Table 2.

Illustrated example of the assessment tool derivation process based on a case study with the pygmy blue-tongue skink (PBTS) (Tiliqua adelaidensis) (adapted from [10] under a Creative Commons Attribution CC BY 4.0 license).

Step in Tool Derivation Process	Process Performed	Outcome of Step	Examples of Considerations
Framework Choice	Contrasted frameworks available and selected Welfare Quality Protocol due to clear guidance on criteria provided	Selection of Welfare Quality Protocol	Frameworks considered were Five Domains and Welfare Quality due to previous experience with use
Source Indicators	Species- relevant criteria were derived by local consultation with keepers and veterinarians, and review of broader literature on lizards	Derived 39 animal or resource-based indicators, aligned against the 12 Welfare Quality criteria and 4 principles	Animal cleanliness was derived as a measure of comfort around resting- the literature suggests that scat piling may be disturbed if welfare poor [88]
			The PBTS is omnivorous, and their diet should consist of 50% vegetables, 25% fruits, and 25% invertebrates such as snails. This needs consideration under the criteria “appropriate diet” for the category of ‘good feeding’ [89]
Review of indicators to determine those for inclusion	Review of derived indicators by team comprising zoo veterinarian, researchers, animal welfare officer and two reptile keepers	Removed one animal-based indicator to yield 38 indicators. Tool was made up of predominately animal-based indicators (77%) and was designed to be	Tail autotomy was initially identified but removed since in the particular genus of interest, Tiliqua, there are reduced planes and tail drop is therefore unlikely [90]
		part of a longer audit-type assessment.	Tool primarily comprised animal-based indicators as suggested in EU Welfare Quality documentation
		Adapted grading system from Sherwen at al. 2018 [3] with scoring for from 0–2 representing high risk to low risk for resource-based indicators, and poor to good welfare for animal-based measures. The overall score was determined by summation and determination of percent out of the maximum score possible.
		Scores were not weighed since there was no evidence at hand to determine relative importance of the criteria in terms of indicators of animal welfare
Evaluate validity, reliability and practicality	Pilot study performed on a breeding pair of pygmy blue-tongue skinks and their enclosure through manual observation	Observation took 2 h for this pair. This is likely suitable for an audit-type assessment but would have been excessive for a daily check	Noted that some criteria could not be observed but were not necessarily a sign of compromised welfare. For example, food was not presented at the time of observation so food intake and hunting behaviour could not be assessed. This highlighted the need to consider incorporating information from records since these criteria had been observed previously the same week.
		Observation performed in winter	Understanding that due to ectothermic physiology and lizards being dormant, behavior may differ in contrast to summer assessment. In spite of this animals scored 79% implying good welfare (based on a 60% threshold).
Develop Final Tool and Implement	Review of derived indicators by team comprising zoo veterinarian, researchers and reptile keepers	Three resource-based indicators were added following observations to yield tool with 41 indicators	Three resource-based indicators were added; enclosure cleanliness, maintenance and group size
		Ensure welfare assessment done at a time when food available so the feeding- related criteria can be assessed	After reflective process, decided to continue work to expand the derived tool out to other reptiles with a focus at taxa level
	Determine identity of scorers and that tool most appropriate for audit-style assessment	Implement and continue to refine tool based on feedback from users and considering corroboration with other indicators e.g., information from health records.