Biodiversity effects on ecosystem functioning: disentangling the roles of biomass and effect trait expression

This study proposes and applies a novel decomposition of the net biodiversity effect to disentangle how plant diversity influences ecosystem functioning through two distinct pathways: total community biomass and the expression of effect traits, revealing that these mechanisms can exert contrasting or opposing impacts on specific functions like nitrogen retention, soil hydraulic conductivity, and forage digestibility.

The Gist

The Big Picture: Why Does a Mixed Garden Work Better?

Imagine you have a garden. You can plant it with just one type of flower (a monoculture), or you can plant a wild mix of flowers, grasses, and herbs (a diverse community).

For a long time, ecologists knew that diverse gardens usually produce more total food (biomass) than single-plant gardens. They figured out why: some plants are just naturally better at growing, and when you mix them, the best ones take over (Selection), or the plants help each other use sunlight and water more efficiently (Complementarity).

But here is the problem: Ecosystems do more than just grow plants. They also clean water, hold soil together, and provide nutritious food for cows. These are different "functions."

This paper asks a new question: Does biodiversity help these other functions in the same way it helps total growth? Or, does it change the quality of what the plants do, independent of how much they grow?

The Two Ways Biodiversity Works

The authors propose that biodiversity affects nature in two distinct ways, like a chef running a restaurant:

1. The "Quantity" Effect (Total Biomass): This is about how much food the kitchen produces. If you have more chefs (plants) working together, you get more meals. This is driven by how the plants interact to grow bigger.
2. The "Quality" Effect (Effect Trait Expression): This is about the flavor or nutritional value of the food. Even if you have the same amount of food, a diverse kitchen might produce a more nutritious meal because the chefs are using different techniques or ingredients that change the outcome.

The paper creates a mathematical "recipe" to separate these two effects. They want to know: Is the garden better because it's bigger, or because the plants are doing something smarter per unit of weight?

The Four Ingredients of the Recipe

The authors break down the "Net Biodiversity Effect" (the total benefit of having a mix) into four specific ingredients:

1. Selection on Quantity (The "Star Player" Effect):

   • Analogy: Imagine a sports team. If you mix a team of average players with one superstar, the team wins mostly because that one superstar dominates the game.
   • Science: The most productive plant species takes over the garden, making the whole thing bigger.

2. Complementarity on Quantity (The "Teamwork" Effect):

   • Analogy: Imagine a relay race where everyone runs a different leg. They don't just run faster individually; they pass the baton perfectly so the whole team goes further than the sum of their parts.
   • Science: Plants use different resources (like deep vs. shallow roots) so they don't fight, allowing the whole garden to grow bigger than expected.

3. Selection on Quality (The "Wrong Star" Effect):

   • Analogy: Imagine you need a team of accountants. You hire a mix of people, but the one who dominates the team is actually a great artist but a terrible accountant. The team gets bigger, but the quality of the accounting drops.
   • Science: A plant that grows huge might not be good at the specific job you need (like holding nitrogen). If it takes over, the whole community's "quality" drops.

4. Complementarity on Quality (The "Magic Mix" Effect):

   • Analogy: Imagine mixing red and blue paint. Individually, they are just colors. But mixed together, they create purple—a new color that neither could make alone.
   • Science: When plants grow together, they might change their behavior (plasticity) or interact in a way that makes the per-unit quality of the ecosystem better (e.g., the soil becomes more permeable just because the roots are mixed).

What They Found: Three Real-World Examples

The researchers tested this on a greenhouse garden with grasses and herbs. They looked at three different "jobs" the plants had to do:

1. Holding onto Nitrogen (Cleaning the Water)

• The Result: The diverse garden held onto nitrogen better because the plants grew bigger (Quantity Effect).
• The Twist: However, the quality of the plants actually got worse.
• Why? The plants that grew the biggest were legumes (like beans). They are great at growing, but they are bad at holding onto nitrogen per gram of weight because they get their nitrogen from the air, not the soil. So, while the garden was huge, the plants inside were "lazy" at cleaning the water. The "Quantity" win was almost cancelled out by the "Quality" loss.

2. Letting Water Flow Through Soil (Hydraulic Conductivity)

• The Result: The diverse garden let water flow through the soil much better.
• The Reason: This was entirely because the garden was bigger. More plants = more roots = more holes in the soil for water to pass through.
• The Twist: There was no "Quality" magic. The plants didn't change their behavior; they just grew more.

3. Making Good Food for Cows (Forage Digestibility)

• The Result: The diverse garden produced food that was harder to digest.
• The Reason: This was a battle between two forces.
  • Complementarity (Good): When mixed, the plants adjusted their traits to make the food slightly better.
  • Selection (Bad): The plants that took over the garden were the ones that made the worst food (tough, fibrous grasses).
• The Outcome: The "Bad Star" effect was stronger than the "Magic Mix" effect, so the final food quality dropped.

The Takeaway

This paper teaches us that biodiversity is a double-edged sword.

Just because a diverse ecosystem is "bigger" doesn't mean it's "better" at every job. Sometimes, the plants that grow the fastest are the worst at the specific task we need (like cleaning water or feeding cows).

By separating the size of the community from the quality of its work, scientists can finally understand why some diverse ecosystems fail to perform certain tasks. It's not just about having a big garden; it's about having the right mix of plants doing the right things, even if they aren't the biggest ones.

In short: Diversity doesn't just make nature bigger; it changes how nature works. Sometimes that's great, and sometimes, the plants that take over are the wrong ones for the job.

Technical Summary

1. Problem Statement

While the positive relationship between plant biodiversity and total community biomass (primary productivity) is well-established, the mechanisms by which biodiversity influences other ecosystem functions remain poorly understood.

• The Gap: Current Biodiversity and Ecosystem Functioning (BEF) research often conflates two distinct pathways:
  1. Quantity: Biodiversity increasing total biomass (driven by response and interaction traits).
  2. Quality: Biodiversity altering the mass-specific contribution of species to a function (driven by effect traits).
• The Challenge: Standard BEF partitioning (e.g., Loreau & Hector, 2001) decomposes Net Biodiversity Effects (NBE) into selection and complementarity effects on biomass. However, for functions like nutrient retention or forage quality, the "mass-specific contribution" (effect trait expression) may change due to phenotypic plasticity or species interactions, independent of total biomass. Existing frameworks struggle to disentangle whether a change in ecosystem function is due to having more biomass or different functional performance per unit of biomass.

2. Methodology

Theoretical Framework

The authors propose a new decomposition of the Net Biodiversity Effect (NBE) that separates impacts on total biomass from impacts on effect trait expression.

1. Definition of Biomass-Scalable Functions ($\Phi$):

   • The authors define a "function" $\Phi$ as the community's contribution to an ecosystem variable $V$.
   • They assume $\Phi$ is a biomass-scalable function: $\Phi = c \times \Sigma B$, where $\Sigma B$ is total community biomass and $c$ is the mass-specific community contribution (the average effect trait expression per unit of biomass).
   • Note: For variables not naturally proportional to biomass (e.g., soil N concentration), the authors provide a mechanistic conversion (Box 1) to derive a biomass-scalable function (e.g., converting N concentration to N retention capacity).

2. Decomposition of NBE:
   The NBE is defined as the ratio of observed functioning ($\Phi$) to expected functioning based on monocultures ($\Phi_E$). The authors derive a multiplicative decomposition (Eq. 7) into four distinct components:

   $$NBE = \underbrace{\frac{m(\eta)}{m(\eta_E)}}_{CE} \times \underbrace{\frac{m(K, \eta)}{m(K, \eta_E)}}_{SE} \times \underbrace{\frac{m(c, B)}{m(c_E, B)}}_{CETE} \times \underbrace{\frac{m(c_E, B)}{m(c_E, B_E)}}_{SETE}$$

   • CE (Complementarity Effect on Biomass): Reflects whether species coexist to produce more total biomass than expected (sum of relative yields > 1).
   • SE (Selection Effect on Biomass): Reflects whether high-biomass monoculture species dominate the polyculture.
   • CETE (Complementarity Effect on Effect Trait Expression): Reflects non-additive changes in mass-specific contributions (e.g., phenotypic plasticity or synergistic trait interactions) independent of biomass shifts.
   • SETE (Selection Effect on Effect Trait Expression): Reflects whether species with high mass-specific functional potential (in monoculture) dominate the polyculture.

Experimental Application

• Data Source: An experiment with 8 perennial grassland species grown in monocultures and 6-species polycultures (28 communities) for 15 months.
• Functions Analyzed:
  1. Nitrogen (N) Retention Capacity: Derived from N leachate concentrations (converted to a biomass-scalable function).
  2. Soil Hydraulic Conductivity Improvement: Derived from changes in soil hydraulic conductivity relative to bare soil.
  3. Forage Digestibility: Treated as a mass-specific trait (humid matter content) rather than a biomass-scalable function, allowing analysis of CETE and SETE only.

3. Key Contributions

1. Novel Partitioning Framework: The paper introduces a mathematical formalism that explicitly separates biodiversity effects on functional quantity (biomass) from functional quality (effect trait expression). This resolves a long-standing ambiguity in BEF literature.
2. Mechanistic Conversion of Variables: It provides a rigorous method (Box 1) to transform raw ecosystem variables (like soil nutrient pools) into biomass-scalable functions, preventing artifactual biodiversity effects caused by non-linear relationships between biomass and ecosystem variables.
3. Four-Component Decomposition: By extending the classic Loreau & Hector (2001) partition, the authors identify four distinct mechanisms (CE, SE, CETE, SETE) rather than just two, allowing for the detection of counteracting pathways (e.g., positive biomass effects masked by negative trait effects).

4. Results

The application of this framework to the three ecosystem functions revealed contrasting mechanisms:

• Nitrogen (N) Retention Capacity:
  • Net Effect: Neutral (NBE $\approx$ 1).
  • Mechanism: A positive Complementarity Effect on Biomass (CE) (more biomass in mixtures) was completely offset by negative effects on trait expression.
  • Specifics: Both CETE and SETE were negative. Species with low N retention per unit of biomass (likely N-fixing legumes that rely less on soil N uptake) became dominant (negative SETE), and phenotypic adjustments in rich communities reduced per-unit uptake rates (negative CETE).
• Soil Hydraulic Conductivity Improvement:
  • Net Effect: Positive (NBE > 1).
  • Mechanism: Driven almost entirely by total biomass (positive CE and SE).
  • Specifics: There were no significant biodiversity effects on effect trait expression (CETE and SETE were near 1). Increased root density from higher biomass improved soil permeability, but the functional traits themselves did not change synergistically.
• Forage Digestibility:
  • Net Effect: Complex (Positive trait complementarity outweighed by negative selection).
  • Mechanism: Since digestibility is independent of total biomass, only trait expression components were analyzed.
  • Specifics: A strong positive CETE indicated that species richness enhanced digestibility via trait adjustments (likely due to light competition). However, a negative SETE showed that the community dynamics favored species with lower inherent digestibility, partially canceling out the positive complementarity.

5. Significance

• Resolving Contradictions: The framework explains how biodiversity can have a net neutral effect on an ecosystem function (like N retention) while simultaneously driving strong positive and negative mechanisms that would remain hidden in a standard analysis.
• Trait-Mediated Insights: It highlights that biodiversity effects are not just about "more plants" but also about "different plants" and "plastic plants." It distinguishes between effects driven by interaction traits (which determine biomass) and effect traits (which determine the function per unit of biomass).
• Predictive Power: By separating quantity and quality, the method improves the ability to predict ecosystem responses to global change. For instance, if a function relies heavily on specific effect traits (like N retention), simply increasing biomass (via diversity) may not improve the function if the dominant species in the mixture have low mass-specific performance.
• Future Directions: The authors suggest that coupling this partitioning with direct measurements of functional traits (e.g., root length density, specific leaf area) will further refine the understanding of the mechanistic links between biodiversity, traits, and ecosystem services.

In conclusion, this paper provides a critical methodological advancement for BEF research, moving beyond the "biomass-only" paradigm to a more nuanced understanding of how biodiversity drives ecosystem functioning through both the quantity of biomass and the quality of trait expression.