Ducks change wintering patterns due to changing climate in the important wintering waters of the Odra River Estuary

Introduction

Migration distance has declined in several species of aquatic (and other) birds as a result of climate change (Musil et al., 2011; Lehikoinen et al., 2013; Meller, 2016). The distances that birds migrate from their breeding areas in northern and eastern Europe to their central European wintering areas are shorter during mild winters (Lehikoinen et al., 2013; Pavón-Jordan et al., 2015); conversely, birds may change their wintering sites to warmer regions during colder periods because they may perceive local manifestations of larger scale weather (Newton, 2008). Reducing migration distance can provide several benefits associated with earlier arrival at the breeding grounds and greater survival (Coppack & Both, 2002; Jankowiak et al., 2015a; Jankowiak et al., 2015b). Food resources of wintering sites may also influence migration decisions (Cresswell, 2014; Aharon-Rotman et al., 2016). Although winter site fidelity is usually very strong among waterfowl (Newton, 2008), this can change in response to weather, habitat, and competition (Cresswell, 2014). Changing winter sites may often be a trade-off between the costs of finding a new site and the benefits it offers (Aharon-Rotman et al., 2016). At sub-zero temperatures, shallow waters freeze over; forcing birds to expend more time and energy searching for food in deeper, open waters. Three functional groups of waterbirds forage in the shallow waters of offshore lagoons: piscivores, herbivores and bottom-feeders, for example, large numbers of waterbirds gather to forage in the Odra River Estuary (hereafter ORE) (Marchowski et al., 2015; Marchowski, Jankowiak & Wysocki, 2016). Two groups of waterbirds—bottom-feeders and piscivores—are among those most commonly wintering here. During winter, the study area is subject to wide variation in temperatures, often making surface waters subject to freezing (Van Erden & De Leeuw, 2010). Yet even relatively small variations in temperature, causing ice cover to form or disappear, can lead to the displacement of waterbirds. Changes in abundance and community structure of birds in the ORE may reflect the impact of climate change. Analysis of the dates of the appearance of ice-related phenomena in the Szczecin Lagoon and of their frequency over time reveals a distinct pattern illustrating recently observed trends in climate warming (Girjatowicz, 2011). In this paper, we are looking for how the abundance of some species in the ORE changes due to climate warming. We are investigating whether climate change will differentially influence foraging patterns, and consequently overwintering patterns, of these two groups of birds. Thus, increasing temperatures due to climate change, and the shorter time interval with ice cover will result in increasing numbers of bottom-feeders because they will then migrate shorter distances. Frosts in the study area are never so severe that the water freezes completely to the bottom even in the shallows. But even if ice does cover the surface of shallow water bottom-feeders have no access to food and have to move to warmer areas because feeding areas where sedentary mussels are abundant tend to be in shallow waters (Marchowski et al., 2015). Piscivores, on the other hand, can still feed in such conditions (frozen surface of shallows, unfrozen deeper areas - further from the shore, where there are no mussels but there are fish) and remain in the area, e.g., observation of large aggregations of piscivores during harsh winters (Smew Mergellus albellus and Goosander Mergus merganser) (Kaliciuk, Oleksiak & Czeraszkiewicz, 2003; Czeraszkiewicz, Haferl & Oleksiak, 2004; Marchowski & Ławicki, 2011; Guentzel et al., 2012; Marchowski & Ławicki, 2012; Marchowski, Ławicki & Guentzel, 2013). This has consequences for conservation management plans in protected areas. Two different groups of birds react differently to climate warming, showing different patterns of moving closer to their breeding grounds, as a consequence in our area should be more bottom-feeders. Thus, changes in the importance of wintering ground due to changing surface-water freezing patterns expected under global warming regimes are likely to have important consequences for very large numbers of these birds.

Study area

The study area in the south-western Baltic Sea forms the Polish part of the Odra River Estuary system, which includes the Great Lagoon (the Polish part of the Szczecin Lagoon), Świna Backward Delta, Kamień Lagoon, Dziwna Strait and Lake Dąbie (522.58 km², Fig. 1). The area comprises four interconnected Important Bird Areas (IBA) and also a Natura 2000 area (Wilk et al., 2010). The average and maximum depths of the estuary are 3.8 and 8.5 m, respectively; the dredged shipping lane passing through the estuary from the Baltic Sea to the port of Szczecin is 10.5 m deep (Radziejewska & Schernewski, 2008). Waters of the Szczecin Lagoon, Kamień Lagoon, and Lake Dąbie are brackish. Salinity in the central part of the estuary varies from 0.3 psu to 4.5 psu (mean = 1.4 psu) and declines with increasing distance from the sea (Radziejewska & Schernewski, 2008). Average winter temperature is 0.3°C (Weatherbase, 2016). The ORE is subject to high levels of eutrophication (Radziejewska & Schernewski, 2008). Communities of benthic organisms are typical of freshwater bodies, and the fauna includes large populations of zebra mussels Dreissena polymorpha, which was introduced in the mid-19th century. By the 1960s, the biomass of zebra mussels in the Szczecin (Great) Lagoon was estimated at 110,000 metric tons (Wiktor, 1969; Wolnomiejski & Woźniczka, 2008). The distribution of the zebra mussel is extremely uneven (see the map in Marchowski et al., 2015). The average density of the zebra mussel in the ORE is 0.18 kg /m², but the vast majority occupies around 10% of the area, where the mean density is 2.05 kg/m² (Stańczykowska, Lewandowski & Czarnoleski, 2010). Fish are mainly freshwater species such as roach Rutilus rutilus, bream Abramis brama, pike Esox lucius, perch Perca fluviatilis and ruff Gymnocephalus cernua; there are also anadromous fish including smelt Osmerus eperlanus and occasionally herring Clupea harengus among others (Wolnomiejski & Witek, 2013).

Methods

Results

Below, we present the results of our analyzes, the main one jointly examines two whole groups: piscivores and bottom-feeders and the second one where each species was analyzed. Bird numbers by feeding group were different in their relationships with ice cover; bottom-feeding species in the study area were more sensitive to lower temperatures and left sooner when colder weather increased ice cover, whereas a number of piscivorous species had a higher tolerance to the extent of ice cover (Table 2). Interactions between feeding group and season, feeding group and ice cover, and feeding group and maximum ice extent on the Baltic sea were all important (Table 2). However, the strongest effects were interactions with ice cover, then interaction with the season; this translates into increases in the importance of the site for bottom-feeders. The effect of maximum ice extent was very small (Table 2). Population indices in the ORE changed over the last 25 years, increased in the case of bottom-feeding species but decreased for piscivorous species (Table 2). Ice cover across the whole Baltic Sea had the same, though weak, impact on both functional groups of birds, numbers of birds in the ORE declined with expanding ice cover in the Baltic (Table 2).

Where particular species are concerned, the situation is more complex. The population indices of Scaup and Tufted Duck in the ORE exhibited an increasing trend (β_Scaup = 0.026 ± 0.010s.e., p = 0.011; however for Tufted Duck it was marginally insignificant β = 0.008 ± 0.004 s.e., p = 0.058), despite the general decline in their entire northern and western European populations; numbers of both species in the ORE were adversely affected by ice cover in that region (β_Scaup =  − 0.065 ± 0.011s.e., p < 0.001; β_Tufted = 0.032 ± 0.005s.e., p < 0.001) but not by ice cover in the whole Baltic (β_Scaup =  − 0.099 ± 0.074s.e., p = 0.186; β_Tufted = 0.010 ± 0.029s.e., p = 0.735). Relative numbers of Pochard in the ORE have declined (β =  − 0.009 ± 0.003s.e., p = 0.005), but so has the whole northern European population, abundance was negatively impacted by ice cover in both the study area (β =  − 0.015 ± 0.003s.e., p < 0.001); and in the entire Baltic (β =  − 0.053 ± 0.023s.e., p = 0.023). For Goldeneye, the index for the ORE population was unchanged (β =  − 0.004 ± 0.004s.e., p = 0.275), despite the increase in the European population, abundance was negatively impacted by ice cover in the study area (β =  − 0.016 ± 0.004s.e., p < 0.001), but not by ice cover in the entire Baltic (β =  − 0.044 ± 0.027s.e., p = 0.107). Relative numbers of Coot in the ORE remained unchanged (β < 0.001 ± 0.002s.e., p = 0.915), despite the slight increase in the European population, abundance was negatively impacted by ice cover in both the study area (β =  − 0.014 ± 0.003s.e., p < 0.001), and in the entire Baltic (β =  − 0.038 ± 0.016s.e., p = 0.019). The ORE population index for Smew decreased (β =  − 0.020 ± 0.008s.e., p = 0.024), despite the increase in its flyway population, abundance in the ORE was unaffected by ice cover either in the study area (β =  − 0.013 ± 0.010s.e., p = 0.183), or in the Baltic as a whole (β =  − 0.079 ± 0.061s.e., p = 0.204). Finally, relative numbers of Goosander in the ORE remained unchanged (β =  − 0.005 ± 0.008s.e., p = 0.572), like those of the whole population wintering in north-western and central Europe; abundance in the ORE was unaffected by ice cover either in the study area (β = 0.018 ± 0.009s.e., p = 0.063) or in the Baltic as a whole (β =  − 0.111 ± 0.057s.e., p = 0.057). The details relating to all these species are listed in Table 1, presented in Fig. 2A and Table S3. Table 3 summarizes the changes in the importance of the ORE for wintering populations of diving waterbirds in the last 25 years. There was a negative relationship between year and winter ice (β_intercept = 82.011, s.e. = 28.49, β_winter =  − 00.40, s.e. = 0.0142, Chi-square = 8.001, df = 1, p = 0.005; Fig. 2B).

Discussion

The phenomenon of freezing in our study area has decreased over time (Girjatowicz, 2011, Fig. 2B), so that target birds species should tend to feed more recently more often than in the past. However, two functional groups of waterbirds—bottom-feeders and piscivores—react differently to ice cover, a factor that is directly connected to climate change; this has consequences for the wintering patterns of these species. Bottom-feeding birds (Scaup, Tufted Duck, Pochard, Goldeneye, and Coot) tend to be more sensitive to ice cover in the study area than piscivores (Smew and Goosander). Piscivores can survive in colder areas, closer to their breeding ranges, but bottom-feeders have to move further south and west. This phenomenon indicates that piscivores are declining in our study site because they are shifting further north and east in order to stay closer to their breeding areas. Bottom-feeders are increasing their number for the same reason—they, too, are moving further north and east—but in their case, the result is a greater number in our study area and a smaller one in areas further west and south. Bottom-feeding birds forage in the ORE mainly on mussels of the genus Dreissena (Marchowski et al., 2015; Marchowski, Jankowiak & Wysocki, 2016); the highest quality of this food resource is found in water 1–2 m deep (Wolnomiejski & Witek, 2013). Shallow water freezes over faster, displacing birds to deeper unfrozen areas where food is accessible only with difficulty. In addition, when ice cover is present, the abundance of food in unfrozen areas declines owing to its greater exploitation, because the birds congregate on a limited area. In the case of piscivorous birds, increasing ice cover does not affect their numbers negatively. The ORE is never completely covered by ice: the shipping lane between Świnoujście and Szczecin is kept free of ice (Girjatowicz, 1990; Girjatowicz, 2005), and there are always other areas free of ice, especially at the mouths of the small rivers flowing into the estuary. These ice-free areas may still abound in fish and provide food for piscivores. In general, there is growing importance of the study area for the bottom-feeders. The main analysis jointly examines two whole groups—piscivores and bottom-feeders; the results of it are the positive effect of not freezing for bottom-feeders and negative effect for piscivores, and also a positive effect of the season for bottom-feeders, and negative effect for piscivores (see Table 2). In the other analysis, which examined each species separately, there are differences between them (see Fig. 2A). As far as the most numerous species—Scaup and Tufted Duck—are concerned, the results tally with those of the first analysis, i.e., their numbers are increasing (correlation with season). For the less abundant species, the result of the second analysis is different: Pochard numbers, for example, are decreasing (negative effect). In the first analysis, abundance and season were negatively correlated in piscivores. In the second analysis, in which we examined species separately, abundance and season were independent in the Goosander, while abundance and season were negatively correlated in the Smew. In fact, the strongest effect of the first years of the study relates to the piscivores, especially Smew. If we take the entire period (1992–2016), the Smew declines in numbers (confer with model results), which was due to their being very abundant in the early 1990s. From 2002–2016, Smew numbers are stable.

An interesting result is the negative effect of maximum ice cover in the entire Baltic Sea on the numbers of all species in our study. This is unexpected since our study area is in the warmer south-western Baltic, where one would anticipate an increase in the number of waterbirds in such circumstances (Alerstam, 1990). The explanation for this relationship is not easy and certainly goes far beyond the scope of this work, but we can speculate on possible scenarios. On maps with the maximum range of ice cover in the Baltic Sea we can see clearly that when the northern Baltic, i.e., the Gulf of Bothnia and the Gulf of Finland, is completely frozen over, the entire Pomeranian Bay (SW Baltic) (see the map—Fig. 1) together with the ORE is also covered with ice (Finnish Meteorological Institute, 2017). These areas freeze over quickly because of their shallowness and low salinity, the latter being due to the considerable influence of fresh water from the Odra river basin. Consequently, during harsh winters, birds from northern Baltic move to the south and west, but they by-pass our study area as it is covered by ice. Under such circumstances, there may sometimes be better conditions for waterbirds in areas farther north, e.g., the southern coast of Sweden, where there is no ice cover (Finnish Meteorological Institute, 2017). Worth noting here, however, is that such cold weather causing the entire Pomeranian Bay and Odra River Estuary to freeze over is rare and becoming rarer (EEA-The European Environment Agency, 2017).

The global temperature has risen about 1°C over the last 130 years, and Northern Hemisphere temperatures of the last 30 years have been the highest in over 800 years (Stocker et al., 2013). The extent and duration of ice cover in the Baltic have decreased on average by 50% over the last 36 years (Schröder, 2015). There is evidence that the range and occurrence of migratory birds have changed in response to climate change and that some species have shortened their migratory movements by wintering closer to their breeding areas (Musil et al., 2011; Lehikoinen et al., 2013; Pavón-Jordan et al., 2015; Meller, 2016). Assuming continued climate warming, the negative correlation between numbers of bottom-feeding birds and the number of days with ice cover indicates that the ORE is becoming more important for this group of birds. Climate change seems to be the primary reason for increases (in the study area) in numbers of Scaup and Tufted Duck and decreases in numbers of Smew; this corresponds with the findings of Lehikoinen et al. (2013) in the case of Tufted Duck and of Pavón-Jordan et al. (2015) in the case of Smew. Our results are important for conservation planning. Declines in the flyway populations of species such as Scaup and Tufted Duck, even though the importance of our study area to these species is increasing, but at the same time, there is an increase in exposure to locally emerging threats. The biggest threats to these species in the area include fishery bycatches (Žydelis et al., 2009; Bellebaum et al., 2012). The ecology of diving birds makes this type of threat responsible for the extra mortality of all species covered by this study. Comparison of a species’ estimated total population numbers (Nagy, Flink & Langendoen, 2014) with numbers for the ORE is interesting, since local trends and European trends do not always concur. The different responses of particular species to the factors investigated are also worth examining. We grouped the species by trends in the study area and discuss these for each species below.

Conclusion

Climate change can influence the distribution of overwintering waterbirds. Apart from climate changes, however, feeding ecology, interspecific competition, fishery and other human-related disturbance may be also important and should be taken into consideration (Quan, Wen & Yang, 2002; Žydelis et al., 2009; Clavero, Villero & Brotons, 2011; Eglington & Pearce-Higgins, 2012). Protected areas covered by our study will be more important for some species (Scaup and Tufted Duck) but less so for others (Smew). Taking into account the large abundance of the target species regularly present in the ORE, conservation measures applied here will have a large impact on entire populations and will be particularly important for Scaup. Shifts in species distributions should be accounted for in future management plans for Special Protection Areas of the European Natura 2000 network. Our results add new insight to the problem of wintering waterbirds protection and can help to shape conservation policy in the southern Baltic.

Supplemental Information