Introduction

The migratory strategy of long-distance migrating herons is poorly understood (Bernick 2005; Kushlan and Hancock 2005; Zwarts et al. 2009; Van der Winden et al. 2010). Only in Purple Heron Ardea purpurea has the post-breeding migratory strategy been well studied using satellite transmitters (Van der Winden et al. 2010). Tracked individuals of this species migrated using flapping flight, mainly at night and without refuelling en route, with only a few short diurnal stops, covering the ca. 4000 km from the Netherlands to the Sahel in 5–7 days.

The main aim of this study was to investigate the post-breeding migration strategy of Night Herons from southern Poland, one of the northernmost stable breeding populations of this species in Europe (latitude 50°N; Betleja 2001). This implies a flight distance from Poland to the Sahel of at least 4,000 km (Betleja, 2001; Zwarts et al. 2009). As flapping flight incurs relatively high energy costs, we expected that, because of the higher wing loading (4.8 kg/m2) in comparison with Purple Heron (3.6 kg/m2), Night Heron should make at least one long stopover to refuel during migration, as do other species with a high wing loading (Alerstam et al. 2007; Fig. 2 in Van der Winden et al. 2010; e.g., Beekman et al. 2002). We tested this prediction by equipping adult Night Herons with GPS/GSM transmitters to track their migration.

Methods

Three adult Night Herons were tagged with GPS/GSM-transmitters (DUCK III, Ecotone); the battery was charged by solar panel. The birds were captured in 2012 in mist-nets (Ecotone) close to the breeding colony in carp ponds at Brzeszcze-Nazieleńce (Poland, Upper Vistula Valley, 50°0′33.80″N, 19°7′53.41″E). The extra weight of the transmitters and Teflon harness varied from 3.4 to 3.9 % of body mass. The transmitters provided accurate information on geographical location and recorded the birds’ positions every 3, 6, 12, or 24 h (usually every 6 or 12 h). The duration of the intervals depended on the level of battery charge and was manually controlled via an online panel.

Stopovers are defined as long rests (>12 h, usually several days) during migration at suitable foraging sites (Van der Winden et al. 2010, Rappole 2013, p. 128). These contrast with short diurnal stops during migration, made at intervals between nocturnal flights and lasting up to 12 h, usually in non-typical foraging habitats. We recorded a few one day long movements that combined nomadic movements, short diurnal stops and stationary foraging. We treated such behavior as a stopover if the distance between the two furthermost points of such movements was ≤50 km.

The home range at stopovers was estimated as the area within the minimum convex polygon (according to Van der Winden et al. 2012). To obtain the total area of the home range, the core areas of separate staging areas were summed (i.e. neglecting the spaces between the non-staging areas). We estimated the habitat type with Google Earth.

Results

Night Herons started their migration in September and October (Fig. 1). The behaviors of only two birds (SL03, SL01) could be determined, because the battery of SL02 was not charging properly during the first phase of migration and the transmitter stopped working. Bird SL01 reached Sicily in three days (a distance of approximately 1,565 km), where it made a stopover of at least 11 days. Due to battery charging failure, the transmitter stopped working after 19 October. Bird SL03 reached central Italy (a distance of approximately 936 km) in two days and then flew to Corsica (368 km, 1.5 days), from where it proceeded to northern Algeria (465 km, 4.5 days). At each of these locations, bird SL03 made a long stopover (Fig. 1, see below). The transmitter stopped working as the bird was flying over the Sahara Desert (the battery did charge properly).

Fig. 1
figure 1

Fixes of three adult Night Herons during post-breeding migrations from Poland to Africa in 2012. Note that the fixes overlap. 4–23 Oct: staging periods at particular stopovers, 128 ha:  home range on stopovers. The transmitters of all birds repeatedly failed during migration after these last fixes were obtained

Full size image

The average speed recorded between two points during flight during migration for an interval of six hours was 54.5 km/h (range 44–65 km/h, N = 7, two birds). The fastest migration rate achieved by SL01 was 1,118 km in 36 h, with only one short diurnal stop (about 6 h) after crossing the Adriatic Sea. SL03 flew the 1,058 km from Corsica to Algeria in 40 h.

The birds migrated mostly at night. Transmitter data indicate that the birds started migrating at least one hour before sunset and landed before sunrise. The migrating birds rested during short diurnal stops made between night flights. At such times, transmitter data indicated movements of up to 60 m in any direction. Only one of seven short diurnal stops was by a lake; the other resting places were in bushes or trees and even in a forest (170 m from the edge).

Four stopovers lasting 17 days on average (range 9–20 days, SL03, Fig. 1) were in river valleys, by small ponds, in gravel pits, and in wadis. The maximum range of movements during stopovers, which were limited to linear displacements along river valleys or wadis, varied from 13 to 58 km. The home range at particular stopovers varied between 150 ha and 13,796 ha.

Discussion

Our results show a hitherto unknown post-breeding migration strategy of adult Night Herons: traveling slowly with several long stopovers en route. The data for one bird, tracked until it reached the Sahara, indicate that the autumn migration lasted at least two months. These birds migrated according to a strategy that is significantly different to that of Purple Herons, which migrate fast, without stopovers en route (Jourdain et al. 2008; Van der Winden et al. 2010). Data obtained from the satellite tracking of American Bitterns, though not as detailed, indicate that these birds migrate more slowly than Purple Herons but faster than Night Herons (Van der Winden et al. 2010; Huschle et al. 2013). More data are needed to be sure that the patterns of autumn migration we found in this study are general and hold for birds breeding across Europe.

The differences in migration strategy between Purple Herons and Night Herons could be due to differences in wing loading, which is significantly lower in the former than in the latter (Alerstam et al. 2007; Fig. 2 in Van der Winden et al. 2010). The costs of flight are therefore higher in Night Herons and this may compel them to make long stopovers en route (Newton 2008). Our results are consistent with the data of Van der Winden et al. (2010) on Fig. 2, which indicate that birds with a wing loading higher than ca 5 kg/m2 migrate by soaring or by flapping flight with at least one long refueling stop en route (e.g., Beekman et al. 2002, Chevallier et al. 2011).

In contrast to Purple Herons, Night Herons made long stopovers during migration. Van der Winden et al. (2010) suggested that Purple Herons were forced into a rapid travel strategy without stopovers, as the availability and predictability of foraging habitat were insufficient along their migration route. Owing to the lack of stopovers, Purple Herons also avoided competition with resident herons. Unlike Purple Herons, Night Herons can forage both during the day and at night (Cramp and Simmons 1977; Kushlan and Hancock 2005), which reduces competition with other primarily daytime-feeding species of herons. During stopovers, Night Herons used unpredictable foraging habitats—mainly wadis in submontane environments where water is present after autumn rains.

Night Herons, like Purple Herons and American Bitterns (Van der Winden et al. 2010; Huschle et al. 2013), migrate mostly at night, and spend short diurnal stops resting. The types of habitat in which Night Herons made their short diurnal stops indicate that these birds did not feed there.

The measured velocity of Night Herons (55 km/h) during migration was slightly higher than that of Purple Herons (46 km/h; Van der Winden et al. 2010), but similar to that of American Bitterns (52 km/h; Huschle et al. 2013).

The insufficient battery charging, a problem affecting two of the transmitters, could have been due to the fact that during the day the birds perched mainly among bushes or reeds, which limited the access of sunlight to the solar panel charging the battery. Insufficient battery charging could also have been caused by the transmitter being covered by the long feathers on the herons’ upper back. The transmitter of SL03 stopped working as the bird was flying over the Sahara Desert (the battery did charge properly, there were no multiple locations from the same place); therefore, we do not know whether the bird died. The other bird species tracked by Ecotone GPS/GSM transmitters sent data as they were crossing the Sahara desert and the Sahel zone (information obtained from Ecotone). Therefore, the reason that the transmitter of SL03 stopped working was not identified.

The extra weight of the transmitters carried by the Night Herons was relatively large (3.4–3.9 % of bird body mass). Transmitters representing 3 % of the bird’s body mass (the acceptable range is 3–5 %, Fair et al. 2010; Bridge et al. 2011) resulted in a roughly 5 % increase in energy expenditure for flight (Vandenabeele et al. 2011). Therefore, in Night Herons, which migrate by flapping flight—energetically one of the most expensive modes of locomotion in birds—the additional weight can affect migratory behavior by, for example, reducing migratory speed and/or prolonging stopovers. Owing to this potential transmitter effect, as well as the small sample size, these results should be treated as tentative until more data have been gathered.