Disentangling site-specific and shared local adaptation in a classic system of repeated evolution

By employing an expanded transplant framework on lake-stream stickleback, this study reveals that site-specific local adaptation contributes more to fitness variation than shared habitat-level adaptation, challenging the assumption that repeated divergence alone captures the full scope of evolutionary adaptation.

The Gist

Imagine you are a chef trying to figure out what makes a restaurant successful. You have two main types of restaurants: Seafood Shacks (by the ocean) and Mountain Cabins (in the woods).

For years, scientists have studied these restaurants by comparing the Seafood Shacks to the Mountain Cabins. They noticed that Seafood Shacks always serve great fish, and Mountain Cabins always serve great trout. They concluded: "Ah! The ocean makes you good at fish, and the mountains make you good at trout. This is a universal rule!"

But here is the problem: Not all Seafood Shacks are the same. One might be on a rocky cliff, another in a calm bay, and a third on a sandy beach. Do they all perform exactly the same? Or does the "Rocky Cliff" restaurant have a secret sauce that the "Calm Bay" restaurant lacks?

Traditional science often ignores these small differences, assuming they are just "noise" or mistakes. This paper says: "Wait a minute. Those 'mistakes' might actually be the most important part of the story."

The Experiment: A Real-World Taste Test

To test this, the researchers (Marius Roesti and his team) didn't just look at the menus; they actually moved the chefs to see how they performed in different kitchens.

They used Stickleback fish (tiny, spiny fish) from a big lake and four different streams.

1. The Setup: They raised baby fish in a lab (like a cooking school) so they all had the same upbringing. This ensured that any differences later were due to their "genetic recipe," not their childhood.
2. The Swap: They took these fish and put them into real enclosures in:
   • The big Lake (the "Seafood Shack" environment).
   • Two different Streams (two different "Mountain Cabin" environments).
3. The Groups: In every enclosure, they mixed:
   • Locals: Fish born in that specific spot.
   • Foreign Stream Fish: Fish from a different stream (same habitat type, but different location).
   • Foreign Lake Fish: Fish from the big lake (a totally different habitat).

The Results: The "Home Court" Advantage

The results were surprising and changed how we think about evolution.

1. The Big Picture (Lake vs. Stream):
As expected, the Lake fish did great in the Lake, and the Stream fish did great in the Streams. This confirmed that fish adapt to the general difference between lakes and rivers.

2. The Twist (Stream vs. Stream):
Here is where it gets interesting. When they put a fish from Stream A into Stream B, it did okay. But when they put a fish from Stream A into its own Stream A, it did amazingly well.

In fact, the advantage of being a "local" fish in its own specific stream was twice as big as the advantage of being a stream fish compared to a lake fish.

The Analogy: The "Special Sauce"

Think of it like this:

• Shared Adaptation (The Big Difference): Being a "Mountain Chef" is better than being a "Sea Chef" in the mountains. This is the 33% of the success story.
• Site-Specific Adaptation (The Local Secret): But being the specific "Chef of the Rocky Creek" is twice as important as just being a "Mountain Chef" in general. This is the 67% of the success story.

The fish from Stream A had evolved a "secret sauce" specifically for the rocks, water flow, and bugs of their specific creek. When they moved to Stream B, they lost that edge, even though Stream B was also a "stream."

Why Does This Matter?

For a long time, scientists thought that if you saw differences between two fish from the same type of habitat (like two different streams), it was just random noise or bad luck. They thought, "Oh, they're both stream fish, so they should be the same."

This paper says: No, they aren't the same.

• Evolution is hyper-local: Nature doesn't just adapt you to "a stream"; it adapts you to your specific stream.
• Conservation: If we lose one specific stream population, we aren't just losing "a fish." We are losing a unique, irreplaceable evolutionary masterpiece that can't be replaced by fish from the next valley over.
• Predictability: We can't just predict evolution by looking at big categories (Lake vs. Stream). We have to look at the tiny, local details to understand how life evolves.

The Bottom Line

This study is like realizing that while all "Italian Pizzas" are better than "Sushi" in Italy, the best pizza is the one made by the specific chef in the specific neighborhood who knows exactly how the local oven and local flour work.

The "local" adaptation is the real hero, and it's been hiding in plain sight, mistaken for random noise, for decades.

Technical Summary

1. Problem Statement

• The Limitation of Repeated Divergence Studies: Evolutionary biologists traditionally infer local adaptation by observing repeated phenotypic or genetic divergence between populations in contrasting habitats (e.g., lake vs. stream). However, this approach only detects shared adaptation (traits common to all populations within a habitat type) and inherently misses site-specific adaptation (unique adaptations to local conditions within the same habitat type).
• The "Noise" vs. Adaptation Dilemma: Differences between populations in similar habitats are often dismissed as non-adaptive noise (drift, constraints) or alternative solutions to the same selection pressure. However, these differences may actually represent adaptive responses to unique, fine-scale environmental conditions.
• Limitations of Traditional Transplant Experiments: Standard reciprocal transplant experiments typically involve pairwise exchanges between two divergent habitats. While effective at measuring overall local adaptation across a habitat divide, they cannot disentangle the fitness contributions of shared habitat-level adaptation versus site-specific adaptation among populations within the same habitat type.

2. Methodology

The authors developed an extended multi-population transplant framework to overcome these limitations, applied to the threespine stickleback (Gasterosteus aculeatus) system in the Lake Constance watershed (Western Europe).

• Study System:
  • Lake Population: A single, panmictic (well-mixed) lake population.
  • Stream Populations: Four independently evolving inlet stream populations (NID, RUG, GRA, OBE).
• Experimental Design:
  • Common-Garden Rearing: To isolate heritable evolutionary differences from plasticity, the authors reared fish for two generations ($F_2$) in a controlled laboratory environment. This eliminated maternal effects and environmental carryover.
  • Transplant Sites: Three field sites were used: one lake site and two independent stream sites (NID and RUG).
  • Stocking Strategy: 55 enclosures were established across the three sites. Each enclosure contained a balanced mixture of 13 fish:
    • Lake Site: Local lake fish + Foreign stream fish (from all 4 streams).
    • Stream Sites: Local stream fish + Foreign stream fish (from the other 3 streams) + Foreign lake fish.
  • Total Sample: 715 individually marked fish were stocked and monitored for approximately seven weeks.
• Fitness Metrics:
  • Survival: Binary outcome (alive/dead) after 7 weeks.
  • Body Condition: Fulton's condition index ($K$), calculated as mass/length$^3$, measured at the start and end of the experiment.
  • Composite Fitness: A metric combining predicted survival probability and predicted post-transplant body condition.
• Statistical Analysis:
  • Qualitative Inference: Used Generalized Linear Mixed Models (GLMM) for survival and Linear Mixed Models (LMM) for body condition to test for local vs. foreign advantages.
  • Decomposition of Adaptation: The study partitioned total local adaptation in streams into two components:
    1. Shared Adaptation: Fitness advantage of foreign stream fish over foreign lake fish (adaptation to stream conditions generally).
    2. Site-Specific Adaptation: Fitness advantage of local stream fish over foreign stream fish (adaptation to specific site conditions).

3. Key Contributions

• Novel Experimental Framework: The study extends the classical reciprocal transplant design by including multiple populations from the same habitat type alongside a contrasting habitat. This allows for the explicit mathematical separation of shared habitat-level adaptation from site-specific adaptation.
• Direct Fitness Measurement: By using common-garden reared $F_2$ individuals, the study provides direct evidence of evolutionary (heritable) adaptation rather than plastic responses, addressing a major gap in previous stickleback transplant studies that often used wild-caught fish.
• Quantification of Non-Parallelism: The study provides a quantitative framework to determine whether non-parallel divergence (differences between similar habitats) is adaptive or non-adaptive.

4. Key Results

• Evidence for Lake-Stream Divergence: As expected, stream fish outperformed lake fish in stream habitats, and lake fish outperformed stream fish in the lake habitat, confirming broad-scale local adaptation.
• Evidence for Site-Specific Adaptation: Crucially, local stream fish significantly outperformed foreign stream fish at both stream sites. This demonstrates that populations within the same nominal habitat type are uniquely adapted to their specific local sites.
• Magnitude of Effects:
  • The fitness advantage of local stream fish over foreign stream fish (site-specific adaptation) was twice as large as the advantage of foreign stream fish over foreign lake fish (shared stream adaptation).
  • Decomposition of Total Fitness: When analyzing the total fitness advantage of local stream fish over foreign lake fish:
    • 67% was attributed to site-specific adaptation (unique local conditions).
    • 33% was attributed to shared adaptation (general stream vs. lake divergence).
• Survival and Condition: These patterns held true for both survival rates (highest at the lake site, lowest for lake fish in streams) and body condition, though the effect sizes were slightly smaller for body condition than for survival.

5. Significance and Implications

• Re-evaluating "Noise": The study suggests that substantial variation often dismissed as non-adaptive "noise" in comparative studies is actually adaptive, driven by fine-scale ecological heterogeneity.
• Conservation Value: If most fitness-relevant variation is site-specific, then populations within the same habitat type represent distinct evolutionary outcomes with independent conservation value, rather than redundant replicates.
• Evolutionary Predictability: The findings challenge the view that evolution is highly predictable based solely on broad ecological contrasts. While adaptation to broad habitat types (lake vs. stream) is repeatable, the specific adaptive trajectories are highly idiosyncratic to local conditions.
• Methodological Shift: The paper argues that relying solely on repeated divergence to infer adaptation is insufficient. Future studies must incorporate multi-population transplant designs to fully capture the spectrum of local adaptation, from broad habitat-level selection to fine-scale site-specific selection.