1 3 J Comp Physiol B (2015) 185:539–546 DOI 10.1007/s00360-015-0897-5 ORIGINAL PAPER Timing matters: corticosterone injections 4 h before ovulation bias sex ratios towards females in chickens Sara E. Pinson 1 · Jeanna L. Wilson 1 · Kristen J. Navara 1 Received: 15 September 2014 / Revised: 10 February 2015 / Accepted: 1 March 2015 / Published online: 15 March 2015 © Springer-Verlag Berlin Heidelberg 2015 ratios were instead biased towards females. These results suggest that the timing and magnitude of the corticosterone elevation are both critical not only to whether a sex ratio bias occurs, but also the direction of the bias. Keywords Primary sex ratio · Offspring sex · Corticosterone · Stress · Maternal effects Introduction Birds have shown a unique ability to bias sex ratios even before significant embryonic development has taken place, and it is likely these biases occur before fertiliza- tion occurs. These primary sex ratio manipulations occur in many species of birds (reviewed in Pike and Petrie 2003; Alonso-Alvarez 2006) and in relation to a variety of fac- tors, including maternal body condition, social status, pres- ence of helpers, mate quality, and food availability and quality (reviewed in Pike and Petrie 2003; Alonso-Alvarez 2006). Even though many of the factors associated with sex ratio biases have been identified, we know little about the mechanisms controlling sex ratio manipulation in birds. In birds, the female is the heterogametic sex, contribut- ing either a W or Z chromosome to offspring. During the first meiotic segregation, which completes 2–4 h prior to ovulation (Olsen and Fraps 1950; Johnson 1996), one sex chromosome is retained in the oocyte and the other is segregated to the polar body with no further potential for development. It has been proposed that a female bird can control primary sex ratios through a variety of mechanisms due to her heterogametic nature (Alonso-Alvarez 2006). Primary sex ratio biases could occur just prior to ovulation through non-random chromosome segregation, so that the desired chromosome remains in the oocyte, or at the time Abstract Birds have the ability to influence offspring sex prior to egg laying and may use hormones to medi- ate these skews. Corticosterone is of particular interest as a mediator of offspring sex because, as the primary stress hormone in birds, it regulates responses to environmental and social stimuli that trigger sex ratio biases. In previous studies in birds, chronic elevations of corticosterone stimu- lated female-biased sex ratios while acute pharmacological elevations that were provided 5 h prior to the expected time of ovulation stimulated male-biased sex ratios. Here, we aimed to determine the magnitude of corticosterone neces- sary to influence offspring sex and to further pinpoint the timing of the hormonal influence. Because high-dose injec- tions of corticosterone stimulated male-biased sex ratios in hens, we hypothesized that females receiving acute phar- macological elevations of corticosterone would produce more male offspring while females with acute physiologi- cal elevations would produce an intermediate proportion of males compared to controls. We tested our hypotheses in laying hens by elevating corticosterone in the physiological or pharmacological range through injections of corticos- terone administered 4 or 5 h prior to the expected time of ovulation. Contrary to our hypothesis, a physiological dose of corticosterone provided 5 h prior to the expected time of ovulation did not bias offspring sex ratios when compared to controls. Further, when corticosterone injections were given at 4 h prior to the expected time of ovulation, sex Communicated by G. Heldmaier. * Kristen J. Navara knavara@uga.edu 1 Department of Poultry Science, University of Georgia, 110 Cedar Street, 203 Poultry Science Building, Athens, GA 30602, USA 540 J Comp Physiol B (2015) 185:539–546 1 3 of ovulation through selective internal ovulation and sec- ondary ovulation, where any oocyte containing the unde- sired chromosome is discarded. If the sexes of the follicles are predetermined, follicles programmed to be the desired sex could be preferentially selected into the ovulatory hier- archy or grow at a different rate than other follicles; in this case, the ultimate sex ratio skew would also occur via non- random chromosome segregation. Hormones are good candidates as mediators for adjust- ing offspring sex ratios because they are important regu- lators of behavioral, biochemical, and physiological func- tions involved in reproduction. Indeed, biases in primary sex ratios have been observed in birds with elevated repro- ductive and stress hormones (reviewed in Navara 2013). Acute elevations of progesterone several hours prior to ovulation stimulated the production of more females (Cor- rea et al. 2005) while acute or chronic elevations of tes- tosterone have stimulated male-biased sex ratios (Veiga et al. 2004; Rutkowska and Cichón 2006; Goerlich et al. 2009; Pinson et al. 2011b). Sex ratio biases have also been observed following acute (Gam et al. 2011; Pinson et al. 2011a) and chronic (Pike and Petrie 2005, 2006; Bonier et al. 2007; Goerlich-Jansson et al. 2013) elevations of the adrenocortical hormone corticosterone. Corticosterone is the primary steroidal stress hormone in birds and is of particular interest as a potential mediator of primary offspring sex ratios for several reasons. Corticos- terone is released from the adrenal glands in close proxim- ity to the ovary, and is elevated during and participates in ovulation (Etches and Cunningham 1976). Corticosterone also regulates catabolic activities and causes weight loss (Norris 2007), and thus would likely be elevated in many situations that lead to sex ratio biases, such as low mater- nal body condition (Wiebe and Bortolotti 1992; Nager et al. 1999; Whittingham and Dunn 2000; Alonso-Alvarez and Velando 2003; Pike and Petrie 2005; Pike 2005; Whitting- ham et al. 2005) and low food availability and quality (Kil- ner 1998; Clout et al. 2002; Arnold et al. 2003; Rutstein et al. 2004). Studies in several avian species link chronically elevated concentrations of corticosterone in female circulation to female-biased sex ratios (Pike and Petrie 2005, 2006; Bon- ier et al. 2007). White-crowned sparrows and peafowl with corticosterone levels that were naturally elevated over a long period of time produced more female offspring (Pike and Petrie 2005; Bonier et al. 2007). White-crowned spar- rows and Japanese quail with experimentally elevated cor- ticosterone via implants also produced more female off- spring (Pike and Petrie 2006; Bonier et al. 2007). A recent study in domestic hens suggests that corticosterone exerts influences on sex ratios, but via interactions with body mass (Aslam et al. 2014). Short-term elevations of corticosterone also exert effects on primary sex ratios in birds. Zebra finches and domestic laying hens injected with high doses of cor- ticosterone 5 h prior to the expected time of ovulation, which is just prior to chromosome segregation, pro- duced significantly more male offspring than untreated females; this likely occurred through non-random chro- mosome segregation of sex chromosomes, because the corticosterone treatment did not increase the incidences of laying gaps or decrease fertility (Gam et al. 2011; Pinson et al. 2011a). It is still unclear whether a physio- logical dose of corticosterone provided at 5 h prior to the expected time of ovulation, or whether providing corti- costerone even closer to ovulation, would elicit a similar sex ratio skew. The purpose of this current study was twofold. First, by elevating corticosterone within the physiological range at 5 h prior to the expected time of ovulation, we tested whether an acute physiological increase of corticosterone at the expected time of chromosome segregation was suf- ficient to influence offspring sex in white leghorn hens. Additionally, to test whether treatment closer to ovulation would induce a similar skew, we elevated corticosterone either in the physiological or pharmacological range at 4 h prior to the expected time of ovulation. We hypothesized that there is a critical dose and time necessary for produc- ing a sex ratio bias in hens. Materials and methods Single-comb laying hens (Hy-Line International) were housed in a single room in individual cages, provided ad libitum access to food and water, and maintained on a stim- ulatory light cycle (14 h L: 10 h D). Laying hens ( Gallus domesticus ) were selected for this study because of their regular ovulation–oviposition cycles and ease of predicting ovulation timing on a daily basis. Ovulation occurs within 30 min of oviposition of the previous egg (Johnson 2000), so egg laying patterns can be used to predict the ovulation timing of the target egg in the domestic hen. Additionally, the timing of meiotic segregation has been well docu- mented and segregation of the chromosomes completes 2–4 h prior to ovulation (Olsen and Fraps 1950). It is this critical time period, 2–4 h prior to ovulation, that we tar- geted with our treatments. Since we showed previously that an injection 5 h prior to the expected time of ovulation elicits an effect on sex ratios, we aimed to repeat this treat- ment and also test a physiological dose at the same time point as well as both a physiological and a pharmacological dose 1 h later (4 h prior to the expected time of ovulation, Fig. 1). 541 J Comp Physiol B (2015) 185:539–546 1 3 Experimental design We conducted two experiments. Experiment 1 (5 h injection protocol) was conducted when the hens were 33–36 weeks old. Hens were artificially inseminated twice weekly with pooled semen from 36 roosters to ensure egg fertilization. Eggs were collected manually every 2 h during peak laying hours and egg collection times were recorded for each indi- vidual hen so that egg laying patterns could be determined. Once the timing of ovulation could be predicted for an indi- vidual hen, she was randomly assigned to one of the follow- ing treatment groups: high corticosterone (HCORT: 1.5 mg corticosterone dissolved in 1 ml peanut oil, n = 56), low corticosterone (LCORT: 0.5 mg corticosterone dissolved in 1 ml peanut oil, n = 54) or control oil (C: 1 ml peanut oil, n = 51). We also collected eggs from 176 un-injected (UN) hens. Injections were administered subcutaneously in the back of the neck 5 h prior to the predicted time of ovulation (Fig. 1). On the day following injection, we monitored both whether an egg was laid and the time the eggs were laid. Of those injected, seven HCORT hens, six LCORT hens, and six C hens did not lay an egg. On average, >90 % of these eggs were laid within 2 h of the predicted laying time. We also monitored numbers of infertile eggs, fertile eggs that had not developed at all, and eggs in which embryos had died earlier in development. Of injected hens, four HCORT hens, one C hen, and one LCORT hen laid infertile eggs. This brings the sample sizes for Experiment 1–45 HCORT embryos, 47 LCORT embryos, and 44 C embryos. Blood samples were collected from a separate set of hens treated in the manner to those described above (Fig. 1). We did not quantify offspring sexes from eggs laid by these hens because the process of blood collection represents a stressor and may have stimulated endogenous corticoster- one production. Blood samples were collected from subsets of hens in the all groups 20 min, 1, 2, and 4 h after injec- tion, which coincided with 4.6, 4, 3, and 1 h prior to the predicted time of ovulation for Experiment I. Blood sam- ples were collected from the brachial vein within 3 min of initial handling to avoid variation in plasma hormone con- centrations due to handling stress (Romero and Reed 2005) and were then centrifuged at 16,000 × g and the plasma was frozen at − 20 °C until hormone analysis. Experiment 2 (4 h injection protocol) was conducted on the same population of hens several weeks after the completion of Experiment 1 to ensure that the treatment from Experiment 1 was no longer influencing the hens. Hens were again randomly divided among the same three treatment groups as in Experiment 1: (HCORT: n = 69, LCORT: n = 64, C: n = 61. Eggs were also collected from an additional 210 un-injected hens (UN). Injections were administered at 4 h prior to the expected time of ovulation. For this experiment, 19 C hens, 12 HCORT hens, and 13 LCORT hens did not lay eggs. These numbers were higher than in Experiment 1, indicating that hens were either aging reproductively or we did not time the experiment as accurately; however, rates of missing eggs did not differ among treatment groups. We found infertile eggs for one HCORT hen, three LCORT hens, and two C hens as well as one LCORT embryo that had died earlier in development and could not be sexed. For this experiment, 75 % of eggs were laid within 2 h of the target time, and this percentage did not differ based on treatment group. Final sample sizes were 56 HCORT eggs, 47 LCORT eggs, and 40 C eggs. Blood samples were also collected from a subset of hens and coincided with 3.6, 3, and 2 h prior to the expected time of ovulation as well as the predicted time of ovulation (Fig. 1). For both experiments, oocytes took approximately 24 h to transverse the oviduct and were collected at ovi- position (Fig. 1). All collected eggs were incubated for 8 days at 37.5 °C and 58 % relative humidity in a Nature- form incubator prior to embryo sexing. We chose to allow eggs to incubate to 8 days to maximize the amount of tis- sue available for molecular sexing. Because these birds have extremely low rates of infertility and early embryonic death, it is unlikely that any sex ratio adjustments observed in this study resulted from sex-specific embryonic death. Molecular sexing Embryos were manually removed from eggs and stored in 70 % ethanol solution until genomic DNA was extracted from embryonic tissue using a standard salt extraction method (Lambert et al. 2000). Portions of CHD-W and CHD-Z genes from the sex chromosomes were amplified Fig. 1 Injections were administered 5 h ( Exp. I ) or 4 h ( Exp. II ) prior to the expected time of ovulation to increase maternal plasma corti- costerone during the critical period of chromosome segregation of the oocytes. Target eggs were collected the following day, approximately 29 h after injection. In separate sets of hens, blood samples were col- lected 20 min, 1, 2, and 4 h after injection 542 J Comp Physiol B (2015) 185:539–546 1 3 using polymerase chain reaction (PCR) with primers 2550F and 2718R (adapted from Fridolfsson and Ellegren 1999). For PCR amplification, we used a reaction volume of 25 ul, containing 3 mM MgCl 2, 40 mM Tris–HCl, 100 mM KCl, 200 uM dNTP Mix (Bio-Rad Laboratories, Hercules, CA, USA, cat# 170-8874), 5 pmol each primer, 0.18U iTaq© DNA polymerase (Bio-Rad Laboratories, Hercules, CA, USA, cat# 170-8870), 500 ng DNA, and water. PCR was performed in a Bio-Rad thermal cycler as described by Pin- son et al. (2011a). We visualized PCR products on a 3 % agarose gel stained with ethidium bromide. Radioimmunoassay Extraction and radioimmunoassay of plasma corticoster- one were completed as described by Wingfield and Farner (1975) and Etches (1976). Briefly, a small amount of tri- tiated hormone (approximately 1000 cpm) was added to each 20 ul plasma sample for later calculation recovery efficiency after which corticosterone was extracted from samples using 3 ml of diethyl ether. Samples were then snap frozen and supernatant was collected in a fresh tube and dried under an N 2 stream. Plasma samples were resus- pended in 300 ul of phosphate-buffered saline (PBS) gel and duplicate aliquots of 100 ul were added to assay tubes. An additional 50 ul sample was used to determine extrac- tion efficiencies. To each assay tube and to additional tubes containing a graduated curve of corticosterone, 50 ul of tri- tiated hormone (approximately 10,000 cpm) and 50 ul of rabbit-derived anti-corticosterone antibody (MP Biomedi- cals, Solon, OH, USA, cat#07-120016) were added. All assay tubes were then incubated for 18 h after which 500 ul of a solution containing dextran-coated charcoal was added to each tube. Tubes were incubated for 10 min and centri- fuged at 4500 rpm for 10 min to separate bound and free fractions. Supernatant was decanted into scintillation vials and radioactivity was counted after 4 ml of scintillation fluid was added. The final concentration for each sample was corrected for its individual extraction recovery percent- age. All samples were run in 15 radioimmunoassays (Exp. 1: seven assays, Exp. 2: eight assays). Average intra-assay variation was 2.67 % (Exp. 1: 2.42 %, Exp. 2: 2.9 %) and interassay variation was 9.54 %. The average extraction recovery was 86.25 % (Exp. 1: 85.25 %, Exp. 2: 87.25 %). From hens injected 5 h prior to the expected time of ovulation (Experiment I), blood samples were analyzed from 29 UN hens ( n = 5, 10, 6, and 8 at 20 min, 1, 2, and 4 h, respectively), 30 C hens ( n = 7, 8, 7, and 8 at 20 min, 1, 2, and 4 h, respectively), 32 LCORT hens ( n = 8 for all collection times), and 31 HCORT hens ( n = 7, 9, 7, 8 at 20 min, 1, 2, and 4 h, respectively). From hens injected 4 h prior to the expected time of ovulation (Experiment II), blood samples were analyzed from 46 UN hens ( n = 11, 12, 11, and 12 at 20 min, 1, 2, and 4 h, respectively), 45 C hens ( n = 10, 13, 10, and 12 at 20 min, 1, 2, and 4 h, respec- tively), 45 LCORT hens ( n = 12, 11, 10, and 12 at 20 min, 1, 2, and 4 h, respectively), and 45 HCORT hens ( n = 12, 12, 10, and 11 at 20 min, 1, 2, and 4 h, respectively). Statistical analyses Logistic regression analyses were used to compare the pro- portions of male offspring produced among treatments for both experiments. We used individual logistic regression to determine whether the proportion of males produced by any treatment group differed significantly from a 50:50 sex ratio of the same sample size. All hormone data were non-normally distributed and were log-transformed for statistical analyses. The effect of treatment on plasma corticosterone concentrations was analyzed at each time point with analysis of variance (ANOVA), using Fisher’s protected least significant differ- ence (PLSD) to determine differences among individual treatment groups. Statistical analyses were performed using Statview software (SAS Institute, Cary, NC, USA). Results Plasma corticosterone concentrations We aimed to elevate corticosterone concentrations into both the physiological and pharmacological ranges at each time point. Values of corticosterone concentrations were considered to lie in the physiological range if they were at or below the maximum concentration of corticoster- one that has been produced in response to a stress chal- lenge, which for chickens is as high as 60 ng/ml (Edens and Siegel 1975). For the pharmacological dose, we used the same treatment of corticosterone (1.5 mg of corti- costerone dissolved in 1 ml peanut oil) as we used pre- viously in the same strain of birds to induce a sex ratio skew. Interestingly, while this treatment initially resulted in a pharmacological elevation within 20 min (120 ng/ ml), concentrations quickly dropped such that they were significantly lower at 1 h following injection compared to those in the previous experiment (69 vs. 160 ng/ml). As a result, our attempts at a pharmacological injection at the − 5 h time point in this study resulted in a high physi- ological elevation at the critical time when sex chromo- somes were likely influenced (Fig. 4), and we now tested the effect of two different physiological elevations (high and low) on sex ratios produced by these hens. The same injection doses given at 4 h prior to the expected time of 543 J Comp Physiol B (2015) 185:539–546 1 3 ovulation produced similar corticosterone concentrations to those in the previous experiment at both 20 min and 1 h following injection (150 vs. 160 ng/ml). The reasons for this difference are unknown; however, for the second experiment, when injections were administered 4 h prior to the expected time of ovulation, we were able to make the intended comparisons between a pharmacological and physiological elevation of corticosterone. For both experiments, treatment significantly affected plasma corticosterone 20 min (Expt. 1: F 3,23 = 26.62, p < 0.0001, Expt. 2: F 3,27 = 631, p < 0.0015), 1 h (Expt. 1: F 3,31 = 106, p < 0.0001, Expt 2: F 3,44 = 313, p < 0.0001) and 2 h (Expt. 1: F 3, 24 = 21.15, p < 0.0001, Expt 2: F 3,41 = 184.56, p < 0.0001) after injection. Corticoster- one concentrations in both HCORT and LCORT birds were significantly elevated compared to controls 20 min, 1, and 2 h after injection (HCORT: p < 0.01 at all time points; LCORT: p < 0.05 at all time points), falling back to uninjected levels 4 h after injection in birds injected 5 h prior to the expected time of ovulation ( p = 0.08) but not those injected 4 h prior to the expected time of ovulation ( p ≥ 0.35 for all; Fig. 2a, b). Offspring sex ratio As with corticosterone concentrations, sex ratios of off- spring did not differ between control and un-injected individuals in either Expt. 1 (5 h) or Expt 2 (4 h) (Expt 1: χ 2 = 1.20, p = 0.27, Expt 2: χ 2 = 0.003, p = 0.96). Thus, we combined control and un-treated birds for all sex ratio analyses to increase sample sizes. When injected 5 h prior to the expected time of ovulation, LCORT females pro- duced 53.2 % males, which was not significantly differ- ent from the ratios produced by UN and control females combined ( χ 2 = 0.58, p = 0.44) or the hypothetical 50:50 ratio ( χ 2 = 0.25, p = 0.62). HCORT birds produced 35.6 % males which also was not significantly different from UN and C females combined ( χ 2 = 1.90, p = 0.17) or a hypothetical 50:50 ratio ( χ 2 = 1.92, p = 0.17). UN and C females produced sex ratios that were not statistically dif- ferent from a hypothetical 50:50 (UN: χ 2 = 0.02, p = 0.89; C: χ 2 = 2.14, p = 0.15) or each other ( χ 2 = 1.47, p = 0.23) (Fig. 3). When injected 4 h prior to the expected time of ovu- lation, HCORT females produced 37.5 % males, which was not significantly different from the ratios produced by UN and control females combined (50 %, χ 2 = 3.01, p = 0.08). LCORT females produced 38.3 % males, which also was not significantly different from the ratios produced by UN + Control females ( χ 2 = 2.29, p = 0.13). When LCORT and HCORT groups were combined, however, they produced significantly fewer males compared with UN and control females ( χ 2 = 4.57, p = 0.03) (Fig. 3). Discussion In this experiment, a dose of corticosterone given 5 h prior to the expected time of ovulation that induces a Fig. 2 a Plasma corticosterone concentrations (mean + standard error) of hens in the untreated (UN), control oil (C), and physiologi- cal corticosterone (LCORT) groups 20 min ( n = 5, 10, 6, 8), 1 h ( n = 7, 8, 7, 8), 2 h ( n = 8 for all groups), and 4 h ( n = 7, 9, 7, 8) after injection 5 h prior to the expected time of ovulation ( Exp. I ). b Plasma corticosterone concentrations (mean + standard error) of hens in the untreated (UN), control oil (C), physiological corticoster- one (LCORT), and pharmacological corticosterone (HCORT) groups 20 min ( n = 11, 12, 11, 12), 1 h ( n = 10, 13, 11, 12), 2 h ( n = 12, 11, 10, 12), and 4 h ( n = 12, 12, 10, 11) after injection 4 h prior to the expected time of ovulation ( Exp. II ). Hormone concentrations were compared only among treatment groups at each time point, not between time points. Statistical differences were determined using log-transformed values in ANOVA because hormone data were not normally distributed, but actual values are shown in the graph. Differ- ent letters above the bars denote statistical differences 544 J Comp Physiol B (2015) 185:539–546 1 3 physiological rise of corticosterone did not stimulate a skew in offspring sex ratios. Corticosterone given in a high dose at 5 h prior to the expected time of ovulation, i.e., the time that meiosis was expected to be resumed, produced numerically but not significantly lower offspring sex ratios. Similarly both high and low doses given 4 h prior to the expected time of ovulation again resulted in a numerically lower offspring sex ratios, but these results were not sig- nificant unless the data of the low dosage and high dosage 4 h corticosterone treatment groups (LCORT and HCORT) were taken together. Therefore, these results must be interpreted with caution, and another larger study would help to validate these results. Still, this trend agrees with previous studies in which physiological doses of a chronic nature stimulated production of more females in other spe- cies (peafowl: Pike and Petrie 2005, 2006; White-crowned sparrows: Bonier et al. 2007; homing pigeons: Goerlich- Jansson et al. 2013). Interestingly, the sex ratio bias we saw in Expt. 2 was in the opposite direction than predicted based on previous studies in which corticosterone was provided in an acute elevation. Injecting birds with pharmacological doses of corticosterone at 5 h prior to the expected time of ovulation (resulting in levels of approximately 150 ng/ml in chick- ens and 70 ng/ml in zebra finches) previously resulted in the production of more male than female offspring (zebra finches: Gam et al. 2011; chickens: Pinson et al. 2011a). We would have expected that corticosterone injections in the same species (chickens) given at the same time (5 h) and at 4 h prior to the expected time of ovulation would stimulate a similar male bias, if any. However, none of our treatments in the current experiment truly mimicked the treatment that was given in Pinson et al. (2011a). In the current study, corticosterone concentrations in the 5 h HCORT birds had already dropped to 69 ng/ml, while the same dose given at the same time in Pinson et al. (2011a) and also at 4 h prior to the expected ovulation in the current study resulted in a concentration of 160 ng/ml at the same time point (1 h after injection). We cannot rule out that something went wrong when solutions were initially mixed at this time point; however, we also noticed that the rates of metabolism for birds at 5 vs. 4 h prior to the expected time of ovulation were quite different, suggesting that there was Fig. 3 a Comparison of the proportion of male offspring collected in Exp. I from eggs laid by hens in the untreated (UN), control oil (C), pharmacological corticosterone (HCORT) and physiological corticos- terone (LCORT) treatment groups that were treated 5 h prior to the expected time of ovulation. b Comparison of the proportion of male offspring collected in Exp. II from eggs laid by hens in the untreated (UN), control oil (C), physiological corticosterone (LCORT), and pharmacological corticosterone (HCORT) treatment groups that were treated 4 h prior to the expected time of ovulation. Statistical differ- ences between the treatment groups were determined using logistic regression analyses Fig. 4 Hormone profiles resulting from physiological and pharma- cological corticosterone injections at 5 and 4 h prior to the expected time of ovulation. Data from the HCORT injection at 5 h prior to ovulation indicated by the dotted line were taken from Pinson et al. (2011a), and all others from the current experiment 545 J Comp Physiol B (2015) 185:539–546 1 3 physiological variation in our hens between the 5 and 4 h time points prior to the expected time of ovulation in terms of how they metabolized corticosterone, thus it is also pos- sible that this variation was responsible for the disparity in HCORT-induced corticosterone concentrations between the two studies and between the two time points in the cur- rent study. In future studies, levels of corticosterone bind- ing globulins should also be assessed at these time points. However, the patterns of corticosterone in the current study and the related sex ratio effects suggest that both corticos- terone dose and time, in relation to chromosome segrega- tion and ovulation, that it is elevated are critical factors involved in sex ratio manipulation. It remains impossible at this point to isolate a mecha- nism by which these doses are working differently. We know that a high dose given 5 h prior to the expected ovula- tion stimulates the production of significantly more males, and a corticosterone treatment (when the high and low doses were combined) stimulated the production of more females in the current study. The dynamics of corticoster- one concentrations in this study and in Pinson et al. (2011a) are shown in Fig. 4. Linear interpolation of the Pinson et al. (2011a) data suggests that corticosterone concen- trations were elevated in the pharmacological range from 4.5 to 1.5 h prior to the expected time of ovulation. In our 5 h birds, only the HCORT dose hits the pharmacological range, and this likely dropped into the physiological range within 1.5 h after injection, meaning only concentrations in the Pinson et al. (2011a) study were pharmacologically elevated between 4 and 4.5 h prior to ovulation. Perhaps this is why only those hens produced a male-biased sex ratio after HCORT treatment. Our 4 h dose did hit the phar- macological range, but was also quickly metabolized and reached physiological levels by 2 h prior to the expected time of ovulation. The dynamics of corticosterone con- centration over time are shown in Fig. 4, in which we also included data from an earlier study (Pinson et al. 2011a), where hens had been injected with a high (pharmacologi- cal) dose (1.5 mg) of corticosterone at 5 h prior to ovula- tion. The time frame when there was a divergence in the levels of corticosterone between treatments that did elicit a bias (the Pinson et al. 2011a study and the two 4 h treat- ments in the current study) and those that did not (The two 5 h treatments in the current study) occurred just when sex chromosomes were supposed to segregate, at 2–4 h prior to ovulation. Perhaps this is when corticosterone needs to be elevated to a specific level. More work needs to be done to understand how the dosage and time may interact to pro- duce different effects on sex ratios. This study also highlights the importance of adequate sample sizes when conducting sex ratio studies. Because this work was conducted on chickens, we had the luxury of starting the study with 50–60 hens per treatment group, substantially larger sample sizes than in many published studies conducted on wild birds. However, because of the large confidence intervals associated with binary distribu- tions, even sex ratios as low as 38 % males were not sig- nificantly different from those with 50 % using our sample sizes. In addition, our 5 h control group contained 39 % males, but was not significant different from un-injected birds at that time point. Only the two corticosterone treat- ments at 4 h showed a significant effect, and that was only when they were combined to produce an adequate sample size to detect differences as low as 12 %. This suggests that it may be impossible to truly detect small differences in sex ratios using species from which large samples cannot be collected, and when reporting negative effects, we need to be careful to ensure that the power is large enough to detect differences that may be small, but still biologically relevant. The next step is to test the 4 h LCORT and HCORT treatments with higher sample sizes to determine whether both treatments induce female-biased sex ratios on their own. In addition, we need to examine levels of other hor- mones and mediators that have been shown to influence offspring sex, such as testosterone (Pinson et al. 2011a) and progesterone (Correa et al. 2005). Finally, we need to perform additional tests to determine exactly how hormone concentrations change after injections that occur 5 and 4 h prior to ovulation, particularly focusing on the 2–4 h win- dow prior to ovulation. If we can pinpoint precisely when corticosterone needs to be elevated to skew sex ratio in a particular direction, we can then examine how the pro- cess of meiotic segregation is changing at that time, per- haps by examining how corticosterone treatment changes expression of genes related to spindle fiber formation or movement. Ultimately, the goal is to understand how environmental and social challenges translate into physiological messages that can exert control of the sexes of the offspring. Results of several studies have demonstrated that long-term eleva- tions of corticosterone stimulate sex ratio skews. However, it is impossible to pinpoint the mechanism responsible using long-term treatments with corticosterone, and birds do not generally produce stable, long-term elevations of corticosterone in response to natural challenges, but instead produce numerous peaks of corticosterone over a short period of time (Rich and Romero 2005). Thus, it is impor- tant to understand how transient peaks of corticosterone influence sex ratios, and how these peaks change across the diurnal cycle. It appears that there is a critical level of corticosterone necessary to manipulate offspring sex, and that this level must be reached within a relatively short time window. This could explain the lack of extreme variation in avian sex ratios, and perhaps provides a level of control over how much environmental and social changes may influence sex ratio adjustment at the population level. 546 J Comp Physiol B (2015) 185:539–546 1 3 Acknowledgments We thank J. Cartmill and B. Wills for technical assistance. Financial support was provided by the US Poultry and Egg Association, Grant F027. This experimental protocol was approved by the University of Georgia Institutional Animal Care and Use Commit- tee, AUP #2008-10079. References Alonso-Alvarez C (2006) Manipulation of primary sex-ratio: an updated review. Avian Poult Biol Rev 17:1–20 Alonso-Alvarez C, Velando A (2003) Female body condition and brood sex ratio in Yellow-legged Gulls Larus cachinnans . Ibis 145:220–226 Arnold K, Griffiths R, Stevens D, Orr K, Adam A, Houston D (2003) Subtle manipulation of egg sex ratio in birds. Proc R Soc Lond B Biol Sci 270:S216 Aslam MA, Groothuis TGG, Smits MA, Woelders H (2014) Effect of corticosterone and hen body mass on primary sex ratio in lay- ing hen ( Gallus gallus ), using unincubated eggs. Biol Reprod. doi:10.1095/biolreprod.113.115352 Bonier F, Martin P, Wingfield JC (2007) Maternal corticosteroids influence primary offspring sex ratio in a free-ranging passerine bird. Behav Ecol 18:1045–1050 Clout MN, Elliott GP, Robertson BC (2002) Effects of supplementary feeding on the offspring sex ratio of kakapo: a dilemma for the conservation of a polygynous parrot. Biol Conserv 107:13–18 Correa S, Adkins-Regan E, Johnson PA (2005) High progesterone during avian meiosis biases sex ratios toward females. Biol Lett 1:215 Edens FW, Siegel HT (1975) Adrenal responses in high and low ACTH response lines of chickens during acute heat stress. Gen Comp Endocrinol 25:64–73 Etches R (1976) A radioimmunoassay for corticosterone and its application to the measurement of stress in poultry. Steroids 28:763–773 Etches R, Cunningham F (1976) The effect of pregnenolone, proges- terone, deoxycorticosterone or corticosterone on the time of ovu- lation and oviposition in the hen. Br Poult Sci 17:637–642 Fridolfsson A, Ellegren H (1999) A simple and universal method for molecular sexing of non-ratite birds. J Avian Biol 30:116–121 Gam AE, Mendonça MT, Navara KJ (2011) Acute corticosterone treatment prior to ovulation biases offspring sex ratios towards males in zebra finches Taeniopygia guttata J Avian Biol 42:253–258 Goerlich V, Dijkstra C, Schaafsma S, Groothuis TGG (2009) Testos- terone has a long-term effect on primary sex ratio of first eggs in pigeons—in search of a mechanism. Gen Comp Endocrinol 163:184–192 Goerlich-Jansson VC, Müller MS, Groothuis TGG (2013) Manipu- lation of primary sex ratio in birds: lessons from the hom- ing pigeon ( Columba livia domestica ). Integr Comp Biol 53:902–912 Johnson A (1996) The avian ovarian hierarchy: a balance between fol- licle differentiation and atresia. Poult Avian Biol Rev 7:99–110 Johnson A (2000) Reproduction in the female. Sturkie’s Avian Phys- iol 5:569–596 Kilner R (1998) Primary and secondary sex ratio manipulation by zebra finches. Anim Behav 56:155–164 Lambert J, Benoit B, Colvin G, Carlson J, Delville Y, Quesenberry P (2000) Quick sex determination of mouse fetuses. J Neurosci Methods 95:127–132 Nager R, Monaghan P, Griffiths R, Houston D, Dawson R (1999) Experimental demonstration that offspring sex ratio varies with maternal condition. Proc Natl Acad Sci USA 96:570 Navara KJ (2013) The role of steroid hormones in the adjustment of primary sex ratio in birds: compiling the pieces of the puzzle. Integr Comp Biol 53:923–937 Norris DO (2007) Vertebrate endocrinology, 4th edn. Elsevier, Bos- ton, p 313 Olsen M, Fraps R (1950) Maturation changes in the hen’s ovum. J Exp Zool 114:475–489 Pike TW (2005) Sex ratio manipulation in response to maternal con- dition in pigeons: evidence for pre-ovulatory follicle selection. Behav Ecol Sociobiol 58:407–413 Pike T, Petrie M (2003) Potential mechanisms of avian sex manipula- tion. Biol Rev 78:553–574 Pike T, Petrie M (2005) Maternal body condition and plasma hormones affect offspring sex ratio in peafowl. Anim Behav 70:745–751 Pike T, Petrie M (2006) Experimental evidence that corticosterone affects offspring sex ratios in quail. Proc Royal Soc B: Biol Sci 273:1093–1098 Pinson SE, Parr C, Wilson JL, Navara KJ (2011a) Acute corticos- terone administration during meiotic segregation stimulates females to produce more male offspring. Physiol Biochem Zool 84:292–298 Pinson SE, Wilson JL, Navara KJ (2011b) Elevated testosterone dur- ing meiotic segregation stimulates laying hens to produce more sons than daughters. Gen Comp Endocrinol 174:195–201 Rich EL, Romero M (2005) Exposure to chronic stress downregulates corticosterone responses to acute stressors. Am J Physiol Regul Integr Comp Physiol 288:R1628–R1636 Romero L, Reed J (2005) Collecting baseline corticosterone samples in the field: is under 3 min good enough? Comp Biochem Phys- iol-Part A: Mol Integr Physiol 140:73–79 Rutkowska J, Cichón M (2006) Maternal testosterone affects the primary sex ratio and offspring survival in zebra finches. Anim Behav 71:1283–1288 Rutstein A, Slater P, Graves J (2004) Diet quality and resource allo- cation in the zebra finch. Proc Royal Soc Lond Ser B: Biol Sci 271:S286–S289 Veiga J, Viuela J, Cordero P, Aparicio J, Polo V (2004) Experimen- tally increased testosterone affects social rank and primary sex ratio in the spotless starling. Horm Behav 46:47–53 Whittingham L, Dunn P (2000) Offspring sex ratios in tree swal- lows: females in better condition produce more sons. Mol Ecol 9:1123–1129 Whittingham L, Dunn P, Nooker J (2005) Maternal influences on brood sex ratios: an experimental study in tree swallows. Proc Royal Soc B: Biol Sci 272:1775 Wiebe K, Bortolotti G (1992) Facultative sex ratio manipulation in American kestrels. Behav Ecol Sociobiol 30:379–386 Wingfield JC, Farner D (1975) The determination of five steroids in avian plasma by radioimmunoassay and competitive protein- binding. Steroids 26:311–327