Lions, construction, and cognitive bias, oh my! (Part 1)

At Zoo Atlanta, our mission is to pair innovation with responsible animal care. When those values meet, we can better understand the animals in our care and continue improving their welfare.  

One recent example of this is our cognitive bias study with our three African lions – Hondo, Hatari, and Azizi – and was designed to explore how our big cats think and feel during periods of change. The project began during the renovation of the lions’ habitat, which expanded their space and improved guest viewing. While those upgrades were exciting, they came with temporary disruptions. For several months, the lions were out of the public eye with access to one of their outdoor spaces (instead of two), and were having to adapt to new sounds, people, and routines. Research shows that animals exposed to extended or unpredictable construction noise can display behavioral changes consistent with stress or altered welfare states (Kight & Swaddle, 2011; Jakob-Hoff et al., 2019). Thus, we wanted to know what effect this construction might have on their emotional state and how we could help them cope better. 

You may remember our previous research post by Dr. Marieke Gartner on cognitive bias, where she introduced cognitive bias testing as a way to peek into an animal’s overall state of mind to ask whether they see their world as “glass half-full” or “half-empty.” Our first study explored optimism and pessimism in our Ambassador Animal birds, helping us understand how season, show schedules, and daily routine might influence their outlook. Building on that foundation, we wanted to see if the same idea could work with a very different species. Unlike birds, we can’t change a lion’s environment as easily, or have them directly engage with something, so we had to adapt our methods.  

Cognitive bias tests rely on an animal’s strongest senses, as that’s how they best interpret the world around them. For cats, smell is everything, and we began there. Each lion was to learn that one scent in one corner of their indoor area meant a high-value reward, while another meant a low-value reward. Once they could tell the difference, we would introduce “in-between” or ambiguous scents that they hadn’t encountered before to see how they’d respond when the outcome was uncertain.  

Our first plan was to pair watered down goat milk (a favorite) with soiled shavings from our zebras and giraffes for the high reward, while we used part of their daily diet (less rewarding than milk, but still yummy) with cologne, which was already approved by our Veterinary Team for scent enrichment. However, the shavings were like lion catnip, with the boys being far too busy reacting and rubbing on everything to make a choice!  

As our scent cues were too disruptive, we pivoted to using visual cues. We came to this decision for two reasons. First, the Lion Care Team already uses shape-based target training as part of daily husbandry, so the lions were familiar with responding to distinct objects (such as an X or an O). Second, cats can distinguish both color and shape cues, with studies showing that domestic cats possess dichromatic vision and are able to discriminate blues, yellows, and some greens, having vision like red-green colorblind humans (Loop, 1970; Tanaka et al., 2017). Cats have also demonstrated visual discrimination and object recognition across shapes and patterns, particularly when linked with food rewards (Loop & Bruce, 1978). While they might not see black and white as distinct colors in the human sense, they can absolutely perceive one as darker and the other as lighter. Pulling from this research, we used 3D-printed shapes made by our Education Department: a black dodecahedron (low reward), a white pyramid (high reward), and cubes in three shades of gray as ambiguous cues. 

Sessions took place at the mesh, and we used both high and less-preferred foods as reinforcers and relied on the lions’ gaze to indicate their selection. Other cues, like body movement, proved inconsistent, so a clear look toward one object/reinforcer counted as a choice even if the lion’s body was angled away. Additionally, each session involved three people: a keeper who presented either the black or white object, and two welfare scientists who reinforced the lions’ choice with the appropriate reward. Once a lion could accurately identify which object led to which reward at least 80% of the time, they would be ready for testing. So far, only Hondo has reached that benchmark, but his brothers aren’t far behind. This means we can now continue to the next exciting step of testing Hondo and seeing if our methods will aid in understanding the affective state of one of our lions. 

We had hoped to keep the study running while habitat construction was underway, but progress sometimes moves faster than plans, and renovations wrapped up before we could continue. Now that the lions are settled back into their usual rhythm, our focus shifts to refining methods, reviewing initial results, and preparing for what comes next. As we move forward, we’ll continue to refine our approach and build on what the lions have taught us about our methodology, keeping the spirit of curiosity that started it all. At Zoo Atlanta, every question we ask brings us one step closer to understanding the animals in our care and making their world, and ours, a little better. 

1. Jakob-Hoff R, Kingan M, Fenemore C, Schmid G, Cockrem JF, Crackle A, Bemmel EV, Connor R, Descovich K. Potential Impact of Construction Noise on Selected Zoo Animals. Animals (Basel). 2019 Jul 31;9(8):504. doi: 10.3390/ani9080504. PMID: 31370284; PMCID: PMC6721009. 
2. Kight, C.R. and Swaddle, J.P. (2011), How and why environmental noise impacts animals: an integrative, mechanistic review. Ecology Letters, 14: 1052-1061. 
3. Loop, M. S. (1970). Color vision of the cat. Vision Research, 10(2), 185–192. 
4. Loop, M. S., & Bruce, L. L. (1978). Cat visual discrimination learning: Effects of stimulus complexity. Animal Learning & Behavior, 6(2), 173–178. 
5. Tanaka, T., Mikami, A., & Saito, A. (2017). Behavioral evidence for dichromatic color vision in the domestic cat. Experimental Brain Research, 235(12), 3753–3761. 

Alexz Allen
Research Associate