Evolving initial conditions: an alternative developmental route to morphological diversity

This study demonstrates that phenotypic diversity in Lake Malawi cichlid vertebral counts can evolve through changes in the initial size of the pre-somitic mesoderm rather than by altering the underlying somitogenesis process, highlighting initial conditions as a significant, previously overlooked driver of morphological evolution.

The Gist

The Big Question: How Do Animals Get Different Shapes?

Imagine you are baking two different cakes. One is a tall, slender tower cake, and the other is a short, wide bundt cake. Usually, we assume that to get these different shapes, you have to change the recipe. Maybe you add more flour, change the baking time, or use a different oven temperature.

In biology, scientists have long believed that when animals evolve different body shapes (like a snake having hundreds of bones versus a frog having very few), they must be changing the "recipe" of their development. They thought the genes and the timing of how the body builds itself had to change.

The Discovery: It's Not the Recipe, It's the Starting Dough

This paper, written by researchers at the University of Oxford, studies two types of fish from Lake Malawi (a group of fish famous for evolving rapidly). One fish is short and stout (Astatotilapia calliptera), and the other is long and snake-like (Rhamphochromis sp. 'chilingali').

The long fish has 38 vertebrae (backbone bones), while the short fish only has 32. That's a 25% difference!

The researchers asked: Did the long fish evolve a new "recipe" to make more bones? Did they speed up the clock that builds the bones? Or did they change the machinery?

The answer was surprising: No.

They found that the "recipe" (the genes), the "clock" (how fast the bones form), and the "machinery" (how the cells move) were exactly the same in both fish. The only difference was how much dough they started with.

The Analogy: The Conveyor Belt Factory

Imagine a factory that makes cookies.

• The Conveyor Belt: This represents the fish's body growing longer.
• The Cookie Cutter: This represents the "segmentation clock," a biological timer that cuts out one vertebra (cookie) every 65 minutes.
• The Dough: This is the Pre-Somitic Mesoderm (PSM), the raw tissue waiting to be turned into bones.

In the short fish, the factory starts with a small pile of dough. The cookie cutter runs down the line, cutting out 32 cookies, and then runs out of dough. The factory stops.

In the long fish, the factory starts with a huge pile of dough. The cookie cutter runs at the exact same speed and cuts cookies at the exact same rate. But because they started with so much more dough, the cutter keeps going for seven hours longer, eventually cutting out 38 cookies before the dough runs out.

What This Means for Evolution

The researchers discovered that the long fish didn't need to invent a new way to make bones or change the speed of the machine. They just needed to start with a bigger pile of raw material.

1. The "Initial Conditions": In science, "initial conditions" are the starting state of a system. In this case, the long fish simply had a larger body axis (the "dough") before the bone-making process even began.
2. The Timing: The long fish started the bone-making process later than the short fish. This extra time allowed them to grow a bigger pile of dough before the "cookie cutter" started working.
3. The Result: Because the process itself (the cutting) was identical, the bones looked the same and formed at the same speed. The only difference was the total number of bones, simply because the process ran longer due to the larger starting size.

Why Is This Important?

This is a big deal for how we understand evolution.

• It's Easier to Change the Start: Imagine trying to redesign a complex machine (changing the recipe). It's hard and risky. But imagine just adding more raw material to the start of the line. That's much easier and safer.
• Rapid Change: This explains how animals can evolve big differences very quickly. They don't need to wait millions of years to invent new genes. They just need to tweak the very beginning of development to get a bigger "starting pile," and the rest of the process naturally produces a different result.
• A New Perspective: It suggests that many of the amazing differences we see in nature might not be because animals are building things differently, but because they are starting with different amounts of "stuff" to build with.

The Takeaway

Think of evolution not just as rewriting the instruction manual, but sometimes as simply pouring more batter into the pan before you turn on the oven. The oven (the developmental process) stays the same, but the final cake comes out much bigger.

This paper shows us that sometimes, the secret to becoming a giant or a dwarf isn't changing how you grow, but simply starting bigger.

Technical Summary

1. Problem Statement

The central question addressed is how phenotypic diversity, specifically in vertebral counts, evolves. While it is well-established that morphological evolution often arises from modifying the underlying developmental processes (e.g., changing the tempo of the segmentation clock, altering gene regulatory networks, or shifting the duration of morphogenesis), the role of developmental initial conditions has been largely overlooked.

According to dynamical systems theory, the initial state of a system can significantly dictate its trajectory and final outcome. However, in evolutionary developmental biology (evo-devo), initial conditions are rarely treated as an evolvable parameter distinct from the process itself. The authors hypothesize that phenotypic divergence (specifically in somite/vertebral count) might occur not by evolving the segmentation mechanism, but by evolving the initial size and state of the tissue at the onset of the process.

2. Methodology

The study utilizes a comparative approach between two closely related Lake Malawi cichlid species with distinct vertebral counts:

• Species A: Astatotilapia calliptera (putative ancestor, ~32 somites).
• Species B: Rhamphochromis sp. 'chilingali' (elongated species, ~38 somites).

Key Experimental Approaches:

• Temporal Quantification: Embryos were staged in hours post-fertilization (hpf) to determine the onset and duration of somitogenesis. Somite formation rates were calculated by counting somites over time.
• Morphometrics: Using high-resolution confocal microscopy and DAPI/phalloidin staining, the authors measured:
  • Pre-somitic mesoderm (PSM) volume, length, width, and depth.
  • Cell counts within the PSM.
  • Nascent somite size and volume.
  • Axial elongation rates.
• Statistical Analysis: Linear regression, ANOVA, and Principal Component Analysis (PCA) were used to compare growth trajectories and morphometric variables between species.
• Molecular Profiling: Hybridization Chain Reaction (HCR) was used to visualize and quantify the spatial expression of key genes involved in PSM morphogenesis and patterning (tbxta, tbx16, msgn, tbx6, MyoD) to determine if molecular mechanisms differed.

3. Key Results

A. Conservation of the Segmentation Process

• Rate: The tempo of somitogenesis (time per somite) is statistically identical between the two species (~65–67 minutes per somite).
• Mechanism: Both species form somites sequentially (one by one) starting from a single somite. There is no evidence of simultaneous segment formation or a change in the segmentation clock frequency.
• Molecular Patterning: The spatial expression patterns of key regulators (tbxta, tbx16, msgn, tbx6, MyoD) are indistinguishable between species at early and mid-segmentation stages. The "wavefront" velocity and nascent somite sizes are also conserved.

B. Divergence in Initial Conditions

• Onset Timing: R. sp. 'chilingali' initiates somitogenesis 5 hours later than A. calliptera (38 hpf vs. 33 hpf), indicating a heterochronic shift in earlier developmental stages (blastulation/gastrulation).
• PSM Size: At the onset of somitogenesis, the PSM in R. sp. 'chilingali' is significantly larger:
  • Volume: ~61% larger.
  • Cell Count: ~57% more cells.
  • Length: ~19% longer.
• Growth Trajectories: While the initial size differs, the growth rates (decay of PSM volume over time) and morphogenetic dynamics are parallel and statistically indistinguishable between the two species. The difference in PSM size is established before somitogenesis begins and persists as a constant offset rather than a divergent slope.

C. Causal Link to Vertebral Count

• The larger initial PSM in R. sp. 'chilingali' provides a larger reservoir of progenitor tissue. Because the rate of somite production is conserved, this larger initial tissue pool simply extends the total duration of the process (7 hours longer), resulting in the production of 6 additional somites without altering the segmentation mechanism itself.

4. Key Contributions

1. Identification of "Initial Conditions" as an Evolvable Parameter: The paper provides the first empirical evidence that phenotypic diversity (vertebral count) can arise solely from changes in the initial state (PSM size) of a developmental process, while the process itself remains evolutionarily conserved.
2. Decoupling Process from Outcome: It demonstrates that a 25% difference in somite count can be achieved without evolving the segmentation clock, the wavefront, or the morphogenetic mechanisms of axial elongation.
3. Refinement of Evo-Devo Models: The study challenges the assumption that morphological evolution requires the modification of the core developmental machinery (e.g., gene network architecture or kinetic rates). Instead, it highlights that "boundary conditions" or "initial states" set prior to the process can drive significant phenotypic divergence.

5. Significance

• Evolutionary Mechanism: This work proposes a parsimonious mechanism for rapid phenotypic evolution, particularly over short timescales (like the <1 million years of the Lake Malawi cichlid radiation). Evolving the initial conditions of a highly constrained process (like somitogenesis) may be evolutionarily "easier" than rewiring the complex genetic networks governing the process itself.
• Theoretical Implications: It bridges dynamical systems theory with evo-devo, suggesting that the "initial conditions" of a developmental system are a primary target for natural selection.
• Broader Applicability: The authors suggest this mechanism may apply to other traits where developmental processes are tightly coupled and constrained (e.g., the phylotypic stage). It also opens the door to investigating "boundary conditions" (spatial/structural constraints) as drivers of evolution, distinct from internal process modifications.

In conclusion, the study establishes that evolutionary change in vertebral count in these cichlids is driven by the evolution of the initial size of the pre-somitic mesoderm, rather than the evolution of the somitogenesis process itself.