Shorebirds of the Yellow Sea
Importance, threats and conservation status
Wetlands International, 2002
ISBN 90 5882 009 2
Regular shorebird surveys have been conducted in South Korea since 1993. Most counting has been concentrated at nine major sites along the west and south coasts (Ganghwa Do, Yong Jeong Do, Namyang Man, Asan Man, Mangyeung Gang Hagu, Dongjin Gang Hagu, Seosan Reclaimed Area, Geum Gang Hagu and Suncheon Man) (Yi & Kim in prep.). Fewer counts have been made at a number of other sites (Yi & Kim in prep., Moores 1999a). From 1993 to 1996, counts were conducted once or twice during both northward and southward migration at the first six major sites. From 1997 onwards, count coverage was expanded to include the last three major sites and the count frequency was increased to at least once a month during each migration period (Yi & Kim in prep.). A limited amount of counting has been conducted during the non-breeding season (Moores 1999a). Coverage of the South Korean coastline, both spatially and temporally, is considered to be good. The locations of the major South Korean sites that have been surveyed for shorebirds are shown in Figure 4. No shorebird count data is available from North Korea, except for limited information from the Mundok Migratory Bird Wetland Reserve.
The coverage and frequency of shorebird surveys in China have been considerably less than in South Korea. Most recent shorebird counting has been conducted in conjunction with shorebird ecology training courses held at a number of coastal nature reserves (Barter et al. 1997a, 2000b, 2000h, 2000i, 2001, 2002); Zhu et al. 2000), but earlier counts at Chongming Dao and in the Huang He Delta were carried out by the East China Waterbird Ecology Group (Wang & Tang 1990a, 1990b; Wang et al. 1991, 1992), and at the Yancheng National Nature Reserve by nature reserve staff (Wang & Liu 1994; Wang 1997; Wang & Barter 1998). Visiting overseas birdwatchers have made numerous counts at Shi Jiu Tuo (Happy Island) and their data have been obtained either directly from them or via web sites.
Locations of the major sites surveyed in South Korea
Most count data from China are for the northward migration period when it is estimated that approximately one-third of the intertidal areas have been surveyed at least once; the southward migration and the non-breeding periods have received much less coverage (Figure 5).
Counts in South Korea were mostly carried out when birds were concentrated at high tide roosts, which are generally easily accessible by vehicle. In China, shorebird surveying was complicated by the extensive nature of the intertidal flats, which are often inaccessible to vehicles, and the fact that the intertidal flats, at most sites, were rarely completely covered at high tide. Typically, counting was carried out by a number of teams that each covered an assigned section of the coastline, often walking many kilometres to reach the tide edge and the birds. Counts were made at all stages of the tide. The exceptions were Yalu Jiang NNR and Tianjin Municipality, where counts were made from the seawall of birds pre-roosting before they flew into shrimp and fish ponds at high tide.
Estimates of species numbers obtained from high tide counts are generally assumed to accurately reflect actual numbers present. However, this may not be true. Colwell & Cooper (1993) have shown that the optimal stage of the tide for counting can vary for different species. Rappoldt et al. (1985) found that underestimation of numbers occurs during counting of roosts. It seems that high tide counts are more likely to result in conservative estimates of numbers present rather than overestimates.
Extent of shorebird surveys in China
In a study carried out in the UK that compared high and low tide counts in 39 estuaries (Musgrove 1998), it was shown that low tide counts recorded 85% of the total number obtained at high tide, with most species being undercounted during low tide conditions. However, some species, such as Bar-tailed Godwit, Dunlin and Grey Plover, exhibited little difference, whilst Red Knot and Greenshank were counted in significantly higher numbers at low tide. It was found that there was a greater difference between high and low tide counts when larger intertidal areas were being counted, presumably due to detectability problems at long distances. Rappoldt et al. (1985) make the point that missing of birds during wide-scale counting often leads to significant underestimates.
Therefore, the counting conditions in China could be expected to result in underestimates of numbers, especially for those species which were distributed widely across the intertidal areas (e.g. Eurasian and Eastern Curlews, and Kentish and Lesser Sand Plovers). The focus on counting intertidal areas would also lead to undercounting of species that use coastal non-tidal wetlands, e.g. Black-tailed Godwit, the Tringa sandpipers, Pied Avocet, Black-winged Stilt and Grey-headed Lapwing.
There is no measure of the precision of the count estimates, as replicate counts were not made during the South Korean and Chinese surveys.
The Ramsar Convention (Convention on Wetlands [Ramsar, Iran, 1971]) has developed eight criteria to identify internationally important wetlands (Ramsar Convention Bureau 2000). Two of these (Criteria 5 and 6) specifically relate to assessing whether a wetland supports internationally important numbers of waterbirds:
"A wetland should be considered internationally important if it regularly supports 20 000 or more waterbirds.
A wetland should be considered internationally important if it regularly supports 1% of the individuals in a population of one species or subspecies of waterbird."
The objective nature of these particular criteria makes them easy to apply to the identification of internationally important shorebird sites in the Yellow Sea, although the "1% criterion" does require an estimate of population numbers for the East Asian-Australasian Flyway.
The Convention acknowledges that these criteria will be applied to wetlands of varying size and provides the guideline that "wetlands identified as being of international importance …. should form an ecological unit, and may thus be made up of one big area or a group of smaller wetlands" (Ramsar Convention Bureau 2000).
The ecological unit concept is relatively easy to apply in South Korea, as count sites generally comprise bays and estuaries.
The application of the concept in China is often difficult because of the large extent of the intertidal areas and the "open-ended" nature of the wetlands surveyed. Thus, in China administrative regions have been used to delimit a site as the great majority of count data was collected during surveys of nature reserves or municipalities. The exception was the island of Shi Jiu Tuo.
The term "regularly supports" is used in both criteria and the Convention provides the following guidance in the interpretation of this phrase (Ramsar Convention Bureau 2000):
"a wetland regularly supports a population of a given size if:
- the requisite number of birds is known to have occurred in two thirds of the seasons for which adequate data are available, the total number of seasons being not less than three; or
- the mean of the maxima of those seasons in which the site is internationally important, taken over at least five years, amounts to the required level (means based on three or four years may be quoted in provisional assessments only)."
In its guidance the Bureau also adds that:
"In some instances, however, for species occurring in very remote areas or which are particularly rare, or where there are particular constraints on national capacity to undertake surveys, areas may be considered suitable on the basis of fewer counts. For some countries or sites where there is very little information, single counts can help establish the relative importance of the site for a species."
In China, most sites have been counted less than five times, some only once. Where sites have been visited more than once the objective has normally been to get a better understanding of migration strategies, rather than to obtain an estimate of total numbers of a species using a site during the migration or non-breeding periods. South Korean sites have been counted more frequently, but some only once or twice.
Due to these data limitations, it has been decided to use the highest recorded count to determine whether a site holds internationally important numbers of a particular species.
The use of the highest count is likely to provide a conservative estimate of the total numbers of a species using a staging site due to:
- underestimation of numbers present (see above);
- the fact that birds are often continuously moving through a site, with the result that the total number of individuals of a species using the site during a defined period, e.g. northward migration, will exceed the highest count (Howes & Bakewell 1989; Wetlands International - Oceania 2000). This effect is often called "migration turnover".
It is considered to be very unlikely that additional count data would lead to any of the identified sites losing their internationally important status; in many cases shorebirds occur in concentrations far in excess of the 1% level and it highly improbable that these are exceptional counts.
The availability of population estimates for shorebird species in the East Asian-Australasian Flyway (Bamford et al. in prep.) means that the 1% criterion is the most appropriate one to use in assessing the importance of wetlands in the Yellow Sea. The employment of this criterion has the significant advantage that it enables identification of suites of sites that are important for individual shorebird species. The numerical values of the 1% criteria used for the 36 species covered in the species accounts are listed in Table 2.
The importance of the Yellow Sea for a particular species is best measured by the proportion of its flyway population that uses the region. This can be calculated from estimates of the numbers using the Yellow Sea and the flyway population of each species.
As northward migration is the only period for which there is sufficient data, estimates of the number of individuals of a species using the Chinese and South Korean coasts of the Yellow Sea have only been made for this time. The following steps were used:
- Calculation of the number of individuals of a species occurring in each of the different regions of the Chinese coastline by multiplying the species density (shorebirds/km2), obtained from the site survey in the region, by the regional intertidal area (see Figure 3). A weighted average density was calculated when more than one site had been surveyed in a region. The regional estimates were added together to obtain an estimate of the total number of individuals of a species using the Chinese coastline. In order to avoid the possibility of overestimation, allowances were made in situations where a species concentrated in exceptionally large numbers at a particular site. The average density for China was used for the Shandong Peninsula.
- Addition of the South Korean estimate (from Yi & Kim in prep.) to the Chinese estimate obtained in step 1.
The resulting estimate of the total number of individuals of a species passing through the Yellow Sea is likely to be conservative because:
- counts probably underestimate actual numbers present at a site;
- the maximum species count at a site is an underestimate of the total number passing through the site because it does not take migration turnover into account;
- North Korea, whose intertidal flats represent approximately 11% of the Yellow Sea area, is not included in the calculation.
Double-counting of the same bird at more than one site will lead to inflation of estimates but the resulting amount is probably outweighed by the underestimation resulting from the effect of the three factors listed above.
The true importance of the Yellow Sea for a species on northward migration is most appropriately measured by the proportion of the flyway breeding population it is supporting, as immature birds mostly remain on or near the non-breeding areas during the migration periods and breeding season. The flyway breeding population for a species has been approximately calculated by assuming that breeding adults comprise 80% of the species estimated flyway population. This follows Wilson & Barter (1998) who used the 80% factor to calculate the sizes of the Great Knot, Red Knot and Bar-tailed Godwit flyway breeding populations. The factor is probably too high for most species and its use will result in an underestimate of the proportion of the flyway breeding population using the Yellow Sea; this will be compounded by the probably conservative estimate of the species numbers (see above). A factor of 100% has been used for species that may breed in their first year (e.g. Black-tailed Godwit, Marsh Sandpiper, Dunlin and Kentish Plover) (del Hoyo et al. 1996).
|Black-tailed Godwit Limosa limosa||1 600|
|Bar-tailed Godwit Limosa lapponica||3 250|
|Little Curlew Numenius minutus||1 800|
|Whimbrel Numenius phaeopus||550|
|Eurasian Curlew Numenius arquata||350|
|Eastern Curlew Numenius madagascariensis||380|
|Spotted Redshank Tringa erythropus||400|
|Common Redshank Tringa totanus||650|
|Marsh Sandpiper Tringa stagnatilis||900|
|Common Greenshank Tringa nebularia||550|
|Spotted Greenshank Tringa guttifer||10|
|Terek Sandpiper Xenus cinereus||500|
|Grey-tailed Tattler Heteroscelus brevipes||400|
|Ruddy Turnstone Arenaria interpres||310|
|Asian Dowitcher Limnodromus semipalmatus||230|
|Great Knot Calidris tenuirostris||3 800|
|Red Knot Calidris canutus||2 200|
|Sanderling Calidris alba||220|
|Red-necked Stint Calidris ruficollis||3 150|
|Sharp-tailed Sandpiper Calidris acuminata||1 600|
|Dunlin Calidris alpina||9 500|
|Curlew Sandpiper Calidris ferruginea||1 800|
|Spoon-billed Sandpiper Eurynorhynchus pygmaeus||40|
|Broad-billed Sandpiper Limicola falcinellus||180|
|Red-necked Phalarope Phalaropus lobatus||1 000|
|Eurasian Oystercatcher Haematopus ostralegus||100|
|Black-winged Stilt Himantopus himantopus||200|
|Pied Avocet Recurvirostra avosetta||300|
|Grey-headed Lapwing Vanellus cinereus||100|
|Northern Lapwing Vanellus vanellus||600|
|Grey Plover Pluvialis squatarola||1 250|
|Kentish Plover Charadrius alexandrinus||950|
|Little Ringed Plover Charadrius dubius||250|
|Lesser Sand Plover Charadrius mongolus||600|
|Oriental Plover Charadrius veredus||700|
|Oriental Pratincole Glareola maldivarum||750|
|NM northward migration||PNR National Nature Reserve|
|SM southward migration||MBWR Migratory Bird Wetland Reserve|
|L Lake||> greater than|
|R River||< less than|
|EAAF East Asian-Australasian Flyway||>> much greater than|
|PNR Provincial Nature Reserve|
|n., ne…..w., sw. Also, when used as an adjective, e.g. s. Yakutia.|