Husbandry Refinement

While well-being is a qualitative experience, and thus, difficult to measure, we can use first principles to build upon our fundamental knowledge. The smallest unit of absolute known truth in a subjective condition, such as personal well-being, are the categories of experiences an animal has throughout their day. For an animal participating in medical research, we review seven categories of their daily experience, as shown in the diagram below:

In this review, we are interested in understanding the answers to these four questions within each category:

1. Do they have access to basic needs?
2. Do they have access to species-typical behaviors?
3. Do they have access to agency?
4. Do they have access to joy?

We define the primary terms in these questions as follows:

1. Basic Needs are the fundamental requirements required for animals to thrive in their environment including access to food, water, and shelter.
2. Species-Typical Behaviors are behaviors that are typically shared by all members of the same species and are a product of the brainstem, implying that they are strongly influenced by evolution.
3. Agency refers to the feeling of control over actions and their consequences. Controlling different aspects of our lives can help us feel safe and motivated. Ensuring the choices we offer to animals are truly voluntary, including the opportunities to change their minds about those choices is one way we can ensure access to agency.
4. Joy is defined generally as great happiness or pleasure, which is difficult to measure quantitatively. We therefore look at the behaviors most associated with this state and note those frequencies exist most often when an animal has access to their most preferred activities.

As an example, when we assessed the well-being of our three animal species under the category of HOUSING, we identified a gap in our sheep housing:

To address this gap and design better animal housing, we reviewed the natural history of sheep and their full range of species-typical behaviors. The prior housing included indoor barn stalls with access to outdoor corrals. Fig. 3

The top species-typical behavioral goals for sheep that we identified in our review were flocking, hopping, and grazing (see ethogram below). Barn stalls and corrals allowed the sheep the ability to flock and hop at a frequency comparable to animals in their natural environment. However, grazing behavior was not readily met in the barn and corrals under the definition in our ethogram.

To ensure our sheep have access to sufficient frequency of grazing behavior, we needed to provide them with a grass pasture, in addition to their barn and corral. Sheep are also considered a prey species, which means we need to provide substantial security, such as fencing and shelters, to protect them from predators. Fig. 5

To assess whether our changes achieved our behavioral goals, we collected the frequency of the target behaviors we observed in the prior and improved environments and compared them. In the graph below, you can see our sheep were able to access grazing behavior once they were provided the ideal environment to do so. We also observed our sheep standing at the gate to return to the barn when temperatures rose throughout the day. From these data collection sessions and observations, we learned their preference to be outdoors in the mornings and a strong preference to return to their air conditioned barn before noon.

Comfortable housing for sheep, therefore, includes a variety of environments for them to choose from. As a result of our well-being assessment, in addition to the existing air-conditioned/heated barns with bedded stalls and protected outdoor corrals, we added a protected outdoor pasture with shelters to help satisfy the behavioral needs of the sheep.

The next step to continue our well-being assessment is asking our four questions in the next categories - food and water. As shown in the assessment tables, we previously found gaps in providing choice for foraging and watering enrichments.

Before we dive into how we addressed these gaps (questions #3 - Agency and #4 - Joy), let's first check out how we approached questions #1 - Basic Needs and #2 - Species-Typical Behaviors in the feed and water categories.

Everyone has the right to basic needs. Access to a healthy diet, clean water, and protective yet comfortable housing is an absolute minimum. Our team of veterinarians designed diverse, healthy diets and helped our behavior team identify ways to deliver those diets in species-appropriate ways. Fig. 9

Novel foraging devices and environments allow us to stimulate cognition while accessing species-typical behaviors. For example, some species may prefer using water lixits over drinking from water pools, depending on what is most comfortable to them. Another example is that foragers need to perform more fine motor hand movements to pluck leaves or peel fruits, while grazers need to stroll long distances as they trim the grass. Fig. 10

Pigs, for instance, are built to root in the earth with their powerful snouts. Ensuring that they have materials to root around for food, as well as build nests with, is an important part of meeting the needs of their species-typical behavioral repertoire. Figs. 11-12

By providing not just access to water and daily calories but also delivering them through engaging, species appropriate methods, we can successfully check off both questions #1 - Basic Needs and #2 - Species-Typical Behaviors under the categories of food and water. Notably, we are absolutely against any food or water restrictions as a behavioral motivation strategy and do not engage in the practice.

This brings us back to questions #3 - Agency and #4 - Joy. How do we as an animal care team provide food and water in such a way that allow the animals agency and joy?

One exciting enrichment initiative we introduced this year to help us with this challenge was communication buttons and icons, which allows the animals to share their preferences with us. These communication tools helped us address our goal to bring animals increased agency and joy.

In Figs. 13-14, Blue Belle and Coraline press their “outside” button to get access to their corrals and then choose between receiving “grape” and “watermelon” snacks.

In Fig. 15, Mars makes a clear preference for eating strawberries that day. Choice gives animals agency. The choice of which activities to engage in is likely to bring joy, as may be evidenced in this video, where sheep selected the opportunity to go out to pasture with their trainers and displayed play behavior. Fig. 16

In Figs. 17-18, Blue Belle and Umbreon request “scritches” from their favorite trainers. In fact, each species is at various stages of communicating which foods, toys, level of human interaction, location, and even bedding they would like for the day.

Using icons on a tablet, we can obtain information about what types of tasks our non-human primates would like to engage in. They can tell us which pieces of structural enrichment they want present in their home, which foraging activities and puzzle feeders they find most engaging, and which reinforcements they find most motivating. When given the option among four different foraging devices, Mars, the rhesus macaque, requests destructible enrichment. Fig. 19

We can verify that the primates understand these associations by initially asking them to match to the sample presented. This type of association testing asks the animal to select the picture of the physical object or food being presented to them, receiving that item, and then being additionally rewarded with their most preferred item. Mars demonstrates this by accurately matching various pieces of produce to the correct food icon on his screen. He receives that item upon correct selections, followed by his preferred reinforcer of the day - apples! Fig. 20

We will continue to expand this program by introducing more buttons/icons, and thus, more choices around preferred foraging and water enrichments, social interactions, favorite trainers, and data collection spaces. Once we've created these associations between concepts and corresponding icons, our next goal will be to show animals what activities are on their schedule for the day. Having a predictable schedule that they self-elect increases autonomy.

In this Refinement update, we looked at some examples of animal well-being assessments performed at Neuralink. Regular refinement reviews like these hold us accountable to the animals contributing so much to us and the people we love. Check back in soon to see how we utilize environmental enrichment to further enhance the experience of the animals in our care.