See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/337841397

Migration Patterns and Characteristics of Eurasian Wigeons (Mareca penelope) Wintering in Southwestern Based on Satellite Tracking

Article in ZOOLOGICAL SCIENCE · December 2019 DOI: 10.2108/zs180207

CITATIONS READS 4 373

9 authors, including:

Tomoko Doko Wenbo Chen Keio University Keio University

24 PUBLICATIONS 136 CITATIONS 18 PUBLICATIONS 117 CITATIONS

SEE PROFILE SEE PROFILE

Noriyuki M Yamaguchi M. Fujita University of Nagasaki Okinawa Prefectural Museum and Art Museum

44 PUBLICATIONS 462 CITATIONS 32 PUBLICATIONS 457 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Evaluating the future impacts of climate change on the movement of migratory birds (Raptors in East Asia) View project

Landscape ecological studies on habitat of Oriental honey-buzzard based on satellite-tracking data View project

All content following this page was uploaded by Tomoko Doko on 13 December 2019.

The user has requested enhancement of the downloaded file. Migration Patterns and Characteristics of Eurasian Wigeons (Mareca penelope) Wintering in Southwestern Japan Based on Satellite Tracking

Authors: Tomoko Doko, Wenbo Chen, Naoya Hijikata, Noriyuki Yamaguchi, Emiko Hiraoka, et. al. Source: Zoological Science, 36(6) : 490-503

Published By: Zoological Society of Japan URL: https://doi.org/10.2108/zs180207

BioOne Complete (complete.BioOne.org) is a full-text database of 200 subscribed and open-access titles in the biological, ecological, and environmental sciences published by nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Complete website, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/terms-o-use.

Usage of BioOne Complete content is strictly limited to personal, educational, and non - commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research.

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use ZOOLOGICAL490 SCIENCE 36: 490–503 (2019) T. Doko et al. © 2019 Zoological Society of Japan

Migration Patterns and Characteristics of Eurasian Wigeons (Mareca penelope) Wintering in Southwestern Japan Based on Satellite Tracking

Tomoko Doko1,2*†, Wenbo Chen1,2*†, Naoya Hijikata3, Noriyuki Yamaguchi4, Emiko Hiraoka5, Masaki Fujita6, Kiyoshi Uchida5, Tetsuo Shimada7, and Hiroyoshi Higuchi8

1Nature & Science Consulting Co., Ltd., Room 302, 3F, Chuo-dairoku-kannai Bld., 1-2-1 Furocho, Naka-ku, Yokohama 231-0032, Japan 2Graduate School of Media and Governance, Keio University, 5322 Endoh, Fujisawa 252-0082, Japan 3Hijikata Office, Prier Sanbankan 102, Sekido 5-11-12, Tama 206-0011, Japan 4Graduate School of Fisheries and Environmental Sciences, Nagasaki University, 1-14 Bunkyo-machi, Nagasaki 852-8521, Japan 5Satoyama Natural History Research Group, 1-11-11 Midori, Abiko, Chiba 270-1153, Japan 6Department of Anthropology, National Museum of Nature and Science, 5-21-5, Shirokanedai, Minato-ku, Tokyo 108-0071, Japan 7The Miyagi Prefectural Izunuma-Uchinuma Environmental Foundation, 17-2 Shikimi, Wakayanagi, Kurihara, Miyagi 989-5504, Japan 8Research and Education Center for Natural Sciences, Keio University, 4-1-1 Hiyoshi, Kohoku-ku, Yokohama 223-8521, Japan

Understanding migration ecology of Eurasian wigeons (Mareca penelope) is crucial for effective population management, mitigating conflicts with human, and habitat conservation. The objectives of the present study were 1) to determine their migration patterns of Eurasian wigeons in the East Asian flyway, and 2) to identify the key breeding and stopover sites. From 2007 to 2016, a total of the 64 wigeons, which wintered in Japan, were equipped with satellite transmitters. Most Eurasian wigeons migrated to breeding sites in Russia either (a) via a continental route through China, (b) via the Kamchatka Peninsula, or (c) via Sakhalin Island. In spring, many of the Eurasian wigeons (60.98%) migrated via the continental route. In autumn, most Eurasian wigeons (57.14%) migrated through Kamchatka. These differences may be attributable to the influence of Okhotsk Sea air mass on migration decisions due to loop migration. Similarly to the migration of Mallards, Eurasian wigeons employed a “long-stay and short-travel” migration strategy. Eurasian wigeons mainly nested between latitude between 43° to 75°N. From the present findings and the published literature, Eur- asian wigeons that winter in Japan are considered to migrate to Russia, China, and the United States during the breeding season, although the main breeding area is in northeastern Russia. A total of 296 important sites to Eurasian wigeons were mapped, and 118 location names with geographic coordinates, and the top five most frequently used sites were identified in each season.

Key words: breeding sites, Eurasian wigeon, Mareca penelope, migration pattern, stopover sites

50° and 75° N across northern Europe and Asia, from Iceland INTRODUCTION and northern Britain across Scandinavia and northern Eurasian wigeons (Mareca penelope) breed between Russia to the Pacific coast (Kear, 2005). The species is strongly migratory, leaving its breeding grounds in late sum- mer to winter across almost the entire temperate zone of * Corresponding authors. E-mail: doko@nature-science-consulting. co.jp (TD); Europe and Asia, with concentrations in coastal areas chen@nature-science-consulting. including Japan and China (Kear, 2005). The wintering sites co.jp (WC) in Far East are Japan, Korea, China, and South Asia, and † These authors contributed equally to this work. Southeast Asia (Nechaev and Gamova, 2009). Wintering doi:10.2108/zs180207 populations likely mix in the breeding grounds and the major

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use Migration patterns of Eurasian wigeons 491

moulting areas (Kear, 2005). The population trend for tion of Eurasian wigeons used in that analysis was not Eurasian wigeons that winter in Eastern Asia is declining Japan, and (2) Eurasian wigeons that winter in Japan may (Miyabayashi and Mundkur, 1999; BirdLife International, have different migration routes compared to other wintering 2017). Yet, ducks have been heavily hunted in Russia, a areas, as suggested in Boere et al. (2007). major breeding ground (Panov, 1973), and there are increas- The objectives of the present study were 1) to determine ing potential conflicts with humans, as Eurasian wigeons the migration patterns and characteristics of Eurasian may facilitate the transmission of highly pathogenic avian wigeons along the East Asian Flyway, and 2) to identify key influenza virus during their migration (World Organisation for sites for breeding and stopovers. These characteristics and Animal Health, 2014), feed on agricultural crops (Lane et al., trends are discussed with reference to the existing literature 1998) or other economically important food products, such and in comparison to other duck species, including Northern as seaweed (Kodama et al., 2014), and increase the poten- pintails (Anas acuta), and mallards (Anas platyrhynchos) tial for aircraft strikes (Ministry of Land Infrastructure and that winter in Japan. Tourism, 2012). Thus, migration studies are needed for MATERIALS AND METHODS effective population management, mitigation of conflict with humans, and habitat conservation for this duck species. Satellite tracking However, the migration pathways of Eurasian wigeons in Eurasian wigeons were captured using net traps or hands each East Asia are almost completely unknown. There are only a few detailed studies on the migration of Eurasian wigeons in Europe (Owen and Mitchell, 1988; Donker, 1959; Boyd, 1956). Through intensive bird banding conducted in Japan from 1961 to 1995, 194 recoveries of Eurasian wigeons were obtained from eastern Russia, con- centrated especially from Sakhalin and Kamchatka (Yamashina Institute for Orni- thology, 2002). On the other hand, 2829 recoveries of wigeons collected through Bird Ringing Centre of Russia demon- strated that Eurasian wigeons wintering in Japan exhibit a different migration pattern, which likely involves migration to Far East Russia during breeding seasons, com- pared to all other wintering populations (Boere et al., 2007). These ringing studies can provide data on released and recov- ered point localities, but cannot explain their migration routes, unlike satellite telemetry. Within Asia, two major flyways—the Central-South Asian Flyway and the East Asian Flyway—are recognized for Anati- dae (Miyabayashi and Mundkur, 1999). Based on this information, Eurasian wigeons wintering in Japan belong to the East Asian Flyway. Palm et al. (2015) attempted to visualize the East Asian- Australasian Flyway (which is the same as the East Asian Flyway), based on satellite and GPS telemetry data and dynamic Brownian bridge movement models. The East Asian-Australasian Flyway in Palm et al. (2015) described waterfowl space use probabilistically. Eurasian wigeons were included as one target species, but the probabilistic map shows the overall habi- tats of multi-species of waterfowl. More- over, the information for Eurasian wigeons wintering in Japan was likely missing in Palm et al. (2015), probably due to the fol- Fig. 1. Locations where Eurasian wigeons were captured and marked with satellite lowing two reasons: (1) the capture loca- transmitters in southwestern Japan, 2007–2016.

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use 492 T. Doko et al.

winter (December to March) from 2007 to 2016 at six locations in version 3.1.3 (R Development Core Team, 2006) with R packages southwestern Japan (Fig. 1). The six locations were at (1) Hitotsuse “zoo” (Zeileis and Grothendieck, 2005; Zeileis et al., 2015) and River at Hitotsusegawa prefectural sports center, Shintomicho, “changepoint” (Killick et al., 2014; Killick and Eckley, 2014) was Miyazaki Prefecture (32°3′9.028″ N, 131°27′52.009″ E), (2) Kota used to determine the threshold dates of the migration status. This pond, Miyazaki City, Miyazaki Prefecture (32°3′25.174″ N, study adopted the combination of the “cpt.meanvar” method for 131°24′29.609″ E), (3) Muromi River, Fukuoka City, Fukuoka Pre- latitude, based on Doko et al. (2016). These threshold dates were fecture (33°33′7.164″ N, 133°55′36.991″ E), (4) Kayoicho Park, plotted against backgrounds showing the latitude, longitude, and Kasuya City, Fukuoka Prefecture (33°36′38.879″ N, 130°29′14.855″ daily distance with the above-mentioned simple location indicators. E), (5) Miyama Park, Tamano City, Okayama Prefecture If the threshold dates were incorrectly identified, the dates were (34°30′53.913″ N, 130°19′26.688″ E) and (6) Koyaike Park, Itami manually edited. City, Hyogo Prefecture (34°47′15.475″ N, 135°23′36.351″ E). The Hitotuse River and Kota pond were close each other and consid- Calculation of migration indices and statistical tests ered to belong to the watershed of Hitotuse River. The Kayoicho The migration indices, including the migration status’ start and Park and Muromi River were close to Hakata Bay. end dates, the duration of the migration status, the duration and Satellite transmitters (platform transmitter terminals; PTTs) number of stopover sites, the total travel distance between the win- were attached to the back of the birds with a harness system, via tering and breeding sites, and geographical extent of latitude and the methods described by Yamaguchi et al. (2008). The total weight longitude were individually calculated for each marked bird from the including solar-powered PTTs (5–18 g) and harness systems (4–5 base dataset. The terminal sites in Russia were assumed to be g) corresponded to 1.2–3.2% of the mean Eurasian wigeon body breeding sites in this study. Because distribution for threshold dates mass (700 g) from Johnsgard (1978). The PTTs were manufactured either by North Star Sci- ence and Technology LLC (King George, VA, USA) or Microwave Telemetry Inc. (Columbia, MD, USA). Five different transmission cycles of PTTs were programmed. In total, 64 Eurasian wigeons with PTTs (23 females and 41 males) were marked. Our sample included 33 adults, seven juveniles, and 24 birds of unknown age. The PTT locations were estimated using the Argos System (Argos, 1996; CLS (Collecte Localisation Satellites), 2007). Data collected using the Argos System included the date, time (Greenwich Meridian Time; GMT), and longi- tude and latitude (WGS84 datum) with location class (LC) accuracy. The Douglas Argos-Filter Algorithm (Douglas et al., 2012) was used to remove implausible locations.

Identification of dates to differentiate migra- tion status by the MATCHED method The migration status of Eurasian wigeons was classified according to the MATCHED (Migratory Analytical Time Change Easy Detection) method (Chen et al., 2016; Doko et al., 2016). In general, bird migration can be classified according to migration status, which include the wintering period, spring migration, breeding period, and autumn migration. To determine the migration status, their periods should be individually determined. Using MATCHED method, it is possible to determine threshold dates of migration status individually based on satellite-tracked data. The only higher quality LC Class 3 (< 0.25 km), Class 2 (0.25– 0.5 km), Class 1 (0.5–1.5 km), and Class 0 (> 1.5 km) locations were retained. If there were several location data in a single date, a single data with better location classes was selected. If there were multiple locations of same high quality within a duty cycle, the earliest record was selected. Through this procedure, the dataset was compiled into 6431 records in total (hereinafter, the base dataset). After projection to Universal Transverse Mercator (UTM) coordinates, geographic coor- dinates in meters were calculated to compute Fig. 2. Regional zones of eastern Russia and eastern Asia related to migration status the distance between two points. R software of Eurasian wigeons via satellite tracking, 2007–2016.

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use Migration patterns of Eurasian wigeons 493

was skewed, the median and 25th and 75th percentiles were calcu- (2011) (Fig. 2). Based on DEM (Digital Elevation Model), water- lated across all birds, instead of mean values. As for duration (days) sheds were calculated by ArcMap ver. 10® (ESRI Inc.), and bound- and number of stopover sites, and the travel distance, mean value aries of watersheds were used as the boundaries delineating the and standard deviation (SD) were calculated. Basically, duration of regions. If the watersheds belonged to the same mountain ridge, each migration status was calculated based on the birds that com- these watersheds were integrated into one watershed. If the target pleted a migration cycle and whose start dates and end dates could region is an island, the boundary of that island is considered to be identified. represent to delineate the boundaries of the region. These regions The travel distance was calculated for individual birds using the were (43°29′32.008″ N, 142°58′21.07″ E), Not Hokkaido bird’s accumulated distance from the record before the migration (34°6′41.608″ N, 132°34′47.137″ E), Hamhung (39°5′18.705″ N, start date until the end date of migration. One record before the 128°44′34.259″ E), Tumen River (42°40′21.149″ N, 131°28′18.055″ migration start date (in general, one day before, if there were no E), Amur River (48°11′29.92″ N, 134°49′51.466″ E), Chukotka missing dates) was included because first-day movement of (63°42′25.434″ N, 169°24′19.59″ E), Kamchatka (57°7′23.332″ N, Eurasian wigeons tended to be long, e.g. even over 1000 km some- 162°5′12.587″ E), Kuril Island (47°23′46.812″ N, 151°47′3.579″ E), times, and we considered it unreasonable to ignore it. Magadan (59°58′52.117″ N, 153°21′19.966″ E), Sakhalin The differences in (1) the stopover duration, (2) the number of (51°16′20.32″ N, 143°8′4.87″ E), and Kolyma (66°41′50.666″ N, stopovers, and (3) the flight distance were tested using the 158°14′54.066″ E) (the latitude and longitude represented the mean Wilcoxon rank-sum test (Wilcoxon, 1945) from three perspectives: values). Here, the “Not Hokkaido” region includes Kyushu, , (i) between spring migration and autumn migration, (ii) between and islands of Japan. Because the movement pattern male and female in the spring migration, and (iii) between male and between Hokkaido and areas except Hokkaido in Japan showed a female in the autumn migration. The exactRankTests (Torsten and different tendency for the choice of migration routes, zoning of Kurt, 2015) package, R software, version 3.1.3 was used. Japan was separated by “Hokkaido” region and “not Hokkaido” Geographical extent of latitude and longitude were calculated region in this study. for distinctive migration statuses separately. Minimum, maximum, and average values were calculated. RESULTS Movement patterns Identification of spring migration routes and autumn migration Of the 64 marked birds with PTTs, transmitters of four routes Based on the base dataset, the migration routes for the spring birds (4.6%) stopped functioning during the wintering sea- and autumn migrations were drawn separately for each marked bird son shortly after deployment. The remaining 60 birds using ArcMap ver. 10® (ESRI Inc.) in the Mercator projection. (93.75%) were grouped in two categories: (1) a group which departed from Japan and made landfall elsewhere (40 birds, Identification of key sites 62.50%), and (2) a group that did not depart Japan during For each bird, the mean centers for the wintering areas, stop- the tracking period (20 birds, 31.25%). over areas, and breeding areas were calculated using Microsoft Of 40 Eurasian wigeons that departed Japan, 29 were Excel 2010 to visualize areas used by Eurasian wigeons. By this successfully tracked until the start of the breeding season. A procedure, when a bird remained at a site for >1 transmission total of 28 birds (43.75%) arrived in Russia during the breed- cycle, multiple locations were reduced to a single point, repre- ing season, and one bird (1.56%) arrived in China during the sented by the mean latitude and longitude of locations. The stop- over areas were categorized according to their duration of use and breeding season. Breeding sites of 11 birds (17.19%) that location. A total of 296 records was finally compiled to represent the departed Japan could not be determined because signals mean centers. This record means that 296 total areas across all ceased shortly after they made landfall elsewhere, and it birds for all seasons were summarized based on a larger number of was not clear whether their final location was a stopover or locations. The mean centers were mapped in ArcMap ver. 10®. The top five important locations Table 1. Start date, end date, and duration of the wintering period, spring migration, breeding were summarized in order to dem- period, and autumn migration of Eurasian wigeons. onstrate the key sites related to Start date End date Duration (days) Eurasian wigeons. Key sites in this study were defined as the top five Q1 ~ Q3 Median n Q1 ~ Q3 Median n Q1 ~ Q3 Median n most frequently used important Wintering period 7-Sep ~ 27-Nov 10-Nov 6 4-Apr ~ 24-Apr 9-Apr 45 133 ~ 167 151 5 areas of Eurasian wigeons within Spring migration 3-Apr ~ 24-Apr 8-Apr 41 18-May ~ 29-Jun 8-Jun 33 39 ~ 83 56 26 each season of spring migration, breeding, and autumn migration, Breeding period 28-May ~ 30-Jun 9-Jun 28 5-Sep ~30-Sep 23-Sep 13 81 ~ 117 94 10 separately. Because our capture Autumn migration 7-Sep ~30-Sep 25-Sep 12 3-Nov. ~ 5-Dec 11-Nov 5 26 ~ 86 52 5 locations of Eurasian wigeons did not represent all wintering sites in Japan, we did not calculate the key sites for the wintering season. Table 2. Duration (days) and number of stopover sites of Eurasian wigeons. The mean number ± SD are shown. The numbers in parentheses indicate sample sizes. The numbers in [] indicate Analysis by regions median values. Eurasian wigeons passed the Duration at stopover sites (days) Number of stopover sites countries of Japan, Russia, China, North Korea and South Korea. Spring migration Autumn migration Spring migration Autumn migration Based on the migration pattern and female 13.4 ±12.2 (58) [10] 15.8 ±13.8 (13) [13] 3.1 ± 2.2 (19) [2] 3.3 ± 2.5 (4) [2] regional characteristics, a total of 11 male 16.1 ±13.0 (70) [12] 15.4 ±14.3 (18) [10] 2.7±1.6 (26) [3] 2.3 ±1.4 (8) [2] regions (hereinafter, region), were identified, similar to Hupp et al. total 14.9 ±12.7 (128) [10] 15.6 ±13.9 (31) [12] 2.9 ±1.9 (45) [3] 2.6 ±1.8 (12) [2]

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use 494 T. Doko et al.

breeding site. (ii) between males and females during spring migration (W = Among 20 birds that remained in Japan during the track- 76, P = 0.4389, Wilcoxon rank-sum test), or (iii) between ing periods, 11 birds (17.19%) moved to northern area (mainly males and females during the autumn migration (W = 4, P = Hokkaido) during spring migration.

Calculation of indices of migration Table 3. Distance moved for spring migration and autumn migra- The wintering duration was longest (151 days), followed tion of Eurasian wigeons. The mean number ± SD are shown. The st by the breeding duration (94 days) (Table 1). The 1 quartile numbers in parentheses indicate sample sizes. The numbers in [] (Q1), the 3rd quartile (Q3) and median dates with sample indicate median values. sizes (n) are shown. The duration of the wintering period needed to be estimated due to lack of complete sets of win- Spring migration (km) Autumn migration (km) tering period. If the start date of the winter season (x year) female 5816 ± 2020 (9) [4538] 8334± 0 (1) [8334] was unknown but the first observation year (x year)’s end male 5246 ±1944 (14) [5288] 4722 ±1267 (4) [5060] date of wintering and start date of the next wintering (x + 1 total 5469 ±1934 (23) [5206] 5445±1953 (5) [5682] year) were known, the first observation year x( year)’s end date of wintering and start date of the next wintering (x + 1 year) were treated as complete sets of wintering, assuming that the wigeons would depart from their win- tering area on the same date as the previ- ous year. All data from the base dataset, which enabled us to identify threshold dates, were used. The duration of both spring and autumn migration was approxi- mately 52–56 days. On average, during both spring and autumn migration marked Eurasian wigeons used approximately three stop- overs and remained at each for 14–16 days (Table 2). All data from base dataset was used in Table 2. There was no signifi- cant difference in the duration of stopovers (i) between spring and autumn migrations (W = 2288, P = 0.8247, Wilcoxon rank- sum test), (ii) between males and females during the spring migration (W = 1728, P = 0.1476, Wilcoxon rank-sum test), or (iii) between male and females during the autumn migration (W = 129, P = 0.6417, Wilcoxon rank-sum test). There was no significant difference in the number of stopovers (i) between the spring and autumn migrations (W = 210, P = 0.9645, Wilcoxon rank-sum test), (ii) between males and females in the spring migration (W = 250, P = 0.9505, Wilcoxon rank-sum test), or (iii) between males and females in the autumn migration (W = 20, P = 0.4990, Wilcoxon rank-sum test). Marked birds migrated an average of 5469 km and 5445 km during spring and autumn migration, respectively (Table 3). In order to remove the effects of wigeons that summered in Japan and one wigeon which remained in China for the breeding season, data of wigeons which migrated and arrived to breeding sites in Russia Fig. 3. Spring migration routes of Eurasian wigeons marked with satellite transmitters was used for this analysis. There was no at wintering sites, Japan, 2007–2016. The spring migration routes were plotted with 56 significant difference in migration distance lines in total, which is all data in which wigeons started spring migration, and departed (i) between the spring and autumn (W = from wintering sites in Japan, including 40 wigeons which migrated outside of Japan, 53, P = 0.8162, Wilcoxon rank-sum test), and 16 wigeons which moved inside of Japan in a northerly direction.

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use Migration patterns of Eurasian wigeons 495

0.4000, Wilcoxon rank-sum test). Of the 29 birds that reached breeding sites, 11 wigeons (three females and eight males; 37%) used multiple sites during the breeding season. The mean dura- tion of use at breeding sites was 22 and 23 days for females and males, respectively. The mean dis- tance Eurasian wigeons moved from one breeding site to another was 189 and 181 km for females and males, respectively.

Identification of spring and autumn migration routes The spring and autumn migra- tion routes are illustrated in Fig. 3 and Fig. 4, respectively. The num- ber of lines represents individual birds. In general, three major migration routes existed during spring and autumn: (a) a continen- tal route through China, (b) migra- tion along the Kuril Islands to the Kamchatka Peninsula, and (c) migration to Sakhalin Island then across the to Russia (Table 4). The routes for 15 Eurasian wigeons in spring and one Eurasian wigeon in autumn could not be clearly determined, either because their PTTs ceased transmitting during migration or they remained in Japan. In spring, most birds (60.98%) migrated via the continental route through China. However, in autumn, most birds (57.14%) returned to Japan via migration through the Kamchatka Peninsula. In spring migration, there were 19 different routes connecting between wintering and breeding sites (Fig. 5). A total of eight Fig. 4. Autumn migration routes of Eurasian wigeons marked with satellite transmitters at win- regions (plus one location) were tering sites, Japan, 2007–2016. The autumn migration routes were plotted with 15 lines in total, used as terminal sites: Tumen which is all data in which wigeons departed from breeding sites in Russia after breeding sea- sons. River, Amur River, Sakhalin, Magadan, Chukotka, Kamchatka, Kuril Islands, Kolyma, and Songhua River (location). The most com- Table 4. Counts and proportion of the major migration routes for spring migration and autumn mon region for breeding was migration of Eurasian wigeons. The counts are the number of individual wigeons which took that Chukotka (nine Eurasian wigeons, route. The numbers in parentheses indicate proportion in percentage. 31.03%), followed by Magadan, Spring migration Autumn migration Amur River, and Sakhalin (4 Eurasian wigeons, 13.79% for Route via continent 25 (60.98) [9 females and 16 males] 2 (14.29) [1 female and 1 male] each). A total of 29 birds migrated Route via Kamchatka 6 (14.63) [2 females and 4 males] 8 (57.14) [3 females and 5 males] and arrived at terminal sites either in Russia or China. After departure Route via Sakhalin 10 (24.39) [5 females and 5 males] 4 (28.57) [1 female and 3 males] of wintering sites on Honshu, 15 total 41 (100.00) 14 (100.00) birds crossed

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use 496 T. Doko et al.

Fig. 5. Pattern of routes for spring and autumn migration of Eurasian wigeons among regions in eastern Asia as determined from satellite telemetry, 2007–2016. Numbers of wigeons that migrated along a particular route are indicated next to arrows and arrow thickness is propor- tional to numbers of wigeons that followed a route. Number of wigeons that arrived terminal sites in each region is indicated within boxes with proportion of those in parentheses within boxes. Dashed lines indicate marked wigeons directly between two regions, and bypassed interme- diate regions.

whereas 13 birds moved first to Hokkaido and 1 bird moved 26 June 2010, and 29 April to 26 July, 2007). to the Kuril Islands. Among Eurasian wigeons whose spring migration was via the continental route through China, 11 Analysis by regions individuals migrated through Tumen River and 12 birds Eurasian wigeons from different wintering locations through the Amur River. Eurasian wigeons that migrated via often used the same breeding regions. Also, Eurasian Sakhalin Island mainly went to Magadan region (10 birds). wigeons from the same winter areas such as Hitotuse River In autumn migration, there were eight different routes and Kayoicho Park dispersed to different breeding sites. connecting between breeding sites and wintering sites (Fig. Eight, 10, and nine regions were used for breeding, 5). Eurasian wigeons departed from Chukotka (four birds), spring migration, and autumn migration, respectively (Fig. the Kolyma River (one bird), Magadan (three birds), and 6). Individual regions were served for multiple purposes. For Kamchatka (two birds). The routes used in autumn were instance, Amur River was used for breeding sites as well as same as used during spring, except a route from Magadan stopovers for spring and autumn migrations. Except for the to Kamchatka used by two birds. Five Eurasian wigeons Hamhung region, all regions had more than single purpose arrived at wintering sites in Japan; one bird arrived at Kyushu use. The Hamhung region was used only for spring migra- using route via the continent, and four birds arrived at tion stopovers. No region was used only for breeding. Hokkaido using route via Kamchatka. The most commonly used region (Fig. 7) during breed- Generally, most Eurasian wigeons selected the breed- ing was the Amur River (31%), followed by Chukotka (22%) ing sites in Russia, but one bird (PTT 121319, male) arrived and Kamchatka (16%). Hokkaido was the most frequently at Songhua River, China (46°N, 125°E), and stayed there used region during spring migration (22%), followed by the during the breeding season (24 July to 11 Oct, 2013). Two Amur River (21%) and Sakhalin (15%). The most frequently birds (PTT 34831, and 38612, both males) stayed Tumen used region during autumn migration was Kamchatka (28%), River region (42–43°N, 130–132°E), that are located in followed by Amur River (17%) and Sakhalin (14%). Amur Russia but close to the national boundary between Russia River was ranked in the top two as the most frequently used and China, during breeding seasons (at least from 31 May to regions for breeding, spring migration and autumn migration.

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use Migration patterns of Eurasian wigeons 497

Fig. 6. Stacked column percentage chart for Eurasian wigeons’ use of different regions of eastern Russia and eastern Asia for dif- ferent migration status. Numerals indicate counts of individual wigeons.

Identification of key sites A map of the identified wintering sites, stopover sites, breeding sites, and summering sites is in Fig. 8. The breed- ing sites were composed of 52 presumable breeding sites and six potential breeding sites. The terminal sites in China and along the border of China/Russia were defined as “potential breeding sites,” which included four points near the Songhua River and two points near the Tumen River (Fig. 8). There was significant overlap in geographical loca- tions of breeding sites and stopover sites. Similarly, many wintering sites were also used as migration stopovers. With a few exceptions, the Chukotka region was mainly used only during breeding. There were 118 identical location names summarized from 296 records in mean centers, as detailed geographic locations with latitude and longitude with location names, and migration status (Table 5). Similarly to the trend of Eurasian wigeons’ use of regions (Fig. 6), some locations, e.g. Amur River, Saroma Lake, Zaliv Shelekhova, and so on, were served for multiple purposes. The top five breeding, spring migration, and autumn migration sites were identified (Table 6, Fig. 8). The Anadyr’ River, Piltun bay, Amur River, Apuka River, and Chernaya River, were the most important breeding sites. The Lake Khanka, , Ussuri River, Amur River, and Notsuke Cape were the five important spring migration stop- overs. The Nerpichye Lake, Amur River, Perevolochnyy Zaliv, Neyiskiy bay, and Lake Khanka were the important autumn migration stopovers. Fig. 7. Pie chart for proportion of Eurasian wigeons’ use of differ- ent regions of eastern Russia and eastern Asia during breeding season, spring migration, and autumn migration, separately. The number indicates proportion of counts in percentage.

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use 498 T. Doko et al.

autumn migration, which likely reflects the smaller number of active PTTs. There may be seasonal differences in migration as the majority of Eurasian wigeons in spring migrated via the continental route, and in autumn most birds migrated via the Kamchatka Peninsula. Similar to Northern Pintails (Hupp et al., 2011), there was no movement of Eurasian wigeons from the Kamchatka Peninsula to Sakhalin Island in autumn. Eurasian wigeons that departed from Kamchatka Peninsula moved to Hokkaido directly through Sea of Okhotsk or through Kuril Islands. Loop migrations, in which birds take different routes in autumn and spring due to different wind conditions, are a relatively common phe- nomenon in ducks (Newton, 2008). Like- wise, northeasterly winds, generated by Okhotsk Sea air mass from June to September (Britannica ONLINE JAPAN, 2018), are considered to help Eurasian wigeons from Kamchatka Peninsula to Hokkaido during autumn migration. Similarly to the migration of mallards wintering in Japan (Yamaguchi et al., 2008), Eurasian wigeons employed a “long-stay and short-travel” migration strategy. Eurasian wigeons need to replen- ish energy reserves during migration by making relatively long time stopovers of 14–16 days before their next movement, the same as Mallards. Eurasian wigeons have a relatively small body size (approxi- mately 700 g from Johnsgard (1978)) com- pared to large migratory birds such as eagles or swans, and the proportion of body weight to wing area may influence migration strategy (Yamaguchi et al., 2008). Also, smaller species carry smaller Fig. 8. Important migration sites of different migration status of Eurasian wigeons iden- stores, and therefore they depend on tified via satellite tracking. The point data of locations in the map was plotted based on high-quality foraging sites along the fly- mean centers containing 296 records. There were 58 breeding sites (52 presumable way (Arzel et al., 2006). Hence, the top five breeding sites plus six potential breeding sites), seven summering sites, 62 wintering stopovers identified during our study are sites, 35 autumn migration stopover sites, and 134 spring migration stopover sites in considered to correspond to high-quality total. Potential breeding sites included four points around Songhua River and two points foraging sites. around Tumen River. The number in the map corresponds to the rank of the top five Most Eurasian wigeons bypassed the frequently used locations in Table 6. The parenthesis showing number means that [ ] for breeding, < > for spring stopover, and { } for autumn stopover: [1] Anadyr’ River, [2] Kuril Islands during spring migration Piltun bay, [3] Amur River, [4] Apuka River, [5] Chernaya River, <1> Lake Khanka, < 2 > between Hokkaido and the Kamchatka Ishikari River, < 3 > Ussuri River, <4> Amur River, < 5> Notsuke Cape, {1} Nerpichye Peninsula in favor of more direct transoce- Lake, {2} Amur River, {3} Perevolochnyy Zaliv, {4} Neyiskiy bay, and {5} Lake Khanka. anic migration. This was similar to the spring migration of Northern pintails (Hupp et al., 2011). The transoceanic route via Sea of Okhotsk from northern Honshu to the west coast of DISCUSSION Kamchatka is about 400 km shorter than that via Kuril Migration routes of East Asian Flyway of Eurasian Islands (Hupp et al., 2011). wigeons In spring, all Eurasian wigeons that departed from Although there was considerable individual variation, regions other than Hokkaido mainly from Kyushu either there were three main migration routes: (a) via the continen- directly crossed Sea of Japan or moved to Hokkaido. For tal route through China, (b) via the Kamchatka Peninsula, those that crossed the Sea of Japan, the majority went and (c) via Sakhalin Island. There was less variation during directly to the Tumen River region. Few Eurasian wigeons

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use Migration patterns of Eurasian wigeons 499

Table 5. Identified important locations with regions, longitude, Altyn River A/S 162.129 58.035 and latitude of Eurasian wigeons during migration status of winter- ing, spring migration, breeding, autumn migration, and summering. Apuka River B/S 168.958 61.137 The codes in the status indicate the following: A: Autumn migration, Avacha Bay S 158.417 52.996 B: Breeding, S: Spring migration, W: Wintering, and J: Summering Azhahackye Lake A 161.907 56.144 in Japan. Bolshaye Lake B/S 156.49 52.518 Region Location Status Lon Lat Bukhta Anapka S 164.574 60.053 Amur River A/B/S 135.583 49.043 Kamchatka River B 160.097 56.229 Bukhta Mosolova B 140.628 51.11 Kamchatka Nerpichye Lake A/B/S 162.546 56.378 Kadi River B 140.905 51.814 Opnla River B 156.6 52.116 Protoka Siy Lake B/S 136.439 49.792 Pustaya River S 163.336 60.685 Shchastya Bay B/S 141.359 53.306 Tigil River S 158.257 58.013 Simmi River S 135.859 49.408 Ukolka River A 162.746 57.227 Tatar Strait B 140.563 50.09 Utka River B 156.113 53.11 Amur River Amurakiy Liman B 141.174 52.375 Zaliv Korfa A 165.957 60.361 Ussuri River S 133.251 45.978 Alaurylenveem River A 163.726 60.837 Lake Khanka A/S 132.477 44.843 Iturup A 147.738 44.974 Songhua River B/S 126.82 46.351 Matua S 153.268 48.077 Kuril Is. Ozero Dzhevdukha Lake A 139.369 52.748 Shumshu B 156.263 50.755 Lake Orel B 139.778 53.352 Urup B 149.87 45.78 Ozero Dal’zha Lake B 139.437 52.995 Kava River B/S 147.522 59.488 Ozero Udyl’ Lake B 140.28 52.267 Obkhodnoy River S 149.42 59.761 Chernaya River B 168.249 63.966 Okhotskoe more gulf S 145.934 59.403 Penzhina River A/B 166.68 62.635 Ola River S 151.436 59.803 Chukotka Talovka River B 165.228 62.218 Magadan Perevolochnyy Zaliv A/S 154.218 59.479 Algan River B 171.358 64.036 Takhtoyama River S 154.771 60.233 Anadyr’ River B 173.329 64.991 Tonyskaya Guba Gulf B 152.156 59.284 Gyeongpo Provincial Park S 128.874 37.671 Zaliv Babushkina S 153.736 59.153 Yangyang Namdae-river S 128.647 38.098 Zaliv Shelekhova A/B/S 159.544 61.659 Kumuya Bay S 127.605 39.219 Biwa Lake S 136.139 35.214 Hamhung Orang River S 129.743 41.413 Hakata Bay W/J 130.423 33.594 Namutechon River S 128.372 40.105 Kotogawa A 131.333 34.028 Taehwa River S 129.354 35.57 Osaka Bay J 135.386 34.669 Kunashiri Island S 145.54 43.713 Sea of Japan W 138.288 37.343 Ishikari River S 141.786 43.268 Shinji Lake A 133.243 35.391 Kutcharo Lake S 142.369 45.144 Shou River S 136.977 36.766 S 142.065 44.839 Cape Oma S 140.916 41.509 S 143.568 42.765 Shinano River S 138.879 37.603 Toshibetsu River S 139.868 42.389 Sarugaseki River S 141.538 39.355 Yudonuma S 143.533 42.614 Kamogawa S 133.158 33.913 Not Hokkaido Notsuke Cape S 145.284 43.606 Biwa Lake S 136.055 35.152 Hokkaido Tofutsu Lake A/S 144.376 43.919 Hitotsuse River S/W 131.471 32.069 Sarufutsu River S 142.211 45.213 Kuzuryu River S 136.147 36.171 Akkeshi Lake S 145.074 43.053 S 140.084 40.21 Iwanai plains S 140.589 43.001 Jusan Lake S 140.357 41.046 Saroma Lake S/J 143.909 44.12 Lagoon Hachiro S 140.035 40.042 Lake Oikamanai A 143.486 42.541 Koyaike Park W 135.397 34.79 Rekihune River A 143.376 42.33 Miyama Park W 133.946 34.515 Furen Lake W/J 145.408 43.299 Shinji Lake S 133.077 35.442 Koetoi River J 141.775 45.37 Shinonsen Town J 134.463 35.617 Kotogawa W 131.141 34.046

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use 500 T. Doko et al.

Aniva Bay S 142.702 46.745 traveled through the Korean Peninsula, similar to the spring migration of Mallard (Yamaguchi et al., 2008). Several Chyvo bay A 143.282 52.592 Eurasian wigeons traveled to the Hamhung region in North Katangli S 143.294 51.377 Korea, and followed the coastline of Korea. Eurasian Langeri River S 143.336 50.163 wigeons that traveled Korean Peninsula or Hamhung mainly Nab’il S 143.36 51.276 moved to the Tumen River. Whether they used the Korean Nabil’ skiy bay A/S 143.291 51.633 Peninsula or not, most Eurasian wigeons that migrated via Neyiskiy bay A/S 143.147 52.032 continent visited the Tumen River region, followed by the Amur River region, which highlights the importance of these Ozero Nevskoye Lake B/S 143.733 49.328 Sakhalin regions as stopovers. A direct flight from southwestern Ozero Tunaycha S 143.385 46.686 Japan to Tumen River region across the Sea of Japan Reka Lyutoga River S 142.316 46.829 (approximately 1400 km) can save energy because it is a Sakhalinsk S 142.922 53.645 shorter flight than the one via Korean Peninsula (approxi- Severny gulf A/B 142.567 54.286 mately 2500 km). As often occurs during with satellite transmitters, the Terpeniya Peninsula S 144.464 48.972 number of functional PTTs diminished steadily following Tym River S 142.654 51.003 their deployment. Signal loss mainly occurred before autumn Piltun bay B/S 143.203 53.037 migration, reducing the sample size in that season. How- zaliv Baykal Lake S 142.321 53.351 ever, the failure within one year was shorter than expected Mys Sedlovidnyy Cape B 132.282 43.066 longevity of PTTs, e.g. 879 days in Roshier and Asmus (2009). Possible reasons include (1) the PTTs failed prema- Zaliv Nakhodka Bay A/S 133.009 42.813 turely (Meyburg and Fuller, 2007), or (2) the birds died either Sobon-po Lake S 130.576 42.328 naturally (Meyburg and Fuller, 2007), due to adverse effect Tumen River Bukhta Narva Bay S 131.51 42.992 of transmitters (Barron et al., 2010), or were captured or shot Bukhta Perevoznaya Bay S 131.578 43.014 by hunters outside of Japan (Panov, 1973). Ozero Ptich’ye B/S 130.743 42.507 Eurasian wigeons that departed from the same winter- ing sites in Japan migrated to a variety of breeding sites, Tumen River A 130.673 42.386 using a complex network of stopovers, similar to the Kolyma River S 157.164 66.427 Kolyma observed migration of Mallards (Yamaguchi et al., 2008). Omolon River B 159.333 66.968 This indicates that if an outbreak of avian influenza occurs among Eurasian wigeons in Japan, there is a risk that the viruses would spread, even if the outbreak locations are geographically limited. Table 6. List of the most important locations of Eurasian wigeons’ migration which winter in Japan. The numbers in parentheses indi- cate counts of samples. The samples for breeding are counted by Multiple sites used by individuals during the breeding birds’ number. The samples for stopovers are counted by numbers season of stopovers. Although the majority (63%) of Eurasian wigeons had a single breeding site, a relatively large number (37%) used Use frequency Breeding Spring Autumn multiple sites during the breeding season. The duration of stopover stopover use at these sites averaged approximately 22–23 days. The the highest Anadyr’ Lake Nerpichye possible reasons are (1) failure to breed, (2) disturbance at frequency River (4) Khanka (10) Lake (5) the initial breeding site, and/or (3) movement to moulting sites (Arzel, 2006; Mathers, 1997; Krechmar, 1994). Because the second highest Piltun Ishikari Amur frequency bay (3) River (10) River (4) the incubation period of 22–25 days (Johnsgard, 1978) and fledging period of 40–45 days (Johnsgard, 1978) are the third highest Amur Ussuri Perevolochnyy reported, the short stay at a breeding site, e.g., 22–23 days 1 frequency River (2) River (8) Zaliv (4) is not sufficient for a female to raise young until they can fly the fourth highest Apuka Amur Neyiskiy (Krechmar, 1994). Hence, if females have several breeding frequency River 1 (2) River (6) bay (2) sites, their breeding is likely failed. Since Eurasian wigeons’ flightless duration during moulting is reported to continue the fifth highest Chernaya Notsuke Lake frequency River 1 (2) Cape (4) Khanka 2 (1) from 21–25 days to one month (Krechmar, 1994) and the above average duration on one breeding site is similar, some 1Regarding use frequency of breeding sites, Kamchatka River and breeding sites identified in this study could potentially func- Zaliv Shelekhova were used by two wigeons each, and therefore tion as moulting sites as well. those were ranked as the fifth highest frequency as well.2 Regard- ing use frequency of autumn stopover, Lake Khanka was ranked as Breeding sites in Russia the fifth highest frequency along with Nabil’skiy bay, Shinji Lake, Among 11 regions, the three most major breeding Zaliv Shelekhova, Zaliv Nakhodka Bay, Altyn River, Tofutsu Lake, Penzhina River, Severny gulf, Kotogawa, Azhahackye Lake, Chyvo regions were Amur River, Chukotka, and Kamchatka. On bay, Iturup, Lake Oikamanai, Ozero Dzhevdukha Lake, Rekihune contrary, Kolyma region was the least used breeding site. River, Alaurylenveem River, Tumen River, Ukolka River, and Zaliv Based on use frequencies of the breeding sites, Anadyr’ Korfa. River, Piltun bay, Amur River, Apuka River, Chernaya River,

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use Migration patterns of Eurasian wigeons 501

Kamchatka River and Zaliv Shelekhova are key breeding vation Association, 1995; MacKinnon et al., 2000a). sites. Hence, at least Songhua River is considered to be one According to Nechaev and Gamova (2009), the breeding of the breeding sites, as it belongs to the Heilongjiang site distribution of Eurasian wigeons in the Far East is as region, one breeding area of this species (China Wildlife follows. The east breeding sites ranged at Bering Sea coast Conservation Association, 1995; Li, 1996). The Songhua at eastside, Taimyr Peninsula, Lena River and downstream River is the largest tributary of the Amur River. As for Tumen of Kolyma River. The southern extent of their breeding River region, these locations are included within Northeast ranges from south of Altai Mountains, and south of Transbai- China and their latitude is over 35° N (MacKinnon et al., kal region, ranging from Chaun Bay and Anadyr’ River to the 2000a, b), it is possible to state that they are the breeding midstream and downstream of Amur River; Kamchatka sites of Eurasian wigeons. Peninsula, Sakhalin, Chantal Islands, probably Lower Cis- Migration to China by mallards (Yamaguchi et al., 2008) Amur River area, and Kuril Islands. Generally, most fre- and Northern pintails (Hupp et al., 2011) wintering in Japan quently used breeding sites in our study are included in the has been reported. In addition, through map interpretation in breeding site reported in Nechaev and Gamova (2009) for Yamaguchi et al. (2008), one mallard’s terminal site was also areas of Anadyr’ River, Kamchatka Peninsula (Kamchatka the Songhua River. Thus, the Songhua River is considered River and Apuka River), Sakhalin (Piltun bay), and Amur to be breeding area for both of Mallards and Eurasian River. Frequency of use of Kolyma River and Kuril Islands wigeons. was low for Eurasian wigeons that wintered in Japan. How- Clearly, there is an information gap about breeding sites ever, the usage of these two regions is in agreement with in China. The geographical locations below 60° N, which is Nechaev and Gamova (2009). It is thought that Eurasian written in Johnsgard (1978), is considered to be “blank wigeons wintering in areas other than Japan use these areas” for information regarding breeding sites. Our findings regions more often. Probably our identified locations are can contribute to identify detailed locations for potential new considered to be more detailed at spatial scale than Nechaev breeding sites of this species. and Gamova (2009). There have been other reported obser- vations of breeding and migration of Eurasian wigeons in Eurasian wigeons summering in Japan Russia: South-West Kamchatka (Gerasimov and Zavgarova, The summering of Eurasian wigeons (seven wigeons) in 2008), Kamchatka River (Lobkov, 1986), Lower Cis-Amur Japan was observed. While a few wigeons (two wigeons) did River area (Babenko, 2000; Nazarov, 2004), Sakhalin such not move for migration, majority of wigeons (five wigeons) as zal Piltun, and Aniva Bay (Nechaev, 1991), Kolyma moved drastically (e.g., travelling 1422 km from Kyushu (Kischinskii, 1980), and South Ussuriland (Panov, 1973). island to Hokkaido island, typically their destination was Locations we identified as breeding sites are consistent with Hokkaido). There is no evidence that Eurasian wigeons that these reports. remained in Japan bred from our data or existing literature. Compared to other ducks, the Amur River and Ussuri Generally, there are two reasons for duck species summer- River were commonly used by the Mallard (Yamaguchi et al., ing in Japan: (1) injured/sick conditions, and (2) juveniles 2008), and the Anadyr’ River, Kamchatka Peninsula, and (Ujihara, 2015). However, these reasons were not applicable lower Kolyma River Basin were commonly used by the to our data. Also, many cases that Eurasian wigeons not Northern Pintail (Hupp et al., 2011). In general, Eurasian attached to transmitters that remain in Japan during the wigeons are more widely distributed during breeding season summer were reported (for instance, Wild Bird Society of than Mallards and Northern Pintails. Since the breeding Japan, 1996; Minowa et al., 1999; Okuyama and Fujimaki, sites of Mallards are inside of Far East Eurasia but limited to 2001). Thus, it is evident that Eurasian wigeons sometimes the Eurasian continent, and the ones of Northern Pintails are remain in Japan during the breeding season, even if there is mainly Far East Russia, e.g., Kamchatka, Chukotka, Kolyma, no influence of transmitters, although the reason is unknown. and Magadan. More or less, these distributions of Mallards Future research is needed to reveal why many of Eurasian and Northern Pintails are considered to be included inside of wigeons remain in Hokkaido (especially Saroma Lake 44° N breeding range of Eurasian wigeons. According to Onuma et 143° E, Furen Lake 43° N 145° E, and Koetoi River 45° N al. (2017), Mallards and Northern Pintails might play an 141° E) during the summer. important role in introducing avian influenza virus into Japan. Hence, sharing the same breeding grounds with Key sites identified along the East Asian Flyway Mallards and Northern Pintails indicates potential risk fac- The Amur River is the most important site, because it is tors for Eurasian wigeons to infect the virus. the only stopover which was ranked in top five frequently used stopovers by both spring and autumn seasons. More- Potential breeding sites in China or China/Russia bor- over, the Amur River has been identified as breeding sites of der mallards (Yamaguchi et al., 2008) and Northern pintails Breeding sites of Eurasian wigeons in China, such as (Hupp et al., 2011). In addition, the Ussuri River is consid- Inner-Mongolia and Heilongjiang (China Wildlife Conserva- ered important for both Eurasian wigeons and mallards tion Association, 1995), breeding behavior in Heilongjiang (Yamaguchi et al., 2008). (43°–53° N, 121°E) (Li, 1996), Northeast of China, even in the northwest of China (MacKinnon et al., 2000a, b) have Locations served for multiple purposes of Eurasian been reported. Eurasian wigeons winter in the south of wigeons China, such as the south of Yellow River, Hainan Island, and Basically, all regions were used for more than one pur- Taiwan (the south of latitude 35° N) (China Wildlife Conser- pose. There was no geographic region used only for breed-

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use 502 T. Doko et al.

ing. Instead, there were considerable overlaps between REFERENCES breeding sites and migration stopovers (31°–66° N and 127°–169° E). Japanese regions were used as migration Argos (1996) User’s manual, Service Argos, Inc., Landover stopover sites, wintering sites, and summering. Arzel C, Elmberg J, Guillemain M (2006) Ecology of spring- This indicates that Eurasian wigeons have high adapt- migrating Anatidae: a review. J Ornithol 147: 167–184 Babenko VG (2000) Ptitsy Nizhnego Priamur’ya [Birds of Lower ability and ecological flexibility in their selection of use sites. Cis-Amur River area], Prometei Press, Moscow (in Russian) This enables them to use common locations different pur- Barron DG, Brawn JD, Weatherhead PJ (2010) Meta analysis of poses. Even though there was no clear geographical bound- transmitter effects on avian behaviour and ecology. Methods ary between wintering sites and stopovers, or between Ecol Evol 1: 180–187 breeding sites and stopovers, there was wide separation BirdLife International (2017) Mareca penelope (amended version of between wintering sites and breeding sites. High ecological 2016 assessment). The IUCN Red List of Threatened Species flexibility in selection of breeding sites by Eurasian wigeons 2017: e.T22680157A111892532. In: https://www.iucnredlist.org/ was reported by Krechmar (1994) around Anadyr, northeast- species/22680157/111892532 ern Russia. Boere G, Galbraith CA, Stroud D (2007) Waterbirds Around the World, Scottish Natural Heritage, Inverness Britannica ONLINE JAPAN (2018) Okhotsk Sea air mass, Britannica Implication of habitat range of breeding sites Japan, Tokyo Generally, Eurasian wigeons have been reported to Chen W, Doko T, Fujita G, Hijikata N, Tokita K-iI, Uchida K, et al. breed “about 60° to 70° north latitude, and more rarely far- (2016) Migration of Tundra Swans (Cygnus columbianus) Win- ther south to Lake Baikal and northern Sakhalin” (Johnsgard, tering in Japan Using Satellite Tracking: Identification of the 1978), and “between 50° and 75° N across Scandinavia and Eastern Palearctic Flyway. Zool Sci 33: 63–72 northern Russia to Pacific coast” (Kear, 2005). Based on China Wildlife Conservation Association (1995) Zhongguo niao lei existing literature regarding Eurasian wigeons in China tu jian =: Atlas of birds of China (Mandarin Chinese Edition), (China Wildlife Conservation Association, 1995; Li, 1996; Henan ke xue ji shu chu ban she, Beijing (in Chinese) MacKinnon et al., 2000a, b), previously known breeding CLS (Collecte Localisation Satellites) (2007) Argos User’s Manual range (Johnsgard, 1978; Kear, 2005), and our findings of On Line. In: http://www.argos-system.org/manual/ Doko T, Chen W, Higuchi H (2016) Development of matched (migra- breeding sites at 42°–67° N, it is probably safe to state that tory analytical time change easy detection) method for satel- (1) the range below 35° N is not likely used for breeding lite-tracked migratory birds. ISPRS Ann Photogramm Remote sites, because it is considered to be wintering range, (2) Sens Spatial Inf Sci III-2: 61–68 range of latitudes between 35° to 42° N are blank areas due Douglas DC, Weinzierl R, Davidson SC, Kays R, Wikelski M, Bohrer to lack of information of breeding evidence (nests, eggs or G (2012) Moderating Argos location errors in animal tracking young), and (3) latitude between 43° to 75° N is considered data. Methods Ecol Evol 3: 999–1007 to be the breeding area. Gerasimov YN, Zavgarova YR (2008) Observations of Anatidae In addition, one Eurasian wigeon that was ringed in spring migration along South-West Kamchatka in 2007. In “The Japan migrated to U.S.A. (Yamashina Institute for Ornithol- Biology and Conservation of the the Birds of Kamchatka, Issue ogy, 2002). Hence, Eurasian wigeons which winter in Japan 8” Ed by Centre of Wildlife Protection Publ, pp 83–89 (in Russian) are considered to migrate to Russia, China, and U.S.A. dur- Hupp JW, Yamaguchi N, Flint PL, Pearce JM, Tokita K-i, Shimada ing the breeding season, although the main breeding areas T, et al. (2011) Variation in spring migration routes and breeding for the species are in northeastern Russia. distribution of northern pintails Anas acuta that winter in Japan. ACKNOWLEDGMENTS J Avian Biol 42: 289–300 Johnsgard PA (1978) Ducks, Geese, and Swans of the World, Uni- We thank the following people who helped with the field work: versity of Nebraska Press Lincoln, Lincoln and London Yuji Inoue, Fumitaka Iseki, Nobuhiko Kataoka, Fumio Katsuno, Kear J (2005) Bird Families of the World: Ducks, Geese and Swans: Mitsuyo Kido, Wataru Kitamura, Kenji Maruyama, Toru Nakahara, Species Accounts (Cairina to Mergus), Oxford University Shiro Nakatake, Fumihito Nakayama, Kazuhisa Oue, Hiroto Okabe, Press, Oxford Ran Ohya, Ryuji Sakane, Sayaka Someya, Katsushige Sakai, Killick R, Eckley IA (2014) changepoint: An R Package for Change- Sumire Sakai, Hisashi Sugawa, Naoki Sugihara,Kazuo Takatsu, point Analysis. J Stat Softw 58: 1–19 Kouki Uchinomiya, Shingo Uemura, Takahiko Yamamoto, and Killick R, Eckley I, Haynes K, Fearnhead P (2014) changepoint: An Midori Yamane. We also thank Dr. Yuzo Fujimaki for intensive trans- R package for changepoint analysis. In: http://cran.r-project. lation work of Russian references and Dr. Jerry Hupp for English org/web/packages/changepoint/index.html proofreading. This study was funded by the Ministry of the Environ- Kischinskii AA (1980) Birds at Koryaku Highlands, Nauka, Moscow ment of Japan and the Ministry of Education, Culture, Sports, (in Russian) Science and Technology, Japan (Grant/ Award Number: The Kodama T, Shiraishi H, Fuchigami A (2014) About the causes for University of Tokyo 21st Century COE Program “Biodiversity and disappearance of Glue Hatay in fishing grounds at estuarine Ecosystem Restoration Research Project”). Capturing birds was region in the Sea of Ariake District. Bull Fukuoka Fisheries Mar authorized by the local governments of capturing sites. This license Technol Res Cent 13–23 (in Japanese) was issued based on the facts that safety and security of birds can Krechmar AV (1994) Eurasian wigeon (Anas penelope) in North- be guaranteed. License numbers are listed as follows: Miyazaki eastern Asia. Zoologicheskii Zhurnal 73: 68–79 (in Russian) Prefecture (24940-1053), Fukuoka Prefecture (21 JI DAI 118-14 Lane SJ, Azuma A, Higuchi H (1998) Wildfowl damage to agricul- GOU, 23 JI DAI 11-9 GOU, 24 JI DAI 64 GOU-11, 25 JI DAI 64 GOU- ture in Japan. Agr Ecosyst Environ 70: 69–77 11, 25 JI DAI 45 GOU-11), Hyogo Prefecture (JI DAI 79~83 GOU, JI Li X (1996) The status and conservation of wildfowl (anatidae) in DAI 116~122 GOU), and Okayama Prefecture (DAI 10-7-1~6 GOU, Heilongjiang Province, China. J Northeast For Univ 7: 39–42 DAI10-11-1~8GOU). Lobkov EG (1986) Breeding Birds of Kamchatka, Academy of Sci- ences (Far Eastern Scientific Centre), Vladivostok (in Russian)

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019 Terms of Use: https://bioone.org/terms-of-use Migration patterns of Eurasian wigeons 503

MacKinnon JR, Phillipps K, He F-Q (2000a) A Field Guide to the Palm EC, Newman SH, Prosser DJ, Xiao X, Ze L, Batbayar N, Birds of China, Oxford University Press, Oxford Balachandran S, Takekawa JY (2015) Mapping migratory fly- MacKinnon JR, Phillipps K, He F-Q (2000b) A Field Guide to the ways in Asia using dynamic Brownian bridge movement mod- Birds of China (Chinese translation), Hunan Jiaoyu Chuban-she, els. Mov Ecol 3: 3 Changsha (in Chinese) Panov EN (1973) The birds of South Ussuriland (fauna, biology, Mathers RG, Montgomery WI (1997) Quality of Food Consumed by behaviour), Nauka, Novosibrsk (in Russian) over Wintering Pale-Bellied Brent Geese Branta bernicla hrota R Development Core Team (2006) R: A Language and Environment and wigeon Anas penelope. Biol Environ Proc Roy Irish Acad for Statistical Computing, R Foundation for Statistical Comput- 97B: 81–89 ing, Vienna Meyburg B-U, Fuller MR (2007) Satellite tracking. In “Raptor Roshier DA, Asmus MW (2009) Use of satellite telemetry on small- Research and Management Techniques” Ed by DM Bird, KL bodied waterfowl in Australia. Mar Freshwater Res 60: 299– Bildstein, Hancock House Publishers Ltd, Surrey, BC Canada, 305 pp 242–248 Torsten H, Kurt H (2015) exactRankTests: Exact Distributions for Ministry of Land Infrastructure and Tourism (2012) Bird strike data Rank and Permutation Tests. In: http://cran.r-project.org/web/ in 2012. In: http://www.mlit.go.jp/common/000994317.pdf packages/exactRankTests/index.html Minowa Y, Tamura M, Kuwabara K (1999) Bird list of Ujihara O, Ujihara M (2015) An Identification Guide to the Ducks of estuary, Chiba Prefecture, Central Japan. In “Avifauna at the Japan, Seibundo Shinkosha Estuary Basin of the Tamagawa River, Central Japan - Birds in Wilcoxon F (1945) Individual comparisons by ranking methods. Tideland Areas of Tokyo Bay” Ed by K Kuwabara, The Tokyu Biometrics Bull 1: 80–83 Foundation for Better Environment, Tokyo, pp 96–149 (in Wild Bird Society of Japan (1996) Wild birds information and obser- Japanese) vation records (1994. 8 - 1995. 7) STRIX 14: 205–211 (in Miyabayashi Y, Mundkur T (1999) Atlas of Key Sites for Anatidae in Japanese) the East Asian Flyway, Wetlands International Tokyo Japan World Organisation for Animal Health (2014) Information received and Kuala Lumpur Asia Pacific on 25/12/2014 from Dr Evgeny Nepoklonov, Vice-minister, Min- Nazarov Y (2004) Birds of Vladivostok and suburbs, Far Eastern istry of Agriculture, Ministry of Agriculture, Moscow, Russia. In: Univ Press, Vladivostok (in Russian) http://www.oie.int/wahis_2/public/wahid.php/Reviewreport/ Nechaev VA (1991) Birds of Sakhalin Island, USSR Academy of Review?reportid =16826 Sciences, Far Eastern Branch, Institute of Bioloby and Pedol- Yamaguchi N, Hiraoka E, Fujita M, Hijikata N, Ueta M, Takagi K, et ogy, Vladivostok (in Russian) al. (2008) Spring Migration Routes of Mallards (Anas platy- Nechaev VA, Gamova T (2009) Birds of Russian Far East (An anno- rhynchos) that Winter in Japan, Determined from Satellite tated catalogue), Dal’nauka, Vladivostok (in Russian) Telemetry. Zool Sci 25: 875–881 Newton I (2008) The Migration Ecology of Birds, Academic Press, Yamashina Institute for Ornithology (2002) Atlas of Japanese London Migratory Birds from 1961 to 1995. Ministry of Environment of Okuyama M, Fujimaki Y (2001) Birds of Tokachi District, Hokkaido Japan, Tokyo (in Japanese with English caption) 10. Anatids of shallow lakes along the lower reaches of the Zeileis A, Grothendieck G (2005) zoo: S3 Infrastructure for Regular Tokachi River. Res Bull Obihiro Univ Nat Sci 22: 135–142 and Irregular Time Series. J Stat Softw 14: 1–27 Onuma M, Kakogawa M, Yanagisawa M, Haga A, Okano T, Neagari Zeileis A, Grothendieck G, Ryan JA, Andrews F (2015) zoo: S3 Y, et al. (2017) Characterizing the temporal patterns of avian Infrastructure for Regular and Irregular Time Series (Z’s Ordered influenza virus introduction into Japan by migratory birds. J Vet Observations). In: http://cran.r-project.org/web/packages/zoo/ Med Sci 79: 943–951 index.html Owen M, Mitchell C (1988) Movements and migrations of wigeon Anas penelope wintering in Britain and Ireland. Bird Study 35: (Received January 8, 2019 / Accepted June 14, 2019) 47–59

Downloaded From: https://bioone.org/journals/Zoological-Science on 12 Dec 2019

Terms of Use:View https://bioone.org/terms-of-use publication stats