A Hundred and One Lesbian Acts: Calculating the Frequency of Homosexual Behavior

where a, b, C, D, and E represent the number of nests with 2-6 eggs respectively

—formulas used in estimating the number of

female homosexual pairs in Gull populations²⁷

While studying Kob antelopes in Uganda, scientists recorded exactly 101 homosexual mounts between females. In Costa Rica, 2 copulations between males were observed during a study of Long-tailed Hermit Hummingbirds. In which species is homosexuality more frequent? The answer would appear to be obvious: Kob. However, simply knowing the total number of homosexual acts observed in each species is not sufficient to evaluate the prevalence of homosexuality. For example, it could be that the Kob were observed for a much longer period of time than the Hummingbirds, in which case the greater number of same-sex mounts would not necessarily reflect any actual difference between the two species. What we really need is a measure of the rate of homosexual activity—that is, the number of homosexual “acts” performed during a given period of time. To determine this, we have to know the duration of the study period for each species and how many animals were being observed. In this case, 8 female Kob antelopes were studied for a total of 67 hours, whereas 36 male Long-tailed Hermit Hummingbirds were observed over several hundred hours—so indeed the Kob have a much higher rate of same-sex activity (both in general and per individual), on the order of many hundreds of times higher than the Hummingbirds.

Rate of occurrence is only one measure of frequency, however. It could be that sexual activity in general is much rarer in Hummingbirds than in Kobs, in which case comparing absolute numbers or rates gives a distorted or incomplete picture. A more meaningful comparison would be to look at how many heterosexual acts are performed during the same time period and express the frequency of homosexual activity as a proportion of all sexual activity. In fact, sexual activity is incredibly common among Kob and much rarer in Long-tailed Hermits: during the same study period, 1,032 heterosexual mounts among Kob were tabulated while only 6 heterosexual matings in Hummingbirds were observed. Thus, homosexual mounts constitute only 9 percent of all sexual activity among the Kob, whereas one-fourth of all copulations in Long-tailed Hermits are between males. This is diametrically opposed to the frequency rate or absolute count of homosexual activity in the same species.²⁸

These two cases offer a good example of the many complications that arise when attempting to answer the question “How common or frequent is homosexuality in animals?” The most valid answer—clichés aside—is, “It depends.” It depends not only on the measure of frequency being used, but also on the species, the behaviors being tabulated, the observation techniques that are employed, and many other factors. In this section we’ll explore some of these factors and try to arrive at some meaningful generalizations about the prevalence of homosexuality in the animal kingdom.

One broad measure of frequency is the total number of species in which homosexuality occurs. Same-sex behavior (comprising courtship, sexual, pair-bonding, and parental activities) has been documented in over 450 species of animals worldwide.²⁹ While this may seem like a lot of animals, it is in fact only a tiny fraction of the more than 1 million species that are known to exist.³⁰ Even considering the two animal groups that are the focus of this book—mammals and birds—homosexual behavior is known to occur in roughly 300 out of a total of about 13,000 species, or just over 2 percent. However, comparing the number of species that exhibit homosexuality against all known species is probably an inaccurate measure, since only a fraction of existing species have been studied in any depth—and detailed study is usually required to uncover behaviors such as homosexuality. Scientists have estimated that at least a thousand hours of field observation are required before more unusual but important activities will become apparent in a species’ behavior, and relatively few animals have received this level of scrutiny.³¹ Unfortunately, it is not known exactly how many species have been studied to this depth, although it has been estimated that perhaps only 1,000–2,000 have begun to be adequately described. Using these figures, the proportion of animal species exhibiting homosexual behavior comes in at 15–30 percent—a significant chunk.³²

In fact, the percentage is probably even higher than this, when we consider how easy it is for common behaviors to be missed during even the most detailed of study. A caveat of any scientific endeavor, particularly biology, is that much remains to be learned and observed, and many secrets await discovery—and this is especially true where sexual behavior is concerned. Nocturnal or tree-dwelling habits, elusiveness, habitat inaccessibility, small size, and problems in identifying individual animals are just some of the factors that make field observations of sexuality in many species exceedingly difficult.³³ Consider heterosexual mating, a behavior that is known to occur in all mammals and birds (and most other animals), usually with great regularity.³⁴ Yet in many species this activity has never been seen: “Despite literally thousands of hours of observations made by biologists over many years in the West Indies, Hawaii, and elsewhere, actual copulation in humpback whales has yet to be observed.”³⁵ Lucifer hummingbirds, northern rough-winged swallows, black-and-white warblers, red-tailed tropic birds, and several species of cranes (such as wattled and Siberian cranes) are just a handful of the birds in which heterosexual mating has never been recorded. In some cases, opposite-sex mating has been observed, but only a handful of times at most: in magnificent hummingbirds and black-headed grosbeaks, for example—the latter a common North American bird—copulation between males and females has only been seen once during the entire history of the scientific study of these species. Heterosexual copulation in Victoria’s Riflebirds was not documented until the mid-1990s (and then only several times), even though the species has been known to Western science for nearly a century and a half. During a ten-year study of Cheetahs, no opposite-sex matings were seen over the course of 5,000 hours of observation, and copulation has only been observed a total of five times in the wild during the entire scientific study of this animal. Similar patterns are characteristic of other species: in the akepa (a Hawaiian finch), only five copulations were witnessed during five years of study, only five heterosexual matings were seen in a four-year study of Spotted Hyenas, and only three matings in a three-year study of Agile Wallabies. Nests and eggs of many birds such as swallows and birds of paradise have yet to be discovered, while the first nest of the marbled murrelet was found in 1959, more than 170 years after discovery of the species by Western science.

And of course new revelations about heterosexual behavior are being made all the time: female initiation of mating activity in Orang-utans, for example, was not documented until 1980 in spite of nearly 22,000 hours of observation over the preceding 20 years (and prior extensive field studies often failed to report any heterosexual copulations). As recently as 1996, the existence of polygamous trios in the tanga’eo or Mangaia kingfisher (of the Cook Islands near New Zealand) were uncovered for the first time, and the full extent of heterosexual mating by Common Chimpanzees with animals outside their group was not understood until 1997. Multiple heterosexual matings by female Harbor Seals were not verified until 1998; even then, the behavior was never directly observed during three years of study (including continuous, 24-hour videotape surveillance of captive animals over an entire breeding season), and had to be verified indirectly through DNA testing.³⁶ If direct observation by scientists were used as the sole criterion for the existence of a behavior, we would have to conclude that many species never engage in heterosexuality (or in certain forms of heterosexuality)—yet we know this cannot be true. So the fact that homosexuality has not been seen in many animals does not necessarily mean that it is absent in those species—only that it has yet to be observed.

Ironically, many species in which heterosexuality has rarely or never been observed are ones in which homosexual activity has been recorded. No information on the heterosexual mating system of wild Emus was available prior to 1995, for example, although homosexual copulation in the same species had been observed in captivity more than 70 years earlier. Heterosexual mating has never been observed in Black-rumped Flameback Woodpeckers—although homosexual copulation has—while some studies of Nilgiri Langurs, Harbor Seals, Northern Quolls, and Gray-capped Social Weavers failed to record any instances of opposite-sex mounting, although same-sex mounting did occur. Similarly, documentation of sexual activity between male Walruses—including photographs—preceded by almost a decade comparable descriptions and photographic evidence of sexual activity between males and females. In Acorn Woodpeckers—a species that regularly engages in same-sex mounting—only 26 heterosexual copulations were recorded in over 1,400 hours of observation devoted specifically to recording opposite-sex mating. Likewise, heterosexual copulations in Australian Shelducks (a species in which females sometimes form homosexual pairs) were observed only nine times during nearly a decade of study, and on only three of these occasions was a complete behavioral sequence involved. Because of the difficulty of observing heterosexual copulation, the mating system of Killer Whales is still poorly understood and, according to one scientist, “may never be known with certainty.” Homosexual activity in the same species has already been documented, although its study is also still in its infancy.³⁷ Obviously, then, an activity can be part of the regular behavior of a species and still be completely missed by observers or documented only rarely, in spite of conscientious and in some cases exhaustive observational regimens (both in the wild and in captivity).

Scientists have often characterized homosexuality in animals as “extremely rare” or “quite common,” for example, or as occuring “regularly” or “infrequently” —often without any numerical or contextual information. Yet such statements are virtually meaningless without a common standard of measurement and an agreed-upon point of reference. In an attempt to standardize the evaluation of homosexual behavior, therefore, many scientists have collected quantitative information—usually tallies of particular behaviors (sexual, courtship, pairing, etc.). In a few cases, the difficulty of field observations has precluded the direct observation of both heterosexual and homosexual activity, and several innovative techniques have been developed to calculate the frequency of same-sex activity based on indirect measures. The sex of Gulls, for example, is often difficult to determine under field conditions, and in colonies that may contain tens of thousands of breeding pairs, the task of determining which couples are homosexual and which are heterosexual is a daunting one. However, once researchers discovered that lesbian pairs typically lay supernormal clutches, the frequency of same-sex pairs could be much more easily tallied by counting the number of nests with double the usual number of eggs. One ornithologist even developed a mathematical formula (see the beginning of this section) for estimating the total number of lesbian pairs in a population based on a sample of clutch sizes, taking into account same-sex pairs that lay smaller than supernormal clutches (or heterosexual pairs that lay larger than average clutches).³⁸ Likewise, determining the sex of mating Dragonflies can be extremely challenging while the animals are still alive, since they copulate in flight. Scientists discovered, though, that insects (both male and female) usually suffer head injuries from being clasped by a male during mating. These injuries can easily be recognized and counted once individual Dragonflies are collected—revealing that an average of nearly 20 percent of males in 11 different species (and more than 80 percent of males in some species) experience homosexual copulations.³⁹

Even when quantitative information is available, assessments of frequency are often subjective and contradictory. For example, one scientist observed 24 homosexual copulations between Pukeko (7 percent of all sexual activity) and classified the behavior as “common,” while another zoologist observed nearly identical numbers and proportions of same-sex mounts in Pronghorns (23 mounts, 10 percent of all mounting activity) yet classified the behavior as “rare.”⁴⁰ The problem is that there are many different ways of measuring and interpreting the frequency of same-sex behaviors. Besides tallying absolute numbers of particular activities or determining the proportion of all sexual activity that is homosexual, frequency rates and activity budgets can also be calculated, along with the percentage of the population that engages in same-sex activity. Frequency rate refers to the number of homosexual acts performed during a given time period, either by each individual or within a group of animals as a whole. For example, among Hanuman Langurs each female participates in a homosexual mounting, on average, once every five days, while in some populations of Ostriches courtship between males occurs at the rate of two to four times a day. An activity or time budget, on the other hand, refers to what proportion of an individual animal’s activity or time is devoted to homosexual interactions. For instance, a male Killer Whale spends more than 10 percent of his time interacting socially and sexually with other males; about 15 percent of courtship display time in male Regent Bowerbirds involves displaying to other males; more than a third of some male White-handed Gibbons’ time is devoted to homosexual interactions; while 10 percent of a male Crab-eating Macaque’s interaction with other males involves mounting activity.⁴¹ The percentage of the population that engages in homosexual activity varies widely, from only one or two individuals in a flock of several thousand Gulls, to virtually the entire population of male Bighorn Sheep, and everything in between. Of course, some of these individuals also engage in heterosexual activity (exhibiting various degrees of bisexuality), while others are more or less exclusive in their homosexual relations; this will be discussed in more detail in chapter 2, where the question of sexual orientation is explored.

Because of the diversity and complexity of homosexual expression across a wide range of species, it is not always a straightforward matter to calculate various measures of frequency such as these. Three different species exemplify some of the issues that are involved in just one measure of frequency, the proportion of sexual activity that is homosexual. Observed versus actually occurring behaviors in Giraffes, seasonal variations in sexual activity in Mountain Sheep, and alternative standards of reference in Gray Herons complicate the calculation of homosexual frequency in each of these species. During an exhaustive study of Giraffes in the Arusha and Tarangire National Parks of Tanzania, researchers recorded 17 homosexual mounts and 1 heterosexual mount in more than a year (and 3,200 hours) of observation. Thus, 94 percent of all observed mounting activity was same-sex. Does this reflect the actual proportion of homosexual activity in Giraffes? Certainly more than one heterosexual mating occurred during that time, since over 20 calves were born that year in one population alone. However, these populations did have relatively low birth rates, and heterosexual mating appeared to be genuinely rare. In addition, if heterosexual matings were being missed by the observers, probably homosexual ones were as well (unless same-sex mountings consistently took place in a more visible setting than opposite-sex ones). This means that the same proportion of homosexual activity could have been involved regardless of whether some mountings were missed.⁴² In Mountain Sheep, there are sharp seasonal differences in the proportion of same-sex activity. During the rut (about two months out of the year), approximately a quarter of all mounts are between males; during the rest of the year, virtually all mounts are homosexual, although only a small fraction of rams’ interactions with each other involves mounting.⁴³ Thus, homosexual activity is more common during the rut if frequency rates, time budgets, or absolute numbers are tallied, but more prevalent outside the rut if proportions of sexual activity are calculated. Gray Herons, like many other birds, often engage in promiscuous copulations: males try to mate with birds, both male and female, other than their partner. One study revealed that about 8 percent of such promiscuous copulation attempts were homosexual. However, if the total number of copulations—both promiscuous and between bonded partners—is taken as the point of reference, the proportion of homosexual mounting drops to 1 percent.⁴⁴ As these three examples show, differences between populations, seasons, and behaviors (among other factors) must be taken into consideration when assessing frequency.

Recognizing that measures of frequency often obscure important behavioral distinctions between (and within) species and may reflect widely divergent observational methodologies, it is nevertheless still useful to compare the prevalence of various homosexual activities across a wide variety of animals. The following summaries focus on the proportion of three behavioral categories—courtship, sexual, and pair-bonding—that occur between animals of the same sex, as well as the percentage of the population that engages in homosexual activity. These measures are the most widely available for a large number of animals, and they lend themselves fairly well to cross-species comparisons. Interestingly, similar average proportions are obtained for a number of these measures, even though they represent many different species and behaviors.⁴⁵

In those animals where homosexual activity occurs, an average of about a quarter of individuals in the population (or of a given sex) engage in same-sex activity—ranging from 2–3 percent of male Ostriches and female Sage Grouse, to nearly half of all male Giraffes and Killer Whales, to entire troops of Bonobos. Concerning specific behaviors, an average of about 25 percent of courtship activity occurs between animals of the same sex in those species that exhibit homosexual courtship interactions—ranging from less than 5 percent in Herring Gulls and Calfbirds, to more than 50 percent in Dwarf Cavies and Giraffes. A nearly identical proportion of sexual activity, just over one-quarter, occurs between animals of the same sex in those species where homosexuality has been observed—from as little as.3 percent in Silvery Grebes and 1–2 percent in Dusky Moorhens and Tasmanian Native Hens, to more than half of all sexual activity in Bison and Bonnet Macaques, and a whopping 94 percent of observed mountings in some populations of Giraffes (as mentioned above). Finally, an average of 14 percent of all pairs are homosexual in those species that have some form of same-sex (and opposite-sex) pair-bonding: ranging from as few as.3–.5 percent in Herring Gulls and Snow Geese, to more than a quarter of all consortships (on average) in Japanese Macaques, to more than half of all pair-bonds among young Galahs.

Combining these three behavioral categories yields a figure of just over 20 percent: roughly one-fifth of all interactions, on average, are homosexual in mammal and bird species that have at least some form of same-sex courtship, sexual, and/or pair-bonding activities. If one figure could be said to represent the overall frequency of homosexual activity in animals, this would probably be the one—but it is virtually impossible to come up with a truly “representative” number. A figure such as this collapses a multiplicity of behaviors (both between and within species), it represents only a fraction of the animals in which homosexuality has been documented, it glosses over many observational and theoretical uncertainties (not the least of which is widely differing sample sizes), and it misleadingly equates often radically unlike phenomena (different forms of heterosexual and homosexual behaviors, disparate social contexts, and so on). A less satisfying, but ultimately more meaningful, “formula” for understanding frequency is to recognize that there is no one overall frequency, no single formula. As in all aspects of animal homosexuality, different species exhibit an extraordinary range of rates, quantities, periodicities, and proportions of same-sex behavior—a diversity that is equal to the variation in the behaviors themselves. We can make tallies for particular species, develop formulas for certain populations or behaviors, and calculate percentages, time budgets, and so on—thereby trying to gain some overall impressions regarding the prevalence of homosexuality in animals. In the end, though, we must acknowledge that our measures are at best imperfect—and what we are attempting to quantify is, in many senses, incalculable.