Keller L., Gordon E. The Lives of Ants

Подождите немного. Документ загружается.

In all of this, the males are once again the losers. If they are to

pass on copies of their genes to posterity, their only hope is to

ﬁnd one of the rare queens hatched from a fertilized egg, a thing

which can happen, given that there are always exceptions to the

rule and about 10 per cent of the female reproducers are not

born via parthenogenesis. Even so, the males’ reproduction rate

is pretty low. The only way they can improve it is if the queen

dies, because then the workers make haste to replace her by

producing young queens. Being unmated, they can of course

only do this via parthenogenesis; but that is immaterial, since

they are passing on their paternal genes to the future queens,

who will transmit them in their turn to their own progeny. All

things considered, the males hand on ver y little of themselves to

posterity and appear to be duped all along the line.

Double cloning

Obviously, for Cataglyphis cursor males, it ’s a mug’s game: the

queen wins hands down and, even if the males do on occasion

manage to pass on their genes, overall their reproductive success

is very poor. However, there is a species, the little ﬁre ant, in

which the queen produces her royal lineage by cloning but the

male reproducers do not let her get away with exploiting them.

They have hit on a smart solution: to prevent the queens from

monopolizing reproduction for their sole beneﬁt, the fathers

turn out sons in their own image—that is, they too have recourse

to cloning. This is unique in the animal kingdom, where such a

mode of reproduction is almost unheard of; and when it does

occur, it appears to be the sole preserve of females.

It was by the greatest of ﬂukes that we discovered this utterly

atypical behaviour, our real focus having been on the invasive-

ness of the little ﬁre ant, Wasmannia auropunctata. Our object was

to ﬁnd out the history of introduced populations of this polygyn-

188

THE LIVES OF ANTS

ous ant, in which new colonies arise by budding, a process

whereby young queens establish their new colony not far from

the nest where they were born and from which they then poach

workers.

Working with Denis Fournier and Arnaud Estoup of the

Institut national de la recherche agronomique (INRA) at

Montferrier-sur-Lez, France, and with colleagues from the Uni-

versity of Toulouse III and a laboratory at the Institut de

recherche pour le de

veloppement (IRD) in New Caledonia, we

undertook a study of Wasmannia auropunctata ants collected in

French Guyana. This was how we discovered that, in any given

population, not only were there great genetic differences be-

tween the queens and the workers, but that all the queens had

the same genotype. On closer inspection, we realized that the

queens also reproduced asexually. However, unlike Cataglyphis

cursor, they do it not by parthenogenesis but by cloning. The

difference lies in the fact that in parthenogenesis the maternal

alleles can recombine. The genes of the queen exist in fact in two

forms, A and B, say, which means that the daughter may receive

different copies, AA, BB, or AB. Because of this, mother and

daughter do not have identical genomes. And there lies the

difference with cloned reproduction, which always reproduces

a daughter who is an exact replica of the mother.

All this was surprising enough; but the ants’ genes had many

more surprises in store for us. As we analysed the genomes of the

workers, we realized that they had all inherited the same paternal

genes. Could this possibly mean they all had the same father?

That would imply that a single male had been smart and proliﬁc

enough to have contrived to mate with every single queen. This

seemed too implausible to be true, even though the nests were

quite close to each other. The only other possible explanation

was to suppose that all the males in the species had the same

genotype. To test this hypothesis, we conducted genetic analyses

of all the individuals produced in colonies where we knew the

189

ANYTHING GOES

genotype of the queen and her mate. These analyses unambigu-

ously conﬁrmed that only the workers are produced by ‘normal’

sexual reproduction. All the new queens were genetically identi-

cal to their mothers, while young males were similar to their

fathers (see Figure 4).

queen

male

future queen worker male

r=0.5 r=1r=0.5r=1

Figure 4 Cloning In the ﬁre ant Wasmannia auropunctata, reproduction

breaks all the rules. Only workers come from ‘normal’ sexual mating (2). Their

degree of kinship (r) with their mother is therefore equal to 0.5.

On the other hand, mothers produce young queens by cloning (1); that is to say,

they transmit to them the entirety of their genome. So queens, both mothers and

daughters, are exact genetic copies, with a degree of kinship equal to 1.

Males also reproduce by cloning, producing sons with the same genome as

themselves (3), so their degree of kinship is also equal to 1.

Because of this dual cloning, the only one of its kind known in the whole

animal kingdom, Wasmannia auropunctata queens and males are not related. It is

as though they each belong to different species.

190

THE LIVES OF ANTS

This can only be possible if the males are also produced by

fathers cloning themselves. This makes all the males true twins.

On the face of it, this seems not easy as it is the queen who is

laying the eggs. However, nature has found the knack of helping

the males to perform this little conjuring trick. At fertilization,

their spermatozoa, which are kept in the queen’s spermatheca,

penetrate the eggs and reach their nuclei. Once they have man-

aged this, some of them apparently eliminate the maternal

genome and take its place. Is this not a ﬁne example of genetic

identity theft?

So, in the little ﬁre ant, there is no intermingling of reproduct-

ive castes: queens produce queens, males produce males, and

there is not the slightest possibility of mixing bloodlines. This

way of reproducing is not without consequences, for it makes

impossible any gene ﬂow, that is mixing of genes, between males

and queens. The only individuals who could have contributed to

the genetic diversity of the population are the workers, for they

have inherited chromosomes from both parents. Given, though,

that they are sterile and have no progeny, there is no role for

them to play. The upshot is that the queens and the males behave

as though they belonged to two separate species.

This mode of reproduction is absolutely without parallel

anywhere in the animal kingdom. Of course, some species of

ﬁsh, amphibians, and insects reproduce by natural cloning; but

it ’s always the females among them who do this, by eliminating

the paternal genome. It never happens the other way round. So

the little ﬁre ant has come up with an innovation and has taken

asexual reproduction one step farther. It is likely that during

evolution these ants behaved at ﬁrst as Cataglyphis cursor still

do, the queen mother producing young queens by parthenogen-

esis. But then the males, almost deprived of any possibility of

reproducing, got their own back by devising a daring strategy for

handing on their genes, for all the world as though they were

saying to their partners, ‘You think you can pass on your total

191

ANYTHING GOES

genome intact to your daughters? Well, in that case, we’ll do

likewise with our sons!’

Could this be an ingenious solution to the recurring problem

of sex rivalry? Be that as it may, the little ﬁre ant Wasmannia

auropunctata certainly gives us a new glimpse of the richness and

complexity of methods of reproduction among social insects.

THE LIVES OF ANTS

192

Part VII

Sociogenetics

Drawing 7 One genome, two phenotypes

Pheidole

ants have the

same genome, but their genes, depending on how they are expressed, can

produce different castes.

Genes and family structure

There is nothing like genetics for ﬁnding out about the

private lives of ants and discovering the various modes of

reproduction that different species have adopted. But

advances in molecular biology can take us even farther

into their secrets, going beyond the general to the par-

ticular. Nowadays it is possible to bring to light subtle

variations between the genomes of individual ants within

the same species and to link genetics to behaviour.

Two questions exercise the minds of specialists in the life sciences

today. Does behaviour have a genetic basis? If so, how can the

genes be identiﬁed? Hitherto, in their attempts to ﬁnd answers to

these questions, biologists have relied on their favourite labora-

tory animals: mice, the fruit ﬂy Drosophila,oreventheworm

Caenorhabditis elegans. But these solitary organisms are not very

helpful in studying genes involved in social behaviour. Recent

studies of social insects have produced ﬁndings that are much

more promising. Work on the red imported ﬁre ant has actually

identiﬁed a genetic element which does directly and indisputably

inﬂuence the social organization of colonies—a world ﬁrst.

The ﬁre ant Solenopsis invicta really is very special, for it

can have two types of organization, the monogynous and the

195

polygynous social forms. In addition to the difference in numbers

of queens per colony, there are other differences between the two

types of organization. A young queen striking out on her own

from a monogynous society does not hesitate to ﬂy up to several

hundred metres above the ground in search of a suitable mate,

before going on elsewhere to found her family, whereas queens

from polygynous societies ﬂy neither so far nor as high, and once

copulation has taken place, they either go back to their original

nest or take up residence in some other polygynous nest, where

they set about reproducing.

Thus within a single species there can be two quite different

types of family structure associated with marked behavioural

differences. This we found strange and intriguing and it made

us wonder whether there was gene ﬂow between the two modes

of social organization. In other words, was it possible for cross-

breeding to happen between individual ants belonging to a

monogynous nest and a polygynous one? Or did the queens’

relationships never change, those from monogynous nests mat-

ing only with males from the same background as themselves

and the polygynous ones doing the same?

In a collaborative study with Kenneth Ross of the University of

Georgia, we discovered that in fact the males were not all that

selective. In particular, many of those that had been born in

single-queen colonies mated with females from polygynous

nests. This is the reason why the genomes of individuals from

the two types of nests are, broadly speaking, very similar. There

is, however, one exception to this, and it is a highly signiﬁcant

one. We discovered that there was a particular gene which made

it possible to differentiate individuals belonging to monogynous

colonies from the others. This gene is called ‘Gp-9’, meaning

‘General protein-9’.

Gener al protein-9 can have two forms, two alleles, designated

B and b. Given that females have two copies of this gene, this

gives three possible combinations: the genotype can be either BB

196

THE LIVES OF ANTS

or bb, individuals having these being called ‘homozygote’, or else

it may be Bb, in which case the term used is ‘heterozygote’.

Populations should therefore contain a mixture of these three

genotypes. But we found that in monogynous colonies all

queens and workers were homozygotes and all were BB. In

polygynous nests, however, all the queens were heterozygotes

(Bb), while the workers were either BB or Bb.

This ﬁnding was amazing, because it contradicted Mendel’s

genetic laws: according to the principles of heredity drawn up by

‘the pea plant man’, there must be something wrong with it. In

polygynous colonies, we should also have found BB and bb

queens as well as bb workers. But there weren’t any.

The characteristics of the alleles offer one possible explanation

for this. To have two copies of b present is lethal, and any queen

or worker carrying them do not live long after reaching adult-

hood, perhaps some weeks or even just a few days. So there is

nothing surprising in there being no bb queens to be found in the

nests, since they all die young.

But why are there no BB queens in polygynous colonies? Was

there perhaps some morphological or other peculiarity which

might be working against them? We decided to investigate

whether there was any link between the genotype of the queens

and their phenotype. And we found that in fact there was. It

turned out that BB queens were the heaviest and most fertile of

all and that bb queens (we did actually ﬁnd a few young ones still

alive) were the lightest and least fertile. The Bb queens were

somewhere in between.

All this was greatly perplexing. If BB queens are so fertile, why

are they absent from polygynous nests? We did in fact ﬁnd some,

only among the young unmated queens, but this does not really

alter the problem, for in theory they should have been very much

in the majority. After much pondering, it occurred to us that the

b allele might turn the Gp-9 into a ‘selﬁsh gene’. This is the term

used for a gene that favours its own propagation to the detriment

197

GENES AND FAMILY STRUCTURE