Niche Complementarity Hypothesis

Niche Complementarity Hypothesis is a foundational principle in community and ecosystem ecology proposing that multispecies assemblages capture limiting resources more completely and achieve higher aggregate rates of primary production, nutrient retention, and biomass accumulation than single-species monocultures because coexisting species occupy differentiated ecological niches or engage in positive interspecific facilitation. Tracing its conceptual roots to Charles Darwin's Principle of Divergence (1859) and classic Lotka–Volterra coexistence theory, the hypothesis was formalized within modern biodiversity–ecosystem functioning (BEF) science during the 1990s and early 2000s by David Tilman, John Lawton, Shahid Naeem, Michel Loreau, and Andy Hector. In 2001, Loreau and Hector (Nature) introduced an algebraic partitioning equation—analogous to the Price equation in evolutionary genetics—that decomposes the net biodiversity effect (ΔY) of a species mixture into two additive components: the Complementarity Effect (N ΔRY M̄), driven by niche differentiation and interspecific facilitation, and the Selection Effect (N Cov(ΔRY, M)), driven by competitive dominance of species with high monoculture yields (closely linked to the mass-ratio hypothesis).

In Chapter 4 of On the Origin of Species (1859), Charles Darwin noted that an experimental plot of turf sown with several distinct genera of grasses yielded a greater weight of dry herbage than a uniform plot sown with a single grass species—an early statement of overyielding driven by structural and physiological divergence among coexisting plants. During the mid-twentieth century, agronomic intercropping researchers (such as C. T. de Wit, 1960) quantified this phenomenon using the Relative Yield Total (RYT) and Land Equivalence Ratio (LER) in cereal–legume polycultures, showing that mixtures regularly exceed the weighted mean yield of their component monocultures (RYT > 1).

During the 1990s, concern over global species loss prompted controlled manipulative experiments testing whether declining species richness directly impairs ecosystem processes:

• The Ecotron facility experiment and the Cedar Creek grassland biodiversity experiments in Minnesota (Tilman, Wedin, and Knops, 1996) established replicated plots spanning 1 to 16 or 32 plant species.
• Across eight European countries, the BIODEPTH project and later the Jena Experiment in Germany and BEF-China subtropical forest experiment demonstrated a consistent positive scaling of aboveground biomass with the logarithm of species richness (Y = a + b ln S).

A fiery scientific debate (the late-1990s "biodiversity debate" led by Michael Huston, 1997, and Lonnie Aarssen, 1997) challenged whether this positive richness–productivity curve reflected genuine niche complementarity among species or merely a statistical sampling (selection) artifact: plots seeded with more species had a higher chance of including a single tall, fast-growing species that competitively dominated the plot. Resolving this controversy required both mechanistic resource-competition models (Tilman, Lehman, and Thomson, 1997) and a rigorous mathematical partitioning method (Loreau and Hector, 2001).

In contemporary synthesis (Barry et al., 2019), complementarity comprises four biological mechanisms that relax intraspecific competition relative to interspecific competition (α_ii > α_ij):

1. Spatial Resource Partitioning

Coexisting species extract resources from distinct vertical strata:

• Belowground rooting depth stratification: Deep taproot forbs (Lupinus, Centaurea) draw water and subsoil nitrate (NO₃⁻) from 30–120 cm depth, whereas fibrous-rooted C₃ and C₄ grasses concentrate fine roots in the upper 0–15 cm topsoil.
• Aboveground canopy architecture: Tall sun-adapted canopy plants with steep leaf angles transmit diffuse photosynthetically active radiation (PAR) to shade-tolerant understory plants, maximizing stand light capture.

2. Temporal (Phenological) Partitioning

Species with staggered seasonal phenologies partition uptake across the growing season. Cool-season C₃ grasses (Poa pratensis, Agropyron smithii) assimilate soil nitrogen and moisture in spring (April–May), whereas warm-season C₄ grasses (Schizachyrium scoparium, Andropogon gerardii) peak in mid-to-late summer (July–August), lengthening the community's seasonal window of primary production.

3. Chemical and Trophic Resource Partitioning

Within the same rooting zone, plants partition chemical forms of nitrogen and phosphorus. Tracer studies using ¹⁵N-labeled substrates show that co-occurring tundra and grassland species specialize on nitrate (NO₃⁻), ammonium (NH₄⁺), or dissolved free amino acids (glycine), while differing in phosphatase enzymes that cleave specific soil organic phosphates.

4. Biotic Facilitation and Pathogen Dilution

Positive interspecific interactions actively increase resource supply or survival:

• Abiotic facilitation and niche construction: Symbiotic dinitrogen (N₂)-fixing legumes (Rhizobium nodules) enrich soil mineral nitrogen for neighboring grasses; hydraulic lift by deep-rooted woody plants moistens surface soil; and cluster-rooted species exude carboxylates (citrate, malate) that mobilize sorbed soil phosphorus.
• Janzen–Connell Host Dilution Effect: In monocultures, host-specific soil oomycetes (Pythium, Phytophthora), nematodes, and foliar rusts accumulate to growth-suppressing densities. In mixtures, physical spacing of conspecific hosts dilutes specialist pathogen transmission, releasing plants from negative soil feedback.

To separate niche complementarity from the sampling/selection effect in experimental mixtures, Michel Loreau and Andy Hector (2001) formulated a rigorous additive decomposition of the Net Biodiversity Effect (ΔY). Consider a mixture sown with N species in initial proportions RY_E,i (typically equal proportions RY_E,i = 1/N, with ∑ RY_E,i = 1). Let:

• M_i = yield (biomass per unit area) of species i grown in a pure monoculture;
• Y_O,i = observed yield of species i within the multispecies mixture, so total observed mixture yield is Y_O = ∑_i=1^N Y_O,i;
• Y_E,i = RY_E,i M_i = expected yield of species i in mixture under the null hypothesis that per-capita interspecific and intraspecific interactions are identical, so total expected mixture yield is Y_E = ∑_i=1^N RY_E,i M_i;
• RY_O,i = Y_O,i / M_i = observed relative yield of species i in mixture;
• ΔRY_i = RY_O,i − RY_E,i = deviation of species i's relative yield from its sown proportion.

The Net Biodiversity Effect ΔY is the difference between observed total mixture yield and expected mixture yield:

ΔY = Y_O − Y_E = ∑_i=1^N Y_O,i − ∑_i=1^N RY_E,i M_i = ∑_i=1^N ΔRY_i M_i

Using the standard statistical identity connecting the sum of products of two variables to their means and sample covariance, Loreau and Hector (2001) obtained the additive partitioning equation:

ΔY = N · ΔRY_mean · M_mean + N · Cov(ΔRY_i, M_i)

where ΔRY_mean = (1/N) ∑ ΔRY_i and M_mean = (1/N) ∑ M_i.

Non-Transgressive vs. Transgressive Overyielding

Ecological theory distinguishes two strictness tiers of mixture performance:

1. Non-transgressive overyielding (Y_O > Y_E): The mixture yields more biomass than the weighted average of its constituent monocultures (ΔY > 0). Either a positive Selection Effect (SE > 0) or a positive Complementarity Effect (CE > 0) can produce non-transgressive overyielding.
2. Transgressive overyielding (Y_O > max_i(M_i)): The mixture yields more biomass than the single most productive monoculture among all constituent species. Because a pure selection effect can at most allow the top monoculture species to take over 100 percent of the mixture (yielding Y_O → max(M_i)), true transgressive overyielding is unambiguous proof of positive niche complementarity or facilitation (CE > 0).

One of the most striking empirical findings of long-term biodiversity experiments (Cardinale et al., 2007; Reich et al., 2012) is that the relative importance of the Selection Effect and the Complementarity Effect shifts systematically over time:

• Years 1–2 (Establishment Phase): In newly sown grassland or tree plots, the Net Biodiversity Effect ΔY is small and often dominated by the Selection Effect (SE), as fast-establishing colonizer species suppress slower perennials.
• Years 3–20+ (Maturation and Soil Feedback Phase): Over subsequent years, the Selection Effect declines toward zero or becomes negative, whereas the Complementarity Effect (CE) grows steadily larger, eventually accounting for >80–95 percent of ΔY and driving strong transgressive overyielding. At Cedar Creek and Jena, long-term soil nitrogen accumulation under legume–C₄ grass mixtures, vertical root segregation, and host-specific soil pathogen accumulation in monocultures caused the richness–productivity power-law exponent b (Y ∝ S^b) to more than double between Year 2 (b ≈ 0.11) and Year 13 (b ≈ 0.28; Reich et al., 2012).

Why can't the Selection Effect alone produce transgressive overyielding?

In the absence of niche complementarity or facilitation, every species competes for the same limiting resource pool in the same manner (so relative yield total RYT = ∑ RY_O,i ≤ 1). Even if the single most productive species (M_max) competitively excludes all other species in the mixture (RY_O,max = 1, RY_O,j = 0), the total mixture biomass can at most equal M_max. Exceeding M_max (transgressive overyielding) requires RYT > 1, which corresponds to ΔRY_mean > 0—a positive Complementarity Effect.

How did Jeremy Fox (2005) refine the Loreau and Hector (2001) partitioning equation?

Jeremy W. Fox (2005) noted that in the two-term Loreau–Hector equation, a high-yielding species (M_i) might achieve a high relative yield in mixture (RY_O,i > RY_E,i) without reducing its neighbors' relative yields if it occupies an unexploited niche. Fox split the covariance term into two sub-components: the Dominance Effect (where one species gains relative yield at the expense of other species, ∑ ΔRY = 0) and Trait-Dependent Complementarity (where high-M_i species expand their yield without suppressing neighbors).

How does the Niche Complementarity Hypothesis relate to the More-Individuals Hypothesis?

Whereas the Niche Complementarity Hypothesis treats species richness as the independent driver (S → resource partitioning → higher ecosystem productivity Y), the more-individuals hypothesis in macroecology examines the reverse causal arrow across broad geographic energy gradients (higher primary productivity/energy availability Y → larger total community abundance N_total → lower local extinction rates → higher equilibrium species richness S). In natural ecosystems, both causal pathways operate interactively across spatial and temporal scales.

1. Loreau, M., & Hector, A. (2001). "Partitioning selection and complementarity in biodiversity experiments." Nature, 412(6842), 72–76. https://doi.org/10.1038/35083573
2. Tilman, D., Lehman, C. L., & Thomson, K. T. (1997). "Plant diversity and ecosystem productivity: theoretical considerations." Proceedings of the National Academy of Sciences, 94(5), 1857–1861. https://doi.org/10.1073/pnas.94.5.1857
3. Cardinale, B. J., Wright, J. P., Cadotte, M. W., Carroll, I. T., Hector, A., Srivastava, D. S., Loreau, M., & Weis, J. J. (2007). "Impacts of plant diversity on biomass production increase through time because of species complementarity." Proceedings of the National Academy of Sciences, 104(46), 18123–18128. https://doi.org/10.1073/pnas.0709069104
4. Reich, P. B., Tilman, D., Isbell, F., Mueller, K., Hobbie, S. E., Flynn, D. F. B., & Eisenhauer, N. (2012). "Impacts of biodiversity loss escalate through time as redundancy fades." Science, 336(6081), 589–592. https://doi.org/10.1126/science.1217909
5. Barry, K. E., Mommer, L., van Ruijven, J., Wirth, C., Wright, A. J., Bai, Y., Connolly, J., De Deyn, G. B., de Kroon, H., Isbell, F., Milcu, A., Roscher, C., Scherer-Lorenzen, M., Schmid, B., & Weigelt, A. (2019). "The future of complementarity: disentangling causes from consequences." Trends in Ecology & Evolution, 34(2), 167–180. https://doi.org/10.1016/j.tree.2018.10.013