Evolutionary ecology studies how interactions with other organisms and the physical environment shape inherited change, and how that change can alter ecological relationships in return. Its gold standard is a testable explanation that connects ecological conditions, variation, inheritance and reproductive consequences without mistaking every visible change for evolution.
This guide helps you distinguish adaptation from plasticity, test local advantage, follow selection through a population and recognise a genuine feedback between evolution and ecology. The worked examples are original fictional models with stated assumptions. Their numbers teach the reasoning; they are not field estimates or recommendations to manipulate wild populations.
Choose your evolutionary ecology route
Connect environments, inherited variation, ecological consequences and feedback through explicit comparisons.
Route 1 · Chapters 1–3: Separate ecological and inherited change
Route 2 · Chapters 4–6: Test spatial and frequency effects
Route 3 · Chapters 7–9: Follow trade-offs and feedback
Route 4 · Chapters 10–11: Compare interacting lineages and rescue
Route 5 · Chapters 12–14: Transfer, diagnose and continue
Chapter index
Separate ecological and inherited change · Chapters 1–3
Test spatial and frequency effects · Chapters 4–6
Follow trade-offs and feedback · Chapters 7–9
Compare interacting lineages and rescue · Chapters 10–11
Transfer, diagnose and continue · Chapters 12–14
Imagine a pond whose grazers consume algae. One question asks whether grazing favours inherited algal characteristics that reduce consumption. Another asks whether a change in algal composition alters grazer abundance or nutrient cycling. The first follows ecology into evolution. The second follows evolutionary change back into ecology. A complete feedback requires evidence connecting both directions over time.
The distinction matters because two observations can occur together without forming a loop. A warmer season might change algal composition and grazer abundance independently. Alternatively, grazer abundance could change first, producing a temporary algal response without inherited change. A diagram containing arrows is a hypothesis about those relationships, not evidence that all the arrows operate.
Begin by naming the entities and the time scale. Are we following individuals, genotypes, species or functional groups? Does one time step represent a day, a generation or a season? An increase in one species within a community is ecological sorting; an increase in an inherited variant within a population is a different kind of change. Both can matter, but they should not be merged into one unspecified process.
For the pond, a useful starting record includes total algal abundance, the frequency of the focal inherited types, grazer abundance and the resource conditions. Each variable answers a different question. A single measurement such as water clarity may be a useful outcome, but it cannot identify the full causal chain by itself.
The Gold Standard of Ecology introduces populations, communities and ecosystems. The Gold Standard of Evolution explains inherited population change. Evolutionary ecology connects those foundations through explicit comparisons: which ecological condition changes selection, and which inherited difference changes an ecological outcome?
Keep these questions separate until the evidence joins them. This makes the explanation stronger, because it becomes clear which part has been demonstrated and which part remains to be tested.
An organism can respond to its surroundings during its lifetime. Plants may develop different forms under different light conditions; animals may change behaviour when a predator appears. Such responses can affect survival and reproduction, but the response itself does not demonstrate a change in inherited population composition.
Phenotypic plasticity describes differences in phenotype expressed by a genetic background across environments. Evolutionary change concerns inherited differences across generations. Plasticity can itself have an inherited basis, so the concepts are connected rather than mutually exclusive. The task is to identify which claim the observations support.
Original common-environment comparison
Suppose fictional plants from a dry site average ten units of leaf thickness, while plants from a wet site average six. When descendants from both sites are grown under the same conditions, both average eight. That result is compatible with a substantial environmental contribution to the original difference. It does not establish that the populations are genetically identical in every respect.
Now imagine descendants retain averages of nine and seven in the common environment. An inherited contribution becomes a stronger possibility, but maternal effects, developmental carry-over and other persistent differences still need consideration. The design determines which alternatives have been reduced; the phrase “common environment” does not automatically eliminate them all.
A further comparison measures both populations in dry and wet conditions. If their responses differ, the environmental effect depends on population background. A single common environment might therefore reveal only one part of the pattern. Plotting the responses makes the limitation easier to see than listing two means without their conditions.
The specialist guide to how the same genes can produce different traits in different environments develops reaction norms and these inference limits. The evolutionary-ecology extension asks whether the measured responses affect reproductive contribution in the environments organisms actually encounter.
Do not call a response adaptive merely because it looks appropriate. A trait can change in the expected direction without improving performance, or improve one performance measure while reducing another. The relevant comparison needs consequences as well as appearance.
Natural selection involves differences in reproductive contribution associated with inherited variation. An ecological explanation should identify the resource, interaction or physical condition that changes those contributions. “The environment selected the best organisms” is too vague: best at what, compared with whom, during which life stage?
Original two-type calculation
Assume a closed, deterministic, asexual teaching population with two faithfully inherited types. Begin with fifty of type A and fifty of type B. Under the stated conditions, each A leaves 1.2 descendants in the next census and each B leaves 0.8. The next expected counts are sixty and forty. A rises from frequency 0.50 to 0.60 while the total stays at one hundred.
In general, if p is A’s initial frequency and wA and wB are the supplied descendant contributions, its next frequency is p × wA divided by p × wA + (1 − p) × wB. The denominator is the mean contribution per initial individual. Keeping it visible prevents a change in frequency from being confused with an absolute increase.
Suppose instead the contributions are 0.6 for A and 0.4 for B. The next counts are thirty and twenty. A again reaches frequency 0.60, but the population halves. Relative advantage does not guarantee population growth. Both types can decline while one becomes more common.
To explain a real pattern, we would need evidence for the inherited difference, the relevant life-cycle contributions and the ecological cause. A frequency increase alone could also involve migration, sampling or drift. The teaching model deliberately excludes those processes so that the selection arithmetic can be inspected.
Changing the environment can reverse the supplied contributions. That does not mean the original calculation was wrong; it means the advantage was conditional. A strong explanation attaches the condition to the result rather than turning a local comparison into a permanent hierarchy of organisms.
Populations from different places often differ. Geography alone does not show that each population is adapted to its home environment. Local adaptation is investigated through performance comparisons that distinguish the origin of organisms from the environment in which they are tested.
Original reciprocal comparison
Take two fictional populations, North and South, and two test environments. In the northern environment, expected recruits per starting individual are four for North and two for South. In the southern environment, they are three for North and six for South. The local population outperforms the foreign population within each environment. This crossing pattern supports a local-advantage interpretation under the model.
Notice that North also performs better at home than away, while South does too in this example. Those home-versus-away comparisons are not the same as local-versus-foreign comparisons. Environments can differ in overall quality, so an organism might perform better in a favourable foreign site while still outperforming foreign organisms at its own less favourable site.
For example, change North’s values to two at home and five away, and South’s to one in the north and eight in the south. North now performs better away than at home, yet the local population still wins within each site: two exceeds one in the north, and eight exceeds five in the south. The choice of comparison changes the verbal conclusion.
Ågren and Schemske’s reciprocal-transplant study of Arabidopsis illustrates a real research design comparing populations in contrasting native environments over multiple years. Its measured fitness components and experimental duration belong beside its conclusions; a single trait difference would not provide equivalent evidence.
The existing specialist on migration–selection balance, clines and local adaptation takes this question into spatial genetic models. Use it when the next task is explaining how local selection interacts with incoming genetic material.
Dispersal moves organisms or propagules. Gene flow involves successful transmission of inherited material between populations. A migrant that arrives but leaves no descendants can affect ecological interactions without contributing genes to the receiving population. The two effects should be measured separately when both matter.
Original arrival-and-selection account
Suppose a local asexual population contains eighty A individuals and twenty B individuals. Twenty B migrants arrive before the next reproductive interval. The new frequency of A is 80/120, or two thirds. Assume A then contributes 1.5 descendants per individual and B contributes one. Expected descendants are 120 A and forty B, so A’s frequency becomes 0.75.
Without migration, the same contributions would produce 120 A and twenty B, giving A a frequency of about 0.857. Local selection favours A in both calculations, but migration changes the composition on which selection acts. The result is not captured by saying either migration or selection is the sole process operating.
Order matters. If reproduction occurs first and migrants enter afterwards, the next census includes a different sequence of events. A model must match the biological timing of arrival, survival, reproduction and observation. Using the same parameter names in a different order can describe a different life cycle.
Dispersal can introduce useful variation, counter local divergence or affect population size. Which consequence dominates depends on the incoming types, their reproductive success and the receiving environment. A general statement that connectivity is always beneficial or always harmful discards the very mechanisms the subject studies.
In field evidence, distinguish movement records, ancestry estimates and contemporary reproductive contributions. A map showing spatial genetic differences can reflect historical separation as well as current selection. Conversely, similar frequencies do not prove the absence of selection if gene flow or other processes counteract its effects.
The practical repair is to write separate sentences for movement, inherited contribution and local performance. Then identify the observations supporting each sentence. This keeps a plausible spatial story from outrunning the available evidence.
Density describes how many organisms occupy a defined area or volume. Frequency describes the proportion belonging to a type. A population can change density while preserving type frequencies, or change frequencies while preserving total density. Fitness can depend on either quantity, and the mechanisms need not be the same.
Original rare-type model
Let p be the frequency of A in a fictional two-type population. Supply contributions wA = 1.2 − 0.4p and wB = 0.8 + 0.4p. These equations are assumptions chosen to make rarity advantageous. At p = 0.25, A contributes 1.1 and B contributes 0.9. The next A frequency is 0.275/(0.275 + 0.675), approximately 0.289.
At p = 0.75, A contributes 0.9 and B contributes 1.1. Its next frequency is 0.675/(0.675 + 0.275), approximately 0.711. In both cases the frequency moves towards one half. At one half, the supplied contributions are equal. This illustrates a possible balancing mechanism rather than proving that every rare type enjoys an advantage.
Now imagine multiplying both contributions by a common density-dependent factor of 0.5. The next frequencies remain the same because that common factor cancels in the ratio, while expected population size is reduced. Density dependence can therefore alter abundance without altering this particular frequency update. If density affects the types differently, it need not cancel.
Potential biological mechanisms include competition for different resources or consumers responding to common prey. The equations alone do not identify which mechanism operates. To test one, measure the proposed mediator and compare conditions that distinguish it from alternatives.
An equilibrium frequency is also not a promise of persistence. Small populations can lose a type through random events, environments can change, and the assumed advantage may weaken at different densities. Keep deterministic tendency, stochastic outcome and empirical mechanism as separate parts of the account.
A trait can improve performance in one ecological task while carrying a cost in another. Defence may reduce consumption while slowing growth; dispersal may improve access to new sites while increasing travel mortality. A convincing trade-off explanation specifies the alternatives, the constraint and the relevant outcomes rather than assuming that every useful trait must have a particular cost.
Original defence comparison
Imagine two asexual algal types. In the absence of grazers, an undefended type multiplies by 1.6 per interval and a defended type by 1.2. With grazers present, survival through the grazing stage is 0.40 for the undefended type and 0.80 for the defended type. Assume reproduction precedes this survival stage and all other effects are absent.
Their combined contributions are 1.6 × 0.40 = 0.64 and 1.2 × 0.80 = 0.96. Defence wins in relative contribution under the grazing conditions, despite slower multiplication beforehand. Yet both combined contributions are below one, so relative success alone does not establish persistence.
Now reduce grazing enough that survival becomes 0.90 and 0.95. Contributions become 1.44 and 1.14. The undefended type now has the larger contribution. The ranking changes because the ecological consequence of defence changed relative to its supplied growth cost.
These are complete calculations for a deliberately short life cycle. A real comparison may require several ages, dormancy, variable resource levels or more than one reproductive event. Omitting those stages can turn a short-term advantage into an inaccurate lifetime claim.
Life History Theory develops survival and reproductive schedules, while Behavioural Ecology examines behaviour under constraints and changing conditions. Evolutionary ecology connects these consequences to the ecological interactions that shape inherited variation.
Ask what observation could reveal the proposed cost. A negative correlation is not automatically causal evidence, and differing resource acquisition can conceal an allocation cost. The explanation should name the missing comparison rather than relying on the appealing language of balance.
To demonstrate an ecological consequence of evolution, compare populations that differ in inherited composition while controlling or accounting for other relevant differences. Otherwise, a difference attributed to evolved traits may actually reflect population density, body size, previous environment or another correlated feature.
Bassar and colleagues’ guppy research tested how locally adapted population types affected ecosystem structure and function. The study provides a concrete example of asking what organisms do to their environment, alongside asking how their environment shaped them. Its experimental comparisons matter more than a general claim that everything is connected.
Original resource-removal example
Suppose type A removes two resource units per individual per interval and type B removes one. At a fixed population size of one hundred, a population containing twenty A and eighty B removes 120 units. A population containing eighty A and twenty B removes 180 units. The composition difference produces a sixty-unit difference under the supplied linear rule.
Now halve the size of the second population. Forty A and ten B remove ninety units, which is less than the first population removes. Looking only at total resource removal would reverse the apparent comparison. Type composition and abundance both affect the outcome, so both need to be represented.
The per-individual rule may itself change with resource availability or crowding. If resources become depleted, multiplying a constant feeding rate by population size can exceed what is physically available. A useful model checks this boundary and explains how the process changes near it.
For evidence, distinguish the total effect of a population from effects measured at matched density or biomass. These comparisons answer different questions and can both be useful. A conservation question about whole-system impact may need the total effect; a mechanism question about inherited feeding differences may need controlled contrasts.
The strongest account reports the comparison clearly and follows the proposed mediator. If the claim concerns nutrient recycling, measure the relevant nutrient process rather than treating any change in abundance as equivalent evidence.
An evolutionary effect on an ecosystem is one direction of a possible feedback. To close the loop, the altered ecological state must in turn affect subsequent evolutionary change. A useful investigation therefore follows a sequence: inherited composition changes an ecological variable, that variable changes relative contribution, and composition changes again.
Original two-state feedback model
Imagine A is favoured when a resource is abundant: its contribution is 1.3 compared with B’s 1.0. At p = 0.50, A’s next frequency is 0.65/1.15, approximately 0.565. Assume a more A-rich population removes the resource more quickly, as in the preceding chapter.
Now stipulate that resource depletion changes contributions to 0.8 for A and 1.1 for B. Starting from p approximately 0.565, the next frequency becomes about 0.486. The ecological consequence of A’s increase has altered the direction of selection in this toy sequence. The arithmetic illustrates a feedback hypothesis; the rule switching between resource states is an explicit supplied assumption.
To turn this into a stronger model, represent resource renewal and depletion continuously or through a justified discrete rule. Track population size as well as composition. Check whether the resulting dynamics stabilise, oscillate or decline rather than assuming that every negative feedback produces a harmless equilibrium.
Yoshida and colleagues’ predator–prey experiment compared laboratory systems differing in prey genetic variability and found that rapid prey evolution altered population cycles. It supplies an empirical reason to investigate evolutionary dynamics on ecological time scales. It does not imply that all predator–prey cycles require the same evolutionary mechanism.
A decisive research question asks what changes when the proposed evolutionary pathway is constrained, while other aspects of the comparison remain interpretable. Such experiments are difficult because changing genetic diversity can change several properties at once. The limitations should guide additional measurements rather than being hidden behind the word feedback.
Two species can strongly affect one another without both undergoing reciprocal evolutionary change. Predation, parasitism, competition and mutualism describe ecological relationships. Coevolution requires an additional claim about evolutionary responses in interacting lineages. Demonstrating that one lineage responds to another establishes only part of that claim.
Original reciprocal-response test
Suppose an archived host population and a later host population are tested against both an archived parasite population and a later parasite population. This creates four comparisons rather than a single “before versus after” observation. The design can ask whether later hosts resist earlier parasites better and whether later parasites perform differently against the corresponding host backgrounds.
Imagine infection probabilities of 0.60 for early hosts with early parasites, 0.30 for late hosts with early parasites, 0.70 for early hosts with late parasites and 0.50 for late hosts with late parasites. These invented values indicate that host background and parasite background both matter. They do not, by themselves, establish the historical selective causes of the differences.
Replication, inheritance, environmental controls and knowledge of the populations’ histories are still needed. A persistent difference may reflect adaptation to another aspect of the environment, migration or chance. A useful design compares the candidate interaction with alternatives rather than treating every reciprocal performance pattern as a complete historical explanation.
Reciprocal evolutionary change also need not be an endless escalation. Trade-offs, spatial variation and changing frequencies can alter which combination is favoured. A defence can be costly when the relevant antagonist is absent; an exploitative strategy can lose its advantage when the host composition changes.
For mutualisms, measure benefits and costs to both partners under the relevant conditions. The fact that both species are present does not demonstrate that both benefit in every context. Likewise, a benefit observed now does not by itself explain how the relationship originated.
The practical habit is to name each direction separately: ecological effect, inherited response and reciprocal selective consequence. Then state which direction the evidence actually supports.
A population facing environmental change can decline while a better-performing inherited type becomes more common. Evolutionary rescue concerns cases where evolutionary change contributes to reversing a decline that would otherwise threaten persistence. It requires more than evidence that selection occurred or that the population later recovered.
Bell and Gonzalez’s experimental research tested conditions affecting evolutionary rescue. Population size, available variation and the severity of environmental challenge are central to interpreting such outcomes. The possibility of rescue is a research question, not a reason to assume that threatened populations will adapt quickly enough.
Original declining-population case
Begin with ninety poorly matched individuals and ten better-matched individuals in a closed asexual model. Their fixed multiplication factors are 0.7 and 1.2 respectively. After one interval the expected counts are sixty-three and twelve, giving seventy-five in total. The better-matched type rises from ten per cent to sixteen per cent while total abundance falls.
After five intervals, the expected counts are 90 × 0.7⁵, about 15.13, and 10 × 1.2⁵, about 24.88. Total abundance is about 40.01. By ten intervals the counts are about 2.54 and 61.92, giving about 64.46. The deterministic trajectory has begun to recover because the growing type increasingly dominates.
Real populations are not guaranteed to follow those expectations. A rare beneficial lineage can be lost by chance. Mate limitation, environmental fluctuations or density effects can change growth. Fractional expected individuals in a model do not ensure that a small actual population survives a bottleneck.
Recovery might also result from immigration, improved conditions or plastic responses. To attribute it to inherited change, investigate those alternatives and establish the relevant contribution. A graph that falls and rises is insufficient on its own.
This is why predictions of rescue need demographic and genetic evidence together. Monitoring frequency without abundance can celebrate adaptation while missing an approaching population collapse.
Many classroom examples vary one condition while holding everything else fixed. That is useful for understanding a mechanism. Real changes may alter several conditions simultaneously, making it necessary to distinguish the isolated effect from the net outcome.
Original changed-context exercise
Return to the defence model. Without grazers, multiplication factors were 1.6 for undefended organisms and 1.2 for defended organisms. Under heavy grazing, survival probabilities were 0.40 and 0.80, giving contributions of 0.64 and 0.96. Starting with equal numbers, the defended type reaches frequency 0.96/(0.64 + 0.96) = 0.60 in the next census.
Now suppose a new environment reduces grazing but also reduces the resource available for reproduction. Supply multiplication factors 1.0 and 0.9, and survival probabilities 0.90 and 0.95. Contributions become 0.90 and 0.855. Starting equally, the defended frequency becomes 0.855/1.755, approximately 0.487. The relative ranking reverses, while both types decline in absolute expectation.
The original reasoning transfers: multiply the correctly ordered stage contributions, then normalise by total contribution to obtain frequency. The original numerical ranking does not transfer because two biological relationships changed. Keeping the procedure is different from keeping the answer.
Next ask what would be observed if the defended type also arrived from another population. Its local frequency could rise despite the local contribution disadvantage. A census after immigration would then combine movement and selection. Sampling before and after relevant stages would help distinguish the processes.
Finally, consider measurement. If defended organisms are easier to detect, observed proportions may differ from actual proportions. Changing water clarity could alter that detection difference. A model about biology and an observation model about detection answer different parts of the same field question.
A complete transfer explanation therefore names the changed ecological conditions, recalculates the life-cycle outcome and identifies any new observation problem. It does not simply announce that the system is more complicated.
Does a useful trait prove adaptation?
No. Current usefulness, inherited variation and historical origin are different claims. A trait may be useful in a new context without having evolved through selection for that use. Investigate the relevant alternatives, including plasticity, ancestry and correlated effects, before presenting an adaptive history as established.
Does rapid change have to be evolutionary?
No. Abundance, behaviour and community composition can change rapidly without demonstrated inherited change within the focal population. Evolution can also occur rapidly. Speed alone does not distinguish the processes; evidence about variation, inheritance and transmission does.
Does local adaptation mean each population is best at everything at home?
No. The claim concerns defined performance comparisons in specified environments. One population may perform well overall, and a particular trait may not represent complete reproductive contribution. Distinguish local-versus-foreign from home-versus-away comparisons and state the life stages measured.
Does an increasing advantageous type mean the population is safe?
No. Relative advantage can occur while all types decline. Track abundance, demographic rates and the probability of persistence alongside frequencies. A deterministic increase in the frequency of a rare type does not eliminate the possibility of its stochastic loss.
When is a feedback claim too strong?
When one direction has been demonstrated and the return direction has merely been assumed. An evolved trait affecting nutrient cycling is an ecological consequence. To establish a feedback, show how the resulting ecological change affects subsequent evolutionary dynamics under the relevant conditions.
What is a useful first repair for a confusing study?
Write the life cycle and the observation schedule. Mark migration, reproduction, survival, environmental change and sampling. Then identify the denominator for each reported rate. Many apparent contradictions arise because studies counted different stages or compared frequencies with totals, rather than because they tested incompatible biological theories.
Start with a two-type population on paper. Supply descendant contributions, calculate next counts and frequencies, and change the environment. Keep the total visible. This simple exercise builds the distinction between relative advantage and population growth before adding the complications of age structure, dispersal or resource feedback.
Next, construct a two-population, two-environment performance comparison. Explain both local-versus-foreign and home-versus-away results. Change the overall quality of one environment without changing which population wins within it. If you can explain why the two comparisons now differ, you have understood the structure rather than memorised a slogan about local adaptation.
For research reading, choose one primary study linked above and draw its actual comparison. Identify what was varied, what was measured and which alternative explanation the design reduces. Separate experimental findings from the authors’ broader interpretation. A strong paper can support an important conclusion without answering every related question.
For field applications, begin with appropriately collected existing records or observations that do not disturb organisms. Useful records include the sampling unit, date, stage, effort, environmental conditions and uncertainty in identification. A high-quality observation is one another reader can interpret; a large collection of unlabelled counts can be much less useful than it appears.
The connected route begins with Ecology and Evolution. Continue to migration–selection balance and local adaptation for spatial processes, phenotypic plasticity for environmental responses, and Life History Theory for survival and reproductive schedules.
The Gold Standard of Education connects this work with learning that survives a changed task. How X Works offers wider mechanism-based reading, while The Gold Standard master connects the subject collection. Keep the final explanation concrete: which ecological relationship changes inherited contribution, what changes as a result, and what evidence would show the explanation is wrong?