0038  0.0388 ± 0.0129xy      10 to 20  0.0126 ± 0.0031x  0.0128 ± 0.0058y  0.0071 ± 0.0024  0.0329 ± 0.0107y      20 to 30  0.0185 ± 0.0066x  0.0416 ± 0.0190x  0.0093 ± 0.0046  0.0668 ± 0.0233x      > 30  0.0 ± 0.0y3  0.0346 ± 0.0370xy  0.0299 ± 0.0299  0.0681 ± 0.0520xy      P2  < 0.001  0.009  0.58  0.04      Lin2  0.37  0.04  0.10  0.04      Q2  0.95  0.93  0.04  0.35  BW loss, %      0 to 4  0.0061 ± 0.0043  0.0 ± 0.03  0.0 ± 0.03x  0.0064 ± 0.0052x      4 to 8  0.0114 ± 0.0028  0.0248 ± 0.004  0.0040 ± 0.0016y  0.0365 ± 0.0151y      8 to 12  0.0206 ± 0.0103  0.0358 ± 0.0135  0.0307 ± 0.0125z  0.0808 ± 0.0393z      + 12  0.0 ± 0.03  0.0 ± 0.03  0.0508 ± 0.0508z  0.0504 ±0.0543xyz      P2  0.55  0.84  0.001  0.003      Lin2  0.09  <0.001  <0.001  <0.001      Q2  0.11  –4  0.051  –4  Item  Lame at unloading  Non-ambulatory  Dead  Total compromised1  Time on truck, h      < 10  0.0029 ± 0.003 x  0.0124 ± 0.0078x  0.0082 ± 0.0058x  0.0228 ± 0.0107x      10 to 20  0.0102 ± 0.0047x  0.0242 ± 0.0111x  0.0035 ± 0.0016x  0.0438 ± 0.0143x      20 to 30  0.0274 ± 0.0147y  0.0027 ± 0.0029y  0.0179 ± 0.0073y  0.0485 ± 0.0198x      > 30  0.1236 ± 0.0622z  0.0891 ± 0.0511z  0.0554 ± 0.0226z  0.3042 ± 0.1129y      P2  < 0.001  < 0.001  < 0.001  < 0.001      Lin2  < 0.001  0.04  < 0.001  < 0.001      Q2  0.67  0.78  0.59  0.81  Midpoint temperature, °C      < 0  0.0149 ± 0.0081xy  0.0132 ± 0.0074x  0.0160 ± 0.0060x  0.0463 ± 0.0166x      0 to 20  0.0067 ± 0.0033x  0.0120 ± 0.0051x  0.0083 ± 0.0020xy  0.0269 ± 0.0087y      > 20  0.0289 ± 0.0132y  0.0372 ± 0.0155y  0.0020 ± 0.0020y  0.0704 ± 0.0228x      P2  0.001  0.001  0.10  < 0.001      Lin2  0.09  0.03  0.009  0.26      Q2  0.13  0.13  0.83  0.01  Temperature range, °C      0 to 10  0.0154 ± 0.0049x  0.0141 ± 0.0067y  0.0093 ± 0.0038  0.0388 ± 0.0129xy      10 to 20  0.0126 ± 0.0031x  0.0128 ± 0.0058y  0.0071 ± 0.0024  0.0329 ± 0.0107y      20 to 30  0.0185 ± 0.0066x  0.0416 ± 0.0190x  0.0093 ± 0.0046  0.0668 ± 0.0233x      > 30  0.0 ± 0.0y3  0.0346 ± 0.0370xy  0.0299 ± 0.0299  0.0681 ± 0.0520xy      P2  < 0.001  0.009  0.58  0.04      Lin2  0.37  0.04  0.10  0.04      Q2  0.95  0.93  0.04  0.35  BW loss, %      0 to 4  0.0061 ± 0.0043  0.0 ± 0.03  0.0 ± 0.03x  0.0064 ± 0.0052x      4 to 8  0.0114 ± 0.0028  0.0248 ± 0.004  0.0040 ± 0.0016y  0.0365 ± 0.0151y      8 to 12  0.0206 ± 0.0103  0.0358 ± 0.0135  0.0307 ± 0.0125z  0.0808 ± 0.0393z      + 12  0.0 ± 0.03  0.0 ± 0.03  0.0508 ± 0.0508z  0.0504 ±0.0543xyz      P2  0.55  0.84  0.001  0.003      Lin2  0.09  <0.001  <0.001  <0.001      Q2  0.11  –4  0.051  –4  1Sum of dead, non-ambulatory, and lame at unloading. 2P: P-value of the main effect in the categorical model; Lin: P-value of the linear effect in the continuous model; Q: P-value of the quadratic effect in the continuous model 3No animals were reported for this category. 4Procedure failed to converge. x-zWithin a column and item, means without a common superscript differ (P < 0.05). View Large Figure 1. View largeDownload slide Effect of time cattle spent on truck (A), midpoint (mean) temperature (B), temperature range (C), and the extent of BW loss (D) on the likelihood of becoming lame, non-ambulatory (downer) and dead during long haul transport (>400 km). Total was the sum of lame, non-ambulatory (Downer) and dead animals during the journeys. Figure 1. View largeDownload slide Effect of time cattle spent on truck (A), midpoint (mean) temperature (B), temperature range (C), and the extent of BW loss (D) on the likelihood of becoming lame, non-ambulatory (downer) and dead during long haul transport (>400 km). Total was the sum of lame, non-ambulatory (Downer) and dead animals during the journeys. Ambient Temperature, Season and Loading Time Cattle were more likely to become lame and non-ambulatory at midpoint ambient temperatures above 20°C (Table 4; P < 0.05). In contrast, cattle were more likely to die at lower temperatures (linear P = 0.009), mortality being greater when the midpoint temperature was below zero compared with temperatures above 20°C (P = 0.05; Table 4). However, inspection of the fitted likelihoods indicated that mortality increased sharply when temperature was below −15°C (Figure 1B). As a result of a greater incidence of cattle becoming non-ambulatory at high temperature and greater mortality at low midpoint temperature, the incidence of total compromised animals was least at moderate temperatures between −15 and 30°C (quadratic P = 0.01; Table 4 and Figure 1B). Temperature range within a journey also affected cattle welfare outcomes during transport (Table 4). Cattle were more likely to become non-ambulatory with increasing temperature range above 20°C (linear P = 0.04) whereas mortality increased sharply with temperature ranges above 35°C (linear P = 0.04; Figure 1C). In agreement with findings from the effects of midpoint ambient temperature, the probability of animals becoming non-ambulatory was greatest in the summer (P ≤ 0.05; Table 5) when 81.5% of all animals transported were fat cattle (data not shown). The probability of death during transport was greater in fall compared with summer (P ≤ 0.05; Table 5) when a greater number of feeder cattle were transported (32% of all cattle; data not shown). Table 5. Proportion of lame, non-ambulatory, and dead cattle (% of animals per trailer) in summer, fall, winter, and spring during long haul transport (>400 km) Item  Summer  Fall  Winter  Spring  P-value  Lame at loading  0.0020 ± 0.0014  0.0082 ± 0.0027  0.0061 ± 0.0035  0.0 ± 0.01  0.20  Lame at unloading  0.0108 ± 0.0044  0.0122 ± 0.0049  0.0057 ± 0.0037  0.0196 ± 0.0099  0.38  Non-ambulatory  0.0318 ± 0.013x  0.0174 ± 0.0076y  0.0090 ± 0.0056y  0.0045 ± 0.0036y  0.002  Dead  0.0040 ± 0.0020x  0.0192 ± 0.0042y  0.0101 ±0.0045xy  0.0068 ±0.0048xy  0.02  Total compromised2  0.0498 ± 0.0147x  0.0543 ± 0.0160x  0.0257 ± 0.0098y  0.0289 ±0.0120xy  0.04  No. animals  99,105  109,390  49,290  29,606  –3  Item  Summer  Fall  Winter  Spring  P-value  Lame at loading  0.0020 ± 0.0014  0.0082 ± 0.0027  0.0061 ± 0.0035  0.0 ± 0.01  0.20  Lame at unloading  0.0108 ± 0.0044  0.0122 ± 0.0049  0.0057 ± 0.0037  0.0196 ± 0.0099  0.38  Non-ambulatory  0.0318 ± 0.013x  0.0174 ± 0.0076y  0.0090 ± 0.0056y  0.0045 ± 0.0036y  0.002  Dead  0.0040 ± 0.0020x  0.0192 ± 0.0042y  0.0101 ±0.0045xy  0.0068 ±0.0048xy  0.02  Total compromised2  0.0498 ± 0.0147x  0.0543 ± 0.0160x  0.0257 ± 0.0098y  0.0289 ±0.0120xy  0.04  No. animals  99,105  109,390  49,290  29,606  –3  1No animals were reported for this category. 2Sum of dead, non-ambulatory, and lame at unloading. 3No statistical analysis was done. x,yWithin a row, means without a common superscript differ (P ≤ 0.05). View Large Table 5. Proportion of lame, non-ambulatory, and dead cattle (% of animals per trailer) in summer, fall, winter, and spring during long haul transport (>400 km) Item  Summer  Fall  Winter  Spring  P-value  Lame at loading  0.0020 ± 0.0014  0.0082 ± 0.0027  0.0061 ± 0.0035  0.0 ± 0.01  0.20  Lame at unloading  0.0108 ± 0.0044  0.0122 ± 0.0049  0.0057 ± 0.0037  0.0196 ± 0.0099  0.38  Non-ambulatory  0.0318 ± 0.013x  0.0174 ± 0.0076y  0.0090 ± 0.0056y  0.0045 ± 0.0036y  0.002  Dead  0.0040 ± 0.0020x  0.0192 ± 0.0042y  0.0101 ±0.0045xy  0.0068 ±0.0048xy  0.02  Total compromised2  0.0498 ± 0.0147x  0.0543 ± 0.0160x  0.0257 ± 0.0098y  0.0289 ±0.0120xy  0.04  No. animals  99,105  109,390  49,290  29,606  –3  Item  Summer  Fall  Winter  Spring  P-value  Lame at loading  0.0020 ± 0.0014  0.0082 ± 0.0027  0.0061 ± 0.0035  0.0 ± 0.01  0.20  Lame at unloading  0.0108 ± 0.0044  0.0122 ± 0.0049  0.0057 ± 0.0037  0.0196 ± 0.0099  0.38  Non-ambulatory  0.0318 ± 0.013x  0.0174 ± 0.0076y  0.0090 ± 0.0056y  0.0045 ± 0.0036y  0.002  Dead  0.0040 ± 0.0020x  0.0192 ± 0.0042y  0.0101 ±0.0045xy  0.0068 ±0.0048xy  0.02  Total compromised2  0.0498 ± 0.0147x  0.0543 ± 0.0160x  0.0257 ± 0.0098y  0.0289 ±0.0120xy  0.04  No. animals  99,105  109,390  49,290  29,606  –3  1No animals were reported for this category. 2Sum of dead, non-ambulatory, and lame at unloading. 3No statistical analysis was done. x,yWithin a row, means without a common superscript differ (P ≤ 0.05). View Large Shrink Animals in loads that lost more than 8% of loaded weight during transport were more likely to die compared with those that shrank 8% or less (P < 0.05; Table 4). Despite the fact that the proportion of lame and non-ambulatory cattle was not related to shrinkage in the categorical model, the linear model indicated that the likelihood of becoming lame (linear P = 0.09) and non-ambulatory (linear P < 0.001) increased with shrinkage (Table 4). Inspection of the fitted likelihoods vs. shrink shown in Figure 2A indicates that the probability of becoming totally compromised (dead, non-ambulatory and lame) during transport increased most sharply above 10% of BW lost. Figure 2. View largeDownload slide Relationship between space allowance within Canada with the likelihood of becoming lame, non-ambulatory (Downer) and dead during long haul transport of cattle (>400 km). Space allowance was measured through the allometric coefficient (k) calculated for each compartment from the formulae k = space allowance (m2/animal) / BW 0.667.4 Figure 2. View largeDownload slide Relationship between space allowance within Canada with the likelihood of becoming lame, non-ambulatory (Downer) and dead during long haul transport of cattle (>400 km). Space allowance was measured through the allometric coefficient (k) calculated for each compartment from the formulae k = space allowance (m2/animal) / BW 0.667.4 Compartment of the Trailer There were no differences among compartments in the probability of animals becoming lame or non-ambulatory onboard (Table 6) although the compartment where lame animals came from was known for only 15 out of 34 cases. However, animals transported in the nose were less likely to die (P < 0.05) and the compartment in which dead animals were found was known for 30 out of the total 32 dead animals and for 60 out of the total 65 non-ambulatory animals. Table 6. Proportion of lame, non-ambulatory, and dead cattle (% of total animals transported in each compartment) reported during long haul transport (≥400 km) Item  Lame at unloading  Non-ambulatory  Dead  Total compromised 1  Nose  0.0038 ± 0.0038  0.0191 ± 0.0085  0.0 ± 0.02x  0.0229 ± 0.0093  Deck  0.0062 ± 0.0028  0.0296 ± 0.0060  0.0087 ± 0.0033y  0.0445 ± 0.0074  Belly  0.0057 ± 0.0026  0.0160 ± 0.0043  0.0162 ± 0.0043y  0.0378 ± 0.0066  Back  0.0036 ± 0.0025  0.0196 ± 0.0059  0.0125 ± 0.0047y  0.0356 ± 0.0080  Doghouse  0.0091 ± 0.0064  0.0135 ± 0.0078  0.0046 ± 0.0046xy  0.0271 ± 0.0111  P-value  0.90  0.34  < 0.001  0.52  Item  Lame at unloading  Non-ambulatory  Dead  Total compromised 1  Nose  0.0038 ± 0.0038  0.0191 ± 0.0085  0.0 ± 0.02x  0.0229 ± 0.0093  Deck  0.0062 ± 0.0028  0.0296 ± 0.0060  0.0087 ± 0.0033y  0.0445 ± 0.0074  Belly  0.0057 ± 0.0026  0.0160 ± 0.0043  0.0162 ± 0.0043y  0.0378 ± 0.0066  Back  0.0036 ± 0.0025  0.0196 ± 0.0059  0.0125 ± 0.0047y  0.0356 ± 0.0080  Doghouse  0.0091 ± 0.0064  0.0135 ± 0.0078  0.0046 ± 0.0046xy  0.0271 ± 0.0111  P-value  0.90  0.34  < 0.001  0.52  1Sum of dead, non-ambulatory, and lame at unloading. 2No animals were reported for this category. w,yWithin a column, means without a common superscript differ (P ≤ 0.05). View Large Table 6. Proportion of lame, non-ambulatory, and dead cattle (% of total animals transported in each compartment) reported during long haul transport (≥400 km) Item  Lame at unloading  Non-ambulatory  Dead  Total compromised 1  Nose  0.0038 ± 0.0038  0.0191 ± 0.0085  0.0 ± 0.02x  0.0229 ± 0.0093  Deck  0.0062 ± 0.0028  0.0296 ± 0.0060  0.0087 ± 0.0033y  0.0445 ± 0.0074  Belly  0.0057 ± 0.0026  0.0160 ± 0.0043  0.0162 ± 0.0043y  0.0378 ± 0.0066  Back  0.0036 ± 0.0025  0.0196 ± 0.0059  0.0125 ± 0.0047y  0.0356 ± 0.0080  Doghouse  0.0091 ± 0.0064  0.0135 ± 0.0078  0.0046 ± 0.0046xy  0.0271 ± 0.0111  P-value  0.90  0.34  < 0.001  0.52  Item  Lame at unloading  Non-ambulatory  Dead  Total compromised 1  Nose  0.0038 ± 0.0038  0.0191 ± 0.0085  0.0 ± 0.02x  0.0229 ± 0.0093  Deck  0.0062 ± 0.0028  0.0296 ± 0.0060  0.0087 ± 0.0033y  0.0445 ± 0.0074  Belly  0.0057 ± 0.0026  0.0160 ± 0.0043  0.0162 ± 0.0043y  0.0378 ± 0.0066  Back  0.0036 ± 0.0025  0.0196 ± 0.0059  0.0125 ± 0.0047y  0.0356 ± 0.0080  Doghouse  0.0091 ± 0.0064  0.0135 ± 0.0078  0.0046 ± 0.0046xy  0.0271 ± 0.0111  P-value  0.90  0.34  < 0.001  0.52  1Sum of dead, non-ambulatory, and lame at unloading. 2No animals were reported for this category. w,yWithin a column, means without a common superscript differ (P ≤ 0.05). View Large Space Allowance Space allowance was also related to animal welfare outcomes but only in larger compartments [i.e. in the deck and belly (compartment × k-value interaction P < 0.001)] because mortality within Canada was greater in the deck and belly when transported at low space allowance (linear P = 0.007) and also at high space allowance in the deck (k > 0.035; quadratic P < 0.001; Figure 2). Animals in the deck also showed a sharp increase in the probability of becoming non-ambulatory at k-values below 0.015 (linear P = 0.003) but this was not the case for the other compartments (P > 0.10; data not shown). Experience of Truck Drivers The years of experience of a driver hauling cattle affected all animal welfare outcomes with the exception of mortality rate (Table 7). The probability of animals becoming non-ambulatory or totally compromised during transport was greater in drivers with 0 to 5 yr compared with 6 or more years of cattle hauling experience (P < 0.01). The proportion of totally compromised animals decreased 6-fold as the experience of truck drivers increased from <2 to >10 yr of cattle hauling experience (Table 7). Table 7. Probability of animals to be lame, non-ambulatory, and dead (% of animals) according to the experience of drivers hauling cattle during long haul transport (>400 km)   Experience of truck drivers, yr    Item  0 to 2  3 to 5  6 to 10  > 10  P-value  Lame at loading  0.0024 ± 0.0017x  0.0218 ± 0.0069y  0.0 ± 0.01z  0.0020 ± 0.0014x  <0.001  Lame at unloading  0.0178 ±0.0075xy  0.0362 ±0.0138x  0.0063± 0.0041yz  0.0 ± 0.02z  <0.001  Non-ambulatory  0.0389 ± 0.0068x  0.0371 ± 0.0090x  0.0062 ± 0.0036y  0.0080 ± 0.0028y  <0.001  Dead  0.0130 ± 0.0039  0.0109 ± 0.0049  0.0124 ± 0.0051  0.0070 ± 0.0026  0.61  Total compromised 2  0.0932 ± 0.0300x  0.0865 ± 0.0283x  0.0303 ± 0.0122y  0.0157 ± 0.0059y  <0.001  No. of animals  84,822  45,817  48,334  100,610  –    Experience of truck drivers, yr    Item  0 to 2  3 to 5  6 to 10  > 10  P-value  Lame at loading  0.0024 ± 0.0017x  0.0218 ± 0.0069y  0.0 ± 0.01z  0.0020 ± 0.0014x  <0.001  Lame at unloading  0.0178 ±0.0075xy  0.0362 ±0.0138x  0.0063± 0.0041yz  0.0 ± 0.02z  <0.001  Non-ambulatory  0.0389 ± 0.0068x  0.0371 ± 0.0090x  0.0062 ± 0.0036y  0.0080 ± 0.0028y  <0.001  Dead  0.0130 ± 0.0039  0.0109 ± 0.0049  0.0124 ± 0.0051  0.0070 ± 0.0026  0.61  Total compromised 2  0.0932 ± 0.0300x  0.0865 ± 0.0283x  0.0303 ± 0.0122y  0.0157 ± 0.0059y  <0.001  No. of animals  84,822  45,817  48,334  100,610  –  1No animals were reported for this category. 2Sum of lame at unloading, non-ambulatory and dead animals. x-zWithin a row, means without a common superscript differ (P ≤ 0.05). View Large Table 7. Probability of animals to be lame, non-ambulatory, and dead (% of animals) according to the experience of drivers hauling cattle during long haul transport (>400 km)   Experience of truck drivers, yr    Item  0 to 2  3 to 5  6 to 10  > 10  P-value  Lame at loading  0.0024 ± 0.0017x  0.0218 ± 0.0069y  0.0 ± 0.01z  0.0020 ± 0.0014x  <0.001  Lame at unloading  0.0178 ±0.0075xy  0.0362 ±0.0138x  0.0063± 0.0041yz  0.0 ± 0.02z  <0.001  Non-ambulatory  0.0389 ± 0.0068x  0.0371 ± 0.0090x  0.0062 ± 0.0036y  0.0080 ± 0.0028y  <0.001  Dead  0.0130 ± 0.0039  0.0109 ± 0.0049  0.0124 ± 0.0051  0.0070 ± 0.0026  0.61  Total compromised 2  0.0932 ± 0.0300x  0.0865 ± 0.0283x  0.0303 ± 0.0122y  0.0157 ± 0.0059y  <0.001  No. of animals  84,822  45,817  48,334  100,610  –    Experience of truck drivers, yr    Item  0 to 2  3 to 5  6 to 10  > 10  P-value  Lame at loading  0.0024 ± 0.0017x  0.0218 ± 0.0069y  0.0 ± 0.01z  0.0020 ± 0.0014x  <0.001  Lame at unloading  0.0178 ±0.0075xy  0.0362 ±0.0138x  0.0063± 0.0041yz  0.0 ± 0.02z  <0.001  Non-ambulatory  0.0389 ± 0.0068x  0.0371 ± 0.0090x  0.0062 ± 0.0036y  0.0080 ± 0.0028y  <0.001  Dead  0.0130 ± 0.0039  0.0109 ± 0.0049  0.0124 ± 0.0051  0.0070 ± 0.0026  0.61  Total compromised 2  0.0932 ± 0.0300x  0.0865 ± 0.0283x  0.0303 ± 0.0122y  0.0157 ± 0.0059y  <0.001  No. of animals  84,822  45,817  48,334  100,610  –  1No animals were reported for this category. 2Sum of lame at unloading, non-ambulatory and dead animals. x-zWithin a row, means without a common superscript differ (P ≤ 0.05). View Large DISCUSSION Despite the fact that significant relationships were found between transport conditions and the probability of animals dying or becoming non-ambulatory, caution should be taken when interpreting results from the present study. The significant relationships among transport conditions and welfare outcomes are due to the large dataset used for analysis. However, the practical importance might be questionable because of the very low occurrence of dead, non-ambulatory and lame animals. The strength of such measures of welfare lies in the fact that they occur at the more extreme ends of poor welfare. Descriptive Data Previous publications from the survey linked to the data presented herein characterized the conditions during commercial long haul transport of cattle departing from or arriving to the province of Alberta, Canada (González et al., 2012a). Commercial space allowances (González et al., 2012b), and differences in transport conditions and shrinkage among cattle categories were presented elsewhere (González et al., 2012c). These studies demonstrated a wide range of conditions and management during commercial transport of cattle in North America. For instance, cattle were transported at ambient temperatures as low as −42°C and as high as 45°C, and temperature ranges (maximum minus minimum) within the same journey was up to 46°C (González et al., 2012a). Cattle experienced delays of up to 20 h within a single journey most frequently as a result of border inspection, driver rest stops, and waiting to unload cattle at their destination. Cattle were transported for up to 2,560 km and remained up to 45 h on truck without access to feed and water. Average and maximum shrinkage was 5.3 and 21.8% of BW, respectively. It is important however to point out that extreme values observed occurred at very low frequencies (González et al., 2012a–c). The allometric coefficient ranged between 0.0109 and 0.157 according to the number of axles of the vehicle, compartment within the trailer, and cattle category with both over and under stocking (González et al., 2012b). Significant differences among cattle categories were also observed in other important aspects related to animal welfare. For example, feeder cattle experienced the longest delays within the journey and spent the longest time on truck compared with other cattle categories. Feeders also experienced greater shrinkage (González et al., 2012c) and were transported more densely compared with fat cattle (González et al., 2012b). The present study assessed the way those values and practices affected cattle welfare outcomes during commercial transport through the incidence of animal mortality, and occurrence of non-ambulatory, and lame cattle. Mortality is the final result of complete failure to cope with the stress and environmental conditions during transport and therefore indicative of suffering and extremely poor welfare. Non-ambulatory cattle may be subjected to a life threatening situation due to trampling by other animals. Animals that are lame at the time of loading may experience additional stress over the course of the journey as a result of the physical activity required to maintain balance and remain standing. Mature cattle (cull and fat) may suffer more if they were lame at the time of loading because older animals stand throughout the journey in contrast to younger stock (Kent and Ewbank, 1986). However, the degree of stress depends on the severity of lameness at the time of loading and the length of time animals have to cope with the condition. However, the severity and time of occurrence of lameness were not measured in the present study and results on the incidence of lame animals upon unloading should therefore be viewed with caution. Type of Cattle The present study found that cull cattle were the most likely to experience stress during transport due to the greater proportion of totally compromised and lame cattle identified at the time of loading. Calves were more likely to be stressed compared with feeders and fat cattle because of the increased likelihood of mortality and becoming non-ambulatory. The greater mortality in calves was largely caused by 2 loads (approximately 155 dairy calves transported 1,350 km) with 2 fatalities each plus 5 and 3 non-ambulatory animals in each load. The drivers hauling these loads indicated that the condition of the calves at loading and unloading was ‘fair’ and ‘poor’ for one load and ‘good’ and ‘fair’ for the other, respectively. This highlights that cattle condition upon loading is an important risk factor associated with the incidence of mortality and becoming non-ambulatory. Finally, feeders were more affected by transport than fat cattle which had the least mortality rates suggesting that fat cattle destined for slaughter are able to cope with the challenge of transport better than other cattle categories. Similar patterns among cattle categories were observed for shrinkage because the extent and rate of shrinkage was greatest in cull, intermediate in calves, less in feeders, and least in fat cattle (González et al., 2012c). Fat cattle may be most able to cope with transport for multiple reasons in addition to their expected better body condition. Firstly, a smaller proportion of fat cattle were marketed through auction markets compared with the other categories. Loading at auction markets was associated with a greater likelihood of becoming non-ambulatory or dying during transport in the present study. However, the causes and consequences of these results are confounded (i.e., whether loading at the auction causes greater mortality or the greater mortality from markets is a consequence of more ‘high risk’ cattle, or both). Marketing through auction markets may extend the time period without access to feed and water depleting body reserves (González et al., 2012c). Marketing through an auction may also expose cattle to mixing with unfamiliar animals and an increase in the frequency of loading and unloading. This increased requirement for handling is greater in feeders and calves than fat cattle for export because US regulations require that all cattle not going directly to slaughter must be unloaded for veterinary inspection at the border. Loading and unloading has been suggested as one of the most stressful events associated with cattle transport (Kent and Ewbank, 1986) and a great contributor to bruising in cattle (Jarvis et al., 1995). In contrast, most fat cattle are transported from feedlots to slaughterhouses directly avoiding auction markets and border inspections (González et al., 2012a,c). A second reason that feeders and calves in this study had greater mortality may be that they were transported at a smaller space allowance compared with fat and cull cattle (González et al., 2012b). Other factors may have also contributed to greater mortality in feeders including the extra time they spent on the truck and greater shrinkage they experienced compared with fat cattle even though ambient temperature during transport was similar (González et al., 2012c). It has been hypothesized that fat cattle may experience less fear-associated stress during human handling and transportation because intensive rearing in feedlots accustoms them to humans as opposed to cattle coming from farm/ranches and that would typically be raised under extensive conditions (Grandin, 1997). Mortality rates in the present study are similar to those reported in studies conducted in North America and other parts of the world. For example, Warren et al. (2010a) inspected 1,363 loads of fat slaughter cattle arriving to a packing plant in Ontario (Canada) and reported very similar values for dead (0.008%) and non-ambulatory (0.010%) cattle; however, lameness (0.158%) at arrival was reported to be greater than in the present study. The 2007 Cow and Bull Beef Quality Audit in the United States recorded the rates of mortality (0.04%) and moribund (0.24%) cattle for 5,500 animals (145 loads) arriving at packing plants. Both of these values are similar to rates reported for cull cattle in the present study as the audit did not include feedlot cattle. In an 8-yr study conducted in Europe, Malena et al. (2006, 2007) reported a mortality rate for road transported cattle to be 0.007% for fat cattle, 0.027% for calves (Večerek et al., 2006a; Malena et al., 2007) and 0.039% for dairy cattle (Večerek et al., 2006b; Malena et al., 2007). These values are similar to our study although transport distances were much shorter in the European study and their values also included animals dying in lairage. The fact that cull cattle were more likely to be lame at the time of loading suggests that lameness may be one of the reasons for being culled on farm. Because of this, pre-transport inspection, management, and planning of marketing should be done accordingly to ensure the welfare of this category (Grandin, 2001). The 1999 Cow and Bull Beef Quality Audit in the United States reported that 0.8% of mature cattle were non-ambulatory (Roeber et al., 2001) but the value declined to 0.24% in 2007 (NMCB-BQA, 2007) which is similar to cull cattle in the present study. The 1999 and 2007, US audits found between 16 and 31% of cattle with some degree of lameness at packing plants (Roeber et al., 2001; NMCB-BQA, 2007). In 2007, the US audit reported that 4% of mature cattle were ‘very disabled’ (NMCB-BQA, 2007). The greater incidence of lameness in the US audits compared with the present study may be a result of older animals, the fact that observations of lameness were done later in the holding pens, study differences in the definition of ‘lameness’, a greater proportion of cull dairy cattle, and animals in the US audits being shipped from both short and long distances. Lame, sick, weak, and old animals are more likely to be transported in small numbers from farm/ranch to local slaughterhouses (short hauls) in stock trailers owned by producers to avoid welfare problems and deaths. In contrast, there were only 3 loads of cull dairy cows in the present study but no lameness or mortalities were reported (plus 30 hauls of cull beef cows and 10 loads of cull beef bulls). It is important to point out that we did not collect information about the degree of lameness because of the subjectivity involved with lameness scores and that the appraisal of lameness by truck drivers might be biased because of the difficulty for assessment under commercial situations. Therefore, results concerning the number of lame animals should be taken with extreme caution in the present study. In agreement with suggestions by Grandin (2001), fitness for travel does not seem to be an important issue in fat slaughter beef cattle because of the relatively low appearance of non-ambulatory animals although this value has not been previously reported in any published scientific paper. Transport Duration The likelihood of becoming lame or dying in our study increased significantly in cattle that spent more than 20 h on truck, although the sharpest increase was observed above 30 h. It is important to note that only 4.7% of the journeys in the present study transported animals for longer than 30 h (González et al., 2012a). This is likely related to the longer duration of stressors, energy expenditure to maintain balance and physiological conditions, and time without feed and water. All these factors may lead to fatigue and dehydration as energy and water become limiting or in deficit. Such threshold values may indicate transport times over which point cattle can no longer cope with the stresses of transport. Interestingly, the 30-h threshold would be in agreement with data from Knowles et al. (1999), who reported a sharp increase in the number of cattle lying down after 28 h of transport, indicative of fatigue. Furthermore, cattle that lied down in that study had greater plasma cortisol concentrations after the 31-h journey. In contrast to results of the present study, Warren et al. (2010a) did not find a significant relationship with transport duration in fat slaughter cattle although onboard mortality was similar to the present study. Although under very different conditions compared with the present study, mortality rate increased by 3-fold when fat cattle, dairy cows and calves were transported longer than 100 km compared with shorter than 100 km (Malena et al., 2007). However, time on truck is more important than distance during long haul transport because of delays within the journeys may be significant as previously reported (González et al., 2012a,c). In contrast, the present study did not find greater mortality in cattle transported for up to 20 h although the likelihood of death increased by at least 2-fold in cattle transported between 20 and 30 h and by at least 7-fold if transported for longer than 30 h. Differences among studies may be due to vehicle design, animals and transport conditions thereby making comparisons difficult. Ambient Temperature, Season and Loading Time In the present study, the greatest mortality rate was found at midpoint temperatures below −15°C and in the fall compared with summer. Greater mortality at low temperatures was unexpected because such findings have not been previously reported. One possible explanation is that the effects of feed withdrawal and physical activity (standing and maintenance of balance) are exacerbated by cold weather transport. This is supported by Schrama et al. (1995) who showed that fasted animals had less heat production and greater energy cost of physical activity at low compared with high temperatures. Sustaining such increased energy demands for heat production, physical activity and stress for long periods of time during cold weather may be problematic for some cattle categories that already have low energy reserves upon loading or that are less able to adapt their physiology such as younger stock. The use of winter boards (panels on the trailer that restrict airflow) on the trucks was not found to be a frequent commercial practice in our study and potential reasons were discussed (González et al., 2012a). However, the use of winter boards reduced the incidence of dark cutters during winter transport in an Ontario study (Warren et al., 2010b). The Canadian Codes of Practice recommends adding abundant bedding to trailers for insulation if temperatures fall below 10°C. Mortality also increased sharply when the temperature range within a journey was above 30°C. However, it is important to point out that such high values were only documented in 2% of the journeys in our study (González et al., 2012a). Malena et al. (2006) reported the greatest mortality for fat cattle during both summer and winter months. In contrast, the greatest mortality observed in our study occurred during the fall which may be associated with the increased transport of ‘high risk’ cattle categories such as feeders and calves, as opposed to fat cattle which were more frequently transported in the summer (González et al., 2012b). The incidence of non-ambulatory animals increased at ambient temperatures above 30°C and was greatest during the summer compared with other seasons. The threshold of 30°C is in agreement with Knowles (1999) who suggested that this threshold should not be exceeded in cattle. However, Randall (1993) suggested that there should be no maximum or minimum temperature values recommended for transport because of the ability of cattle to maintain thermo-neutrality under varying conditions. In addition, results from a companion study indicated that temperature has to be considered along with length of transport because shrinkage, and consequently dehydration, is a function of both (González et al., 2012c). Shrink Mortality was greatest and the likelihood of becoming totally compromised increased sharply in those hauls that transported cattle experiencing shrinkage above 10% of BW. These values were reached with the combination of both long time on truck and high ambient temperature for all cattle categories with the exception of fat cattle destined for slaughter (González et al., 2012c), for example over 30 h on truck and above 20°C. A plausible hypothesis for this finding is that the extent of shrinkage is linked to the amount of body water an animal losses and to their level of dehydration. The greater the shrinkage the lower the reserves of body water to respond to stress and high environmental temperature (sweating and cooling). Animals may eventually die if coping mechanisms fail including sweating or basic cellular functions. Compartment of the Trailer Welfare outcomes were also different among compartments of the trailer which might be related to different microclimate conditions and space allowance (González et al., 2012b). For instance, animals in the nose were allowed more space than recommended and were less likely to die compared with cattle in the belly where smaller space allowance seemed to be frequent (González et al., 2012b). Similarly, White et al. (2009) reported that feeder cattle transported in the nose were less likely to be treated for health problems after arrival compared with those loaded in the deck and belly, and tended to have less mortality after arrival compared with the other compartments (P = 0.15). However, the authors speculated that air quality may be better in the nose because the solid front panels may create different airflow compared with the other compartments. Space Allowance Previous studies indicated that there is an optimum space allowance during transport so cattle can achieve support from each other to avoid falls, struggles and excessive displacement within the compartment (Eldridge and Winfield, 1988; Knowles, 1999). Therefore, space allowances that are too small or too large are expected to increase the incidence of lame, non-ambulatory, and dead animals. However, to date no studies have been done from which to determine the appropriate space allowances that would aid in reducing the incidence of lame, non-ambulatory, and dead animals across different BW ranges. The present study reported a sharp increase in onboard mortality at small space allowance (i.e., high density and k-value below 0.015) for animals in the belly and deck. There were 7 dead animals at k-value equal or below 0.015 out of a total of 2,909 animals transported below such threshold, and a total of 25 dead animals were at k-value 0.020 or less. This may be due in part to the fact that animals in these compartments were consistently loaded at greater densities than the other compartments in our study and that these compartments have the largest size and hold the largest number of cattle (González et al., 2012b). The likelihood of being trapped under other cattle at small space allowance may be greater in larger compartments and groups. Similarly, excessive space allowance (k-value > 0.035) may also result in greater mortality in large compartments because it may result in further displacement of animals during abrupt changes in direction or speed of the vehicle. Thus, space allowance may play a more important role in larger compartments but this is pure speculation and more research is needed in this area. The greater mortality at small space allowance may also be related to the reduced ability to breathe and or dissipate heat because of reduced air flow and lower body surface exposure for sweating. Tarrant et al. (1992) reported that the number of struggles, falls, and bruising increased sharply when the allometric coefficient reached approximately 0.015 in cattle transported for 24 h. In contrast, Eldridge and Winfield (1988) reported greater bruising at both too large (k-value = 0.027) and too small (k-value = 0.017) space allowance. Based on these previous studies, Petherick and Phillips (2009) suggested that a k-value of 0.02 across all BW would be close to the optimum stocking density so that animals can receive support from each other and minimize displacements. Our data agree with this observation and suggest that coefficients <0.015 and >0.035 should be avoided, especially in large compartments such as the deck and belly compartments. It is important to point out that only 1% of all surveys recorded k-values equal to or less than 0.015 in the belly and 0.7% in the deck. Similarly, only 0.02% of all surveys recorded k-values ≥ 0.035 in the belly and 1% in the deck (data not shown) but these values were associated with increased mortality. Experience of Truck Drivers The number of years of experience a driver has had hauling cattle had great impact on the welfare of the cattle. For example, drivers with >10 yr of experience had decreased incidences of non-ambulatory and totally compromised cattle on their trucks than those with <10 yr of experience. This may be attributed to greater knowledge of and compliance with regulations and recommendations, and training received to deal with hauling cattle. In addition, experienced drivers may have better driving quality (i.e., smoother cornering, breaking, shifting gears, changing direction of travel), as well as parking in appropriate places for resting (e.g., in the shade or protected areas during inclement weather). Smoother driving has been reported to allow cattle to change positions as needed to maintain balance, and to reduce the frequency of stumbles and falls (Tarrant et al., 1992). In addition, more experienced drivers may better distribute animals among compartments to maintain the most appropriate space allowance. In contrast to our results, the likelihood of getting dark cutters in slaughter heifers and steers increased with increased experience of truck drivers whereas receiving training reduced it (Warren et al., 2010b). In conclusion, incidence of lame, non-ambulatory, and dead animals during commercial transport suggests that animal welfare could be improved through better management of journeys longer than 30 h and experiencing ambient temperatures below −15°C and above 30°C, as well as improving cattle distribution among compartments of the trailer to maintain the allometric coefficient between 0.015 and 0.035, and using experienced truck drivers. However, these ‘threshold’ values and results should not be extrapolated to different conditions because other factors may change the strength of such relationships with welfare outcomes including vehicle design, ventilation, cattle management, and breed. Potential correlations or co-linearity among transport and animal conditions may complicate the interpretation of results in the present study. For instance, there was a relationship between mortality and time on truck, ambient temperature, shrinkage, and cattle category. Feeder cattle were more likely to die but also spent more time on truck, and had greater shrinkage compared with fat cattle (González et al., 2012c). This study was supported by the Alberta cattle industry to study the complexity of factors affecting animal welfare during commercial transport. Objectives also included obtaining benchmark information about the current status of the transport industry practices and the relationship with other economic and welfare outcomes such as shrinkage and loading density reported in companion papers. This proactive approach provides a framework and example for other livestock industries and regions to gain understanding of commercial practices and their implications. However, collaboration and support from the industry is critical and should therefore be seriously engaged. LITERATURE CITED AFAC (Alberta Farm Animal Care) 2005. Humane handling of beef cattle: Standards for the care of unfit animals.  Alberta Farm Animal Care Association, Calgary, Canada. Eldridge G. A. Winfield C. G. 1988. The behaviour and bruising of cattle during transport at different space allowances. Aust. J. Exp. Agric.  28: 695– 698. Google Scholar CrossRef Search ADS   González L. A. Schwartzkopf-Genswein K. Bryan M. Silasi R. Brown F. 2012a. Benchmarking study of industry practices during long haul transport of cattle in Alberta, Canada. J. Anim. Sci.  90: 3606– 3617. Google Scholar CrossRef Search ADS   González L. A. Schwartzkopf-Genswein K. Bryan M. Silasi R. Brown F. 2012b. Space allowance during commercial long distance transport of cattle in North America. J. Anim. Sci.  90: 90: 3618– 3629*. Google Scholar CrossRef Search ADS   González L. A. Schwartzkopf-Genswein K. Bryan M. Silasi R. Brown F. 2012c. Factors affecting body weight loss during commercial long haul transport of cattle in North America. J. Anim. Sci.  90: 3630– 3639. Google Scholar CrossRef Search ADS   Grandin T. 1997. Assessment of stress during handling and transport. J. Anim. Sci.  75: 249– 257. Google Scholar CrossRef Search ADS PubMed  Grandin T. 2001. Perspectives on transportation issues: The importance of having physically fit cattle and pigs. J. Anim. Sci.  79( E Suppl): E201– 207. Google Scholar CrossRef Search ADS   Jarvis A. M. Selkirk L. Cockram M. S. 1995. The influence of source, sex class and pre-slaughter handling on the bruising of cattle at two slaughterhouses. Livest. Prod. Sci.  43: 215– 224. Google Scholar CrossRef Search ADS   Kent J. E. Ewbank R. 1986. The effect of road transportation on the blood constituents and behaviour of calves. III. Three months old. Brit. Vet. J.  142: 326– 335. Google Scholar CrossRef Search ADS   Knowles T. G. 1999. A review of the road transport of cattle. Vet. Rec.  144: 197– 201. Google Scholar CrossRef Search ADS PubMed  Knowles T. G. Warriss P. D. Brown S. N. Edwards J. E. 1999. Effects on cattle of transportation by road for up to 31 h. Vet. Rec.  145: 575– 582. Google Scholar CrossRef Search ADS PubMed  Malena M. Voslářová E. Kozák A. Bělobrádek P. Bedáňová I. Steinhauser L. Večerek V. 2007. Comparison of mortality rates in different categories of pigs and cattle during transport for slaughter. Acta Vet. Brno.  76: S109– S116. Google Scholar CrossRef Search ADS   Malena M. Voslářová E. Tomanová P. Lepková R. Bedáňová I. Večerek V. 2006. Influence of travel distance and the season upon transport-induced mortality in fattened cattle. Acta Vet. Brno.  75: 619– 624. Google Scholar CrossRef Search ADS   NMCB-BQA. 2007. Executive Summary of the 2007 National Market Cow and Bull Beef Quality Audit. Beef Cattle Edition. Accessed December 28,  2010. http://www.bqa.org/CMDocs/bqa/2007AuditBeef.pdf. Petherick J. C. Phillips C. J. C. 2009. Space allowances for confined livestock and their determination from allometric principles. Appl. Anim. Behav. Sci.  117: 1– 12. Google Scholar CrossRef Search ADS   Randall J. 1993. Environmental parameters necessary to define comfort for pigs, cattle and sheep in livestock transporters. Anim. Prod.  57: 299– 307. Roeber D. L. Mies P. D. Smith C. D. Belk K. E. Field T. G. Tatum J. D. Scanga J. A. Smith G. C. 2001. National market cow and bull beef quality audit-1999: A survey of producer-related defects in market cows and bulls. J. Anim. Sci.  79: 658– 665. Google Scholar CrossRef Search ADS PubMed  Schrama J. W. Roefs J. P. Gorssen J. Heetkamp M. J. Verstegen M. W. 1995. Alteration of heat production in young calves in relation to posture. J. Anim. Sci.  73: 2254– 2262. Google Scholar CrossRef Search ADS PubMed  Tarrant P. V. Kenny F. J. Harrington D. Murphy M. 1992. Long distance transportation of steers to slaughter, effect of stocking density on physiology, behaviour and carcass quality. Livest. Prod. Sci.  30: 223– 238. Google Scholar CrossRef Search ADS   Večerek V. Malena M.Jr Malena M. Voslářová E. Bedáňová I. 2006a. Mortality in dairy cows transported to slaughter as affected by travel distance and seasonality. Acta Vet. Brno.  75: 449– 454. Google Scholar CrossRef Search ADS   Večerek V. Šímová V. Malena M. Voslářová E. Malena M.Jr 2006b. Effect of calf diseases on mortality during transport for slaughter. Acta Vet. Brno.  75: 625– 630. Google Scholar CrossRef Search ADS   Warren L. A. Mandell I. B. Bateman K. G. 2010a. An audit of transport conditions and arrival status of slaughter cattle shipped by road at an Ontario processor. Can. J. Anim. Sci.  90: 159– 167. Google Scholar CrossRef Search ADS   Warren L. A. Mandell I. B. Bateman K. G. 2010b. Road transport conditions of slaughter cattle: Effects on the prevalence of dark, firm and dry beef. Can. J. Anim. Sci.  90: 471– 482. Google Scholar CrossRef Search ADS   Warriss P. D. 1990. The handling of cattle pre-slaughter and its effects on carcass and meat quality. Appl. Anim. Behav. Sci.  28: 171– 186. Google Scholar CrossRef Search ADS   American Society of Animal Science
TI - Relationships between transport conditions and welfare outcomes during commercial long haul transport of cattle in North America
JF - Journal of Animal Science
DO - 10.2527/jas.2011-4796
DA - 2012-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/relationships-between-transport-conditions-and-welfare-outcomes-during-Ya0cCQ7eae
SP - 3640
EP - 3651
VL - 90
IS - 10
DP - DeepDyve
ER -