The phylogenetic signal in primate ontogenies, with special attention to dental development

This study quantifies the phylogenetic signal across 35 primate traits and reveals that dental development, particularly permanent teeth, exhibits the strongest evolutionary conservation, challenging the expectation that earlier-developing traits are more phylogenetically constrained.

The Gist

The Big Question: Who Are We, Really?

Imagine you are looking at a family photo album. You can see that cousins look alike, siblings look alike, and parents look like their kids. In biology, this is called phylogenetic signal. It's the idea that because species share a common ancestor, they tend to share traits.

But here's the tricky part: Are they alike because of their family tree, or because they are all just trying to survive in the same environment?

Scientists have spent a lot of time studying adult animals (like how big a monkey's brain is or how long it lives). But this paper asks a different question: What about how they grow up? Do baby monkeys grow up in a predictable, "family-like" way, or is their development a chaotic mess that changes easily?

The Experiment: A Race of 157 Primates

The author, Paola Cerrito, gathered data on 35 different milestones in the lives of 157 different primate species (from tiny marmosets to huge gorillas).

She looked at four main categories of "growing up":

1. Motor Skills: When do they start crawling? When can they use their hands like adults?
2. Cognitive Skills: When do they start playing social games?
3. Life History: When do they get weaned? When do they have their first baby? How long do they live?
4. Dental Development: When do their baby teeth fall out? When do their adult teeth come in?

She used a mathematical tool (called Blomberg's K) to measure how "sticky" these traits are to the family tree.

• High K (Sticky): The trait is very consistent across the family. If you know the family, you can predict exactly when this happens.
• Low K (Slippery): The trait changes easily. Even close relatives might do this at very different times.

The Surprising Results

1. The "Slippery" Traits: Life History

The traits that were the least consistent (the most "slippery") were things like how long they live and when they have their first baby.

• Analogy: Think of these traits like a fashion trend. Even if you and your cousin are related, you might dress very differently because you are reacting to the current "season" (environment). Evolution can change these things quickly to fit new situations.

2. The "Sticky" Traits: Teeth

The most consistent traits were teeth. Specifically, the timing of when teeth erupt is very hard-wired into the primate family tree.

• Analogy: Think of teeth like the engine of a car. No matter if the car is a red sports car or a blue truck (different species), the engine parts are built to a very specific, unchangeable blueprint. If you know the engine model, you know exactly how it works.

The Winner: The most "sticky" trait of all was the eruption of the lower canine tooth. It is so predictable that if you find a fossil of a lower canine, you can guess the animal's age with incredible accuracy.

3. The Big Twist: "Later is Stronger"

The author had a hunch that things happening early in life (like baby teeth) would be more fixed because they are "hard-wired" by genetics. She thought things happening later (like adult teeth) would be more flexible because they are influenced by the environment.

She was wrong.

• The Reality: Later-developing traits (permanent teeth) were actually MORE conserved (stickier) than early ones (baby teeth).
• The Metaphor: Imagine a construction project.
  • Baby teeth are like the scaffolding. It's temporary, it's a bit wobbly, and different construction crews might build it slightly differently depending on the weather.
  • Adult teeth are like the foundation and the steel beams. Once the scaffolding is gone, the permanent structure is poured with concrete. It is rigid, unchangeable, and follows the architect's (evolution's) plan perfectly.

Why Does This Matter?

1. The "Train" Problem

In the introduction, the author uses a great analogy about trains. Imagine you are on Train A and you see Train B next to you. Is Train A moving? Is Train B moving? Or are both moving?

• In evolution, scientists often try to say, "Humans weaned their babies earlier than chimps." But to say that, you need a fixed point to measure against.
• If you measure against weaning, you might be wrong because weaning is "slippery" (it changes easily).
• If you measure against adult teeth, you are using a "fixed rail." Because adult teeth are so consistent, they are the best anchor to tell us if other things (like brain growth or weaning) are actually speeding up or slowing down in human evolution.

2. Reading the Fossil Record

When we find a fossil, we can't see the brain or the behavior. We mostly see bones and teeth.

• Because adult teeth are so "sticky" and predictable, they are the best clues we have to figure out how fast an ancient human ancestor grew up.
• Because baby teeth are "slippery," they might actually tell us more about how flexible a species was in adapting to its environment (like when they stopped breastfeeding).

The Bottom Line

This paper tells us that not all parts of growing up are created equal.

• Some parts of growing up (like life span and weaning) are flexible and change quickly to fit the environment.
• Other parts (like adult teeth) are rigid and follow the family tree strictly.

By understanding which parts are "sticky" and which are "slippery," scientists can finally stop guessing and start accurately measuring how humans and our ancestors evolved over millions of years. It turns out that the permanent teeth are the most reliable timekeepers in the primate family.

Technical Summary

1. Problem Statement

Comparative evolutionary biology relies on the assumption that closely related species share traits due to common ancestry. However, distinguishing between traits shaped by shared phylogeny versus those shaped by adaptive pressures or environmental plasticity is challenging. While the phylogenetic signal of adult phenotypes has been extensively studied (e.g., Blomberg et al., 2003), the phylogenetic signal of ontogenetic trajectories (the timing of developmental events) remains poorly understood.

A critical issue in primate evo-devo and human evolution studies is the lack of a reliable "temporal anchor." Researchers often use specific developmental milestones (e.g., age at weaning, brain maturation, or dental eruption) to assess whether other traits are accelerated or delayed. However, if the anchor trait itself is evolutionarily labile (variable), such comparisons are invalid. The author seeks to quantify which developmental traits are phylogenetically conserved (constrained) versus labile to identify robust anchors for comparative studies.

2. Methodology

Data Collection:

• Dataset: Age at event occurrence (in days, measured from conception) for 35 variables across 157 primate species.
• Categories: Motor development, Cognitive development, Life-history traits, and Dental development.
• Sources: Literature review of 40+ studies. Data for males and females were averaged; gestation duration was added to post-natal ages to standardize all data to "age from conception."
• Sample Sizes: Varied significantly, ranging from 14 species (Onset of crawling) to 157 species (Gestation duration).

Statistical Analysis:

• Phylogenetic Signal ($K$): Calculated using Blomberg's $K$ statistic via the phytools R package. A value of $K=1$ indicates evolution under Brownian Motion (BM); $K>1$ indicates strong phylogenetic conservatism; $K<1$ indicates lability.
• Significance Testing: Significance was determined by comparing observed $K$ against a null distribution generated by 100,000 random permutations of data across the tree.
• Phylogeny: Used the species-level mammalian phylogeny from Upham et al. (2019).
• Sample Size Bias Control: To address uneven sample sizes, the author performed subsampling analyses (capping samples at 11, 20, and 25 species) to ensure results were not artifacts of sample size.
• Evolutionary Models: Fitted six evolutionary models (Brownian Motion, Ornstein–Uhlenbeck, Early Burst, Pagel's $\lambda$, $\delta$, and $\kappa$) using the geiger R package. Models were compared using Akaike's Information Criterion (AIC).
• Regression Analysis: Constructed linear models to test the relationship between phylogenetic signal ($K$) and the average age of trait emergence, testing for interactions with tooth type (deciduous vs. permanent) and jaw location.

3. Key Results

Phylogenetic Signal by Category:

• Dental Development: Exhibited the strongest phylogenetic signal ($K$ range: 0.7–2.6). This supports the prediction that morphological traits are more conserved than behavioral or life-history traits.
• Life-History Traits: Exhibited the weakest signal (most labile), with $K$ values ranging from 0.51 to 1.01.
• Motor/Cognitive: Showed intermediate to high signal ($K \approx 1.0–1.15$).
• Extreme Values:
  • Most Conserved: Eruption of the mandibular canine ($K = 2.64$).
  • Most Labile: Maximum longevity ($K = 0.51$).

Timing and Conservation (Counter-Intuitive Finding):

• Contrary to the hypothesis that earlier-developing traits are more conserved, the study found a positive correlation between the age of emergence and phylogenetic signal within dental traits.
• Permanent teeth (later developing) showed significantly stronger phylogenetic signals than deciduous teeth (earlier developing).
• The best statistical model for explaining variance in dental $K$ was simply the average age of emergence (slope = 0.6), without needing to include tooth type or jaw as interaction terms.

Evolutionary Model Fit:

• Brownian Motion (BM) was the best-supported model for the majority of traits, suggesting gradual divergence.
• Exceptions: Six life-history and one motor trait (including maximum longevity, gestation duration, and age at sexual maturity) were best explained by rate-scaling models (Pagel's $\lambda$ and $\kappa$).
  • These traits showed high $\lambda$ (strong phylogenetic structure) but low $\kappa$ (changes concentrated on short branches/early in divergence), indicating episodic evolution rather than gradual accumulation.

4. Key Contributions

1. Identification of Robust Temporal Anchors: The study identifies the eruption of the mandibular canine (and other permanent teeth) as highly phylogenetically constrained, making them superior anchors for comparative developmental studies compared to life-history traits or deciduous teeth.
2. Refutation of the "Early is Conserved" Hypothesis: The paper challenges the assumption that early developmental stages are the most evolutionarily stable. Instead, it demonstrates that later-developing permanent teeth are more constrained than earlier-developing deciduous teeth.
3. Decoupling of Signal and Evolutionary Mode: The results show that high phylogenetic signal ($K$) does not always equate to Brownian Motion. Some traits with high signal evolve via episodic shifts (rate-scaling), while others with lower signal may still follow BM.
4. Methodological Framework: Provides a practical tool for researchers to select appropriate developmental markers for reconstructing the ontogeny of extinct hominins, distinguishing between traits likely to reflect adaptive plasticity versus deep phylogenetic constraints.

5. Significance and Implications

• Human Evolution & Fossil Record: Many studies use the eruption of the first permanent molar (M1) as a standard anchor to argue for "early weaning" or "delayed maturation" in humans. This study suggests M1 is not the most constrained trait; the mandibular canine is more stable. Consequently, interpretations of human developmental acceleration based on M1 relative to weaning may need re-evaluation.
• Deciduous Teeth as Markers of Plasticity: Because deciduous teeth show lower phylogenetic signal ($K < 1$), their timing may be more responsive to environmental factors (e.g., nutrition, weaning age). This makes them potentially useful markers for inferring life-history strategies (like weaning age) in fossil hominins, rather than just being "noise."
• Evo-Devo Applications: The findings underscore the necessity of distinguishing between "hard" constraints (permanent dental timing) and "soft" constraints (life history, deciduous timing) when modeling evolutionary trajectories. It provides a quantitative basis for selecting which developmental milestones are reliable for cross-species comparison.

In conclusion, Cerrito provides empirical evidence that phylogenetic constraint varies significantly even within a single developmental system (dentition), offering a refined framework for interpreting primate and human evolutionary history.