BIRDS DO IT. Bees do it -- but not always. And, while educated fleas do it, a number of other insects never do it.

It, as biologists know and as Cole Porter meant, is sex. Or, more specifically, sexual reproduction.

The animal species that do not do it practice an alternative form of reproduction without benefit of males. It is called parthenogenesis, the Greek word for virgin birth.

Although the reality of virgin births among animals has been established for about 250 years, only recently have biologists discovered that the phenomenon is fairly widespread. It is practiced by an estimated one out of every thousand species, an elite group that includes not only insects but scores of higher species.

Biologists have also begun to grapple more deeply with a variety of controversial theoretical questions. One is startlingly fundamental: If parthenogenesis works, why is there sex? Why do most species have males? Why bother with fathers?

The question is especially challenging in view of what seems to be a great advantage enjoyed by asexual parthenogenetic species: In most such species, every member is female and can produce young, while among ordinary sexual species, only half are female and can do so. Parthenogenetic animals, therefore, would seem to be able to multiply twice as fast as they could if they had to devote half of their resources, such as food and space, to creating and maintaining males.

It seems obvious that any species able to increase its numbers faster than a competing species should drive the rival out of their common ecological niche and come to dominate, all other factors being equal. And yet, the fact remains that an estimated 99.9 percent of all species keep males around. This suggests males must be good for something.

Most parthenogenetic species consist entirely of females who produce only daughters. They include lizards in New Mexico, salamanders from many parts of the United States and several fish species from around the world. Even the bees that do it are partially parthenogenetic. Females -- the workers and queens -- are produced by sexual reproduction but males -- the drones -- hatch from unfertilized eggs. Male bees have only mothers, never fathers.

In other non-sexually reproducing species, males are a sometime thing. These creatures, such as aphids, the familiar garden pests, live through the spring and summer as all-female societies, reproducing one male-less generation after another through virgin birth. (Birth is the only word, for aphids are live-bearers.) Then, as fall approaches and in a process that is not well understood, males are produced. The males participate in one generation of ordinary sexual reproduction that yields eggs which remain dormant through the winter. Spring warmth hatches a new, all-female season.

Biologists also have recently found that parthenogenetic strains can arise fairly easily within sexual species -- adding to the puzzle of why fatherhood survives. Of 30 species of fruit fly examined, 27 have given rise to parthenogenetic strains. In fact, most of the known parthenogenetic species are mutant forms of closely related sexual species. Some biologists think many species repeatedly give rise to nonsexual mutant lineages. Some guppy-like species that reproduce sexually have spawned several parthenogenetic lineages from the same parent stock.

Laboratory studies have shown that the unfertilized eggs of many species can be induced to develop into mature individuals. Often it takes no more than a minor stimulus to the egg, either mechanical or chemical. Unfertilized frog eggs, for example, pricked with a pin dipped in lymph will "think" they have been fertilized and begin developing into a frog.

At the Agricultural Research Service's laboratories in Beltsville, scientists have even produced parthenogenetic turkeys this way, the only known example of virgin birth among birds. The turkeys, however, were sickly males that could not perpetuate their lineage.

A limited form of parthenogenesis has also been found among mammals. One strain of mice is prone to spontaneous parthenogenesis, but the embryos die early in development. Scientists have induced artificial parthenogenesis in several strains of mice. At Yale University, for example, chemically treated ova have been induced to begin embryonic development without the male's sperm. Again, however, the embryos died before reaching full development.

Parthenogenesis is even known in human beings. There are recorded instances in which an ovum in a woman's ovary began dividing as if it had been fertilized. Often development goes awry and, instead of producing a normal embryo, the result is an ovarian tumor. There is no scientifically documented instance of human parthenogenesis that produced a full-term baby.

The virgin conception and birth of Jesus, a key article of faith for many Christians, has no scientific documentation. The fact that Jesus was a male, however, argues against his birth being a result of parthenogenesis. The genetic factor that produces a human male is the Y chromosome, which can come only from the male's contribution to conception: a sperm bearing a Y chromosome. Ova, produced by women, contain only X chromosomes. An ovum that developed parthenogenetically would have no Y chromosomes and would, therefore, be female.

Scientists agree that if Jesus had no human father, his conception and birth would have had to be miracles.

The fact that mammalian parthenogenesis seems invariably to fail leads biologists to suspect that their reproduction, including that of human beings, is a more complex, perhaps more subtly controlled, phenomenon than among simpler animals.

These studies have established that fertilization by sperm actually consists of two separate events -- the contribution of the father's genes and the triggering of cell division. Artificial virgin birth amounts to simulating the triggering effect alone.

Ever since Darwin, biologists have been impressed that most species bother with fathers. The orthodox explanation goes as follows:

Species that reproduce by virgin birth produce offspring that inherit only the female's genes. Thus all the individuals are clones of the founding mother and, theoretically, possess exactly the same physical traits.

Among sexual species, on the other hand, each individual receives half its chromosomes from each of two genetically different parents. As a result, individuals within such species vary widely in their physical traits and have different abilities to survive in various habitats.

The value of genetic variability shows itself whenever the species' environment changes or when the species migrates to a new habitat. A drought, for example, could wipe out the whole population of a parthenogenetic species. By contrast, a similar sexual species would have differing susceptibilities and some individuals might, by chance, have genes that make them drought resistant. Though many individuals might die, the resistant ones would survive to replenish the species.

Thus, under this theory, the reason for males is to maintain genetic variability in a species. Evolution, to put it teleologically, created sex because it was a good way of helping a species to survive over the long haul.

(It should be noted that males serve an additional function in species where parental care of the young is important to their survival. Parenting is not a factor, however, for most lower species where the young fend for themselves. Thus the raison d'etre for males in these cases must lie elsewhere.)

In recent experiments biologists have found evidence that the orthodox explanation represents too clean-cut a distinction between the relative advantages of sexual and asexual species. Parthenogenetic species, it turns out, are not nearly as good at reproduction as they might seem on a theoretical basis. And, in fact, they are not nearly as genetically similar as was once supposed.

Detailed studies have shown that most of the eggs and embryos produced by asexual species fail to develop. The best that has been found in nature is an all-female fruit fly species from South America in which only about half the eggs hatch -- canceling out the theoretically doubled reproductive rate of a parthenogenetic species. All the other species studied do less well. Thus they have little chance of out-competing their ecological rivals.

At the same time, many asexual species are not nearly as invariant as the term clone might suggest. While offspring of an asexual mother do inherit the same genes, they are not always in the same position on the chromosome and, as a result, the offspring may develop differently. Also, asexual individuals, like sexual individuals, are subject to mutations that can create entirely new genes out of old ones. Thus some parthenogenetic species may have enough genetic variability to escape an environmental shift that would otherwise doom the entire population.

Even so, there appear to be limits to such evolution. All the known examples seem to have evolved fairly recently, at least on the geological scale. Most sexual animals today belong to larger taxonomic groups that can be traced back to extinct forms that lived millions of years ago. When parthenogenetic species are traced back, the ancestor is almost invariably found to be a sexually reproducing species that is still living.

Such recent insights in the study of virgin birth, taken together, have greatly complicated a theoretical picture that once seemed much simpler. If there is any consensus on parthenogenesis, it is that the phenomenon is a biological aberration that can arise fairly readily, at least among lower animals, and produce species well adapted to a specific ecological niche.

Many biologists see parthenogenetic species as good colonizers of newly opened ecological niches -- habitats suddenly created by an environmental shift. This is because all members share much the same set of adaptations, despite some genetic variability. A female that happened to have the traits enabling her to survive in the new habitat could populate it fairly quickly without needing to take a male along.

But such species do not often do well in competition with established species. Short-term heavy hitters, some biologists call them. When sexual species evolve a lineage adapted to the new habitat, they may quickly doom the parthenogenetic species.

Natural virgin birth, then, is real but transitory, an experiment of nature that works only under special conditions. But it keeps reappearing, reminding biologists that there are still times and places in which males are useless and unnecessary.