Introduction

The thickness of avian eggshells and their elemental composition (mostly trace/toxic chemical elements) are among the tools most frequently used in biomonitoring studies of bird eggs obtained from virtually all types of habitats worldwide. The eggs for such studies originate mainly from marine/ocean ecosystems (Burger 1994; Miljeteig et al. 2012; Kim and Oh 2014) to various terrestrial habitats (Mora 2003; Dauwe et al. 2005; Ding et al. 2019), including fresh water (Rodriguez-Navarro et al. 2002; Kitowski et al. 2017), urban (Hui 2002) and industrial areas (Migula et al. 2000). The avian eggshells (or their fragments, varying in size and presence of the eggshell membrane) often used in biomonitoring studies, such as those of Falconiformes (falcons, eagles), Charadriiformes (gulls, terns and waders), Galliformes and Passeriformes, exhibit very considerable variation in their colouring, which is manifested primarily by the presence of dark brown/red spotted areas (pigment spots) on the lighter (unmaculated) background/plain region of the shell.

The actual function of eggshell maculation continues to be a topic of intense debate, and both signalling (e.g. quality of parents, mechanism against brood-parasitic species) and structural explanations have been put forward (reviewed in Maurer et al. 2011; Brulez et al. 2016). Earlier studies suggested that eggshell maculation primarily performed a structural function related to Ca availability whereby pigment spots demarcated the thinner areas of the shell (reviewed in Gosler et al. 2005, 2011; Garcia-Navas et al. 2011; Cherry and Gosler 2010). Such localised eggshell thinning at pigment spots has been reported in a few species representing altricial (= passerines) (Gosler et al. 2005; Higham and Gosler 2006), semi-precocial (= Charadriiformes; including gulls and waders; Maurer et al. 2011; Bulla et al. 2012) and semi-altricial (= Falconiformes; Jagannath et al. 2008) species. This is invoked as evidence for the structural function of maculation in a mechanism for compensating Ca-deficiency and/or for strengthening the eggshell (see Gosler et al. 2005; Jagannath et al. 2008; Cherry and Gosler 2010). But on the other hand, localised eggshell thinning at pigment spots is not universal among all birds that lay maculated eggs, since the opposite pattern is found in galliform birds (precocial). The eggshells of the latter are thicker at the pigment spots, and the spots are formed mostly as a superficial layer, which can be manually rubbed off (Poole 1965; Baird et al. 1975; Orłowski et al. 2017, 2021; Rosenberger et al. 2018). However, it is not clear whether this is a consequence of the different reproductive strategy of Galliformes per se, or rather a consequence of the distant phylogenetic relationship of this order as a sister group to Neoaves (Eoet al. 2009). This could have resulted in different physiological processes of egg formation.

The brightly coloured eggs of the Japanese Quail (Coturnix japonica), i.e. the ones with pale brown shells and relatively small and more demarcated pigment spots, have thicker shells at the pigment spots than the dark eggs of this species, i.e. those with dark brown shells and more extensive pigment spots (Orłowski et al. 2017). A spot’s colour is also an indicator of the shell’s thickness, as is the case in passerines, in which darker spots marker yet thinner shell than paler spots (Gosler et al. 2005).

A common feature of all birds is that during embryonic development, Ca is mobilised exclusively from the innermost eggshell layer, i.e. the mammillary tips, whilst the two outer layers, i.e. the palisade (middle) and vertical (outermost crystal layer) layers, do not contribute to Ca resorption (Bond et al. 1988; Blom and Lilja 2004). However, across the spectrum of altricial–precocial birds, there are apparent differences in the structures of their eggshells. In small passerines, the eggshell is considered incomplete, as it lacks the vertical layer of larger species, and consists almost entirely of a highly vesiculated squamatic zone (Mikhailov 1997, discussed in Gosler et al. 2005). Bond et al. (1988) compared changes in the fine structure of the inner shell surface of hatched eggs of different bird species across the altricial–precocial spectrum. They found that fewer changes occurred in the eggshells of altricial species than in those of precocial species, the difference presumably being associated with a smaller Ca requirement for the developing altricial embryos. Karlsson and Lilja (2008) showed that the eggshells of precocial species have the highest number of mammillary tips, which entails a greater need for Ca in developing embryos, compared with altricial and semi-precocial species.

It has been suggested that the secretory products of the shell glands vary between different orders of birds (Board and Love 1980). Specifically, Board and Love (1980) claimed that in birds, there occur two distinct patterns of Mg concentrations across the eggshell thickness: that in Galliformes has two Mg peaks in the cone layer and the outer edge of the shell, whereas in the 19 other orders of birds there is only one Mg peak in the cone layer. Nevertheless, it should be stressed that these authors did not specify the advancement of embryonic development in these eggs (cf. Board and Love 1980). The distributions of three major micronutrients (Ca, Mg and P), and very likely certain trace elements as well, are not uniform across the eggshell thickness and are presumably species-specific (Board and Love 1980; Gonzalez and Hiraldo 1988; Damaziak and Marzec 2022). In part, this might be due to anatomical differences in the proportions of the different shell layers in various species (Panheleux et al. 1999). In particular, Mg and P tend to occur in greater concentrations in the outer shell layers, whereas Ca is deposited in the inner layers, which are eroded during embryonic development.

From a biochemical/elemental composition perspective, the key aspect is that the eggshell colouration of avian eggs involves a background base colour that is due to two pigments: protoporphyrin (responsible for brown and red colours) and biliverdin (responsible for blue and green colours) (Mikšik et al. 1996; Gorchein et al. 2009). Both these pigments can form chelates and bind various divalent metal cations, such as Fe, Zn, Cu, Cd or Hg, but almost all other metals can be incorporated into their molecules (Mikšik et al. 1996; Casini et al. 2003; Martínez and López-Rull 2024). Protoporphyrin and biliverdin occur in varying concentrations in the shells of most birds’ eggs, irrespective of their colouration (Mikšik et al. 1996). Both pigments occur concurrently in spotted and plain eggshell regions, although their distribution varies between these two regions (Mikšik et al. 1996; Cassey et al. 2012; discussed in Orłowski et al. 2017).

From the perspective of investigating elemental compositions and the resulting recommendations for studies in environmental assessments (including methodological issues), and also for drawing biological/ecological inferences in light of the varying habitat conditions experienced by breeding populations of birds, the most important point is that, apart from the eggshells of three precocial galliform species, the elemental composition of eggshells at the pigment spots and in the adjacent plain regions, and also the relationship between eggshell pigmentation and trace-element content remain unknown (cf. Mikšik et al. 1996). The few investigations in this respect that have been carried out in recent years are limited to galliform birds, whose eggshells differ anatomically from those of other bird orders (as stated above), and that the compositions of the majority of chemical elements examined vary significantly between the spotted and plain eggshell regions. Specifically, in Japanese Quail eggs with maternal levels of resources, i.e. freshly laid eggs prior to embryonic growth, the concentrations of major elements (Ca and Mg) and trace elements (Cu and Cd) were significantly higher in the spotted than in the adjacent plain eggshell regions. In particular, the Cd concentration was 2.5-fold higher in the pigment spots compared with the plain shell regions. Moreover, the Pb concentration was 3.1 times higher in the plain shell areas than in the spotted ones, and the Mn, Fe and Co concentrations measured in the spotted and plain shell regions were variable and dependent on egg colouration (Orłowski et al. 2017). Our later research involving eggshells of Black Grouse (Lyrurus tetrix) showed that concentrations of 25 out of 45 elements measured in shells of non-embryonated eggs varied significantly between the spotted and plain shell regions, and that the concentrations of most elements, including rare earths, were consistently higher in the spotted regions (Orłowski et al. 2021). A similar analysis for Capercaillie showed that the concentrations of eight out of 16 elements measured in the shells of non-embryonated eggs (Be, Cr, Mo, Ni, Ru, Sm, Zn and Zr), and as many as 14 out of 16 elemental concentrations measured in post-hatched shells (Al, B, Be, Cr, Fe, K, Mn, Na, Ni, Re, Ru, Sm, Tb and Zn) varied significantly between these two neighbouring shell regions. Almost all these elemental concentrations were higher in the spotted shell region in both non-embryonated and post-hatched eggshells (Orłowski et al. 2021). Our principal explanation for these differences is that the spotted eggshell region exhibits a greater propensity for accumulating different chemical elements, mostly via their binding to dark pigment (protoporphyrin) molecules, and consequently to the associated pool of metals present in spotted shell regions.

Retrospectively, these explanations are well established in earlier studies that highlighted a direct link between the darker pigmentation of eggs and elevated pollution levels (Jagganath et al. 2008; Hanley and Doucet 2012; Ding et al. 2019). Female birds in poor physiological condition and suffering a high level of oxidative stress, such as those exposed to pollution, may lay eggs with higher concentrations of protoporphyrin (Moreno and Osorno 2003; Hargitai et al. 2015 and references therein) or biliverdin or both pigments simultaneously (Jagganath et al. 2008; Hanley and Doucet 2012), which translates into the darker pigmentation and/or the greener background colour of such eggs, respectively. Finally, it is worth stressing that the incorporation of certain chemical elements into a spotted shell region may be a physiological mechanism enabling female birds to excrete excesses of some trace elements (see Burger 1994). Initially, this idea did not distinguish within-egg differences in the accumulation of elements relating to the presence of pigment spots or eggshell colour. Most importantly, however, the key question remains unanswered: whether variation in elemental composition between neighbouring eggshell regions of different colour, i.e. in spotted and plain areas, occurring in birds laying cryptic/maculated eggs and representing distant avian groups according to their mode of development (sensu Karlsson and Lilja 2008), is associated with the anatomical structure of the shells/pigment spots. This question should therefore be resolved by investigating a broader range of birds representing species from the three major developmental modes of birds, i.e. precocial, semi-precocial and altricial, with respect to the maculation of their eggs. Thus, the objectives of our study are twofold, i.e. to measure (1) the shell thicknesses, and (2) the concentrations of 6 staple macro/micronutrients (C, Ca, K, Mg, P and Na) and 11 trace elements (Fe, Cd, Pb, Hg, Co, Mo, Cu, Cr, Al, Mn and Zn) in the spotted and plain shell regions of the eggs of seven species of birds: four semi-precocial species (Black-headed Gull Chroicocephalus ridibundus, Mediterranean Gull Ichthyaetus melanocephalus, Sandwich Tern Thalasseus sandvicensis and Black Tern Chlidonias niger), two precocials (Western Capercaillie Tetrao urogallus, Black Grouse) and one altricial species (Great Reed Warbler Acrocephalus arundinaceus). The justifications for re-examining Western Capercaillie and Black Grouse eggshells in this study are i) to standardise the method(s) of analysing the elemental composition across all the samples of eggs of all seven species, and ii) to analyse three staple macro/micronutrients (C, Ca and P), not previously analysed in these species.

As the egg samples for this study were collected at random, they differed in embryonic development status. The analytical material thus included both unhatched eggs (with an undetermined embryonic development status) and post-hatched eggshells. We, therefore, aimed to determine the general differences in the elemental composition/shell thickness of two adjacent shell regions varying in colour, and to compare the levels of elements—in particular if these were high and of environmental concern—with previously published data measured in avian eggshells.

A well-known characteristic of eggshells is embryo-induced thinning (Castilla et al. 2010; Orłowski and Hałupka 2015), which leads to concurrent changes in their elemental composition (Orłowski et al. 2016; Ding et al. 2019). We emphasise, however, that our analytical material did not allow us to determine such detailed directional changes in the elemental composition/thickness of shells during embryonic development. This problem will require detailed studies in the future using precisely assigned samples of eggs/eggshells representing at least two extremes of embryonic development in various birds with different developmental modes. Ultimately, taking into account the results of earlier studies, which showed the significant impact of the presence of the inner shell membrane on the measurement of shell thickness (Maurer et al. 2012; Santolo et al. 2024) and the concentration of some eggshell elements (see Discussion), we have treated this problem very carefully in the data analysis and present inferences and recommendations for future studies.

Methods

The origin of the eggs/eggshells

The analytical materials used in this study were sampled (in different years between 2006 and 2018) in various localities in Poland, in most cases by the co-authors of this article, who were involved in the study of particular species. All the unhatched eggs were frozen until further processing. The post-hatched eggshells of terns and gulls were collected outside the nests of these species, so there was no additional interference in the nesting colonies of these birds.

Post-hatched eggshells (n = 4) of Black-headed Gulls, and post-hatched eggshells of Mediterranean Gulls (n = 10) were collected (by PK) during research on the reproductive biology and chick-ringing of the latter species on the Mietków Reservoir [Zbiornik Mietkowski] (50° 57′ 6.68″ N 16° 35′ 55.57″ E), SW Poland, conducted as part of the ongoing nationwide monitoring of the Mediterranean Gull (Zielińska et al. 2007). Additionally, after the Black-headed Gulls’ nesting colony had been inundated by flood water, their eggs (n = 12), floating on the surface of the water some distance from the colony, were sampled (GO). For the last 35 years, the area of the Mietków Reservoir was intensively used as a gravel pit (closed in December 2022), which resulted in elevated levels of trace metals (Pb, Cd and Cu) in the water, bottom sediments and the internal organs of the Black-headed Gulls nesting there (Orłowski et al. 2007).

Post-hatched eggshells (n = 15) of Sandwich Terns were sampled (SB) from the colony on the Mewia Łacha Bird Sanctuary reserve (54° 22′ 4.97″ N, 18° 56′ 44.89″ E), at the mouth of the River Vistula, N Poland, during a conservation project and the ringing of chicks of this species in 2014. There is a recent study on Hg accumulation in eggshells and egg contents of Sandwich Terns, and two other species, the Common Tern Sterna hirundo and European Herring Gull Larus argentatus, nesting in this area (Grajewska et al. 2015).

Post-hatched eggshells (n = 9) and deserted eggs (n = 4; floating on the water surface) of Black Terns were sampled (AG) in two nesting colonies situated in oxbow lakes in the Bug Valley (52° 39′ 54.73″ N, 21° 59′ 43.97″ E; 52° 40′ 40.71″ N 21° 54′ 38.11″ E), east-central Poland, and in one colony in the Stawy Milickie fish ponds (Żeleźniki, 51° 27′ 49.94″ N 17° 24′ 5.7″ E), SW Poland, where we sampled post-hatched eggshells.

The post-hatched eggshells of Western Capercaillie (n = 10) and Black Grouse (n = 10) were collected (DM) during an extensive reintroduction programme of these species in the Bory Dolnośląskie (Lower Silesian Forest) (51° 24′ 26.62″ N 15° 8′ 4.11″ E), SW Poland (for more details, see Merta et al. 2016; Orłowski et al. 2020).

Unhatched eggs (n = 10) of Great Reed Warbler were sampled (AD) on the Stawy Milickie fish ponds (Stawno, 51° 27′ 49.94″ N 17° 24′ 5.7″ E), SW Poland. They made up part of the analytical material used in earlier studies of eggs of this species (Drobniak et al. 2014).

Egg/eggshell processing and measurement

The unhatched eggs were thawed, then opened, and their contents poured out. After drying, the shells were stored in plastic containers at room temperature. Shell thicknesses were measured (by one co-author, JR), both with the attached inner shell membrane (if present) and without it to the nearest 10 μm using a micrometre (Insize 3580-25A) with a 0.2 mm spline diameter. If it was possible to remove the shell membrane, the thickness was measured without it. If a shell was thin or if the membrane adhered strongly to the shell interior (as in the case of the Great Reed Warbler eggs), removing the membrane caused the shell to crumble, so the thickness was measured without removing it.

The shell thickness was measured over the equatorial region of the egg, i.e. from approximately halfway between the equator of the egg and the blunt/sharp pole. The measurement was carried out in pairwise fashion, namely at a pigment spot and in the adjacent background colour region, the latter always being situated within c. 1 mm from the pigment spot. The measurements for each egg were made on three to six pigment spots and the same number of background colour regions.

Before being prepared for chemical analysis, all the eggshell samples were rinsed twice with doubly distilled water. For the chemical analysis, samples of relatively large pigment spots and adjacent background colour areas were cut out from each eggshell over the entire longitudinal section of the egg (from the sharp pole to the blunt one) using surgical instruments (tweezers, scissors, and scalpel), and then pooled into an Eppendorf tube to obtain the appropriate sample mass. These samples thus represented the shell material collected from one entire eggshell. Consequently, two paired samples were obtained from the pigmented and background colour regions of each eggshell, which were then analysed for their elemental composition.

Chemical analysis

The elemental composition was determined in air-dried eggshell samples of pigment spots and adjacent non-pigmented regions in the Laboratory of Biogeochemistry and Environmental Protection, Biological and Chemical Sciences Centre, University of Warsaw, Poland.

The sample weights used for the analysis (depending on the material supplied) ranged from 7 to 20 mg (the sample masses of particular species are available on request from the authors). In the chemical analysis, we tried to use only those eggshell fragments without inner membranes. Even so, some shell fragments were so small that it was often impossible to determine whether the membrane was present or not. Consequently, the small eggs of Great Reed Warblers and some other small eggshell fragments from which the eggshell membranes could not be removed (see above) were also included in chemical analysis.

The total C content was assessed by the combustion method in a CHNS Flash 2000 elemental analyser for sample weights of c. 1.5 mg. The concentrations of other elements were assessed by elemental analysis using spectral methods. Prior to these analyses, the samples were mineralised in 0.5 ml of 65% ultrapure nitric acid (Supelco Suprapur) in pre-cleaned PTFE containers at 200 °C using a Milestone Ultra Wave microwave apparatus, after which the mineralizates were diluted to a volume of 14 ml. The concentrations of Ca, K, Mg and Fe were determined by flame atomic absorption spectrometry (Contraa 700—Analytic Jena). A mixture of nitrous oxide and acetylene was used to measure the Ca concentrations, and a mixture of acetylene and air for K, Mg and Fe. The concentrations of the other elements, i.e. Na, Al, P, Ca, Cr, Mo, Mn, Co, Cu, Hg, Pb, Cd and Zn, were measured by inductively-coupled plasma mass spectrometry (ICP-MS Nexion 300d).

The quality control and accuracy of the elemental analysis was checked by determining the individual elements against standard reference materials. But because we had no certified reference material (CRMs) based on eggshells, we instead used CRMs based on other biological matrices, namely bovine liver (BCR1845R) and Polish Virginia Tobacco Leaves ICINCT (LGC standards).

The concentrations of the chemical elements were expressed in milligrams per kilogram (mg kg−1 or parts per million; ppm) of dry mass (d.w.), and those of C and Ca in %.

Statistical analysis

The primary aim of the statistical analysis was to assess the differences in (1) shell thickness and (2) concentrations of 17 chemical elements between the two adjacent eggshell regions varying in colour, i.e. the pigment spots and neighbouring plain eggshell regions in seven species of birds. To assess these differences, we used the t-test for dependent samples, where in a paired manner we compared the eggshell thickness data and elemental concentrations for each species. Some of our comparisons of thickness data were conducted for eggshells with an inner shell membrane, and these are summarised in Table 1.

Table 1 Eggshell thicknesses (average; 95% CL; number of measurements) measured in the equatorial egg region in plain areas and in adjacent pigment spots (expressed in mm) in seven species of birds, and the results of the T-statistic (t-test) for paired comparisons between these two areas of the same shell; n.d.—not determined (sample too small)
Full size table

The major aim of our t-test analysis was to assess the potential differences in eggshell elemental concentrations measured in the plain and pigment spot eggshell regions in seven bird species. Importantly, our two earlier papers had already documented an analogous relationship, namely, that there were significant differences in the concentrations of chemical elements between the pigment spots and plain eggshell regions (Orłowski et al. 2017, 2020). Formally, therefore, those earlier studies should be treated as having yielded preliminary outcomes, which were specified beforehand (sensu Streiner and Norman 2011). Thus, any correction for multiplicity or a multiple hypothesis test correction seems to be too conservative and should be avoided, also so as to avoid the risk of making a type 2 error (Pike 2011; Streiner and Norman 2011; Armstrong 2014). Consequently, we give the actual p-value for each t-test result (electronic supplementary material Table S1). As recommended, therefore, the reader can re-assess our results in the context of future research and its repeatability (Feise 2002; Armstrong 2014; Di Leo and Sardanelli 2020).

Finally, however, for documentary purposes and because of the large number of paired comparisons, we applied false-discovery rate (FDR)-based multiple comparison procedures to adjust the original p-values using the classical one-stage method (Benjamini and Hochberg 1995) in the statistical software spreadsheet (Pike 2011). The FDR-adjusted p-values were calculated from seven (species) tests (k = 7) comparing elemental concentrations between the plain and spotted eggshell regions. The results of the actual and FDR-adjusted p-values (q-values) are summarised in electronic supplementary material Table 1. Before the analysis, both the eggshell thickness and elemental concentrations were checked for normality (Kolmogorov–Smirnov test), and if necessary, data (only for some elemental concentrations) were transformed to meet the parametric assumption. As some of the concentrations failed to meet the assumptions of a normal distribution, we log-transformed the following variables/concentrations: Black-headed Gull (Cd, Al), Mediterranean Gull (Cd, Pb. Co, Al, and Mn), Sandwich Tern (Cd), Black Tern (Mo, Zn), Western Capercaillie (Mg, Cd, Cu), Black Grouse (Cu) and Great Reed Warbler (Mg, Hg, Co, Al, Zn). In addition, the percentage data of C concentrations in Western Capercaillie and Black Grouse were square-root transformed to improve the normality of their distribution.

In the statistical analysis for Fe concentrations below the detection limit, i.e. 0.03 ppm d.w. (n = 8, 4 and 3 for Sandwich Terns, Black Terns and Great Reed Warblers, respectively), this value was used for the calculation. This corrected value was at least 149 times lower than the lowest measured Fe concentration (4.46 ppm Fe d.w.) in these three species.

The statistical analyses were performed using Statistica ver. 7.0 (StatSoft 2006) and Excel software. The statistical significance level was 0.05.

Results

Eggshell thickness variability

Our egg/eggshell samples contained both unhatched eggs and post-hatched eggshells (Table 1). For two species—Black-headed Gull and Black Tern—the changes in eggshell thickness due to embryonic development could be assessed based on a general classification into unhatched eggs and post-hatched eggshells. Compared with the unhatched eggs without membranes, the post-hatched eggshells without membranes in Black-headed Gull were 6.63% and 7.77% thinner in the plain area and pigment spots, respectively; while in Black Tern, the analogous thicknesses were 6.48% and 6.54% less for eggshells without membranes, and 5.03% and 5.09% less for eggshells with membranes (Table 1).

In the seven species of birds, we observed two major patterns in eggshell thickness variation for an individual egg associated with the presence of pigment spots. Generally, these patterns remained consistent, despite the presence of an eggshell membrane strongly increasing the eggshell thickness (Table 1).

Pattern 1: eggshells are thinner at pigment spots than in adjacent plain areas

The first pattern, that eggshells are thinner at pigment spots than in adjacent plain areas (confirmed by the results of the t-test for paired comparisons), characterised Black-headed Gull eggs: 1.53% thinner in unhatched eggs, and 2.73% thinner in post-hatched eggshells (Table 1). In Mediterranean Gulls, the thickness of post-hatched eggshells without membranes was 1.98% less at the pigment spots, while in Sandwich Tern this value was 1.04% less without membranes and 2.27% less with membranes. Lastly, the eggshells of Great Reed Warblers at the pigment spots were 1.87% thinner than in the adjacent plain areas (Table 1).

Furthermore, in Black Terns, the shell thickness of unhatched eggs without membranes was 0.92% less at the pigment spots than in the adjacent plain areas, 1.27% less in unhatched eggs with membranes, 0.99% less in post-hatched eggshells without membranes, and 1.32% less in post-hatched eggshells with membranes (Table 1). However, the statistical analysis did not confirm these differences, most likely because of the small sample size of the particular groups of eggs.

Pattern 2: eggshells are thicker at pigment spots than in adjacent plain areas

The opposite pattern, that eggshells are thicker at pigment spots than in adjacent plain areas, was characteristic of the post-hatched eggshells without membranes of the two galliform species, i.e. Western Capercaillie and Black Grouse, in which at the pigment spots, the eggshells were 2.11% and 0.93% thicker, respectively (Table 1). Most probably, owing to the small sample size and/or potential variation due to the less intensive colouration, the differences in the Black Grouse eggshells were insignificant (Table 1). Our previous data on Black Grouse, however, based on the much larger sample of post-hatched eggshells (n = 84), confirmed that these were 3.26% thicker at the pigment spots (average ± SD = 0.190 ± 0.021 mm) than in the neighbouring plain region (0.184 ± 0.019 mm) (Orłowski et al. 2021).

Elemental composition of eggshells within the pigment spots and adjacent plain areas

Our measurements of the concentrations of 17 chemical elements in 280 eggshell samples from pigment spots and adjacent plain areas in the seven bird species (40 samples per species) resulted in 680 elemental concentrations for each species.

Statistical analysis using the t-test for paired comparisons showed that for all seven species, the concentrations of chemical elements within the pigment spots and the adjacent plain areas varied significantly (at p ≤ 0.05) in 41 out of the 119 pairs of comparisons (= 17 elements × 7 species) (Fig. 1). Once the multiple testing using the FDR approach had been checked, 27 of these pairs met the assumption of the FDR-rate adjusted p-value (Fig. 1; electronic supplementary material Table S1).

Fig. 1
figure 1

Average (± SE) concentrations of 17 chemical elements (C and Ca in %, the other elements in ppm d.w.) measured in maculated eggs of seven bird species in the plain regions (background colour) and the adjacent pigment spots; Black-headed Gull Larus ridibundus (BG), Mediterranean Gull Larus melanocephalus (MG), Sandwich Tern Sterna sandvicensis (ST), Black Tern Chlidonias niger (BT), Western Capercaillie Tetrao urogallus (CA), Black Grouse Tetrao tetrix (GR) and Great Reed Warbler Acrocephalus arundinaceus (RW); the asterisks denote the significance level obtained in t-tests for paired comparisons between plain regions and pigment spots on the same shells: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001; the results of paired comparisons that did not meet the FDR-adjusted p-values are shown in brackets (*), see details of the statistical analysis in electronic supplementary material Table S1

Full size image

In each species, we found statistically significant differences in elemental concentrations between these two eggshell regions, but the differences were specific to particular species and varied among the elements. Importantly, however, a clear and consistent trend of higher elemental concentrations in the pigment spots than in the plain eggshell regions emerged, particularly in the case of six elements: C, Ca, Pb, Cu, Cr and Al (Fig. 1).

The following statistically significant differences between the two eggshell regions were found for the various species.

Black-headed Gull: significant differences in concentrations between the plain region and pigment spots were detected for four elements: Ca, Cr, Mn and Na. Specifically, the concentrations of Ca, Cr and Mn were, respectively, 8.8%, 11.6% and 26.7% higher in the pigment spots than in the plain region, whereas the Na concentration was 4.3% higher in the plain eggshell region than in the pigment spots (Fig. 1).

Mediterranean Gull: the concentrations of eight elements (C, Ca, Mg, Fe, Cr, Al, Mn and P) varied significantly between the plain region and pigment spots. All the concentrations were higher in the pigment spots than in the plain region by 1.6% (C), 2.5% (Ca), 13.5% (Mg), 32,5% (Fe), 35.0% (Cr), 33.7% (Al), 57.4% (Mn) and 27.5% (P) (Fig. 1).

Sandwich Tern: the concentrations of three elements (C, K and Cr) varied significantly between the plain eggshell region and pigment spots. Specifically, the C and Cr concentrations were 9.6% and 22.2% higher, respectively, in the pigment spots than in the plain region, whereas the K concentration was 23.1% higher in the plain region than in the pigment spots (Fig. 1).

Black Tern: the concentrations of four elements (K, Fe, Mn and Cr) varied significantly between the pigment spots and the plain eggshell region. All these concentrations were 25.7% (K), 23.3% (Fe), 49.5% (Mn) and 10.1% (P) higher, respectively, in the plain region than in the pigment spots (Fig. 1).

Western Capercaillie: the concentrations of nine elements (C, Ca, K, Pb, Cr, Al, Mn, Zn and P) varied significantly between the two eggshell regions, and in all cases, the concentrations were higher within the pigment spot region. Specifically, compared to the plain eggshell region, the C concentration within the pigment spots was higher by 1.7%, and for the remaining elements the differences were as follows: Ca (9.6%), K (6.8%), Pb (33.8%), Cr (29.1%), Al (48.8%), Mn (30.8%), Zn (35.5%) and P (38.8%) (Fig. 1).

Black Grouse: here the findings were similar, in which concentrations of eight elements (C, Pb, Cu, Cr, Al, Mn, Zn and P) varied significantly between the two eggshell regions; in all cases, the concentrations were higher in the pigment spot region. Specifically, compared to the plain eggshell region, the C concentration in the pigment spots was higher by 0.93%, whereas for the remaining elements the differences were as follows: Pb (42.7%), Cu (90.6%), Cr (25.2%), Al (56.7%), Mn (54.1%), Zn (32.9%) and P (44.7%) (Fig. 1).

Great Reed Warbler: the concentrations of four elements (Mo, Cu, P and Na) varied significantly between the two eggshell regions. Only the Cu concentration was 16.8% higher in the pigment spot region. In contrast, the concentrations of Mo, P and Na were 28.2%, 25.9% and 8.9% higher, respectively, in the plain eggshell regions than in the neighbouring maculated ones (Fig. 1).

Discussion

Our study reveals for the first time in a quantitative manner the differences in eggshell thickness and elemental composition between pigment spots and adjacent plain eggshell regions in bird species from three developmental modes, i.e. altricial, semi-precocial and precocial. From a broader biological/biogeochemical perspective, our data increase knowledge of the heterogeneity of elemental composition between adjacent spotted-plain regions of eggshells, previously demonstrated in galliform birds, which differ in colour for many chemical elements. However, owing to the limited number of samples or species used in the analysis, it is impossible to outline the phylogenetic patterns for all extant bird orders or families.

These differences, however, are distinctive of particular species and vary among the elements. A clear and consistent trend of higher element concentrations within pigment spots compared to the plain eggshell region has emerged, particularly for five elements, namely C, Pb, Cu, Cr and Al. The results for the other elements are variable. In part, this could be due to the heterogeneity of the egg samples in terms of their embryonic development and the presence of the inner eggshell membrane (discussed below), as well as to differences in elemental composition between particular eggshell layers, principally those related to the varying distribution of three major elements (Ca, Mg and P) (Board and Love 1980; Gonzalez and Hiraldo 1988; Cusack et al. 2003; Damaziak and Marzec 2022).

The most important and novel finding of our study is that even though the eggshells of Charadrids i.e. gulls and terns (semi-precocial species) and of a passerine (altricial species) were thinner at the spots (hence there was less shell material) than those of the precocials (with thicker eggshells at the spots), the spotted regions of all these eggs contained disproportionally higher concentrations of most major and minor chemical elements. This, therefore, seems to be a general rule among all birds, regardless of the spot-plain eggshell thickness and/or their developmental mode. In other words, this finding confirms that the thickness of the eggshell itself does not affect the concentrations of elements, and that the key issue here is the presence of a protoporphyrin layer(s) along with the accompanying pool of chemical elements within the spotted eggshell region. At the same time, this finding provides overwhelming evidence that even in the relatively thinner eggshells and the smaller amount of shell material within the spotted region of the semi-precocial and altricial species compared with the precocials, the dark pigment present within the spots plays a crucial role as a carrier of both major and trace elements. In particular, our data include the concentrations of several elements representing the major mineral and organic constituents of avian eggshells, such as C, Ca, Na, Mg, P and K, measured in a direct quantitative manner (earlier data were based on electron probe analysis; Board and Love 1980; Cusack et al. 2003). However, it should be recalled that certain pigment patches, as in Charadrids (gulls and terns), are not all distributed externally, but can be located at varying depths within the shell thickness (Harrison 1966; Mikhailov 1997). Thus, it is impossible to determine the elemental composition in a specific layer of the eggshell as we used a sample consisting of material from a full cross-section of the entire eggshell. Moreover, it should be stressed that the eggshell samples used in the chemical analysis consisted of a mixture of both pigmented layers and unmaculated deeper eggshell layers. Thus, it was impossible to determine the actual elemental composition of the various eggshell layers.

In general, we hypothesised that, apart from basic biological/ecological differences like those related to diet, environmental pollution, habitat or the variable advancement of embryonic development, a working explanation for the inter-specific differences in elemental composition can also be based on anatomical differences and the accompanying different compositions, mostly of C, Ca, Mg and P, owing to the different thicknesses of the individual shell layers. This is highlighted by the fact that the overall highest concentrations of K, Mg, P, Na were detected in eggshells of an altricial species (a small passerine, Great Reed Warbler), in which the shell structure differs from other orders of birds, as there is no vertical crystal layer (external zone), present in the eggs of larger species, and is composed almost entirely of a highly vesiculated squamatic zone (Mikhailov 1997). Indeed, a recent investigation on two related thrush species (Turdus merula and T. philomelos) showed significant differences (based on SEM–EDS analysis) in the content of Mg, S and Si between the three major eggshell layers (Damaziak and Marzec 2022).

Moreover, we speculate that the relatively higher concentrations of K, Cr, Zn, P and Na in the shells of Great Reed Warbler eggs than in Charadrid eggshells could be due to the blue colour resulting from the higher content of bilirubin and/or biliverdin in the eggshells of the former (sensu Mikšik et al. 1996; Martínez and López-Rull 2024). Nevertheless, we wish to highlight the fact that the relationships between specific eggshell pigments, such as protoporphyrin, bilirubin and/or biliverdin under a high environmental contaminant load, including trace elements, deserve further detailed investigation.

Statistically significant differences were found for some of the elements, despite the relatively small differences between the two eggshell regions (as illustrated in Fig. 1), but this was due to the small scatter of the data. At the same time, while there were relatively large differences in the concentrations of some elements between both eggshell regions, e.g. Cd in Black-headed Gull (54.4% higher in the pigment spots) or Hg in Sandwich Tern (32.0% higher in the pigment spots), statistical analysis failed to confirm their significance. In part, this could also have been due to the small size of the eggshell sample used in the study.

The presence of the eggshell membrane

The simultaneous analysis of two different eggshell traits—thickness and elemental composition—was motivated by a search for general principles during the investigation of avian eggs. Although we found that maculation was very closely associated with differences in shell thickness and affected the elemental composition, the obscuring impact of the eggshell membrane needs to be elucidated. This is a critically important point in our discussion, in the context both of eggshell thickness measurements and of elemental composition. Such problems were highlighted in previous studies, in which both the thickness and/or elemental composition of eggshells were measured (Maurer et al. 2012; Grajewska et al. 2015; Peterson et al. 2017; Santolo et al. 2024). Generally, in most earlier studies, eggshell thicknesses were measured with the inner eggshell membrane attached (e.g. Maurer et al. 2012), this being due to the difficulty of removing this part of the egg. The data in Table 1 show that the presence of the eggshell membrane appears to increase the eggshell thickness by as much as 47% (e.g. in Black Tern). We believe that these results are most likely a reflection of uneven membrane detachment as the shell dries (see Maurer et al. 2012). These findings corroborate recent observations by Santolo et al. (2024), who reported that in 13 species from 7 families of waterbirds, the presence of the eggshell membrane increases the measured eggshell thickness by 7.9–20.6%.

More importantly, the inner eggshell membrane also has a key influence on the elemental composition of an eggshell. Certain trace metals within the eggshell, such as Hg, were contained in the inner membrane and the material adhering to the eggshell rather than in the calcium-rich hardened eggshell (Grajewska et al. 2015; Peterson et al. 2017). For instance, the study by Grajewska et al. (2015) on Sandwich Tern eggshells showed the following Hg concentrations in the eggshells (0.0006 ppm Hg d.w.), membrane (0.021 ppm Hg d.w.) and eggshell with membrane (0.0501 ppm Hg d.w.). Similar findings were obtained for American Avocet (Recurvirostra americana) membranes, in which Hg concentrations were 13.2 times higher than in the hardened portion of the eggshell without the membrane (Peterson et al. 2017). Grajewska et al. (2015) stress that the shell is composed mainly of calcium carbonate, whereas the membrane is made of keratin. The shell, being made up of inorganic substances, accumulates the least Hg. In our study, we measured respective average Hg concentrations of 0.073 and 0.096 ppm d.w. in the plain and spotted regions of Sandwich Tern eggshells (Fig. 1). These values are higher than those reported previously for this species (Grajewska et al. 2015).

Our further observations (M. Suska-Malawska, M, Sulwiński—unpubl.) strongly suggest that even a tiny sliver of membrane on an eggshell fragment was enough to increase the C content. In the Japanese Quail, the C content in an eggshell without membrane, eggshell with membrane and membrane was 15.5%, 16.3% and 46.0%, respectively; and in Domestic Chicken (Gallus gallus domesticus), the analogous values were 13.4%, 13.6% and 48.6%, respectively. Therefore, the amount of membrane in a sample can have a much greater impact on the result than the differences resulting from the presence of a spot. Eggshells consist mainly of calcium carbonate, so theoretically they should contain about 12% C. A result clearly above this value (e.g. 16–20%) probably implies that it was magnified by the presence of the eggshell membrane. In the case of some other elements, such as metal cations like Zn or Fe, which form chelates with shell pigments, the presence of the eggshell membrane should not affect the results (M. Suska-Malawska, M, Sulwiński—unpubl.). Thus, we strongly encourage assessing this problem in the future using eggs of a known developmental stage.

Our findings demonstrate that the presence of an eggshell membrane significantly affects the measurement of eggshell thickness. This indicates that the membrane, often adhering to the shell, plays a crucial role in determining the overall thickness recorded in such measurements, so highlighting the need to account for its presence in accurate eggshell thickness assessments.

Environmental concerns due to trace-element enrichment

In the context of the potential environmental concerns raised as a result of trace-element enrichment, and taking into account species-specific data (Fig. 1) along with previously published data on trace elements measured in avian eggshells (Burger 1994; Mora 2003; Dauwe et al. 2005; Kim and Oh 2014; Kitowski et al. 2017), our data indicate high shell concentrations of Cd, Pb and Cu in the eggs of Black-headed Gulls, and of Pb and Al in the eggs of Mediterranean Gulls, which nest sympatrically on the Mietków Reservoir. In general, however, eggshells of Black-headed Gulls from industrialised areas of southern Poland have much higher concentrations of Cd (up to 5.25 ppm Cd d.w.), Pb (up to 66.41 ppm Pb d.w.) and Cu (up to 9.64 ppm Cu d.w.) (cf. Migula et al. 2000). Moreover, we find high concentrations of Al in Mediterranean Gull eggshells, and significantly higher Al levels in the pigment spots (average = 121.5 ppm Al d.w.; range = 34.9–262.3 ppm Al d.w.) compared with the adjacent plain region (average = 90.9 ppm Al d.w.; range 20.7–247.8 ppm Al d.w.). Generally, literature data show that eggshell Al concentrations were much lower and rarely exceeded 34.8 ppm Al d.w. (cf. Kitowki et al. 2017). Although earlier articles put strong emphasis on the high toxicity of Al (mostly due to acidic precipitation) causing shell defects, they surprisingly lack information about actual Al concentrations in eggshells (Nyholm and Myhrberg 1977; Nyholm 1981), so that further inferences are not possible. Al becomes toxic at high dietary levels in excess of 1000 ppm d.w (reviewed in Sparling and Lowe 1996). We found that, even though Black-headed Gulls nested sympatrically with Mediterranean Gulls and had a diet similar to that of the latter species, their eggs had a nearly two- to three-fold lower Al shell concentration. It is possible, therefore, that females may accumulate Al in their bodies beyond the breeding grounds.

Furthermore, we measured high Mn concentrations in Black Terns—44.4 and 29.7 ppm Mn d.w. in the plain and spotted eggshell regions, respectively (Fig. 1). Such a high eggshell Mn concentration is most likely related to the distinctiveness of Black Tern breeding/foraging sites, i.e. oxbows, which are particularly susceptible to Mn accumulation, up to 1,190 ppm Mn d.w., in the bottom sediments (Obolewski and Glińska-Lewczuk 2013). Among the species compared here, we also obtained remarkably high Mo eggshell concentrations in Sandwich Terns of 0.626 and 0.670 ppm Mo d.w. in the plain and spotted eggshell regions, respectively (Fig. 1). However, these levels are relatively low compared with other biological material of animal origin (Eisler 1989) or even Baltic Sea water (Kunzendorf et al. 2001), they may indicate elevated Mo accumulation as a result of cyanobacteria and algal blooms enhanced by a higher nitrogen load (eutrophication) in the Baltic Sea.

Furthermore, our results suggest that the developing embryo can draws more Ca from non-spot areas than spot areas (as in Black-headed Gull), which could mean that at high level of heavy/trace metals, the spot areas might concentrate these pollutants.

To conclude, our results indicate a strong relationship between the biogeochemical background and the elemental composition of some of the chemical elements examined in this study. This implies that environmental factors play a key role in determining the specific elemental composition of the eggshells we examined.

Summary

The function of eggshell maculation is an intensely debated topic, and both signalling and structural explanations have been put forward (reviewed in Maurer et al. 2011). Our study appears to confirm the structural function of eggshell spotting, implying small-scale variation in elemental composition between neighbouring maculated and plain eggshell areas.

Our data reveal a relationship, practically unknown hitherto, between an egg’s colouration and the content of chemical elements in the shells of several species of wild birds. We highlight the fact that both eggshell thickness and elemental composition differ greatly between the spotted and plain regions of the shells in the species pool examined in this study. Nevertheless, further detailed investigation of other groups of bird orders having maculated eggs, such as Gruiformes, Falconiformes and Passeriformes, are highly desirable.

Valuable data were acquired in this study, which make an important contribution to the development of basic research in the field of biology, developmental physiology and ecotoxicology of birds, as well as in the context of the use of birds’ eggs in the biomonitoring of trace metals and minerals. Avian eggs, including those of species whose shells exhibit a very broad spectrum of colouration in the form of dark pigment spots (such as the eggs of gulls or terns), are often used in ecotoxicological studies and environmental monitoring. In general, however, the potential differences resulting from the variability of this colouration have been completely overlooked. The results of this project clearly demonstrate that there are significant differences in the concentrations of many chemical elements between the lighter background and the pigment spot regions. Therefore, it is important for future research on the chemical composition of avian eggshells that samples are standardised and that shells or fragments with a similar colour and pigment spot distribution are chosen.

Conclusions and research implications

Our data are useful for setting up further hypotheses, especially in the context of ecotoxicology, developmental physiology or population viability among birds with varying diets. Elucidating the significance of this would make such research an attractive contribution to the study of avian eggs.

Two major conclusions with clear further research implications emerged from our study.

First, that all chemical analyses of eggshells and measurements of their thickness require standardised eggshell sampling procedures in order to unify their colouration and embryonic status. We realise, however, that such an assessment of colouration is in many cases impossible, especially when only tiny eggshell fragments are available. We strongly recommend further multidisciplinary investigations of the basic relationships between the colouration and chemical composition of eggshells. They should include the eggs of different bird species with variable distributions of pigment layer(s)/spots within the shell thickness. We also highlight the need to measure other major chemical elements in the background colour and pigment spot regions in other groups of birds, and to link these with within-egg variability in shell thickness and size-related traits of eggs.

Second, in view of the fact that protoporphyrins and biliverdin occurred concurrently in both pigmented and unmaculated (plain) shell regions (Mikšík et al. 1996), further studies to determine the elemental composition of isolated shell material from both these shell regions and varying in colour are essential.