Concept

Ecological succession

T-089Home BU-305Threads systems · evolution

Statement

How communities assemble and change over time.

Why it matters

Where competitive-exclusion and r-k-selection describe the fate and strategy of individual species, ecological succession is the unit's account of how entire communities change composition over time as a predictable, ordered process, not simply a random accumulation of species. It is the temporal complement to island-biogeography's spatial account of community assembly, and it directly explains why keystone-species and metapopulation-dynamics both have to be understood as operating within a community whose species composition is itself a moving target, not a fixed backdrop.

Hypotheses

Species differ systematically in their colonising ability, growth rate, and tolerance of the physical conditions present at different stages of a site's development.Without such differences, there would be no reason for community composition to change in any particular order over time; succession's characteristic, repeatable sequence depends specifically on early-arriving species differing predictably from later-arriving ones in exactly these traits. Earlier-arriving species alter local conditions (soil development, shade, microclimate, resource availability) in ways that affect which species can subsequently establish.This is the mechanistic link between successive stages: without species themselves modifying the environment they occupy, there would be no reason later species should depend on, or be enabled by, the presence of earlier ones, and the sequence would instead simply reflect whichever species happened to arrive by chance. Succession is not assumed to proceed toward a single, universally fixed climax community independent of local conditions; contemporary ecology generally treats the "climax" concept as a relatively stable, but not necessarily unique or permanently fixed, later-stage community whose specific composition depends on local climate, soil, and disturbance history.

Proof

\[ \text{Primary succession begins on substrate with no pre-existing soil or biological legacy (e.g. bare rock, new volcanic deposits); secondary succession begins where soil and some biological legacy already remain after a disturbance.} \]

This distinction sets a very different starting point for the sequence that follows: primary succession must begin with organisms capable of establishing on bare substrate and initiating soil formation, while secondary succession can begin considerably further along, since soil structure, seed banks, and root systems commonly survive the disturbance itself. A

\[ \text{Pioneer species (e.g. lichens and mosses in primary succession) establish first, tolerating harsh conditions unsuitable for most other species, and begin altering the local environment (Hypotheses).} \]

Pioneer species are characteristically fast-colonising, stress-tolerant organisms whose activity (weathering rock, trapping wind-blown particles, fixing nitrogen, contributing organic matter on death) gradually builds the rudimentary soil and microclimate conditions that subsequent, less stress-tolerant species require to establish at all. A

\[ \text{Each subsequent stage's species further modify local conditions, enabling colonisation by species adapted to those newly modified, and generally less physically extreme, conditions.} \]

This is Step 2's logic reapplied repeatedly: grasses and small shrubs, once soil depth and organic content are sufficient, modify shading, moisture retention, and nutrient availability further, in turn enabling colonisation by larger shrubs and eventually trees, each successive wave of species depending on the environmental modification contributed by the wave before it. A

\[ \text{Succession slows and community composition approaches a comparatively stable, self-replacing state (a climax or near-climax community) once further species replacement no longer occurs at appreciable rate.} \]

Once the resident species are themselves capable of regenerating under the conditions they and their predecessors have collectively created, rather than being progressively displaced by newly enabled colonists, net compositional turnover slows substantially, though ongoing small-scale disturbance and gap-phase replacement (individual trees dying and being replaced) can continue within this broadly stable state (Hypotheses, t3). B

Result

\[ \text{Pioneer colonisation} \to \text{environmental modification} \to \text{replacement by better-adapted successors} \to \text{a relatively stable later-stage community} \]

Reading. Community change over time at a given site is not random accumulation but a broadly predictable, directional sequence, driven mechanistically by earlier colonists altering conditions in ways that specifically favour their own eventual replacement by different, later-arriving species.

Scope. The general sequence (pioneer to progressively more complex, longer-lived species) is widely applicable; the specific species involved, the sequence's exact length, and the nature of any climax state are all strongly dependent on local climate, substrate, and disturbance regime, not fixed or universal.

Corollaries & converses

r-k-selection's contrast between fast-reproducing, poor-competitor ("r-selected") and slow-reproducing, strong-competitor ("K-selected") life histories maps closely onto succession's own sequence, with pioneer species characteristically r-selected and later-successional species characteristically more K-selected.
Disturbance (fire, storm, land clearance) resets succession locally, either back to bare substrate (restarting primary succession) or to an intermediate stage with residual soil and biological legacy still present (initiating secondary succession, Step 1), which is why disturbance frequency is itself a major determinant of the community composition actually observed at a given site.
Converse: observing a site's current community composition, together with knowledge of local successional sequences, allows a rough inference of how long that site has been undisturbed and roughly how far along the successional sequence it has progressed.

Fails without

Drop systematic species differences in colonising ability and tolerance (Hypotheses): if species did not differ predictably in these traits, there would be no consistent reason particular species should arrive early versus late, and site composition over time would instead reflect essentially random colonisation order rather than the directional, repeatable sequence actually observed across independent sites of similar type.
Drop environmental modification by earlier species (Hypotheses): without pioneer and intermediate species altering local conditions, later-successional species requiring those modified conditions (deeper soil, more shade, higher organic content) would have no mechanism enabling their eventual establishment, and succession would stall indefinitely at the pioneer stage rather than progressing further.

Common errors

Assuming succession always converges on one single, universally fixed climax community type regardless of local conditions; contemporary ecology treats climax composition as strongly dependent on local climate and substrate, not as a single global endpoint (Hypotheses, t3).
Confusing primary succession (starting from bare substrate with no soil, Step 1) with secondary succession (starting where soil and biological legacy already remain after disturbance), which proceeds considerably faster precisely because it skips the early, slow soil-building stages primary succession requires.
Treating the climax or late-successional stage as entirely static once reached; ongoing small-scale disturbance and gap-phase replacement typically continue within an overall stable community (Step 4), rather than compositional change stopping altogether.
Assuming succession is driven purely by physical environmental change (climate, geology) rather than substantially by the biological activity of the organisms present at each stage modifying conditions for their successors (Step 2–3).

Discussion

Henry Cowles's studies of vegetation zonation on the sand dunes of Lake Michigan, published from 1899, and Frederic Clements's subsequent, more strongly deterministic theory of succession culminating in a single, climatically determined climax community, are usually credited as founding this area of ecology; Clements's strict single-climax view was later substantially revised by Henry Gleason and others, who argued succession and community composition were considerably more individualistic, contingent, and locally variable than Clements's original framework proposed, closer to the qualified, less strictly deterministic view (Hypotheses, t3) generally held today.

The intermediate disturbance hypothesis, related to but distinct from succession itself, proposes that species diversity within a community is often maximised not at either extreme of disturbance frequency (very frequent or very rare) but at some intermediate level, since intermediate disturbance can maintain a mix of early-, mid-, and late-successional species together rather than allowing any single successional stage to dominate exclusively.

Common misconception: that succession always proceeds toward greater species diversity at every stage. While diversity commonly increases through early and intermediate stages as more niches and structural complexity become available, some late-successional or climax communities can actually show reduced diversity relative to intermediate stages, as a small number of highly competitive, dominant species come to exclude many of the smaller, more opportunistic species characteristic of earlier stages (competitive-exclusion).

Worked examples

\[ \text{Bare volcanic rock} \to \text{lichens \& mosses} \to \text{grasses \& herbaceous plants} \to \text{shrubs} \to \text{fast-growing pioneer trees} \to \text{shade-tolerant climax forest} \]

This standard primary-succession sequence follows Steps 1–4 directly: lichens weather rock and begin organic matter accumulation, enabling grasses once minimal soil depth exists, which further build soil and organic content enabling shrubs, then trees, with each stage's presence a direct prerequisite (via environmental modification, Step 2–3) for the stage that follows. A

\[ \text{each stage's colonists make conditions progressively more suitable for the next stage's colonists, not for their own species' continued dominance} \]

Reading. The sequence's driving logic is not that later species outcompete earlier ones directly under unchanged conditions, but that earlier species change the conditions in ways that specifically favour their own eventual successors over themselves.

Scope. The identical structural logic (pioneer, then progressively larger and longer-lived successors) applies to primary succession sequences in very different physical settings (volcanic rock, glacial retreat, sand dunes), even though the specific species differ substantially by region.

Further example. A researcher measures the Shannon-Wiener diversity index (\(H'\)) of a plant community at two time points during secondary succession on a clear-cut site. At Year 5 (early succession), the relative abundances of the four dominant species are: Species A = 0.60, Species B = 0.25, Species C = 0.10, Species D = 0.05. At Year 20 (late succession), the relative abundances of the same four species (now including late-successional trees) are: Species A = 0.25, Species B = 0.25, Species C = 0.25, Species D = 0.25. Calculate the Shannon-Wiener index for both time points and determine the change in diversity.

\[ H' = -\sum_{i=1}^{S} p_i \ln(p_i) \]

The Shannon-Wiener index quantifies diversity based on both species richness (\(S\)) and evenness (distribution of individuals among species). The natural logarithm (\(\ln\)) is the standard convention for this index in ecology. A

\[ H'_{\text{Year 5}} = -(0.60 \ln(0.60) + 0.25 \ln(0.25) + 0.10 \ln(0.10) + 0.05 \ln(0.05)) \]

Substitute the relative abundances (\(p_i\)) for the early succession community into the formula. Note that \(\ln(p_i)\) is negative for \(p_i < 1\), so the sum inside the parenthesis is negative, and the final \(H'\) will be positive. A

\[ H'_{\text{Year 5}} = -(-0.3065 - 0.3466 - 0.2303 - 0.1498) = 1.0332 \]

Calculate the individual terms: \(0.60 \ln(0.60) \approx -0.3065\); \(0.25 \ln(0.25) \approx -0.3466\); \(0.10 \ln(0.10) \approx -0.2303\); \(0.05 \ln(0.05) \approx -0.1498\). Summing these gives \(-1.0332\). Negating the sum yields \(H' = 1.0332\). A

\[ H'_{\text{Year 20}} = -(0.25 \ln(0.25) + 0.25 \ln(0.25) + 0.25 \ln(0.25) + 0.25 \ln(0.25)) \]

Substitute the relative abundances for the late succession community. Since all four species have equal abundance (\(p_i = 0.25\)), this represents maximum evenness for \(S=4\). A

\[ H'_{\text{Year 20}} = -4(0.25 \ln(0.25)) = -\ln(0.25) = \ln(4) \approx 1.3863 \]

Calculate the term: \(0.25 \ln(0.25) \approx -0.3466\). Multiplying by 4 gives \(-1.3863\). Negating the sum yields \(H' = 1.3863\). This is the theoretical maximum \(H'\) for a community with 4 species. A

\[ \Delta H' = H'_{\text{Year 20}} - H'_{\text{Year 5}} = 1.3863 - 1.0332 = 0.3531 \]

Calculate the change in diversity. The positive value indicates an increase in Shannon diversity from early to late succession in this specific case, driven by increased evenness. A

\[ H'_{\text{Year 5}} = 1.033, \quad H'_{\text{Year 20}} = 1.386, \quad \Delta H' = +0.353 \]

Reading. In this example, diversity (as measured by Shannon index) increased during succession. This is a common pattern where early successional communities are dominated by a few highly competitive or tolerant species (low evenness), while later communities support a more even distribution of individuals among multiple species. However, diversity can also peak in mid-succession and decline in late succession if dominance by a few large trees reduces evenness significantly.

Scope. The Shannon index assumes that all individuals are counted and that relative abundance is a meaningful measure of diversity. It does not account for spatial structure or species identity, only abundance distribution. The result holds for this specific dataset; other successional trajectories may show different patterns depending on species interactions and environmental constraints.

Problems

A forest is clear-cut but the soil, root systems, and seed bank remain largely intact. Predict whether the site will undergo primary or secondary succession, and estimate qualitatively whether recovery will be faster or slower than succession on newly exposed bare rock, referencing Step 1.
Solution
Because soil and biological legacy (root systems, seed bank) remain, this is secondary, not primary, succession (Step 1). Recovery is expected to be considerably faster than on bare rock, since the slow, early soil-building stages that primary succession must complete from scratch (Step 2) are largely already accomplished; the site can begin from a considerably more advanced starting point in the sequence.
Two nearby sites of similar climate undergo succession after similar disturbances, but end up with somewhat different dominant tree species in their later stages. Explain how this observation is consistent with the Result, using the Hypotheses' t3 qualification.
Solution
Because succession's later stage (climax or near-climax community) is not assumed to converge on one single, universally fixed community type independent of local conditions (Hypotheses, t3), some variation in exact species composition between sites of broadly similar but not identical climate, substrate, and disturbance history is expected, rather than being inconsistent with the general successional model described in Steps 1–4.
A grassland is subjected to very frequent fire disturbance, preventing shrubs and trees from ever establishing, while a nearby grassland with rare fire disturbance progresses to shrubland and eventually forest. Explain both outcomes using the Result and the concept of disturbance interacting with succession.
Solution
Succession's directional sequence (Steps 1–4) requires enough time between disturbance events for each stage's species to establish and begin modifying conditions for the next stage; frequent fire repeatedly resets the frequently burned grassland back to an early stage before shrubs or trees can establish, holding it at the grassland stage indefinitely, while the rarely burned grassland has sufficient undisturbed time for the full sequence (grassland to shrubland to forest) to proceed toward a later successional stage.
On a new lava flow, soil organic carbon builds up as \(C(t) = C_{\max}\left(1 - e^{-kt}\right)\) with \(C_{\max} = 4.0\%\) and \(k = 0.020\ \text{yr}^{-1}\). Find the soil carbon after 50 years and the time needed to reach half of \(C_{\max}\). Why does accumulation slow down as succession proceeds?
Solution
\(C(50) = 4.0\%\,(1 - e^{-0.020\times50}) = 4.0\%\,(1 - e^{-1}) = 4.0\% \times 0.632 = 2.5\%\). Half of \(C_{\max}\) is reached when \(e^{-kt} = 0.5\), i.e. \(t_{1/2} = \ln 2/k = 0.693/0.020 = 35\ \text{yr}\). The rate \(dC/dt = k\,(C_{\max} - C)\) falls as the soil fills: litter inputs level off as the plant community matures, while decomposition losses grow with the amount of carbon already stored, so the pool approaches a steady state where inputs balance decomposition.
Two plots of equal area are undergoing secondary succession. Plot A has a species richness of 10 and an evenness index (Pielou's \(J'\)) of 0.80. Plot B has a species richness of 15 and an evenness index of 0.60. Calculate the Shannon-Wiener index (\(H'\)) for both plots and determine which plot has higher diversity.
Solution
Pielou's evenness index is defined as \(J' = H' / \ln(S)\). Therefore, \(H' = J' \times \ln(S)\). For Plot A: \(S = 10\) \(J' = 0.80\) \(H'_A = 0.80 \times \ln(10)\) \(\ln(10) \approx 2.3026\) \(H'_A = 0.80 \times 2.3026 = 1.8421\) For Plot B: \(S = 15\) \(J' = 0.60\) \(H'_B = 0.60 \times \ln(15)\) \(\ln(15) \approx 2.7081\) \(H'_B = 0.60 \times 2.7081 = 1.6249\) Comparison: \(H'_A = 1.842\) \(H'_B = 1.625\) Plot A has a higher Shannon-Wiener index, indicating higher diversity despite having fewer species, because its evenness is significantly higher.