Age mediates access to land ﬁ ll food resources and foraging pro ﬁ ciency in a long-lived bird species

in ﬂ uences land ﬁ ll foraging pro ﬁ ciency and dominance over resources. Adult storks visited land ﬁ lls on 57% of the days, while juveniles only visited land ﬁ lls on 29% of the days. There was strong competition for food at land ﬁ lls, with adults exerting dominance over juveniles, foraging predominantly in areas with higher food availability and outcompeting juveniles in food acquisition. Juveniles had signi ﬁ cantly lower food intake rates in the best foraging areas and showed less aggressiveness, being forced to use adjacent lower quality areas. Overall, juveniles had limited access to land ﬁ ll resources, suggesting that land ﬁ ll diet specialization is mediated by age-related improvements in foraging expertise and increased competitiveness developed during maturation. Thus, land ﬁ ll use is shaping foraging strategies and species behaviour from an early age, with potential consequences for population dynamics.

Humans have greatly transformed ecosystems over recent centuries, changing the availability of foraging resources for wildlife across the world (Ellis et al., 2010;Foley et al., 2005).Food waste disposal at landfills and dumps currently provides a major source of predictable anthropogenic food subsidies (PAFS; Oro et al., 2013), which is exploited by a growing number of species (Oro et al., 2013;Plaza & Lambertucci, 2017).The superabundance of food and easy access, combined with high spatial and temporal predictability (Oro et al., 2013;Plaza & Lambertucci, 2017;Thyberg & Tonjes, 2016), are substantially reducing foraging times and energetic costs of species and shaping the foraging strategies of individuals (Patenaude- Monette et al., 2014;Soriano-Redondo et al., 2021;van Donk et al., 2019).
In recent decades, the exploitation of anthropogenic food waste has often been associated with dietary shifts in opportunistic and scavenger species (Bialas et al., 2020;Newsome et al., 2015;Oro et al., 2013;Tauler-Ametller et al., 2017).These changes are linked with positive fitness effects, such as improvement in body condition, and increased breeding performance and survival leading to population growth (Newsome et al., 2015;Oro et al., 2013;Plaza & Lambertucci, 2017;Weiser & Powell, 2010).Nevertheless, foraging in landfills can produce contrasting and detrimental effects, through increased risk of pathogen infections, poisoning or ingestion of foreign bodies, affecting the health and survival of individuals (Plaza & Lambertucci, 2017).Additionally, the concentration of organic waste in restricted areas attracts high densities of individuals, especially birds (Blanco, 1994;Novaes & Cintra, 2013;Soriano-Redondo et al., 2021;Tauler-Ametller et al., 2017), and probably increases intraspecific and interspecific competition mediated by density-dependent processes, which may result in reduced food acquisition with detrimental consequences for individual fitness (Araújo et al., 2011).
Increased competition for food resources can have a strong effect on population dominance structure, in which dominant individuals outcompete subordinates in food acquisition, forcing them to move to alternative foraging areas (Kaufmann, 1983;Tibbetts et al., 2022).Bird dominance phenotypic attributes often include age, sex and body size (Lundberg, 1985;Richner, 1989).Older individuals are usually dominant, as young ones lack the skills to recognize or compete for profitable foraging sites, due to inexperience and physical immaturity (Marchetti & Price, 1989;Wunderle, 1991).Experience can be especially relevant in the choice of foraging grounds, where adults repeatedly and efficiently exploit the same areas from learned predictable habitat features (Votier et al., 2017;Wakefield et al., 2015).In young birds, foraging specialization is developed during individual exploratory behaviour in early life and is later refined as they grow older and gain more experience (Bolnick et al., 2003;Campioni et al., 2020;Marchetti & Price, 1989;Wunderle, 1991).
In long-lived species with slow maturation, physical development and individual learning can further delay the acquisition of adult-like foraging skills, increasing the differences in foraging proficiency between adults and naïve juveniles (Grecian et al., 2018;Mendez et al., 2017;Riotte-Lambert & Weimerskirch, 2013).Ultimately, this might prevent the specialization on some food sources and the development of foraging expertise (Diamond & Bond, 1991;Marchetti & Price, 1989;Wunderle, 1991).Age segregation in the use of the best foraging sites or resources can have significant spillover effects, lowering long-term survival and thus individual fitness, with consequences for the demography of populations (Rotics et al., 2021;Saether et al., 2013).While exploitation of human food subsidies, such as landfill waste, has enabled the rapid population growth of generalist species, the influence of age and experience on an individual's ability to compete for space and resources remains poorly understood.This knowledge will help us understand how species adapt to new resources and changes in food availability.
The white stork, Ciconia ciconia, is a very adaptable and opportunistic long-lived species with slow maturation, for which early-life learning and experience play an important role in the acquisition of foraging skills (Boche nski & Jerzak, 2006;Elliott et al., 2020).In the last few decades, the breeding population in southern Europe has increased considerably, driven by their trophic plasticity and ability to exploit new anthropogenic resources, mainly those available at landfill sites (Molina & Del Moral, 2006;Rosa et al., 2005, p. 41).Thousands of storks now use landfills (Catry et al., 2017;Gilbert et al., 2016;Soriano-Redondo et al., 2021), and this time-and energy-saving strategy improves feeding efficiency when compared to foraging on natural prey (Soriano-Redondo et al., 2021).
The migratory behaviour of white storks has been changing and an increasing number of individuals no longer carry out their annual autumn migration from Europe to sub-Saharan Africa, remaining in Iberia throughout the year (Catry et al., 2017;Gilbert et al., 2016).Abundant year-round food available at landfills is one of the factors that may have contributed to the suppression of migratory behaviour and to the steep increase in the resident population (Catry et al., 2017;Tortosa et al., 2002).However, while Iberian adult white storks are predominantly resident, juveniles are still mostly migratory (Ac acio, 2021;Ac acio et al., 2022;B ecares et al., 2019).Therefore, understanding the use of landfill resources, by adult and juvenile storks, can help unravel the processes acting on individual movement behaviour and on development of foraging strategies, including exploitation of anthropogenic food sources.
In this study, we analysed age-related differences in landfill use by white storks.Combining GPS tracking data, monthly counts in landfills and behavioural data from video recordings at landfills, we investigated the influence of age and experience in determining (1) landfill attendance, (2) access to food resources and (3) foraging proficiency (behaviour time budgets, feeding success and agonistic interactions), during the premigratory period.We hypothesized that landfill foraging skills are progressively acquired with age; hence we expected adults to use landfills more often and show higher feeding success.

GPS Deployment
We used GPS tracking data from 68 adult and 67 juvenile white storks tagged in southern Portugal between 2018 and 2020.Storks were tagged with 'Flyway 50' GPS/GSM loggers from Movetech Telemetry (Thetford, U.K.; four different models varying slightly in weight), 'Ornitrack-50' GPS/GSM loggers from Ornitela (Vilnius, Lithuania) and 'Bird solar tags' GPS/GSM loggers from e-obs GmbH (Grünwald, Germany).Adult birds were captured for tagging at multiple landfill sites using leg loop traps, or at their nests with a remotely activated clap net.Birds tagged at landfills were further confirmed as breeding adults by identifying their nests from the GPS data and visiting them to verify the presence of eggs and/or chicks.Juvenile birds were retrieved from their nests for tagging 50e55 days after hatching and returned afterwards.All birds were measured and ringed, and the tracking devices were deployed as backpacks with a Teflon harness and programmed to transmit locations every 20 min.

GPS Data Set Selection
This study focused on the white stork premigratory period, when both adults and juvenile birds that recently fledged gather at landfills, probably facilitating learning from conspecifics but also competition for resources (Araújo et al., 2011;Mueller et al., 2013).For all storks, only the premigratory period of the tagging year was included.That period for adult storks was set between 1 June (corresponding to the earliest fledging day for juveniles) and the day migration started or 30 September for nonmigratory birds.The premigratory period for juvenile storks was set between the fledging day (when they moved more than 50 m from their nest) and the day migration started or the last day alive for birds that did not survive to migration.Juveniles that died soon after fledging (N ¼ 4) and never moved beyond the vicinity of the nest were not included.To establish the start of migration, we used the spatiotemporal displacement method described in Soriano-Redondo et al. (2020), which combines movement displacement and spatial and temporal thresholds to identify the beginning and end of migratory movements.Following this method, the departure date was defined as the first of 3 consecutive days a stork moved more than 60 km between roosts after leaving the breeding area (calculated as the 90% kernel of June GPS locations).Lastly, in order to include only ground GPS locations in the analysis that represented the use of possible foraging sites, all locations at the nests or in flight with ground speed above 1.39 m/s (Marcelino et al., 2021) were excluded.

Classification of Landfill Areas
All landfill sites in Portugal and Spain used by the tracked white storks were initially identified from the GPS data and later confirmed in the field.Furthermore, from 2018 to 2020, we carried out monthly visits, between June and September, to five landfills in southern Portugal, which were the most used landfills by the tracked storks ( Evora, Ermidas do Sado, Beja, Barlavento and Sotavento; Fig. 1).In all visits, the exact location where the waste was discarded (dump site) was visually confirmed and mapped, enabling the classification of three landfill areas with a decreasing gradient of food availability.The landfill 'core area' was defined as the area within a radius of 25 m around the dump site, where piles of fresh waste were highly clumped and food availability was the highest.The area between 25 m and 50 m from the dump site was defined as the 'buffer area', where the waste was spread over a larger area and compacted by the landfill machinery, making organic matter less accessible and gradually reducing the amount of food available.The remaining landfill area, more than 50 m away from the dump site, was defined as the 'outer area', being the area with the lowest food availability.

GPS Data Analysis
Age-related differences in landfill attendance were investigated by comparing the daily and overall landfill attendance of adult (N ¼ 68) and juvenile (N ¼ 67) storks.For each individual, daily landfill attendance was determined as the proportion of days of the premigratory period with at least one GPS location at a landfill, and overall landfill attendance was calculated as the proportion of GPS locations obtained during the premigratory period at landfill sites.
Age-related differences in access to food resources were investigated by comparing the proportion of GPS locations of adult and juvenile storks in areas where food was available within landfills (core and buffer area attendance) during the premigratory period.This was only possible to determine for the 61 adult and 47 juvenile storks that used the five landfills where food availability areas (core, buffer and outer areas) were identified.To determine core area attendance, the GPS locations of each bird and at each landfill were first classified as 'inside the core area', where food resources were most abundant, or 'outside the core area'.The subset of locations obtained outside the core area was then considered to determine buffer area attendance, where food resources were less abundant; thus the GPS locations were classified as 'inside the buffer area' or 'inside the outer area'.

Stork Counts and Video Recording at Landfills
In 2020, monthly visits to the five landfills in southern Portugal enabled us to determine the number of adult and juvenile white storks in each area (core, buffer and outer area) of the landfills.Storks were counted three times during the visits, with intervals of approximately 30 min, to account for variability in numbers due to stork arrivals and departures.The average monthly number of storks using each landfill and each area of the landfills was calculated.At each count, the proportion of storks actively feeding was visually estimated, and the monthly average proportion of storks foraging in each landfill area was determined.
In July 2020, using a Canon Power Shot SX50 HS camera, B.H.M. recorded on video the behaviour of 165 adult storks and 124 juveniles during 3 consecutive minutes for each individual, across the different food availability areas of the five landfills.As the number of juveniles at each landfill site was always very low, especially in the core areas, only those that were observed simultaneously and confirmed as different individuals on a given visit were recorded.Hence, only a maximum of five juvenile storks were recorded in some landfill areas.In contrast, as adults were always numerous, five adult storks were randomly selected and recorded in each area per visit.Since it was not possible to reliably identify birds between different visits, to prevent pseudoreplication, recordings were performed at the five different landfills located hundreds of kilometres apart and visits to the same landfill on consecutive days were avoided.We obtained 55 videos for adult storks in each landfill area, while for juvenile storks, we obtained 17 videos in the core area, 55 in the buffer area and 52 in the outer area.For both counts and videos, birds were classified as adults (>2 years) or juveniles (first year) through direct observation of phenotypic characteristics (Van den Bossche et al., 2002).Adults have a red beak, lack glossy feathers and have extensive moult, while juveniles have a dark beak, glossy feathers and no moult.It was not possible to distinguish further age classes based on phenotypic characteristics.Entry to the landfills and filming was authorized by the local authorities, which provided access to the best place for counting and filming while avoiding disturbing the birds.The visits were all completed during the morning, as this is the time when most waste is dumped at landfills.

Video Recording Analysis
Age-related differences in landfill foraging proficiency were examined by comparing information obtained through video recordings of the behaviours of 165 adult and 124 juvenile storks.All videos were analysed by B.H.M. and were used to determine behaviour time budgets, feeding success and agonistic interactions of each stork.For behaviour time budgets, the predominant behaviour in each 10 s period of the 3 min videos was identified using an ethogram (Table A1), to facilitate interpretation and guarantee consistency.All behaviours were then merged into three main categories ('foraging', 'alert' or 'inactive'; Table A1), and were used to determine the individual proportion of foraging, alert and inactive periods.For feeding success, the total number of food units ingested during the 3 min videos was estimated.To quantify food intake, all ingested items were classified into size categories using bill length as a reference and considering a food unit equivalent to one-quarter of the bill size.Thereafter, all estimated corresponding food units were counted to determine total food intake.For agonistic interactions, the overall aggressiveness during the 3 min videos was quantified.All agonistic encounters were identified and classified as 'aggressor type' or 'victim type', and used to determine the individual proportion of interactions as the aggressor.

Statistical Analysis
Data were analysed using R version 4.2.1 (R Core Team, 2022).To evaluate age-related differences in landfill attendance and understand how often tracked storks visited landfills during their premigratory periods, we fitted two generalized linear models (GLMs), with the 'glm' function.The first model included the daily landfill attendance (individual proportion of days visiting landfills during the premigratory period) and the second model included the overall landfill attendance (individual proportion of GPS locations in landfills during the premigratory period), both as binomial response variables.Both models included age (adult or juvenile) as the explanatory variable and a quasibinomial distribution due to residual overdispersion.
To evaluate age-related differences in access to food resources and understand how tracked storks were able to use the landfill areas with higher food availability during their premigratory periods, we fitted two generalized linear mixed models (GLMMs), with the 'glmer' function in the 'lme4' package (Bates et al., 2015).The first model included the landfill core area attendance (individual proportion of GPS locations in the core area) and the second model included the landfill buffer area attendance (individual proportion of GPS locations in the buffer area) as binomial responses.In both models, age (adult or juvenile) was specified as the explanatory variable, while bird ID and landfill site were set as random effects to account for variations in access to food resources by individuals that used more than one landfill.
To examine the foraging proficiency of adult and juvenile storks at landfills, we first fitted three GLMMs to specifically explore the influence of age and landfill area on time spent performing each of the main behaviours observed when using landfills.The models included the foraging time (individual proportion of foraging periods), alert time (individual proportion of alert periods) and inactive time (individual proportion of inactive periods) as binomial responses.For the three models, age (adult or juvenile), landfill area (core, buffer or outer area) and their interaction were specified as explanatory variables, while landfill site was set as a random effect.Second, we explored the influence of age and landfill area on feeding success.The GLMM model included food intake (number of food units ingested) as a Poisson response, while age (adult or juvenile), landfill area (core or buffer area) and their interaction were specified as explanatory variables, and landfill site was included as a random effect.Food intake was not examined in the outer area as it rarely took place, reflecting the low food availability in this area.Finally, we fitted one GLMM to explore the influence of age and landfill area on agonistic interactions.This model included aggressiveness (individual proportion of interactions as the aggressor) as the binomial response.Age (adult or juvenile), landfill area (core or buffer area) and their interaction were specified as explanatory variables, while landfill site was included as a random effect.There were no agonistic interactions observed in the outer area; hence this was excluded from the analysis.Whenever necessary, we performed post hoc Tukey tests to assess differences between landfill areas, using the 'emmeans' function in the 'emmeans' package (Searle et al., 1980).

Ethical Note
The white stork is not endangered or threatened and is common in Portugal, where this study was conducted.The procedure was approved by the Institute for Nature Conservation and Forests in Portugal (licence numbers: 548/2018/CAPT, 248/ 2019/CAPT and 365/2020/CAPT).Storks were trapped, handled and tagged by trained researchers and released at the point of capture.The tag and harness together weighed 50e90 g, representing 1.1e3.7% of the bird's body mass.Most birds were resighted in the days following tag deployment and throughout the breeding season, and no abnormal behaviour or adverse effects due to tagging were observed.Additionally, to reduce the long-term effects of the loggers, all devices were deployed with a weak-link harness design.The four Teflon straps that made up the harness were secured together with a biodegradable cotton thread, acting as the weak link that deteriorates over time until it breaks and the entire harness falls off.

Landfill Use During the Premigratory Period
Tracking data for 68 adult and 67 juvenile white storks tagged in southern Portugal were obtained during the premigratory period.This included 7499 adult stork tracking days (mean 111 ± 22 days per individual) and 2952 juvenile stork tracking days (mean 44 ± 22 days per individual).In total, there were 259,995 adult and 63,232 juvenile GPS locations recorded.Storks foraged mainly in areas around their nests and at landfills in southern Iberia (Fig. 1).The majority of the tracked storks visited landfills (96% of adults and 76% of juveniles), while the remaining birds (4% of adults and 14% of juveniles) only foraged in natural areas.In addition, 78% of the adult birds were residents, staying in Iberia during the full    annual cycle, while all juveniles alive at the end of the premigratory period migrated to Africa.Monthly counts at landfills confirmed the regular use of these sites by thousands of storks during the premigratory period (Fig. A1).Adults were always present in large numbers, yet their numbers increased steeply between June (mean 578 ± 330 individuals) and August (mean 2133 ± 1472 individuals), decreasing towards the end of the period in September (mean 1273 ± 307 individuals).Juveniles were only present between June and August in relatively small numbers (maximum count was 172 individuals), in proportions always lower than 3% of the total number of storks counted (Fig. A1).

Landfill Attendance and Access to Food Resources
Tracking data showed clear age-related differences in daily landfill attendance during the premigratory period (Table 1).
The distribution of storks in the three areas of the landfills (core, buffer and outer areas), representing the decreasing gradient of food availability, was also strongly associated with bird age (Table 2).The proportion of adult stork GPS locations in landfill core feeding areas was nearly twice as high as that of juveniles (predicted values ± SE: adults ¼ 0.22 ± 0.29; juveniles ¼ 0.12 ± 0.31; P < 0.001; Fig. 2c).In the adjacent buffer areas, the proportion of GPS locations was also considerably higher for adults (predicted values ± SE: adults ¼ 0.36 ± 0.36; juveniles ¼ 0.24 ± 0.37; P < 0.001; Fig. 2d).Juvenile storks were mainly restricted to outer areas of the landfill sites (Fig. 3).Stork counts confirmed that the core and buffer areas were dominated by adults, whereas juveniles were mostly absent from these areas and dispersed in the outer areas (Fig. A2).Storks were more aggregated in areas with higher food availability and their density decreased towards the outer areas (average density in the core, buffer and outer areas ¼ 0.167, 0.083 and 0.001 storks/ m 2 , respectively), along the decreasing gradient of food availability.

Landfill Foraging Proficiency
Video recording analysis showed that foraging proficiency was influenced by bird age and landfill area (Table 3, Table A2).Behaviour time budgets showed that adult birds spent most time foraging in core areas and were mostly inactive in other landfill areas, while juveniles spent most of their time alert in core and buffer areas, and less time inactive than adults in outer areas.Adults spent almost twice as much time foraging in core areas as juveniles (predicted values ± SE: adults ¼ 0.67 ± 0.18; juveniles ¼ 0.35 ± 0.21; P < 0.001); in buffer areas, adults and juveniles had similar foraging levels (adults ¼ 0.42 ± 0.18; juveniles ¼ 0.40 ± 0.18; P ¼ 0.909); and in outer areas foraging was almost nonexistent (adults ¼ 0.01 ± 0.32; juveniles ¼ 0.03 ± 0.25; P ¼ 0.091; Fig. 4a).In contrast, the proportion of time alert was more than twice as high in juveniles as in adults, both in core (predicted values ± SE: adults ¼ 0.26 ± 0.11; juveniles ¼ 0.62 þ 0.14; P < 0.001) and buffer (adult ¼ 0.21 ± 0.11; juvenile ¼ 0.41 ± 0.10; P < 0.001) areas.In outer areas, the time spent alert was low for both age classes, but significantly higher for juveniles (adults ¼ 0.01 ± 0.34; juveniles ¼ 0.08 ± 0.14; P < 0.001; Fig. 4b).Finally, the proportion of time inactive in core areas was low for both ages, but significantly higher for adults (predicted values ± SE: adults ¼ 0.07 ± 0.19; juveniles ¼ 0.03 ± 0.52; P ¼ 0.013); in buffer areas, it increased considerably for both ages, but remained higher for adults (adults ¼ 0.37 ± 0.17; juveniles ¼ 0.19 ± 0.17; P < 0.001); and in outer areas, almost all individuals observed were inactive but fewer juveniles were inactive than adults (adults ¼ 0.98 ± 0.26; juveniles ¼ 0.89 ± 0.19; P < 0.001; Fig. 4c).Stork counts at landfills reinforced the results of the video recordings, confirming that adult foraging birds were concentrated in the core areas while juveniles occasionally managed to forage in buffer areas (Fig. A3).
Regarding feeding success, adults showed higher food intake compared to juveniles.Adults consumed nearly four times as many food units as juveniles in core areas (predicted values ± SE: adults ¼ 8.6 ± 0.1; juveniles ¼ 2.4 ± 0.2; P < 0.001).The number of food units consumed by adults in buffer areas was significantly lower than in core areas and similar to juveniles (adults ¼ 3.2 ± 0.1; juveniles ¼ 3.6 ± 0.1; P ¼ 0.737; Fig. 5).

DISCUSSION
By combining multiple fine-scale methods, this study reveals that, in a long-lived opportunistic species, age determines landfill attendance, food access and foraging proficiency in landfill sites.In line with our predictions, adult white storks visited landfills more often than juveniles and were always present at these sites in very high numbers.They were also more likely to occur and forage in areas where resources were highly available, while juveniles were displaced to outer areas with less food availability.Additionally, adults had higher feeding success in core areas and showed higher dominance over resources.Together, these findings show that the resources available at landfills are predominantly used by adult storks, which outcompete subordinate juveniles.This supports the hypothesis that competitive abilities for foraging in landfills are probably developed as storks age.
Juvenile white storks showed similar preferences to adults in selecting foraging sites and were able to locate and attend landfills just after fledging, which is in line with what would be expected in birds using social information and behavioural cues from older birds (Franks et al., 2018;Marchetti & Price, 1989).The number of storks using landfills increased from June to August, which coincided with the end of the breeding season, after juveniles fledged and when breeding adults were no longer constrained to the nests.Stork numbers at landfills started to decrease in September, as most migratory individuals (especially juveniles) initiated their journey then (Arizaga et al., 2018;Blanco, 1996).Thus, during the premigratory period (between June and September), thousands of adult and juvenile storks congregated at landfill sites, particularly in the small core and buffer areas where the waste is dumped.This behaviour may promote changes in foraging intensifying intraspecific competition for limited food resources (Real et al., 2017), and leading to the establishment of hierarchies in the population dominance structure (Kaufmann, 1983;Tibbetts et al., 2022).We show evidence of considerable age differences in storks using landfill sites during the premigratory period.Adult storks were twice as likely as juveniles to visit landfills daily (57% and 29% of the days, respectively).However, the overall proportion of GPS locations in landfills was similar for adults and juveniles (27% versus 20%), indicating that juveniles spent more time at landfills during their landfill days.Within these sites, storks were spatially segregated by age, with adults predominantly using areas with higher food availability (core areas) and juveniles being forced to use areas with almost no food.The lower attendance at landfills by juvenile storks, although proportionately as intensive as adults, together with their inability to access optimal landfill areas, is probably related to their lack of skills to compete with older birds, which are normally acquired through learning and foraging experience, as well as physical development (Diamond & Bond, 1991;Grecian et al., 2018;Marchetti & Price, 1989;Mendez et al., 2017;Wunderle, 1991).
Adult storks used landfills more efficiently, primarily foraging in core areas and resting and preening in the other areas.Juveniles spent most of their time alert, looking for foraging opportunities and food, and avoiding confrontations with adults.Overall, adults had significantly greater feeding success and consistently demonstrated higher aggressiveness to ensure dominance over resources, consuming nearly four times more food than juveniles in core areas.Owing to the low densities of juvenile storks and the inability to identify them individually between different landfill visits, we were unable to completely eliminate the possibility of pseudoreplication influencing estimates.Still, both adult and juvenile stork estimates exhibited comparable levels of variability, indicating that the eventual impact of pseudoreplication was minimal.Therefore, this study shows that, as expected from dominance hierarchies in food access (Richner, 1989;Tibbetts et al., 2022), age is a determining attribute in the hierarchical structure of white storks, with adults dominating and monopolizing landfill food resources.
The use of landfill food waste has been shown to provide a wide range of benefits to white stork populations, most notably enabling individuals to save foraging time and energy (Soriano-Redondo et al., 2021), improve breeding success and boost population growth (Bialas et al., 2020;Djerdali et al., 2008;Gilbert, 2015;L opez-García et al., 2021).Furthermore, the year-round availability of food waste in landfills, particularly when other resources are scarce, is facilitating the use and defence of nests by adults during the nonbreeding season, probably allowing for the establishment of resident populations (Catry et al., 2017;Gilbert et al., 2016).Indeed, most of the Portuguese breeding population is now resident (62% by 2015), no longer carrying out the annual migration to Africa (Catry et al., 2017).
In the first few months of independent life, the development of foraging skills is critical for juvenile survival (Daunt et al., 2007;Lindstr€ om, 1999;Orgeret et al., 2016;Saether et al., 2013).Adult dominance in access to food waste may force juveniles to seek resources outside landfills, where they will be more exposed to environmental seasonality and natural resource depletion, which is known to induce migratory decisions in birds (Newton, 2007).Hence, the high intraspecific competition at landfill sites during the premigratory period suggests that food availability is restricted to adults and the ability to compete for these resources is unlikely to override the urge to migrate among inexperienced birds (Chernetsov et al., 2004).With fewer conspecific adult storks foraging in natural habitats, there will be fewer social learning opportunities for juveniles, delaying the acquisition of vital foraging skills (Franks & Thorogood, 2018;van Schaik, 2010).Furthermore, as the number of nonmigratory adults increases due to year-round food availability at landfill sites (Catry et al., 2017;Tortosa et al., 2002), the resulting lack of experienced birds migrating and leading the way to profitable natural foraging areas may be driving juveniles into suboptimal foraging grounds during the winter.Ultimately, the reduced number of adult birds on wintering grounds may even contribute to the inefficient exploitation of resources by juvenile birds, which will probably affect their body condition and increase mortality.
Overall, our findings provide strong evidence that anthropogenic food subsidies can affect age classes differently, with adults benefiting more than juveniles.This age-structured access to landfill resources may have potential carryover effects on population dynamics (Oro et al., 2013;Plaza & Lambertucci, 2017).After significant population declines until the 1980s, white storks are now increasing across Europe (BirdLife International, 2016).In recent decades, the exponential growth of this species in some southern European countries, mostly assisted by landfill resources, has increased the potential for humanewildlife conflicts in agricultural and urban areas (Molina & Del Moral, 2006;Rosa et al., 2005, p. 41), as well as the transmission of pathogens and diseases from landfill sites (H€ ofle et al., 2020).Therefore, to avoid future conflicts, specific management measures are needed to reduce the amount of organic food waste available at landfill sites.

Figure 1 .
Figure 1.Total ground GPS locations of (a) adult (N ¼ 68) and (b) juvenile (N ¼ 67) tracked white storks during their premigratory periods between 2018 and 2020.Black circles and dots indicate landfills and nesting sites, respectively.

Figure 2 .
Figure 2. Predicted values and 95% confidence intervals of the estimates from the GLMs and GLMMs explaining the influence of age (adult or juvenile) of tagged white storks in determining (a) daily landfill attendance and (b) overall landfill attendance (N ¼ 68 adults, N ¼ 67 juveniles); and in determining (c) landfill core area attendance and (d) landfill buffer area attendance (N ¼ 61 adults, N ¼ 47 juveniles) during the premigratory period.Orange and purple areas represent the density distribution of attendance for adult and juvenile birds, respectively.Dots represent raw individual data.

Figure 3 .
Figure 3. Kernel utilization distributions (KUD) calculated from the combined data of GPS-tracked adult (N ¼ 61) and juvenile (N ¼ 47) white storks that used the five landfills in southern Portugal during the 2018e2020 premigratory periods.Density of GPS locations decreases along the decreasing gradient of food availability (core > buffer > outer areas).Colours represent specific KUD contours; landfill areas are represented by black dashed circles.

Figure 4 .
Figure 4. Predicted values and 95% confidence intervals of the estimates from the GLMMs explaining the influence of age (adult or juvenile) and landfill area (core, buffer or outer area) in determining (a) foraging, (b) alert and (c) inactive time budgets (N ¼ 55 adults per landfill area; N ¼ 17, 55 and 52 juveniles per landfill area) of white storks using landfills during the premigratory period.Orange and purple areas represent the density distribution of behaviour time budgets for adult and juvenile birds, respectively.Dots represent raw individual data.

Figure 5 .Figure 6 .
Figure5.Predicted values and 95% confidence intervals of the estimates from the GLMMs explaining the influence of age (adult or juvenile) and landfill area (core or buffer area) in determining food intake (N ¼ 55 adults per landfill area; N ¼ 17 and 55 juveniles per landfill area) of white storks using landfills during the premigratory period.Orange and purple areas represent the density distribution of food intake for adult and juvenile birds, respectively.Dots represent raw individual data.

Table 2
Parameters of the GLMMs explaining the influence of age (adult or juvenile) of GPS-tagged white storks in determining access to landfill resources during the premigratory period Data were collected from 61 adult and 47 juvenile storks.The reference level for age is 'adult'.Significant P values are shown in bold.

Table 1
Parameters of GLMs explaining the influence of age (adult or juvenile) of GPS-tagged white storks in determining landfill attendance during the premigratory period Data were collected from 68 adult and 67 juvenile storks.The reference level for age is 'adult'.Significant P values are shown in bold.B.H.Martins et al. / Animal Behaviour 207 (2024) 23e36

Table 3
Parameters of the GLMMs explaining the influence of age (adult or juvenile) and landfill area (core, buffer or outer area) in determining foraging proficiency of white storks using landfills during the premigratory period Data were collected from 55 adult storks per landfill area and 17, 55 and 52 juveniles per landfill area.The reference level for age is 'adult' and for landfill area is 'core area'.Significant P values are shown in bold.