A biosphere represents the parts of the Earth where life exists. This ecological system includes life in parts of the atmosphere, hydrosphere and lithosphere.

Ecological organization works as a nested model. Individual organisms make up populations, and interacting populations form communities. Communities then interact with physical environments in ecosystems. Taken together, all ecosystems form the biosphere. This hierarchy helps us model natural systems at different scales and explore how a disturbance at one level spreads to others.

An individual organism is a single living entity that can carry out the processes of life. Each individual belongs to a species.

Under the biological species concept, a species is a group of organisms that can interbreed and produce fertile offspring. Here, fertile offspring is the key phrase. Mating by itself doesn’t show that two organisms belong to the same biological species.

The concept works well for organisms that reproduce sexually. It’s harder to apply to fossils, geographically separated populations or organisms that reproduce asexually. These limitations don’t change the definition required here, but they do explain why taxonomists also draw on anatomical and genetic evidence.

Classification arranges organisms into groups based on shared characteristics. Without it, the immense diversity of life would make identification and communication unmanageable. Group membership can also support cautious predictions: species in the same genus are expected to share several traits.

Binomial naming

Each scientific species name has two parts. The first gives the genus, a taxonomic group of closely related species. The second identifies the species within that genus. The genus begins with a capital letter, while the second name is written in lower case. When typed, both words are italicized, as in Vulpes vulpes. When handwritten, each word is underlined separately.

Common names differ across languages and regions. The binomial system instead gives scientists a shared, relatively precise label.

A taxonomist is a scientist who identifies, names and classifies organisms. Identification is the process of determining the species of an individual organism.

To do this, taxonomists can compare a specimen with authenticated material held in a museum, herbarium or another reference collection. They may also carry out deoxyribonucleic acid surveys. DNA is the hereditary molecule whose sequence can be compared among organisms, making genetic evidence especially useful when species look very similar.

A dichotomous key is an identification tool that gives two contrasting choices at each step. Start with the specimen’s observable features, choose the statement that matches, then follow its direction until you reach a species name. Objective features, such as leaf arrangement, should be used instead of vague descriptions such as “large”. If neither choice fits, go back to the previous step—an early mistake will send the identification down the wrong branch.

Identification applications compare photographs, sounds or location data with stored records. They’re quick and useful during fieldwork, but any suggestion should be checked against a key, an expert-verified database or a reference specimen. Errors may result from poor image quality, juvenile forms or incomplete databases.

A population is a group of organisms of the same species living in the same area at the same time, and which are capable of interbreeding. This makes it an interbreeding unit. A species may consist of one population or several populations separated geographically.

For a local ecosystem investigation, clearly define the study area and time, identify the population under study, and record the relevant biotic and abiotic variables. Where possible, state the independent variable, dependent variable and controlled variables. Field conditions such as light or temperature often can’t be controlled; monitor them instead. Otherwise, an apparent biological pattern could be caused by an environmental change that wasn’t recorded.

A biotic factor is a living part of an ecosystem that affects organisms, such as competitors, predators, prey, parasites or pathogens. An abiotic factor is a non-living physical or chemical condition that can influence organisms.

Together, these factors shape a population's distribution—the spatial pattern showing where its members occur. A plant, for instance, may tolerate the soil's moisture and acidity but still be absent because competitors outcompete it or grazers feed on it heavily. In contrast, suitable food cannot make up for an abiotic condition outside the organism's tolerance range.

Distribution data show an association, but don't necessarily prove causation. A convincing investigation measures the organism and plausible environmental factors at the same locations, while also considering how variables interact.

Temperature, sunlight, pH, salinity, dissolved oxygen and soil texture can limit where species live. Their effects differ. Temperature alters metabolic rates, while light controls photosynthesis. pH influences enzyme function and nutrient availability; salinity affects water balance. In aquatic environments, dissolved oxygen constrains respiration. Soil texture determines drainage, aeration and water retention.

Measuring abiotic factors

Take measurements at representative sites using calibrated instruments. Keep depth, height, time and exposure period consistent. Suitable methods include:

• measuring temperature with a thermometer or temperature probe, recorded in kelvin (K) or degrees Celsius ($^\circ\text{C}$)
• measuring light intensity with a light meter, recorded in lux (lx)
• measuring pH with a calibrated pH probe; pH has no unit
• measuring salinity with a conductivity or salinity probe, commonly reported in grams per kilogram ($\text{g kg}^{-1}$)
• measuring dissolved oxygen with an oxygen probe, commonly reported in milligrams per litre ($\text{mg L}^{-1}$)
• determining soil texture by separating or estimating the proportions of sand, silt and clay.

Conditions vary from place to place, so take repeated measurements at multiple points. Measurements should also be repeated at suitable times to account for daily and seasonal changes. A data logger is an electronic instrument that automatically records sensor readings at set intervals. It can detect fluctuations that a short visit might miss, though its probes still need calibration and secure, consistent placement.

Plot the abundance of the study species against each quantified factor. The pattern may show where tolerance limits lie. However, correlated factors and biotic interactions mean that a single graph rarely proves one cause.

A niche describes the particular set of abiotic and biotic conditions and resources upon which an organism or a population depends. A species’ ecological niche is its role within an ecosystem. It covers every biotic and abiotic interaction that affects population growth, survival and reproduction, including the way the species obtains food.

For the Eurasian otter, Lutra lutra, these niche parameters include access to unpolluted freshwater or coastal water and bankside cover for resting and breeding. Fish and crustaceans provide food, while water temperature and oxygen conditions must remain tolerable. Its niche also involves interactions with prey and competitors, as well as vulnerability to parasites and human disturbance. So, a niche is far more than a location: habitat tells us where the otter lives; its niche describes the conditions, resources and interactions that allow it to live there.

Population interactions can change abundance, behaviour and selection. For each participant, the effect can be shown as beneficial (+), harmful (−) or broadly neutral (0).

• Herbivory occurs when an animal consumes living plant tissue (+/−). For example, browsing by snowshoe hares can reduce the survival of young willows. As shoots become scarce, hare numbers are limited. Selection favours plants with defensive chemicals or rapid regrowth, while herbivores that can detoxify those chemicals gain an advantage.
• Predation occurs when a predator captures and consumes prey (+/−). Owls reduce vole numbers, but the resulting vole scarcity later restricts owl reproduction. Better camouflage may be favoured in prey and sharper senses in predators. Prey may also become more vigilant or change their feeding times.
• Parasitism occurs when a parasite takes resources from a host and harms it, usually without killing it immediately (+/−). Ticks can reduce deer condition and reproductive success. This places host resistance and parasite traits for attachment or immune evasion under opposing selective pressures.
• Mutualism is an interaction between species in which both benefit (+/+). Mycorrhizal fungi receive sugars from tree roots and increase the tree's uptake of mineral nutrients and water. Selection may favour traits that improve exchange or partner recognition.
• Disease is impaired functioning caused here by infection with a pathogen. Chytrid fungal infection can reduce amphibian survival. Transmission may increase as host contact rises, while selection favours resistant hosts and pathogen variants that can infect them.
• Competition occurs when organisms require the same limited resource, reducing access for at least one competitor (−/−). Intraspecific competition takes place within one species, whereas interspecific competition occurs between species. Competing mangrove seedlings may be selected for faster growth or greater salt tolerance. Their behaviour may also shift toward different feeding places or times.

Modelling predator–prey feedback

An increase in prey makes more food available to predators, so predator numbers may rise after a delay. Greater predation then reduces prey abundance. The resulting food shortage lowers predator abundance, allowing the prey population to recover. This is negative feedback because the response counteracts the initial change.

These models can predict broad responses to disturbance, but they simplify age structure, migration, alternative prey, weather, disease and time delays. The predicted cycle should be treated as a testable expectation, not a timetable.

Carrying capacity is the maximum size of a population determined by competition for limited resources. It is commonly shown as $K$, where $K$ is carrying capacity (individuals; a dimensionless count).

A limiting factor is an environmental condition or resource that restricts population growth because of its shortage or intensity. Biotic resources that affect $K$ include food and prey, as well as nesting sites and access to mates. Abiotic resources or conditions include water, light, mineral nutrients, oxygen and space; temperature and suitable pH also play a role.

Carrying capacity can change. Drought may reduce the water available, lowering $K$. Habitat restoration, by increasing shelter or food, may raise it. The relevant capacity therefore applies to a specified population in a particular place and period.

A density-dependent factor is a limiting factor that has a stronger effect as population density rises. When density is high, organisms compete more intensely for limited food or space. Predators find prey more easily, while pathogens spread more readily between hosts. As a result, death rates may rise or birth rates may fall, pushing the population back toward carrying capacity.

This is a negative feedback mechanism: a process in which a change triggers responses that oppose it and move the system toward equilibrium. When population density falls, competition and transmission may weaken, so survival and reproduction can recover.

A density-independent factor has an effect that is not determined by population density. Examples include a severe freeze, wildfire or toxic spill. Such a factor can cause a major population change, but it does not normally regulate abundance around $K$ because its strength does not increase consistently as the population becomes denser.

Exponential growth occurs when a population increases at a rate proportional to its current size. If resources are abundant and no limiting factors act effectively, more individuals reproduce during each interval. The result is a J-shaped curve, with both population size and the absolute rate of increase growing progressively larger.

Growth slows as density-dependent limits become stronger. An S-shaped, or logistic, curve starts gradually, becomes steeper while resources remain plentiful, then levels off near $K$. Instead of staying exactly constant, the population may fluctuate around this equilibrium.

A boom-and-bust pattern describes rapid population growth followed by a sharp crash. This can happen when a population overshoots the available resources. Growth continues for a time, resources become depleted, then mortality rises. Unlike a stable S-curve, the decline may be severe. Recovery can also be slow if the habitat has been damaged.

Growth curves provide useful models for predicting the direction and approximate pattern of change. However, real populations rarely follow perfectly smooth curves because limiting factors, carrying capacity and time delays change.

Humans have steadily weakened many of the natural limits on population growth. Large predators no longer control most human populations. Agriculture, sanitation, vaccination and medicine have reduced food shortages and disease, while refrigeration, transport and water engineering have eased local resource scarcity. Technology also allows people to settle in conditions that would otherwise be unsuitable.

But those ecological consequences haven’t disappeared; they’ve shifted. As populations and consumption rise, more water, biomass, minerals and energy are extracted. Habitat conversion, waste, pollution and greenhouse-gas emissions damage ecosystems. This may reduce the carrying capacity of other species and threaten long-term human sustainability.

Technology can therefore improve survival in the short term while making delayed environmental limits more severe. A population model that leaves out consumption per person and ecosystem degradation will give a misleading picture of sustainability.

Assessing human carrying capacity is difficult because the human niche is broad and keeps changing. Most non-human populations rely mainly on resources found within their ecosystem, and their numbers may fluctuate around equilibrium. Humans, by contrast, transport food, water, fuels and materials across great distances. A city can therefore exceed its local ecological capacity by relying on ecosystems elsewhere.

Technology affects which resources are usable and how efficiently people use them. Consumption varies greatly between societies and over time, so the same number of people can place very different demands on the environment. Human habitats also change rapidly through urbanization, environmental degradation and climate change.

Estimates therefore depend on assumptions about diet, living standards, trade, technology, waste and acceptable ecosystem damage. Such estimates are disputed and can only show carrying capacity under present conditions—a useful “now” value rather than an unchanging global ceiling.

Population abundance is the amount of a species found within a defined area. It may be expressed as a count or another suitable measure, such as cover. Since a complete census is often impractical, ecologists instead sample representative parts of the habitat.

• Random sampling gives every possible sampling location an equal chance of being selected. This reduces deliberate placement bias and works well in a relatively uniform area.
• Systematic sampling takes samples at regular spatial intervals. It provides even coverage and is useful when position or distance may affect the results. However, a repeating environmental pattern could introduce bias.
• Transect sampling records organisms along a line or strip that crosses an environmental gradient. Continuous recording provides more detail, while interrupted sampling at fixed intervals is quicker.

The procedure should fit the question. Random sampling can estimate mean abundance across a broadly homogeneous field, whereas systematic sampling provides even spatial coverage. To investigate change from one condition to another, use a transect—for example, to measure the effect of increasing distance from a footpath. In every case, replication and a sufficient sample size improve reliability.

A quadrat is a frame used to mark out a known area for ecological sampling. Random quadrat sampling suits plants and other non-mobile organisms, since individuals stay in place while the samples are collected.

Place a coordinate grid over the study area and use a random-number generator to choose coordinates. Put equal-sized quadrats at the selected locations, using the same counting rule each time. Then calculate population density, the number of individuals per unit area:

If the samples are representative, multiply mean density by the total habitat area to estimate population size.

Percentage cover estimates the proportion of a quadrat's area covered by the target organism. For a gridded quadrat, count the occupied squares and follow a consistent rule for partial squares. This method works well for mats, grasses and clonal organisms, where separating individuals is difficult.

Percentage frequency is the number of occurrences divided by the number of possible occurrences, expressed as a percentage:

Cover and frequency provide estimates of abundance, but neither gives actual population size.

To measure change along a gradient, set quadrats continuously or at fixed intervals along a transect, then plot abundance against distance. Quadrats are less suitable for very large organisms, tiny cryptic organisms or mobile species. They also cause problems when connected plant stems make an “individual” impossible to define. Seasonal invisibility and patchy distribution can affect the estimates too.

Capture–mark–release–recapture is a sampling method used to estimate a mobile population. It works from the proportion of marked individuals found in a later sample. First, capture a sample and mark each individual harmlessly. Release them at the capture site, allow enough time for mixing, then collect a second sample using the same effort.

The Lincoln index gives an estimate of total population size:

$P=\frac{MN}{R}$

Here, $P$ is estimated population size (individuals; a dimensionless count). $M$ is the number caught and marked initially (individuals; a dimensionless count), $N$ is the total number caught in the second sample (individuals; a dimensionless count) and $R$ is the marked individuals found in the second sample (individuals; a dimensionless count).

For example, if $M=48$, $N=40$ and $R=12$, then $P=(48\times40)/12=160$ individuals. When $R$ is low, the estimate is large and becomes especially sensitive to chance.

The method assumes:

• the population is closed to births, deaths, immigration and emigration
• marks are retained, recognized and do not alter survival or capture probability
• marked individuals mix fully back into the population
• marked and unmarked individuals are equally likely to be captured
• sampling effort is representative and the two samples are taken close enough in time.

Breaking these assumptions introduces bias. Lost marks or immigration usually reduce $R$, which inflates the estimate. Trap-shy marked animals have the same effect. Trap-happy animals increase $R$ and can produce an underestimate. Reliability can be improved by using adequate samples and short intervals. Trapping should be standardized, marks should be harmless and durable, and estimates should be repeated.

A community is a collection of interacting populations within the ecosystem. It contains several species rather than merely several individuals.

A local pond community includes populations of submerged plants, algae, insect larvae, snails, fish, amphibians, fungi and bacteria. Feeding, competition, disease and decomposition link these populations. Water, sediment and temperature affect the community, but they aren’t members because they are non-living. Instead, they form part of the wider ecosystem.

Habitat is the location in which a community, species, population or organism lives. A complete description may give the geographical location, physical position and ecosystem type, as well as the environmental conditions needed for survival.

For instance, a freshwater mussel’s habitat might be the gravel bed of a cool, oxygenated river reach within a named catchment. Simply writing “river” is too broad. The physical micro-location and necessary conditions must also be included.

Habitat and niche aren’t the same. Habitat refers to the place occupied; niche covers how an organism obtains resources and interacts with biotic and abiotic conditions. One local ecosystem can contain many habitats and overlapping communities.

An ecosystem is a community and the physical environment with which it interacts. An open system is a system across whose boundary both energy and matter can enter and leave.

In a pond ecosystem, solar radiation and organic matter may enter. Organisms, water and dissolved nutrients can cross the boundary too. Heat escapes, water evaporates, gases diffuse and organisms migrate. Inside the boundary, matter moves among the water, sediment and organisms. Energy passes through biological processes before eventually dispersing as heat.

The boundary chosen forms part of the model. A single pond could be treated as an ecosystem, as could a wetland complex or an entire drainage basin, provided that its components, boundary, inputs, outputs and interactions are clearly stated.

Sustainability is a system property that allows its functions and processes to continue over time. In a steady-state ecosystem, inputs and outputs balance over an appropriate period. Major storages therefore fluctuate around relatively stable values instead of rising or falling indefinitely.

The system is dynamic rather than motionless. Organisms are born and die, nutrients move, and seasonal populations change. Even so, feedback and recycling can preserve the system’s overall organization. Some ancient woodland and lake ecosystems have persisted for very long periods despite constant internal change.

A flow diagram can test this idea. It compares inputs and outputs that cross the system boundary while also showing internal storages and transfers. If humans keep removing biomass faster than it regenerates, the storage shrinks and the system is no longer in steady state. Models can therefore show the direction of change, though incomplete measurements and changing boundaries limit exact prediction.

A tipping point is a critical threshold where one relatively small additional change pushes a system into a substantially different state. The original ecosystem may collapse, replaced by a new equilibrium that different feedbacks maintain.

Take a shallow lake receiving nutrient-rich agricultural runoff. Moderate enrichment boosts algal growth. Once a threshold is crossed, dense blooms block light and submerged plants die. Decomposition consumes oxygen, while bottom sediments release more nutrients. This new turbid, algae-dominated state can persist even when nutrient inputs fall because internal positive feedback now maintains it.

This is why models can’t always predict disturbance precisely. Thresholds may be uncertain, responses delayed and feedbacks hidden. Even so, models can identify warning indicators—declining plant cover, falling oxygen and rising algal biomass—and show why acting before the threshold is safer than trying to restore the ecosystem afterward.

A keystone species has a disproportionately large effect on community structure compared with its abundance. If it is removed, food webs may be reorganized and habitat complexity may fall. In severe cases, its loss can contribute to ecosystem collapse.

In North Pacific kelp systems, sea otters, Enhydra lutris, feed on sea urchins. When otter numbers decline, urchin populations can rise and overgraze the kelp. This removes both food and structural habitat used by many species. Otter recovery can help restore kelp-dominated conditions.

Beavers, Castor canadensis, build dams that create wetlands. Their ponds and flooded margins change water flow, sediment and oxygen conditions. They also affect the habitat available to fish, amphibians, birds and aquatic plants. Removing beavers can drain these habitats, leaving a simpler community.

Keystone status depends on context. A species’ effect must be demonstrated in the particular ecosystem concerned. Protecting a keystone species can be efficient, but it cannot replace the protection of habitat, ecological processes and the wider community.

Biosphere integrity describes the living world's capacity to maintain ecological processes through its genetic diversity, species and functioning ecosystems. In the planetary boundaries model, which identifies a safe operating space for Earth systems, change to biosphere integrity has already passed a critical threshold.

Species diversity and ecosystems depend on each other. Different species carry out ecological roles; intact ecosystems provide the habitats and niche requirements those species rely on. Human activities disrupt both sides of this relationship through habitat conversion, exploitation, pollution, invasive species and climate change.

Extinction rates offer key evidence. Observed rates sit far above long-term background rates, showing that species are disappearing faster than evolutionary processes can replace them. Extinction can't be reversed. Even before complete ecosystem collapse becomes visible, the loss of interacting species can weaken how an ecosystem functions.

The boundary is a model, not a precise prediction of a single collapse date. It brings together evidence at a planetary scale and signals increasing systemic risk. Uncertainty about the exact threshold doesn't mean the risk is absent.

Avoiding critical tipping points calls for reversing deterioration, not simply slowing it down. Protecting ecosystem integrity keeps habitats connected and maintains ecological interactions, as well as the abiotic conditions species depend on. This preserves their niche requirements for feeding, shelter, movement, breeding and survival.

Effective action may involve preventing further habitat conversion, reducing overexploitation and pollution, controlling damaging invasive species, reconnecting fragmented habitats or restoring degraded ecosystem processes. Species-focused measures are still useful when populations are already very small. However, they work best alongside ecosystem protection.

The guiding question for this topic now comes into focus. Models of populations, feedback, carrying capacity and ecosystem flows can predict likely directions under human disturbance and identify thresholds. They can’t remove uncertainty because niches, technologies, feedback delays and environmental conditions change. Their proper role is to compare plausible outcomes and support precautionary decisions before change becomes irreversible.

A clade is a taxonomic group made up of a common ancestor and all of its descendants. Every member evolved from that ancestor, so the group shows evolutionary relationships.

A cladogram is a branching model used to represent hypotheses about evolutionary relationships. Each branching point stands for a common ancestor. Groups that share a more recent branching point are inferred to be more closely related. The position of the tips across the page doesn’t matter; the branching pattern does.

Classifying organisms by clades ties each group directly to common ancestry. It can use evidence from DNA and proteins, while new evidence may lead scientists to revise relationships. Clades can also predict inherited traits that their members may share. A cladogram is still a model, though. Its results depend on which organisms and evidence were sampled, and different datasets may produce competing trees.

A taxon is a named group in biological classification. In the traditional hierarchy, taxa run from broad to narrow: kingdom, phylum, class, order, family, genus and species.

This ranking system doesn’t always reflect evolutionary divergence. Early classification depended heavily on visible similarities. However, unrelated organisms may independently evolve similar features when they face similar selective pressures. Closely related organisms can also look very different if they adapt to different niches. Horizontal gene transfer and hybridization add further complications, as does disagreement over species boundaries.

Fixed ranks may suggest that two families or orders show equivalent amounts of evolutionary difference, even when they don’t. Genetic evidence can move a species into another group. It may also show that a traditional taxon leaves out some descendants of its supposed common ancestor. Clade-based classification represents branching ancestry more accurately, though names and groupings may need to be revised.

The fundamental niche describes the range of conditions and resources in which a species could survive and reproduce if there were no limiting factors. It represents the species' potential niche under the abiotic conditions it can tolerate.

The realized niche of a species is the actual mode of existence, which results from its adaptations and competition with other species. Predation, disease, competition and access to partners may narrow it compared with the fundamental niche.

In Britain, the Eurasian red squirrel, Sciurus vulgaris, can physiologically use a broad range of woodland. However, its realized niche is becoming concentrated in some coniferous and upland forests. Competition from introduced eastern grey squirrels is weaker there, and the risk of disease transmission is lower. So, the suitable abiotic space is broader than the area the population actually occupies.

This distinction affects conservation. A climate-only model might identify the fundamental conditions, yet overestimate future distribution when competitors or pathogens occupy the same area. Realized-niche models are often more realistic, though interactions may change as populations move.

A life cycle describes the developmental and reproductive stages an organism passes through. Species vary in their age at maturity and reproductive behaviour, as well as in offspring number, parental investment and lifespan.

An $r$-strategist is a species adapted to exploit temporary resources or newly available habitats. It reproduces rapidly and produces many offspring, with limited provision for each individual’s survival. Common tendencies include early maturity, a short lifespan and rapid population increase. Unpredictable, density-independent disturbances also have strong effects.

A $K$-strategist is a species adapted to relatively stable communities. It produces few offspring but invests substantially in each one, giving them a high probability of survival. These species tend to mature later, live longer and provide repeated parental care. Density-dependent factors regulate their populations near $K$.

Typical tendencies along the $r$–$K$ life-cycle continuum.

The two strategies are ends of a continuum, not rigid boxes, so a species may show traits from both. Life-cycle strategy is also linked to succession. Rapid colonizers can exploit early, disturbed stages, whereas slower-reproducing competitors tend to do better in stable later-stage communities.

Classification, niche requirements and life cycles each reveal a different side of vulnerability. Classification identifies close relatives that may share susceptible traits. It also shows when the loss of a species would wipe out an evolutionarily distinct lineage. Niche information pinpoints the resources, conditions and interactions that need protection, while life-cycle knowledge shows how quickly a population may recover and which stage human activity disrupts.

Climate warming can shift seasonal timing. Earlier spring temperatures, for example, may cause woodland plants to flower before their usual pollinating insects emerge. Pollination and seed production may then fall. The insects also lose an important food source, which affects their growth and reproduction. Both species may remain in the same habitat, yet the timing mismatch still disrupts their realized niches and life cycles.

Species that are long-lived, mature late and produce few offspring generally recover slowly from hunting, bycatch or habitat loss. Rapidly reproducing species may recover faster, though they can still decline when disturbance removes an essential niche requirement. Classification, niche information and life-cycle knowledge therefore improve predictions of human impact. Monitoring remains essential because ecological interactions and responses to climate can change.