This is a thought experiment:
If we form a population with only a single founder pair, can this population survive? What would happen? Would this inbreeding cause the population to go extinct? Could such a population continue to exist indefinitely?
Populations founded by a single pair are quite common in the lab
The Drosophila Genetic Reference Panel, one such example of high artificial inbreeding, is a series of inbred lines where the researchers mated brother and sister flies for 20 generations to form the lines, which means they will be highly inbred. When you do these though you often lose many of the initial lines because of inbreeding depression. Inbreeding depression is reduced fitness as a result of increased homozygosity in the lines - deleterious mutations are more frequently expressed. I have tried to find the number of initial pairs they used to make the ~200 DGRP lines but come up unsuccessful, but I've heard it was around 1500 pairs, although that may be well off the mark*.
What would happen to your mice?
Just like the DGRP lines, there is a high risk that the population would collapse. However, if they do get through a few generations of inbreeding, and are kept in steady conditions (i.e. selection doesn't change once the lines are formed), then they have an ok chance of persisting. If you started with multiple founder pairs then there's a better chance you will still have some (very inbred) mice in a few generations. The success rate will depend on the frequency of deleterious mutations and strength of their effects.
What is inbreeding?
Inbreeding is increased homozygosity in the population than would be expected under random mating. For example, if the frequency of two alleles ($A_1$ and $A_2$) at a gene are $p = 0.1$ and $q = 0.9$ respectively, the expected frequency of $A_1A_1$ homozygotes is
$p^2 = 0.1^2 = 0.01$
the expected frequency of $A_2A_2$ homozygotes is
$q^2 = 0.9^2 = 0.81$
and the expected frequency of heterozygotes is
$A_1A_2$ is $2pq = 2 imes 0.1 imes 0.9 = 0.18$
If the true frequency of heterozygotes ($A_1A_2$) is less than 0.18, then there is inbreeding.
Is there a benefit to inbreeding?
Well inbreeding can be a good thing. Imagine all of those lines that did not collapse. It is possible that a population formed out of those would have higher fitness than the original population, because deleterious mutations may have been purged (lines carrying deleterious mutations are more likely to collapse). So a highly inbred population can prosper. However, that population would also have lower genetic variance. Genetic variance places limits on the potential for adaptation, low genetic variance means low potential to respond to selection. So, if selection is not constant, then inbred populations are more likely to go extinct as they are unable to adapt to novel scenarios.
Inbreeding in nature
One classical example of high inbreeding in nature is in cheetahs, where it is beleived they went through a strong genetic bottleneck ~10,000 years ago. The problems faced by cheetahs today really highlights the problems of high inbreeding because they are struggling to adapt. Genetic bottlenecking is a serious long-term problem for many species including the Javan rhino, the Kakapo, Mountain Gorillas, and others; even if we can stop poaching and habitat loss wiping out these species in the short-term, they will struggle to adapt to changes in the climate etc…
"As a species, cheetahs have famously low levels of genetic variation. This can probably be attributed to a population bottleneck they experienced around 10,000 years ago, barely avoiding extinction at the end of the last ice age. However, the situation has worsened in modern times. Habitat encroachment and poaching have further reduce cheetah numbers, consequently snuffing out even more genetic variation and leaving cheetahs even more vulnerable to extinction."
* Supporting material from the DGRP paper would suggest that it was indeed formed from 1500 isofemale lines.
There have been several studies with similar structure and background as your experiment, such as "Inbreeding Effects on Reproductive Outcome: A Study Based on a Large Sample from the Endogamous Vadde of Kolleru Lake, Andhra Pradesh, India" from 1992.
It was an experiment to investigate this hypothesis:
Sanghvi's hypothesis on long term effects of inbreeding was tested in Kotas. Kota is a numerically small tribal population in the Nilgiri district, Tamil Nadu State, India. Consanguineous marriages are common in this tribe. A total of 95 couples were taken for this study and necessary data were collected on a set proforma. Of the 95 couples, 28 (29.5%) were consanguineously related. The inbreeding coefficient for autosomal genes is 0.022 and for sex-linked genes is 0.03. Inbreeding effects on reproductive losses were examined through an exponential regression model. Although the regression coefficient B values are positive, they are insignificant, suggesting no consistent relationship between degree of consanguinity and the reproductive losses. The estimates of genetic load is 1.8 lethal equivalents per gamete and the average B/A ratio is 5. These findings empirically support the Sanghvi's contention.
The abstract of the experiment from 1992:
Inbreeding effects on reproductive outcome vis-à-vis fertility, prenatal loss and prereproductive mortality, and secondary sex ratio of live and dead children were examined in a large sample of 2078 women of the Vadde fishing population of Kolleru Lake in Andhra Pradesh, India. Demographically, this population is a single endogamous unit. By using an exponential regression model with the proportion of offspring survival as a dependent variable and the inbreeding coefficient as an independent variable, I further examined the inbreeding effects. The results were compared with results from other fishing groups and other southern Indian and non-Indian populations. The results among the Vadde were consistent with those found for other groups of Telugu-speaking fishermen and several other southern Indian populations in that the effects were neither perceptible nor significant. The average B value and the number of lethal equivalents found for the highly inbred southern Indian populations in general and for the Vadde in particular were much smaller than those from other parts of the world, providing empirical support to Sanghvi's hypothesis on long-term effects of inbreeding.
The hypothesis was also supported by another investigation, looking at The effect of inbreeding on mortality and morbidity among Telugu-speaking populations of Kharagpur, West Bengal, India.
Increased mortality and morbidity including congenital malformations among the offspring of consanguineous marriages have been widely reported in human populations from different parts of the world. However, there are few studies on the effect of the intensity of inbreeding and different degrees of inbreeding on mortality and morbidity. The present study is an attempt to examine the effects of inbreeding on mortality and morbidity including congenital disorders in different levels of inbreeding among Telugu-speaking populations of Kharagpur, West Bengal, India, based on data collected through extensive pedigrees. The study reveals that the frequency of spontaneous abortions and stillbirths is higher in the offspring of consanguineous marriages than in that of non-consanguineous marriages. A similar effect is also observed in the infant mortality rate, which is known to have a genetic component, but is not seen in the mortality rate of children and juveniles. The rate of morbidity is consistently higher in the offspring of consanguineous marriages with a sex bias in favour of inbred females. The increased morbidity rates in inbred individuals tend to be inversely correlated with the increase in average autosomal inbreeding coefficient. This appears to strengthen Sanghvi's hypothesis of a decline in the frequency of deleterious genes with intensification of inbreeding through generations. The present study also confirms an increase in genetic disorders with an increase in inbreeding in almost all populations.
Sources & Further Reading:
The Naked Mole Rat presents a oddity in many of its reproductive behaviors when compared to other African Mole Rats. Heterocephalus glaber presents several behaviors that are directly in contrast to many of its relatives. For one H. glaber is a eusocial mammal and splits its reproductive responsibilties up through division of labour. In a colony there is generally only one breeding female and 2-3 breeding males. 1 The breeding female is morphologicaly different from the rest of the females in the colony and often displays agressive shoving behaviors towards the other females in the colony. 2
The breeding female of a colony of H. glaber also produces many more young than related species. Related species such as the Damarand Mole Rat ( C. damarensis ) produce an average of 2-4 pups per litter, with a maximum of 6. Colonies of H. glaber produce an average of 12 pups per litter and have been observed to produce up to 27 in a single pregnancy. 3 This is linked to the fact that the H. glaber colony sizes are much higher than those of the other African mole rats. The species with the next largest size of mole rat is C. damarensis which has a mean colony size of 11 individuals per colony. The largest C. damarensis colony discovered had 41 individuals The mean colony size of H. glaber colonies is 80 individuals, and the largest colony discovered contained 271 rats. 4
H. glaber also inbreeds to a high degree, much higher than most other mammals. The whole colony is constructed of one single family, and breeding is almost always between brothers and sisters. 5 Indeed, the average relatedness between the offspring of the Breeding female and the rest of the colony is often as high as 0.5, a value that in most other species is reserved for parent-offspring relationships. Meanwhile the relatedness between a parent and offspring is often as high as 0.81. The maximum relatedness is 1.0, a score given to monozygotic twins. 6
Althought H. glaber puts the burden of creating new young for the colony upon only one female, care of the young is spread out among many of the different members of the colony. Studies have shown that from just before birth to till four weeks after huddling of all members of the colony increases, and young often beg for feeces from all members of the colony. Furthermore, nonbreeding males are often observed nudging and grooming the pups.
But perhaps the most interesting part divergence of H. glaber from its fellow African mole rats is the way in which queens sexually suppress the other mole rats in the colony. Of all the mole rats in the colony, only the breeding females and males are able to breed. Non breeding females have prepubescent ovaries and males do not create gametes in quantities anywhere near what the breeding males do. 7 The Damarand mole rat ( C. damarensis ) also has a single breeding male and female in the colony however the non breeding males and females seem to be fully capable of breeding, and seem meerly to'choose' not to, rather than being, like in H. glaber unable to. 8
Why male and female cells behave differently after being reprogrammed into stem cells
Vincent Pasque from KU Leuven, Belgium, and Kathrin Plath from UCLA led an international study into how specialized cells are reprogrammed to induced pluripotent stem cells (iPS). The researchers discovered that female and male cells behave differently after the reprogramming process and that this is due to their different number of X chromosomes.
Thanks to a promising technique known as cell reprogramming, a patient's skin cells could be used, for example, to restore vision in the future. The technique allows scientists to make stem cells from a patient's own cells. These stem cells—known as induced pluripotent stem cells, or iPS cells—can become almost any specialised cell type of the human body, including the cells we need to see.
Working with iPS cells has numerous advantages compared with other methods. For one thing, adult cells can be used to make iPS cells, so researchers do not require embryos. Furthermore, the reprogrammed cells have the patient's DNA, so their body will be less likely to reject them.
"Cell reprogramming is revolutionising medicine, yet surprisingly little is known about how it actually works," says stem cell expert Vincent Pasque, assistant professor at KU Leuven, Belgium.
Together with Kathrin Plath from UCLA, Vincent Pasque from KU Leuven led an international study into how adult cells are reprogrammed to iPS cells. To investigate the different stages of the reprogramming process, the researchers separated male and female cells to examine them side by side. The researchers discovered that female and male cells behave differently after the reprogramming process due to their different number of X chromosomes—two in female cells and one in male cells.
Vincent Pasque explains: "In a normal situation, one of the two X chromosomes in female cells is inactive. But when these cells are reprogrammed into iPS cells, the inactive X becomes active. So the female iPS cells now have two active X chromosomes, while males have only one. We now know that this explains the different behaviour of male and female cells after reprogramming because, over time, female cells lose one of the two X chromosomes and start behaving more or less like male iPS cells."Microscope picture of a colony of iPS cells obtained by reprogramming a specialized cell for two weeks. Credit: Vincent Pasque
The differences between male and female iPS cells are most pronounced for one of the key processes in the field of epigenetics: DNA methylation, a modification that changes the activity of a DNA segment without changing its sequence. The researchers found that DNA methylations are erased in female iPS cells but not in male cells.
"Our study also indicates that the path towards iPS cells involves the use of specific control switches on DNA called enhancers," adds Dr. Constantinos Chronis from UCLA.
Vincent Pasque concludes: "Our results show that studying male and female cells separately is key to a better understanding of how iPS cells are made. And we really need to understand the process if we want to create better disease models and to help the millions of patients waiting for more effective treatments."A colony of iPS cells obtained by reprogramming fibroblasts. Credit: Vincent Pasque
Honey Bee Colony
Honey bees are social insects that live in colonies. Honey bee colonies consist of a single queen, hundreds of male drones and 20,000 to 80,000 female worker bees. Each honey bee colony also consists of developing eggs, larvae and pupae.
The number of individuals within a honey bee colony depends largely upon seasonal changes. A colony could reach up to 80,000 individuals during the active season, when workers forage for food, store honey for winter and build combs. However, this population will decrease dramatically during colder seasons.
Honey bee colonies depend upon diversity of population for survival, as each caste of bee performs specific tasks. Thus, while queens are extremely powerful within their societies, they cannot establish new colonies without the help of drones and workers, who provide fertilization, food and wax to construct the hive.
All members of a honey bee colony undergo complete metamorphosis, passing through the egg, larval and pupal stages before becoming adults. Honey bee larvae are legless grubs that eat honey, nectar or pollen. Larvae shed their skin and molt several times before they enter the pupal stage. After another molt, these pupae will emerge as adult honey bees and begin to perform specialized tasks for the colony.
Queens are the only members of a colony able to lay fertilized eggs. An egg-laying queen is important in establishing a strong honey bee colony, and is capable of producing up to 2,000 eggs within a single day. Queens mate early in life and store up millions of sperm within their bodies. While they are capable of living up to five years, they only often only live two to three years producing eggs.
Worker honey bees are the largest population within a colony. Worker bees are entirely female, but they are unable to produce fertilized eggs. If there is no queen they do sometimes lay unfertilized eggs, which become male drones. Worker bees use their barbed stingers to defend the colony, but after attacking, the barbs attach to the victim's skin, tearing the stinging bee's abdomen, resulting in death.
Workers are essential members of honey bee colonies. They forage for pollen and nectar, tend to queens and drones, feed larvae, ventilate the hive, defend the nest and perform other tasks to preserve the survival of the colony. The average life span of worker bees is approximately six weeks.
Drones, or male honey bees, have only one task: to fertilize new queens. Drones mate outdoors usually in midair and die soon after mating. Some honey bee colonies will eject surviving drones during fall when food for the colony becomes limited.
Honey bee swarming is a natural part of a developing their colony. Honey bees swarm as a result of overcrowding within a hive. To create a swarm, an old honey bee queen leaves the hive with about half of the hive's worker bees, while a new queen remains in the old hive with the rest of the workers. In the wild, honey bees swarm most in late spring and early summer, at humid times of the day. While swarming is part of the healthy life cycle of every honey bee colony, beekeepers often attempt to reduce the incidence of swarming in domesticated bees.
A honey bee swarm may contain hundreds or thousands of worker bees and a single queen. Swarming honey bees fly temporarily, and then cluster on shrubs and tree branches. The clusters rest there for several hours to a few days, depending on weather conditions and the amount of time needed to search for a new nesting site. When a scout honey bee locates a good location for the new colony, the cluster immediately flies to the new site.
Generally, honey bee swarms do not harm people. Swarming honey bees do not have young or a nest to defend during the swarm, and as such, their incentive to sting is reduced.
However, a swarm of bees will attack when provoked, as workers attempt to protect their queen. Should a persistent swarm of bees appear near your home or garden, it may be necessary to contact a pest control expert to assist in relocating or exterminating the swarm. Honey bees are a protected species in some areas, so check with a professional pest control expert before taking any action yourself.
Trans + Gender = Civilizational Collapse
Before anyone gets too excited, this is not an essay about people with non-standard sexuality. This is an essay about the power of words.
There is a peculiar fashion, like so many such fashions driven by academia, for elements of words to acquire a kind of magical glamour. Not whole words: grammatical elements of words. The first example to consider is ‘meta’.
I’m sure we’ve all heard people saying something like ‘that is so meta!’. Few people would be able to give any half-way convincing definition of this adjective, but from a grammatical point of view it is a repeated pattern: a word elements, here a prefix (meta) through some mysterious process gets detached from its original lexical use, acquires glamour, and is firstly freed up to form a slew of new compounds, and finally achieves greatness as an independent adjective, but one which is pretty much meaningless.
‘Meta’ is a good example to use because in this case there is no mystery as to its origins. I lived in Greece for a while and was surprised to discover that in Modern Greek, ‘meta’ is a perfectly ordinary everyday word meaning ‘after’. In Classical Greek, the situation is a little bit more complicated, as it originally meant ‘in the middle of’, being connected with ‘mesos’, which means middle, but by Hellenistic times, which is what concerns us for this vocabulary item, the meaning had already settled down to ‘after’.
In the English-speaking world, we today tend to think of Classical Greek philosophy as a kind of arm-wrestling match between the two titans, Plato and Aristotle, but this is a rather recent development, caused mainly by a revival of interest in Plato in Oxford University during the Victorian period. Before that, for approximately two thousand years, there was only ‘The Philosopher’: Aristotle, who was assumed to have solved most of the problems of natural philosophy, and who could be cited as an authority without any need to take the views of other thinkers into consideration. This unrivalled position explains, for instance, why Thomas Aquinas, seeking to establish a logical basis for Catholic theology, adopted what is on the face of it the unlikely project of seeking to prove that Aristotle’s philosophy agreed on every single point with Catholic doctrine. The Arab investigation of Greek thought during the Golden Age of the Caliphate had a similar aim, but the congruence they sought to prove was with the Koran. Whatever your cosmological tastes, then, the origin of all true philosophy was clearly The Philosopher.
There were some obvious difficulties with this. Aristotle had not left a single, unified system of thought (like Kant, for instance), or a few distinguished standalone volumes (David Hume), or even, like Heidegger, a shelf of works so dense and difficult to understand that nobody can really tell if they are connected into a coherent system or not. What we have from Aristotle is a ragged collection of works of different kinds. Some are whole volumes, finished and perfect, but others are often said to be lecture notes, meaning that they cannot really be understood unless you add a lot of assumptions about what he meant to say. Some of them are also now believed to have been written by other people.
Be that as it may, a standard Greek edition of Aristotle was established, later called the Corpus Aristotelicum in Latin translation, and still the basis of all modern editions. This divides the surviving works into five sections, and still includes some 13 works now thought to be spurious. In between sections 2 and 4 there is the shortest section, consisting of one book only, called ‘Metaphysics’. So the order of the sections is Logic, Physics, Metaphysics, Ethics and Politics, and Rhetoric and Poetics. The Metaphysics is internally disorganized, perhaps consisting of lecture notes amended by later editors, but nevertheless set the standard for a whole style of philosophy ever since, dealing as it does with the big philosophical issues of the nature of existence and the nature of meaning.
Aristotle called this kind of philosophy ‘first philosophy’, and we can see why. If these first issues are not solved, then how can you say anything at all about the world with certainty? He did not call it ‘metaphysics’, and indeed could not have done this, as the standard edition of his works was only compiled after his death. The word ‘metaphysics’ means ‘after the Physics’, and refers to that shortest section that comes after the longest one, the Physics.
While Latin, the language of the Western Church, survived in the West, Greek, now the language of the Eastern Church, did not, and through the middle ages it was only Latin that kept the flame of learning alive in Western Europe. Hence the expression ‘It’s all Greek to me’, said to come from a notation made by mediaeval monks copying Latin manuscripts and coming across a passage of Greek. They would write ‘Graecum est non legitur’: It is Greek, and cannot be read. But Aristotle’s writings, together with a scrappy and incomplete collection of other Greek works, survived in Latin manuscripts, and the book on first philosophy was still known in Latin as the ‘Metaphysica’. In due course, once the Renaissance had revived interest in classical civilization, this was in turn translated into ‘metaphysics’ in English, and the natural interpretation of this impressive word (meta having no meaning in Latin) was that this was some kind of extension of physics into a different area of study. ‘Metaphysics’ and ‘metaphysical’ thus became established English words, and would probably have languished in obscurity as recondite learned language had not John Dryden and Samuel Johnson decided to label a gang of poets they disapproved of as ‘metaphysical’, a simple piece of sarcasm suggesting that they were over-learned and over-abstract. Like other such attempts, this cleverness backfired, and the Metaphysical Poets are now in all histories of English literature, and John Donne, chief among them, outshines Dryden and is widely read.
There was then a gap of over two hundred years before it occurred to anyone to separate the two halves of ‘metaphysics’ again. There are a few other English words starting with ‘meta’ that have a slightly different origin, deriving from different and earlier meanings of the Greek word. So ‘metamorphosis’ survived directly in Latin and then in English from Greek, in which the sense of ‘meta’ is ‘trans-‘, as in ‘transformation’. The medical terms ‘metabolism’ and ‘metastasize’ have a similar origin. But the Oxford English Dictionary cites the first modern usage of the prefix meta- only from 1917, and the original uses were fairly respectable. The mathematician David Hilbert proposed ‘metamathematics’, directly on the model of the mistaken etymology of ‘metaphysics’, and ‘metatheorem’, ‘meta-data’, and other well-defined technical uses soon followed. People probably sort of assumed there was some connection with metamorphosis and metamorphic rock. But the genie was now out of the bottle, and by the 1970s the glamorous new-old term was being added to all sorts of other words to make vaguely scientific-sounding and above all glamorous, fashionable formulations. What exactly is metapsychology? Metafiction? Metaethics? Metahistory? I’m sure the inventors of these individual terms did so with the best of intentions (‘Metahistory’ is the title of a 1973 book by Hayden White presenting a theory about the writing of history), but the net effect was to reduce the prefix meta- to meaninglessness. Meta-feminism. Metamodernism.
There seems to be a common end-point in this strange linguistic process – firstly, the shift from a bound word element to a free-standing adjective, and secondly an idea of something superlative. Metahistory as an improved form of history. When all the other meaning has been sucked out by the proliferation of inaccurate or unjustified uses, that sense of improvement is all that is left. Metaphysics is felt to be superior to physics in some undefined way: deeper, probably, and more permanent, and so by analogy meta-ethnic, meta-racism, meta-gender, and similar recent formulations have practically nothing left of anything you could really call meaning, but signal instead a spurious depth and seriousness. An implied claim to universal truth.
I have chosen ‘meta’ to start with firstly because its entire history is well understood and easy to demonstrate, but also because it is a neutral term in the game of political correctness. At the other end of that spectrum is a word element which has a different kind of etymology: phobia.
This is not originally a prefix or suffix but an independent word, also Greek, and still retaining its original meaning as an irrational and obsessive fear of something, used thus in psychiatric terms like claustrophobia or agoraphobia. There is a general phenomenon of watering-down that often happens with psychiatric terms. They are devised to describe extreme and disabling psychological states. The origin of modern psychiatric treatment is sometimes said to be the need to treat large numbers of shell-shocked soldiers in World War I, when it was observed that they had the same symptoms, regardless of their personal experiences. These symptoms were highly disabling, and the same goes for clinical depression, paranoid schizophrenia, and other familiar conditions. But when someone says ‘I’m depressed today’, they don’t mean that they are curled up in bed facing the wall and unable to face the idea of any human contact, perhaps for years on end, as the truly clinically depressed experience. They are just a bit unhappy about something. ‘Don’t be so paranoid’. It is a natural process, as starting with Freud, psychiatry provided a new terminology for symptoms we could all understand as exaggerated forms of our own experience. A more recent example of specialized psychiatric vocabulary passing into ordinary language is ‘dysphoric’. A slippery diagnostic principle has become a matter for virtue signalling: I’m feeling so dysphoric right now! Who would even have been able to guess what this was supposed to mean a few years ago?
But this is not what has happened with ‘phobia’. At a social level, there do indeed seem to be people who have irrational, obsessive fears about other groups of people. A particular kind of anti-Semitism is one example. There are people who basically devote their lives to hating and fearing Jews, blaming them for all their personal difficulties and for all the ills of society. To give a current example, there are people who are convinced that ‘the Jews’ must be responsible for the Covid-19 epidemic. They don’t need evidence. They just know it. But the term Jew-phobia has not arisen. Similarly, there is a certain kind of feminist activist described perfectly by the actress Emma Watson at the UN in 2014 as practising ‘man-hating’. In Sweden in the 2000s, a national network of women’s centres called ROKS (Riksorganisationen för kvinnojourer och tjejjourer i Sverige) developed a parallel society on that basis, with runaways and other young women in distress welcomed into an environment where not only did they never meet or interact with a man, but where the older women in charge indoctrinated them into a ‘men are animals’ ideology according to which there was an international conspiracy of ‘powerful men’ that routinely kidnapped, tortured and murdered women for fun, so that it was only by staying within the confines of ROKS that they would be protected from the roving murder gangs, often disguised as policemen. But is any of this described as ‘man-phobia’ or ‘androphobia’? These words have not become established, even if this looks like a clear-cut case of an irrational and obsessive fear.
Instead, ‘phobia’ has taken on a new life by changing its lexical function, so that in compound words it no longer means what it has always meant. Do people who use the word ‘Islamophobia’ really mean that people suffer from an obsessive, irrational fear like people who suffer from claustrophobia or another psychiatric condition? A crippling disability? I suppose that in some mental ward somewhere there is some unfortunate who is convinced that there are Muslims everywhere, responsible for all their misfortunes, and who can’t stop thinking about this, as hampered by this belief as someone whose agoraphobia stops them leaving the house, but I have never met such a person, and the people who are nowadays labelled Islamophobic don’t fit that description either.
I suppose in this case the link is the term ‘homophobic’, and while it’s also an inaccurate term, it has a bit more of a justification, insofar as there evidently are, and certainly used to be, a lot of overtly heterosexual men with some kind of psychological complex about gay men, either because of fear of their repressed homosexual feelings, or because of the need to maintain denial of their own gay experiences. But irrational and obsessive fear? Really? Again, maybe, somewhere, there is someone like that, but it is an overstatement. Created in the 1960s by a psychologist called George Weinberg, the word was a weapon in the war for gay liberation, but should surely have been laid to rest by now. Instead, we find that Wikipedia defines homophobia as encompassing ‘a range of negative attitudes and feelings toward homosexuality or people who are identified or perceived as being lesbian, gay, bisexual or transgender (LGBT).’ This is a very convenient definition for those wishing to use this as a term of abuse, as this ‘range’ is not defined. In fact, rather than describing a phobia in any real sense, it is a term of ridicule and abuse that can be applied to anyone saying anything at all, even once, that could be interpreted as claimed in the definition, according to a flexible set of criteria changed at will by the thought-leaders and activists involved. What this boils down to is that -phobia as a political term now just means ‘having a politically incorrect attitude towards’.
There are a few other word elements we could look at, notably post-, which like meta, just means ‘after’, and which has had a similar trajectory, so that post-structuralism is seen as some kind of ultra-structuralism or meta-structuralism, deeper and altogether superior, with the same kaleidoscopic variety of other formations. It even sporadically makes it as an independent adjective and noun, referring to the post-modernists and post-structuralists collectively (‘the posts’) – but it hasn’t really passed into the language like that. ‘I am post’ remains incomprehensible. But with the discussion of phobia we at last arrive at ‘trans’, for ‘transphobia’ has joined the short but distinguished list of neo-phobias, and even more mysteriously. Since the expression ‘trans’ only acquired the sanctification of independent adjective status in the 21 st century, it seems even less likely that there is a widespread mental disorder based on irrational and obsessive fear of the trans. This is doubly so as it is not at all clear what the term ‘trans’ is supposed to mean.
Dictionaries typically give three or four definitions of the prefix ‘trans-‘, but from an etymological point of view they boil down to two, both from Latin, and cognate with Greek ‘meta’ in some of its senses. Either it means ‘on the other side of’, or it means ‘undergoing change’. The first use is the original one, but is now by far the rarer, being limited to chemistry, where it refers to chemical bonds being on the other side of a line of symmetry, and geography, the Romans having named two of their European provinces Trans-Alpine Gaul and Cis-Alpine Gaul. This terminology has been borrowed a few times, with Transjordania and Cisjordania meaning territories on one side or the other of the river Jordan, and Transkei and Ciskei referring in the same way to the two banks of the river Kei. The prefix cis-, meaning ‘on this side of’ is also used in chemistry, this being an analogy to the geographic use.
Far more common is the meaning of ‘trans-‘ as to do with change, as in translate, transform, transaction, transition, and many other familiar words. It is immediately clear that none of these words have opposites with cis-. A moment’s thought shows why. If we say that something is on the other side of a barrier, then we can conceive of something on this side of it. But if we say that something changes, then there is really no need to have a word for the state in which no change has taken place. If we do want to form a negative, we negate the whole word. If something is not translated, it is not ‘cislated’ it is untranslated or just not translated. ‘Cisformed’ is incomprehensible untransformed or not transformed is fine. Un-transacted. With nouns, though, even this kind of negation is impossible, not exactly for grammatical but for logical reasons. An iconic ‘trans’ word, the title of Lou Reed’s 1972 album that included ‘Walk on the Wild Side’, is ‘Transformer’. So what is the opposite of that? Cisformer? Not transformer? Like most common nouns, it has no negation.
So how has it come about that when applied to sex and gender, the word ‘trans’ is widely believed to have an opposite in ’cis’? Cis-gender, cis-sexual, even cis-man and cis-woman? There is only one explanation. In this usage, both ‘trans’ and ‘cis’ are meaningless. In fact, they share the special kind of meaninglessness we have already seen in the Wikipedia definition of homophobia. As a purely ideological term, ‘trans’ can mean whatever the fashion of the moment decrees. Currently, one of its main uses is in the expression ‘the trans community’, a term which suggests that there are lots of different kinds of trans people (or ‘folks’ as they are often called for reasons that escape me: is ‘people’ a politically incorrect term too?) and they are all joined together in one big happy family. So I have a question. What about transvestites? Are they part of the ‘trans community’? And if not, why not?
There has been an attempt to unperson the transvestite population.
Paradoxically, it seems that it is only by disavowing their ‘trans’ name that transvestites are allowed into the ‘trans community’. The word ‘transvestite’ dates from the 1920s and therefore predates both ‘transsexual’ (1949) and ‘transgender’ (1966), and there was no mystery about what it meant. There were female transvestites, often lesbians who dressed as men in public, but the bulk of transvestites were heterosexual or mainly heterosexual men who got a sexual thrill from dressing up in women’s clothes at home. They might occasionally go to some party or special event, or risk an outing to the pub, but it was mainly a private passion indulged at home. Since the ‘born in the wrong body’ ideology had not yet been invented, it never occurred to anyone that indulging in this harmless fetish was the start of an important journey towards changing into women or anything like that. Transvestites don’t want surgery or hormones and are content with the thrill of occasionally dressing up. Since this does not fit the narrative of trans ideology, according to which any small boy who shows an interest in putting on a dress is to be whisked off for a lifetime of drugs and surgery ‘before it is too late’, happily heterosexual adult transvestites (often married, and borrowing their wives’ clothes) are something of an embarrassment. However, it seems that if they stop calling themselves transvestites and start calling themselves ‘cross-dressers’ then they are allowed to be part of the trans community. Since ‘cross-dresser’ and ‘transvestite’ mean exactly the same thing, the only difference seems to be the use of the prefix ‘trans’.
As with the definition of homophobia above, the point seems to be that the term ‘trans’ has no objective or core meaning, but rather a vague meaning whose application changes as fashion changes. In particular, apart from the problem with transvestites, it has the advantage of not making any specific demands for who may be included. This suits one important group: the allegedly trans-sexual men, often young men, who are identified by some experts as auto-gynephilic, the suggestion being that they have a paraphilia or fetish that means they are fixated on the image of themselves as women, accentuating this by dressing as women, but not desiring any kind of hormones or treatment…. In other words, they are transvestites, but rather exhibitionistic ones, and with the key difference from an ideological point of view that on the basis of this fetish or whatever you want to call it, they now claim actually to be women. Bruce Jenner suddenly coming out as a transvestite would have been much less exciting, after all. Since they are simultaneously heterosexual men from a biological point of view, this produces the odd result that they often also claim to be lesbians. This demographic is very important in the trans community, and also now in what is left of the women’s movement, with institution after institution coming to terms with the fact that these aggressive young men are taking over. But a real head-scratcher is that they don’t consider themselves to be cross-dressers either. Pious explanations are given of the difference between transvestites and these male-bodied trans women, and it is pretty clear that the former are not regarded as part of the community, while the latter are its core membership. There doesn’t seem to be much of a difference between them apart from that.
So once again we find that a fairly ordinary word component, trans, has been separated from its original meaning and taken on a new life as a carrier of glamour and magic. We have seen from looking at the trans/cis pair that ‘trans’ as a sexual marker is now pretty much a meaningless term. But many would object that ‘trans’ is just short for ‘transgender’. Since from purely lexical point of view, ‘trans’ does have a core meaning, perhaps it is this formulation that should properly be identified as meaningless.
The main reason that ‘gender’ has become such a fashionable term in the English-speaking world is that the English language has only a vestigial gender system. Swedish, too, has a vestigial gender system, though it functions differently, and Sweden is the other main site of gender activism in the world, having influenced the rest of Scandinavia through pan-Nordic institutions. Apart from those countries, Israel and Germany stand out, but that is because of the high proportion of English-speakers there.
In English, gender is not a grammatical but a semantic concept. Boys are called ‘he’ because of the meaning of the word, and not just because of grammatical convention. Compare this with French. In French, every noun has a fixed grammatical gender, and French people are programmed to accept this as an inherent part of the word, like its pronunciation and spelling. ‘Table’ and ‘chair’ are both feminine. ‘Floor’ and ‘ceiling’ are both masculine. This is obviously not their sex it is their gender. From a biological point of view there is nothing female about tables and chairs, and nothing male about floors and ceilings. The gender of the noun then determines the form and sometimes the pronunciation of adjectives attached to it, and also the pronouns used to refer to it. So the noun for ‘person’ is feminine and takes feminine adjectives and is replaced by the feminine pronoun elle or she even when it is referring to a man.
Discussions of gender have been very slow to get off the ground in France and in many other countries with gender-based grammar like this (basically the whole of Europe except for Sweden, Finland, Estonia and Hungary, plus Armenia, Georgia and Turkey in Europe’s near-abroad) because the idea of changing gender seems inherently impossible. If you decide to say le table and declare that the table is now masculine, it doesn’t seem perverse or iconoclastic or daring or anything: it just looks wrong. Not morally or politically wrong: just incorrect.
In English, though, once the notion of gender-fluidity was introduced, it took off like a rocket. The idea seems to be that since we have three genders, masculine, feminine and neuter, corresponding to he, she and it, a better way of conceiving of gender is as a spectrum, a beautiful rainbow with masculine and feminine now only the end-points, and nearly everyone somewhere in between. Indeed, from the way that ‘gender’ is used, you could infer that that is the main difference between sex and gender: that sex is fixed but that gender is excitingly variable. If that were true, though, then surely the natural thing would be for all those claiming to be somewhere near the middle of this distribution to demand to be called ‘it’, and that clearly does not happen. The only people who use the pronoun ‘it’ for people claiming to be neither male nor female are those seeking to insult them, and the reason is clear. The semantic category it refers to objects other than people or ‘near-people’ such as pets. It is not at all the mid-point of a beautiful rainbow, because when it comes to gender, the beautiful rainbow does not exist. There are no French chairs that are 60% female and 40% male, or female on weekdays but male at the weekends. This is not because of the tyranny of the patriarchy, it is just a fact of life. Language has evolved like this. There is nothing you can do to change it. ‘Male’ versus ‘female’ is one of the most basic categories of life, meaning not just human life but the life of animals and plants as well.
This is confirmed by a look at a couple of languages that, unusually, have no grammatical gender at all. English and Swedish have each evolved from older versions of the language that did have gender, which is why they retain some vestigial gender structures, but Finnish and Chinese are gender-free. Not only is there no agreement between nouns and adjectives on the basis of gender, there is no gender distinction in pronouns either. Does that mean that the semantic categories don’t exist either?
In Finnish, the pronoun hän means ‘he’ or ‘she’ without distinction. In speech, it is also often used for ‘it’, but there is also a separate inanimate/non-human pronoun ‘se’. Students of historical linguists teach us that this animate/inanimate distinction is actually the origin of grammatical gender, with the sexual distinction added on afterwards. Other Uralic languages work in the same way. Swedish gender activists have long thought that this is wonderful and on the model of Finnish, have tried to introduce a neutral pronoun hen to replace their blatantly sexist words for ‘he’ and ‘she’. But unfortunately they are confusing grammatical and semantic concepts of gender. My mother was a Finn, and despite speaking English every day for fifty years, until the end of her life she would confuse he and she, even when referring to her own children. But that doesn’t mean that she didn’t know whether we were boys or girls. Her concept of basic sexual categories was not affected at all by this linguistic peculiarity. But because she had not been brought up doing it, she had the same difficulty with ‘he’ and ‘she’ that an English-speaker has remembering to make the table feminine and the floor masculine in French. It could never be 100% automatic.
Chinese is an interesting example too. The third-person singular pronoun is tā, indifferently he, she, and it. But even though there is a total absence of grammatical gender in Chinese, we see again that this has no effect on the inherent human sexual categories. There are three different characters used to write this word in contemporary Chinese, with no change in pronunciation or grammar, and the three characters mean he, she, and it. Furthermore, the character for ‘she’ is an invention of the early twentieth century, influenced by Western languages, so for thousands of years the language was like Finnish, with just one word for he and she but as with my mother, the absence of grammatical gender in their lives made no difference to their understanding of sexual categories. Indeed, Imperial China was just about the most patriarchal and sexist country you could imagine, with wives obeying and daughters ignored. This is an indication of how deeply misguided the gender activist project is.
We can extract two simple principles from all of this. The first is a simpler way of putting the distinction between semantic and grammatical gender:
Gender is a characteristic of words and not of things or people
And the second is an observation based on the table/chair/floor/ceiling example, though we could have looked at almost any other examples:
Gender is inherently meaningless.
The gender of nouns in French can be explained by referring back to Latin, but that doesn’t get us any closer to any actual meaningful way that the table is female. The table has no sex, but the word for table has a gender. In English, the same word, ‘table’, has the default neuter gender, meaning only that it takes the neuter ‘it’. To qualify for female gender the table would not only need female sexual characteristics, it would also need to be a human being or a provisional human being. In other words, it would be its sex that determined this use of ‘she’. Gender in English is a marker of sex and nothing else.
Despite all the talk of gender-fluidity, what we see when we look at actual gender systems in real human languages, including English, is a simple binary logic, with another, independent animate/inanimate (or human/not human) distinction sometimes applied as well. It would appear that these semantic distinctions are innate and pre-linguistic, as they are still present among speakers of languages that make no formal gender distinctions. Some readers, however, will feel that what I have been discussing is not gender at all: not, anyway, what they mean by gender. So let us turn to the contemporary misconceptions.
It is amazing to me that nobody ever mentions the origin of the modern conception of gender as a social construct. This is now the established idea of gender among those who feel they can safely ignore all the facts about gender in language. The United Nations Entity for Gender Equality and the Empowerment of Women, for instance, gives the following long-winded definition:
Gender: refers to the social attributes and opportunities associated with being male and female and the relationships between women and men and girls and boys, as well as the relations between women and those between men. These attributes, opportunities and relationships are socially constructed and are learned through socialization processes. They are context/ time-specific and changeable. Gender determines what is expected, allowed and valued in a women or a man in a given context. In most societies there are differences and inequalities between women and men in responsibilities assigned, activities undertaken, access to and control over resources, as well as decision-making opportunities. Gender is part of the broader socio-cultural context. Other important criteria for socio-cultural analysis include class, race, poverty level, ethnic group and age.
This is another one of those definitions that is at once so broad and so vague that you need a guide to interpret it. What are the last two sentences doing there, for instance ? Is the idea that class, race, poverty level, ethnic group, and age are part of the definition of gender? Or has it just been cut and pasted in by mistake? What it appears to say is that gender is just one among many factors that must be taken into account in a complete analysis of socio-cultural context, which is a very peculiar way to go about a definition. And how does this definition apply to the many gender-trenders who want to have some non-specific or third gender that is not just a question of male and female?
We could try a more intellectual definition, maybe. Catharine MacKinnon is a leading – the leading, actually – feminist legal theorist, and she offers this definition of gender.
Stopped as an attribute of a person, sex inequality takes the form of gender moving as a relation between people, it takes the form of sexuality. Gender emerges as the congealed form of the sexualization of inequality between men and women. (p. 6)
By what stretch of the imagination do these two definitions refer to the same thing? Once again, MacKinnon makes the elementary mistake of defining gender in terms of men and women (in her defence, this was 1987, before these terms had been made obsolete), but what we can see is simply that the definitions don’t matter. Ask ten gender experts to define gender, and you will get ten different answers – different not just in their details, but in their frames of reference and general concepts. And yet they will all nod and agree with each other and look happy. It is merely using the word ‘gender’ that unites them: as it is merely using the word ‘trans’ that unites a more recent but very similar group. Who cares what any of it means? But where did it all come from? ‘Gender’ used to refer only to grammatical gender. At some point in the 1950s or 1960s a sea change came about, and the word ‘gender’ began to be used in relation to sexual identity. What happened?
There is a problem of chronology, which, however, a little reading soon dispels. The great authority on gender for a generation was the New Zealander John Money, who literally wrote the book on the subject in the form of the 1972 university textbook (with Anke Erhardt) Man & Woman, Boy & Girl, a comprehensive survey of the current state of knowledge of the biology of sex with an emphasis on scientific explanations of various physical and chromosomal abnormalities. Money was not just a theorist. In 1965 he was a co-founder of the ground-breaking Johns Hopkins Gender Identity Clinic, and supervised what at that time were called sex-change procedures. He seemed very modern and liberating, and introduced a lot of trendy vocabulary, most of which (ever use the word ‘ycleption’?) has disappeared but he wrote about gender identity and gender roles, and was such a dominant figure in the field that most people have assumed that he invented these terms and thus set up some underlying though unclear theoretical frame of reference to go with them.
However, in a footnote in his textbook, he refers the reader for gender to another author working in the same field, Robert J. Stoller, whose ideas could not be hiding in plainer sight, housed as they are in a thick two-volume compendium entitled Sex and Gender. The problem of chronology arises because while Money was a prolific author who evidently worked fast, Stoller took several years to compile this work, which is a study of the case histories of 85 patients and 63 family members from the clinical experience of the Gender Identity Research Clinic of UCLA, where he was Professor of Psychiatry, and his personal clinical practice from the late 1950s on. The UCLA body was not a clinic in the normal sense, but largely a discussion group among the various professionals involved in treating intersex people and other problem cases, though they did interview clients too. Everyone working in this emerging field obviously knew each other and kept in touch, and while John Money kept up a stream of neologisms and pet theories, Stoller’s work is a slow, solid attempt to create a new theory of sex and gender. It is recognizably the theory used today, in which sex is a biological given, but gender is something created by the circumstances of life, including family dynamics and social expectations. The first volume of Sex and Gender appeared in 1968, but the second only came out in 1974, and Stoller wrote that he considered them a single study. He writes approvingly of Money, sharing his belief that one can socialize a genitally female child as a boy and vice versa (for instance in cases of ambiguous genital development), a view which is much less popular today. However, while Money had a concept of gender role, as something culturally determined and elastic, Stoller gravitated towards the idea of a core gender identity, established as part of the developmental process and then permanent.
Apart from influencing John Money and other clinicians, Stoller also found a follower in the feminist sociologist Ann Oakley, whose Sex, Gender and Society from 1972 took up Stoller’s conception of gender and introduced it into feminist discourse, where it has certainly flourished. In a 2016 study called Intersexualization the Clinic and the Colony, Lena Eckert devotes a chapter to the history of the concept of gender in feminism, and writes (page 78) that ‘Sex and Gender… was crucial for feminism, since the gender -concept has been derived from this book.’ So it is Stoller, not Money, who started the whole gender bandwagon rolling.
However, what nobody seems to have grasped at the time is that Stoller, a professional psychoanalyst, saw as a main goal of his theories an alignment with orthodox Freudian theory. Eckart understands this, but has missed the explanation given by Stoller in a preface which seems to have been written for a one-volume reprint of both parts of Sex and Gender from 1978. He explains the rationale for separating sex and gender in Freudian terms, and before going into a more general discussion of ‘The Interpretation of Dreams’ and ‘Three Essays on Sexuality’, actually gives a specific reference in Freud’s writings for the origin of this theoretical separation: ‘The Psychogenesis of a Case of Homosexuality in a Woman,’ (1920). This is an account of a case in which Freud was called in as an act of desperation by a bourgeois Viennese family whose 18-year-old daughter had fallen in love with an older woman of questionable reputation, and the aim of the treatment was to cure the girl’s ‘genital inversion’. Freud’s description of the case is in general a sympathetic one. He gives up on the assigned task, suggesting that the girl be treated by a woman doctor, and sees the homosexual attachments of the girl in a neutral way, without moralizing. But his analysis of the case reads like a caricature today. The girl, we read, had a normal Oedipal development, but Freud was able to get to the bottom of her case thanks to her descriptions of her dreams. She suffered from a bad case of penis-envy, and when her mother became pregnant, she reacted by ‘turning into a man’. What Freud makes of all this is a theory of homosexuality which is also a theory of human sexuality more generally. This is that a person’s sexual identity has three parts: somatic sexual characteristics, psychical sexual characteristics, and kind of object-choice. Stoller quotes this passage, and adds after ‘psychical sexual characteristics’:’(what we are now calling gender)’.
It is not immediately clear what this passage means, but fortunately Freud explains it in the article. By ‘sexual characteristics’ he means simply male or female. A person may be physically male or female, psychically male or female, and choose a male or female sexual object. So a man may, for instance, be of a feminine psychical type (have a female attitude, Freud says), and yet be attracted to women. In the case under consideration, the woman is of masculine attitude or psychical type, and is also attracted to women. This rock-paper-scissors approach explains all the mysteries of homosexuality. ‘If we take these findings into account, then the supposition that nature in a freakish mood created a ‘third sex’ undoubtedly falls to the ground’.
So, to be clear, in these two examples, according to Stoller’s interpretation of Freud, the man has a female core gender identity but is nevertheless heterosexual as he chooses a female object, and the woman has a male core gender identity and is homosexual as she chooses a female object. It is amazing that this theory should have been transplanted – one must assume by mistake – into feminist discourse and spawned the monster of congealed sexualization and all the rest of the mumbo-jumbo that has followed on. It provides no basis whatsoever for the assumption that gender is something infinitely elastic and malleable. The homosexual woman is psychically a man, and that’s it. If anything, he believes, and with some evidence if we think of intersex people, that it is the physical sex rather than the ‘gender’ that is uncertain, suggesting for instance that a possible treatment for female homosexuality might be to remove the hermaphroditic ovaries and replace them with single-sex ones – although to be fair he does say there is little prospect of this being applied in practice.
Once you have grasped what Stoller means by core gender identity – a simple matter of whether you feel you are a boy or a girl – his idea, based as it is on decades of observation and questioning – does not look so bad on the whole. ‘Born in the wrong body’ goes out of the window, as he buys into the Freudian account of development, so that, while core gender identity is formed early on, it is only by the ‘genital stage’ (i.e. puberty) that the process of sexual development is complete, and it is ultimately the progress of the Oedipus Complex that decides the outcome. But the idea – shared by Stoller – that people develop a fixed core gender identity that cannot be changed is a compassionate one. As a psychiatrist, Stoller was not himself directly involved in surgery, but he was part of a professional movement that saw surgeries originally intended to correct the genitals of intersex babies applied to adults. But he was cautious, and as these treatments became available, he wrote of the corresponding appearance of a new type of mental illness, which he called transsexualism – a delusional obsession with sex-change surgery, as it was called at the time, as the solution to all one’s emotional and psychological problems. This was another idea with legs, as Money’s clinic at Johns Hopkins was closed down in 1979, partly because a percentage of patients were coming back after surgery and saying they wanted their body parts back, but mainly because the unrealistic expectations of deluded patients had led them to believe that all their problems would be solved by their surgery, and long-range studies showed that the treatment had no measurable effect like that – the simplest marker being that the very high rate of suicide attempts reported among people presenting themselves for treatment (about 40%) was unchanged after it.
You can agree or disagree with Stoller’s Freudian theory (I suspect there would be few takers today), but at least you can understand what it is. Just as people can be divided into male and female physical types, they can be divided into masculine and feminine gender identities. No beautiful rainbow spectrum appears. There is indeed a complete absence of post-modern psycho-babble.
So what has evidently happened is that feminists, and then gender activists and the ‘trans community’ have made a strange move. They were early adopters of a theory they didn’t understand, and which if they had understood would have run a mile from: one which says just about the direct opposite of what they want it to say. So their reaction, collectively and perhaps instinctively, was to treat their new possession, the word ‘gender’, as a magical, totemic source of glamour and excitement rather than as an analytical term. On that basis it became a key element in the progress and professional success of their complicated movement. Ann Oakley had evidently read Stoller (though not his still-unpublished preface) and uses his terminology correctly but this tiresome convention was soon abandoned. The theory of gender is the quintessential Dead White Male theory, after all.
So what we have instead is typified in Judith Butler’s Gender Trouble from 1990, a work which uses the word ‘gender‘ on practically every page, but nowhere offers a definition of what it means. It evidently means whatever she wants it to mean in the particular sentence she is writing, with no need for consistency or rigour, since none of her readers know what it means either. There is an academic debate involving ‘doing gender’ versus ‘undoing gender’. Just nod and agree. In all the explosion of writing on gender, the meaningful part can be identified by performing the experiment of replacing ‘gender’ with ‘sex’ and seeing if it still makes sense.
Since the great revival of interest in sexology in the 1960s, a lot of scientists who should know better have also adopted the word ‘gender’. It thus reverts to its use as a euphemism for ‘sex’. What genuine advantage is there in talking about ‘gender identity’ instead of ‘sexual identity’, ‘gender role’ instead of ‘sex role’. Even the deliciously fashionable ‘gender dysphoria’ would make more sense as ‘sexual dysphoria’. By adopting the language of gender, something done fifty years ago and in good faith, doctors and scientists unwittingly made it possible for gender studies ideologues to sound as if they were talking about something scientific – and even to demand that the doctors and scientists should defer to them as the real experts on the subject.
Apart from what seems to be the core idea that gender is infinitely flexible in meaning, the ideologues seem to have three ideas which are not stated directly but which can be inferred. 1. Gender is not the same thing as sex but is better than sex. 2. Gender is the product of patriarchal society and therefore something to be resented. 3. Gender represents the glorious victory of women/lesbians/trans folks/insert flavour of the month here. This seems to be roughly the emotional complex. But from this contradictory and chaotic mix we are forced to a simple conclusion. If we are going to remove the real historical origin of gender theory and replace it with nothing, we are left with a theory that means nothing. There is, in fact, no such thing as gender anymore.
3. Civilizational collapse
It is not just the word ‘transgender’ that is the problem. Any number of pseudo-intellectual terms have little by little crept in and invaded what used to be respectable areas of study, or driven them out completely and replaced them with mountains of treacherous and dangerous nonsense. Diversity: used by people who insist on total conformity. Inclusion: code for excluding non-believers. Micro-aggression – naturally enough, defined as ‘demeaning meta-communications’.
If you use meaningless language, preferring it in fact to plain and factual explanation, it is not education. It is not even really indoctrination, since if the doctrines themselves can’t be understood, how can you indoctrinate your followers? It is more like a religious system of education, and of course there is plenty of that about and we know how it works. The Mormons have places called universities, offering a wide range of courses, leading to qualifications that look just like real degrees but in order to enrol in these institutions you need to believe – not pretend to believe, mind you: actually believe – that there was an Ancient Egyptian-speaking civilization in Central America whose scriptures were read by Joseph Smith wearing a magic helmet and that the Garden of Eden is located in Missouri. Not that the official learning of the West has been much better. The learned Jesuit Athanasius Kircher, a man of extraordinary talents who was believed by many to be the most knowledgeable man in the world, taught that volcanoes were eruptions from the fires of hell beneath the ground. Centres of Islamic learning today teach that sorcerers can capture djinns and force them to fly to heaven to spy on the angels. The proof of this is shooting stars, which are the weapons the angels use to drive the djinns away. Hundreds of millions of people believe this as a result of this ‘education’. Are these the standards of enquiry we wish to revive?
There is a very great danger from one particular direction. The American universities that are driving the steady decline of intellectual standards are increasingly run not by academics but by professional administrators, and the courses of study they go through to qualify for these well-paid positions have long since been invaded by the politically correct but mentally challenged purveyors of social justice. It was long assumed that the exact sciences could not possibly be taken over by the peddlers of magical slogans and pointless, destructive projects, but this is now happening. Astronomy, for some reason, is a target, and job applicants must not just explain something about astronomy. They must offer plans for making astronomy more diverse, this being the chief goal of the non-scientists responsible for administering the universities. Probably the most extreme case of intellectual corruption is in Sweden. It is the most gender-equal country in the world by all sorts of measures, but the feminists who have taken over the government and the whole of higher education there believe in a ‘discourse of subordination’, according to which Swedish society is gender-segregated and male supremacist. They claim that they want to see the same number of female professors as males in all subjects, and I suppose they will achieve it, but obviously this is not going to satisfy them. There are any number of fresh sources of outrage lined up and ready to go.
All this cascades downwards. It is not just higher education that is in the hands of the anti-rationalists. Some university students, after all, do rebel or even just keep their mouths shut and fail to believe as they go through their obligatory training in micro-diversity and meta-ethnic discourse theory. Primary school children are a much safer bet. So page 45 of California’s 2019 K-3 guidelines on gender identity, aimed at teachers of five-year-olds, states: ‘While students may not fully understand the concepts of gender expression and identity, some children in kindergarten and even younger have identified as transgender or understand that they have a gender identity that is different from their sex assigned at birth.’
‘Give me the child for the first seven years and I will give you the man’ is a saying that was alleged by Voltaire to be a Jesuit maxim. The Social Justice Warrior knows better. ‘Leave him to me. I will give you the genderqueer transman.’ That is the sacrament of their faith.
We sincerely thank Dr. Atsushi Takashima, an assistant professor at the Yona Field of the experimental forest of the University of the Ryukyus for valuable advice on our field research, the staff at Yona Field for their support and assistance during our fieldwork, and Dr. Firouzeh Javadi for her advice on English expression. This work was in part supported by Grant-in-Aid for JSPS Research Fellow (JSPS KAKENHI Grant Number JP 19J20022).
|eth13133-sup-0001-VideoS1.movvideo/mov, 7.5 MB||Video S1|
|eth13133-sup-0002-VideoS2.movvideo/mov, 2.7 MB||Video S2|
|eth13133-sup-0003-VideoS3.movvideo/mov, 3.7 MB||Video S3|
|eth13133-sup-0004-VideoS4.movvideo/mov, 7.4 MB||Video S4|
Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
Reproductive sharing in relation to group and colony-level attributes in a cooperative breeding fish
The degree to which group members share reproduction is dictated by both within-group (e.g. group size and composition) and between-group (e.g. density and position of neighbours) characteristics. While many studies have investigated reproductive patterns within social groups, few have simultaneously explored how within-group and between-group social structure influence these patterns. Here, we investigated how group size and composition, along with territory density and location within the colony, influenced parentage in 36 wild groups of a colonial, cooperatively breeding fish Neolamprologus pulcher. Dominant males sired 76% of offspring in their group, whereas dominant females mothered 82% of offspring in their group. Subordinate reproduction was frequent, occurring in 47% of sampled groups. Subordinate males gained more paternity in groups located in high-density areas and in groups with many subordinate males. Dominant males and females in large groups and in groups with many reproductively mature subordinates had higher rates of parentage loss, but only at the colony edge. Our study provides, to our knowledge, the first comprehensive quantification of reproductive sharing among groups of wild N. pulcher, a model species for the study of cooperation and social behaviour. Further, we demonstrate that the frequency of extra-pair parentage differs across small social and spatial scales.
In group-living species, reproductive conflict can exist at multiple levels of the social structure. Group members may disagree over the distribution of reproduction within the group, as dominant individuals seek to monopolize reproduction while subordinate group members attempt to parasitize dominant reproduction, potentially facing punishment or eviction when doing so [1,2]. Individuals within a group may also come into conflict when group members mate with neighbouring individuals, as the offspring of neighbours may parasitize the resources and parental care of current group members . Both within-group factors, such as group size and composition, and between-group factors, such as the density and location of neighbouring groups, influence the ability of individuals to pursue reproduction within their group [4,5] as well as pursue reproductive opportunities with neighbouring individuals [6,7]. Therefore, it is difficult to understand variation in individual behaviour or reproductive success without exploring how both within-group and between-group social structure affect the quality and quantity of reproductive opportunities.
Within-group attributes can influence the degree of reproductive sharing within the group (reproductive skew [1,2]) by altering the ability of dominant individuals to skew reproduction in favour of their own fitness interests. For example, dominants may be less effective at reproductively suppressing subordinate reproduction when there are many same-sex subordinates acting as reproductive competitors [4,5], or when these same-sex subordinates are close in size to the dominant . Furthermore, the frequency of extra-pair parentage in a group may be related to the size difference between the socially bonded male and female males who are much larger than their socially bonded female may be better at preventing the female from reproducing with other males  or females with larger, potentially higher quality males may be less inclined to pursue extra-pair matings .
Colony-level factors can also have a profound influence on reproductive dynamics. High densities of neighbouring territories facilitate intergroup forays, providing females with opportunities to mate with neighbouring males or to lay eggs in other female's territories . Further, competition with neighbours may be intense in densely inhabited areas and it may not always be possible to prevent one's social mate from mating with a neighbour . In addition to density, the spatial location of a territory relative to other territories may influence reproductive patterns. Territories on the edge of a colony tend to be more exposed to predators [11,12]. If individuals occupying these edge territories must perform more predator defence or practice more vigilance, they may have less time and energy available for reproductive activities . Additionally, because individuals from edge territories may be inferior competitors relative to individuals in the colony centre , high-quality individuals may exploit poor-quality competitors and reproduce with their socially bonded mate . While it has been well established that territory location can have a significant impact on the number of offspring produced [12,16,17], far less is known about how the spatial location of a territory in relation to other territories influences extra-pair parentage.
Finally, within-group dynamics are not independent of the dispersal and reproductive opportunities offered by neighbouring groups [1,2,18]. An individual's ability to control its mate's activity is probably reduced when there is an increased availability of high-quality extra-pair mates in neighbouring groups . Similarly, when there are a high number of opportunities to mate outside of the social group, dominants may be less concerned with suppressing subordinate reproduction within their own group. Further, if subordinates risk eviction to reproduce within the group, the willingness to engage in reproduction may rise if there are many groups to disperse to, or if the quality of their group is low relative to surrounding groups [18–20]. Hence, to better understand why and how patterns of reproductive sharing arise, within- and between-group factors need to be considered in tandem.
To determine how between-group traits, within-group characteristics and the interaction between these two levels of social structure impact rates of extra-pair parentage, we examined within-group and colony-level demographics in relation to parentage in 36 wild groups of Neolamprologus pulcher, a cooperatively breeding cichlid native to Lake Tanganyika, East Africa. These fish form colonies of 2–200 territorial groups, each with a dominant pair and 1–15 subordinate group members who help raise the offspring of the dominant pair . Individuals primarily interact with other members of their own social group and those belonging to neighbouring social groups located within a 3 m radius [22,23]. Individuals strongly prefer to settle on territories in the centre of the colony and will often remain subordinates on centre territories rather than become dominants on edge territories . While there is physiological evidence of reproductive suppression in subordinates of both sexes [24,25], previous laboratory studies have shown that subordinate males and females can reproduce within their home groups [5,25–28] and one study found evidence of subordinate female reproduction in the wild . Here, we investigated the possibility that dominants may lose parentage to neighbours, to subordinate group members, or to both.
We predicted that rates of extra-pair paternity (EPP) would be greater on the edge of the colony and at higher densities [6,7]. By contrast, we predicted that extra-pair maternity (EPM) would be low regardless of territory location or density, because females in this species can recognize and destroy competitor eggs . Regarding within-group characteristics, we predicted that parentage loss would increase as group size and the number of same-sex subordinates increased [4,5] and that subordinates closer in size to the dominant would be more likely to reproduce . Finally, we predicted that a larger size difference between the dominant male and female would allow the dominant male to exert more control over group dynamics, facilitating subordinate or neighbouring female reproduction and suppressing subordinate or neighbouring male reproduction .
2. Material and methods
(a) Study site and collection
Between February and April 2013, we sampled 36 N. pulcher groups found in Kasakalawe Bay, Lake Tanganyika, East Africa (8°46′ S 31°46′ E) using SCUBA. Groups all had free-swimming offspring and were located in seven different colonies or subpopulations ranging in depth between 11 and 13.5 m. In order to minimize disturbance in the colony the sampled groups comprised less than 10% of the total number of groups in these colonies. Colonies consisted of a clustering of distinct social groups each defending a territory. Colonies are separated from each other by large open expanses of sand and rocky rubble uninhabited by N. pulcher [21,30]. Each group's territory was defined as the area of rocks defended by the dominant male and female, who are the largest individuals in the group . The dominant and subordinate fish were identified as part of the focal group if they swam repeatedly under the territory rocks (shelter) without eliciting aggression from other fish within the territory boundaries. We measured territory size for each focal group and mapped the distance to all conspecific neighbours within a 3 m radius. A group was considered to be on the edge of the colony if half or more of that group's territory bordered an unoccupied area, with no other conspecific group within 10 m.
Using fence nets and hand nets, we captured all individuals in these 36 groups that were close to sexual maturity (more than 30 mm standard length (SL): the tip of the snout to the base of the caudal fin ). We then injected 20 ml of a clove oil mixture (one part clove oil three parts 70% ethanol one part water) into the brood chamber (a central rock used for breeding and shelter) to euthanize any free-swimming young (less than 10 mm SL). Groups breed year round in this species, and young collected for this study were all estimated to be two to six weeks old. We brought all the collected fish to the surface, where we measured adult body mass (to nearest 0.001 g) and body length (SL to nearest 0.01 mm). Mature fish were euthanized by immersion in a lethal concentration of benzocaine and the gonads were removed from adults and weighed. The entire carcass of free-swimming young and a section of the dorsal fin from adult fish were preserved with 95% ethanol for parentage analysis.
(b) Parentage analysis
We analysed parentage in 36 groups (24 groups from the centre and 12 groups from the edge of a colony) with a total of 397 offspring (mean: 11.0 ± 0.8 s.e. offspring per group). We genotyped adults and young using six loci previously used on this species [5,29] (US783  Tmo11, Tmo13, Tmo25  LOC101  UME003 ). We extracted genomic DNA from preserved fin clips and young (QIAGEN DNeasy Blood and Tissue Kits) and used it in multiplexed PCR reactions, which involved six primer pairs end-labelled with the fluorescent dyes (FAM, HEX and NED QIAGEN Multiplex Kit). We amplified samples using a Eppendorf thermocycler with a touchdown program set to the following parameters: 15 min at 95°C 8 cycles of 30 s at 94°C, 90 s at 55°C (decreasing 0.5°C per cycle), 60 s at 72°C 21 cycles of 30 s at 94°C, 90 s at 50°C, 60 s at 72°C 30 min at 60°C 60 s at 10°C. We ran amplified samples on an ABI DNA sequencer (3100 Genetic Analyzer) and scored the profiles using G ene M apper A nalysis software.
We used CERVUS v. 3.0 to assign maternity and paternity at the 95% confidence level. The loci were highly variable with an average of 25 alleles per locus (see the electronic supplementary material, table S1). One locus (LOC101) significantly deviated from Hardy–Weinberg equilibrium owing to a high number of null alleles and so we removed it from our analyses. The combined exclusion power of the remaining five loci averaged 0.998 for a single parent and over 0.9999 for a parent pair.
(c) Statistical analysis
Most offspring (98.2%) genotyped had at least one assigned parent in the social group, indicating that both young and group members had been correctly identified and that, unlike some other cichlid species, this species does not appear to adopt unrelated fry . When two of our 36 focal groups were less than 3 m apart, we tested adults from the other, nearby focal group as potential parents. When young could not be assigned to any collected potential father or mother, we determined the minimum number of mothers or fathers that had contributed to that group's parentage by counting the number of alleles at each locus that did not come from the assigned mothers or fathers collected with the brood. As a conservative estimate, we assumed all unknown parents were heterozygotes and contributed two alleles per locus.
We examined the factors correlated with the number of offspring assigned to the dominant male/female within the sampled group and to the subordinates within the sampled group (generalized linear mixed effects models GLMER with Poisson distribution R v. 3.0.2, glmmadmb package [37,38]). We also tested factors influencing the number of parents contributing to a brood using cumulative link models (CLM) in the ordinal package . For all models, we tested the fixed effects of location within the colony (centre versus edge), density of neighbouring groups (number of conspecific groups within 3 m), group size, number of same-sex subordinates within the group, number of offspring collected (to control for variation in brood size) and relative size difference between the dominant pair ([dominant male SL − dominant female SL]/dominant male SL). The colony of each focal group was included in GLMER models as a random effect, but could not be used in a cumulative link mixed model because the model would not converge. For all models, two-way interactions were tested and non-significant interactions were removed from the models. Models for males and females were run separately. None of the within-group characteristics (e.g. group size, dominant size) were significantly different between groups from the centre versus edge of the colony, nor did they vary with the density of neighbouring groups (electronic supplementary material, table S2). Density also did not vary with location within the colony (linear mixed effects (LME): F1,28 = 1.16, p = 0.29).
It is possible that dominants were misidentified in the seven groups in which the dominants gained no parentage. In all other groups, dominants gained more than 50% of parentage and it is unlikely they were misidentified. In one of the five groups where the dominant male gained no paternity, two extremely small subordinate males (30 mm and 32 mm SL) gained all paternity in that group. In two additional groups, extensive behavioural observations confirmed the identity of the dominant male. Therefore, we reran our analyses excluding the other four groups with no dominant paternity (n = 2) or maternity (n = 2). Our results were largely robust to these exclusions (see the electronic supplementary material, tables S3 and S4) and therefore the full dataset is presented below.
(a) Parentage loss
Across our 36 study groups, dominant males fathered 76.3% of collected offspring (303 out of 397 offspring genotyped). The dominant male fathered all offspring in 12 groups, the majority of offspring in 19 groups (mean paternity ± s.e: 72.8% ± 3.4), and no offspring in five groups. Of the 94 offspring that could not be assigned to the dominant male collected with the social group, 34 were assigned to a subordinate male in the same social group (mean 10.8% ± 3.7 s.e. of offspring within their social group), four were assigned to a subordinate male from another sampled group, one was assigned to a dominant male from another sampled group, and 55 could not be assigned to any male sampled (figure 1). These unassigned offspring were probably fathered by males in groups that we did not sample or by males within the group that had been recently evicted or eaten by a predator. The number of fathers per brood ranged from 1 to 4 (mean: 2.1 ± 0.2 s.e.).
Figure 1. The per cent of offspring (pooled across all 36 sampled groups) belonging to the dominant male or female (black), and to subordinate males and females in the social group (light grey). The per cent of all instances of extra-group reproduction are shown in dark grey and represent, a combination of reproduction assigned to known individuals outside the sampled group and reproduction that was unable to be assigned.
The dominant females mothered 81.6% of offspring (324 out of 397 offspring genotyped). The dominant female collected with the focal group was the mother of all offspring in 18 of the social groups, the majority of offspring in 16 groups (mean maternity ± s.e: 75.2% ± 5.4), and no offspring in two groups. Of the 73 offspring not assigned to the dominant female, 23 were assigned to a subordinate female in the same group (mean 4.7% ± 1.6 s.e.), two were assigned to a dominant female from another sampled group, and 48 were not assigned to any female sampled (figure 1), meaning that these offspring were probably mothered by females in groups that we did not sample or by females within the group that were recently evicted or eaten by a predator. The number of mothers per brood ranged from 1 to 4 (mean: 1.6 ± 0.1 s.e.). The degree of maternity loss across social groups was not correlated to the rate of paternity loss (LME: F1,28 = 0.19, p = 0.85).
(b) Subordinate reproduction
Subordinate males reproduced in almost a third of the sampled groups (11 out of 36) and subordinate females reproduced in a quarter of groups (9 out of 36). Of the reproductively mature subordinates genotyped, 35% of male subordinates (13 out of 37 males) and 19% of female subordinates (9 out of 47 females) achieved some degree of reproductive success. Subordinate females as small as 36 mm SL reproduced and males as small as 30 mm SL reproduced, which for males is below the average size previously reported for sexual maturation . The likelihood of a male subordinate reproducing was not influenced by his body size relative to the dominant male ([SL dominant − SL subordinate]/SL dominant generalized linear models GLM with binomial error distribution: Z24 = 0.95, p = 0.34) however, female subordinates were more likely to reproduce if they were close in size to the dominant female (GLM: Z35 = −2.22, p = 0.03). There was no influence of relative gonad size (gonad mass relative to body mass, GSI) on whether a subordinate of either sex reproduced (GLM: males: Z24 = 1.06, p = 0.29 females: Z35 = −0.82, p = 0.41).
(c) Group and colony-level factors related to parentage loss
Dominant males lost significantly more paternity and subordinate males gained significantly more paternity in larger groups, but this occurred only in territories on the edge of the colony (table 1: location by group size interaction). There was no strong influence of group size on dominant and subordinate paternity in groups in the centre of the colony (table 1). Subordinate males also gained more paternity in higher density areas and in groups with more males (table 1). The number of fathers per brood increased with the density of neighbouring groups and was higher on the edge of the colony compared with the centre (table 1). Groups in high-density areas and on the edge of the colony also had a higher number of offspring that could not be assigned to any collected male group member (GLMER density: Z29 = 1.96, p = 0.05 location: Z29 = 2.97, p = 0.003), suggesting that neighbouring males sired a larger proportion of offspring in those areas. Although gonadal investment (GSI) of dominant males did not vary with territory density, males on edge territories had relatively larger gonads than males with territories in the centre of the colony (LME location: F1,33 = 4.49, p = 0.04 density: F1,33 = 1.39, p = 0.25).
Table 1. Allocation of paternity within groups. (Test statistics (with d.f.) and p-values from full models measuring the effect estimates of location (centre versus edge), density of neighbouring groups, group size, relative size of dominant male to dominant female ([dominant male SL − dominant female SL]/dominant male SL), and the number of male subordinates on the number of offspring assigned to dominant and subordinate males within the sampled group (GLMER) and number of fathers contributing to the brood (CLM). Total number of offspring within each brood was included as a fixed effect to control for variation in brood size. Asterisks indicate significance, assessed at α = 0.05.)
On the edge of the colony, groups with many subordinate females had dominant females that lost more maternity, subordinate females that gained more maternity and more mothers per brood compared with edge groups with fewer female subordinates (figure 2a and table 2: location by female subordinate interaction). By contrast, centre groups with a high number of subordinate females had dominant females that retained more maternity, subordinate females that gained less maternity, and fewer mothers per brood compared with centre groups with fewer female subordinates (figure 2b and table 2). Interestingly, dominant females who were much smaller than the dominant male retained a larger portion of the maternity of their brood compared with dominant females who were close in size to the dominant male (table 2). This result was driven by differences in dominant male size, as females paired with larger dominant males suffered less loss of maternity (figure 3 GLMM Z21 = 2.05, p = 0.04).
Table 2. Allocation of maternity within groups. Test statistics (with d.f.) and p-values from full models measuring the effect estimates of location (centre versus edge), density of neighbouring groups, group size, relative size of dominant male to dominant female ([dominant male SL − dominant female SL]/dominant male SL), and the number of female subordinates on the number of offspring assigned to dominant and subordinate females within the sampled group (GLMER) and the total number of mothers contributing to the brood (CLM). Total number of offspring within each brood was included as a fixed effect to control for variation in brood size. Asterisk indicate significance, assessed at α = 0.05.
Figure 2. (a) In the 12 groups on the edge of the colony, the number of offspring mothered by the dominant female decreased as the number of female subordinates in the group increased (GLMER: Z6 = −2.97, p = 0.003). (b) By contrast, in the 24 groups in the centre of the colony, the number of offspring mothered by the dominant female increased as the number of female subordinates in the group increased (GLMER: Z16 = 3.38, p < 0.001).
Figure 3. Across all 36 sampled groups, dominant females mothered more offspring (after controlling for variation in brood size) when they were mated with a large male, both in terms of absolute size (SL, as shown in this figure) and in terms of size relative to the dominant male (table 2). Data presented are least squared means of the regression model testing predictors of the number of offspring mothered by the dominant female.
EPP and EPM were common in this study, accounting for 24% and 18% of the total number of sampled offspring, respectively. For both males and females, extra-pair offspring were sired by subordinates within the group as well as by individuals from neighbouring groups. Both within-group characteristics, including the size of the dominant male, and between-group characteristics, such as the density of neighbouring groups, significantly influenced reproductive dynamics within groups. Further, the effect of some within-group characteristics, such as group size, varied with colony-level characteristics. Collectively, these results suggest that understanding of the reproductive dynamics within colonial groups requires comprehension of both within- and between-group social dynamics.
This study provides strong evidence of extensive subordinate reproduction in wild groups of this well-studied fish species. We uncovered high levels of subordinate reproduction, with 35% of collected male subordinates and 19% of collected female subordinates achieving some degree of reproductive success. Previous studies based on much smaller sample sizes [29,40] found no occurrences of subordinate male reproduction and only one occurrence of subordinate female reproduction in the wild, leading to the mistaken conclusion that subordinate reproduction is rare in free-living N. pulcher groups.
High levels of subordinate reproduction in a cooperatively breeding system call into question current explanations for the fitness benefits of helping in these fish. If subordinates obtain high levels of reproduction, what is perceived to be subordinate alloparental care may in fact represent direct parental care . Further, if subordinates reproduce within their own groups, then lost reproductive opportunities resulting from delayed dispersal, considered to be one of the main costs of helping , may have previously been overstated. Our study demonstrates that the evolution of apparent helping behaviour in N. pulcher may be at least partly explained by direct reproductive benefits. Although some cooperatively breeding species have higher levels of subordinate reproduction than those reported in this species, this degree of subordinate reproduction is rare for cooperative breeders, as subordinates in many cooperatively breeding species are reproductively suppressed, either behaviourally or physiologically, by dominants . In birds, the evolution of cooperative breeding is associated with lower rates of extra-group parentage [42,43], but cooperative breeding can be maintained even in highly promiscuous species if subordinates receive direct benefits while helping in the group . Given the high level of extra-group parentage found in this study, some baseline subordinate reproduction may be necessary for assuring continued subordinate help and membership in the group .
The social environment can also determine the level of within-group subordinate reproduction by altering the ability and motivation of subordinates to reproduce [20,46]. On the colony edge, the portion of the brood mothered by subordinate females increased as the number of female subordinates within the group increased however, in the colony centre, groups with many female subordinates actually had comparatively less subordinate female reproduction. Similarly, subordinate males reproduced more frequently in larger groups, but only on the edge of the colony. Compared with dominants in the colony centre, dominants on the edge of the colony may be less capable of suppressing subordinate reproduction or they may concede more reproduction to subordinates because they are more reliant on subordinates for defence against predators .
Alternatively, subordinates in central territories may choose not to reproduce, as getting evicted following reproduction  would result in losing the opportunity to inherit a high-quality territory in the centre of the colony. This would be particularly costly for subordinates in large groups, which are more productive and more likely to persist from year to year [47,48]. However, if there are lower fitness benefits to inheriting low-quality territories on the colony edge , the benefits of reproducing for subordinates may outweigh the costs of getting evicted due to reproduction. Similarly, subordinate males may reproduce more frequently in denser areas because eviction may be less costly when there are a high number of potential groups to join. Further, increased reproduction by neighbouring males may mean that subordinates on the colony edge and in denser areas are less related to the offspring in their group compared with subordinates in the centre or in less dense areas. Therefore, subordinates in the centre or less dense areas may gain greater indirect fitness benefits, future fitness benefits, or both, by providing alloparental care, compared with subordinates on the edge or in denser areas.
In addition to high rates of subordinate reproduction, we found evidence of EPM in about half of the sampled groups. Dominant females paired with large males mothered a larger portion of offspring compared with dominant females paired with smaller males. Across a variety of taxa, larger and older males are higher quality mates, have more reproductive success and are more likely to be extra-pair mates [15,49–51]. Dominant females paired with smaller males may be more likely to pursue mating opportunities with higher quality males in neighbouring groups and spend less time guarding against reproductive competitors . Large dominant males may also be more successful at cuckholding neighbouring males and mating with females in neighbouring territories. They may choose this strategy, rather than court females in their own territory, because reproduction of multiple females within a territory can increase within-group conflict [52,53], and siring offspring on other territories allows males to parasitize the care and resources of other individuals .
On the edge of the colony, a higher proportion of collected offspring were sired by males other than the collected dominant or subordinate males, suggesting that neighbouring males steal more paternity from a given resident male on the edge of the colony compared with the centre. Resident males on the edge of the colony may have a reduced ability to prevent other males from mating with their social mate, either due to decreased quality or to increased time spent vigilant against predators [11–14] however, edge males may also adopt different mating strategies. In wild guppy populations (Poecilia reticulata), EPP is much more common in high-predation environments compared with low-predation environments . Experiments manipulating predation pressure in guppies found that, under perceived high-predation pressure, males reduce their courtship displays and sneak copulations more often , while females become less selective regarding potential mates . This is probably because courtship displays and mating selection tend to attract predators . This may also explain why males on the edge of the colony had the largest testes edge males may invest substantially more in sneaking paternity in neighbouring nests compared with males in the centre.
Here, we provide support that both within- and between-group factors are linked to reproductive patterns in N. pulcher. Further, we found that the influence of within-group characteristics may depend upon colony-level social structure. Our results demonstrate that studies of reproductive dynamics within groups should consider the influence of neighbouring groups, which offer additional reproductive opportunities and can alter how individuals pursue reproductive opportunities within their own group. Our results suggest that individuals occupying lower quality edge territories may adopt different reproductive strategies to accommodate increased predation risk. Further, subordinates may attempt to reproduce when living in lower quality edge territories to compensate for lowered kin-selected benefits of alloparental care or the costs of living in a suboptimal habitat. We suggest that the patterns of extra-pair parentage and the selective pressures governing the pursuit of extra-pair parentage may vary over relatively small spatial and social scales. Further research exploring behavioural and reproductive exchanges within and between groups would be useful for elucidating the extent to which individuals vary their decisions based on the presence of neighbouring groups.
Neolamprologus pulcher is neither endangered nor threatened. The physical territories were minimally disturbed during collection and new groups occupied the vacated territories within 1–2 days post-collection. All methods, including euthanasia techniques, were approved by the Zambian Department of Fisheries, the Ohio State University IACUC (protocol ID 2008A0095) and the Animal Research Ethics Board of McMaster University (Animal Utilization Protocol no. 10-11-71). Our procedures adhered to the guidelines of the Canadian Council for Animal Care, and the Animal Behavior Society.
All data presented in this manuscript are currently deposited in Dryad (http://dx.doi.org/10.5061/dryad.rv5mg).
J.K.H., I.Y.L., C.M.O., A.R.R., S.B. and I.M.H. assisted with study design and conception. J.K.H. collected field data, conducted the parentage and statistical analysis and wrote the majority of the manuscript. I.Y.L., C.M.O., A.R.R. and S.E.M. aided with coordination of the field season, data collection and revision of the manuscript. K.A.G. collected field data. H.L.G. aided with parentage analysis and manuscript revisions. S.B. and I.M.H. assisted with statistical analysis, facilitation of the field season and manuscript writing.
We have no competing interests.
This research was supported by NSERC Discovery grant, an Ontario Innovation Trust and a Canadian Foundation for Innovation grant to S.B., a Journal of Experimental Biology Travelling Fellowship to C.M.O., and McMaster School of Graduate Studies and Canadian Society of Zoologists research grants to A.R.R. C.M.O. was supported by an E.B. Eastburn Postdoctoral Fellowship. J.K.H. and I.Y.L. were supported by T he Ohio State University Fish Systematics Endowment, and the SciFund Challenge . J.K.H. was supported by the American Academy of Underwater Sciences and Animal Behavior Society . A.R.R. was supported by a Margo Wilson and Martin Daly Ontario Graduate Scholarship. S.B. is supported by the Canada Research Chairs Program.
Doublesex and sex differentiation in A. gambiae
Sex differentiation in insects follows a common pattern in which a primary signal activates a central gene that induces a cascade of molecular mechanisms that control alternative splicing of the doublesex (dsx) gene 13,14 . Although the molecular mechanisms and the genes involved in regulating sex differentiation in A. gambiae are not well understood, except that Yob1 functions as a Y-linked male determining factor 15 , available data indicated an important role of dsx in determining sexual dimorphism in this mosquito species 16 . In A. gambiae, dsx (Agdsx) consists of seven exons, distributed over an 85-kb region on chromosome 2R, a gene structure similar to that of Drosophila melanogaster dsx (Dmdsx) and other insect orthologs, and is alternately spliced to produce the female and male transcripts AgdsxF and AgdsxM, respectively. The female transcript consists of a 5′ segment common with that of males, a highly conserved female-specific exon (exon 5) and a 3′ common region, while the male transcript comprises only the 5′ and 3′ common segments. The male-specific isoform contains an additional domain at the C terminus that is transcribed as a noncoding 3′ untranslated region in females (Fig. 1a).
(a) Schematic representation of the male- and female-specific dsx transcripts and the gRNA sequence used to target the gene (shaded in gray). The gRNA spans the intron 4–exon 5 boundary. The protospacer-adjacent motif (PAM) of the gRNA is highlighted in blue. Scale bar, 200 bp. Coding regions of exons (CDS) are shaded in black, noncoding regions in white. Introns are not drawn to scale. UTR, untranslated region. (b) Sequence alignment of the dsx intron 4–exon 5 boundary in six of the species from the A. gambiae complex. The sequence is highly conserved within the complex suggesting tight functional constraint at this region of the dsx gene. The gRNA used to target the gene is underlined and the protospacer-adjacent motif is highlighted in blue. (c) Schematic representation of the HDR knockout construct specifically recognizing exon 5 and the corresponding target locus. DSB, double-strand break. (d) Diagnostic PCR using a primer set (blue arrows in c) to discriminate between the wild-type and dsxF allele in homozygous (dsxF −/− ), heterozygous (dsxF +/− ) and wild-type (dsxF +/+ ) individuals.
To investigate whether dsx is a suitable target for a gene drive to suppress population reproductive capacity, we disrupted the intron 4–exon 5 boundary of dsx (Fig. 1b) to prevent the formation of functional AgdsxF while leaving the AgdsxM transcript unaffected. We injected A. gambiae embryos with a source of Cas9 and a single-guide RNA (sgRNA) designed to recognize and cleave a sequence overlapping the intron 4–exon 5 boundary, in combination with a template for homology-directed repair (HDR) to insert an eGFP transcription unit (Fig. 1c). Transformed individuals were intercrossed to generate homozygous and heterozygous mutants among the progeny. HDR-mediated integration was confirmed with diagnostic PCR using primers that spanned the insertion site: a large amplicon for the HDR event and a smaller amplicon for the wild-type allele enabled facile confirmation of genotypes (Fig. 1d). The knock-in of eGFP resulted in the complete disruption of the exon 5 (dsxF − ) coding sequence and was confirmed by PCR and genomic sequencing of the chromosomal integration (Supplementary Fig. 1). Crosses of heterozygous individuals produced wild-type, heterozygous and homozygous individuals for the dsxF − allele at the expected Mendelian ratio 1:2:1, indicating that there was no obvious lethality associated with the mutation during development (Supplementary Table 2). Larvae heterozygous for the exon 5 disruption developed into adult male and female mosquitoes with a sex ratio close to 1:1. However, half of dsxF −/− individuals developed into normal males whereas the other half had both male and female morphological features, as well as a number of developmental anomalies in the internal and external reproductive organs (intersex phenotype). To establish the sex genotype of these dsxF −/− intersex mosquitoes, we introgressed the mutation into a line containing a Y-linked visible marker (RFP) and used the presence of this marker to unambiguously assign sex genotype among individuals heterozygous and homozygous for the null mutation. This approach revealed that the intersex phenotype was observed only in females that were homozygous for the null mutation. We saw no phenotype in heterozygous mutants, suggesting that the female-specific isoform of dsx is haplosufficient. Examination of external sexually dimorphic structures in dsxF −/− genotypic females (n > 50) showed several phenotypic abnormalities, including the development of dorsally rotated male claspers (and absence of female cerci) and longer flagellomeres associated with male-like plumose antennae (Fig. 2 and Supplementary Table 3). Analyses of the internal reproductive organs of the same set of insects revealed the absence of fully developed ovaries and spermathecae instead these were replaced with male accessory glands and in some cases ( ∼ 20%) by rudimentary pear-shaped organs resembling unstructured testes (Supplementary Fig. 2). Males carrying the dsxF − null mutation in heterozygosity or homozygosity showed wild-type levels of fertility as measured by clutch size and larval hatching per mated female, as did heterozygous dsxF − female mosquitoes. Intersex XX dsxF −/− female mosquitoes, although attracted to anesthetized mice, were unable to take a blood meal and failed to produce any eggs (Fig. 3). The drastic phenotype of dsxF −/− in females indicates that exon 5 of dsx has a fundamental role in the previously poorly understood sex differentiation pathway of A. gambiae mosquitoes and suggested that its sequence might represent a suitable target for gene drives designed for population suppression.
(a) Morphological appearance of genetic males and females heterozygous (dsxF +/− ) or homozygous (dsxF −/− ) for the exon 5 null allele. This assay was performed in a strain containing a dominant RFP marker linked to the Y chromosome, whose presence permits unambiguous determination of male or female genotype. Anomalies in sexual morphology were observed only in dsxF −/− genetic female mosquitoes. This group of XX individuals showed male-specific traits, including a plumose antenna (red arrowhead) and claspers (blue arrowheads). This group also showed anomalies in the proboscis and accordingly they could not bite and feed on blood. Representative samples of each genotype are shown. (b) Magnification of the external genitalia. All dsxF −/− females carried claspers, a male-specific characteristic. The claspers were dorsally rotated rather than in the normal ventral position.
Male and female dsxF −/− and dsxF +/− individuals were mated with the corresponding wild-type sexes. Females were given access to a blood meal and subsequently allowed to lay individually. Fecundity was investigated by counting the number of larval progeny per lay (n ≥ 43). Using wild type (wt) as a comparator, we saw no significant differences ('ns') in any genotype other than dsxF −/− females, which were unable to feed on blood and therefore failed to produce a single egg (****P < 0.0001 Kruskal–Wallis test). Vertical bars indicate the mean and the s.e.m. Blue and red indicate the crosses of male or female dsxF mutants, respectively, to wild type, whereas the gray dots represent wild-type-only crosses.
Building a gene drive to target dsx
We used recombinase-mediated cassette exchange to replace the 3xP3::GFP transcription unit with a dsxF CRISPRh gene drive construct that comprised an RFP marker gene, a transcription unit to express the guide RNA (gRNA) targeting dsxF, and cas9 under the control of the germline promoter of zero population growth (zpg) and its terminator sequence (Fig. 4a and Supplementary Fig. 3). The zpg promoter has improved germline restriction of expression, resulting in increased female fertility compared with the vasa promoter used in previous gene drive constructs 17 (Supplementary Fig. 4). Successful cassette exchange events that incorporated dsxF CRISPRh into the target locus were confirmed in those individuals that had swapped the GFP for the RFP marker (n = 17 G1 individuals) (Supplementary Fig. 3). During meiosis the Cas9–gRNA complex cleaves the wild-type allele at the target sequence and the dsxF CRISPRh cassette is copied into the wild-type locus by HDR ('homing'), disrupting exon 5 in the process. The ability of the dsxF CRISPRh construct to home and bypass Mendelian inheritance was analyzed by scoring the rates of RFP inheritance in the progeny of heterozygous parents (referred to as dsxF CRISPRh /+ hereafter) crossed to wild-type mosquitoes. High dsxF CRISPRh transmission rates were observed in the progeny of both heterozygous dsxF CRISPRh /+ male (95.9% ± 1.1% s.e.m. n = 87) and female mosquitoes (99.4% ± 0.5% n = 33) (Fig. 4b). The fertility of the dsxF CRISPRh line was also assessed to unravel potential negative effects due to ectopic expression of the nuclease in somatic cells and/or parental deposition of the nuclease into the newly fertilized embryos (Fig. 4c). These experiments showed that while heterozygous dsxF CRISPRh /+ males showed a fecundity rate (assessed as larval progeny per fertilized female) that did not differ from that of wild-type males, heterozygous dsxF CRISPRh /+ females had reduced fecundity overall (mean fecundity 49.8% ± 6.3% s.e.m., P < 0.0001). We noticed a greater reduction in the fertility of heterozygous females when the drive allele was inherited from the father (mean fecundity 21.7% ± 8.6% P < 0.0001) (n = 15) rather than the mother (64.9% ± 6.9% P < 0.001) (n = 28) (Supplementary Fig. 5). This could be explained by assuming a paternal deposition of active Cas9 nuclease into the newly fertilized zygote that stochastically induces conversion of dsx to dsxF − , either through end-joining or HDR, in a substantial number of embryonic cells, which in females results in a reduced fertility. Consistent with this hypothesis, some heterozygous females (9 of 31 examined) receiving a paternal dsxF CRISPRh allele showed a somatic mosaic phenotype that included, with varying penetrance, the absence of spermatheca and/or the formation of an incomplete clasper set (Supplementary Fig. 2c).
(a–c) Male and female mosquitoes heterozygous for the dsxF CRISPRh allele (a) were analyzed in crosses with wild-type mosquitoes to assess the inheritance bias of the dsxF CRISPRh drive construct (b) and for the effect of the construct on their reproductive phenotype (c). (b) Scatter plot of the transgenic rate observed in the progeny of dsxF CRISPRh /+ female or male mosquitoes that gave progeny when crossed to wild-type individuals (n ≥ 33). Each dot represents the progeny derived from a single female. Both male and female dsxF CRISPRh /+ showed a high transmission rate of up to 100% of the dsxF CRISPRh allele to the progeny. The transmission rate was determined by visually scoring offspring for the RFP marker that is linked to the dsxF CRISPRh allele. The dotted line indicates the expected Mendelian inheritance. Mean transmission rate (± s.e.m.) is shown. (c) Scatter plot showing the number of larvae produced by single females (n ≥ 35) from crosses of dsxF CRISPRh /+ mosquitoes with wild-type individuals after one blood meal. Mean progeny count (± s.e.m.) is shown (****P < 0.0001 Kruskal–Wallis test).
Assessment of dsx gene drive in caged insects
Using a mathematical model that includes the inheritance bias of the construct, the fecundity of heterozygous individuals, the phenotype of intersex, and the effect of the paternal deposition of the nuclease on female fertility (Online Methods), we found that the dsxF CRISPRh had the potential to reach 100% frequency in caged population in 9–13 generations considering a starting allele frequency of 12.5% and stochasticity (Fig. 5a). To test this hypothesis, we mixed caged wild-type mosquito populations with heterozygous individuals carrying the dsxF CRISPRh allele and monitored progeny at each generation to assess the spread of the drive and to quantify effect(s) on reproductive output. We started the experiment in two replicate cages, each with an initial drive allele frequency of 12.5% (300 wild-type female mosquitoes with 150 wild-type male mosquitoes and 150 dsxF CRISPRh /+ male individuals). The initial drive allele frequency that we selected minimizes the stochastic loss of the drive (Supplementary Fig. 6) and represents a realistic field release scenario, being severalfold lower than that used in non-invasive genetic control strategies 18 . All of the eggs produced by the entire cage population were counted, and then 650 eggs were randomly selected to seed the next generations. The larvae that hatched from the eggs were counted and screened for the presence of the RFP marker to score the number of the progeny containing the dsxF CRISPRh allele in each generation.
Two cages were set up with a starting population of 300 wild-type females, 150 wild-type males and 150 dsxF CRISPRh /+ males, seeding each cage with a dsxF CRISPRh allele frequency of 12.5%. (a) The frequency of dsxF CRISPRh mosquitoes was scored for each generation. The drive allele reached 100% prevalence in both cage 2 (blue) and cage 1 (red) at generation 7 and 11, respectively, in agreement with a deterministic model (black line) that takes into account the parameter values retrieved from the fecundity assays. Twenty stochastic simulations were run (gray lines) assuming a maximum population size of 650 individuals. (b) Total egg output deriving from each generation of the cage was measured and normalized relative to the output from the starting generation. Suppression of the reproductive output of each cage led the population to collapse completely (black arrows) by generation 8 (cage 2) or generation 12 (cage 1). Parameter estimates included in the model are provided in Supplementary Table 5.
During the first three generations we observed an increase of the drive allele from 25% to ∼ 69% in both caged populations, but at generation 4 the outcomes in the two cages diverged. In cage 2 the drive reached 100% frequency by generation 7 in generation 8, no eggs were produced and the population collapsed. In cage 1 the drive allele reached 100% frequency at generation 11 after remaining at around 65–70% for generations 4 through 8. In generation 12 the cage 1 population also failed to produce eggs (Fig. 5b). While the dynamics of spread of the gene drive in the two caged populations was different, both sets of finding fall within the prediction range of our mathematical model (Fig. 5).
Potential for resistance to dsx gene drive
We monitored the occurrence of mutations at the drive target site in generations 2, 3, 4 and 5 to identify the occurrence of nuclease-resistant, functional variants. Amplicon sequencing of the target sequence from pooled samples containing a minimum of 359 mosquitoes, which were collected in generations 2–5, revealed several low-frequency indels present at the target site (up to 1.16% frequency among nondrive alleles), none of which appeared to encode a functional AgdsxF transcript (Supplementary Fig. 7). In addition, none of the variants identified showed any signs of positive selection, which would be expected to cause them to increase in frequency as the drive progressively increased in frequency over generations, suggesting that the selected target sequence has rigid functional or structural constraints. This hypothesis is supported by the exceptionally high conservation of exon 5 in A. gambiae mosquitoes 19,20 and the presence of a strictly regulated splice site that is crucial in mosquito reproductive biology. Furthermore, large-scale resequencing of 765 wild-caught mosquitoes from eight sub-Saharan African countries 20 revealed only a single rare SNP within the drive target site, present at 2.9% frequency (Supplementary Fig. 8). This naturally occurring variant could block the spread of the drive. To investigate this hypothesis, we tested whether this SNP variant was as susceptible to cleavage in vitro by Cas9 as the wild-type sequence, using the sgRNA from our gene drive construct. We found that the gRNA in our gene drive construct efficiently cleaved both the wild-type and the SNP sequence variant, which may indicate that our gene drive would be able to spread even if this conserved SNP was present (Supplementary Fig. 9). However, it is important to note that we cannot state that our drive target site is 'resistance-proof', since at scale, and over time, it is possible that nuclease-induced mutations could be produced that do restore sufficient function to the gene to be positively selected. This notwithstanding, targeting gene drives to functionally constrained sequences is clearly advantageous, as evidenced by the population collapse effected by this gene drive in both caged mosquito populations. Distinct, highly conserved sequences may have varying levels of functional constraint, and the relative strength of selection for maintaining sequence conservation versus the strength of selection imposed by the gene drive will ultimately determine their suitability as targets for gene drives.
Our data not only provide important functional insights into the role of dsx in A. gambiae sex determination, but also represent a substantial step toward the development of effective gene drive vector-control measures that aim to suppress insect populations. The intersex phenotype of dsxF −/− genetic females shows that exon 5 is crucial for the production of a functional female transcript, as was initially hypothesized on the basis of the expression profile of the dsx splice variants in the two sexes 16 . Furthermore, the observation that heterozygous dsxF CRISPRh /+ females are fertile and produce almost 100% inheritance of the drive might indicate that most of the germ cells in these females are homozygous and, unlike somatic cells, do not undergo autonomous dsx-mediated sex commitment 21 .
An Introduction to Honey Bee Biology
Honey bee colonies contain three distinct types, or castes, of individuals. Each hive contains a single female queen, tens of thousands of female workers, and anywhere from several hundred to several thousand male drones during the spring and summer months.
Honey Bee Castes: Queen Bees
Queen bees are the largest individuals in most colonies and carry out many important functions in the hive. The queen is responsible for laying a constant supply of eggs to build up and maintain the hive’s population at adequate numbers. In a good year, a queen may lay as many as 200,000 eggs!
The queen also produces chemicals called pheromones that control and organize many of the behaviors of her colony. Each queen has her own distinct pheromone profile, which allows her colony to recognize her, defend her and meet her needs to keep the hive safe and strong.
Honey Bee Castes: Worker Bees
Worker bees are by far the most numerous caste in hives and, as their name implies, carry out all of the work needed to keep the colony fed and healthy. During their first days as mature adults, workers tend to perform tasks inside the hive, such as cleaning and capping cells.
As they mature, worker bees begin to perform more tasks inside the hive, including feeding the queen and developing brood, drawing out new comb, and managing food stores. The oldest and most experienced workers tend to perform the most dangerous chores: guarding the hive against intruders and foraging outside the hive for pollen and nectar.
Honey Bee Castes: Drone Bees
The only males found in the hive, drones perform only one task during their lifetime: mating with new queens. When a drone reaches sexual maturity at about two weeks of age, he begins taking mating flights. These flights usually take place in spring and summer afternoons and last about 30 minutes.
Newly matured queens and drones from several hives typically join in these flights. In most instances, the queens mate with multiple drones and store the drones’ sperm in an organ called the spermatheca. The queen will then use this stored genetic material to fertilize her eggs for the rest of her life.
A cultural culprit
It’s not unprecedented for human genetic diversity to take a nosedive once in a while, but the Y-chromosome bottleneck, which was inferred from genetic patterns in modern humans, was an odd one. First, it was observed only in men – more precisely, it was detected only through genes on the Y chromosome, which fathers pass to their sons. Second, the bottleneck is much more recent than other biologically similar events, hinting that its origins might have something to do with changing social structures.
Certainly, the researchers point out, social structures were changing. After the onset of farming and herding around 12,000 years ago, societies grew increasingly organized around extended kinship groups, many of them patrilineal clans – a cultural fact with potentially significant biological consequences. The key is how clan members are related to each other. While women may have married into a clan, men in such clans are all related through male ancestors and therefore tend to have the same Y chromosomes. From the point of view of those chromosomes at least, it’s almost as if everyone in a clan has the same father.
That only applies within one clan, however, and there could still be considerable variation between clans. To explain why even between-clan variation might have declined during the bottleneck, the researchers hypothesized that wars, if they repeatedly wiped out entire clans over time, would also wipe out a good many male lineages and their unique Y chromosomes in the process.
INFORMATION SHEET 3
The following is a discussion of the members of a honey bee colony, their development and their duties within the colony.
The vast majority of adult honey bees in any colony are female worker bees. The jobs of the worker bees are: tending and feeding young bees (larvae), making honey, making royal jelly and beebread to feed larvae, producing wax, cooling the hive by fanning wings, gathering and storing pollen, nectar and water, guarding the hive, building, cleaning and repairing the comb, and feeding and taking care of the queen and drones. In part, the job the worker honey bee performs on any given day depends on its age.
As insects, honey bees pass through four distinct life stages: the egg , larva , pupa and adult . The process is called complete metamorphosis, which means that the form of the bee changes drastically from the larva to the adult. Passing through the immature stages takes 21 days for worker bees. On the first day, the queen bee lays a single egg in each cell of the comb. The egg generally hatches into a larva on the fourth day. The larva is a legless grub that resembles a tiny white sausage. The larva is fed a mixture of pollen and nectar called beebread . On the ninth day the cell is capped with wax and the larva transfor ms into the pupa. The pupa is a physical transition stage between the amorphous larva and the hairy, winged adult. The pupa doesn't eat. On day 21, the new adult worker bee emerges.
The male members of the colony, the drones , are somewhat larger and make up only about five percent of the hive population. Drones are fed royal jelly , and develop in a slightly larger cell than worker bees from unfertilized eggs. Drones remain in the pupal stage for 15 days, so they don't emerge until day 24. Drones have huge compound eyes that meet at the top of their head and an extra segment in their antennae. In comparison to worker bees, drones have wider bodies and their abdomens are rounded rather than pointed. Drones, like all other male bees and wasps, do not have stingers.
There is only one queen in a honey bee colony. She is slightly larger than a worker bee, with a longer abdomen. She does not have pollen baskets on her legs. Eggs destined to become queens are laid in a larger cell, and the larvae are fed only royal jelly. The adult queen's sole duty is to lay eggs, up to 2,000 a day! She is fed by the workers and never leaves the hive except to mate.
Queen bees also have stingers and use them in battles with each other for dominance of the colony. If a new queen emerges from her incubation cell and is detected by the current queen, the "old lady" often goes over and kills her rival. In this way, the stability of the colony is maintained. When a queen gets old or weak and slows her production of queen substance, she is generally replaced by a new queen. New queens are also produced in colonies about to swarm.
Virgin queen bees take what is known as a " nuptial flight " sometime within the first week or two after emerging from the pupal chamber. The new queen flies out of the hive and begins to produce a perfume-like substance called a " pheromone ." The drones in the area are attracted to the pheromone and the queen will mate with as many as 20 of them. After mating, the drones die.
Once the queen has mated, she heads back to the hive to start laying eggs in beeswax chambers that the workers have created especially for this purpose. A queen can lay her own weight in eggs every day and, since she can maintain the sperm she has collected for her lifetime in a special pouch in her body, she can continue laying eggs indefinitely. The fertilized eggs laid by a queen become female worker bees and new queens. The queen also lays some unfertilized eggs, which produce the drones. Since they come from unfertilized eggs, the drones carry only the chromosomes of the queen.
The drones could be called the couch potatoes of the insect world. While they wait for an opportunity to mate with a virgin queen, they are fed and cared for by workers, and only occasionally fly out of the hive to test their wings. If no opportunity to mate arises by fall, the drones are ejected from the nest by the workers and left to fend for themselves.
On average, queen bees live for about a year-and-a-half, although some have been known to survive for up to six years. While she is alive and active, the queens are constantly cared for by workers acting as attendants. In cases where a queen dies prematurely and the colony had no new queen to replace her, some worker bees develop the ability to lay eggs but, because they cannot mate, they produce only drones and the colony eventually perishes.
When the colony starts to become too crowded, some of the bees split off to form a new colony. This is called swarming . First the eggs for new queens are laid in their special larger cells. "Swarming" occurs when part of the colony breaks off with the old queen and flies off looking for another place to call home. The bees engorge themselves on their honey reserves before leaving so as to have sufficient energy to make it to a new location. There can be multiple swarms from one hive, since new queens can also emerge and fly off with part of the worker force.