Parthenogenesis is a type of asexual reproduction and applies only to animals (the botanical counterpart is called apomixis; note that in this post, I will only be talking of animals, not plants, fungi, prokaryotes, bacteria, etc.). It was first discovered by Owen in his 1849 book/monograph “On parthenogenesis or the successive production of procreating animals from a single ovum”. As an aside, I want to go a bit into Owen’s work and legacy, and how parthenogenesis played into them. He is often viewed as being a creationist – this is false. He was a critic of natural selection, but he has always thought in terms of evolution. His observations on the cyclic parthenogenesis of aphids (see later) led him to believe that new species can arise within a few generations and not the eons of time that Darwin stressed. His hypotheses did not depend on external factors, such as natural selection, but on the potential inside the organism (developmental possibilities). This is, of course, wrong, and the fact that he infused his science with his own pantheistic religious beliefs did not help his legacy and has led to the view of him being a creationist. Just to make it clear, he did believe in evolution and a natural origin of species, but not in the way Darwin had expressed it.
Parthenogenesis literally means a virgin birth; in biological terms, it simply means that an unfertilised ovum will give rise to a fully-functioning adult. That’s the broadest definition, and the one we will be using in this post; keep in mind that in the narrowest sense, parthenogenesis implies unisexual populations of virgin females producing clones (as seen in some lizards). A species can be obligately parthenogenic (can’t reproduce sexually at all) or facultatively parthenogenic (has the ability to switch between sexual and parthenogenic reproduction). In all, there are about 2000 species recorded to reproduce at least facultatively parthenogenetically under natural conditions and many more definitely exist, but remain unobserved/unsampled.
Of the vertebrates, parthenogenesis has never been reported for mammals; bird eggs reported as being of parthenogenic origins never hatch. It’s only found in several unique fish, amphibians and lizards. In those, it is present due to polyploidy and hybridisation. At the cellular level, the difference between parthenogens and sexuals is that in the latter, meiosis is followed by fusion of a male and female gamete. In parthenogenesis, meiosis is changed so that only one particular set of chromosomes is transferred in a non-random fashion. Before going on to look at examples from the invertebrates, let’s examine the cytological underpinnings of parthenogenesis in vertebrates.
The root problem of parthenogenesis in vertebrates is that haploid eggs are useless. The egg cell must either remain diploid (after undergoing meiosis!) or somehow be able to become diploid again by itself. This is achieved in several ways. The simplest way is to simply stop meiosis from happening. This is called apomixis, and the oocyte is then produced just by mitosis, skipping over recombination and segregation of chromosomes, leading to clones. Another way is to double the number of chromosomes before meiosis happens – so when the ploidy halves, the result is still a diploid oocyte. Again, clones are produced here because recombination and segregation take place between identical homologous chromosomes, not strange ones. The most common way to maintain diploid levels though is automixis. This is when meiosis occurs, and the resultant haploid oocyte either duplicates its genome (becoming diploid) or fusion with one of the polar bodies occurs. This way obviously does not produce clones.
Let’s look at mammals specifically. The non-existence of parthenogenic mammals can easily be explained. Simply said, we need a set of chromosomes from the male sperm in order for development to proceed. This can be proven by experiments done in rabbits. One thing that a germ cell doesn’t have is centrioles (they’re lost during gametogenesis). In the rabbit egg, centrioles are built anew – this requires no external input, and is one of the steps that kicks off embryonic cell divisions. Theoretically then, the lack of centrioles is not a limit to the egg cell and it should proceed alone with its development (as it would do in a hymenopteran egg that will become a male) – but it doesn’t. There are no parthenogenic rabbits, not even experimentally induced, because the chromosome from the sperm (and the genomic imprinting, i.e. that certain genes will not switch on unless paternal genes are present) is missing. This is an issue with mammals and not with the parthenogenic fish because mammals have an overabundance of a protein called mos. When absent, automixis occurs (it usually blocks cleavage formation in the oocyte, therefore when present, the oocyte can’t go through automixis).
Just to make one thing clear: we can artificially induce parthenogenesis in mice by activating the egg, but the embryo will die and not grow to adulthood. Only the first few divisions will be achieved, until the lack of imprinting leads to abortion – experiments have shown that histones, the structures around which DNA is coiled, are lost, as the instructions on how to rebuild them are present in the sperm. Having said that, there is one case where a mouse adult has been developed parthenogenetically in the lab. What they did was basically simulate imprinting in the egg cell by simply transferring the relevant nuclear genes. I say simply, but this is an extremely difficult protocol and cannot yet be applied to anything other than lab mice. What it does show is that imprinting is absolutely crucial to mammalian embryonic development – and that this can only occur with the presence of a male nucleus. The case is pretty much closed: natural parthenogenesis in mammals is impossible.
The best example of parthenogenesis is in the hymenopterans: any ant or bee or wasp male got there by parthenogenesis, since hymenopterans are haplodiploid (males come from unfertilised eggs, females from fertilised eggs). This ‘standard’ parthenogenesis is called arrhenotoky and is also found in all thrips (one of the reasons why they are such big pests: they can settle very quickly in a new environment, even if no males are around; this applies to all invasive species as well), and termite queens do it when founding a new colony. There are even more extreme types: worker ants of a couple of ant species can lay eggs which give rise to other females without fertilisation – parthenogenesis. Specifically, haplodiploid parthenogenesis is called thelytoky; a thelytokous population has no males. Although rare in ants, it did arise multiple times independently in other hymenopterans (for example in some honeybees, where workers will undergo thelytoky in the absence of a queen).
The reason why thelytoky is interesting is because of sex determination. For a hymenopteran to be female, the ovum must become diploid – hence why they are normally fertilised. The easiest way to explain thelytoky in hymenopterans (and keep in mind this is still a hypothesis in need of concrete proof) is to say that during or after meiosis (during oogenesis), two haploid nuclei fuse together, forming a diploid egg without need for sperm.
There is another form of parthenogenesis in hymenopterans, found only in some species of chalcidoids: paternal genome elimination (PGE), in which males develop from fertilised eggs, but lose the paternal genome at some point during development, becoming haploid again.
Of the parthenogenic animals, most are facultatively so: they can either undergo sexual or asexual reproduction as they please. The factors that influence the decision are numerous. One pattern that is very clear in these organisms is that parthenogenesis has a geographic distribution; this was first realised by Vandel in 1928 (“La parthénogénèse géographique: contribution a l’étude biologique et cytologique de la parthénogénèse naturelle”). The parthenogenic populations of a species tend to occur at high altitudes, high latitudes and on the geographical edges of a species’ range. This is easily explainable: parthenogenesis, as a form of asexual reproduction, doesn’t have the two-fold cost of producing males, therefore it allows for more rapid settling in new and/or extreme habitats (where extremeness is measured relative to that species’s specific attributes); natural selection takes care of keeping the gene pool successful. This seems simple now, but it took some time to realise it: White, in his 1973 book “Animal Cytology and Evolution”, had this to say about geographical parthenogenesis: “It is not entirely clear, however, how forms whose genetic system must be very inflexible manage to become adapted to new environments when they do get transported to them: the apparent ecological versatility in space seems to be at variance with their lack of ecological versatility in time.” Sexual reproduction is more ideal for environments where there are plenty of biotic interactions – there are virtually no parthenogens in the tropics, for example – because the genetic variation is needed to stay ahead of the curve (think of the Red Queen). In the Arctic, many species are parthenogenic, simply because mating requires too much energy.
To look at this geographical aspect of parthenogenesis, we’ll use phasmids (stick insects). Parthenogenesis is widespread among them (it evolved independently in each lineage, not an ancestral state; for example, there are 21 species of Timema, and parthenogenesis evolved independently five times in the genus), with some being obligately parthenogenic. Parthenogenesis in them is related to hybridisation and polypoidy: some stick insects can have up to 100 chromosomes! This sexual freedom is one of the reasons for their success: a single female phasmid egg can potentially give rise to a new successful population in a new environment, even if it is far away. For example, there are is an all-female population of phasmids on the Scilly Isles, UK. It is not endemic – it was brought there by humans. From New Zealand. And they are now extremely successful – the fact that they’re all female means they are also parthenogenic. In New Zealand itself, one phasmid species is descended from a tiny population. After the last glaciations, parthenogenesis allowed this population to expand its range, leading to the situation now: it dominates New Zealand and even has a colony on a nearby island. Staying in Australia, parthenogenesis is the reason why there are many species inhabiting the Australian continent, despite it being mostly an arid wasteland – the aridisation occurred fairly recently in the Pleistocene. The unique thing about the place is that there have been large changes in climatic conditions, and this is reflected by the parthenogenic populations there, many of which went extinct, only to be replaced by other parthenogens. What we see nowadays is really just a snapshot of a very dynamically changing biome.
Other facultative parthenogens include aphids. For example, the pea aphid undergoes sexual reproduction only once in autumn and parthenogenesis from early spring to the autumn. When such a clear temporal pattern is present, we refer to the species as a cyclical parthenogen (another name is heterogony). In the case of the pea aphid, the environmental conditions (shorter days and colder temperatures) induces the development of sexuparae – females that give birth to males and females. The presence of males is the odd thing here, and they’re formed by an extra mini-meiosis of the egg cell. Aphid sex is XX/X0 (female/male) – this extra meiosis basically produces an extra egg that is missing one chromosome (the second X fails to attach to the spindle and is lost), and thus will be male. The sexual forms will then mate and produce eggs that are tolerant to cold, as to be able to survive in winter. The first generation animals are called fundatrices, and will then only reproduce parthenogenetically (the aphids from early spring to autumn are therefore all clones), producing between 3 to 15 generations until the next autumn. Note these different types of aphid are morphologically variable, even though they belong to the same species (one of the reasons why Owen refused the long time scale required by natural selection: he thought that every year, a new species of aphid was born, so to speak). Another interesting note about aphids is their so-called ability of ‘telescoping generations‘, where the parthenogenic embryos in the female will already have their own offspring developing inside them (like those Russian dolls). This is another reason why aphids are so widespread and successful: even though they are clones, the successive generations are already exposed to the conditions that await them. This cyclical parthenogenesis in aphids is ancient: aphids are ~200 Ma and Cretaceous fossils already have somewhat reduced ovipositors, implying that sex already had a smaller role even back then.
Other cyclical parthenogens include cladocerans and monogonont rotifers. At the cellular level, they incredible control over meiosis – this can be seen directly from the genome, where there are duplications in the genes involved in meiosis and the cell cycle. This genomic repertoire is needed to have this kind of reproductive variability.
A small jump into theory is suitable here. These aphids, and all other facultative parthenogens, exhibit occasional sex. Theoretically, it is assumed that the two-fold cost of sex is dependent on the frequency that sex happens: obligately sexual species reproduce only sexually and thus must pay the full cost. However, those who rarely have sex get the best of both worlds: when needed, they can double their population number quickly, and then get the benefit of genetic variation by having sex once. This was first realised (theoretically) by Sewall Wright in a 1939 paper called “Statistical genetics in relation to evolution”.
Another trend obvious in cyclic parthenogens is that they switch to sexual reproduction when resources become limited. For example, water fleas will reproduce at fast rates parthenogenetically until their food source becomes depleted, at which point they mate sexually, producing varied offspring that can colonise other areas. The same is true for the aphids.
There are cases where parthenogenesis occurs, but also needs sperm to proceed. For example, the planarian Schmidtea polychroa. Planarians are hermaphrodites that undergo internal fertilisation (on a related note, sexual reproduction is hilarious in them: they fight and whoever loses must get pregnant) – but no self-fertilisation. In S. polychroa, whether an individual is sexual or not depends on the ploidy level: diploids are sexual, polyploids (tri- to pentaploid) are parthenogenic. Regardless of their sexual type though, they mate and exchange sperm. Without the sperm, there will be no initial zygotic divisions and the embryo will not develop. The reason why we call it parthenogenic is because in terms of genetic contribution, there is nothing from the father: paternal alleles always disintegrate or are thrown out. This type of parthenogenesis is variably called sperm-dependent parthenogenesis (biologists are not a creative bunch), pseudogamy or gynogenesis. And it’s worth mentioning that the occurrence of sexual and parthenogenic forms is also geographically dependent: S. polychroa is a European species, and south of the Alps are exclusively sexual populations, whereas parthenogens occur everywhere else.
Another variant of parthenogenesis is hybridogenesis, which occurs in some fish and amphibians. Here, haploid oocytes are produced and fertilised by the sperm. The male contribution persists for one generation only, since the next batch of eggs will delete the paternal chromosomes during meiosis, resulting in the exclusive retention of the maternal genome. However, what must be kept in mind is that accidents can always happen and some male genes can always leak through. The observant among you will realise that there is a problem with these reproduction types: a clonal population consisting solely of females will eventually go extinct, since there are no males available to provide sperm. In fact, these species are not fertilised by males of the same species, but of a different, closely-related sexual species with which they coexist. The parthenogenic females compete for those males as normal.
So far, all our examples have been of parthenogenesis that evolved specifically in a particular lineage. There are cases when a lineage may be forced to reproduce parthenogenetically by parasitism. These parasites are intracellular bacteria living in the reproductive tissues of their hosts (invertebrates, especially arthropods). They not only cause thelytokous parthenogenesis, but can also lead to feminisation of males and may even kill male eggs. The reason is that they are transferred maternally through the oocyte’s cytoplasm. Males are useless to them, so in order to survive, they need to make sure that the population is heavily female-biased. The most well-known of these parasites is Wolbachia, a type of Gram-negative Rickettsiales bacterium that infects hymenopterans and other haplodiploids; it’s estimated that 16% of all insect species are infected with Wolbachia. It has become so deeply entrenched in its hosts that in some cases, the host is unable to reproduce if it is removed. Cytologically, it has several ways of working. It can prevent haploidy by making sure mitosis fails to separate chromosomes during the first anaphase. Alternatively, mitosis proceeds, but in the second prophase, two mitotic nuclei are fused together, bringing diploidy back. Other examples include Cardinium and several Flavobacteria in arthropods, or Xiphinematobacter in Xiphinema nematodes. Keep in mind that these parasites are generally benign, not virulent and do not affect the physiology of their arthropod hosts.
In addition to parasitic parthenogenesis, toxicity in the environment can also make animals reproduce parthenogenetically – although it must be said that this has never been conclusively observed in the wild (i.e. discounting other factors that may have caused parthenogenesis), but has been done many times in the lab, as far back as Jacques Loeb’s experiments on sea urchins in the late 19th Century. He found that toxins like UV light, ammonia, chlorine, acids, alkalis and alcohols can cause sea urchin eggs to start developing, with no need for sperm. The way this works is by messing with the egg membrane, reducing its surface tension. This kicks off a chain reaction (usually started by sperm) and causes embryonic growth.
Parthenogenesis has occasionally been cited as being proof of Dollo’s Law – the concept of the irreversability of evolution (introduced by Dollo in a 1893 essay, “Les lois de l’évolution”). Before we can decide that though, we need to look at how parthenogenesis actually evolved: the ancestors of parthenogens are always sexual species. What does the transition look like (besides the already-mentioned infection-induced parthenogenesis)? The most elegant transitional type is tychoparthenogenesis, where unfertilised eggs will accidentally produce an adult. Accidentally, because this happens due to errors in meiosis resulting in diploid instead of haploid oocytes (so they count as being automatically fertilised, basically). Where females are the homogametic species (e.g. XX, not XY), this will result in a female adult. The only problem with this hypothesis is that even though we can artificially increase the success rate of tychoparthenogenesis, in nature the success rate is extremely low – so how would tychoparthenogenic populations survive against much more successful sexuals? The only reasonable solution is to say that many females will fail to find a mate and thus will only reproduce tychoparthenogenetically, eventually leading to female-only populations.
One clear pattern that can be highlighted in this context is the positive correlation between flightlessness in insects and parthenogenesis. Parthenogenesis is rare in beetles, except in the weevils. Same with Lepidoptera and bagworm moths; it is almost non-existent in the odonates, the best fliers of the animal world; stick insects, on the other hand, are flightless and many of them are parthenogenic. A possible explanation for these observations is that in the less-mobile taxa, there will be more failure than success in mating – females and males will not be able to reach other – and this then plays into the tychoparthenogenesis model outlined above.
If you try to make a list of parthenogenic animals, what you will notice is that parthenogens occur most often in single species (for example, only one species of decapod crustacean, the marbled crayfish, is parthenogenic), not at high taxonomic levels, hinting that parthenogenesis is an evolutionarily short-lived strategy. The exceptions to this include the bdelloid rotifers, darwinulid ostracods or oribatid mites, where parthenogenesis has been maintained for geologically significant amounts of time (e.g. the oribatid mites: ~100 Ma!). Other examples of ancient asexuals can be found within the beetles, aphids, walkingsticks (e.g. in Timema, the sister group to the rest of the phasmids, some asexual lineages have been dated at ~2 Ma) and brine shrimp (e.g. Artemia).
One possible consequence of widespread parthenogenesis is vestigialisation of traits that are sexually selected for – things like mating songs or display structures. For example, parthenogenic crickets cannot recognise the mating call of their sexual ancestor. One must be careful though: something like hearing is crucial to an insect. The parthenogenic cricket may not be as highly tuned as its sexual counterparts (e.g. they have less auditory receptor cells), but it still has functioning hearing organs because it has to watch out for predators (in this case, ultrasound emitting bats).
Nonetheless, vestigialisation of sexual characters is constantly threatening for parthenogens because of Muller’s Ratchet. In 1949, Muller proposed that if selection pressure on a trait is removed, random mutations will degrade the trait down until positive selection can be established again. A perfect example of this principle, applied to parthenogenesis and sexual characters, can be seen in mating behaviours, which often don’t have any use other than sex. In fruit flies, it has a genetic basis (based on many genes). The good thing about fruit flies is that they are common lab animals and thus always cultured. Some are facultatively parthenogenic, and this parthenogenesis can be artifically selected for. In 1961, such a strain was set up specifically for this purpose. In 1971, the females of this strain were put with males. Some mated, many didn’t – but they still recognised the behaviour. In 1981, they could not mate at all. Genetic tests showed that this was because there had been too many mutations affecting those genes – sexual behaviour had essentially become vestigialised (and this was the first real proof of Muller’s Ratchet).
Just to summarise, here’s a list of constraints that prevent the evolution of parthenogenesis. The largest problem, known from the vertebrates, is how to activate the egg – it’s usually frozen in the metaphase II part of the cell cycle and needs sperm to activate it. The fact that the centriole is not inherited and must be built anew could also be a limiting factor. The sex determination system is important, as parthenogenesis requires a heavily female-biased sex ratio. In mammals, genomic imprinting by the sperm is critical to starting fetal development.
Parthenogenesis occurs in less-mobile taxa, the relation being that these have trouble finding mates, and parthenogenic taxa are often derived by hybridisation and polyploidy. It is favoured by taxa living in borderline habitats, where they can use the energy that would have been allocated for mating to more useful tasks, as well as allowing a quick population increase – this is also the reason why invasive insect species and pests are parthenogenic.
For a thought experiment, think of the implications of parthenogenesis on species concepts (are parthenogens really new species, for example? How can we detect them?), keeping in mind hybrid and suture zones. As a starting point, you could consider this fact: in ancient lakes, where there are many endemics, none of the endemics occur in families where parthenogenesis is common.
As a small addendum, not really important but that does involve parthenogenesis, I want to mention the most complex life cycle known in the insects: that of Micromalthus, a very rare archostematan beetle. They live and feed on decaying wood. The larva-like female gives birth parthenogenetically to live female young. These are active, with long legs and hair. They then metamorphose to the second instar, which is legless, and this type persists until pupation. Under some conditions, the females will parthenogenetically lay a single large egg, though. Out of this egg comes a thick, C-shaped legless larva. This is a haploid male and once it grows to adulthood, it will eat its mother. Even more intriguingly, a female might sometimes produce both the egg and the live young at the same time. The evolution of the life cycle is hypothesised as having something to do with endosymbiotic bacteria: the beetles need them to digest the wood, and they, like Wolbachia, may screw around with the reproductive system.
As an other addendum, I want to bring up the marbled crayfish. It is a parthenogenic decapod crustacean that is now gaining much popularity in epigenetics and especially cancer research: it can live on a single food pellet for 3 years. It has a generation time of 6 months and produced 400 eggs per batch. These eggs are, of course, genetically identical to their mother. It can be brought up in a single microplate, so no housing conditions are needed. Each egg can be grown separately. It is thus perfect for precise experimentation, and the genetical identicness due to parthenogenesis also allows for extremely ideal experimental procedures.
Lost Sex, a 2009 multi-author volume edited by Schön, Martens and van Dijk (Publisher: Springer) is exclusively dedicated to parthenogenesis. While I did not have the time to read all of it, the one chapter I consulted was excellent and if it is indicative of the rest of the book, then I fully recommend it!