Although on-farm welfare assessments exist solely to help understand and improve the welfare of the animals involved, the use of certain WIs themselves can significantly impact welfare. Certain WIs, such as those requiring physiological samples or handling of the animals out of water, can be significantly invasive and disturbing to the animals under assessment. Conversely, certain WIs benefit from causing little to no disturbances, with some indicators being carried out on animals blissfully unaware they are under assessment. Another consideration closely linked to the practicality of WIs are what personnel are required to conduct the assessments on-site; direct, animal-based WIs commonly require sampling methods in which multiple farm staff are likely to be involved in the netting, crowding, capturing, and actually sampling / scoring stages. Certain animal-based WIs, however, require no staff on-site as they can either be recorded remotely (e.g., with remote sensors or live-feed cameras), or the capturing process itself can be automated (e.g., through computer vision algorithms on feed cameras).

Key difference between WIs being intrusive vs. invasive:

Intrusive WIs: Indicators that can cause stress/discomfort to the animal, but do not require significant interventions (e.g., measuring body weight for growth, body condition.

Invasive WIs: Indicators that specifically require physical interventions where the animal's skin and/or body is penetrated (e.g., collecting blood samples, tissue/gill biopsies, implanting tags).

Production parameters:

While still often requiring some form of sampling, many production parameters at least tend to act as non-invasive OWIs, through virtue of the fact that they reflect the current physical/physiological health of the animals without requiring physical intervention.

Mortality rate: One of the most commonly used, health-related OWI, a sharp increase of mortalities is one of the most definitive indications that aquatic animal welfare has already been severely compromised. However, it is also retrospective as an WI; 

Sampling considerations - Mortality rates are simple and recorded as part of daily routines on almost all farms; sudden sharp increases in mortality rates can be used as an OWI. It is, however, as "crude" of an WI as possible. Only measurable at the population level, and by the time an animal has died, it is too late to respond or treat the animal.

Growth rate: Intrinsically linked to health and nutritional needs of the animal. Growth rates vary greatly between different life stages / species, and can be affected by a multitude of factors. Most importantly for certain species, growth can be impacted by factors that are not related to welfare. Owing to this imprecision, poor or reduced growth rates must be linked with other poorly scoring OWIs (i.e., those indicative of chronic stress or disease) in order to link it with a welfare problem.

Behaviour as an OWI:

There are a whole suite of behaviours that can be evaluated as OWIs (see below), at either the individual or group level. The importance of an aquatic animal's behaviour as a WI will depend on the context, species-specific behavioural repertoire, and the ability of an individual to adapt at any given moment (see allostasis in Module 1). If using behaviour to evaluate the animal's ability to cope with its surroundings, it is specifically the change of an animal's behaviour in response towards a certain stimuli or change over time that provides insights. 

Unique insights provided by behavioural WIs: early warning signs, potential to be recorded/captured remotely and even automated.

Feeding responses (incl. appetite/feed intake): While nutrition and the need to have access to food is a well-established necessity for any animal, whether they choose to eat (and how much) depends on multiple interconnected behavioural and physiological factors, namely appetite. Appetite itself arises from various factors, including its energy reserves, how full its stomach is, and seasonal adaptations (which can vary significantly for certain species) alongside the animal's own motivations to feed. Intraspecific competition, stress (e.g., the presence of predators / repeated husbandry disturbances), disease, inadequate environmental conditions, and many other issues can suppress feeding responses.

Swimming performances (darting events, lethargy, schooling / coordination etc.): In species that normally school, a breakdown of this coordinated swimming (particularly individuals separated from the group) can be indicative of stress / illness.

Opercular beat rates (OBR): Gill beat rate increases when the need for oxygen supply to the body increases (likely either due to reduced oxygen levels in surrounding water, or higher metabolic rates due to higher stress levels). This is not necessarily indicative of increase of acute stress of impaired welfare, but may signal that something is wrong (especially when paired with poor WQ parameters or other poorly scored OWIs).

Surface activity (jumping/flashing, gasping etc.): Excessive jumping/flashing out of the water often indicates some form of irritation, either from parasites, poor water quality, or overcrowding. Overcrowding often coincides with surface gasping, indicating a deficiency of oxygen that would be a consequence of the crowding. 

Sampling considerations of behaviours - 

Strengths - Qualitative changes can be quickly, remotely, and remotely captured, placing behaviour as one of the key OWIs for detecting welfare issues (particularly at their early stages of occurring). Variety of approaches can be taken to observing fish behaviour, from personnel observing adjacent to rearing systems, to underwater feed cameras, or echo sounders; each provide different perspectives & dimensions on the behaviour, from vertical, qualitative observations to 3-dimensional recordings of changes in distribution within their enclosure. Abrupt deviations in normal behaviour have been linked with both acute and chronic stressors in aquaculture, typically established signs of disease / poor welfare. 

Weaknesses - Many behavioural OWIs, however, can be difficult to quantify and/or entirely dependent on the motivation and experience of the observer. Quantitative changes (swimming speed, aggression levels, darting events, gill beat frequency etc.) are mostly only achievable by later analysis, making it a less appropriate OWI. There may also be difficulties explaining and quantifying what abnormal behaviours consist of, and there are significant risks of misinterpreting certain behaviours.

Holistic assessments - behavioural analyses (QBA, human intuitive assessments):

Whole-animal, behavioural indicators are outcome-based observations that provide an insight / snapshot into the animals' overall welfare status, typically by assessing how the animal responds or interacts with its environment. Examples of this include human intuitive analyses and Qualitative Behavioural Assessment (QBA). Human intuition typically refers to human observers that, over the long-term, record the lives of animals alongside the conditions they are kept in the farming / enclosures (as a reference point). More recently, the more robust application of QBA within aquaculture has been explored, particularly for what unique insights into welfare it may provide. 

QBA involves describing and quantifying the overall expressive manner / body language in which animals carry out their behaviours; rather than assessing what the animal is doing, QBA focuses on how relaxed / agitated etc. the animal appears to be while it carries out its behaviours.

The following video is from Francoise Wemelsfelder, one of the co-founders of Qualitative Behavioural Assessment (QBA); in this video, she discusses the underlying framework for QBA and how it can provide unique insights into welfare, particularly that of farm animals:

NB: This third-party video is linked for illustration only and copyright and control belong entirely to the originating organisation