Strip a few species from an ecosystem and it may barely flinch. Remove a few more, and something shifts. Take away enough, and the whole system can unravel. The relationship between how many species occupy a system and how well that system functions has been one of ecology's most debated questions for three decades.

It matters far beyond academic curiosity. Ecosystem services—pollination, water purification, carbon storage, disease regulation—underpin human economies and well-being. If biodiversity loss degrades these services, the argument for conservation gains an economic backbone that complements any ethical one.

But the relationship is not as straightforward as more species equals better performance. The evidence reveals a nuanced picture: biodiversity generally enhances ecosystem functioning, but through specific mechanisms that operate differently across contexts. Understanding why diversity matters—not just whether it does—changes how we think about managing ecological systems.

The modern biodiversity-ecosystem function (BEF) research program launched in the 1990s with landmark experiments like Cedar Creek and BIODEPTH. The design was deceptively simple: plant communities with different numbers of species, control everything else, and measure ecosystem processes like biomass production, nutrient cycling, and decomposition. The results, replicated across hundreds of experiments worldwide, pointed in a consistent direction.

More species generally means higher ecosystem functioning. Plots with greater plant diversity produced more biomass, captured more light, cycled nutrients more completely, and resisted invasion more effectively. A major meta-analysis by Cardinale and colleagues in 2012 synthesized 574 experiments and confirmed that species loss reduces the efficiency by which communities capture resources and produce biomass. The signal was robust across trophic levels and ecosystem types.

But here's the critical detail: the relationship saturates. The biggest functional gains come from adding the first few species to a monoculture. Each additional species contributes progressively less. Plotted on a graph, functioning rises steeply at low diversity, then flattens into an asymptote. This means that in a species-rich system, losing one or two species may produce no detectable decline—a phenomenon sometimes called functional redundancy. The system has backup.

This saturating curve carries a dangerous implication. It can lull us into thinking diversity loss is tolerable because the first losses don't seem to matter. But the curve also means we cannot predict where the threshold lies—how many species we can afford to lose before the system crosses from the flat portion into the steep decline. Long-term experiments at Cedar Creek showed that diversity effects actually strengthen over time, suggesting short-term studies may underestimate the true functional importance of richness.

Takeaway

The relationship between biodiversity and ecosystem function resembles an insurance curve: redundancy buffers the system against early losses, but that same buffer makes it dangerously easy to underestimate how close the system is to a tipping point.

Establishing that diversity matters was only half the puzzle. The harder question was why. Two competing mechanisms emerged, and distinguishing between them carries real consequences for how we interpret and apply BEF results.

The first mechanism is the sampling effect (sometimes called the selection effect). In a more diverse mixture, there is a higher probability of including the single most productive species. If that dominant species drives the community's performance, then diversity is just a statistical lottery—a vehicle for ensuring the best competitor shows up. Under this interpretation, what matters is not diversity per se but the identity of key species. Lose the dominant performer and function collapses regardless of how many other species remain.

The second mechanism is complementarity, which encompasses both niche partitioning and facilitation. Niche partitioning means that different species exploit different portions of available resources—one grass roots deeply while another roots shallowly, or species peak in activity during different seasons. Facilitation occurs when one species actively enhances conditions for others, like nitrogen-fixing legumes enriching the soil for neighboring plants. Under complementarity, the mixture outperforms even its best component because species do different things, and those differences compound.

Additive partitioning methods developed by Loreau and Hector in 2001 allowed researchers to separate these effects statistically. The verdict across many experiments: both mechanisms operate, but complementarity tends to dominate in longer-running experiments and in more realistic community assemblages. This is a critical finding for conservation. If sampling effects were all that mattered, we could protect function by safeguarding a handful of key species. But if complementarity drives function—and the evidence increasingly says it does—then the diversity itself is the mechanism, not a proxy for it.

Takeaway

When species perform different ecological roles, the whole becomes greater than the sum of its parts. Protecting function requires protecting the portfolio of differences, not just the star performers.

BEF research offers conservation a powerful new argument. Historically, biodiversity conservation has rested on intrinsic value—species deserve to exist—or on charismatic appeal. These arguments resonate with some audiences but fail to move policy in contexts dominated by economic cost-benefit logic. The BEF framework translates biodiversity into the language of ecosystem services and, by extension, into economic terms.

If diverse ecosystems purify water more reliably, sequester more carbon, resist disease outbreaks more effectively, and maintain productivity under variable climates, then biodiversity loss carries quantifiable costs. This insurance hypothesis—that diversity stabilizes ecosystem services against environmental fluctuation—has received strong empirical support. Diverse grasslands, for instance, show less year-to-year variability in biomass production than species-poor ones, precisely because different species respond differently to drought, heat, and other stressors.

But this utilitarian framing carries risks that ecologists openly debate. If we value species only for their functional contributions, we implicitly accept that functionally redundant species are expendable. The saturating BEF curve could be misread as permission to tolerate moderate diversity loss. Moreover, laboratory and plot-scale experiments may not capture the full complexity of real landscapes, where spatial heterogeneity, trophic interactions, and long time scales magnify the importance of rare and seemingly redundant species.

The most defensible position synthesizes both arguments. Biodiversity has intrinsic value and it underpins the ecosystem services that human societies depend on. These are not competing claims—they are reinforcing ones. The BEF literature gives conservation advocates a tool for engaging stakeholders who respond to utilitarian logic, while the ethical argument guards against the reductionism that purely economic framing can invite. Effective ecosystem management requires maintaining diversity not as a luxury, but as the operating infrastructure of the systems we depend on.

Takeaway

Framing biodiversity as ecosystem infrastructure—not ornament—gives conservation an economic argument, but the strongest case welds utilitarian and intrinsic value together so that neither can be dismissed alone.

The BEF evidence is clear in its broad message: species richness generally enhances ecosystem functioning, and the mechanisms behind this relationship—particularly complementarity—mean that diversity itself is doing real ecological work.

The saturating curve should not comfort us. It tells us that systems can absorb some loss, but it hides the threshold beyond which decline accelerates. We rarely know where that threshold lies until we've crossed it.

For ecosystem management, the practical lesson is systems-level: maintain the diversity of functional roles, protect against homogenization, and treat species richness as infrastructure rather than inventory. The feedback loops that stabilize ecosystems depend on the variety of parts as much as on any single one.