Eggshell coloration reflects both yolk characteristics and dietary carotenoid history of female mallards

Functional Ecology 2013, 27, 1176 1185 doi: 10.1111/1365-2435.12123 Eggshell coloration reflects both yolk characteristics and dietary carotenoid history of female mallards Michael W. Butler *, and Kevin J. McGraw School of Life Sciences, Arizona State University, PO Box 874601, Tempe, Arizona 85287-4601, USA Summary 1. Avian eggshell coloration has frequently been examined in a functional context (e.g. mimicry, camouflage), but in recent years, an interest has emerged in identifying the mechanisms that drive eggshell colour variation. 2. Eggshell coloration is predominately caused by pigment deposition; one such pigment is the antioxidant biliverdin, and deposition of biliverdin into eggshells may be costly to mothers due to depletion of their antioxidant reserves. Previous work has shown that dietary supplementation during laying with another type of antioxidant carotenoid pigments induces females to produce more biliverdin-rich eggshells. However, the impact of pre-laying nutrition including the developmental period early in life on eggshell coloration has not been investigated. 3. Here, we raised female mallards (Anas platyrhynchos) from hatching, supplemented their diets with carotenoids during early-, mid- or late-developmental periods, and at adulthood measured female circulating carotenoid levels, yolk carotenoid levels, and eggshell coloration. We found that carotenoid supplementation during the late stages of development (transitional period from juvenile to adult plumage) promoted the laying of eggs with more biliverdin-rich eggshells. Independent of developmental dietary treatment, females with higher circulating carotenoid levels at the time of egg laying produced more biliverdin-rich eggshells and more carotenoid-rich yolks. When controlling for female identity, we found that more biliverdin-rich eggshells were associated with more carotenoid-rich, but smaller, yolks. We also detected a laying order effect; later-laid eggs had larger, less carotenoid-rich yolks and less biliverdin-rich eggshells. 4. Taken together, these results demonstrate that eggshell coloration reveals carotenoid status of both mothers and yolks and that diet quality more than 1 month prior to laying can affect eggshell coloration in a waterfowl species. As mallards are considered to be capital breeders in terms of lipid stores, our findings provide a new developmental perspective on the carryover of lipid-soluble and antioxidant nutrient reserves for breeding. Key-words: Anas platyrhynchos, auto-communication, biliverdin, capital breeding, carotenoids, developmental plasticity, mallards, yolk size Introduction Behavioural ecologists have traditionally devoted most of their research attention on exaggerated traits to adult animals (e.g. sexual selection; Andersson 1994), but increasingly, studies are focusing on elaborate features of neonates and juveniles (e.g. Boncoraglio, Caprioli & Saino 2012). Such features include vocalizations (Dreiss, Lahlah & Roulin 2010), odours (Celerier et al. 2011) and colourful *Correspondence author. E-mail: butlermw@lafayette.edu Present address. Lafayette College, Easton, Pennsylvania, USA patches (De Ayala et al. 2007) that are used in offspring offspring or parent offspring communication (Shizuka & Lyon 2010). For oviparous species, such communication may occur even earlier in life, via coloration of eggshells. Eggshell coloration has long been viewed as a trait that enhances camouflage (Lee, Kwon & Yoo 2010), plays a thermoregulatory (Maurer, Portugal & Cassey 2011) or structural-integrity role (Gosler, Higham & James Reynolds 2005; Cherry & Gosler 2010), or is the product of a co-evolutionary arms race between brood parasites and hosts (Honza & Polacikova 2008). However, there is 2013 The Authors. Functional Ecology 2013 British Ecological Society

Developmental plasticity and eggshell coloration 1177 mounting interest in the intraspecific quality-signalling role of conspicuous eggshell colours (Moreno & Osorno 2003). For example, producing exaggerated shell coloration is argued to be costly and honestly reveals maternal quality and offspring quality (i.e. variation in egg contents) to fathers, who may vary paternal investment accordingly (English & Montgomerie 2010). To date, all attention on costliness of quality-revealing eggshell colours has been placed on the environmental and physiological conditions of the laying mother (e.g. Hargitai, Herenyi & T or ok 2008; Hargitai, Mateo & T or ok 2010; Honza et al. 2011). For example, investigators have examined body condition (Martınez-De La Puente et al. 2007) and manipulated dietary antioxidant levels (Morales, Velando & Torres 2011; Dearborn et al. 2012) of females during egg production and measured aspects of eggshell coloration (Morales, Velando & Torres 2011), egg size (Walters & Getty 2010), and yolk nutritional profile (Navarro et al. 2011). This emphasis on assessing or manipulating conditions during adulthood makes sense given that this is the time period when shells are pigmented and eggs are laid (shells are generally coloured within 24 h of laying; Sparks 2011), but it is also important to consider conditions during the long period leading up to egg production due to potentially longlasting effects of developmental plasticity (Monaghan 2008). There has been some work to date on developmental plasticity of pigment acquisition and utilization in birds, but it has largely focused on the diet-derived carotenoid pigments that are present in egg yolk (Blount et al. 2003). In contrast, eggshells are coloured by porphyrin pigments (Kennedy & Vevers 1976; Gorchein, Lim & Cassey 2009), which can be synthesized de novo (Maurer, Portugal & Cassey 2011) and thus may carry different costs and benefits of synthesis and allocation to eggs. Many studies on eggshell coloration focus on one class of porphyrins the bilins (e.g. biliverdin, which creates bluegreen shell colour; Riehl 2011) because they can have antioxidant properties (Kaur et al. 2003). Some have hypothesized that it may be costly for mothers to allocate this pigment to eggshells, and thus, colour intensity is a function of maternal antioxidant supplies (Hanley, Heiber & Dearborn 2008). In support of this relationship, biliverdin-based eggshell coloration reflects yolk antioxidant levels (e.g. in spotless starlings, Sturnus unicolor; Navarro et al. 2011), and supplementation of dietary antioxidants such as carotenoids to laying mothers can result in increased biliverdin-pigmented eggshell coloration (bluefooted boobies, Sula nebouxii; Morales, Velando & Torres 2011) and in yolks with increased carotenoid concentrations (zebra finches, Taeniopygia guttata; McGraw, Adkins-Regan & Parker 2005). Moreover, exposure to carotenoids early in life can affect adult carotenoid assimilation ability (Blount et al. 2003), yielding a system wherein nutritional conditions during development may affect adult pigment-associated traits, including allocation of carotenoids to egg yolk and of biliverdin to eggshells. Given the paucity of data on how pre-laying conditions of females affect eggshell coloration, we experimentally tested the effect of developmental nutrition on yolk and eggshell characteristics in female mallards (Fig. 1; Anas platyrhynchos). Mallard eggs range from grayish to greenish in colour (Drilling, Titman & McKinney 2002; Kreisinger & Albrecht 2008), and clutch sizes typically range from 6 to 13 (Drilling, Titman & McKinney 2002). This egg laying comes at a great cost to females, as they lose considerable lipid reserves and half of their body mass during this time (Krapu 1981). Some female mallards can lay up to three clutches in a year, with renesting occurring after prior nest failures (Arnold et al. 2010). We raised female mallards from hatch and supplemented their diets with carotenoids at three different points throughout development, allowing us to evaluate whether a critical developmental window exists for affecting adult egg-laying characteristics. After laying, we measured yolk and egg size, eggshell coloration, and yolk carotenoid concentration as additional indices of reproductive investment (Batt & Prince 1978; Eldrigde & Krapu 1988). First, we predicted that carotenoid supplementation would increase biliverdin-based eggshell coloration and yolk carotenoid concentration (i) based on the previously documented relationship between carotenoids and biliverdin-pigmented eggshells in birds (Morales, Velando & Torres 2011); and (ii) because female mallards are considered to be capital breeders in terms of lipid reserves (Boos et al. 2002), which may also include lipophilic carotenoids. If early-life carotenoid access shapes adult carotenoid physiology (Blount et al. 2003) more so than later-life carotenoid access, carotenoid supplementation early during development would have the greatest effects on egg characteristics. Alternatively, if recent access to carotenoids is more important, then carotenoid supplementation later in development (i.e. closer to adulthood) would have the greatest effects on egg traits. Because previous work with lesser black-backed Fig. 1. Female mallard in nuptial plumage (photo: M.W. Butler).

1178 M. W. Butler & K. J. McGraw gulls (Larus fuscus) and red-legged partridges (Alectoris rufa) showed that carotenoid supplementation to laying mothers had no effect on egg or yolk size (Blount et al. 2002; Bortolotti et al. 2003), but work with Japanese quail (Coturnix japonica) showed that yolks became smaller after carotenoid supplementation (McGraw 2006), we did not develop predictions related to the effects of supplementation on egg or yolk size. Lastly, to test for relationships between carotenoid-related traits in mothers and their eggs, we also quantified circulating carotenoid levels and carotenoid-pigmented bill coloration in females, which is a signal of sexual attractiveness in male mallards (Omland 1996a,b) but has no known signalling function in females. Materials and methods EXPERIMENTAL PROTOCOL AND BLOOD COLLECTION We acquired 46 one-day-old female ducklings from Metzer Farms (Gonzales, CA, USA) in December 2009 and housed them as in Butler & McGraw (2009, 2012). Ducklings were reared indoors in randomly selected groups of five ducklings per cage (60 9 60 9 60 cm) until they were 2 weeks old, then three per cage until they were 4 weeks old, and then two per cage until they were 7 weeks old, at which point all birds were moved outside and individually housed to allow for normal sexual maturation (Butler & McGraw 2009, 2012). Light : dark regime was 13L : 11D when ducklings were housed indoors, and natural photoperiod thereafter (105L : 135D at 7 weeks old to 135L : 105D at 20 weeks old). Individuals were randomly assigned to one of four treatment groups that varied in dietary carotenoid content. Individuals received carotenoid-supplemented diets during either the period of maximal growth (EARLY; 3 6 weeks old; N = 10; Butler & McGraw 2012), minimal growth and minimal nuptial plumage acquisition (MIDDLE; 8 11 weeks old; N = 12), or complete acquisition of nuptial plumage (LATE; 13 16 weeks old; N = 12). CONTROL (N = 12) birds did not receive carotenoid-supplemented diets at any point. We prepared diets by mixing a base diet of dry food (Mazuri Waterfowl Starter: Richmond, IN, USA, weeks 0 7; Mazuri Waterfowl Maintenance thereafter) with ORO-GLO dry pigmenter (2% carotenoids by mass, predominately lutein; Kemin AgriFoods North America, Inc., Des Moines, IA, USA) suspended in sunflower oil to achieve concentrations of 25 lg g 1 of carotenoids (upper quartile of carotenoids in mallard duckling diets in the wild; Butler & McGraw 2010). Whenever any treatment group was receiving carotenoid-supplemented diets, all other individuals received food mixed with sunflower oil as a sham control. We measured body mass to the nearest g four times during a 10-day adult immune assessment period that was part of a separate study (Butler & McGraw In Press). During this immune assessment period, we collected 600 ll of whole blood at four different time points (Days 0, 1, 6 and 10) that was used to quantify circulating carotenoid concentration (see below). We calculated mean adult body mass as the average of the four body mass values that were obtained concurrent with blood collection. Blood was stored on ice for several hrs, centrifuged for 3 min at 10 000 rpm and then stored at 80 C until analysis. EGG COLLECTION AND ANALYSIS We collected all eggs (N = 59) on a daily basis starting when the first female laid an egg (on 27 April; 20 weeks old) until all birds were euthanized at 225 weeks old for a separate study (Butler & McGraw In Press), thus creating a time period when only a subset of individuals were able to lay eggs (EARLY: N = 3 birds; MIDDLE: N = 4; LATE: N = 5; CONTROL: N = 2). Eggs were stored in a refrigerator for 1 week, at which point we used an Ocean Optics (Dunedin, FL, USA) USB2000 spectrophotometer with a PX-2 pulsed xenon light source to measure reflectance from k = 300 700 nm on 5 randomly selected areas of the eggshell. We binned all measurements into 1 nm increments using CLRfiles (CLR version 1.05; Montgomerie 2008) and then used CLRvars (CLR version 1.05; Montgomerie 2008) to calculate brightness (total amount of reflected light; B1) and saturation (proportion of reflectance within a specific portion of the spectrum, see below; S1U and S1v) for each egg. Biliverdin, which is the only pigment in eggshells that produces blue-green coloration (Sparks 2011), has a peak absorbance at 365 nm (Mizobe et al. 1997), so a decrease in eggshell reflectance around this wavelength relative to reflectance at other wavelengths should be due to increased biliverdin concentration. Thus, a reduced S1U and S1v (which measure the saturation in the 300 400 nm and 300 415 nm range, respectively) should correspond to a greater biliverdin concentration; in fact, this relationship has been demonstrated empirically in other bird species (Hargitai, Mateo & T or ok 2010) and corresponded with our visual assessment of eggshell coloration (M.W. Butler, pers. obs.). Additionally, this methodology is analogous to work with other pigments, wherein the colour metrics that are most closely associated with pigment concentrations are those that quantify saturation around the pigments peak absorbance (Butler, Toomey & McGraw 2011). We then weighed each egg to the nearest 001 g, measured egg width and length to the nearest 01 mm and used the formula described in Hoyt (1979) to calculate egg volume. We then cracked each egg into a translucent weighing boat and took digital photographs (Nikon Coolpix P3; Nikon Inc., Melville, NY, USA) of the egg contents with a ruler in the background. With Adobe Photoshop v. 8 software (Adobe Systems Inc., San Jose, CA, USA), we used the lasso marquee to encircle the yolk and measure the number of pixels occupied by the yolk (sensu McGraw 2006). We then measured the area standard in the photo so that we could convert pixel number to mm 2. Because we could not wholly separate yolk from albumen in these eggs, we were unable to quantify yolk mass. We prepared yolks for extraction by placing 200 ll of yolk in a screw-top 8-mL glass tube, drawing an additional 200 ll ddh 2 O into the same pipette tip to remove any yolk residue, and then vortexing the combined yolk and water mixture for 10 s. We stored this suspension at 80 C until further analysis. To extract carotenoids, we added 2 ml of ethanol to the thawed suspension, vortexed the solution for 10 s, added 2 ml of 1 : 1 hexane : methyl tert-butyl ether, vortexed the solution for an additional 10 s and then centrifuged the samples for 5 min at 3000 rpm. We then drew off the supernatant, evaporated it to dryness under nitrogen and resuspended samples in 42 : 42 : 16 methanol : acetonitrile : dichloromethane mobile phase for highperformance liquid chromatography analysis (see methods in McGraw, Tourville & Butler 2008). We primarily detected the carotenoids lutein, zeaxanthin, a lutein isomer and b-cryptoxanthin in mallard egg yolk. Because concentrations of each of these carotenoids were significantly positively correlated with total yolk carotenoid concentration (all r > 092, all P < 00001), we used total yolk carotenoid concentration in all further analyses. To estimate relative carotenoid amount per egg, we multiplied total yolk carotenoid concentration by yolk area (as determined from digital photographs). Because one egg contained only albumen and no yolk, sample sizes are reduced by one for yolk-dependent statistics.

Developmental plasticity and eggshell coloration 1179 PLASMA CAROTENOID AND BILL COLOUR ANALYSES We quantified plasma carotenoid concentration by extracting carotenoids from 50 ll of plasma using a 1 : 1 hexane : methyl tert-butyl ether method and analysing the extracts using high-performance liquid chromatography (McGraw, Tourville & Butler 2008). Detectable amounts of lutein, zeaxanthin and a lutein isomer were present in plasma from birds at all four time points of the immune assessment period, while smaller amounts of b-cryptoxanthin were frequently detected. Within each adult sampling time point, amounts of lutein, zeaxanthin, the lutein isomer and b-cryptoxanthin were positively correlated with total carotenoid titre (all r > 0624, all P < 00001). We thus calculated the average of total circulating carotenoid levels from these four samples as a metric of circulating carotenoid levels at adulthood. Carotenoid-based bill coloration in mallards begins to develop by 10 weeks of age (Drilling, Titman & McKinney 2002) and is completed in all birds by 16 weeks (M.W. Butler, pers. obs.; J. Metzer, pers. comm.). When ducks were 19 weeks old, we measured carotenoid-based bill coloration of adults using the spectrophotometer apparatus described previously. We measured a 1 cm band of the right dorso-lateral surface of the bill between the nares and the bill tip, and again used CLRfiles and CLRvars to calculate the brightness (B1), saturation (S1B: proportion of reflectance between 400 and 510 nm, the peak absorption area of carotenoids), and hue (H4b) scores that are most closely correlated with carotenoid content in the male mallard bill (Butler, Toomey & McGraw 2011). While male (yellow) and female (orange) bill coloration differs, both predominately use lutein and zeaxanthin as pigments (M.W. Butler and K.J. McGraw, unpublished data), and thus, we examined these same coloration metrics for female mallards. Pilot work from females in this study demonstrated that only S1B retained a significant relationship with carotenoid content of the integument in females (r = 0495, P = 005); thus, we report all associated statistics for this colour metric. We also performed analyses using both B1 and H4b, but neither variable explained significant variation in any analysis (all P > 015). Results REPEATABILITY OF EGG CHARACTERISTICS WITHIN FEMALES All egg characteristic metrics were statistically significantly repeatable within females (Table 1). Repeatabilities varied substantially among characteristics, ranging from 023 to 075. EFFECTS OF EARLY-LIFE DIETARY CAROTENOID ACCESS ON EGG CHARACTERISTICS Dietary carotenoid access during development did not significantly affect likelihood of laying (v 2 3 = 184, P = 061), but did significantly affect eggshell brightness, saturation in the ultraviolet region (S1U; Fig. 2) and saturation in the violet region (S1v; all results, Table 2). LATE birds had darker eggshells than all other treatment groups (all P < 0022; effect size relative to CONTROL = 103; EARLY = 115; MIDDLE = 129) and less ultraviolet- (effect size = 146) and violet-saturated (effect size = 129) eggshells than CONTROL (both P < 00044) and MIDDLE (both P < 0027; effect size S1U = 105, effect size S1v = 105) birds (Fig. 2). Dietary carotenoid access during development also significantly affected yolk carotenoid concentration (Table 2), with LATE females having more carotenoid-rich yolks relative to CONTROL birds (P = 00058; effect size = 125). Dietary carotenoid treatment during development did not significantly affect any other aspects of average egg characteristics (Table 2; all other effect sizes <10). STATISTICS These data have been deposited in the Dryad repository: http:// dx.doi.org/10.5061/dryad.pt22p. We performed all statistical analyses with SAS 9.2 (Cary, NC, USA), and residuals from all models were normally distributed. To test for inter-individual differences in egg characteristics, we calculated repeatability (Lessells & Boag 1987) for eggshell hue, brightness, and saturation in the ultraviolet (S1U) and violet (S1v) portions of the light spectrum, egg mass and volume, yolk area, and total yolk carotenoid concentration and amount. We performed a contingency table analysis to test whether likelihood of laying among all 46 females was affected by dietary carotenoid treatment. Within the subset of females that did lay eggs, we then used mixed models with individual as a random effect to test for (i) effects of dietary carotenoid treatment, adult body mass, adult bill coloration and average adult circulating carotenoid levels on eggshell brightness, ultraviolet- and violet-saturation, egg mass and volume, yolk area and yolk carotenoid concentration; (ii) relationships between eggshell coloration and egg characteristics (e.g. yolk carotenoid concentration, egg volume); (iii) relationships between yolk area and yolk carotenoid concentration; and (iv) laying order effects on eggshell coloration and egg characteristics. All comparisons between treatment groups utilized LSMeans estimates and standard errors. Effect sizes were calculated as Cohen s d. Due to relatively small sample sizes, our power to detect a significant effect may have been low in some analyses, and thus, any relationships that are reported as not significant should be interpreted within this context. Table 1. Intra-female repeatability of eggshell characteristics within a clutch for female mallard ducks. All egg characteristics were significantly repeatable (R; Lessells & Boag 1987) within female, although yolk carotenoid metrics were repeatable to a lesser degree than the other variables. See text for descriptions of variables Variable d.f. F-statistic P-value R Eggshell 13, 42 795 <00001 065 brightness (B1) Eggshell 13, 42 323 00020 037 saturation (S1U) Eggshell 13, 42 443 00001 048 saturation (S1v) Eggshell 13, 42 1204 <00001 075 hue (H4b) Egg mass 13, 41 538 <00001 054 Egg volume 13, 41 454 <00001 049 Yolk area 13, 41 898 <00001 068 Yolk carotenoid 13, 42 220 0027 024 concentration Total yolk carotenoid amount 13, 41 212 0034 023 S1U, saturation in the ultraviolet; S1v, saturation in the violet. For this and all subsequent tables, we use boldfaced text to indicate statistically significant P-values.

1180 M. W. Butler & K. J. McGraw S1U (unitless) 0 18 0 17 0 16 0 15 0 14 0 13 Control Early Middle Late Treatment Fig. 2. Differences in eggshell coloration as a function of access to carotenoid-supplemented diets during different periods of development in female mallard ducks (Least Squares Means estimates with 95% confidence interval). Individuals that received carotenoid-supplemented diets when 13 16 weeks old laid eggs with less ultraviolet-saturated eggshells (P < 005: *) than those without carotenoid-supplemented diets (CONTROL) and those that received carotenoid-supplemented diets during 8 11 weeks (MID- DLE) of age. There were no significant differences (all P > 005) among CONTROL, EARLY or MIDDLE birds in the ultraviolet saturation of their eggshells. Table 2. Differences in egg traits at sexual maturity in response to developmental dietary carotenoid manipulations in female mallard ducks. Dietary carotenoid supplementation during different periods of development significantly affected eggshell coloration and yolk carotenoid concentration, but not other egg characteristics, among females Variable d.f. F-statistic P-value Eggshell brightness (B1) 3, 43 354 0022 Eggshell saturation (S1U) 3, 43 393 0015 Eggshell saturation (S1v) 3, 43 332 0028 Egg mass 3, 43 249 0073 Egg volume 3, 43 168 019 Yolk area 3, 42 238 0084 Yolk carotenoid concentration 3, 42 292 0045 Total yolk carotenoid amount 3, 42 170 018 S1U, saturation in the ultraviolet; S1v, saturation in the violet. CORRELATIONS BETWEEN FEMALE METRICS AND EGG CHARACTERISTICS Neither average adult body mass nor female bill coloration was significantly related to egg mass or volume, yolk area or any metric of eggshell coloration (Table 3). Average adult circulating carotenoid level was significantly related to multiple egg characteristics; it was negatively related to eggshell saturation in both the ultraviolet (Fig. 3a) and violet portions of the spectrum, and positively related to total yolk carotenoid concentration (Fig. 3b) and yolk carotenoid amount (Fig. 3c; Table 3). Average adult circulating carotenoid level was not significantly related to egg mass or volume, yolk area or eggshell brightness (Table 3). * EGGSHELL COLORATION AS A PREDICTOR OF YOLK CHARACTERISTICS Eggshell coloration significantly predicted yolk carotenoid concentration, with brighter (F 1,41 = 668, P = 0013), more violet-saturated (F 1,41 = 956, P = 00036) and more ultraviolet-saturated (F 1,41 = 1014, P = 00028) shells associated with yolks that had lower carotenoid concentrations (Fig. 4a). Eggshell coloration also significantly predicted total yolk carotenoid amount. Eggs with more violet-saturated (F 1,41 = 634, P = 0016) and more ultraviolet-saturated (F 1,41 = 697, P = 0012) shells had yolks that were less carotenoid-rich (Fig. 4b). There was no significant relationship between eggshell brightness and total yolk carotenoid amount (F 1,41 = 287, P = 0098). Eggshell coloration also significantly predicted yolk area; brighter (F 1,41 = 2038, P < 00001), more violet-saturated (F 1,41 = 992, P = 00030) and more ultraviolet-saturated (F 1,41 = 1009, P = 000028) shells were associated with yolks that had larger areas (Fig. 5).Yolks with larger areas did not have significantly lower carotenoid concentrations (F 1,41 = 303, P = 0089). EFFECTS OF LAYING ORDER ON EGG CHARACTERISTICS Egg characteristics changed throughout the laying sequence. Eggshells became significantly brighter (F 1,42 = 509, P = 0029) and more saturated in both the violet (F 1,42 = 1111, P = 00018; Fig. 6) and ultraviolet (F 1,42 = 1510, P = 00004) portions of spectrum throughout the laying sequence, and yolks became significantly larger (F 1,41 = 1929, P < 00001; Fig. 7), but with lower carotenoid concentrations (F 1,41 = 3141, P < 00001) and lower overall carotenoid amounts (F 1,41 = 2514, P < 00001). Egg size also significantly increased, as assessed by both mass (F 1,42 = 831, P = 00062) and volume (F 1,42 = 836, P = 00060), throughout the laying sequence. Discussion We uncovered inter-individual differences in eggshell coloration and yolk carotenoid levels in female mallards that were due to dietary carotenoid access during development and that were correlated with circulating carotenoid levels just prior to laying. Yolk and eggshell characteristics also covaried with laying order, with eggshells becoming less biliverdin-rich and yolks becoming larger but less carotenoid-rich over the laying sequence, but still showing high overall repeatability within females. Finally, intensity of eggshell pigmentation was a reliable indicator of more carotenoid-concentrated but smaller egg yolks. Together, these results show that eggshell coloration is linked not only to egg contents, but also to female conditions during development, a period that occurred more than a month prior to egg laying.

Developmental plasticity and eggshell coloration 1181 Table 3. Maternal morphological and physiological traits as predictors of egg traits in laying female mallard ducks. Average adult circulating carotenoid levels, assessed several weeks prior to laying, predicted average eggshell coloration and yolk carotenoid concentration and amount. However, adult body mass and carotenoid-pigmented bill coloration during the same period did not predict any egg characteristics Independent variable Dependent variable d.f. F-statistic P-value Estimate Average adult body mass Eggshell brightness (B1) 1, 42 008 078 0019 Eggshell saturation (S1U) 1, 42 012 073 1E-6 Eggshell saturation (S1v) 1, 42 014 071 1E-6 Egg mass 1, 42 000 097 46E-4 Egg volume 1, 42 007 079 40E-3 Yolk area 1, 42 005 083 015 Yolk carotenoid concentration 1, 42 007 080 59E-3 Total yolk carotenoid amount 1, 42 015 070 104 Bill coloration (S1B) Eggshell brightness (B1) 1, 43 253 012 707 Eggshell saturation (S1U) 1, 43 209 016 028 Eggshell saturation (S1v) 1, 43 208 016 034 Egg mass 1, 43 017 068 423 Egg volume 1, 43 068 041 870 Yolk area 1, 42 178 019 59E3 Yolk carotenoid concentration 1, 42 009 076 505 Total yolk carotenoid amount 1, 42 054 047 15E5 Average adult circulating carotenoid level Eggshell brightness (B1) 1, 42 109 030 60 Eggshell saturation (S1U) 1, 42 683 0012 52E-3 Eggshell saturation (S1v) 1, 42 487 0033 59E-3 Egg mass 1, 42 079 038 105 Egg volume 1, 42 022 064 059 Yolk area 1, 41 051 048 399 Yolk carotenoid concentration 1, 41 1352 00007 398 Total yolk carotenoid amount 1, 41 1240 00011 54E3 S1U, saturation in the ultraviolet; S1v, saturation in the violet. Dietary supplementation of carotenoids during development affected eggshell coloration at adulthood, such that birds that received carotenoid-rich diets from 13 to 16 weeks of age laid eggs when they were 20 22 weeks old that had more biliverdin-rich (less bright, less ultravioletsaturated and less violet-saturated; Hargitai, Mateo & T or ok 2010) shells than those laid by both CONTROL and MIDDLE birds. It is important to note that our sample size in this analysis was modest, with only four MIDDLE birds and two CONTROL birds laying eggs during this time period; however, those two CONTROL birds laid a total of 20 eggs, suggesting that our calculation of within-individual egg characteristic metrics for these birds was robust. Together, these two results suggest that it was carotenoid supplementation during the LATE period in particular that drove differences in eggshell coloration, although we readily acknowledge that confidence in this finding would be greater with a more robust sample size. Unfortunately, our experimental design did not allow us to detect whether treatment-based differences in eggshell coloration were due to carotenoid supplementation during a critical period of development (i.e. 13 16 weeks old) or to recency of supplementation (i.e. approximately 6 weeks prior to egg laying). However, these findings augment the work of Morales, Velando & Torres (2011), corroborating a link between dietary carotenoid levels and eggshell coloration, but extend this relationship to include the ability of dietary carotenoid access over a month prior to laying to affect shell pigmentation. To our knowledge, this is the first study to show that nutrition during development and/ or significantly prior to laying can affect egg coloration. Egg characteristics, including shell coloration and yolk area, were highly repeatable within female mallards. Dearborn et al. (2012) also found that maternal identity was a strong predictor of eggshell coloration in chickens (Gallus gallus domesticus), even under standardized housing conditions and despite differences in feed. This high within-female repeatability is an important prerequisite for both condition-dependent signalling hypotheses (e.g. sexually selected eggshell coloration hypothesis; Moreno & Osorno 2003) and some condition-independent hypotheses (e.g. detection of brood parasitism). Such consistent interindividual differences may be driven by female genetics (Morales et al. 2010) or physiological conditions (Moreno 2006). Although we lack the data to test whether variation in genotype or gene expression was associated with differences in egg characteristics, we did uncover links between maternal physiological state and eggshell characteristics. Specifically, we found that average adult circulating carotenoid levels collected when individuals were 18 20 weeks old positively predicted both average yolk carotenoid concentration and biliverdin-rich eggshell coloration. Inter-individual differences in carotenoid assimilation or transport ability could lead to increased circulating carotenoid levels, and may thus be associated with an increased ability to deposit carotenoids into yolk (Karadas et al.

1182 M. W. Butler & K. J. McGraw (a) (a) (b) (b) (c) Fig. 4. (a) Yolk carotenoid concentration and (b) total yolk carotenoid amount as a function of eggshell coloration. Eggshells with less violet-saturated coloration (i.e. were more biliverdin-rich) contained yolks that had a greater carotenoid concentration and total carotenoid amount. In this and all subsequent figures, note that these are the raw data, uncorrected for female identity. Fig. 3. Average eggshell coloration for each female as a function of average adult circulating carotenoid levels within each female. Females with greater plasma carotenoid levels laid eggs that (a) were less saturated in the violet portion of the spectrum (300 415 nm; i.e. more biliverdin-rich), (b) had yolks with a greater carotenoid concentration and (c) had yolks with a greater total carotenoid amount. 2006). These higher levels of circulating carotenoids may also be linked to antioxidant status, either by directly acting as antioxidants themselves (Woodall, Britton & Jackson 1997; H~orak et al. 2007) or indicating repletion of other antioxidants (Hartley & Kennedy 2004). Because biliverdin can act as an antioxidant (Kaur et al. 2003), increased levels of circulating carotenoids may allow for a greater amount of biliverdin to be directed towards eggshell pigmentation without compromising maternal antioxidant status. The relationship between biliverdin-rich eggshell coloration and greater yolk carotenoid concentration held within individual eggs, even when controlling for female identity, further supporting the putative link between biliverdin production and carotenoid physiology. However, we also uncovered a relationship between more biliverdin-rich eggshells and smaller, but more carotenoid-rich, yolks. Although we did not detect an effect of carotenoid supplementation during development on yolk carotenoid levels, previous work has shown that feeding carotenoidrich diets to laying Japanese quail (Coturnix japonica) induced females to produce more orange (and presumably more carotenoid-rich), but smaller, yolks (McGraw 2006). Because both larger yolks (Dzialowski & Sotherland 2004) and more carotenoid-rich yolks (Saino et al. 2003) typically promote increases in offspring quality (e.g. larger individuals with more robust immune systems), it is unclear how the smaller, carotenoid-rich yolks laid by mallards that had more colourful eggs would affect duckling quality. This finding also raises the possibility that females who obtain a carotenoid-deficient diet may compensate for reduced yolk carotenoid concentration by increasing yolk size (but see Safran et al. 2007). Eggshell coloration, yolk size and yolk carotenoid concentration all varied according to laying order; later-laid eggs were larger, contained larger and less carotenoidconcentrated yolks and had shells that were less biliverdinrich. Such decreases in egg pigment or nutrient levels as a function of laying order are frequently ascribed to resource depletion (Rubolini et al. 2011). Under this framework, intensity of eggshell coloration may indicate either biliverdin (Lopez de Hierro & Neve 2010; but see Hanley & Doucet 2009) or carotenoid (Blount et al. 2002; Saino et al. 2002; Cassey et al. 2005; but see Newbrey et al.

Developmental plasticity and eggshell coloration 1183 Fig. 5. Yolk area as a function of eggshell coloration. Eggs with more violet-saturated (i.e. less biliverdin-rich) coloration contained larger yolks. Fig. 6. Eggshell coloration as a function of laying order. Shells from later-laid eggs were more violet-saturated (i.e. less biliverdinrich) in coloration. If the data point in the upper-right corner is removed, the slope remains significant (F 1,41 = 796, P = 00074). Fig. 7. Yolk area as a function of laying order. Later-laid eggs had larger yolks, although these yolks also had lower carotenoid concentrations, which resulted in an overall lower amount of carotenoids compared with earlier-laid eggs. 2008) limitations of mothers. However, our results also show that females may have been compensating for a reduction in yolk carotenoid concentration by producing larger eggs with larger yolks, and in birds, egg size is a significant predictor of not only size at hatching, but also juvenile survival (Krist 2011). Without more precise data regarding the relative benefits to offspring regarding both yolk carotenoid concentration and/or amount and egg or yolk size, it is difficult to assess how these competing metrics nested within laying order might affect offspring quality. It is possible that a continuum of egg size and carotenoid investment results in offspring that are more phenotypically diverse (i.e. differing levels of immunocompetence, variation in structural size), allowing females to produce a suite of offspring phenotypes that are differentially attuned to a diversity of future habitats (Monaghan 2008; i.e. bet-hedging: Childs, Metcalf & Rees 2010). Female mallards are considered to be capital breeders in terms of lipid reserves (Boos et al. 2002), relying on lipid stores acquired prior to arrival on the nesting grounds to produce clutches (Krapu 1981). However, few studies on nutrition and reproduction consider the developmental period in this capital-breeding context, and to our knowledge, lipid-soluble nutrients like carotenoids have never been subsumed into theories on capital-vs. income-breeding investments. Previous research in other species has demonstrated that carotenoid supplementation during laying yields an increase in yolk carotenoid levels (Blount et al. 2002; Karadas et al. 2006), suggesting an income-breeding investment with regard to these molecules. However, because we found that yolk carotenoid concentration and amount decreased throughout laying, and more importantly that pre-laying carotenoid status predicted subsequent eggshell coloration and yolk carotenoid levels, we suggest that female mallards also use carotenoids as capital resources during breeding. We encourage more work in this area, with an emphasis on the possibilities that other lipidsoluble yolk substances (e.g. vitamin E, steroids) may also be subject to a capital-resource-investment paradigm. Although the primary goal of this work was to evaluate developmental and nutritional mechanisms underlying egg characteristics, our findings have implications for how eggshell coloration may function in mallards. Unlike many of the species for which quality-signalling hypotheses for egg colour have been tested previously (i.e. where paternal care is predicted to respond to maternal egg colour variation; Moreno & Osorno 2003), male mallards provide no paternal care to offspring and in fact often abandon mating sites after insemination in anticipation of the upcoming molt (Drilling, Titman & McKinney 2002). Thus, because of their minimal to nonexistent opportunity to visually assess a female s clutch, it is highly unlikely that eggshell coloration functions as a signal to mallard fathers. However, mallard females engage in intraspecific brood parasitism, with almost one quarter of all nests experiencing egg dumping (Kreisinger et al. 2010). Thus, the high repeatability of eggshell coloration within females suggests that females could use inter-individual differences in eggshell coloration to increase likelihood of detection of brood parasites. Still, eggshell coloration also reflects internal

1184 M. W. Butler & K. J. McGraw yolk carotenoid concentration, demonstrating that eggshell coloration has the potential to signal aspects of egg quality. Hence, our data led to the creation of a new, autocommunication (or self-signalling) hypothesis of avian eggshell coloration (M. Meadows, pers. comm.); that is, a female may use information from her own eggshell coloration to make decisions about her subsequent reproductive investment. While our experiment was not designed to test this hypothesis, it can inform several critical predictions that may be applicable to many avian species. Under this hypothesis, we predict that females would invest more heavily in incubation of more colourful eggs or in post-hatch maternal care of offspring hatched from more colourful eggs. For example, under standardized thermal conditions, a female with more colourful eggs would spend a greater amount of time incubating and less time foraging than a female in similar body condition but with less colourful eggs, due to the greater expected benefit of maximizing optimal nesting conditions for high-quality eggs, even at a cost of self-maintenance. We also predict that self-signalling would be most beneficial in longer-lived species. In such multiply-breeding species, reduced investment in lower-quality young (i.e. reduced somatic cost to mothers) during one breeding season could permit increased investment during future breeding seasons. Lastly, we predict that self-signalling would have the greatest benefit in species that have low paternal care, forcing incubating females to directly allocate time and/or energetic reserves between egg-tending behaviours and self-maintenance behaviours, and thus maximizing the utility of information of egg quality to females. Indeed, correlations between eggshell coloration and female care of offspring (e.g. provisioning rate) have actually been reported before (e.g. in house wrens, Troglodytes aedon, Walters & Getty 2010), but we now must employ experimental approaches to determine whether females adjust maternal behaviour according to their egg colour signals per se. Acknowledgements We would like to thank C. Ross for animal and laboratory assistance, and R. Ligon, M. Meadows, D. Hanley, and two anonymous reviewers for helpful comments on an earlier draft of this manuscript. Funding was provided to MWB by the Animal Behavior Society and the Facilities Initiative Grant for Graduates in the School of Life Sciences and to KJM by the National Science Foundation (IOS-0746364). MWB was partially supported by the Arizona State University Graduate College Dissertation Fellowship during this study. All work with animals was approved under Arizona State University s Institutional Animal Care and Use Committee (protocol number 10-1094R). References Andersson, M. (1994) Sexual Selection. Princeton Universtiy Press, Princeton, NJ. Arnold, T.W., DeVries, J.H. & Howerter, D.W. (2010) Factors that affect renesting in mallards (Anas platyrhynchos). Auk, 127, 212 221. Batt, B.D.J. & Prince, H.H. (1978) Some reproductive parameters of Mallards in relation to age, captivity, and geographic origin. The Journal of Wildlife Management, 42, 834 842. Blount, J.D., Surai, P.F., Nager, R.G., Houston, D.C., Møller, A.P., Trewby, M.L. & Kennedy, M.W. (2002) Carotenoids and egg quality in the lesser blackbacked gull Larus fuscus: a supplemental feeding study of maternal effects. Proceedings of the Royal Society B: Biological Sciences, 269, 29 36. Blount, J.D., Metcalfe, N.B., Arnold, K.E., Surai, P.F., Devevey, G.L. & Monaghan, P. (2003) Neonatal nutrition, adult antioxidant defences and sexual attractiveness in the zebra finch. Proceedings of the Royal Society B: Biological Sciences, 270, 1691 1696. Boncoraglio, G., Caprioli, M. & Saino, N. (2012) Solicitation displays reliably reflect oxidative damage in barn swallow nestlings. Behavioral Ecology and Sociobiology, 66, 539 546. Boos, M., Zorn, T., Le Maho, Y., Groscolas, R. & Robin, J.-P. (2002) Sex differences in body composition of wintering Mallards (Anas platyrhynchos): possible implications for survival and reproductive performance: wintering female Mallards have a higher fat load than males which may compensate for their lower body size. Bird Study, 49, 212 218. Bortolotti, G.R., Negro, J.J., Surai, P.F. & Prieto, P. (2003) Carotenoids in eggs and plasma of red-legged partridges: effects of diet and reproductive output. Physiological and Biochemical Zoology, 76, 367 374. Butler, M.W. & McGraw, K.J. (2009) Indoor housing during development affects moult, carotenoid circulation and beak colouration of mallard ducks (Anas platyrhynchos). Avian Biology Research, 2, 203 211. Butler, M.W. & McGraw, K.J. (2010) Relationships between dietary carotenoids, body tissue carotenoids, parasite burden, and health state in wild mallard (Anas platyrhynchos) ducklings. Archives of Biochemistry and Biophysics, 504, 154 160. Butler, M.W. & McGraw, K.J. (2012) Developmental immune history affects adult immune function but not carotenoid-based ornamentation in mallard ducks. Functional Ecology, 26, 406 415. Butler, M.W. & McGraw, K.J. (in press) Immune function is related to adult carotenoid and bile pigment levels, but not dietary carotenoid access during development, in female mallard ducks. Journal of Experimental Biology, doi: 10.1242/jeb.082982. Butler, M.W., Toomey, M.B. & McGraw, K.J. (2011) How many color metrics do we need? Evaluating how different color-scoring procedures explain carotenoid pigment content in avian bare-part and plumage ornaments. Behavioral Ecology and Sociobiology, 65, 401 413. Cassey, P., Ewen, J.G., Boulton, R.L., Blackburn, T.M., Møller, A.P., Biard, C., Olson, V. & Karadas, F. (2005) Egg carotenoids in passerine birds introduced to New Zealand: relations to ecological factors, integument coloration and phylogeny. Functional Ecology, 19, 719 726. Celerier, A., Bon, C., Malapert, A., Palmas, P. & Bonadonna, F. (2011) Chemical kin label in seabirds. Biology Letters, 7, 807 810. Cherry, M.I. & Gosler, A.G. (2010) Avian eggshell coloration: new perspectives on adaptive explanations. Biological Journal of the Linnean Society, 100, 753 762. Childs, D.Z., Metcalf, C.J.E. & Rees, M. (2010) Evolutionary bet-hedging in the real world: empirical evidence and challenges revealed by plants. Proceedings. Biological Sciences/The Royal Society, 277, 3055 3064. De Ayala, R.M., Saino, N., Møller, A.P. & Anselmi, C. (2007) Mouth coloration of nestlings covaries with offspring quality and influences parental feeding behavior. Behavioral Ecology, 18, 526 534. Dearborn, D.C., Hanley, D., Ballantine, K., Cullum, J. & Reeder, D.M. (2012) Eggshell colour is more strongly affected by maternal identity than by dietary antioxidants in a captive poultry system. Functional Ecology, 26, 912 920. Dreiss, A., Lahlah, N. & Roulin, A. (2010) How siblings adjust sib sib communication and begging signals to each other. Animal Behaviour, 80, 1049 1055. Drilling, N., Titman, R. & McKinney, F. (2002) Mallard (Anas platyrhynchos). URL http://bna.birds.cornell.edu/bna/account/mallard/rec- OMMENDED_CITATION.html [accessed 29 December 2011]. Dzialowski, E.M. & Sotherland, P.R. (2004) Maternal effects of egg size on emu Dromaius novaehollandiae egg composition and hatchling phenotype. Journal of Experimental Biology, 207, 597 606. Eldrigde, J. & Krapu, G. (1988) The influence of diet quality on clutch size and laying pattern in Mallards. The Auk, 105, 102 110. English, P.A. & Montgomerie, R. (2010) Robin s egg blue: does egg color influence male parental care? Behavioral Ecology and Sociobiology, 65, 1029 1036. Gorchein, A., Lim, C.K. & Cassey, P. (2009) Extraction and analysis of colourful eggshell pigments using HPLC and HPLC/electrospray ionization tandem mass spectrometry. Biomedical Chromatography: BMC, 23, 602 606.

Developmental plasticity and eggshell coloration 1185 Gosler, A.G., Higham, J.P. & James Reynolds, S. (2005) Why are birds eggs speckled? Ecology Letters, 8, 1105 1113. Hanley, D. & Doucet, S.M. (2009) Egg coloration in ring-billed gulls (Larus delawarensis): a test of the sexual signaling hypothesis. Behavioral Ecology and Sociobiology, 63, 719 729. Hanley, D., Heiber, G. & Dearborn, D.C. (2008) Testing an assumption of the sexual-signaling hypothesis: does blue-green egg color reflect maternal antioxidant capacity? The Condor, 110, 767 771. Hargitai, R., Herenyi, M. & T or ok, J. (2008) Eggshell coloration in relation to male ornamentation, female condition and egg quality in the collared flycatcher Ficedula albicollis. Journal of Avian Biology, 39, 413 422. Hargitai, R., Mateo, R. & T or ok, J. (2010) Shell thickness and pore density in relation to shell colouration, female characteristics, and environmental factors in the Collared Flycatcher Ficedula albicollis. Journal of Ornithology, 152, 579 588. Hartley, R.C. & Kennedy, M.W. (2004) Are carotenoids a red herring in sexual display? Trends in Ecology and Evolution, 19, 353 354. Honza, M. & Polacikova, L. (2008) Experimental reduction of ultraviolet wavelengths reflected from parasitic eggs affects rejection behaviour in the blackcap Sylvia atricapilla. The Journal of Experimental Biology, 211, 2519 2523. Honza, M., Pozgayova, M., Prochazka, P. & Cherry, M.I. (2011) Bluegreen eggshell coloration is not a sexually selected signal of female quality in an open-nesting polygynous passerine. Die Naturwissenschaften, 98, 493 499. H~orak, P., Saks, L., Zilmer, M., Karu, U. & Zilmer, K. (2007) Do dietary antioxidants alleviate the cost of immune activation? An experiment with greenfinches. The American Naturalist, 170, 625 635. Hoyt, D.F. (1979) Practical methods of estimating volume and fresh weight of bird eggs. The Auk, 96, 73 77. Karadas, F., Grammenidis, E., Surai, P.F., Acamovic, T. & Sparks, N.H.C. (2006) Effects of carotenoids from lucerne, marigold and tomato on egg yolk pigmentation and carotenoid composition. British Poultry Science, 47, 561 566. Kaur, H., Hughes, M.N., Green, C.J., Naughton, P., Foresti, R. & Motterlini, R. (2003) Interaction of bilirubin and biliverdin with reactive nitrogen species. FEBS Letters, 543, 113 119. Kennedy, G.Y. & Vevers, H.G. (1976) A survey of avian eggshell pigments. Comparative Biochemistry and Physiology. B, Comparative Biochemistry, 55, 117 123. Krapu, G.L. (1981) The role of nutrient reserves in mallard reproduction. The Auk, 98, 29 38. Kreisinger, J. & Albrecht, T. (2008) Nest protection in mallards Anas platyrhynchos: untangling the role of crypsis and parental behaviour. Functional Ecology, 22, 872 879. Kreisinger, J., Munclinger, P., Javurkova, V. & Albrecht, T. (2010) Analysis of extra-pair paternity and conspecific brood parasitism in mallards Anas platyrhynchos using non-invasive techniques. Journal of Avian Biology, 41, 551 557. Krist, M. (2011) Egg size and offspring quality: a meta-analysis in birds. Biological Reviews, 86, 692 716. Lee, W.-S., Kwon, Y.-S. & Yoo, J.-C. (2010) Egg survival is related to the colour matching of eggs to nest background in Black-tailed Gulls. Journal of Ornithology, 151, 765 770. Lessells, C. & Boag, P.T. (1987) Unrepeatable repeatabilities: a common mistake. The Auk, 104, 116 121. Lopez de Hierro, M.D.G. & Neve, L. (2010) Pigment limitation and female reproductive characteristics influence egg shell spottiness and ground colour variation in the house sparrow (Passer domesticus). Journal of Ornithology, 151, 833 840. Martınez-De La Puente, J., Merino, S., Moreno, J., Tomas, G., Morales, J., Lobato, E., Garcıa-Fraile, S. & Martınez, J. (2007) Are eggshell spottiness and colour indicators of health and condition in blue tits Cyanistes caeruleus? Journal of Avian Biology, 38, 377 384. Maurer, G., Portugal, S.J. & Cassey, P. (2011) Review: an embryo s eye view of avian eggshell pigmentation. Journal of Avian Biology, 42, 494 504. McGraw, K.J. (2006) Dietary carotenoids mediate a trade-off between egg quantity and quality in Japanese quail. Ethology, Ecology & Evolution, 18, 247 256. McGraw, K.J., Adkins-Regan, E. & Parker, R.S. (2005) Maternally derived carotenoid pigments affect offspring survival, sex ratio, and sexual attractiveness in a colorful songbird. Naturwissenschaften, 92, 375 380. McGraw, K.J., Tourville, E.A. & Butler, M.W. (2008) A quantitative comparison of the commonly used methods for extracting carotenoids from avian plasma. Behavioral Ecology and Sociobiology, 62, 1991 2002. Mizobe, M., Kondo, F., Kumamoto, K., Kanda, Y. & Seguchi, H. (1997) High-performance liquid chromatographic analysis of bilirubin and biliverdin from jaundiced broilers. The Journal of Veterinary Medical Science, 59, 677 680. Monaghan, P. (2008) Early growth conditions, phenotypic development and environmental change. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 363, 1635 1645. Montgomerie, R. (2008) CLR: Colour Analysis Programs v1.05. Analysis. Queen s University, Kingston, ON, Canada. Morales, J., Velando, A. & Torres, R. (2011) Biliverdin-based egg coloration is enhanced by carotenoid supplementation. Behavioral Ecology and Sociobiology, 65, 197 203. Morales, J., Kim, S.-Y., Lobato, E., Merino, S., Tomas, G., Martınez-de la Puente, J. & Moreno, J. (2010) On the heritability of blue-green eggshell coloration. Journal of Evolutionary Biology, 23, 1783 1791. Moreno, J. (2006) Experimental evidence that egg color indicates female condition at laying in a songbird. Behavioral Ecology, 17, 651 655. Moreno, J. & Osorno, J.L. (2003) Avian egg colour and sexual selection: does eggshell pigmentation reflect female condition and genetic quality? Ecology Letters, 6, 803 806. Navarro, C., Perez-Contreras, T., Aviles, J.M., McGraw, K.J. & Soler, J.J. (2011) Blue-green eggshell coloration reflects yolk antioxidant content in spotless starlings Sturnus unicolor. Journal of Avian Biology, 42, 538 543. Newbrey, J.L., Reed, W.L., Foster, S.P. & Zander, G.L. (2008) Layingsequence variation in yolk carotenoid concentrations in eggs of yellowheaded blackbirds (Xanthocephalus xanthocephalus). The Auk, 125, 124 130. Omland, K.E. (1996a) Female mallard mating preferences for multiple male ornaments I. Natural variation. Behavioral Ecology and Sociobiology, 39, 353 360. Omland, K.E. (1996b) Female mallard mating preferences for multiple male ornaments II. Experimental variation. Behavioral Ecology and Sociobiology, 39, 361 366. Riehl, C. (2011) Paternal investment and the Sexually Selected Hypothesis for the evolution of eggshell coloration: revisiting the assumptions. The Auk, 128, 175 179. Rubolini, D., Romano, M., Navara, K.J., Karadas, F., Ambrosini, R., Caprioli, M. & Saino, N. (2011) Maternal effects mediated by egg quality in the Yellow-legged Gull Larus michahellis in relation to laying order and embryo sex. Frontiers in Zoology, 8, 24. Safran, R.J., Pilz, K.M., McGraw, K.J., Correa, S.M. & Schwabl, H. (2007) Are yolk androgens and carotenoids in barn swallow eggs related to parental quality? Behavioral Ecology and Sociobiology, 62, 427 438. Saino, N., Bertacche, V., Ferrari, R.P., Martinelli, R., Møller, A.P. & Stradi, R. (2002) Carotenoid concentration in barn swallow eggs is influenced by laying order, maternal infection and paternal ornamentation. Proceedings of the Royal Society B: Biological Sciences, 269, 1729 1733. Saino, N., Ferrari, R., Romano, M., Martinelli, R. & Møller, A.P. (2003) Experimental manipulation of egg carotenoids affects immunity of barn swallow nestlings. Proceedings of the Royal Society B: Biological Sciences, 270, 2485 2489. Shizuka, D. & Lyon, B.E. (2010) Coots use hatch order to learn to recognize and reject conspecific brood parasitic chicks. Nature, 463, 223 226. Sparks, N.H. (2011) Eggshell pigments from formation to deposition. Avian Biology Research, 4, 162 167. Walters, L.A. & Getty, T. (2010) Are brighter eggs better? Egg color and parental investment by House Wrens. Journal of Field Ornithology, 81, 155 166. Woodall, A.A., Britton, G. & Jackson, M.J. (1997) Carotenoids and protection of phospholipids in solution or in liposomes against oxidation by peroxyl radicals: relationship between carotenoid structure and protective ability. Biochimica et Biophysica Acta, 1336, 575 586. Received 12 February 2013; accepted 24 April 2013 Handling Editor: Tony Williams