No evidence of reactive avoidance of baboons (Papio ursinus and Papio anubis) to the presence of predators

An analysis of extensive camera trap data across four African savanna sites reveals that neither chacma nor olive baboons exhibit reactive avoidance by temporarily vacating areas following predator encounters, suggesting they rely on alternative strategies like proactive avoidance or aggressive retaliation to manage predation risk.

The Gist

Imagine a bustling neighborhood where the residents are baboons and the potential burglars are lions, leopards, and hyenas.

For a long time, scientists have wondered: When a burglar walks through a neighborhood, do the residents pack up their bags and move to a different street for a few days to stay safe? Or do they just lock their doors, keep an eye out, and stay put?

This paper is the story of researchers trying to answer that question using a massive network of motion-sensor cameras (like security cameras) scattered across the African savanna.

The Big Question: Do Baboons Play "Hide and Seek" with Predators?

The researchers wanted to see if baboons exhibit "reactive avoidance."

• Proactive avoidance is like checking the news before leaving the house to see if a storm is coming.
• Reactive avoidance is like seeing a burglar in your driveway and immediately running to your neighbor's house and staying there for three days.

The scientists hypothesized that if a lion (a scary, powerful predator) walked through an area, the baboons would be so spooked that they would avoid that specific spot for up to 72 hours (3 days). They thought different predators might cause different levels of fear:

• Hyenas (opportunistic hunters) might cause a short scare.
• Leopards (climbing predators) might cause a medium scare.
• Lions (the kings of the jungle) might cause the longest, most intense avoidance.

The Investigation: The "Security Camera" Detective Work

To find the answer, the team looked at data from four different "neighborhoods" (three in South Africa and one in Tanzania). They had millions of photos taken by cameras that snap a picture whenever something moves.

They used a clever statistical trick called a randomization test. Think of it like this:

1. They took all the times a lion was seen.
2. They shuffled those times around randomly, like mixing up a deck of cards.
3. They asked: "If the baboons didn't care about lions at all, how often would we see them in the area just by chance?"
4. Then, they compared the real data to this random shuffle.

If the real number of baboon sightings dropped significantly after a lion sighting compared to the random shuffle, it would prove the baboons were hiding.

The Plot Twist: The Baboons Didn't Budge!

Here is the surprising result: The baboons didn't run away.

No matter which predator showed up (lion, leopard, or hyena), and no matter how many hours passed (24, 48, or 72), the baboons kept showing up in the same spots at the same rates as if nothing had happened.

The Analogy: Imagine a burglar walks down a street. In most movies, the neighbors would hide in their basements for three days. But in this study, the neighbors (baboons) just looked out the window, saw the burglar, and then went back to mowing their lawns and playing in the park immediately after.

Why Didn't They Run?

The authors suggest a few reasons why the baboons didn't play "hide and seek":

1. They Have Superpowers: Baboons are tough. They don't just run; they fight back. If a leopard tries to attack, a group of baboon males will gang up and chase it away. They are like a neighborhood watch that doesn't just call the police; they tackle the intruder.
2. They Have Safe Houses: Baboons sleep in trees or on cliffs. If a lion is on the ground, the baboons just climb up. They don't need to leave the neighborhood; they just move to the second floor.
3. They Are Smart: They might have already known the lion was coming (proactive avoidance) and adjusted their plans before the lion arrived, so they didn't need to run away after seeing it.

The "False Alarm" and the Lesson Learned

There was a catch. The researchers also tested this method on Zebras. Zebras are known to run away from lions. When they ran the test on zebras, the method worked perfectly—it showed the zebras hiding.

However, when they reduced the number of zebra photos to match the smaller number of baboon photos, the method stopped working. It couldn't "see" the zebras hiding anymore.

The Takeaway: This teaches us that sometimes, just because we don't see animals running away, it doesn't mean they aren't scared. It might just mean we didn't have enough "security camera" photos to catch the moment they left. The baboons might be hiding, but the cameras weren't looking in the right place or often enough to see it.

In a Nutshell

This study is a reminder that nature is complex. While we often think prey animals always run when predators appear, baboons have a different strategy: Stand your ground, climb a tree, or fight back. They don't necessarily abandon their favorite lunch spots just because a lion walked by; they just make sure they are ready for a fight or a climb.

It's like saying, "Just because you didn't see the neighbors run out the back door when the burglar arrived, doesn't mean they weren't scared—they might just be very good at hiding in plain sight."

Technical Summary

1. Problem Statement and Research Gap

Predation exerts strong selective pressure on prey, driving the evolution of behavioral adaptations classified as proactive (anticipatory, e.g., site selection) or reactive (immediate or delayed responses to a detected threat). While reactive responses like fleeing or signaling are well-documented in primates, long-term reactive avoidance (temporarily vacating an area for hours or days following a predator's passage) remains understudied in primates.

Most existing studies on baboon-predator interactions rely on direct field observations, which are limited by:

• Observer bias (the "predator shelter" effect).
• Inability to monitor nocturnal activity or extended timeframes.
• Difficulty in capturing long-term spatial avoidance patterns.

The study aims to determine if chacma (Papio ursinus) and olive (Papio anubis) baboons exhibit species-specific reactive avoidance (reduced presence) in the 72 hours following the detection of their three main predators: lions (Panthera leo), leopards (Panthera pardus), and spotted hyenas (Crocuta crocuta).

2. Methodology

Data Source and Study Sites:
The researchers utilized data from Snapshot Safari, a large-scale camera trap monitoring program.

• Sites: Four savanna locations: Madikwe Game Reserve (MAD), Tswalu Kalahari Private Wildlife Reserve (TSW), and Associated Private Nature Reserves (APN) in South Africa; and Serengeti National Park (SER) in Tanzania.
• Species: Chacma baboons (South Africa), Olive baboons (Tanzania), and their predators (Lions, Leopards, Spotted Hyenas).
• Dataset: 3,042 baboon detections and 6,958 predator detections after filtering.

Data Processing:

• Independence Filter: A 30-minute time-to-independence filter was applied to predator detections to ensure distinct events.
• Temporal Windows: The study analyzed baboon presence in three 24-hour blocks after predator detection (0–24h, 24–48h, 48–72h) and symmetrically before detection to establish a baseline.
• Exclusion: Camera traps with lifespans <30 days and detections near waterholes (South African sites) were excluded to reduce bias.

Statistical Approach:

• Randomization Test: To test for avoidance, the authors compared the observed number of baboon detections following a predator event against a null distribution generated by randomizing predator dates (1,000 iterations) while preserving diel activity patterns.
• Hypotheses:
  • H1: Baboons avoid spotted hyenas (cursorial, opportunistic) for 0–24h.
  • H2: Baboons avoid leopards (semi-arboreal ambush) for 0–48h.
  • H3: Baboons avoid lions (terrestrial ambush, low mobility) for 0–72h.
• Validation: The method was validated using Burchell's zebras (Equus quagga) in the APN site, a species known to exhibit reactive avoidance of lions.

3. Key Contributions

1. Application of Large-Scale Camera Trap Data: This study demonstrates the utility of non-invasive, large-scale camera trap networks (Snapshot Safari) for investigating long-term predator-prey spatial dynamics in primates, moving beyond short-term observational studies.
2. Methodological Validation: The authors rigorously validated their randomization approach using a known avoidance case (zebras vs. lions) and demonstrated how sample size reduction impacts statistical power, providing a cautionary framework for future camera trap studies on low-density species.
3. Species-Specific Anti-Predator Strategy: The study challenges the assumption that all prey species exhibit long-term reactive avoidance, highlighting that primates may rely on alternative strategies (e.g., mobbing, short-term refuge use) rather than spatial displacement.

4. Results

Primary Findings (Baboons):

• No Reactive Avoidance: Across all four study sites and for all three predator species (lions, leopards, hyenas), there was no statistically significant decrease in baboon presence in any of the 0–72 hour time blocks following a predator detection.
• Pre-Detection Anomalies: Interestingly, in some sites (APN and SER), baboon presence was significantly higher than expected in the 24 hours before a predator detection. The authors suggest this may be due to baboons utilizing ground-level foraging routes more frequently when predators are absent, leading to higher capture rates, rather than a true avoidance behavior.
• Hypothesis Rejection: Hypotheses H1, H2, and H3 were all rejected; no difference in avoidance duration was found based on predator hunting strategy or mobility.

Method Validation (Zebras):

• Confirmed Avoidance: In the APN site, Burchell's zebras showed significant reactive avoidance of lions in the 0–24 hour window ($p = 0.024$), validating the randomization method.
• Sample Size Sensitivity: When the zebra dataset was down-sampled to match the baboon detection numbers ($n=500$), the significant avoidance signal disappeared ($p \approx 0.18$). This indicates that the lack of avoidance signal in baboons could be partly attributed to lower statistical power due to smaller sample sizes compared to zebras.

5. Significance and Conclusion

Ecological Implications:
The absence of long-term reactive avoidance suggests that baboons rely on short-term, flexible anti-predator strategies rather than spatial displacement. These likely include:

• Refuge Use: Retreating to trees or cliffs only during immediate threat.
• Mobbing/Retaliation: Aggressive counter-attacks (particularly against leopards) which deter predators without requiring the group to abandon resource-rich areas.
• Vigilance: Sentinel behaviors allowing for rapid detection and response.

Methodological Implications:
The study highlights a critical limitation in using camera traps for low-density species like primates. The validation with zebras proved that sample size is a critical determinant in detecting behavioral avoidance. The failure to detect avoidance in baboons may reflect a true biological difference (reliance on non-spatial strategies) or a statistical artifact (insufficient detections to overcome noise).

Future Directions:
The authors recommend increasing camera trap density and potentially utilizing arboreal camera traps to better capture the three-dimensional movement of primates. This study underscores the complexity of predator-prey interactions in savanna ecosystems and the need for high-resolution data to distinguish between true behavioral avoidance and statistical limitations.