- Top of page
The range of the great bustard stretches 10 000 km across Eurasia, one of the largest ranges of any threatened species. While movement patterns of the western subspecies of great bustard are relatively well-understood, this is the first research to monitor the movements of the more endangered Asian subspecies of great bustard through telemetry and to link a breeding population of Asian great bustards to their wintering grounds. Using Argos/GPS platform transmitter terminals, we identified the annual movement patterns of three female great bustards captured at their breeding sites in northern Mongolia. The 4000 km round-trip migration we have recorded terminated at wintering grounds in Shaanxi, China. This route is twice as long as has previously been reported for great bustards, which are among the heaviest flying birds. The journey was accomplished in approximately two months each way, at ground velocities of 48–98 km h−1, and incorporated multiple and variable stopover sites. On their wintering grounds these birds moved itinerantly across relatively large home ranges. Our findings confirm that migratory behavior in this species varies longitudinally. This variation may be attributable to longitudinal gradients in seasonality and severity of winter across Eurasia. The distance and duration of the migratory route taken by great bustards breeding in Mongolia, the crossing of an international border, the incorporation of many stopovers, and the use of a large wintering territory present challenges to the conservation of the Asian subspecies of great bustard in this rapidly changing part of the world.
The range of the great bustard Otis tarda, a large lekking bird, stretches from Manchuria to the Iberian Peninsula across the grasslands and steppes of Eurasia (Isakov 1974, Collar 1996). The two subspecies of great bustard, European (O. t. tarda) and Asian (O. t. dybowskii) are geographically isolated and differ in coloration of neck, wing coverts and rectrices, patterning on the back, and extent of specialized display plumes on the chin and neck (Ivanov et al. 1951, Johnsgard 1991). While populations of the nominal subspecies are listed as Vulnerable (VU) worldwide by IUCN (BirdLife International 2012), only 1200–2200 Asian great bustards remain and this subspecies is Red-listed across its range of Russian South Siberia, Mongolia and China (Tseveenmyadag 2003, Goroshko 2008). Breeding grounds in Mongolia now represent the stronghold for this subspecies (Alonso and Palacín 2010). Clarification of threats to the subspecies and its natural history parameters, particularly in Mongolia, is identified as a priority for its conservation (Boldbaatar 1997, Chan and Goroshko 1998).
Detailed movement studies have not previously been undertaken on Asian great bustards, but data from radio and satellite tracking of the European subspecies indicate that great bustards display a wide range of migratory behaviors, including both partial and differential migration (Terrill and Able 1988). In general, migratory distance of great bustards increases longitudinally across Europe from west to east, in correspondence with severity of winter weather conditions and the degree of seasonality. A variety of short seasonal movements have been described in Spanish populations. These include post-breeding migrations by some males of up to 196 km, the distance of which may be dependent on climatic and habitat variables (Alonso et al. 2001, 2009). Some females make autumn/winter movements of up to 110 km (Alonso et al. 2000, Palacín et al. 2009); these migrations are culturally transmitted and condition-dependent (Palacín et al. 2011).
Great bustards in central Europe tend to be sedentary, though short migrations by some populations, or some individuals in a population, have been observed (Bankovics and Széll 2006). Irregular irruptive movements of up to 650 km have been recorded for these populations in response to severe winter weather (Faragó 1990, Block 1996, Streich et al. 2006).
Populations of European great bustards on the Lower Volga River in Russia – the most easterly populations for which tracking data are available – are mostly migratory. Females monitored via satellite telemetry traveled 1100 km over the course of approximately one week to winter in southeast Ukraine (Oparina et al. 2001, Watzke et al. 2001, Khrustov 2009).
Our group investigated the migratory behavior of Asian great bustards in north central Mongolia, approximately 4000 km east and 200 km south of the Volga populations. Given the severely continental climate of northern Mongolia, we predicted that distance migrated would be farther than observed in European populations, in correspondence with the longitudinal trends noted above. Here we present the first data on complete annual movements of this subspecies: the long-distance round-trip migrations of three female Asian great bustards.
- Top of page
Research was carried out on breeding populations of great bustards in east Khövsgöl Aimag, Mongolia (approximately 50°N, 101°E). Birds were found in valleys dominated by low-intensity agriculture (primarily summer wheat) and livestock herding by nomadic pastoralists. In this region of forest-steppe, winters are severe, with average January temperatures around –30°C (Inst. Geografii – Sibirskoe Otdelenie 1989). Nights and cold fronts in winter bring low temperatures of –40 to –50°C.
All work was carried out under permits issued by the Mongolian Ministry of Nature, Environment, and Tourism (no. 4/730, 4/1813, 6/1650) and using methods approved by the Arizona State Univ. Institutional Animal Care and Use Committee (no. 07-924R). We captured one female in 2007 and two additional females in 2008 by spotlighting (Giesen et al. 1982, Seddon et al. 1999, Geyser 2000).
Each bird was fitted with a solar-powered 70 g Argos/GPS platform transmitter terminal (‘PTT’) using a custom-fit backpack harness (modified from Osborne and Osborne 1998, Alonso et al. 2001). Stretchable silicone rope was threaded through bunched teflon ribbon to create a durable harness capable of adjusting to weight changes. The straps of the backpack cross at the breast, where they were stitched to ensure that the harness did not shift location. Points at which the harness was threaded through the transmitter were stabilized with instant glue. Birds were released at the site of capture within 15–30 min. The PTT and harness represent approximately 2% of the females’ body weight, which falls within the range of loads recommended by Kenward (2001).
Each PTT transmitted GPS data (± 18 m accuracy) by radio signal to the Argos system (maintained by CLS, Toulouse, France) deployed on satellites. Duty cycles were tailored to maximize the number of GPS locations transmitted, with the length of day and strength of solar charge to the battery as limiting factors. Locations were recorded every two hours from 6:00 to 20:00 in spring and fall, from 4:00 to 22:00 in summer, and from 7:00 to 19:00 in winter. PTTs also reported speed of movement (± 1 km h−1 accuracy at speeds > 40 km h−1). Upon receipt of a series of radio transmissions, the Argos system also estimates the location of the PTT using Doppler shift calculations, which are transferred in a separate data frame.
A comparison of the movements of individual tagged birds to each other, and to records of bustard migration at geographically similar locations, did not yield observations of consistent delays by any individual. We also did not observe correspondence between failure to breed and timing of spring arrival, which would indicate strong transmitter effects (Barron et al. 2010).
Routes were plotted and distances between points calculated using ArcGIS 10. Minimum convex polygons and kernel density estimations were created using Geospatial modelling environment (Beyer 2011). Departure and arrival dates were determined primarily through scrutiny of GPS-quality transmissions. We used Doppler-shift calculated locations when those allowed us to narrow the range of dates of a bird’s arrival or departure in the absence of GPS-quality data.
- Top of page
All three female birds were roughly the same weight at capture (Table 1). Birds no. 01 and no. 03 were captured in the same valley; bird no. 02 was captured in a valley 50 km distant. Data presented are of migratory movements from date of capture (Table 1) through 1 June 2009.
Table 1. Migratory activity recorded for three female great bustards Otis tarda dybowskii captured in north central Mongolia and harnessed with Argos/GPS satellite transmitters.
|Bird ID||Capture date/weight||Season||Distance flown (km)||Start date||End date||Duration (days)||Mean km flown d−1||Number of GPS points||Mean ground speed ± SD (km h−1)||n*|
|01||14 Jun 2007 3400 g||fall 2007||1954||13–16 Oct||4–12 Dec||49–60||33–40||56||59 ± 2||3|
|01||-||fall 2008||1852||17–19 Oct||4–6 Nov||16–20||93–116||19||59 ± 6||5|
|02||27 Jun 2008 3500 g||fall 2008||1836||12 Oct||31 Oct||19||97||56||87 ± 10||3|
|03||10 Jun 2008 3600 g||fall 2008||2044||17 Oct||18–20 Dec||62–64||32–33||205||76 ± 12||5|
|01|| ||spring 2008||1966||24–26 Mar||28–31 May||63–68||29–31||39||62||1|
|01|| ||spring 2009||1932||12–14 Mar||1 Jun||79–81||24||80||NA||–|
|02|| ||spring 2009||1860||5 Apr||9–13 May||34–38||49–55||52||80 ± 6||2|
|03|| ||spring 2009||2100||5 Apr||9 Jun||65||32||323||60 ± 9||8|
Due to radio interference typical in eastern Siberia and China and poor battery charge especially during winter months, not all logged GPS data were ultimately received by the Argos system. The greatest distance between any two successively received GPS points was approximately 1000 km, from Khövsgöl Aimag in Mongolia to the southern border of Mongolia, over a period of six days (bird no. 01, fall 2008).
Each female migrated from Khövsgöl Province in northern Mongolia in a southeastern direction (approximately 140°) to wintering spots near Xi’an city in Shaanxi Province, China (Fig. 1–3). Data indicate that the marked birds traveled independently of one another. Fall routes deviated from spring routes, but a consistent loop directionality was not detected. Average distance migrated was approximately 2000 km one-way, and was similar among birds and seasons (Table 1).
Figure 1. Map (UTM 47N projection) of the autumn 2007 (o), spring 2008 (+), autumn 2008 (▴) and spring 2009 (x) migratory routes of female great bustard Otis tarda dybowskii no. 01. Each vertex represents a GPS-quality stop location reported by the transmitter. GPS locations during flight were used to construct the path, but are not shown as vertices.
Download figure to PowerPoint
Figure 3. Map (UTM 47N projection) of the autumn 2008 (o) and spring 2009 (+) migratory routes of female great bustard Otis tarda dybowskii no. 02. Each vertex represents a GPS-quality stop location reported by the transmitter. GPS locations during flight were used to construct the path, but are not shown as vertices.
Download figure to PowerPoint
The migratory route of bird 01 in 2008 was similar to her route in 2007 (Fig. 1). In spring 2009 bustard 01 also performed a 50 km roundtrip detour in the direction of another known lek, where she spent 4–8 d before returning along the same path to resume her route northward.
The spring and fall migratory routes of bird 03 exhibited the most variation of the three birds tracked, with a maximum divergence of approximately 170 km (Fig. 2). This bird also took a detour of 60 km in northern Mongolia before returning to her primary lek in spring 2009.
Figure 2. Map (UTM 47N projection) of the autumn 2008 (o) and spring 2009 (+) migratory routes of female great bustard Otis tarda dybowskii no. 03. Each vertex represents a GPS-quality stop location reported by the transmitter. GPS locations during flight were used to construct the path, but are not shown as vertices.
Download figure to PowerPoint
Though distances traveled were similar among birds and seasons, we found five-fold variation among the three birds in the duration of migration. Average duration of a one-way trip was approximately two months (Table 1). In three of four cases, spring migration lasted longer than that bird’s previous autumn migration. In the case of bird 01, spring 2009 migration was almost two months longer than the preceding fall migration (Table 1).
When in flight bird 02 regularly achieved speeds 30% greater than the other two birds, with a maximum ground speed of 98 km h−1. The duration of her migrations was approximately half that of the other two birds (Table 1). Minimum ground speed recorded was 48 km h−1 for bird 03 in spring 2009.
The bustards we monitored used multiple and varied stopover sites, and it is likely that additional locations in which the birds stopped were not detected because of failed transmissions. We did not find fidelity to specific stopover localities. Most routes included a stop on the outskirts of Bayanur, an agricultural oasis in Nei Mongol, China, but stopovers there were spread across 130 km. Individuals occupied some stopovers for only 1–2 d and rarely took longer stops. Stops of approximately 10 d were recorded in Khishig-Öndör sum of Bulgan Aimag and Tarialan sum of Khövsgöl Aimag, Mongolia, and Ordos Prefecture and the Bayanur oasis in Nei Mongol, China. One stop of 45 d was recorded for bird 03 in the Bayanur oasis.
These bustards overwintered in agricultural fields near the confluence of the Wei and Yellow Rivers in Shaanxi Province of China. Individuals tended to progress eastward through a series of non-repeated sites over the course of winter months, resulting in a large overall winter range (Fig. 4). The smallest range was recorded for bird 01 in winter 2008; this dataset also included the fewest observations and a gap in data reception of 107 d (Table 2). Bird 03 gradually moved eastward during the winter months, such that her first major northward movement was 90 km east of her last major southward movement (Table 2). Bird 02 also spent much of the winter moving gradually 50 km to the northeast.
Figure 4. Map (UTM 47N projection) of the minimum convex polygons encompassing GPS locations at which each great bustard was recorded over the winter. Watercourses and urbanized areas are shaded.
Download figure to PowerPoint
Table 2. Wintering areas in China for three great bustards captured in northern Mongolia. MCP stands for minimum convex polygon.
|Bird ID||Winter||Number of GPS points||Significant gaps in data (days)||Area of MCP (km2)||Area of 80% kernel (km2)||Maximum distance between points (km)||Ground speed||n*|
|01||2007–2008||49||12, 20, 18||401.7||63.7||51.7||61||1|
|02||2008–2009||217||11, 16, 10||1450.7||355.7||93.8||64||1|
|03||2008–2009||200||17, 14, 12||1967.6||723.2||95.4||54||1|
Though birds 01 and 03 summer at the same lek in northern Mongolia, their wintering ranges did not overlap (Fig. 4). The ranges of birds 02 and 03 overlapped (Fig. 4), but the core areas used by each bird differed (Fig. 5). In 2008, bird 01 wintered 40 km north of the range she used in the previous winter.
Figure 5. Map (UTM 47N projection) of 80% kernel density estimates of wintering areas used by each tagged great bustard. Watercourses and urbanized areas are shaded.
Download figure to PowerPoint
- Top of page
We are grateful for training and advice on bustard capture and handling shared by many experienced researchers, particularly J. C. Alonso, M. Gilbert, S. Hallager, M. Lawrence and others at the National Avian Research Center (UAE), M. Mace, C. Martín and T. Osborne. We thank T. Katzner for facilitating receipt of Argos transmissions. The Taimen Conservation Fund and many individuals in rural Mongolia extended crucial support especially during the early stages of this research. D. Dorjhürel, D. Erdenetsetseg, G. Natsag and Ü. Tövshin provided field assistance. Comments by B. Hogan, M. Fujitani and M. Toomey improved this manuscript.
This research was funded in part by a US National Security Education Program Boren Fellowship, Wildlife Conservation Society Research Fellowship, and US National Science Foundation Graduate Research Fellowship awarded to AK. Grants from the American Museum of Natural History’s Chapman Fund, Arizona State University’s Graduate and Professional Student Association, Cleveland Metroparks Zoo, and the Rufford Small Grants Foundation supported our 2007–2009 field seasons. Microwave Telemetry, Celestron Optics and the National Wild Turkey Federation generously donated equipment used in this research.