Self-Thinning Rule
The self-thinning rule is an empirical and theoretical scaling law in plant population and marine benthic ecology describing how density-dependent mortality in crowded, even-aged stands causes mean individual biomass to scale as an inverse power function of survivor density along a upper boundary line.
The self-thinning rule (also known as Yoda's โ3/2 power law or the self-thinning boundary line) is a quantitative scaling relationship in population and forest ecology describing the trajectory of density-dependent mortality and individual growth in crowded, even-aged populations of sessile organisms. First formalized mathematically by Japanese plant ecologists Kyoji Yoda, Tatuo Kira, Hideo Ogawa, and Kazuo Hozumi in 1963, the rule states that as an initially dense cohort of seedlings or benthic recruits grows in body size and competes for space, light, and soil nutrients within an ecosystem, smaller and suppressed individuals die while survivors continue to accumulate biomass. On a double-logarithmic plot of mean individual biomass (B, or mean plant weight w) against surviving population density (N), crowded cohorts approach and track an upper limiting boundary line described by the power law B = C ยท Nโฮฑ, where the scaling exponent ฮฑ was classically proposed to equal 3/2 (โ1.5) from Euclidean geometric packing and later re-derived as 4/3 (โ1.33) under fractal allometric metabolic scaling theory.
Historical Formulation and Trajectory of a Crowded Cohort
From Density-Independent Growth to the Self-Thinning Boundary
During the 1950s and early 1960s, researchers at Osaka City University conducted systematic glasshouse and field experiments on monocultures of herbaceous crop species, weeds, and forest trees sown across a wide gradient of initial seed densities (Yoda et al., 1963). Earlier yieldโdensity studies by Kira et al. had established the law of constant final yield, which demonstrates that across a broad range of sowing densities, total standing crop biomass per unit area (Y = B ยท N) converges toward an asymptotic ceiling determined by local resource supply rather than initial seed input. However, when initial density is high enough that growing canopies interlock before reproductive maturity, the stand cannot reach final yield without shedding individuals through intraspecific competition.
Tracing a single even-aged cohort through time on a log10(B) versus log10(N) phase plane reveals three distinct demographic phases:
- Pre-Competition Establishment Phase (Nearly Vertical Trajectory): Immediately after germination or larval settlement, individual crowns and root systems do not overlap. Each juvenile grows at an intrinsic rate governed by temperature, photosynthetically active radiation, and soil moisture while experiencing minimal density-dependent mortality. On the logโlog plot, log10(B) increases rapidly while log10(N) remains nearly constant, producing a vertical upward trajectory.
- Approach and Canopy Closure Phase (Curving Trajectory): As leaf area index (LAI) approaches full canopy closure and root zones interpenetrate, asymmetric competition for light and soil resources intensifies. Growth of overtopped individuals slows, size hierarchies become positively skewed, and suppressed individuals begin to die, curving the trajectory leftward toward the upper boundary.
- Self-Thinning Phase (Linear Boundary Trajectory): Once the stand reaches maximum space-filling capacity, any further increase in the mean biomass (B) of surviving individuals requires a proportional reduction in survivor density (N). The cohort slides upward and to the left along the self-thinning boundary line:
log10(B) = log10(C) โ ฮฑ ยท log10(N)
Equivalently, in power-law form:
B = C ยท Nโฮฑ or Y = B ยท N = C ยท N1โฮฑ
where B is mean dry mass per surviving individual (g or kg), N is survivor density per unit area (individuals mโ2 or stems haโ1), C is an intercept constant reflecting species-specific wood density, leaf morphology, and site productivity (typically log10 C โ 3.5 to 4.4 when B is measured in grams and N in mโ2), and โฮฑ is the self-thinning slope (White & Harper, 1970).
Mechanistic Derivations: Geometric Packing (โ3/2) vs. Metabolic Allometry (โ4/3)
Euclidean Isometric Scaling vs. Fractal Vascular Networks
For more than three decades, ecologists debated whether the canonical self-thinning slope is governed by simple geometric dimensionality (โ3/2) or by vascular hydraulic and metabolic allometry (โ4/3):
- Yoda's Euclidean Geometric Model (ฮฑ = 3/2 = 1.50): Yoda et al. (1963) derived the โ3/2 exponent from two isometric assumptions: (1) plants maintain geometric similarity (isometry) as they grow, so crown projected area A scales with linear plant dimension L as A โ L2 while individual volume and biomass B scale as B โ L3, yielding A โ B2/3; and (2) during self-thinning, the surviving cohort maintains 100% canopy cover of a fixed ground area, so total occupied area N ยท A = constant. Substituting A โ Nโ1 into A โ B2/3 gives B2/3 โ Nโ1, or B โ Nโ3/2 (and total stand biomass Y = B ยท N โ Nโ1/2).
- Biomechanical and Allometric Critique: Because woody stems must resist Euler buckling under gravity and wind drag, taller trees do not remain geometrically isometric; instead, stem diameter D scales allometrically with height H (often H โ D2/3 under elastic similarity), and crown radius scales non-isometrically with stem mass. Using reduced major axis (RMA) and principal component regressions across hundreds of plant datasets, Weller (1987) and Lonsdale demonstrated that empirical thinning slopes vary around โ1.3 to โ1.5 depending on functional group and shade tolerance.
- EnquistโBrownโWest Metabolic Scaling Model (ฮฑ = 4/3 โ 1.33): In 1998, Brian J. Enquist, James H. Brown, and Geoffrey B. West derived an alternative self-thinning exponent from the WestโBrownโEnquist (WBE) fractal vascular branching theory. In the WBE framework, an individual plant's whole-organism metabolic rate and resource uptake rate (Q, including water transpiration and nutrient absorption) scale as the 3/4 power of individual body mass (Q โ B3/4). In a resource-saturated stand where total rate of resource supply per unit area (Rsupply) is fixed, total stand resource use equals the sum of individual metabolic rates: Rsupply = N ยท Q โ N ยท B3/4. At resource equilibrium, N ยท B3/4 = constant, which rearranges directly to:
N โ Bโ3/4 โบ B โ Nโ4/3 (and total stand biomass Y โ Nโ1/3)
Comparative forest inventory analyses across angiosperm and gymnosperm stands indicate that while crown geometry and shade tolerance can steepen local thinning trajectories toward โ3/2 during early pole stages, broad macroecological sizeโdensity relationships across mature forests converge closely toward the metabolic โ4/3 (N โ Bโ0.75) prediction (Enquist et al., 1998).
| Scaling Parameter / Prediction | Yoda's Geometric Model (Yoda et al., 1963) | Metabolic Scaling Theory (Enquist et al., 1998) | Silvicultural Stand Density Index (Reineke, 1933) |
|---|---|---|---|
| Governing Biological Constraint | Isometric Euclidean geometry (A โ L2, B โ L3) and 100% canopy space filling | Fractal vascular hydraulic transport and 3/4-power metabolic rate (Q โ B3/4) | Empirical stem packing of quadratic mean diameter (Dq) vs. stems per hectare (N) |
| Individual Mass vs. Density (B vs. N) | B โ Nโ3/2 (slope = โ1.50) | B โ Nโ4/3 (slope = โ1.33) | Derived via B โ Dq8/3 โ B โ Nโ4/3 to Nโ3/2 |
| Density vs. Individual Mass (N vs. B) | N โ Bโ2/3 (slope = โ0.67) | N โ Bโ3/4 (slope = โ0.75) | N โ Dqโ1.605 (Reineke exponent) |
| Total Stand Biomass Yield (Y = B ยท N) | Y โ Nโ1/2 โ B1/3 | Y โ Nโ1/3 โ B1/4 | Basal area plateaus as Dq increases along maximum SDI line |
Asymmetric Competition for Light vs. Symmetric Root Competition
Size Hierarchies and the Mechanism of Mortality
Self-thinning is driven primarily by size-asymmetric competition, a process analyzed extensively by Jacob Weiner (1990). When plants compete for soil water or dissolved mineral ions (such as nitrate NO3โ or phosphate PO43โ), competition is largely size-symmetric: individuals acquire soil resources roughly in proportion to the surface area or biomass of their root systems. In contrast, competition for unidirectional solar radiation is strongly size-asymmetric: taller individuals with foliage positioned in the upper canopy intercept incoming photons and cast shade on smaller neighbors, which cannot shade taller plants in return.
Because taller plants capture a disproportionate share of photosynthetically active radiation relative to their mass, small initial differences in germination timing, microsite fertility, or seed mass amplify over time into pronounced size hierarchies characterized by high Gini coefficients of individual plant weight. Suppressed individuals drop below their whole-plant light compensation point, where gross photosynthetic carbon fixation no longer offsets autotrophic maintenance respiration (Ra) and tissue turnover. Exhausted of non-structural carbohydrates, these overtopped stems succumb to carbon starvation, drought cavitation, or opportunistic root pathogens, driving the cohort along the self-thinning line.
Applications Beyond Terrestrial Plants: Forestry and Marine Benthos
Reineke's Stand Density Index and Sessile Marine Invertebrates
Thirty years before Yoda et al. (1963), American forester L. H. Reineke (1933) discovered an analogous empirical rule for even-aged forest stands in the United States. Plotting the logarithm of trees per unit area (N) against the logarithm of quadratic mean diameter at breast height (Dq) for fully stocked, undisturbed stands across 14 coniferous and hardwood species, Reineke found a universal upper boundary slope of โ1.605:
SDI = N ยท (Dq / 25.4)1.605
where SDI is the Stand Density Index (standardized to a reference quadratic mean diameter of 25.4 cm, or 10 inches). Because tree biomass B scales with stem diameter roughly as B โ Dq2.4 to Dq2.67, Reineke's relationship (N โ Dqโ1.605) is algebraically equivalent to B โ Nโ1.50 to Nโ1.66. Forest managers use DrewโFlewelling density management diagrams based on Reineke's SDI and Yoda's โ3/2 line to schedule commercial thinnings between the crown closure line (relative density ~0.15) and the zone of imminent competition-induced mortality (relative density ~0.55), maximizing merchantable timber volume without losing trees to natural self-thinning (Drew & Flewelling, 1979).
The self-thinning rule also extends across the biosphere to non-plant taxa. Sessile marine suspension feedersโincluding intertidal barnacles (Semibalanus balanoides), blue mussels (Mytilus edulis), and reef-building scleractinian coralsโcompete asymmetrically for two-dimensional substrate attachment space and boundary-layer planktonic food flux, exhibiting space-regulated self-thinning along slopes near โ3/2. Even mobile stream-dwelling salmonids such as brown trout (Salmo trutta) and juvenile Atlantic salmon (Salmo salar) defending foraging territories exhibit metabolic self-thinning (N โ Bโ0.75 to Bโ0.87) when growing cohorts compete for drift-feeding stations in riffle habitats.
Frequently Asked Questions
Why does the self-thinning slope have a negative sign?
The negative exponent (โ3/2 or โ4/3) reflects the inverse trade-off between individual body size and population density once a habitat is fully occupied. For surviving individuals to grow larger (increasing log B), total population density (log N) must decrease through mortality of suppressed neighbors, producing a line that slopes upward and to the left (or downward from left to right when B is plotted against increasing N).
Does the self-thinning rule apply to uneven-aged old-growth forests?
In multi-cohort or old-growth forests where canopy gap disturbances allow continuous seedling recruitment beneath mature trees, the entire stand does not march along a single even-aged trajectory. However, the size distribution of living stems across diameter classes within an old-growth stand is governed by the same metabolic space-filling constraint, generating an inverse power-law tree size spectrum (N(B) โ Bโ4/3) that mirrors the self-thinning boundary (Enquist et al., 1998).
How do soil fertility and shade tolerance affect the self-thinning line?
In classical self-thinning theory, soil nutrient enrichment increases the rate at which a cohort moves along the self-thinning line without altering the boundary slope or intercept. Shade-tolerant tree species (such as beech Fagus or hemlock Tsuga), however, can pack deeper multi-layered canopies per unit ground area than shade-intolerant pioneers (such as pines Pinus or birches Betula), resulting in a higher boundary intercept (C) and occasionally a slightly shallower thinning slope.
References
- Reineke, L. H. (1933). "Perfecting a stand-density index for even-aged forests." Journal of Agricultural Research, 46(7), 627โ638. https://naldc.nal.usda.gov/catalog/IND43968212
- Yoda, K., Kira, T., Ogawa, H., & Hozumi, K. (1963). "Self-thinning in overcrowded pure stands under cultivated and natural conditions (Intraspecific competition among higher plants XI)." Journal of Biology, Osaka City University, 14, 107โ129. https://www.semanticscholar.org/paper/28edbe1c4d3ea5f40fe51a924bc1bfd83a4524bd
- White, J., & Harper, J. L. (1970). "Correlated changes in plant size and number in plant populations." Journal of Ecology, 58(2), 467โ485. https://doi.org/10.2307/2258284
- Drew, T. J., & Flewelling, J. W. (1979). "Stand density management: an alternative approach and its application to Douglas-fir plantations." Forest Science, 25(3), 518โ532. https://doi.org/10.1093/forestscience/25.3.518
- Weller, D. E. (1987). A Reevaluation of the โ3/2 Power Rule of Plant SelfโThinning. Ecological Monographs. https://doi.org/10.2307/1942637
- Weiner, J. (1990). "Asymmetric competition in plant populations." Trends in Ecology & Evolution, 5(11), 360โ364. https://doi.org/10.1016/0169-5347(90)90095-U
- Enquist, B. J., Brown, J. H., & West, G. B. (1998). "Allometric scaling of plant energetics and population density." Nature, 395(6698), 163โ165. https://doi.org/10.1038/25977