Sex Ratio and Egg Sequence in Ring-Billed Gulls

Abstract

Recently, Ankney (1982) documented a relationship between egg-laying sequence and the sex of newly hatched Lesser Snow Goose (Chen c. caerulescens) goslings. In 4-egg clutches more males (64%) than females hatched from the first 2 eggs laid and more females (72%) than males from the last 2 eggs. Investigations of this phenomenon in other species are necessary if we are to appreciate its importance and adaptive significance to sex-ratio dynamics in birds. During studies of Ring-billed Gulls (Larus delawarensis) nesting on Granite Island (48?43'N, 88?29'W) in Black Bay, northern Lake Superior [see Ryder (1976) for a description of the colony] in 1978, 1979, and 1982, I determined the sex of chicks from three-egg clutches in order to record the secondary (hatching) sex ratio. Stimulation for this study arose from the discovery, in 1978, of female-female pairs of Ringbilled Gulls nesting on Granite Island (Ryder and Somppi 1979). My initial objective was to determine whether or not disproportionately more females hatched from three-egg clutches. If so, that might explain an apparent abundance of females in the breeding population; the consequent skewed secondary sex ratio could result in the formation of femalefemale pairs. Each season I marked each study nest with a numbered wooden block and the eggs, in order of laying, with either a nontoxic black felt marking pen or brown nail enamel dots representing the egg number. The a-egg, b-egg, and c-egg refer to the 1st, 2nd, and 3rd eggs laid in a clutch, respectively. Only nests that were initiated before or during the peak of egg laying and subsequently contained three eggs were used for analyses, because such clutches are considered to be of normal size for Larus gulls (Vermeer 1969, Coulter 1977) and they typically experience the highest hatching success (Ryder and Ryder 1981) relative to clutches of more or fewer eggs. Clutches initiated later are usually smaller and suffer higher egg mortality (Ryder and Ryder 1981). I obtained data from a total of 89 three-egg clutches, but the number of chicks from which I obtained sex and sequence data did not equal 267 (3 X 89). Of the 66 eggs that were not used, 18 embryos died during early incubation, 37 eggs disappeared, and 11 eggs were found destroyed. From each marked clutch, I retrieved chicks that had died in late incubation, during pipping, or soon after hatching, if their position in the clutch was known, and determined sex by gonadal inspection. I collected newly hatched live chicks under permit, determined their sex, and then deposited them in the ornithological holdings of Lakehead University. Data were analyzed using Chi-square and G-tests (Sokal and Rohlf 1981). Significance was set at P < 0.05. Significant dependence occurred between chick sex and egg sequence in 1978 and 1979 but not in 1982 (Table 1). The lack of significance in the earliest starting season of 1982 is attributed to a greater proportion of males produced in b-eggs (57%) relative to 1978 (27%) and 1979 (25%). The combined data, however, show a strong statistical relationship. Although the secondary sex ratio for clutches did not differ from unity in any one study season (1978, x2 = 0.18; 1979, x2 = 3.26; 1982, x2 = 0.60), proportionately more females than males were produced in 1979 relative to 1978 and 1982 (Table 2). The results of this study suggest that the secondary sex ratio of Ring-billed Gulls probably does not contribute to a skewed sex ratio in favor of breeding-age females and, thus, cannot currently be considered an important factor in the origin of female-female pairs in the study population. The data indicate, however, that the allocation of sex by the female to specific eggs in the clutch very probably occurs in Ring-billed Gulls. Although the functional significance of sex allocation is not yet known in birds, it might be adaptive in Ring-billed Gulls for a male to hatch from the a-egg, because potentially it would then be a betterfed, larger, more vigorous son. This would have a