**3. Results**

A total of 1391 trucks, mainly trailers and semi-trailers, stopped at the control post over the study period, transporting a total of 111,536 animals. Figure 1 shows the descriptive statistics of the categorical variables year, month, season and species.

**Figure 1.** Descriptive statistical analysis of the 1391 trucks which stopped at control post CE 07/PS Bitritto (Bari, Italy) from 2010 to 2015. Number of trucks examined by categorical variables (i.e., factors) year, season, month, and species.

The frequency and the percentage of all the trucks transiting across the control post based on category, provenance, nationality of transport company and destination are shown in Table 1 and Figure 2, respectively.



**Figure 2.** Pie chart by provenance, nationality of transport company, and destination for the 1391 trucks which stopped at control post CE 07/PS Bitritto (Bari, Italy) from 2010 to 2015.

The majority of the trucks transported bovines, while heavy cattle (i.e., animals with a body weight of about 550 kg) was the category most frequently transported. The most common provenances were France, Spain, and Ireland, even though the transport companies were mainly from Greece, which was also the most common final destination. There was a decrease in the number of trucks and transported animals after 2012, in particular for bovines and sheep. The transport of bovines tended to drop in August, while the transport of sheep/goats peaked in July and October (Figure S1).

Table 2 shows the descriptive statistics of the numerical variables based on the total shipments. The maximum number of DOAs (14) was recorded in a truck transporting a total of 214 lambs from Spain, and this represented 93% of all DOAs.

**Table 2.** Descriptive statistics of the numerical variables for the 1391 trucks which stopped at control post CE 07/PS Bitritto (Bari, Italy) from 2010 to 2015.


**Legend.** SD: Standard Deviation; Q1: first quartile; Q3: third quartile; NATT: number of animals transported per truck; DOA: Dead on Arrival; DCP: Dead at the Control Post; UFT: Unfit for Continue to Travel.

Space allowance varied accordingly with the species, and was within the range reported in CE 1/2005, which regulates a space allowance from 0.30 to 1.60 m2/animal for bovines, from 0.20 to 0.50 m2/animal for sheep and at least 235 kg/m<sup>2</sup> for pigs of 100 kg. In our study, the minimal values of space allowance resulted in lamb transportation (Median = 0.34, IQR: 0.27–0.34).

The mortality and morbidity rates were 0.025% and 0.010%, respectively, with maximum values for transport of lambs (0.084%, and 0.019%). Cases of mortality and/or morbidity were reported for only 11 out of the 1391 trucks (0.8%). Table 3 shows the number of transported animals, mortality and morbidity rates calculated in relation to the different factors studied.


**Table 3.** Frequency table with overall calculated mortality and morbidity rates and calculated mortality and morbidity rates per year, month, season, species, and category. Total dead = dead on arrival (DOA) + dead at the control post (DCP).


**Table 3.** *Cont.*

**Legend.** UFT: Unfit to Continue Travel.

There was an association between species transported and year (X<sup>2</sup> = 40.25; *p* < 0.001), season (X<sup>2</sup> = 38.28; *p* < 0.001), month (X<sup>2</sup> = 67.19; *p* < 0.001), provenance (X<sup>2</sup> = 751.88; *p* < 0.001), nationality of transport company (X<sup>2</sup> = 1200.74; *p* < 0.001) and destination (X<sup>2</sup> = 62.67; *p* < 0.001) (Figure 3, Figure S2).

**Figure 3.** Association between species and year, month, season, provenance, nationality of transport company and destination.

Consequently, there was also an association between category transported and year (X<sup>2</sup> = 841.29; *p* < 0.001), season (X<sup>2</sup> = 90.83; *p* < 0.001), month (X<sup>2</sup> = 191.79; *p* < 0.001), provenance (X<sup>2</sup> = 1603.70; *p* < 0.001), nationality of transport company (X<sup>2</sup> = 2114.33; *p* < 0.001) and destination (X<sup>2</sup> = 192.41; *p* < 0.001). There was a higher number of trucks transporting sheep/goats in June, July and August, i.e., in summer, than in other months and seasons of the year. Sheep/goats came mainly from Spain and Hungary.

In the univariate logistic model, species of animal being transported and space allowance were the only predictive variable which proved to be associated with a welfare problem (X<sup>2</sup> = 5.780; df = 2; *p* = 0.049, and X<sup>2</sup> = 8.982; df = 1; *p* = 0.003, respectively) (Table S1). Trucks transporting sheep/goats were four times more likely to be associated with a case of DOA, DCP, or UFT than those transporting bovines. For a unit increase in space allowance, the odds in favor of the welfare problem occurring decreased by a factor of 0.24 (OR: 0.24; CI: 0.10–0.61) (Table 4). In the multivariate model, only space allowance remained significant (*p* < 0.05).

**Table 4.** Results of the univariate regression analysis between welfare problem (trucks reported with a case of DOA, DCP or UFT) and species of the animal being transported. Data were collected from trucks (n = 1391) transiting across a control post in Southern Italy from 2010 to 2015.


**Legend.** SE: standar error; OR: odds ratio; CI: confidence interval.

At the GLM, the effect of the year was not significant (*p* = 0.072) on NATT. While the number of transported animals per truck varied significantly depending on species (pigs 106.7 ± 4.7, sheep/goats 268.7 ± 2.2, cattle 41.0 ± 1.0; *p* < 0.001), the effect of the other predictive variables associated with species resulted in significant differences on NATT (Table S2).
