Classification, biodiversity and conservation

What is a species? Three species concepts

Classification is the ordering of all the organisms on Earth into groups. The basic group is the species, but a biologist working with bacteria, with fossils or with two populations that never meet cannot use the same test as a biologist studying birds in one wood, so three species concepts are recognised. The biological species concept defines a species as a group of organisms with similar characteristics that can interbreed to produce fertile offspring, and that do not normally interbreed with other such groups. Interbreeding that produces fertile offspring is the test: a horse and a donkey can mate and produce a mule, but the mule is sterile, so the horse (Equus ferus) and the donkey (Equus asinus) are separate species. A grey wolf (Canis lupus) and a domestic dog (Canis lupus familiaris) can interbreed and produce fertile young, which is why the dog is classified as a subspecies of the wolf and not as a separate species.

The biological species concept breaks down in three situations you should always quote. Organisms that reproduce asexually, such as bacteria, many protoctists and plants that spread by runners or bulbs, never interbreed, so the concept cannot be applied to them at all. Extinct organisms known only from fossils cannot be interbred, so only their structure can be compared. Populations that live in different parts of the world and never meet cannot be shown to interbreed, so in practice the decision is made on other evidence.

The morphological species concept defines a species by a consistent set of measurable physical features that its members share and that separate them from other species. It is the concept a dichotomous key uses: you decide which species an organism belongs to by comparing visible characters such as leaf shape, flower structure or the number of legs. Its great advantage is that it works for fossils, for asexual organisms and for organisms that cannot be kept and crossed in captivity. Its weakness is that structure can mislead in both directions. Cryptic species look almost identical but never interbreed: the common pipistrelle and the soprano pipistrelle, two bats found in the same British woodlands, were treated as one species until their echolocation calls and their DNA were compared. In the other direction, the caterpillar and the adult butterfly of one species look so different that they were once given different names, and in many species the male and the female look different too.

The ecological species concept defines a species by the niche it occupies, that is by the way it uses its environment. Members of one species exploit the same resources in the same way: they feed on the same food, are active at the same time of day, breed in the same places and tolerate the same range of conditions. Two populations are therefore treated as separate species if they occupy clearly different niches, even when they look alike and could be crossed in a laboratory, and the concept can be applied to asexual organisms and to fossils, because a fossil's structure still tells you how it fed and moved. In practice examiners want you to discuss the concepts together: the biological concept is the strongest test when it can be used, and morphology or ecology is used when it cannot.

• Biological species concept: members can interbreed to produce fertile offspring and do not normally interbreed with other groups. Fails for asexual, extinct and geographically separated organisms.
• Morphological species concept: a consistent set of shared physical features. Works for fossils and asexual organisms, but cannot separate cryptic species and can split one species into several.
• Ecological species concept: the niche a population occupies. Works even for organisms that never meet or never interbreed, but needs detailed field data on how each species uses its environment.

Three domains: Archaea, Bacteria and Eukarya

The domain is the highest and broadest taxon. There are three of them: Archaea, Bacteria and Eukarya. This three-domain system is not based on how organisms look but on phylogeny, their evolutionary relationships, and the evidence used to build it is the nucleotide sequence of ribosomal RNA and of the genes that code for it. Ribosomal RNA is ideal for the job because every cell needs ribosomes and so every cell has rRNA genes, yet the sequence changes slowly enough that the differences between two organisms reflect how long ago they shared a common ancestor. Comparing these sequences showed that the prokaryotes are not one group: some prokaryotes are more closely related to eukaryotes than they are to the other prokaryotes, and that is why they are placed in two separate domains.

Archaea and Bacteria are both prokaryotes. Neither has a nucleus, a nuclear envelope or any membrane-bound organelle, both have a single circular molecule of DNA in a nucleoid, both have 70S ribosomes and both are surrounded by a cell wall, except for the bacteria in the genus Mycoplasma. They also both divide by binary fission, not by mitosis. If a question asks you to state one feature that shows that Archaea and Bacteria are prokaryotes, any of those will do.

The syllabus limits the differences between the two domains to three features. Membrane lipids: the lipids of an archaeal membrane are branched hydrocarbon chains joined to glycerol by ether bonds, whereas the lipids of a bacterial membrane are straight-chain fatty acids joined to glycerol by ester bonds. Ether-linked branched lipids are far more stable at high temperature and in strong acid, which is one reason so many Archaea are extremophiles: Sulfolobus lives in hot, acidic volcanic springs and methanogens such as Methanobrevibacter live in the rumen and in the human gut. Ribosomal RNA: the base sequence of archaeal rRNA differs from that of bacterial rRNA, and both the rRNA and the enzymes of transcription and translation of Archaea are more similar to those of eukaryotes than to those of Bacteria, which is the main evidence that Archaea and Eukarya share a more recent common ancestor. Cell walls: a bacterial cell wall contains peptidoglycan (murein), the network of sugar chains cross-linked by short peptides, whereas no archaeal cell wall contains peptidoglycan; archaeal walls are built from proteins, glycoproteins or polysaccharides such as pseudopeptidoglycan, and some Archaea have no wall at all.

The differences have practical consequences. Because penicillin stops the cross-linking of peptidoglycan, it kills growing Bacteria but has no effect on Archaea or on eukaryotes, which is exactly why it is a useful antibiotic: it attacks a structure that human cells do not have. It also means that a patient with an infection caused by an archaeon or a fungus will not respond to penicillin.

The taxonomic hierarchy, the four eukaryotic kingdoms and the viruses

Within the domain Eukarya, organisms are placed in a taxonomic hierarchy, a series of ranked groups in which each level is contained inside the one above it. The ranks, from the broadest to the narrowest, are kingdom, phylum, class, order, family, genus and species. Every organism has one position at each level, and the further down the hierarchy you go, the fewer organisms there are in the group and the more features its members share. Take the grey wolf: domain Eukarya, kingdom Animalia, phylum Chordata, class Mammalia, order Carnivora, family Canidae, genus Canis, species Canis lupus. Take a flowering plant, the wild cabbage: kingdom Plantae, phylum Angiospermophyta (the flowering plants), class Dicotyledonae, order Brassicales, family Brassicaceae, genus Brassica, species Brassica oleracea. Hierarchies like this are also testable: a group at any level is called a taxon, and a taxon must be monophyletic, meaning that all its members descend from one common ancestor and that the ancestor's other descendants are included too.

In the field and in the laboratory you identify an unknown organism with a dichotomous key, a series of paired statements in which each pair offers two mutually exclusive choices, so that each answer takes you to the next pair until you reach a name. A good key uses characters that are present all year and vary little, such as leaf arrangement and flower structure rather than leaf colour, and it never asks about a feature that is absent from some specimens.

Four kingdoms of eukaryotes are recognised here. Protoctista are the eukaryotic organisms that are not fungi, plants or animals: most are single-celled, such as Amoeba, Paramecium and Chlorella, while others are simple multicellular forms without true tissues, such as the seaweeds Ulva and the slime moulds; some are autotrophic and some heterotrophic, and some, like Euglena, can be both. Fungi are heterotrophic eukaryotes that feed by saprophytic absorption: they secrete digestive enzymes onto their food and absorb the soluble products, and their body is a network of fine threads called hyphae forming a mycelium; the cell wall is made of chitin and the storage carbohydrate is glycogen, and they may be saprophytes or parasites but never photosynthesise. Examples are Mucor and Rhizopus, the moulds on bread, Penicillium, the yeast Saccharomyces and the mushroom Agaricus. Plantae are multicellular autotrophs that photosynthesise using chlorophylls a and b inside chloroplasts; their cells have cellulose walls and they store carbohydrate as starch or sucrose; they show cell and tissue differentiation into organs, and they are the bryophytes, ferns, conifers and flowering plants, such as Pinus and Helianthus. Animalia are multicellular heterotrophs that feed holozoically, taking food into the body and digesting it internally; their cells have no cell wall, they store glycogen and lipids, most have specialised nervous and muscular tissue and most are able to move at least at one stage of the life cycle. Examples are Homo sapiens, the honeybee Apis mellifera and the Atlantic salmon Salmo salar.

Viruses cannot be fitted into this system at all, because they are not cells. A virus has no cytoplasm, no ribosomes, no enzymes to release energy and no metabolism of its own, and it cannot reproduce outside a living host cell: it is an obligate intracellular parasite. The syllabus limits their classification to two features of the nucleic acid, which give four groups: whether the nucleic acid is DNA or RNA, and whether it is single stranded or double stranded. A double-stranded DNA virus such as herpes simplex, the smallpox virus or human papillomavirus has its genetic material in the same form as a cell, but a single-stranded DNA virus such as human parvovirus B19 does not. A double-stranded RNA virus such as rotavirus carries its genome as two strands of RNA, and a single-stranded RNA virus such as HIV, SARS-CoV-2 or influenza virus carries one. The group a virus belongs to matters clinically: HIV is a single-stranded RNA retrovirus, so it must first copy its RNA into DNA using reverse transcriptase before the cell can use it, which is why drugs such as zidovudine, which inhibit reverse transcriptase, stop it. Influenza virus is a single-stranded RNA virus whose genome is in several separate pieces, and when two influenza viruses infect the same cell the pieces can be reassorted, producing the new virus strains that appear from time to time.

Ecosystems, niches and the levels of biodiversity

An ecosystem is a relatively self-contained, interacting unit made up of a community of organisms and the environment in which they live, so it includes both the living and the non-living parts and the exchanges between them. A pond, a woodland, a coral reef and a rotting log on a forest floor are all ecosystems, and they are 'relatively self-contained' because energy and matter flow across their boundaries but most of the interactions happen inside them. A niche is the role of a species in its ecosystem: the way it uses its environment. That includes the habitat and microhabitat in which it lives, what it feeds on and how it obtains food, when it is active, where and when it breeds, which other species it interacts with as a competitor, predator, prey, parasite or decomposer, and the range of abiotic conditions it can tolerate, such as temperature, pH, light and water availability.

The niche explains how so many species can live together. Two species that occupied exactly the same niche would compete for exactly the same resources, and the better competitor would survive while the other declined and eventually disappeared: this is competitive exclusion. In practice species divide the available resources, or one is restricted to part of its potential range. On rocky shores the barnacle Chthamalus lives high up the shore and Semibalanus lives lower down; when experimenters removed Semibalanus, Chthamalus extended its range downwards, which shows that Chthamalus could live lower down but is pushed out of that part of the shore by competition. Overlap between niches is also the reason that a new species with a similar niche, such as an introduced predator, can devastate a community.

Biodiversity is not one number. It is assessed at three levels, and a good answer states which level is meant. The first is ecosystem and habitat diversity: the number and range of different ecosystems and habitats in an area, such as the woodland, heath, bog, stream and meadow inside one nature reserve. The second is species diversity: the number of species present, which is the species richness, together with their relative abundance, or evenness, which is how evenly the individuals are divided among those species. Species richness alone is a poor measure: a field containing ten species where one species makes up 90 per cent of the individuals is less diverse than a field containing the same ten species in similar numbers, and this is exactly what Simpson's index of diversity measures. The third level is genetic diversity: the genetic variation within each species, that is the number and range of alleles present in its populations.

Genetic diversity is the level most easily overlooked, and it is the one that decides whether a species can adapt. The allele variants in a population are the raw material of natural selection, so a population with plenty of genetic variation is more likely to contain individuals that survive a new disease, a new parasite or a change in climate. A population that has passed through a bottleneck of a few individuals, as the cheetah has, loses most of that variation, and its members become genetically similar, so the whole species is vulnerable to the same disease and inbreeding produces less healthy offspring. Measure any one level of biodiversity and you can compare areas, but you must sample the same way for the comparison to mean anything.

• Ecosystem: a relatively self-contained interacting unit made up of a community of organisms and the environment they live in.
• Niche: the role of a species in its ecosystem, including where it lives, what it feeds on, when it is active, how it interacts with other species and the conditions it can tolerate.
• Level 1, ecosystem and habitat diversity: the number and range of different ecosystems and habitats in an area.
• Level 2, species diversity: the number of species present (species richness) together with their relative abundance (evenness).
• Level 3, genetic diversity: the genetic variation within each species, which determines how well it can adapt to change.

Sampling: random quadrats, transects and mark-release-recapture

You cannot count every organism in a wood or a field, so you sample and use the sample to estimate the whole area. The estimate is only valid if the sample is representative, and that is why random sampling is used. Picking the places that look interesting, or that are easiest to reach, biases the result: a quadrat dropped beside the path, where the vegetation is trampled, gives a very different picture from one in the middle of the field. In a random sample every part of the area has an equal chance of being selected, so the bias is removed, and the results can then be treated with statistics such as a mean with confidence limits, or compared between two areas with a correlation or a test of difference. The practical method is to lay two tape measures along two edges of the area at right angles, use a random number generator or a table of random numbers to give pairs of coordinates, walk to those points and place the quadrat there without choosing. Sampling is also called systematic when the positions are not random, as in a transect, and then you must say why: you are following a gradient, not estimating an average.

The frame quadrat is a square frame of known area. Its size depends on the organisms: 1 m x 1 m is usual for herbaceous plants and 0.25 m x 0.25 m for a lawn or a rocky shore, while trees may need 10 m x 10 m; if the quadrat holds fewer than about ten individuals of the species of interest it is unreliable, and if it holds hundreds the counting takes too long. Abundance in a quadrat can be measured in three ways, and you must state which you are using: density, the number of individuals counted; percentage cover, the proportion of the quadrat covered, which is best for grass or moss where individual plants cannot be told apart; and frequency, the number or proportion of quadrats in which the species appears, which describes how widely spread it is rather than how many there are. Take enough quadrats to be representative, at least ten to twenty, or keep sampling until the running mean stops changing as new quadrats are added, which shows that the sample is now large enough.

A line transect answers a different question. A tape measure is stretched across the area, usually along a gradient such as from the top of a rocky shore down to the sea, or from the shade of a wood edge into an open field, and the organisms that touch the tape, or the point along the tape at which each species first appears, are recorded. A line transect shows how the distribution of species changes along the gradient, and it is quick, but it samples only one narrow line and tells you nothing about abundance in the surrounding area. A belt transect is a line transect with width: quadrats, or a continuous strip, are placed at regular intervals along the tape, so the abundance of each species is recorded at each position. A belt transect therefore shows both the distribution and the abundance of the species along the gradient, and paired with measurements of the abiotic factors, such as light intensity with a light meter, soil pH, soil moisture or salinity, it is how you find out which factors control where a species can live.

Mark-release-recapture is used for animals that move. A sample of the population is caught, each individual is marked harmlessly, for example with a small spot of non-toxic paint, a numbered tag or a small clip of fur, and then released. Time is allowed for the marked animals to mix back into the population, and a second sample is caught; the proportion of marked animals in the second sample estimates the proportion of marked animals in the whole population. The estimate is the Lincoln index: population size = (number in the first sample x number in the second sample) / number of marked individuals in the second sample. If 60 woodlice are marked and released and a second sample of 80 contains 12 marked woodlice, the estimate is (60 x 80) / 12 = 400 woodlice. The estimate is only as good as its assumptions, and quoting them is how you score the evaluation marks: the mark must not be lost and must not change the animal's chance of survival or of being caught again, the marked animals must mix randomly with the rest, the population must be closed so that no births, deaths, immigration or emigration take place between the two samples, and catching must be random. Break an assumption and you can say which way the estimate moves: if the marks make the animals easier for a predator to see, fewer survive and the second sample is too small, but if marked animals become trap-happy and are caught again more often, too many marked animals appear in the second sample and the estimate is too low.

Simpson's index of diversity, and correlations between variables

Simpson's index of diversity, D, puts species richness and relative abundance into a single number. The formula is provided in the examination and you must apply it: D = 1 - (sum of (n / N)2), where n is the number of individuals of each species and N is the total number of individuals of all species. Work through one example. A quadrat in an unimproved meadow contains 10, 8, 6, 4 and 2 individuals of five plant species, so N = 30. Square each count and add: 100 + 64 + 36 + 16 + 4 = 220, and divide by N squared, which is 30 x 30 = 900, to give 220 / 900 = 0.244. Then D = 1 - 0.244 = 0.76. Compare it with a grazed pasture in which the same number of individuals is divided among only three species as 24, 4 and 2: the sum of the squares is 576 + 16 + 4 = 596, so 596 / 900 = 0.662 and D = 0.34. Both samples contained 30 individuals, so the difference in D is a real difference in diversity and not an effect of sample size, and it comes from the same point: the pasture has fewer species and one of them dominates.

Interpret the value and you have the mark for significance. D lies between 0 and 1. A value close to 1 means high diversity: many species are present and the individuals are divided fairly evenly among them. A value close to 0 means low diversity: one or two species dominate the community and most of the individuals belong to them, as in a field of a single crop, a heavily polluted stream or a bare, recently colonised surface. If a calculation gives an answer outside 0 to 1, you have made an arithmetic mistake, most often by forgetting to subtract the sum from 1. Remember also that D describes only the community you sampled: the number, size and placement of the quadrats and the time of year all affect the value, so comparing two areas only means anything if you sampled both in the same way.

The next question is what causes the difference, and that is where the two correlation tests come in. A correlation describes the relationship between two variables measured together, and in this topic the two variables are usually an abiotic factor, such as light intensity, soil moisture, soil pH, temperature or salinity, or a biotic factor, such as the abundance of a competitor or a predator, and the abundance or distribution of a species. The data are collected by sampling along a transect or in a set of quadrats, recording both measurements for each sample point, and then plotting one against the other.

Pearson's linear correlation is used when both variables are continuous measurements that are expected to be related in a straight line and are approximately normally distributed; it is calculated as a coefficient r that runs from -1 through 0 to +1. Spearman's rank correlation is used when the data do not meet those conditions, for example when one variable is a rank or a score, when the relationship is curved, or when the sample is small; the data are ranked and the coefficient rs is calculated from the differences between the ranks of each pair as rs = 1 - (6 x sum of d2) / (n3 - n), where d is the difference between the two ranks in each pair. In both tests a value near +1 means a strong positive correlation, a value near -1 a strong negative correlation and a value near 0 no relationship, and you decide whether the correlation is significant by comparing the calculated value with a table of critical values at the 0.05 probability level, using the number of pairs. The evaluation point is that correlation does not prove causation: sand dune plants and soil pH may be correlated because both change with distance from the sea, and the presence of a species may itself change the factor you measured. Only a controlled investigation, in which the factor is manipulated and everything else is kept constant, can show that one variable causes the other.

Extinction, why biodiversity matters, and the conservation toolbox

A species becomes extinct when the last individual dies, and a population becomes extinct when its death rate exceeds its birth rate for long enough that it disappears. Four causes are named in the syllabus. Climate change alters temperature and rainfall patterns, and with them the distribution of species and the timing of their life cycles; species must move, adapt or die. Corals on the Great Barrier Reef bleach and die when unusually warm water makes them expel the algae that live in their tissues; the caterpillars that great tits feed their chicks to are now peaking earlier in spring, so in some woods the chicks hatch after the food supply has passed its peak; and because the sex of a sea turtle is decided by the temperature of the sand, warmer beaches produce mostly females. Species on mountains and on islands and those in small fragments of habitat have nowhere to move to. Competition removes species when another species uses the same resources more effectively, whether the competitor is a native species whose numbers have been increased by human activity or an introduced one: the introduced grey squirrel (Sciurus carolinensis) outcompetes the red squirrel (Sciurus vulgaris) for food in British woodlands and also carries the squirrelpox virus, which kills red squirrels but hardly affects grey ones. Hunting by humans removes individuals faster than the population can replace them: the dodo and the passenger pigeon were hunted to extinction, elephants and rhinos are still poached for ivory and horn, pangolins are taken for their scales and meat, and many seabirds, turtles and dolphins are killed accidentally as bycatch in fishing gear. Species with slow reproduction, long lives and few young cannot replace such losses. Degradation and loss of habitat is the largest single cause: draining wetlands, felling tropical forest for timber and oil palm (which is the main threat to the orangutan), building roads and towns, and polluting water and soil all remove the place an organism lives, and a habitat can also be degraded by pollution while still physically existing, as when fertiliser washed off fields causes eutrophication of a lake or when pesticides such as DDT thin the eggshells of birds of prey.

These pressures interact, because loss of habitat splits a species into small populations. A small population loses genetic variation through genetic drift and inbreeding, so its individuals become less healthy and less able to adapt, and it can then be wiped out by a single outbreak of disease, a single bad season or a single fire, even when the original threat has stopped. Once a population has passed that point it continues to shrink of its own accord, which is why conservationists act while numbers are still in the thousands rather than the tens.

There are good reasons to maintain biodiversity. The ethical and aesthetic reason is that other species have a right to exist and people value them; there are also the practical reasons. Economic and medical: wild species give us food, timber, fibres and medicines, and many drugs came from wild organisms, including quinine from the bark of Cinchona trees and salicylic acid from willow (Salix), the starting point for aspirin; species not yet studied may still hold treatments for disease, and wildlife, as in the gorilla-watching and safari industries, brings income to local people who then have a reason to protect it. Ecosystem services: natural systems keep the planet habitable, by storing carbon in forests and peat, cycling nutrients and building soil, pollinating crops by wild insects, controlling pests naturally, purifying water, controlling floods and preventing erosion, as mangroves and coral reefs protect coasts from storms. Genetic resources: the wild relatives of crops carry genes for resistance to disease, pests and drought that plant breeders use when the varieties in farmers' fields lose their resistance. Stability: an ecosystem with more species usually has more alternative pathways for energy and nutrients, so it is more resilient and recovers better from drought, fire or disease.

Conservation works at two levels. In situ conservation protects the species in its own habitat, and it protects the whole community, the interactions between species and the potential for the population to keep on evolving. The main in situ method is the conserved area: national parks and marine parks, nature reserves and protected areas, where the habitat is protected by law and managed by wardens. Examples include the Serengeti National Park and the Ngorongoro Conservation Area in Tanzania, the Great Barrier Reef Marine Park off Australia, and the National Nature Reserves and Wicken Fen in the United Kingdom. Management includes preventing poaching, removing invasive species, controlling grazing or burning to keep the habitat open, giving local people an income from ecotourism so that the reserve is worth more to them than the land would be if it were converted, and providing corridors so that animals can move between fragments. Ex situ conservation keeps the species outside its habitat, as a back-up. Zoos run international captive breeding programmes with studbooks that record the ancestry of every animal, so that pairs can be chosen to keep the captive population genetically diverse, and they fund and carry out fieldwork; the Arabian oryx is the classic success, bred in captivity in several zoos and then reintroduced to the wild in Oman and Jordan, and Przewalski's horse and the Mauritius kestrel have been through the same process. Botanic gardens grow and propagate rare plants, keep collections of living material and seed, and reintroduce species; the Wollemi pine (Wollemia nobilis), discovered in Australia in 1994 as a few trees in a canyon, was propagated by botanic gardens around the world. A frozen zoo stores frozen cells, gametes and DNA in liquid nitrogen at about -196 °C, so genetic material is preserved permanently and can be used decades later for artificial insemination or in vitro fertilisation; the Frozen Zoo at San Diego Zoo Wildlife Alliance and the Frozen Ark in the United Kingdom are examples. A seed bank dries and stores the seeds of many thousands of wild species in a small space, at low temperature, testing samples periodically for viability: the Millennium Seed Bank at Wakehurst, part of the Royal Botanic Gardens at Kew, and the Svalbard Global Seed Vault in Norway, which stores crop seeds inside an Arctic mountain. Seed banks are cheap for the number of species conserved, but not all seeds survive drying, because so-called recalcitrant seeds such as oak and avocado must be stored as living plants or as tissue cultures, and a seed in a freezer does not pollinate flowers, disperse fruit or feed a food chain. Ex situ collections are too small to hold a species' full genetic variation and their animals can adapt to captivity, so they are a back-up for in situ work, not a substitute for it.

Assisted reproduction, invasive alien species, and international action

Endangered mammals often fail to breed in captivity: the pair may not be compatible, a female may be too old to carry a pregnancy safely, or there may be too few animals left to make good pairs. Assisted reproduction gets around each of these problems. In in vitro fertilisation (IVF) eggs are collected from a female, often after hormone treatment to make her release more than one egg at a time, sperm are collected from a male, and fertilisation takes place outside the body in a laboratory dish of culture medium; the resulting embryos can be grown for a few days and, crucially, can be produced from gametes that were collected earlier and stored frozen in a frozen zoo, so a dead animal's genes can still be used. Embryo transfer then puts those embryos into the uterus of a female, who carries the pregnancy and gives birth. Surrogacy is the same idea when the female that carries the pregnancy is not the genetic mother: the embryo is produced from the cells of other parents, often of a different but closely related species, and the surrogate's own genes are not present in the offspring. Domestic horses have carried Przewalski's horse embryos, and domestic cats have carried embryos of their endangered wild relatives, such as the African wildcat. The value of embryo transfer and surrogacy is that a single valuable female can contribute many offspring in one season instead of one, and that embryos from several sets of parents can be carried at the same time, which increases the size and the genetic diversity of a captive population quickly. Frozen gametes and embryos also mean that offspring can be produced years later, and frozen cells have been used to produce a cloned black-footed ferret, adding genetic variation to that species. The limitations are real and examinable: success rates per attempt are low, the reproductive cycles of donor and surrogate must be synchronised with hormones, there is a risk of transmitting disease, and no amount of laboratory work can help a species whose habitat has gone, which is why assisted reproduction is always used with habitat protection and with reintroduction.

An alien species is one that has been moved outside its natural range by human activity. It becomes invasive if it spreads and causes harm, and controlling it is a conservation measure because the damage is done to native species and to whole ecosystems. Reasons to control invasive species: it usually arrives without the predators, parasites and competitors that kept it in check at home, so its population grows unchecked; it outcompetes native species for food, light, space or nest sites, and by competitive exclusion can drive them out; it may be a predator of native species that have no defences against it, as the brown tree snake (Boiga irregularis) has wiped out most of the forest birds of Guam and the cane toad (Rhinella marina), introduced to Australia in 1935 to control the cane beetle, eats native insects and small vertebrates while its skin poison kills the quolls, goannas and snakes that try to eat it; it may carry a new disease or parasite to which native species have no resistance, as the grey squirrel carries squirrelpox virus; it may change the habitat itself, as Japanese knotweed (Fallopia japonica) and Rhododendron ponticum form dense stands that shade out native ground flora, and the zebra mussel (Dreissena polymorpha) filters plankton from lakes and fouls every hard surface; it may hybridise with a native species and swamp its gene pool, which is why the ruddy duck was controlled in the United Kingdom to protect the white-headed duck in Spain; and control and lost production cost a great deal of money, both in agriculture and in repairs to buildings and river banks. Methods of control are physical removal or barriers, chemical control, biological control by introducing a natural enemy (which carries its own risk, since the cane toad was itself introduced as a biological control), and legislation, backed by inspections and public reporting.

Two international bodies coordinate much of this work. The International Union for Conservation of Nature (IUCN) is an international organisation, based in Switzerland, that brings together governments, scientists and conservation groups. Its Species Survival Commission assesses the status of species against criteria based on population size, rate of decline and the size of the range, and publishes the Red List of Threatened Species, in which each species is placed in a category that runs from Least Concern through Near Threatened, Vulnerable, Endangered and Critically Endangered to Extinct in the Wild and Extinct, with a Data Deficient category for species too little known to judge. The Red List matters because it is the shared, evidence-based list that governments, charities and researchers use to decide which species need help most and where money and protected areas should go; the mountain gorilla, for example, was moved from Critically Endangered to Endangered as its numbers rose. The IUCN also advises on the management of protected areas and runs conservation programmes with its members.

The Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) is an international agreement, in force since 1975, to which more than 180 countries are parties. It controls international trade in listed species rather than banning all trade, and species are placed in one of three appendices. Appendix I lists species threatened with extinction, and commercial trade in them is banned: tigers, rhinos and gorillas are examples. Appendix II lists species that are not yet threatened with extinction but would be if trade were not controlled, so trade is allowed only with permits and is monitored to keep it sustainable: many orchids, mahogany and seahorses. Appendix III lists species protected in one country that has asked the other parties to help control its trade. Each member country must pass its own laws to enforce the agreement and appoint a Management Authority, which issues the export and import permits, and a Scientific Authority, which advises whether the trade would harm the species. The weakness of CITES is that it applies to international trade only: it does nothing about habitat loss or about hunting for local use, and it depends on enforcement, which is why illegal trade in ivory, pangolin scales and rosewood continues. The two organisations work together: IUCN assessments show which species are in trouble and why, CITES listings then control the trade that is one of the pressures, and protected areas, ex situ breeding programmes and control of invasive species deal with the others.

• IVF: eggs and sperm are collected and fertilised outside the body, so gametes stored frozen in a frozen zoo can still produce offspring.
• Embryo transfer: laboratory-made embryos are put into the uterus of a female, who carries the pregnancy.
• Surrogacy: a female of the same species or a closely related one carries an embryo that is not hers, so one valuable female can produce many offspring in a season.
• IUCN: assesses species and publishes the Red List, categorising them from Least Concern to Extinct, so that priorities can be set.
• CITES: regulates international trade in listed species through three appendices, with permits issued by a Management Authority and advice from a Scientific Authority.
• Limitations: assisted reproduction has low success rates, CITES covers international trade only, and neither can replace protecting the habitat.