The Influence of Winter and Spring Large-Scale Climate Patterns on the Timing of Male and Female Blackcap Spring Passage at the Baltic Coast

1. Introduction

Spring migration of passerine birds within the Palearctic-African migration system is an evolutionary adaptation that enables them to take advantage of the seasonal abundance of food and breeding sites in spring and summer at northern latitudes, where harsh winter limits birds’ residence year-round [1,2,3,4,5]. To maximise individual fitness, migrants time their migration to match local food resources availability, such as spring vegetation growth and the peak of insects, at their spring destination [3,6]. However, the optimal timing of spring migration might differ between the sexes, because of different sexual selection pressures on males and females [1,7,8,9,10,11,12]. Climate change in the Northern Hemisphere affects the timing of bird migrations [13,14,15,16,17,18]. However, most studies have focused on the interspecific differences in these phenological shifts, while intraspecific variation, including possible sex differences, in responses to climate change, remains poorly understood [19].

In many bird species, males usually arrive at breeding grounds earlier than females, a phenomenon known as protandry [8,9,10,12,20,21]. Changing environmental conditions might influence the degree of protandry, modifying the costs and benefits of early arrival for each sex [2,10,22,23,24]. Although protandry is common among migrants, the difference in arrival times between the sexes is shaped by the selective pressures of a species’ mating system [6,11]. The main mechanism explaining protandry is the mating opportunity hypothesis, where earlier arrival allows males to occupy high-quality territories, which increases their chances of attracting females and thus improves their reproductive success [7,10,25,26]. Evolutionary models support this hypothesis, suggesting that selection for the timing of male and female arrivals is the main driver of asynchronous sex-specific arrival times [6,21,27]. Differences in arrival timing can also be explained by different energy management, as females in spring should maintain sufficient reserves to meet the high nutritional demands of egg production, which often cannot be met only through local foraging [5]. Arriving too early under harsh spring conditions could deplete these reserves, leading to smaller clutches or lower-quality eggs [5,10]. In contrast, sperm production is far less energetically demanding, allowing males to take the risk of arriving early [5,11]. The evolutionary advantages of securing the best territories and mates outweigh the potential costs of adverse weather, especially during warmer springs [7,8,10]. This leads to sex-specific migration strategies, in which males prioritize speed of arrival, while females optimize energy costs to maintain sufficient body condition [5,11].

The earlier arrival of males at stopover sites and breeding grounds may also result from spatial sex segregation during the non-breeding season, when males and females occupy separate wintering grounds or use different migration routes [6,11,12,28,29,30,31,32]. However, studies using geolocators show that even when the sexes overwinter in the same regions, protandry is driven by behavioral differences [11,33]. In particular, males often depart from their wintering grounds earlier and stay ahead of females throughout the migration [6,33]. Furthermore, migration speed may be sex-specific, as in some species males migrate faster due to a higher rate of energy replenishment or morphological adaptations, such as more pointed wings [5,11].

Over the past few decades, the effects of climate change have become increasingly evident in the northern hemisphere, leading to long-term shifts in temperatures and weather patterns [13,15,34]. These climatic changes can directly influence the timing of spring migration [35] and potentially alter the degree of protandry by modifying the costs and benefits of early arrival for each sex [2,22,23,24]. The timing of migration is influenced by environmental factors, such as food abundance or weather conditions, including temperature and precipitation at subsequent locations visited by migrants [17,18,22,36,37]. Large-scale climate indices, such as the North Atlantic Oscillation (NAO), Indian Ocean Dipole (IOD), Mediterranean Oscillation Index (MOI), and Scandinavian Index (SCAND, hereafter referred to as SCI), which shape temperatures, precipitation and circulation patterns over vast regions of Europe and Africa, have been shown to influence the spring migration timing in many species of small passerines [17,18,38,39,40,41].

Among studies on shifts in avian migration phenology, the possibility of different responses of males and females to changes in climatic conditions has attracted some attention. Earlier analyses of species exhibiting sexual dimorphism—including the goldcrest (Regulus regulus), blackbird (Turdus merula), reed bunting (Emberiza schoeniclus), redstart (Phoenicurus phoenicurus), whinchat (Saxicola rubetra), the red-backed shrike (Lanius collurio), the pied flycatcher (Ficedula hypoleuca), and the Eurasian blackcap (Sylvia atricapilla)—showed that between 1976 and 1997, spring migration advanced in both sexes at a similar rate, thus their degree of protandry remained unchanged [10]. However, more recent studies show that, with progressively warmer winters and springs, males begin their migration much earlier than females, leading to an increasing difference in arrival dates, as shown in the pied flycatcher [42,43], the barn swallow (Hirundo rustica) [44], the song thrush (Turdus philomelos) [45], the willow warbler (Phylloscopus trochilus) [46], and the chiffchaff (Phylloscopus collybita) [47]. Uneven shifts in migration timing might result from different responsiveness of the sexes to the same conditions, or from slightly different environmental conditions they experience at non-breeding grounds due to spatial or temporal segregation, or from both mechanisms. The consequent carry-over effects can modify reproductive success, with long-term effects on local population dynamics [12,48].

The Eurasian blackcap (thereafter: blackcap), a common migrant passerine with a prominent sexual dimorphism, shows remarkable flexibility in adjusting its migrations to climate change. This plasticity manifests in reducing the distance of migration and establishing new wintering grounds in the British Isles over recent decades [49,50,51,52], as well as in major changes to their historical ranges, such as the suggested recent shift of migration of the eastern European populations from their traditional wintering grounds in the Eastern Mediterranean region to more western parts of their wintering range [53]. Spring migration phenology of blackcaps has been related to variation in several large-scale climate indices (SCI, NAO, IOD) that operate across their wide wintering grounds and migration routes, but the sexes were treated jointly, so the potential sex differences have not been considered [17,41]. For these reasons, we selected the blackcap to investigate hypotheses regarding a potential increase in protandry and sex-specific responses to climate change, analogous to those in other migratory passerines.

Earlier studies based on about 20 years of data from the late 20th century found no evidence of increasing protandry in the blackcap [10]. Using a dataset twice as long on spring migration of blackcap, we set out to reassess this pattern, expecting an increase in protandry over a longer period, considering more rapid climate change and temperature increase across Europe in recent decades. Thus, our first aim is to identify any long-term trends in the spring migration phenology of male and female blackcaps, using a 44-year data series (1982–2025) from the southern Baltic coast, and to determine if the degree of protandry has changed over this period. Secondly, we aim to determine if any relationships of the spring migration timing to large-scale climate indices differ between male and female blackcaps, which we would expect considering blackcaps’ protandry.

2. Materials and Methods

2.1. Study Species

The Eurasian blackcap (Sylvia atricapilla) is a small passerine easily sexed by plumage and size: females have a brown crown, and males are slightly larger with a characteristic black cap [54]. Blackcaps are commonly breeding across Europe (Figure 1a) in habitats with dense undergrowth, such as parks, gardens, and deciduous or mixed woodlands [55].

They are mainly insectivorous, but their pre-migratory fattening is supported by a highly flexible diet including fruits; during periods of invertebrate scarcity they can switch to pollen, nectar, and fruit [55,57]. Blackcaps exhibit considerable variation in migration habits, ranging from completely resident populations breeding in the south of Europe, to obligatory long-distance, trans-Saharan migrants in the north and east of the breeding range [55,58]. Central Europe features a migratory divide separating southwest-bound and southeast-bound autumn migration routes, with intermediate populations moving south [59,60]. Additionally, a recent evolutionary shift has led some central and northwestern breeding birds to winter in the British Isles rather than in their traditional southern wintering grounds [49,50,51,59,61], and the eastern European populations to change their historical wintering grounds in the Eastern Mediterranean region to more western areas, closer to the breeding grounds [53]. Populations breeding in western Europe are only partially migratory [55,61].

Blackcaps passing through the southern Baltic coast, which are the focus of this study, originate primarily from populations breeding in Scandinavia and the Baltic region, including local birds breeding near the study site (Figure 1a) [60,62]. These populations overwinter across the Iberian Peninsula, northwest Africa, and the eastern Mediterranean region (particularly the Levant), with some individuals migrating as far as West Africa [59,60,62,63,64]. The annual phenology of these populations (Figure 2) includes the spring passage from these wintering quarters to the northern breeding grounds during March – mid-May, the breeding season (May to August, with one or two broods), the autumn migration (mid-August to November), and the wintering period (December–February) [55].

2.2. Study Site and Methods of Catching Birds

Fieldwork was conducted at the Bukowo-Kopań ringing station, situated on the Polish Baltic Sea coast (Figure 1a), as part of the long-term Operation Baltic research project. This coastal area, overgrown by mixed forests with fruit bushes, serves as a stopover site for blackcaps heading in spring to northern and northeastern breeding territories and also supports a local breeding population [59,62]. Following the Operation Baltic standard [65], birds were caught in mist nets daily from dawn until dusk. The number of mist nets remained constant within each spring season, but inter-annual variation ranged from 35 to 61 nets (Supplement, Table S1). Upon capture, each bird received a ring, and birds’ sex and age were recorded before release. The station's habitats and procedures are described in detail by Remisiewicz and Underhill [17].

Our analysis covers spring migration data collected between 26 March and 16 May 1982–2025. Fieldwork was suspended during springs 2011 and 2020 due to logistical constraints and the COVID-19 pandemic, resulting in 42 active seasons. Blackcaps caught after 16 May were mostly local breeding individuals, confirmed by their high recapture rates and the presence of brood patches. We used only the data from the initial capture for each bird within a season, and excluded any unsexed individuals from analyses. We combined all age classes (adults and second-calendar-year birds), working under the assumption that all blackcaps caught during spring were migrating to their breeding grounds.

To identify geographical regions used by the studied blackcap populations at their different life-stages, we analysed a supplementary dataset of 105 ringing recoveries. This dataset includes birds ringed at three Operation Baltic stations (Figure 1a) and subsequently found elsewhere, as well as foreign-ringed individuals controlled at these stations, between September 1960 and June 2025. We mapped these ringing recoveries (Figure 1a) against the species' geographical distribution [56] using QGIS v. 3.44.5 [66].

2.3. Large-Scale Climate Indices

We analysed seven large-scale climate indices that represent climatic conditions across the non-breeding areas and spring migration routes of the studied blackcap populations. The monthly values of these indices were downloaded from online databases listed in Table 1.

To match these indices with the species' annual cycle, we calculated seasonal means from these monthly values for two periods: the wintering period (December–February) and spring migration (March–May) (Figure 2; Table 1). These indices serve as proxies for ecological conditions experienced by blackcaps during these life stages. We provide their brief descriptions below, focused on their positive phases; the negative phase usually brings opposite regional climate effects, but we specify when this pattern occasionally differs.

2.3.1. Mediterranean Oscillation Index (MOI)

The Mediterranean Oscillation Index is the pressure gradient between the western and eastern Mediterranean regions, reflecting opposing changes in pressure, temperature, and precipitation in these areas [67,68]. We used the MOI1, calculated as the normalized sea-level pressure difference between Algiers and Cairo [68,69]. The positive MOI represents high pressure over the southwestern Mediterranean, which in winter leads to stable, dry, and warm weather across the western and central basin [70,71], extending to the Balkan Peninsula [72,73] (Figure 1a). Opposite conditions, with increased storms, higher rainfall, and lower-than-average temperatures occur in the easternmost Mediterranean [70,74]. In spring, the positive MOI maintains dry conditions in the central Mediterranean [75], but with regional exceptions, such as increased precipitation in April and May in Morocco, because moist Atlantic air is forced over the Atlas Mountains [71]. Similarly, in the Balkan Peninsula, a positive MOI in April brings cooler or humid conditions, correlated with fewer forest fires than during negative MOI [72]. The negative MOI in winter and spring allows moist Atlantic air to increase precipitation across the western and central Mediterranean [70,73].

2.3.2. North Atlantic Oscillation (NAO)

The North Atlantic Oscillation is the surface sea-level pressure difference between the Azores High and the Icelandic Low pressure centres [76], which behave oppositely: when pressure rises in one, it decreases in the other [76]. This mechanism shapes the intensity and direction of westerly winds, controlling the transport of heat and moisture across the North Atlantic into Europe [77]. The positive winter NAO (Figure 1b) produces strong westerly winds across the middle latitudes [76], which steer warm and moist maritime air toward Europe [78,79,80]. Consequently, Northern, Central, and Eastern Europe experience warm winters, while Southern Europe, the Mediterranean, and North Africa undergo colder conditions due to the airflow from the north [76,81,82]. During the positive phase, winter precipitation is increased in Northern Europe and the central coast of North Africa, while Western and Southern Europe, the Balkans, and western North Africa remain drier [76,77,82,83]. In northwestern Africa, this drought is further driven by the reinforced Azores High, which frequently brings dry eastward winds [83]. In early spring, the positive NAO maintains these weather patterns [78,84].

2.3.3. Scandinavian Pattern Index (SCI)

The Scandinavian Pattern Index is shaped by a pressure dipole between Scandinavia and Southern Europe, with the main circulation centre over Scandinavia and weaker centres of opposite sign over Western Europe and Eastern Russia or Western Mongolia [85,86,87]. The positive phase is often associated with height centres and blocking anticyclones over Scandinavia and Western Russia (Figure 1c) [85,88]. The positive SCI phase in winter leads to low temperatures across Scandinavia and Western, Central, and Eastern Europe, while Southeastern Europe and the eastern Mediterranean basin experience warm conditions [85,86]. Considering precipitation, Northern, Central, and Eastern Europe are drier than usual [86,87]. At the same time, a low-pressure centre forms over the Mediterranean, which brings rainfall to Southern and Western Europe, particularly over the Iberian and Apennine Peninsulas [85,86,89]. The positive SCI in spring continues its cooling effect on European weather [90], and maintains drier-than-usual conditions in the eastern Baltic Sea basin, affecting the Baltic countries and northwestern Russia [87]. During the negative SCI phase, the spatial extent of the climate conditions remains identical to the positive phase, but the temperature and precipitation effects are reversed [86,87,89,90].

2.3.4. Indian Ocean Dipole (IOD)

The Indian Ocean Dipole reflects the difference in ocean surface temperatures between the western and southeastern tropical Indian Ocean, which directly affects the regional climate and exerts indirect long-distance atmospheric influences (Figure 1d) [91]. The positive IOD occurs when the western Indian Ocean is warmer than usual, and the eastern part is cooler, which influences local weather systems, causing extreme flooding in the Horn of Africa and droughts in Southern Africa [91,92,93]. Indirectly, these massive shifts in tropical heat and moisture generate large-scale atmospheric waves that affect pressure systems thousands of kilometers away. This remote influence brings warm winters to most of Europe and North Africa [94,95]. Precipitation patterns related to positive winter IOD differ regionally. Winter rainfall increases mostly in southern and western Europe, particularly on the Iberian Peninsula and at the western Mediterranean coast of North Africa, while drier conditions occur across parts of North Africa and the eastern Mediterranean basin (Figure 1d) [95]. During spring, IOD usually remains neutral in March and April, but its new cycle begins in May, triggering weather patterns similar to winter [91,96,97]. These patterns stabilise high-pressure systems in Europe and Eurasia, leading to unusually warm temperatures in spring [96]. The warming of the western Indian Ocean also brings increased rainfall in the Horn of Africa [98,99]. The negative winter IOD is directly related to reduced rainfall and drought in this region [91]. Its remote effect includes colder-than-average winters across the western and central Mediterranean basin and North Africa, but warmer conditions in northern and eastern Europe, and drier conditions in southwestern and eastern Europe, the Balkans, and the eastern Mediterranean basin (Figure 1d) [94,95].

2.3.5. Long-Term Trends of Large-Scale Climate Indices and Their Correlations

We checked for multi-year trends in all used large-scale climate indices (Table 1) over the study period using simple linear regression models. The only statistically significant trend occurred for the spring SCI, which showed a significant decrease over 1982–2025 (Figure S1). We also checked if these large-scale climate indices were correlated using Pearson’s correlation coefficient, applying Benjamini and Hochberg's [100] correction for multiple comparisons. Only three correlations were statistically significant (Table S2), and the strongest correlation between the winter NAO and winter MOI was r = 0.68, thus we avoided using these two variables in one multiple regression model. All correlation coefficients were lower than the recommended threshold of |r| = 0.7, indicating that the effects of multicollinearity on the results of models including these climate variables would be negligible [101].

2.4. Methods of Data Analysis

2.4.1. Analysis of Long-Term Trends in Spring Migration Phenology and Protandry

First, we used individual capture data to calculate the dates of percentiles of passage of males and females in each year: the 10th (p10) (beginning), 50th (p50) (median), and 90th (p90) (end of passage), expressed as the day number in the year (1 January = Day 1). We used these dates as parameters reflecting the timing of these three phases of migration of each sex in each year, as in many other studies [10,102,103,104,105]. We excluded four years from the analyses when fewer than 5 females were captured, as females were fewer than males among captures (Table S1). Thus, we finally analysed data from 38 springs during 1982–2025, including 2912 females and 4345 males, in total 7257 blackcaps (Table S1).

To estimate multi-year trends in migration timing for each sex, we used linear regression of the dates of p10, p50 and p90 for each sex against the year number from 1982 (YearN) as the continuous explanatory variable. We considered the degree of protandry as the difference in the dates of these percentiles between males and females, as in Tøttrup and Thorup [10]. To check for protandry and to test whether its degree had changed over the years, we ran three multiple linear regression models, one for each of the time-series of these percentiles’ dates, including the standardised year (YearSt) and sex (SexN, coded as 0=males, 1=females), and their interaction as explanatory variables. We applied the “all subsets” approach using the “dredge” function in the “MuMIn” package [106], and we ranked the models using the Akaike Information Criterion corrected for small sample sizes (AICc). The model with the lowest AICc was considered the best model, which we then interpreted.

2.4.2. Estimating the Effects of Large-Scale Climate Indices on Different Phases of Blackcap Spring Migration

To test the effects of large-scale climate indices on the timing of these three phases of passage, we ran multiple linear regression models where the time series of the dates of p10, p50 and p90 in 1982–2025 were response variables, and the explanatory variables included the year, the four large-scale climate indices (NAO, MOI, SCI, IOD) in winter and spring (Table 1), and the interactions of these climate variables with sex. All explanatory variables were standardised before being included in the models, as they were at different scales. Because these variables were too many to be included in one model, we first ran the models separately for these indices in winter months (December–February) in the "winter models", and indices for spring months (March–May) in separate "spring models". Because winter MOI and winter NAO were correlated at r = 0.68 (Table S2), to avoid multicollinearity, we ran the partial winter models for two sets of variables: one excluding winter MOI, and the other excluding winter NAO. Thus, we ran nine models: two “winter” models and one “spring” model for each of three migration phases. We conducted the model selection using the “all subsets” approach and model ranking according to AICc, using the “MuMIn” package [106], as previously.

Comparing these partial “winter” models, we found that models including winter NAO consistently explained a higher proportion of the variation in the migration timing, as they had higher Adjusted R² (Adj R²) values than the models including winter MOI. Therefore, in the second step, we used the variables selected in the best “winter” models including winter NAO (not winter MOI), and in the best “spring” models.

Next, for each migration phase (p10, p50, and p90), we ran “winter and spring” models, combining the winter and spring climate indices selected by these best models. In these multiple regression models, we included those interactions of these indices with Sex that were retained by the respective “winter” or “spring” best models in the first step. We applied the same “all subsets” approach and AICc model ranking as earlier, to identify the final best “winter and spring” models. To determine the percentage of explained variation in the response variable, we extracted Adj R² for each final model using the “broom” package [107].

We evaluated multicollinearity among predictors in these models using the Variance Inflation Factor (VIF) from the “car” package [108]. To test for potential overfitting, we cross-validated the final best models using a three-step approach recommended by [109]: (1) visual inspection of diagnostic plots to identify outliers and highly influential data points based on Cook’s distance; (2) evaluation of the ratio of Adj R² to the multiple coefficient of determination (R²), where a ratio approaching 1 indicates an optimal model fit; and (3) comparison of Adj R² with the Predictive R² (predR²). The predR² was calculated from the Predicted Residual Error Sum of Squares (PRESS) using a Leave-One-Out Cross-Validation procedure using custom R functions [110]. In this approach, each data point is sequentially removed, the model is refitted, and its ability to predict the omitted observation is evaluated. A minimal difference between Adj R² and predR² indicates that the model is stable and not overfitted [110]. This application of multiple regression, evaluation of multicollinearity and stability of final models followed earlier studies [18,40,41]. The final “winter and spring” models provided estimates of the effect of each selected climate variable on the dates of p10, p50, and p90 of the spring passage. We visualized these effects as scatter plots showing relationships of these variables to the dates of these phases of passage and fitted linear regression lines, using the “ggplot2” package [111]. All statistical analyses were conducted in R 4.4.2 [112].

3. Results

3.1. Protandry and Sex-Specific Long-Term Trends in the Timing of Spring Migration of Blackcaps

During 1982–2025, the dates of the beginning (p10) of spring passage for both sexes shifted earlier by 9 days (Figure 3; Table S3).

The protandry in blackcaps at Bukowo-Kopań occurred at the beginning of spring passage, because p10 was on average 2.2 days earlier for males than for females (Figure 3; Tables S4 and S5). This degree of protandry remained stable over 1982–2025 (Figure 3, Tables S4 and S5), as indicated by the parallel regression lines (Figure 3), and a lack of the YearSt:SexN interaction in the best model for p10 (Table S5). The median dates (p50) of passage did not differ between the sexes (Figure 3), as confirmed by the absence of the sex effect (SexN) in the best model (Table S5). The p50 dates for males shifted earlier by 9 days and for females by 7 days, over 1982–2025 (Figure 3; Table S3), but these advancing trends were similar, as shown by no YearSt:SexN interaction in the best model (Table S5). The dates of the end of migration (p90) showed no trends over time and did not differ between the sexes (Figure 3; Table S3), as confirmed by a lack of the effect of sex (SexN) in the best model (Table S5).

3.2. Effects of Large-Scale Climate Indices on Spring Migration Timing

The “winter” models including NAO (but not MOI) explained 7% – 27% of the variation in the timing of different phases of spring migration (Table S7), while the models including MOI (but not NAO) explained 7% – 25% of this variation (Table S9). Because the “winter” models including winter NAO consistently explained a higher proportion of the variation than those including MOI (Tables S7 and S9), we used the outcome from the first approach in the combined “winter and spring” models. The best “spring” models explained from 0% to 44% of the variation in the timing of p90 and p10, respectively (Table S11).

The best “winter and spring” model (Table S12) explained 48% of the variation in the date of beginning of passage (p10) at Bukowo-Kopań during 1982–2025 (Table S13). We found no differences between the sexes in the relationships of p10 to the climate indices. For both sexes, p10 was early with high winter NAO (Figure 4a, Table S14) and high spring SCI (Figure 4b, Tables S13 and S14), or late with low values of these indices, as these relationships work both ways.

For the median date of migration (p50), the best model explained 39% of its variation in 1982–2025 (Table S13). No sex-specific differences occurred in the effects of large-scale climate indices. The median for both sexes was early with high winter NAO (Figure 4a, Table S14), high spring NAO (Figure 4c, Table S14), and high spring SCI (Figure 4b, Table S14). The best model for the end of passage (p90) explained 7% of its variation over the 44 years (Table S13). The sexes did not differ in their relationships to any of the climate indices. The date of p90 for both sexes was late with high winter IOD, or early with low values of this index (Figure 4d; Tables S13 and S14).

Diagnostic tests confirmed the stability of the final “winter and spring” models (Table S15, Figure S2). Minimum discrepancies between the Adj R² and pred R² and high ratios of Adj R² to R² (range: 0.84–0.94) indicated that the models were not overfitted (Table S15). Visual inspection of diagnostic plots confirmed that the assumptions of the regression models were met, with no highly influential outliers (Figure S2).

4. Discussion

Our study shows that both male and female blackcaps advanced their spring passage through the Baltic coast at a similar rate, hence the degree of protandry remained stable over the last 44 years, opposite to what we expected, considering the pattern in other passerines. Protandry was the most apparent at the very beginning of passage, but decreased as the spring progressed. Annual variation in migration timing was related to large-scale climate indices in both sexes similarly, contrary to our expectations. However, we found that the timing of subsequent migration phases in both sexes was shaped by different climate indices, suggesting that successive cohorts of males and females arrived along different spring migration routes. We discuss these findings in the context of the mating opportunity hypothesis and changing climate conditions across blackcap wintering grounds and migration routes.

4.1. Long-Term Phenological Trends in Both Sexes and Drivers of Spring Protandry

Protandry is common among migratory birds, driven mainly by the selective pressures of the mating system [6,11]. The mating opportunity hypothesis explains that earlier arrival enables males to occupy high-quality territories to improve reproductive success [10,26]. These phenological differences are related to energy management. Females must maintain sufficient fat and protein reserves for egg production, so they optimise energy costs and avoid harsh conditions early in spring that could deplete these essential reserves [5]. In contrast, males prioritise migration speed over energy saving [11]. Thus, blackcap males exhibit a time-minimising strategy and females an energy-maximising strategy during spring migration [113]. Furthermore, the earlier passage of males may result from spatial sex segregation at the non-breeding grounds. In blackcaps, males might winter in greater proportion at more northern latitudes than females, having a shorter migration distance to cross [30,114,115].

We suggest that both mechanisms contribute to the spring protandry we found at the beginning (p10) of passage at Bukowo-Kopań over 1982–2025, in line with earlier studies [10,116]. The protandry among the first migrants supports the mating opportunity hypothesis. Competition for breeding opportunities creates evolutionary pressure on males [7,117], as the first-arriving individuals secure the best territories and attract the earliest females [118,119,120]. In contrast, the timing of the median (p50) and the end (p90) of passage at our station was similar for both sexes, suggesting that the pressure for early male arrival fades as the season progresses. The lack of protandry in the median dates of passage contrasts with results from Helgoland (North Sea, Germany) and Christiansø (Baltic Sea, Denmark), where this phase of passage was significantly earlier in male than in female blackcaps [116]. This discrepancy might result from differences in the composition of populations passing through these stations compared to Bukowo-Kopań, or the different time periods analysed (1982–2025 in our study vs. 1976–1997 in Rainio et al. [116]).

The degree of protandry at the start of passage did not change over our 44-year study period, as males and females advanced their migration timing at a similar rate. Furthermore, the median date of passage advanced similarly for both sexes, despite no clear protandry at that phase of migration. This uniform phenological shift aligns with earlier studies, which also reported no changes in the degree of protandry over 1976–1997, both for the combined data from Helgoland and Christiansø [116] and for Christiansø alone [10]. The similar advancement suggests that earlier spring passage benefits both sexes equally. Mild springs might enable males to reach breeding grounds early to maximise mating opportunities, and allow females to lay eggs earlier, extending the breeding season and increasing chances of raising two clutches for both sexes [20,26].

The advancement of the start of migration by 9 days (for both sexes) and of the median by 7–9 days at Bukowo-Kopań is consistent with the shift to earlier median migration dates reported during 1976–1997 for combined data from Helgoland and Christiansø [116]. Our results also correspond with the negative trends observed at Christiansø alone [10] that indicated an advance in arrival at all migration phases (5%, 50%, and 95%), although these trends were not statistically significant, likely because of a shorter, 22-year time series. More recent data from the Blåvand bird observatory in Denmark also showed significant advancement of the first and last 5% of the passage of blackcaps during 1984–2022; however, unlike our study, this analysis combined males and females [105]. The lack of any multi-year trend in the timing of p90 in our study differs from the significant trend at Blåvand [105], probably because at Bukowo–Kopań the locally breeding blackcaps arriving at the end of spring might have masked any trends. A weaker phenological shift with the progress of the passage is a common pattern across all the mentioned stations [10,105]. This might result from a slower migration of birds in the “tail” of passage, which often includes weaker individuals, which likely were “outcompeted” by the preceding fitter individuals, but might still be able to raise some chicks [121], hence smaller time pressure. The overall shift to earlier passage for the first half of the population suggests that blackcaps respond to increasingly mild conditions on their wintering grounds and along spring migration routes [2,10,22], which increases their chances to raise two clutches during extended breeding season [26]. Similar advancement suggests that earlier spring passage benefits both sexes equally. Mild springs might enable males to reach breeding grounds early to maximise mating opportunities, and allow females to lay eggs earlier, extending the breeding season and increasing chances of raising two clutches for both sexes [20,26].

4.2. Similar Responses of Both Sexes to Large-Scale Climate Indices

Our key finding is a lack of differences in male and female blackcap relationships to any of the large-scale climate indices. This differs from our initial expectations that males should respond stronger to some climate patterns, especially those that shape winter and spring conditions in Europe, as male blackcaps generally winter closer to their breeding grounds than females [6,11,12,28,29,30,31,32]. Unlike the song thrush, in which after warm winters and springs the males arrived in the Baltic region much earlier than the females [45], in the blackcap the relationships of both sexes to climatic conditions were similar, despite the protandry at the beginning of passage (p10) at Bukowo-Kopań. This difference might reflect species-specific phenological plasticity, with the song thrush and both sexes of the blackcap showing the same level of flexibility in response to changing conditions. Both species share part of their wintering grounds in the Mediterranean region, but Redlisiak et al. (2021) analysed regional mean temperatures from southern and south-western Europe. If male and female song thrushes are partly segregated across the wintering grounds, as other passerines, they might be exposed to different local temperatures, which could affect their migration differently. This contrasts with the large-scale climate factors we analysed for blackcaps, which have a climatic footprint over much larger areas of the wintering range of the blackcap, from northern Africa through the Mediterranean, up to central Europe [76,122]. Even though females winter farther south than males due to spatial segregation [11,30], both sexes likely experience the same broad weather conditions dictated by these large-scale oscillations, and adjust their migration timing accordingly. Additionally, female blackcaps appear to be as phenologically adaptive as males, likely because the benefits of advancing spring migration and increasing a chance for second broods balance the risks posed by cold spells in spring in females similarly to the males [20,26]. This uniform response of both sexes is consistent with findings from Helgoland and Christiansø, where the winter NAO did not affect the degree of protandry in blackcaps [116], and with the conclusion that both sexes respond similarly to changing climatic conditions [10]. Protandry in blackcaps appears to be maintained by the selective pressures described by the mating opportunity hypothesis [7,10,20], regardless of the weather conditions in a given season.

4.3. Effect of the North Atlantic Oscillation (NAO) on Spring Migration Timing

Positive winter NAO was related in both sexes to an advanced beginning (p10), and, with a weaker effect, to the median (p50) of passage, but not to its end (p90), a pattern similar to the responses to winter NAO in blackcaps migrating through other locations [10,102,105,116]. Our results for the beginning of passage contrast with a lack of significant relationship between the first arrival dates of blackcaps and winter NAO shown by long-term (1881–2007) observations from the Czech Republic [123]. This suggests that first blackcaps arrive there from areas under a weaker climatic influence of NAO, like the Eastern Mediterranean or southeast of Europe, indicated by ringing recoveries [60], than the first migrants at the more northern and coastal locations in Europe , including Bukowo-Kopań. The relationship between NAO and the beginning of passage at our station suggests that the first migrants winter close to the breeding grounds, in areas where positive NAO brings mild winters, such as western and central Europe [76,82]. Warm winters reduce energy expense for thermoregulation and lead to an earlier availability of insect prey, allowing these short-distance migrants to reach migration condition faster and depart early. The relationship we found between winter NAO and advancement of p50 contrasts with earlier records from Helgoland and Christiansø, where no analogous effect was found [116], possibly because they analysed a shorter time series. The relationship we revealed might also reflect a more recent shift in wintering patterns of the eastern population of blackcap, from eastern to more western wintering grounds [53], which are under climatic influence of NAO. Positive spring NAO was related to the advanced median of passage (p50) for both sexes, in line with earlier results showing analogous shifts for both sexes jointly [17]. Positive NAO brings warm spring and south-westerly winds across central and western Europe, and this effect of NAO is most pronounced in March, about the time that migrant blackcaps moving northward enter this region [76,78,79,84]. Winds from the south-west can assist migrants in their movement to the north-east across Europe. Warm conditions might benefit migrants more indirectly, accelerating development of spring vegetation and increasing insect abundance at stopover sites [22,124].

4.4. Effect of the Scandinavian Pattern Index (SCI) on Spring Migration Timing

Positive spring SCI, which brings warm weather and thus likely favorable food conditions to the central and western Mediterranean [85,86,89] was related to an advanced start (p10) and median (p50) of migration of both sexes at Bukowo-Kopań. This suggests the first half of migrants benefit from these conditions, which allow for quick fuel accumulation and thus shorter stopovers on spring migration. Following the mating opportunity hypothesis [7,10,20], we suggest that the migrants that form the first phase of passage (p10) winter close to the breeding grounds, probably at the northern edges of the central and western Mediterranean region. With positive SCI, they might experience the spring peak in food availability in these areas, which enables an early departure for migration. Birds that form the middle phase (p50) of passage, arriving from more southern wintering areas, might experience such conditions during their stopovers in these areas, which results in an analogous phenological shift. These phase-specific responses align with previously found relationship between the positive spring SCI and advancement of the overall spring passage of the blackcap at Bukowo-Kopań [17]. By dividing the migration into phases, our study reveals that this overall relationship has been primarily driven by a response of the first half of migrants to environmental conditions shaped by SCI.

4.5. Effect of the Indian Ocean Dipole (IOD) on Spring Migration Timing

Positive winter IOD, which brings dry spells to the eastern Mediterranean [95,125], was related to the delayed end of migration (p90) in both sexes, which suggests that blackcaps that form the end of spring passage at Bukowo-Kopań stopover in this region, as indicated by ringing recoveries (Figure 1a)[60]. A lack of rain in the Mediterranean region limits vegetation growth and food availability [126,127], which might extend the time the birds need to build fuel reserves, delay their departure, and ultimately their arrival at the Baltic coast. An earlier study did not find an effect of winter IOD on the overall spring migration of the blackcap at Bukowo-Kopań [17], but analysing the Annual Anomaly of the overall migration might mask the climatic influences on the last migrants (p90). However, multi-species studies showed relationships to the winter IOD in the spring migration timing at Bukowo-Kopań of the common redstart [17] and the chiffchaff [41], which share wintering areas affected by IOD with the blackcap. In the chiffchaff, positive winter IOD delayed the last fraction of passage [41], similarly as we found in blackcaps of both sexes. These analogies indicate that IOD influences similarly several species of long-distance migrants, suggesting they arrive at the southern Baltic coast from the south-east, along similar spring migration routes.

5. Conclusions

We revealed that both male and female blackcaps advanced the first half of their spring migration at the Baltic coast at a similar rate; thus the two-day protandry, manifested especially at the beginning of the passage, remained stable during the 44 years of our study. This stable protandry can be explained by the mating opportunity hypothesis, which suggests that selective pressures preserve the difference in spring timing between the sexes, regardless of the weather conditions varying between seasons. Despite this protandry, both sexes responded uniformly to atmospheric patterns in winter and spring, which suggests they affect similarly broad non-breeding grounds of both sexes. We suggest that females are as phenologically adaptive as males, to maximise common reproductive outcome for both sexes. Such phenological response to climate change apparently benefits the species, considering its almost three-fold increase in numbers across whole Europe, including the Baltic region, since the 1980s [128].