Skip to main content
biology business computer-science philosophy sociology

​
Succession

​
Description

Succession is the development of a system through a sequence of structurally distinct stages, where each stage transforms the substrate in ways that enable (and typically also require) the next stage. The structural shape is initial state → pioneer stage → intermediate stages → climax (or oscillation) → disturbance reset. The defining property is stage-by-stage facilitation: pioneers’ waste, decay, modification, or accumulated capital is what makes the next stage possible. The path is not a smooth gradient from initial to mature; it is a series of qualitatively distinct regimes. The classical case is plant succession after disturbance. On bare rock after glacial retreat, lichens colonize (pioneer stage), slowly weathering the rock and accumulating organic matter; mosses follow, requiring the soil lichens produced; grasses follow the mosses; shrubs follow the grasses; eventually a climax forest community establishes itself. Each stage requires the previous stage’s substrate-modification and transforms conditions in ways that exclude itself from the next stage (the grass that thrives in open sunlight is shaded out by the shrubs it permitted to establish). The diagnostic question — “is this development moving through structurally distinct regimes where each enables the next, or smoothly along a gradient?” — separates succession from continuous growth. Succession’s stages are qualitatively distinct: the climax forest is not a “bigger lichen patch”; the mature platform is not a “larger group of early adopters.” Each stage has its own characteristic dynamics, occupants, and resource-use profiles. Mistaking succession for continuous growth produces predictable failures: planting climax-species on bare rock fails because they require the soil that intermediate stages produce; targeting mainstream customers from day one fails because the product doesn’t yet have the trust-substrate that early-adopter validation builds. The disturbance regime is constitutive. Without disturbance, the system reaches a climax that resists further change; with frequent disturbance, the system never gets past pioneer or early-intermediate stages. The intermediate-disturbance hypothesis (Connell 1978) gives the empirical claim that biodiversity peaks at moderate disturbance frequency — too little disturbance and climax-species dominate, too much disturbance and only pioneer species survive. The same shape recurs in markets, where regulatory or technological disturbance resets parts of the succession trajectory.

​
Triggers

User-initiated: User describes a system developing through staged regimes, asks about lifecycle stages, or evaluates strategy for crossing between adoption stages. Vocabulary cues: “succession,” “stages,” “lifecycle,” “maturation,” “early adopters then mainstream,” “primary/secondary succession,” “pioneer stage,” “climax community.” Agent-initiated: Agent observes a system whose development is not a smooth gradient but a sequence of qualitatively distinct regimes, where each stage’s outputs are inputs to the next stage. Candidate inference: “what stage is this system in; what does it require from the previous stage; what does the next stage require it to do?” Situation-shape signals: Strategy discussions that involve “what comes next” planning. Product-roadmap discussions that recognize different user-cohorts. Ecosystem-management or restoration planning. Organizational-stage diagnostic conversations. Career-development planning. Any discussion where “we’re not ready for X yet” or “we’ve outgrown Y” is the structure.

​
Exclusions

  • Smooth-gradient development — when a system grows continuously without qualitatively distinct regimes (a savings account compounding, a smoothly-improving skill, a steadily-growing fan base without distinct fan-type-transitions), the succession framing imposes stages that aren’t there. The diagnostic test: are there qualitatively distinct regimes with characteristic occupants and dynamics, or just a quantitative scale?
  • Disturbance-dominated regimes — when disturbance is too frequent or severe for stages to develop, the system stays in pioneer or early-intermediate state perpetually (constantly-disturbed parking lots that only support weeds; markets in constant regulatory upheaval that never mature). The succession framing predicts climax-state arrival that the disturbance regime prevents.
  • One-shot or non-recoverable transitions — when a system changes through a single threshold-crossing event without recoverable substrate-modification (a stock split, a corporate acquisition, a sudden regulatory ban), the structure is phase-transition, not succession. The “sequence of stages each enabling the next” requirement fails.
  • Parallel rather than sequential development — when multiple regimes coexist simultaneously rather than transitioning through time (a market with permanently-segmented customer types served by different vendors), the succession framing misreads parallel structure as sequential development.
  • Backward-compatible mature stages — some systems’ mature stages retain the pioneer-stage’s affordances rather than excluding them (Linux supports both modern and ancient user-cohorts; English borrows continuously without succession-displacing earlier vocabulary). When the climax doesn’t exclude pioneers, the succession framing’s “each stage transforms conditions in ways that exclude itself from the next” claim fails.

​
Structure

Internal structure of succession: a table of its component slots and the concepts that fill them.

​
Relationships

Relationship neighborhood of succession: a graph of the concepts it connects to and the concepts it is a part of.
  • cadence — succession has characteristic cadence per domain; the pair captures both the qualitative-stage structure and the temporal-rhythm structure. The succession-stage question and the cadence question together are how curators evaluate “is this a succession?”
  • niche — succession is the temporal mechanism by which niche-structure accumulates; mature succession-stages support many more niches than pioneer stages.
  • seeding — pioneer-stage entrants are seeders; their influence on the substrate shapes which later stages are reachable. Together they describe how small initial inputs lock in trajectories that play out over staged time.
  • phase-transition — succession’s between-stage shifts are often phase-transition-shaped at the boundaries; the system resists transition until something tips it. Reading them together: succession describes the staged sequence; phase-transition describes the shape of the individual stage-shifts.
  • graduation-promotion — graduating from one stage to the next is a succession-event; the pattern of scaffolding-then-adult-form recurs in product-stages, organizational-stages, and developmental-stages.
  • exaptation — pioneer-stage features sometimes get exapted as the system matures; the early-adopter community’s habits become institutional patterns; the founding team’s processes calcify into late-stage bureaucracy.
  • hysteresis — succession trajectories often exhibit hysteresis on disturbance; returning to a pioneer state after climax doesn’t restore the original pre-succession conditions, and the recovery trajectory differs from the original development.

​
Examples

Primary ecological succession after glacial retreat · biology

Glacier Bay, Alaska is the canonical study site (Chapin et al. 1994): bare-rock-to-spruce-forest succession over centuries, with lichens, mosses, alders, and eventually spruce occupying distinct stages.

Neighborhood gentrification · sociology

vacant lots → artist colonies + cheap rent → cafes and galleries → professional class → wealthy displaces founders. Urban-sociology literature names the stages; the displacement of pioneer-stage occupants by climax-stage occupants is structurally homologous to plant succession’s lichen-to-forest transition.
Adizes’ Corporate Lifecycles (1988, later expanded as Managing Corporate Lifecycles) treats organizations as living systems progressing through a sequence of named developmental stages — Courtship, Infancy, Go-Go, Adolescence, Prime, Stable, Aristocracy, Recrimination, Bureaucracy, Death. Each stage has a characteristic structural composition and a characteristic transition-failure mode if mismanaged: Affairs terminates a weak Courtship, Infant Mortality (cash-flow collapse) terminates Infancy, the Founder’s Trap (overextension dependent on the founder) terminates Go-Go, Divorce (interpersonal conflict between operators and entrepreneurs) terminates Adolescence. The diagnostic apex is Prime, the stage at which flexibility and controllability are jointly maximized; declines from Prime are not random degeneration but reliably-staged. Cross-cutting the lifecycle is the PAEI framework — Producing (short-term effectiveness), Administering (short-term efficiency), Entrepreneuring (long-term effectiveness), Integrating (long-term efficiency) — whose relative dominance shifts predictably across stages and whose imbalance signals impending stage-transition. Loss of the E role, for example, is the structural marker of decline from Prime into Aristocracy.Inference: Adizes’ contribution to the catalog’s succession primitive is the organizational instance of the staged-trajectory-with-substrate-modification shape. Each stage’s substrate (people, processes, capital, customer expectations, PAEI composition) is what the next stage requires and what the next stage’s transition-failure mode exploits when the substrate is left unmodified. The model differs from Greiner’s 1972 evolution-and-revolution framework not in disagreeing with its phase-and-crisis shape but in supplying a richer role-dynamic mechanism — PAEI lets the diagnostician name which capability is missing rather than only naming which crisis is hitting. The structural parallel to plant succession is exact: lichens prepare soil for grass; grass prepares soil for shrubs; the Go-Go stage’s improvised processes prepare substrate for Adolescent reorganization. The Adizes Institute’s adoption as a practitioner-training arm institutionalizes the framework as a diagnostic doctrine for organizational-stage recognition.
Apprentice → Journeyman → Master → Senior → Late-Career; each stage has distinct learning-rate and contribution profiles. Mistaking the master-stage’s strengths for apprentice-stage development planning produces predictable career stalls.
Glacier Bay is the textbook primary-succession sequence — bare glacial till colonized over roughly two centuries by pioneer mosses and Dryas, then nitrogen-fixing alder, then Sitka spruce forest — but Chapin et al.’s 1994 study is notable for dissecting the mechanism of each transition rather than just naming the stages. Their central finding was that no single mechanism drives the whole sequence. Life-history traits (seed size, dispersal range, time to reproduction) determine the order in which species arrive: the early pioneers have small wind-dispersed seeds and reach maturity fast, while spruce, with larger seeds and limited dispersal, simply cannot establish in quantity until later. Facilitation — the soil-nitrogen enrichment classically assumed to drive succession — turned out to govern the rate and final productivity, not the sequence. And crucially, the alder stage is double-edged: it facilitates the growth of established spruce (via added soil nitrogen) while inhibiting spruce establishment (shading and root competition suppress seedlings). The classical Clementsian picture, in which each stage uniformly prepares the way for the next, was shown to be one mechanism among several operating simultaneously.This sharpens the concept. Succession’s defining shape — qualitatively distinct stages, each transforming the substrate in ways that enable and require the next — is confirmed, but the mechanism of “enabling” is revealed to be heterogeneous: colonization-ability sets the sequence, substrate-modification sets the rate, and competition/inhibition can simultaneously help one life-stage of the successor while blocking another. The substrate transformation that early occupants impose is not a single forward push but a mix of facilitation and inhibition.Inference: when a system shows staged development, do not assume each stage uniformly prepares its successor. Disaggregate the transition mechanism — what determines the order of stages may differ from what sets the rate, and the same predecessor can facilitate one aspect of its successor while inhibiting another. A stalled transition may reflect inhibition by the prior stage (the alder shading out seedlings) rather than a missing facilitation, and the corrective differs accordingly.
Frederic Clements’s 1916 monograph is the founding treatment of ecological succession as a structured, predictable phenomenon. Clements proposed that a plant community develops through a determinate sequence of stages — pioneer species colonizing bare substrate, intermediate communities each modifying the soil and microclimate in ways that enabled their successors, culminating in a stable climax community whose composition was determined by the regional climate. The whole sequence, in Clements’s framing, behaved as a kind of “superorganism” with its own developmental trajectory.The contribution to the concept is the naming of the sequence as a unit. Before Clements, ecological change was treated as miscellaneous flux; after Clements, “primary succession,” “secondary succession,” “climax community,” and the developmental view of vegetation became the standard vocabulary. Even the parts of his framework that later ecologists rejected (notably the superorganism analogy, challenged by Gleason 1926) were rejected in terms of the staged-development scaffold he established.Inference: A concept’s first articulation often overreaches — Clements’s deterministic climax has been substantially softened by later work — but the productive contribution is the structural identification of the phenomenon as stage-by-stage substrate-modification. Subsequent refinement narrows the scope without dissolving the core shape; the concept survives the deterministic framing’s defeat because the underlying observation (each stage enables the next) was real.
Joseph Connell’s 1978 Science paper presented the intermediate-disturbance hypothesis: biodiversity in successional communities peaks at intermediate frequencies and intensities of disturbance, not at either extreme. With no disturbance, the system reaches a climax dominated by competitively-superior late-stage species that exclude others; with frequent severe disturbance, only fast-colonizing pioneer species survive. Diversity is maximized in the intermediate regime where pioneer, mid-successional, and late-stage species can all coexist because none has time to dominate before the next disturbance reshuffles the substrate.The result formalizes disturbance as a constitutive component of the succession structure rather than an exogenous interruption. The disturbance regime — its frequency, severity, and spatial pattern — is part of what determines which successional stages the system can occupy and for how long. Connell’s empirical examples (tropical rain forests with treefall gaps, coral reefs with storm damage) showed the inverted-U curve of diversity against disturbance frequency that has become a standard reference.Inference: When evaluating a system that develops through stages, the disturbance regime is not a perturbation to model around — it is part of the system’s structure. Asking “what disturbance frequency would this system experience in steady state?” is often the load-bearing question for predicting which stages will be observable. A succession concept used without a disturbance-regime parameter is incomplete.
Connell and Slatyer’s 1977 paper unbundled “succession” into three distinct mechanisms governing how one stage gives way to the next: facilitation (early-stage occupants modify the substrate in ways that make conditions suitable for later-stage species, as Clements had assumed); tolerance (later-stage species can establish regardless of early-stage occupants, simply by being more competitive in the long run); and inhibition (early-stage occupants actively suppress later-stage establishment, and replacement requires the early occupants to die or be removed).The taxonomy mattered because it falsified the assumption that succession is always facilitation-driven. Different communities follow different mechanisms; some show pure facilitation, others tolerance, others inhibition, and many show different mechanisms at different points in the same sequence. The “each stage enables the next” framing turned out to be one possibility among three, and the question “which mechanism is operating here?” became a diagnostic that produced different predictions about disturbance response, restoration strategy, and stability.Inference: When a concept has a generic forward-shape (“A leads to B leads to C”), look for whether the mechanism of the transition is uniform or varies. In ecology, what looked like one phenomenon was three; once disaggregated, the three behaved differently under disturbance and intervention. The same diagnostic applies to other staged-development concepts: ask whether each transition’s mechanism is the same or whether the sequence is mechanistically heterogeneous.
Henry Gleason’s 1926 paper challenged Frederic Clements’s superorganism framing of plant succession. Where Clements treated the climax community as a discrete, internally-coordinated developmental unit with a determinate endpoint, Gleason argued that plant associations are the coincidental products of each species’ independent responses to environmental gradients. Two locations with similar communities had not co-developed; they had each been independently filtered by similar conditions from the available species pool. There is no community-level developmental trajectory, on this view — only individual species’ dispersal, establishment, and competitive outcomes that happen to co-occur.The dispute reshaped what “succession” could mean. Gleason’s individualistic concept dissolved the strict stage-by-stage determinism while preserving the observation that communities change predictably over time. The mechanism was relocated from a community-level developmental program to species-level autecology and stochastic dispersal — a more modest and more empirically defensible position.Inference: When a concept’s first articulation includes a strong holistic claim (“the community develops as a unit”), look for whether the predictive content of the concept actually requires the holism, or whether weaker individualistic mechanisms would produce the same observable trajectory. Gleason’s critique didn’t dissolve succession; it dissolved the superorganism reading of succession, while preserving the staged-change observation. Concepts often survive the defeat of their initial metaphysical framing because the structural observation underneath is more robust than the framing.
Larry Greiner’s 1972 Harvard Business Review article proposed that growing organizations pass through a recurring sequence of stages, each ending in a crisis that forces the transition to the next stage. The phases — creativity, direction, delegation, coordination, collaboration — each have characteristic management structures that enable growth within the phase but produce predictable failures at scale (the creativity phase ends in a leadership crisis; the direction phase ends in an autonomy crisis; and so on). Crossing each crisis requires a qualitative reorganization, not incremental adjustment of the prior phase’s structures.The model is structurally a succession applied to organizations: each stage transforms the substrate (people, processes, communication patterns) in ways that enable the next stage and require the next stage’s structural shift. The “evolution” phases are the relatively-stable periods of incremental growth; the “revolution” crises are the phase-transition events between stages. Organizations that try to extend a stage past its natural ceiling stall; those that attempt the next stage’s structures before reaching the prior stage’s ceiling implement structures the organization does not yet require.Inference: When evaluating an organization’s growth challenges, the question “which phase is this organization currently in, and what crisis ends that phase?” is often more diagnostic than the question “what management practice should we adopt?” The best practices for one phase are the wrong practices for the next, and management literature read without phase-awareness produces advice that fits no phase. Greiner’s contribution is the phase-recognition discipline as a precondition for situated advice.
Thomas Kuhn’s 1962 book introduced a staged model of scientific change: long periods of normal science operating within a shared paradigm; accumulation of anomalies that the paradigm cannot explain; crisis when the anomalies become structurally embarrassing; revolution in which a competing paradigm displaces the previous one; and a new period of normal science under the successor paradigm. The transitions between stages are not smooth; the late-normal-science stage looks productive right up until the crisis fires, and the revolution itself often appears sudden in retrospect even when the underlying drift was gradual.Kuhn’s structural contribution is the recognition that scientific progress is not a continuous accumulation of refinements toward a stable truth, but a sequence of qualitatively distinct regimes each with its own characteristic vocabulary, instruments, exemplar problems, and standards of acceptable explanation. Each paradigm transforms the substrate (the trained scientists, the standard instruments, the catalogued exemplars) in ways that enable normal science within the paradigm and constrain what the next paradigm must address. Late-paradigm anomalies are the substrate-residue that the successor paradigm must inherit.Inference: When evaluating whether a field is in a stable productive phase or approaching a regime shift, the diagnostic is the anomaly inventory: which observations are increasingly difficult to fit into the current paradigm’s vocabulary, and how is the field treating them — as marginal puzzles, as the responsibility of someone else’s specialty, or as load-bearing problems? Late-normal-science discomfort with anomalies is often the leading indicator of an imminent revolution.
sourdough cultures, kombucha, cheese rind communities, gut microbiome after antibiotics; each transitions through stages with characteristic dominant species before stabilizing.
Geoffrey Moore’s 1991 book applied a successional model to technology product adoption. The customer base for a new technology product, in Moore’s framing, develops through structurally distinct stages — innovators, early adopters, early majority, late majority, laggards — each with different motivations, evaluation criteria, and risk tolerance. Crucially, Moore argued that the gap between early adopters (who tolerate incomplete products in exchange for novelty and influence) and the early majority (who require references, completeness, and proven reliability) is a chasm, not a smooth transition. Products that succeed with early adopters routinely fail to cross it because the strategies that won the early-adopter market actively repel the early majority.The structural shape is succession with an explicit between-stage failure mode: each customer cohort’s substrate-modification (the reference cases, the integrations, the supporting ecosystem) is what enables the next cohort, but the early-adopter ecosystem is not the same kind of substrate the early majority requires. Companies must deliberately retool — building a “whole product” with documented references, channel partnerships, and conservative messaging — to cross. Many do not survive the chasm.Inference: When transitioning between stages of any staged-adoption system, ask whether the substrate the prior stage produced is the substrate the next stage requires. Late-adopter discomfort often signals that the prior cohort’s ecosystem (artifacts, conventions, support patterns) does not meet the new cohort’s interface requirements, even when the technology itself is more mature. The cross-stage failure mode is the structural one Moore named; recognizing it as succession-with-substrate-mismatch identifies the corrective.
Everett Rogers’s Diffusion of Innovations synthesized thousands of adoption studies into a successional model of how a new idea spreads through a social system. Adopters fall into five qualitatively distinct categories defined by their position in time and their motivation: innovators (≈2.5%, venturesome, risk-tolerant, networked outside the local system), early adopters (≈13.5%, the respected opinion leaders others watch), the early majority (≈34%, deliberate, adopting just ahead of average), the late majority (≈34%, skeptical, adopting under economic or peer pressure once uncertainty is gone), and laggards (≈16%, traditional, oriented to the past). Plotted cumulatively over time, adoption traces an S-shaped curve — slow start, a take-off near a critical mass of roughly 10-25%, rapid mainstream uptake, then a leveling tail — while the per-period frequency of new adopters is the familiar bell.The mapping to succession is close. Each adopter category is a stage with characteristic occupants and dynamics, not a point on a continuous gradient: the early adopters are a different kind of population than the early majority, with different decision criteria, and the system passes through them in sequence. The mechanism that carries one stage into the next is itself successional — early adopters, as opinion leaders, modify the social substrate (legitimacy, visible reference cases, word-of-mouth) in ways that lower the uncertainty the more conservative later stages require before they will move. The pioneer population literally makes the environment habitable for the mainstream that follows, exactly as a pioneer plant community conditions the soil for its successors.Inference: when adoption of anything new stalls, identify which successional stage the population is in and what substrate-modification the next stage requires. Strategies tuned to one stage (novelty and influence, which win innovators and early adopters) can actively repel the next (the early and late majority need proven references and reduced uncertainty); the transition is not smooth, and treating the audience as a single homogeneous market misreads a staged sequence as a uniform one.
In The Structure of Scientific Revolutions, Kuhn described science as moving through a characteristic sequence: a pre-paradigmatic stage with competing schools, the consolidation of a paradigm and the period of “normal science” working within its puzzle-set, the accumulation of anomalies the paradigm can’t accommodate, a crisis as the anomalies become structural, a revolutionary shift to a new paradigm, and a return to normal science under the new framework.The example instantiates succession in a non-ecological domain: each stage transforms the substrate (theoretical commitments, training, instrumentation, problem-vocabulary) in a way that enables the next, and the regime-change between paradigms has the discontinuous quality Connell and Slatyer described for ecological succession after large disturbances. The Kuhnian sequence has its own characteristic disturbance regime — accumulated anomaly rather than fire or glacial retreat — but the structural shape transfers.
fire-adapted ecosystems (chaparral, longleaf pine, eucalyptus forests) cycle through post-fire stages with characteristic species composition; the burn resets the succession partway, not to bare rock.
Greenfield → Mid-development → Mature → Legacy. Each stage has characteristic architectural pressures: greenfield is constrained by what hasn’t been built; mature is constrained by what has; legacy is constrained by what can’t be removed.