Abstract
Identifying associations between phenotypes and environmental parameters is crucial for understanding how natural selection acts at the individual level. In this context, genetically isolated populations can be useful models for identifying the forces selecting fitness-related traits. Here, we use a comprehensive dataset on a genetically and ecologically isolated population of the strictly marine bird, the brown booby Sula leucogaster, at the tropical and remote Saint Peter and Saint Paul Archipelago, mid-Atlantic Ocean, in order to detect phenotypic adjustments from interindividual differences in diet, foraging behaviour, and nest quality. For this, we took biometrics of all individuals of the colony breeding in 2014 and 2015 and tested their associations with nest quality, diet parameters, and foraging behaviour. While body size was not related to the foraging parameters, the body size of the females (responsible for nest acquisition and defence) was significantly associated with the nest quality, as larger females occupied high-quality nests. Our findings suggest that the small breeding area, rather than prey availability, is a limiting factor, emphasizing the role of on-land features in shaping phenotypic characteristics and fitness in land-dependent marine vertebrates.
Similar content being viewed by others
Introduction
Identifying associations between phenotypic variability and local selective pressures at the population level is crucial for understanding population differentiation and intraspecific diversity, along with predicting sensitivity to environmental changes1,2. Nonetheless, gene flow and genetic drift may hinder the role of local pressures at the population level, making it necessary to identify the patterns of the population structure and stochastic processes before tracking population-specific drivers of local adaptation3,4. In other words, for closely related populations with contemporary genotypes shared by gene flow, phenotypic diversity is adjusted as a balance between adaptation to local conditions and heritage from successful immigrants; so that, the fitness-related traits of genetically isolated populations are mostly dependent on local selective pressures5. Therefore, isolated populations are influenced by local conditions rather than inter-population cohesive forces (e.g., gene flow) and can even deviate from ecogeographical rules (e.g., Bergmann’s rule6), making them useful models for exploring the role of environmental features as selective forces for phenotype adjustment in wild populations7.
The effects of spatiotemporal environmental differences on phenotypic diversity have been extensively demonstrated in natural populations, which have been used as open air laboratories for understanding microevolutionary processes at the individual level7,8. Seminal research projects have demonstrated clear associations between morphological traits and environmental correlates, including the effect of predation pressure shaping colour of male guppies in South American streams9, the effect of rainfall and seed availability on bill size of Darwin’s finches from the Galapagos10,11, and the effect of the perch height and diameter selecting number of subdigital lamellae and leg size of Anolis lizards in Caribbean islands12,13. Currently, a variety of research tools have been applied to identify individual fitness-related traits and their environmental correlates in evolving natural populations from fungal prevalence and infection intensity14 to molar shape of mammals on islands15. However, information on the drivers of local adaptation in marine vertebrates is still scarce.
Field studies addressing evolution issues include important trade-offs that are often absent in laboratory studies due to the inherent complexities found in the natural environment7. Combining traditional and innovative techniques makes it possible to investigate natural selection, even in free-living animals, such as seabirds. Despite their high mobility, seabirds are known to contain population structure associated with colony-specific environmental conditions16,17, presenting gene flow disruption even between sympatric populations18. In this context, genetically isolated seabird populations arise as interesting models for testing individual quality and detecting fitness-related traits due to their high intraspecific diversity, high philopatry, colonial breeding, and dependence on resource availability around colonies19.
The brown booby Sula leucogaster is a strictly marine bird that breeds on oceanic islands and is distributed in tropical and subtropical regions of all ocean basins, exhibiting a strong phylogeographic structure20,21. Brown boobies form nests in small ground depressions and forage on prey available around colonies, so that specialization on local resources can disrupt gene flow and act as immigration filter for boobies from adjacent breeding sites22. This is suggested to be the case of a brown booby colony located in the remote and tropical Saint Peter and Saint Paul Archipelago (hereafter “SPSP”), an emerged portion of the Mid-Atlantic ridge. Boobies breeding at SPSP were demonstrated to be larger and heavier than the remaining brown booby colonies in the southwest Atlantic Ocean6, in addition to present a quite disrupted gene flow in relation to these colonies22. In this context, local adaptation was suggested to explain population isolation of brown boobies from SPSP, indicating that local selective pressures may be promoting individual specialization and population differentiation22.
Due to its geographical location and geological origin, SPSP is influenced by oceanographic conditions similar to seamounts, which spatially constrains prey distribution and limits the potential distribution of predators. This small-scale oceanographic dynamics increases local productivity and concentrates biomass around the archipelago, supplying a rich top predator community of resident species, such as bottlenose dolphins Tursiops truncatus23, brown boobies, black noddies Anous minutus and brown noddies A. stolidus24, and migratory species, such as whale sharks Rhincodon typus25, blackfin tunas Thunnus atlanticus26, and oilfish Ruvettus pretiosus27. Additionally, the small emerged area makes the configuration of the brown booby colony distinct from the typical pattern observed for this species. Colonies of brown boobies are usually small and comprised of scattered, irregularly spaced groups28, but the nest density of SPSP is unusually high due to the limited available area for nesting. Approximately 500 individuals breed in less than 0.006 km2 29,30, and the resulting between-nest distance is only about 1 m31.
Given the environmental and population distinctiveness, we assessed the role of potential selective pressures in shaping phenotypes of brown boobies nesting at the SPSP. For this, we used an observational approach to extensively document the phenotypic variations, nest characteristics, diet, and foraging behaviour parameters. Our aim was primarily to identify current associations between the individual phenotypes and potential selective pressures rather than to describe population-specific microevolutionary processes, such as the strength and direction of selection, which require sampling over a longer time frame1. Genetic and phenotypic distances observed between brown boobies from SPSP and adjacent colonies (i.e., Fernando de Noronha Archipelago and Rocas Atoll) are suggested to be explained by local adaptations due to colony-specific selective pressures, highlighting the role of environmental features in isolating highly mobile top-predators6,22. Due to the comparatively large mean body size observed in brown boobies from SPSP6, we hypothesized that differences in the foraging parameters (i.e., diet and behaviour) and nest characteristics should be observed between the small and large individuals (i.e., a larger body size should be advantageous in some aspects in SPSP).
Results
Global results
In total, 319 brown boobies (160 females and 159 males) had their biometrics assessed. On average, females were larger and heavier than males: 5.1% for culmen length, 7.1% for tarsus length, 4.4% for wing chord, and 25.3% for body mass (Table 1). Stable isotope values ranged from −17.52 to −16.82‰ for δ13C and from 11.24 to 13.33‰ for δ15N, and the isotopic standard ellipses areas (SEAc) were similar for males (0.22‰2) and females (0.18‰2). In total, 72 stomach contents and 307 individual prey from 60 birds (36 females and 24 males) were analysed. The asymptote of the prey species richness was reached with 21 stomach contents and 55 prey. All the seven prey items were identified to the species level, and the tropical two-wing flyingfish Exocoetus volitans (hereinafter referred to as ‘TTF’) was the most important prey item (PSIRI = 62.8%), followed by the bigwing halfbeak Oxyporhamphus micropterus (hereinafter referred to as ‘BH’; PSIRI = 17.6%) (Table 1; Supplementary Fig. 1). Stable isotopes measured from muscle samples of the prey species ranged from −18.28 to −17.31‰ for δ13C and from 7.42 to 11.11‰ for δ15N (Supplementary Table 1). Due to its high prey-specific importance, the TTFs were separated into three food item categories according to their fork length: small (<100 mm), intermediate (100−150 mm), and large individuals (>150 mm), for both the MixSIAR and subsequent %PSIRI analyses. Mixing models from MixSIAR using the three most important food items (large and intermediate TTF, and BH) as potential prey presented similar results: the mean posterior probabilities with a 95% credibility interval for females were 69.7% for the large TTFs, 7.2% for the intermediate TTFs, and 23.1% for the BHs, while for males they were 59.1%, 5.9%, and 35%, respectively (Fig. 1; Table 1; Supplementary Fig. 2).
Foraging behaviour was recorded for 97 brown boobies (52 females and 45 males), summing to 258 trips. Mean (±1 standard deviation) trip duration was 57 ± 27.5 min, while mean maximum distance from the colony and the mean total trip length were 7.1 ± 4.3 km and 27.0 ± 13.7 km, respectively. Nest content had no significant effect on the foraging trip parameters and isotopic data for males or females; therefore, no individual was removed from the male and female datasets in subsequent analyses (Supplementary Fig. 3).
In total, all the 304 nests were sampled for their landscape characteristics. Colony area was calculated using a minimum convex polygon and resulted in 601 m2, representing a density of 0.51 nests.m−2. Nest altitude ranged from 10.7 to 20.7 m a.s.l., and the average between-nest distance was 1.03 ± 0.16 m. Nest altitude was negatively correlated with the mean between-nest distance (τ = −0.15; P = 0.0004), meaning that the nest density was higher in the high-altitude areas. Using 15 m a.s.l. as a threshold, the nest density in areas >15 m was 0.89 nests.m−2 (114 nests in 128.9 m2), and it was 0.63 nests.m−2 (190 nests in 301 m2) in areas <15 m a.s.l. Although information for all the 304 active and non-active nests was used to build the scheme for nest classification, only the 112 active nests were considered in the subsequent analyses, which were 26.2% classified as low-quality, 29.9% as intermediate, and 43.9% as high-quality nests (Fig. 2).
Intrasexual comparisons
Nest classifications were similar when comparing results from the hierarchical scheme and the scoring technique, so that 86% of the nests received the same classification in both approaches. Differences in the mean body size of females occupying nests of different quality were detected considering both the hierarchical scheme (F h = 9.67; P h = 0.0001) and the scoring technique (F s = 5.749; P s = 0.0043). Larger females predominated in high vs. low (P h = 0.0002; P s = 0.003) and high vs. intermediate (P h = 0.02; P s = 0.07) quality nests. In contrast, differences in the mean body size of males in relation to the nest quality were not detected (Fig. 3). Significant differences (P < 0.05) in the foraging trip parameters were not observed between the body size groups of males or females. In addition, significant correlation between the continuous body size index and trip parameters for either gender was not detected (Fig. 4; Table 1). In addition, no significant correlation or differences in the mean values were observed between the body size and δ13C and δ15N values, and the mean overlap of the Bayesian ellipses between the body size groups was 26% for females and 24% for males (Fig. 5). Prey item richness ranged from three to seven species among the body size groups of females and males, and TTF was identified as the food item of high prey-specific importance for both genders, corresponding to 79.6%, 90.7%, and 80.6% of the diet for the small, intermediate and large females and 57.4%, 41.7%, and 76.3% for the small, intermediate and large males, respectively (Table 1). There were no substantial differences in diet composition, inferred by the %PSIRI results, between the small, intermediate and large body sizes for the females or males, but comparisons between the %PSIRI results of the body size groups of the males had low power due to the small sample size (Table 1). Carbon and nitrogen isotopic ratios were statistically similar (P > 0.05) between the body size groups of males and females, and the contributions of potential food items generated by the mixing models were similar among groups (Fig. 1; Table 1). Finally, the highest importance of body size was observed in the regression trees explaining the variance in nitrogen for males (the third most important variable) and carbon for females (the second most important variable); the body size was not the most important variable in any regression tree (Supplementary Fig. 4).
Discussion
Significant differences in the mean body size were detected between the females nesting in low and high-quality nests, but associations of diet and foraging trip parameters with body size of both genders were not observed. Brown boobies breeding in SPSP have marked intersexual differences in nest attendance so that females dedicate significantly more time to incubation, brooding, and chick provisioning31. Furthermore, nest defence is also female-biased in boobies from SPSP31 and can be critical to ensure egg or chick safety and avoid extreme injuries, such as chick predation by cannibalism32. Considering the high colony density found in SPSP, we suggest that spatial constraints for nesting are the major driver for shaping fitness traits in brown boobies, positively selecting large females, which would benefit from their body size during fights for acquisition and defence of high-quality territories for nesting.
Interestingly, the normal distribution curve of female body size was more acute than that of males, but slightly skewed towards the large females, suggesting decreased phenotypic variability and reinforcing the idea of positive selection acting on intermediate and large females. The parameters we used to classify the quality of nests are closely related to breeding success in other colonial bird species33,34,35, and therefore, the predominance of small females in low-quality nests suggests a negative selection on this group. Indeed, mean body size and mass of brown boobies from SPSP are not explained by ecogeographical rules, such as the Bergmann’s rule, and are the highest among brown boobies breeding along the southwest Atlantic Ocean6, suggesting higher breeding success of the large and intermediate individuals in comparison to small ones, and increasing the average body size of birds in this population. Furthermore, the population is genetically distant from adjacent populations (i.e., Rocas Atoll and Fernando de Noronha Archipelago), which could be explained by an adaptation to the local seascape, causing phenotypic differentiation and, consequently, selection against immigrants22.
Absence of intrasexual differences in foraging behaviour could be associated with the abundant food around SPSP. Brown boobies from the Gulf of California spent ~2.5 h at sea during foraging trips and travelled almost 80 km on average, flying up to 28 km from the colony36. In the Caribbean, brown boobies spent ~6 h at sea during foraging trips, travelling an average of 129 km, up to 48 km from the colony37. The relatively small foraging range of brown boobies around SPSP, revealed by the short mean trip duration (57 min), mean maximum total distance (27 km), and maximum distance from the colony (7 km), suggests there is plenty of food available in the immediate surroundings of the archipelago. This is consistent with the seamount-like oceanographic dynamics observed around SPSP, which enhances local primary productivity on a very small spatial scale due to the interaction of the Equatorial Undercurrent with the submarine relief, increasing the residence time of nutrients and generating vortices and small-scale upwellings in the surroundings of the archipelago38,39. Though individual differences in foraging trip parameters in relation to the body size have been demonstrated for other booby species40,41, prey seems to be easily accessible for all boobies at SPSP, and thus obtaining food does not seem to be the main selection pressure on the phenotypes of boobies from SPSP.
Exocoetidae flyingfishes represented more than 95% of the material regurgitated by brown boobies as a whole, emphasizing the importance of this family to sustain a diverse epipelagic community around a tropical remote archipelago. Flyingfishes play a central role in making SPSP an important site for the tropical Atlantic Ocean epipelagic food web, as they spawn around the archipelago and represent approximately 80% of the fish larvae found in its surroundings during the northern summer, of which 73% is composed by the margined flyingfish, Cheilopogon cyanopterus42,43. Interestingly, the margined flyingfish, a 330 mm-length species44, has been identified as the most important prey species for wahoo Acanthocybium solandri, yellowfin tuna Thunnus albacares45, oilfish46 and brown booby47 in the SPSP, but in this study, it was not present in the regurgitated material of males and was only rarely found in the female diet. In turn, TTF is not considered to be an important prey for these top predators, and BH appears to be secondary in importance for seabirds; however, in our study, TTF was dominant in the regurgitated material of both male and female brown boobies. The margined flyingfish is abundant around SPSP from November to April, when it is believed to spawn and use rocks for egg mass attachment44. Indeed, the highest capture rates of pelagic fishes around SPSP are during this period, mainly through the exploitation of yellowfin tuna, which are attracted to the high concentrations of margined flyingfish, while the lowest capture rates are observed during the northern winter, when this study was carried out45,46. Therefore, we suggest that flyingfishes support breeding activities of brown boobies in SPSP, but the main prey species appears to alternate between margined flyingfish during the southern summer47 and TTF during the southern winter (as found here), demonstrating the consequences of the high primary productivity on the food web in SPSP throughout the year, despite this being a tropical remote archipelago surrounded by oligotrophic waters.
Studies addressing fine individual differences could benefit from research tools with even higher resolution and information provided by long-term monitoring programmes. We are convinced that there are individual covariates influencing brown boobies nesting at SPSP, such as personality48, senescence49, and early body condition50, which could influence intrapopulational parameters. Additionally, dataloggers capable of detecting behaviour at a higher resolution, such as accelerometers and time-depth recorders, could provide valuable information on additional attributes, such as underwater activities51 and daily energy expenditure52. Monitoring breeding success of small, intermediate and large individuals, as well as of females using low, intermediate and high-quality nests is crucial to understand how the limited breeding area is determining fitness of brown boobies at SPSP. The relatively long breeding cycle and the logistics to stay at the colony make it difficult, but remote monitoring could be carried out by using cameras, for example.
Nonetheless, our findings demonstrate that the available area in terrestrial environments could be a relevant limiting factor even for strictly marine vertebrates, such as seabirds. This is particularly relevant for seabird conservation, as habitat degradation and human disturbance of colonies have been attributed as major threats to seabirds globally53; thus, decreasing the area available for nesting could make territory disputes fiercer and influence which individuals will occupy high-quality nests. Other marine vertebrate species that depend on land for reproductive activities could also be affected by the individual ability to obtain high quality breeding locations. For example, it has been reported that grey seals Halichoerus grypus breeding on sand dunes do not have breeding site fidelity due to the changes and unpredictability of colony landscape54, but grey seals breeding on rocky islands have shown colony and breeding location fidelity, suggesting that colony landscape heterogeneity promotes interindividual disputes for high-quality breeding areas55. Additionally, climate change can strongly affect ice-dependent species by depleting the ice conditions required for breeding, increasing youth mortality and the frequency of sabbatical years, and forcing dispersal to more suitable breeding areas56,57. Over the past 100 years, the global mean sea level has risen by 0.19 m and is expected to increase ~0.5 m by 2100 compared to the current level58. Based on this estimate, the breeding activities of boobies nesting in the lowest areas at SPSP will be even more difficult than they currently are, possibly resulting in population declines due to both the reduction of suitable nesting areas and a trend of increasing average body size due to the intensification of disputes for high-quality nests. In this context, understanding current limiting factors, individual quality and population plasticity is crucial to predict how natural populations will cope with rapid environmental changes promoted by human disturbance in colonies, such as habitat degradation and global warming.
Material and Methods
Study area
Saint Peter and Saint Paul is a small remote tropical archipelago composed of ten islets with a total area of 0.017 km2 and located at 0°55′N and 29°20′W, on the mid-Atlantic ridge. The archipelago holds colonies of brown noddies, black noddies, and brown boobies, the latter breeding mainly on Belmonte Island, which has a maximum altitude of 21 m and an area of 6000 m2 30. Despite the spatial constraint, the brown booby colony on Belmonte Island has over 100 nests this study31, which are distributed on the sloping stones located on the northwest face of the island (Fig. 6). The colony is heterogeneous regarding landscape features; for example, the nests vary in altitude, distance from the sea, susceptibility to waves, and between-nest distances. As SPSP is located on the Equator, it is directly influenced by the seasonality of the Intertropical Convergence Zone, a low-pressure area that meets easterly surface winds forming a cloudy band with increased precipitation around the globe59. From February to May, the Intertropical Convergence Zone is between 7°S and 8°N and by April, the maximum accumulated precipitation reaches 370 mm on SPSP38. Furthermore, SPSP has a seamount-like seascape in its surroundings and is influenced primarily by the Equatorial Undercurrent, which flows eastward at a depth of ~80 m, slowing when it reaches the archipelago, generating vortices and small-scale upwelling that increase local primary productivity38,60.
Nest characterization
All the active and non-active nests (i.e., those only containing nest material such as small stones and feathers) were georeferenced using the Real-Time Kinematic (RTK) method with a preprocessed base point and a rover station (Topcon Positioning Systems, Inc.), which was placed at the central point of each nest. Non-active nests are also easily identified by the configuration of the nest material, making it possible to locate the centre of the nest to take the point. Radio-based communication between base and rover stations provided real-time corrections when triangulating to GPS satellites so that approximately 20 mm-resolution latitude, longitude, and altitude data were obtained for each nest. After recording data for 304 points (i.e., all the active and non-active nests) in 1580 m2, a three-dimensional elevation map was built with a resolution of 0.19 points.m−2. Additionally, the nests were individually numbered using photographs of the colony.
Given the landscape heterogeneity of the colony, the nest quality was determined using a combination of parameters, including global altitude, distance and altitude in relation to their neighbours, distance from the colony edge, and susceptibility to destruction by wave action obtained for all the 304 nests. From this, nests were classified folowing a hierarchical scheme considering the relevance of each metric for quality (Fig. 7). A threshold of 15 m above sea level (a.s.l.) was established as a criterion of quality, because during an extreme event in 2014, all nests below ~15 m were destroyed by waves, representing almost 80% of the active nests of that year (pers. obs.). Altitude is the most important feature in our scheme because nests >15 m are safe even during storms, adults can depart and arrive easily and safely, and nests do not get flooded during raining periods. Hence, nests >15 m were classified as high-quality nests. Colonial breeding has been suggested to be a strategy against predators, and chick survival seems to be higher in central nesting sites35,61. Peripheral nests were defined as those within a 3 m-range from the outer limit of the colony. Within peripheral nests, there are some nest locations that are usually destroyed by waves, mainly during syzygy tides. SPSP is primarily influenced by the north branch of the South Equatorial Current, which flows northwestward, and by the southeast trade winds, so that waves predominantly reach the archipelago in its south-southeast portion60. However, only the southern part of the colony is exposed to waves, as the east-eastern portion of the colony is protected by the island (Fig. 6). Therefore, nests located in the southern portion of the island were classified as low-quality nests, since they are usually destroyed during syzigy tides. The position relative to neighbouring nests was defined by comparing the altitude of each nest to those of the three nearest nests. This is important because nests in lower places are more susceptible to flooding, soaking eggs and chicks. Nests in the peripheral area with the lowest local altitudes were also classified as low-quality nests. For each nest, distances from the three nearest nests were taken to represent the average between-nest distance. The average between-nest distance for nests <15 m in altitude was 1.2 m (see the ‘Results’ section). Therefore, a nest below the threshold of 15 m was classified as of high quality if it was located outside the peripheral area, it was not at the lowest local altitude among neighbours, and it had an average between-nest distance >1.2 m. All nests that did not fit into the parameters for low or high quality nests were classified as intermediate quality nests.
As an additional approach, each metric was categorized by a binomial score (0: unfavourable, 1: favourable). The five scores were summed up and scaled in order to vary between 0 and 3 (nest quality score (NSQ) = sum of each metric score *3/5). Low quality nests were those with NSQ within [0;1], intermediate quality nests those with NQS within [1;2] and high quality nests those with NSQ within [2;3].
Sampling scheme
This study was performed in accordance with the Guidelines to the Use of Wild Birds in Research of the Ornithological Council. In addition, all the sampling protocols were approved by The National Center for Bird Conservation and Research (Permit No. 3208) and the Chico Mendes Institute for Biodiversity Conservation (ICMBio Permit No. 38723). During June 2013, May–June 2014 and June–July 2015, morphometrics and body mass measurements from breeding boobies were obtained after capturing individuals on the nests by hand or handnet. Body mass was obtained to the nearest 20 g using Pesola® spring scales, and morphometrics were measured as follows: culmen length (exposed culmen) and tarsus length (from middle of the midtarsal joint to the distal end of the tarsometatarsus) with Vernier calipers (0.01 mm) and wing chord (carpal joint to the tip of the longest primary; unflattened wing) with a metal ruler with stop (1 mm). The gender was determined by differences in the skin coloration around the eyes, and the adults were distinguished from juveniles by plumage coloration28. After sampling, each booby was identified with a numbered tarsal metal ring to avoid resampling and aiming monitoring demographic parameters.
Because there is a strong and reverse sexual size dimorphism in brown boobies28, during the fieldwork carried out in 2015, boobies for sampling were divided into subgroups based on their body size and each gender was treated as a distinct dataset. To generate a global body size index, the wing chord, culmen and tarsus length, and body mass were standardized by subtracting the mean and dividing by the standard deviation and synthesized through a Principal Component Analysis. The first principal component (PC1), which explained 66.1% of the total variance, was adopted as a body size index and used for sorting individuals. To group the individuals in relation to their body size, males and females PC1 (body size index) were ranked and split into 20% quantiles: the first quantile (0–20%) was treated as the smallest individuals, the third quantile (40–60%) was treated as intermediate individuals, and the fifth quantile (80–100%) was classified as the largest individuals (Supplementary Fig. 5). Because adults in distinct breeding stages were sampled, the non-parametric Kruskal-Wallis test was used to test for the effect of nest content on the foraging behaviour and diet parameters. The chicks were aged as follows: N1 (0–3 weeks); N2 (4–6 weeks); N3 (7–11 weeks); and N4 (12 weeks to fledgling)62.
Foraging behaviour
For studying foraging movements, miniaturized high-resolution GPS units with integrated chip antennae and rechargeable batteries were used (12.5 g, 19 × 25 × 5 mm; Technosmart, Rome, Italy). Loggers were set to record one position every second when deployed for one day and one position every 10 seconds when deployed for two days. During 30 days in July 2015, boobies were captured for logger deployment between 04:00 and 05:00 h, and GPS transmitters were attached to the three central rectrices using TESA tape. Loggers were packaged within heat-shrinkable waterproof tubes, and the total mass of the equipment (i.e., tube, tape, and logger) did not exceeded 3% of the body mass of the lightest brown booby at SPSP (i.e., a male of 1185 g), as recommended for seabirds63. After logger recovery, data were downloaded using the dedicated GiPSy-4 Utility software (Technosmart, Rome, Italy). Logger deployment and recovery (when biological samples were taken) lasted ~5 and 10 min, respectively, and in both handling procedures, the bird was returned immediately to its nest to minimize stress.
Diet parameters
Diet of boobies was studied during the same aforementioned period of tracking by analysing stable isotopes of carbon and nitrogen in the blood serum and the regurgitated material. For serum isolation, 3-ml blood samples were taken from the tarsal vein during GPS recovery using a sterile syringe/needle and transferred to a non-heparinized tube, which was kept undisturbed at room temperature for 10–30 min to allow for blood clotting. After clotting, the tubes were centrifuged for 25 min at 3000 RPM to isolate 1 ml of the serum, which was then transferred to 2-ml tubes using clean pipette tips and frozen at −4 °C or lower until processing in the laboratory. Bird serum has an isotopic half-life <5 days64 and, thus, the carbon and nitrogen isotopic ratios obtained in this study correspond to the period immediately prior to remote tracking. Spontaneous regurgitation of the stomach content is a usual behaviour in Sulidae species during stressful situations28. During handling, the nondigested regurgitated material was collected and identified at species level, and the fork length was measured using a ruler with 1-mm accuracy. Additionally, muscle samples from prey items, which have an isotopic half-life of approximately 30 days for fish, were taken and stored in anhydrous ethanol for stable isotope analysis65.
In the laboratory, muscle samples were washed in a Soxhlet extractor during three sessions of 8 h each to remove lipids using a 2:1 chloroform:methanol solution as solvent66. Muscle and serum samples were freeze-dried and homogenized, and subsamples of ~0.7 mg were transferred into tin capsules to be analysed using mass spectrometry (measurement precision of 0.2‰) at the Stable Isotope Core, Washington State University (USA). Differences between the sample ratios and the international reference standards (Vienna Pee Dee Belemnite limestone for carbon, and air for nitrogen) were expressed in δ notation as parts per thousand (‰)67.
Statistical analyses
The accuracy of GPS positions used for tracking was <10 m in more than 95% of the location fixes, allowing reconstruction of the foraging trips and extraction of high-resolution foraging parameters, such as the total distance travelled (‘D’, km), maximum distance to the colony (‘Dmax’, km), time spent at sea (‘T’, h), sinuosity of the path (‘Sin’, defined as the ratio D.2Dmax−1), and average flight speed (‘FS’, km.h−1).
Carbon and nitrogen isotopic ratios were used for calculating the isotopic niche width through the Bayesian framework implemented in SIBER package68. Standard ellipse areas (‰2) corrected for small sample sizes (SEAc) were estimated for each body size category along with the pairwise overlap percentage value between ellipses. Proportions of prey contribution to the diet of each group were estimated with Bayesian mixing models as implemented in MixSIAR package69. For this, isotopic ratios of the three most important prey species found in the %PSIRI analysis (see below) were used as source data, and trophic enrichment factors were 1.1 ± 0.5 for δ13C and 2.8 ± 0.5 for δ15N, following an experimental study with a piscivorous seabird and lipid-free tissues70. Posterior distributions for each prey item for each group were obtained after running 1,000,000 MCMC, discarding the first 500,000 as ‘burn-in’69.
Characterization of boobies diet from regurgitated material followed the Prey-Specific Index of Relative Importance (‘%PSIRI’)71, which is based on prey-specific parameters (i.e., ranging from >0% to 100%), such as frequency of occurrence, prey-specific abundance, and prey-specific mass or volume. Because prey may be sampled in various stages of digestion, individual mass was predicted from the fork length by using previously published Bayesian regression parameters for elongated-body fish species72.
Differences of boobies body size between the categories of nest quality (considering both approaches to classify nests) were tested with univariate Analysis of Variance (ANOVA), and normality and homoscedasticity of the residuals were tested with Shapiro-Wilk’s and Levene’s tests, respectively. Similarly, the intrasexual differences between the foraging parameters and stable isotopes were tested with the nonparametric Kruskal-Wallis test. As intrasexual relationships between body size and behavioural and diet variables were expected to be non-linear, two distinct approaches were used to test for the association between the isotopic data and foraging trip parameters with PC1. First, intrasexual differences of foraging behaviour and diet were tested with the nonparametric Kruskal-Wallis test by using body size groups as factors. The second approach was based on correlations between the continuous body size index and behavioural and diet variables by using the nonparametric Kendall’s tau (τ), which measures the correlation between two ranked variables73. Finally, regression trees were used to identify the importance of the body size groups by explaining the variance in each foraging and diet parameter. For classifying variables according to their importance at each regression tree, the Random Forest algorithm implemented in the R package randomForest was applied74 by generating 1000 regression trees with 20 per-tree permutations for out-of-bag data. The classification procedure was carried out by ranking variables according to the resulting error rate caused when each variable was removed from the model (i.e., the higher the error rate was, the higher the importance of the variable was).
Data availability
Data used in this study is deposited in Dryad Digital Repository.
References
Siepielski, A. M., DiBattista, J. D. & Carlson, S. M. It’s about time: the temporal dynamics of phenotypic selection in the wild. Ecol. Lett. 12, 1261–1276 (2009).
Grant, P. R. & Grant, B. R. 40 years of evolution: Darwin’s finches on Daphne Major Island (Princeton University Press, 2014).
Kawecki, T. J. & Ebert, D. Conceptual issues in local adaptation. Ecol. Lett. 7, 1225–1241 (2004).
Bolnick, D. I. & Nosil, P. Natural selection in populations subject to a migration load. Evolution 61, 2229–2243 (2007).
Mayr, E. Geographical character gradients and climatic adaptation. Evolution 10, 105–108 (1956).
Nunes, G. T., Mancini, P. L. & Bugoni, L. When Bergmann’s rule fails: evidences of environmental selection pressures shaping phenotypic diversification in a widespread seabird. Ecography 40, 365–375 (2017).
Reznick, D. N. & Ghalambor, C. K. Selection in nature: experimental manipulations of natural populations. Integr. Comp. Biol. 45, 456–462 (2005).
Hendry, A. P. & Kinnison, M. T. An introduction to microevolution: rate, pattern, process. Genetica 112, 1–8 (2001).
Reznick, D. & Endler, J. A. The impact of predation on life history evolution in Trinidadian guppies (Poecilia reticulata). Evolution 36, 160–177 (1982).
Price, T. D., Grant, P. R., Gibbs, H. L. & Boag, P. T. Recurrent patterns of natural selection in a population of Darwin’s finches. Nature 309, 787–789 (1984).
Grant, P. R. & Grant, B. R. Unpredictable evolution in a 30-year study of Darwin’s finches. Science 296, 707–711 (2002).
Losos, J. B. & Sinervo, B. The effects of morphology and perch diameter on sprint performance of Anolis lizards. J. Exp. Biol. 145, 23–30 (1989).
Glossip, D. & Losos, J. B. Ecological correlates of number of subdigital lamellae in anoles. Herpetologica 53, 192–199 (1997).
Lambertini, C. et al. Local phenotypic variation in amphibian-killing fungus predicts infection dynamics. Fungal Ecol. 20, 15–21 (2016).
Ledevin, R. et al. Phylogeny and adaptation shape the teeth of insular mice. Proc. R. Soc. Lond. B, Biol. Sci. 283, 20152820 (2016).
Friesen, V. L., Burg, T. M. & McCoy, K. D. Mechanisms of population differentiation in seabirds. Mol. Ecol. 16, 1765–1785 (2007).
Friesen, V. L. Speciation in seabirds: why are there so many species…and why aren’t there more? J. Ornithol. 156, 27–39 (2015).
Hailer, F. et al. Long-term isolation of a highly mobile seabird on the Galapagos. Proc. R. Soc. Lond. B, Biol. Sci. 278, 817–825 (2011).
Schreiber, E. A. & Burger, J. Biology of marine birds (CRC Press, 2001).
Morris-Pocock, J. A., Anderson, D. J. & Friesen, V. L. Mechanisms of global diversification in the brown booby (Sula leucogaster) revealed by uniting statistical phylogeographic and multilocus phylogenetic methods. Mol. Ecol. 20, 2835–2850 (2011).
Carboneras, C. Family Sulidae. In Handbook of the birds of the world. Vol. 1. Ostrich to ducks (eds del Hoyo, J., Elliott, A. & Sargatal, J.) 312–325 (Lynx Editions, 1992).
Nunes, G. T. & Bugoni, L. Local adaptation drives population isolation in a tropical seabird. J. Biogeogr. 45, (in press) (2018).
Milmann, L. C., Danilewicz, D., Baumgarten, J. & Ott, P. H. Temporal–spatial distribution of an island-based offshore population of common bottlenose dolphins (Tursiops truncatus) in the equatorial Atlantic. Mar. Mamm. Sci. 33, 496–519 (2017).
Mancini, P. L. & Bugoni, L. Resources partitioning by seabirds and their relationship with other consumers at and around a small tropical archipelago. ICES J. Mar. Sci. 71, 2599–2607 (2014).
Macena, B. C. L. & Hazin, F. H. V. Whale shark (Rhincodon typus) seasonal occurrence, abundance and demographic structure in the mid-equatorial Atlantic Ocean. PLoS One 11, e0164440 (2016).
Bezerra, N. P. A., Travassos, P., Hazin, F. H. V., Viana, D. L. & Macena, B. C. L. Occurrence of blackfin tuna Thunnus atlanticus, Lesson 1931 (Scombridae) in Saint Peter and Saint Paul Archipelago, Brazil. Pan. Am. J. Aquat. Sci. 6, 68–70 (2011).
Viana, D. L. et al. Diet of the oilfish Ruvettus pretiosus (Perciformes: Gempylidae) in the Saint Peter and Saint Paul Archipelago, Brazil. Braz. J. Oceanogr. 60, 181–188 (2012).
Nelson, J. B. Pelicans, cormorants and their relatives: the Pelecaniformes. (Oxford University Press, 2005).
Barbosa-Filho, R. C. & Vooren, C. M. Abundância, estrutura etária e razão sexual do atobá-marrom Sula leucogaster (Pelecaniformes: Sulidae) no Arquipélago de São Pedro e São Paulo, Brasil. Rev. Bras. Ornitol. 18, 157–163 (2010).
Mancini, P. L., Serafini, P. P. & Bugoni, L. Breeding seabird populations in Brazilian oceanic islands: historical review, update and a call for census standardization. Rev. Bras. Ornitol. 24, 94–115 (2016).
Kohlrausch, A. B. Biologia reprodutiva, comportamento e ecologia de atobás (Sulidae): implicações para a evolução do dimorfismo sexual no tamanho. (PhD Thesis, Universidade de São Paulo, 2003).
Neves, F. M., Mancini, P. L., Marques, F. P., Nunes, G. T. & Bugoni, L. Cannibalism by brown booby (Sula leucogaster) at a small tropical archipelago. Rev. Bras. Ornitol. 23, 299–304 (2015).
Thompson, K. R. & Furness, R. W. The influence of rainfall and nest-site quality on the population dynamics of the Manx shearwater Pufinus pufinus on Rhum. J. Zool., Lond. 225, 427–437 (1991).
Svagelj, W. S. & Quintana, F. Breeding performance of the imperial shag (Phalacrocorax atriceps) in relation to year, laying date and nest location. Emu 111, 162–165 (2011).
Minias, P. Evolution of within-colony distribution patterns of birds in response to habitat structure. Behav. Ecol. Sociobiol. 68, 851–859 (2014).
Weimerskirch, H. et al. Species- and sex-specific differences in foraging behaviour and foraging zones in blue-footed and brown boobies in the Gulf of California. Mar. Ecol. Prog. Ser. 391, 267–278 (2009).
Soanes, L. M. et al. Foraging behaviour of brown boobies Sula leucogaster in Anguilla, Lesser Antilles: preliminary identification of at-sea distribution using a time-in-area approach. Bird Conserv. Int. 25, 87–96 (2015).
Soares, J., Oliveira, A. P., Codato, G. & Escobedo, J. F. Local and regional features of surface radiation fluxes over the tropical Atlantic Ocean near São Pedro and São Paulo Archipelago: evidence of small scale upwelling. Nat. Env. Poll. Tech. 11, 541–548 (2012).
Souza, C. S., Luz, J. A. G., Macedo, S., Montes, M. J. F. & Mafalda-Jr, P. Chlorophyll α and nutrient distribution around seamounts and islands of the tropical south-western Atlantic. Mar. Freshw. Res. 64, 168–184 (2013).
Sommerfeld, J., Kato, A., Ropert-Coudert, Y., Garthe, S. & Hindell, M. A. The individual counts: within sex differences in foraging strategies are as important as sex-specific differences in masked boobies Sula dactylatra. J. Avian Biol. 44, 531–540 (2013).
Lewis, S. et al. Sex-specific foraging behaviour in tropical boobies: does size matter? Ibis 147, 408–414 (2005).
Lessa, R. P. et al. Distribution and abundance of ichthyoneuston at seamounts and islands off north-eastern Brazil. Arch. Fish. Mar. Res. 47, 239–252 (1999).
Macedo-Soares, L. C. P., Freire, A. S. & Muelbert, J. H. Small-scale spatial and temporal variability of larval fish assemblages at an isolated oceanic island. Mar. Ecol. Prog. Ser. 444, 207–222 (2012).
Monteiro, A., Vaske-Jr, T., Lessa, R. P. & El-Deir, A. C. A. Exocoetidae (Beloniformes) off north-eastern Brazil. Cybium 22, 395–403 (1998).
Vaske-Jr, T., Vooren, C. M. & Lessa, R. P. Feeding strategy of yellowfin tuna (Thunnus albacares), and wahoo (Acanthocybium solandri) in the Saint Peter and Saint Paul Archipelago, Brazil. Bol. Inst. Pesca 29, 173–181 (2003).
Viana, D. F., Hazin, F. H. V., Andrade, H. A., Nunes, D. M. & Viana, D. L. Fisheries in the Saint Peter and Saint Paul Archipelago: 13 years of monitoring. Bol. Inst. Pesca 41, 239–248 (2015).
Both, R. & Freitas, T. R. O. A dieta de Sula leucogaster, Anous stolidus e Anous minutus no Arquipélago de São Pedro e São Paulo, Brasil. In Ornitologia e conservação: da ciência às estratégias (eds Albuquerque, J. L. B., Cândido-Jr, J. F., Straube, F. C. & Roos, A. L.) 313–326 (Editora Unisul, 2001).
Gosling, S. D. & John, O. P. Personality dimensions in nonhuman animals: a cross-species review. Curr. Dir. Psychol. Sci. 8, 69–75 (1999).
Reed, T. E. et al. Reproductive senescence in a long-lived seabird: rates of decline in late-life performance are associated with varying costs of early reproduction. Am. Nat. 171, E89–E101 (2008).
Barbraud, C. & Chastel, O. Early body condition and hatching success in the snow petrel Pagodroma nivea. Polar Biol. 21, 1–4 (1999).
Cox, S. L. et al. Seabird diving behaviour reveals the functional significance of shelf-sea fronts as foraging hotspots. R. Soc. Open Sci. 3, 160317 (2016).
Elliott, K. H., Vaillant, M. L., Kato, A., Speakman, J. R. & Ropert-Coudert, Y. Accelerometry predicts daily energy expenditure in a bird with high activity levels. Biol. Lett. 9, 20120919 (2013).
Croxall, J. P. et al. Seabird conservation status, threats and priority actions: a global assessment. Bird Conserv. Int. 22, 1–34 (2012).
Weitzman, J., den Heyer, C. & Bowen, D. W. Factors influencing and consequences of breeding dispersal and habitat choice in female grey seals (Halichoerus grypus) on Sable Island, Nova Scotia. Oecologia 183, 367–378 (2017).
Pomeroy, P. P., Twiss, S. D. & Redman, P. Philopatry, site fidelity and local kin associations within grey seal breeding colonies. Ethology 106, 899–919 (2000).
Stenson, G. B. & Hammill, M. O. Can ice breeding seals adapt to habitat loss in a time of climate change? ICES J. Mar. Sci. 71, 1977–1986 (2014).
Chambert, T., Rotella, J. J. & Garrott, R. A. Female Weddell seals show flexible strategies of colony attendance related to varying environmental conditions. Ecology 96, 479–488 (2015).
IPCC. Climate change 2014: synthesis report. In Fifth assessment report of the Intergovernmental Panel on Climate Change (eds Pachauri, R. K. & Meyer, L. A.)73–74 (IPCC, 2014).
Riehl, H. Climate and weather in the tropics. (Academic Press, 1979).
Araujo, M. C. & Cintra, M. M. Modelagem matemática da circulação oceânica na região equatorial. In O arquipélago de São Pedro e São Paulo: 10 anos de estação científica (eds. Viana, D. L., Hazin, F. H. V. & Souza, M. A. C.) 107–113 (SECIRM, 2009).
Forster, I. P. & Phillips, R. A. Influence of nest location, density and topography on breeding success in the black-browed albatross Thalassarche melanophris. Mar. Ornithol. 37, 213–217 (2009).
Simmons, K. L. Ecological adaptations in the life history of the brown booby at Ascension Island. Living Bird 6, 187–212 (1967).
Phillips, R. A., Xavier, J. C. & Croxall, J. P. Effects of satellite transmitters on albatrosses and petrels. Auk 120, 1082–1090 (2003).
Boecklen, W. J., Yarnes, C. T., Cook, B. A. & James, A. C. On the use of stable isotopes in trophic ecology. Annu. Rev. Ecol. Evol. Syst. 42, 411–440 (2011).
Bugoni, L., McGill, R. A. R. & Furness, R. W. Effects of preservation methods on stable isotope signatures in bird tissues. Rapid Commun. Mass Spectrom. 22, 2457–2462 (2008).
Logan, J. M. & Lutcavage, M. E. A comparison of carbon and nitrogen stable isotope ratios of fish tissues following lipid extractions with non-polar and traditional chloroform/methanol solvent systems. Rapid Commun. Mass Spectrom. 22, 1081–1086 (2008).
Bond, A. L. & Hobson, K. A. Reporting stable isotope ratios in ecology: recommended terminology, guidelines and best practices. Waterbirds 35, 324–331 (2012).
Jackson, A. L., Inger, R., Parnell, A. C. & Bearhop, S. Comparing isotopic niche widths among and within communities: SIBER – Stable Isotope Bayesian Ellipses in R. J. Anim. Ecol. 80, 595–602 (2011).
Stock, B. C. & Semmens, B. X. MixSIAR GUI User Manual. R package version 3.1 (2013).
Bearhop, S., Waldron, S., Votier, S. C. & Furness, R. W. Factors that influence assimilation rates and fractionation of nitrogen and carbon stable isotopes in avian blood and feathers. Physiol. Biochem. Zool. 75, 451–458 (2002).
Brown, S. C., Bizarro, J. J., Cailliet, G. M. & Ebert, D. A. Breaking with tradition: redefining measures for diet description with a case study of the Aleutian skate Bathyraja aleutica (Gilbert 1896). Environ. Biol. Fish. 95, 3–20 (2012).
Froese, R., Thorson, J. T. & Reyes-Jr, R. B. A Bayesian approach for estimating length-weight relationships in fishes. J. Appl. Ichthyol. 30, 78–85 (2014).
Legendre, P. & Legendre, L. Numerical ecology. (Elsevier, 2012).
Liaw, A. & Wiener, M. Classification and regression by randomForest. R News 2, 18–22 (2002).
Acknowledgements
We thank Fernanda P. Marques and Patrícia L. Mancini for their support in fieldwork and laboratory procedures, as well as Arel Hadi and the LABSERCA/Unisinos team for assistance in obtaining and processing GIS data. We also thank Marinha do Brasil and Comissão Interministerial para os Recursos do Mar for providing logistical support for fieldwork in Saint Peter and Saint Paul Archipelago. Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio) granted sampling permit No. 38723 and CEMAVE/ICMBio provided metal rings and permits for banding boobies. The manuscript greatly benefited from comments and suggestions by Eduardo R. Secchi, Luciano Dalla Rosa, Alexandre M. Garcia, and Flávio Quintana, and two anonymous reviewers. Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) granted fund No. 405497/2012–1 through the Pró-Arquipélago Program. G.T.N received a PhD scholarship granted by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), and L.B. is a research fellow of the CNPq (Proc. No. 310550/2015–7).
Author information
Authors and Affiliations
Contributions
G.T.N. and L.B. originally formulated the idea and developed methodology, G.T.N. conducted fieldwork, G.T.N. and S.B. performed statistical analyses, and G.T.N. wrote the manuscript. All authors revised and contributed to the final version of the manuscript.
Corresponding author
Ethics declarations
Competing Interests
The authors declare that they have no competing interests.
Additional information
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Nunes, G.T., Bertrand, S. & Bugoni, L. Seabirds fighting for land: phenotypic consequences of breeding area constraints at a small remote archipelago. Sci Rep 8, 665 (2018). https://doi.org/10.1038/s41598-017-18808-7
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-017-18808-7
- Springer Nature Limited
This article is cited by
-
Biogeographic variation on dietary aspects of a widely distributed seabird
Marine Biology (2023)
-
Intraspecific trophic variation in brown booby (Sula leucogaster) from the Southwestern Atlantic
Marine Biology (2023)
-
Foraging behaviour and diet of Brown boobies Sula leucogaster from Tinhosas Islands, Gulf of Guinea
Marine Biology (2021)
-
Did the animal move? A cross-wavelet approach to geolocation data reveals year-round whereabouts of a resident seabird
Marine Biology (2021)
-
Breeding stage, not sex, affects foraging characteristics in masked boobies at Rapa Nui
Behavioral Ecology and Sociobiology (2020)