Regarding the third link in this proposed chain: skewed sex ratios do not in fact automatically result in homosexual pairs—some populations of Western and Herring Gulls with a disproportionate number of females, for example, have few (if any) same-sex pairs.65 Even in populations that do have a surplus of females, only a subset of the birds actually form homosexual pairs: most unpaired female Herring Gulls remain single (lesbian pairs constitute less than 3 percent of all pairs, and sometimes as few as 1 in 350), and some males remain unpaired even in populations with more females than males (indicating that some “extra” females are bypassing heterosexual mates). Granted, scientists were able to “induce” the formation of female pairs in a population of Ring-billed Gulls by removing males. However, this simply demonstrates that many females in this species have a latent bisexual capacity that manifests itself when males are in short supply—not that all same-sex pairs in this (or any other species) result from a shortage of the opposite sex. Moreover, the sex ratio that was required to “trigger” homosexual pairing (77 percent females) was much higher than the proportion of females in naturally occurring populations with homosexual pairs (55 percent females).66 Apparently not all bird species have this latent bisexual capacity either (or at least not to the same extent), since homosexual pairing does not occur in most species that have their sex ratios experimentally manipulated. Removing males (or adding females) to populations of willow ptarmigans, bufflehead ducks, Pied Flycatchers, great tits, Brown-headed Cowbirds, and song, seaside, and savanna sparrows, for example, results in females mating polygamously with males or remaining single (among other strategies) rather than forming same-sex pairs. When alternate heterosexual behaviors such as polygamy are “induced” in usually monogamous species such as these, scientists do not interpret this as evidence that the behavior is somehow “artificial” or that its occurrence is due solely to the experimentally triggered demographic changes. Rather, it is taken to indicate that the species possesses an inherent capacity for polygamy (and more broadly, a flexibility of mating behavior), perhaps expressed at relatively low levels in most populations but manifesting itself on a larger scale under the appropriate conditions.67 Significantly, this interpretation has not generally been afforded homosexual pairing.
Fourth, the evidence for homosexual pairing as a breeding strategy is slim. According to scientists, females bond with same-sex partners in order to raise young that result from copulation (but not pairing) with males (since two-parent care is generally required in these species). However, only a relatively small proportion of females in homosexual pairs actually mate with males and lay fertile eggs: 0–15 percent of Western Gull eggs belonging to female pairs are fertile, while only 4–30 percent of Herring Gulls’ are fertilized, indicating that few such females are actually breeding.68 Most importantly, females that could potentially benefit from same-sex pairing—were it a reproductive strategy—do not generally “avail” themselves of this option. Researchers found that unpaired Herring Gull females that copulate with males do not in fact go on to form homosexual pairs in order to raise any resulting offspring, nor do they even try to form such pairs. Likewise, Ring-billed Gull, Western Gull, and Roseate Tern mothers that have lost their male partners (and are otherwise unable to find another male) do not establish same-sex pair-bonds with available females, even though they supposedly need to find a new mate to assist them with parenting. In addition, some unpaired and homosexually paired Ring-billed females may actually lay eggs in the nests of other (heterosexual) pairs. This shows that: (a) single females need not seek pair-bonding (with birds of either sex) in order to have their young raised by two parents, and (b) at least some females in homosexual pairs lay eggs that they have no intention of caring for themselves.69
Lastly, there is not an absolute correlation between female pairs and supernormal clutches. True, in some species most lesbian pairs lay supernormal clutches, and most supernormal clutches belong to lesbian pairs. However, in many cases female pairs lay “normal”-sized clutches (or lose eggs so they end up with regular-sized clutches), while oversized clutches also regularly result from many other factors. These include egg stealing or adoption, supernumerary clutches laid by one female, nest-sharing by two heterosexual pairs, egg laying by outside females (not paired to the nest owners), and heterosexual trios, among others. In many gulls and other species, the connection between supernormal clutches and homosexual pairs has never been established (e.g., glaucous-winged gulls) or has been refuted (e.g., black-tailed gulls, brown noddies). Hence, studies that show correlations between toxins and increases in supernormal clutches cannot reliably be extrapolated to homosexual pairing unless it has been independently established that female pairs in that species lay larger than average clutches.70
Scientists also frequently point out a “correlation” between the two end points of this chain—toxins and supernormal clutches—without also providing evidence for all the intervening links.71 To show conclusively a relationship between the two phenomena, all the intermediate sequences need to be established, and they should preferably be established for the particular species in question. Sometimes, extrapolations are made in these links between species: that is, toxins are shown to be in the environment of one Gull species, feminization from toxins in another Gull species, skewed sex ratios in a third, female homosexual pairs in a fourth, and supernormal clutches in others—yet rarely (if ever) have all these conditions been shown to coexist in the same species or geographic area.72 Moreover, in many Gull studies this chain is collapsed entirely or rendered circular. If homosexuality occurs in a species, and there is also evidence of contamination or pollutants in the environment, the two are automatically assumed to be linked. Homosexual pairing is regarded as a self-evidently “dysfunctional” phenomenon (typically characterized as “reproductive failure”), hence investigators often feel no need to address the actual details of occurrence or causation in the supposed link to toxins. Indeed, the very existence of homosexuality is often subtly equated with environmental contamination and disease even when no actual pollutants have been discovered in the population in question. Ultimately, female pairing is seen as more than simply a behavioral response to certain demographic parameters, which may or may not be indirectly traceable to certain chemical effects. Rather, it assumes the status of a pathological “symptom” directly induced by man-made toxins, symbolizing the larger havoc that people have wreaked on the environment—nature gone awry as a result of human meddling.73 In the end, homosexuality becomes not merely the result of pollution, but the very “contamination” that is itself poisoning otherwise healthy—that is, purely heterosexual—species.
In summary, then, unavailability of the opposite sex is, at best, a tenuous “explanation” for the occurrence of animal homosexuality. Aside from having questionable theoretical and methodological underpinnings, this explanation is in many cases simply incompatible with the facts. In other cases, while same-sex activity does occur in contexts where opposite-sex partners are unavailable, many additional factors are involved, and many important questions concerning its occurrence remain to be addressed. Why, for example, do only some individuals or species with sex-skewed populations exhibit homosexual activity, while others manifest a wide variety of alternative behavioral responses? And why have social systems that entail sex segregation or skewed sex ratios—and hence that supposedly “favor” homosexual activities—evolved in the first place, and in so many species? Where it is relevant, unavailability of the opposite sex should be seen as only one of many contributing factors—and the beginning of further study of other more complex issues surrounding the occurrence of homosexuality in animals. Unfortunately, this explanation continues to be offered as a final scientific pronouncement on the “cause” of same-sex activity. Not only does this do a disservice to the actual richness of animal behavior, it effectively discourages further investigation of a phenomenon whose true intricacies are just beginning to be understood.