*2.1. Animals and Transportation*

Sheep were video recorded (Figure 2) as a group during 52 short-haul (1–2 h) commercial road journeys. We filmed transport events opportunistically, and then later selected footage approximately 10 min after departure point that would allow us the opportunity to test for differences between experimental treatments (Table 1). Data was collected during routine sheep haulage over two months in Spring 2011 (Study A: Comparison of trailer crate design, breed of sheep and point of origin) and two months in Autumn 2013 (Study B: Comparison of the deck level within a crate).

The sheep filmed on all transport events were located in the fore or middle pen of the upper deck of the front crate, or the fore pen of the lower deck of commercial livestock transport vehicles registered in Western Australia. These pens were selected for their suitability for camera attachment and recording because it was unknown if adequate footage could be obtained from cameras mounted on these vehicles given the stocking densities and prevailing weather. Stocking densities and handling procedures for all transport events were at standard industry recommendations.

We carried out two studies, Study A compared the following: 1. trailer crate design, 2. sheep breed and 3. point of origin. Study B was an investigation of the effects of deck level for two trailer crate designs.

**Table 1.** Description of the treatment comparisons carried out. Bold text indicates the comparison made for each study.


†: Vehicles with the standard crate design had four 1.2 × 2.4 m W × L (2.9 m2) pens per deck and a solid metal floor that was fixed in place. Vehicles with the convertible crate design had two 2.4 × 3.0 m W × L (7.2 m2) pens per deck and a solid metal floor that was not permanently fixed in place. ‡: The routes taken included a period of continuous and stop-start driving, originating from either SY: saleyard (Muchea, Western Australia) to feedlot, or FL: from a registered premises feedlot (Baldivis, Western Australia) to a loading wharf (Fremantle, Western Australia).

#### *2.2. Study A: Comparison of 1. Trailer Crate Design, 2. Sheep Breed and 3. Point of Origin*

Study A1: we compared the behavioural expression of sheep transported on the upper decks of two vehicle types: a standard crate (S) and a convertible crate (C) (Table 1; Figure 3). In the convertible crate, two decks that carry sheep are designed to fold away up against the sides of the vehicle for the cartage of cattle; hence decks 1 and 3 have fixed floors and decks 2 and 4 (upper) have non-fixed, fold-away floors. Additionally, the convertible crate has horizontal side panels rising above sheep head height, whereas the standard crate has side panels with a top side rail that curves inwards (about 250 mm from the side) above the head of the sheep, designed to prevent them from jumping out of the crate. Therefore, there is slight reduction of head room on the standard crate (Figure 3). Only sheep in the upper deck of both vehicle types were analysed in this study.

Study A2: we compared between merino (M) and fat-tail (F) sheep breeds. Merino groups were subjectively identified as principally merino based on physical appearance. The fat-tail group consisted of Awassi, Damara or crossbred mixes of these breeds (Figure 4). Sheep were sourced from a live export pre-assembly feedlot where they had been in the feedlot between 3 and 5 days.

Study A3: we compared merino sheep transported from two points of origin (saleyard or feedlot). Animals sourced from the saleyard had been fasted anywhere between 24 and 72 h These animals were provided with water *ad libitum* and were held in point of origin groups. The animals had been transported between 1 and 2 days before the experimental transport trip and exposed to a new environment and drafted into pens on arrival. Animals sourced for the live export feedlot had feed available up until being brought into the loading yard (2–3 h prior to the transport event). Prior to transport, sheep were held at the feedlot between 3–5 days in the same groups in which they were transported.

**Figure 4.** View of a typical consignment of fat-tail sheep in a convertible crate.

#### *2.3. Study B: Comparison of the Deck Level within a Crate*

Study B: we compared between lower (L) and upper (U) decks for Standard (S) crates (Study B1: SU-SL), lower (L) and upper (U) decks for Convertible (C) crates (Study B2: CU-CL). A final analysis included all four combinations (Study B3). The sides of the upper and lower decks for both crate designs were similar, being composed of horizontal slatted metal. The upper decks for both crate designs had an open roof. Flooring of upper decks (non-slip metal grate over a solid metal sheet) differed according to crate type; the standard crate floor was fixed and the convertible crate floor was suspended, providing slightly less stability. Pens on the lower decks had solid metal sheeting roof.

Flooring of lower decks (non-slip metal grate over a solid metal sheet) also differed according to crate type; the convertible crate had a metal beam (150 × 100 mm W × H) that traversed the front pen, providing structural support for the convertible function of the vehicle (Figure 1). This beam required the sheep to stand astride it. The size and type of the flooring grate varied slightly between vehicles, as did the amount of compacted manure that was trapped between the floor and the grate.

## *2.4. Qualitative Behavioural Assessment*

#### 2.4.1. Selecting Footage for Scoring Clips

Continuous video footage was recorded during transport with a digital camera on both the upper deck (Panasonic SDR-H250, Belrose, NSW, Australia) and lower deck (GoPro Hero3 White Edition, and ContourRoam, Harvey Norman O'Connor WA, Australia) fixed to the railing of the trailer, just above sheep head height. From the footage approximately 1 min duration clips were edited out after the 10 min timepoint of the video so as not to introduce selection bias, as this was the period where the vehicle had commenced travelling at reasonable speed (on highway) and was predicted to be a time where behavioural responses of the sheep would be most marked. Clips were only rejected if the heads of the sheep were not clearly visible and subsequently taken at the next opportunity. Study A: footage was selected from approximately 10–15 min into the transport journey. Study B: clips were selected from upper and lower deck footage at the same time point at approximately 12 min after the truck departed the feedlot. Each clip was edited to highlight the focal groups by increasing the opacity of the surrounding animals in the same frame (Adobe Premiere Pro CS3 and Adobe After Effects CS3, Chatswood, NSW, Australia) and sound was removed thus obscuring evidence from the vehicle.
