Functional and compositional diversity peak at intermediate fire frequencies when modeling the plant-fire feedback

By modeling vegetation-fire feedbacks in Boreal and Mediterranean ecosystems, this study reveals that both compositional and functional plant diversity generally peak at intermediate fire frequencies, underscoring the critical role of these feedbacks in predicting biodiversity responses to global change.

The Gist

Imagine a forest or a shrubland as a giant, crowded apartment building where different plant species are the tenants. Some tenants are tough and can survive anything (like a fire), while others are delicate and need quiet, stable conditions.

This paper is like a computer simulation of what happens to this building when you introduce a recurring event: fire. The researchers wanted to answer a big question: How often should the building catch fire to have the most diverse and interesting mix of tenants?

Here is the breakdown of their findings using simple analogies:

1. The "Goldilocks" Zone of Fire

The researchers found that fire acts like a gymnast on a balance beam.

• Too little fire (The "Stagnant" Building): If the building never catches fire, the strongest, most aggressive tenants (usually tall trees or dominant shrubs) take over. They block the sunlight and push out the weaker, smaller plants. The building becomes a monoculture—boring and low in diversity.
• Too much fire (The "Chaotic" Building): If the building burns down every week, only the toughest, most fire-proof tenants can survive. The delicate ones die out before they can grow back. Again, diversity drops because only a few "survivors" remain.
• Just right (The "Sweet Spot"): The magic happens at intermediate fire frequencies. Fires happen often enough to knock down the bullies (the dominant species) and open up space for new, smaller tenants to move in, but not so often that they wipe out everyone. This is where you get the highest number of different species living together.

2. The "Feedback Loop" (The Self-Driving Car)

A key part of this study is that the plants and the fire talk to each other. It's not just a random event; it's a self-driving car.

• Some plants are like dry kindling (highly flammable). If they take over, they make fires happen more often.
• Other plants are like wet sponges (low flammability). If they take over, they stop fires from spreading.
• Some plants are like "resilient survivors" that can regrow after a fire, while others are like "seeders" that need the fire to release their babies.
  The model showed that this constant back-and-forth between what the plants are made of and how often they burn is what creates the perfect balance for diversity.

3. The Twist: "Richness" vs. "Variety"

The researchers looked at two types of diversity:

• Compositional Diversity: "How many different names are on the tenant list?" (Species Richness).
• Functional Diversity: "How different are their jobs and skills?" (Functional Diversity).

The Surprise: The peak for "how many tenants" and the peak for "how different their skills are" didn't happen at the exact same time.

• The Analogy: Imagine a band. You can have a band with 10 members (high richness), but if they all play the guitar, the music sounds the same (low functional diversity).
• The study found that to get the maximum number of species, you actually need some of them to be similar to each other. They need to share certain traits (like being able to handle fire) to coexist peacefully. If they were all too different, they would fight too hard or die out too easily. So, a little bit of "sameness" is actually necessary to pack the most species into the ecosystem.

4. Why This Matters for the Real World

We are currently living in an era where fires are becoming more frequent and intense due to climate change.

• The Warning: If we let fires happen too often (or if we suppress them completely for too long), we risk pushing our ecosystems out of that "Goldilocks zone." We might lose the delicate balance that allows diverse forests and shrublands to thrive.
• The Takeaway: Nature isn't just about stopping fires or starting them. It's about finding the rhythm. The study suggests that to protect biodiversity, we need to understand that fire is a natural part of the ecosystem's "heartbeat," and managing it requires keeping that heartbeat steady, not stopping it or speeding it up to a heart attack.

In short: Fire is a gardener. If it never prunes the garden, weeds take over. If it burns the garden every day, nothing grows. But if it prunes at the right pace, the garden becomes a vibrant, diverse paradise.

Technical Summary

1. Problem Statement

The relationship between fire regimes and biodiversity is complex and often debated. While fire is a critical driver of ecosystem composition, the exact nature of its impact on plant diversity (both compositional and functional) remains unclear due to the interplay of fire frequency, ecosystem type, and vegetation-fire feedbacks.

• The Gap: Existing theories, such as the Intermediate Disturbance Hypothesis (IDH), suggest diversity peaks at intermediate disturbance levels. However, empirical results are mixed, and many models treat fire as an exogenous (external) disturbance rather than an endogenous process driven by vegetation flammability and recovery strategies.
• The Challenge: There is a lack of comprehensive understanding regarding how functional diversity (variety of traits) and compositional diversity (species richness and abundance) interact across fire frequency gradients, specifically when the feedback loop between plant traits (flammability, fire response) and fire occurrence is explicitly modeled.

2. Methodology

The authors developed and extended a stochastic, spatially implicit competition-colonization model to simulate plant communities under varying fire regimes.

A. Model Framework

• Base Model: An extension of the Tilman (1994) competition-colonization model, previously adapted by Magnani et al. (2023) for three species. The authors scaled this to include a large number of species ($N$).
• Competition Dynamics: Species are ranked hierarchically ($i=1$ to $N$) based on competitive ability (colonization rate $C_i$ vs. mortality rate $M_i$). Superior species can displace inferior ones.
• Fire Dynamics:
  • Fires are modeled as stochastic events (Poisson process) where the fire return time ($T_f$) is inversely proportional to the total flammable vegetation cover.
  • Feedback Loop: The fire return time is endogenous; it depends on the community's composition (specifically the flammability $L_i$ of each species).
  • Fire Response: Each species has a fire response parameter ($R_i$), representing the fraction of cover surviving a fire (ranging from 0 for no response to 1 for full resprouting).
• Equations:
  • Growth: $\frac{db_i}{dt} = C_i b_i (1 - \sum_{j=1}^i b_j) - M_i b_i - b_i \sum_{j=1}^{i-1} C_j b_j$
  • Fire Return Time: $T_f = \frac{1}{\sum L_i b_i + \epsilon}$
  • Post-fire Cover: $b_i(T_f) = R_i b_i$

B. Experimental Design

• Ecosystems: Two distinct biomes were parameterized:
  1. Mediterranean: High flammability, frequent summer fires. Simulated with $N=10$ and $N=50$ initial species.
  2. Boreal: Conifer-dominated, lightning-ignited fires. Simulated with $N=10$ initial species.
• Parameterization: Colonization ($C$) and mortality ($M$) rates were derived from literature and expert data, then stochastically generated for $N$ species based on their hierarchical position. Fire traits ($L$ and $R$) were sampled from uniform distributions to create a wide variety of fire-return times.
• Simulations: 10,000+ runs per scenario (totaling ~64,000 simulations) with different initial conditions to capture alternative stable states.
• Metrics:
  • Compositional: Species Richness ($S$) and Inverse Simpson Index ($I$).
  • Functional: Functional Richness (FR, volume of trait space) and Functional Divergence (FD, density of trait space) calculated using hypervolumes based on traits $C$, $R$, and $L$.

3. Key Contributions

1. Endogenous Fire Modeling: Unlike previous studies that imposed fixed fire frequencies, this model captures the vegetation-fire feedback, where plant traits determine fire frequency, which in turn selects for specific traits.
2. High-Dimensional Trait Space: The model scales from 3 species to large pools ($N=50$), allowing for a more realistic representation of biodiversity and the exploration of complex trait combinations.
3. Decoupling of Diversity Types: The study explicitly investigates the relationship between compositional diversity (species count) and functional diversity (trait variety), revealing that they do not necessarily peak in the same communities.
4. Mechanistic Validation of IDH: The study provides a mechanistic explanation for the Intermediate Disturbance Hypothesis in fire-prone ecosystems, attributing the hump-shaped diversity curve specifically to the inclusion of vegetation-fire feedbacks.

4. Key Results

• Fire Promotes Diversity: Both Mediterranean and Boreal communities showed significantly higher compositional and functional diversity in the presence of fire compared to fire-free scenarios (Tilman model equilibria).
• Hump-Shaped Relationship:
  • Compositional Diversity ($S$ and $I$): Peaked at intermediate fire frequencies. Diversity was low at very low frequencies (dominance by late-successional competitors) and very high frequencies (only highly fire-resistant species survive).
  • Functional Diversity (FR and FD): Also generally peaked at intermediate frequencies, though the patterns were more variable across quantiles and ecosystem types compared to compositional diversity.
• Decoupling of Peaks: The communities with the maximum species richness did not coincide with the communities having the maximum functional diversity.
  • Implication: Maximum species richness may require a degree of functional similarity (species sharing similar fire response strategies) to coexist, whereas maximum functional diversity requires a broader spread of strategies that might limit the total number of coexisting species.
• Role of Feedback: When the vegetation-fire feedback was removed (fixed fire return time), the hump-shaped relationship disappeared, replaced by multimodal or monotonic relationships. This confirms that the feedback loop is the driver of the IDH pattern in these systems.

5. Significance and Implications

• Theoretical Insight: The study resolves inconsistencies in previous literature by demonstrating that the IDH pattern in fire-prone ecosystems is contingent on the endogenous nature of fire. Without vegetation-fire feedback, the classic hump-shaped diversity curve may not emerge.
• Management Relevance: As global change alters fire regimes (increasing frequency and intensity), understanding these feedbacks is crucial.
  • Fire Suppression: May lead to a loss of diversity by pushing ecosystems toward low-frequency states dominated by competitive exclusion.
  • Prescribed Burning: Managing fire at intermediate frequencies could optimize both species richness and functional resilience.
• Modeling Future: The findings suggest that global vegetation models (GVMs) must incorporate dynamic vegetation-fire feedbacks (flammability and post-fire recovery) to accurately predict biodiversity trends under climate change. Ignoring these feedbacks may lead to erroneous projections of ecosystem collapse or biodiversity loss.

In conclusion, the paper demonstrates that intermediate fire frequencies maximize biodiversity in fire-prone ecosystems, but this peak is driven by a complex interplay where compositional and functional diversity are correlated yet distinct, and the mechanism relies fundamentally on the feedback between plant traits and fire occurrence.