Storage Effect
The storage effect is a fluctuation-dependent coexistence mechanism formulated by Peter L. Chesson and Robert R. Warner in 1981 demonstrating that competing species can stably coexist on a single limiting resource when differential environmental responses, environment–competition covariance, and a buffered long-lived life-history stage prevent competitive exclusion during unfavorable years.
The storage effect is a foundational fluctuation-dependent mechanism of species coexistence in community ecology, first discovered and formalized mathematically by Peter L. Chesson and Robert R. Warner in 1981 (as the "lottery model" for coral-reef fishes) and generalized into Modern Coexistence Theory by Chesson (1994, 2000). Whereas classical equilibrium niche theory—summarized by the Competitive Exclusion Principle and David Tilman's Resource-Ratio Hypothesis—predicts that no more than R species can stably coexist at equilibrium on R limiting resources, the storage effect proves that temporal or spatial fluctuations in the environment are not merely random noise, but an active stabilizing niche mechanism. Specifically, when competing species have different environmental optima across years (or patches) and possess a long-lived, buffered life-history stage (such as a soil seed bank, diapausing resting eggs, or long-lived adult trees), population gains made during favorable recruitment years are "stored" in the persistent adult or seed stage to carry the population through unfavorable years—generating negative frequency dependence that rescues any species when it becomes rare (Chesson & Warner, 1981; Chesson, 2000).
Why Environmental Fluctuations Alone Do Not Cause Coexistence
Before Chesson and Warner (1981), ecologists such as G. Evelyn Hutchinson (1961) proposed in the "Paradox of the Plankton" that seasonal weather fluctuations prevent a dominant competitor from excluding inferior competitors simply by continually reversing the direction of selection before competitive exclusion can finish.
Chesson and Warner proved mathematically that in a simple linear, non-overlapping-generation model (without a buffered stage), temporal environmental fluctuations never stabilize coexistence: instead, whichever species has the highest long-term geometric mean growth rate (or mean log growth rate, ⟨ln λi(t)⟩) still drives all other species to extinction with probability 1. For environmental fluctuations to actively stabilize coexistence—conferring a positive invasion growth rate (r̄i > 0) on every species when it drops to low density (Ni → 0) against a resident community at equilibrium—an interaction between environmental favorability, intraspecific competition, and life-history buffering is required.
The Three Mandatory Ingredients of the Storage Effect
Consider a general population model in which the finite per-capita growth rate λi(t) = Ni(t + 1) / Ni(t) of species i in year t is a function of two variables: its density-independent environmental response Ei(t) (how well species i germinates, recruits, or reproduces under the temperature/rainfall of year t) and the intensity of competition Ci(t) (resource depletion by all competing individuals in year t). For example, in the classic Chesson–Warner lottery / annual-plant seed-bank model:
λi(t) = (1 − gi(t)) si + [ gi(t) Yi(t) ] / Ci(t)
where gi(t) is the fraction of seeds in the soil seed bank that germinate in year t, si is the annual survival probability of ungerminated dormant seeds in the soil (0 < si < 1), Yi(t) is per-germinant seed yield, and Ci(t) = ∑j αij gj(t) Nj(t) is total competition from all germinating plants.
Expanding the long-term invasion growth rate r̄i = ⟨ln λi(t)⟩ via a second-order Taylor series around the mean environmental and competitive states, Chesson (1994) proved that the storage effect ΔIi emerges if and only if three biological conditions hold simultaneously:
| Ingredient | Mathematical Expression | Ecological & Physiological Meaning |
|---|---|---|
| 1. Species-Specific Differential Responses to the Environment | Var(Ei) > 0 and Corr(Ei, Ej) < 1 | Species must not respond identically to weather or patch variation: e.g., Species 1 germinates best in cool early-winter rains, whereas Species 2 germinates best in warm late-winter rains (temporal niche partitioning). |
| 2. Covariance Between Environment and Competition (EC Covariance) | Cov(Ei, Ci) > 0 when species i is common (resident), vs. Cov(Ei, Ci) ≈ 0 when species i is rare (invader) | When a species is abundant (resident), a year that is environmentally favorable for its germination (Er high) automatically produces a large cohort of its own seedlings, driving intraspecific competition (Cr) high (self-limitation in its own best years). Conversely, when a species is rare (invader, Ni → 0), its own seedlings are too few to create high competition during its favorable year (Ei high while resident Er and Ci are low)—allowing the rare species to experience a recruitment boom unrestrained by competition. |
| 3. Buffered Population Growth (Subadditivity: γi = ∂2ln λi / ∂Ei∂Ci < 0) | si > 0 (persistent seed bank or long-lived adult survival) | During a bad environmental year (Ei low) that happens to be a good year for the abundant competitor (Ci high), the rare species does not suffer catastrophic negative growth. Because most of its population remains safely dormant in the soil seed bank (si)—or survives as long-lived canopy adults immune to seedling competition—its per-capita losses during bad/high-competition years are bounded (buffered), while its gains during good/low-competition years are unbounded. |
Mathematically, the stabilizing contribution of the storage effect to the invasion growth rate of rare species i against resident r is:
ΔIi = |γi| Cov(Er, Cr) − |γi| Cov(Ei, Ci) > 0
Because all three factors (differential response + positive EC covariance for residents + subadditive buffering γi < 0) multiply together, if any one of the three ingredients equals zero (for instance, if seed survival si = 0 so there is no seed bank and γi = 0), the storage effect vanishes (ΔIi = 0).
Temporal vs. Spatial Storage Effect
Chesson (2000) showed that the exact same mathematical structure operates across space in a heterogeneous metacommunity:
- Temporal Storage Effect: Operates across time (t) within a single locality; buffering is provided by a long-lived persistent stage (seed banks, resting eggs, or long-lived iteroparous adults) that bridges across years.
- Spatial Storage Effect: Operates across habitat patches (x) at a single point in time when propagules disperse across a heterogeneous landscape; buffering is provided by spatial averaging across patches via dispersal, where high seed/larval output from low-competition favorable patches subsidizes the regional pool without requiring multi-year dormancy (Sears & Chesson, 2007).
Landmark Empirical Verifications in Natural Ecosystems
Over the past two decades, field ecologists have directly parameterized Chesson's equations (Ei(t), Ci(t), si, and ΔIi) across multi-year time series to prove that the storage effect maintains biodiversity in the wild:
| Ecosystem & Taxonomic Guild | Buffered "Storage" Stage | Environmental Driver of Differential Ei(t) | Quantitative Empirical Finding |
|---|---|---|---|
| Sonoran Desert Winter Annual Plants (Erodium cicutarium, Pectocarya recurvata, Schismus barbatus; Angert et al., 2009) | Soil Seed Bank (dormant seeds surviving 2–10 years in desert soil, si ≈ 0.3–0.8) | Timing and temperature of autumn/winter desert rainstorms ( modulating germination fraction gi(t) and water-use efficiency vs. RGR trade-offs) | Multi-decade demographic censuses at Tumamoc Hill, Arizona, proved that EC covariance and seed-bank buffering generate a positive storage effect (ΔI > 0) sufficient to prevent competitive exclusion among functional trade-off guilds (Angert et al., 2009). |
| Freshwater Crustacean Zooplankton (Daphnia and Diaptomus copepods in temperate lakes; Cáceres, 1997) | Ephippial / Diapausing Resting-Egg Bank in anoxic lake sediments (viable for 30 to >100 years) | Interannual variation in spring phytoplankton blooms, water temperature, and planktivorous fish recruitment | Carla E. Cáceres (1997) measured emergence from sediment egg banks in Oneida Lake, New York, and used experimental egg-bank removal models to prove that Daphnia pulicaria and Daphnia galeata mendotae can coexist only because the sediment resting-egg bank provides the storage effect. |
| Coral-Reef Fishes & Temperate Forest Trees (Thalassoma wrasses; North American mixed forests; Warner & Chesson, 1985; Usinowicz et al., 2017) | Long-Lived Iteroparous Adults (surviving 10–300+ years with low adult mortality da ≪ 1) | Pelagic larval oceanographic currents (reef fish) or interannual mast-seeding / climate anomalies across latitude (forest trees) | Global ForestGEO synthesis across 10 forests showed that asynchronous interannual seed/seedling recruitment coupled with long adult tree lifespans generates a strong temporal storage effect that increases toward temperate and boreal latitudes, complementing tropical Janzen–Connell spatial stabilization (Usinowicz et al., 2017). |
| North American Sagebrush Steppe Perennial Grasses (Adler et al., 2006) | Perennial Tussock Rootstocks / Clonal Genets | Interannual precipitation and temperature anomalies across a 32-year mapped chart-quadrat record in Idaho | Parameterized integral projection models (IPMs) showed that the temporal storage effect accounts for nearly half of the total stabilizing niche differences maintaining coexistence among four dominant bunchgrasses. |
References
- Chesson, P. L., & Warner, R. R. (1981). "Environmental variability promotes coexistence in lottery competitive systems." The American Naturalist, 117(6), 923–943. https://doi.org/10.1086/283778
- Chesson, P. (1994). "Multispecies competition in variable environments." Theoretical Population Biology, 45(3), 227–276. https://doi.org/10.1006/tpbi.1994.1013
- Cáceres, C. E. (1997). "Temporal variation, dormancy, and coexistence: a field test of the storage effect." Proceedings of the National Academy of Sciences, 94(17), 9171–9175. https://doi.org/10.1073/pnas.94.17.9171
- Chesson, P. (2000). "Mechanisms of maintenance of species diversity." Annual Review of Ecology and Systematics, 31, 343–366. https://doi.org/10.1146/annurev.ecolsys.31.1.343
- Angert, A. L., Huxman, T. E., Chesson, P., & Venable, D. L. (2009). "Functional tradeoffs determine species coexistence via the storage effect." Proceedings of the National Academy of Sciences, 106(28), 11641–11645. https://doi.org/10.1073/pnas.0904512106
- Usinowicz, J., Chang-Yang, C.-H., Chen, Y.-Y., Clark, J. S., Fletcher, C., Garwood, N. C., et al. (2017). "Temporal coexistence mechanisms contribute to the latitudinal gradient in forest diversity." Nature, 550(7674), 105–108. https://doi.org/10.1038/nature24038