3.1. Population changes of Cygnus spp.
C. cygnus and C. columbianus populations searching for the estuary of Nakdonggang did not arrive in a consistent population size; winter populations showed significant changes every year (Figs. 2 and 3). However, a certain pattern of arrival was noted. If many birds were wintering in a certain year, the number decreased significantly the following year, and restoration of the population in the following year occurred in two-year cycles. While the C. cygnus maintained this pattern until recently, the population of C. columbianus has declined sharply since 2007, continuing to find few Nakdonggang estuary since 2015. C. cygnus has also not been showing any tendency to recover the pattern since 2017.
The 15-year-long population record showed that C. columbianus seemed to present a sharp decline in the winter populations, which could be attributed to the Nakdonggang estuary being no longer feasible as their wintering site. There was, however, no sharp decline in C. cygnus populations, rather only a moderate decline was identified. The largest winter population of C. cygnus was 4,219 individuals in 2011, whereas the smallest was 783 in 2009. The wintering population of C. columbianus plummeted from 181 individuals in 2006 to only eight in 2019. Although the tendency of the population decline between the two species was very different, the pattern of population increase and decrease, compared to that in the previous year, was similar. Despite these characteristics, both the species showed the same trend in response to changes in the overall habitat of the wintering site. However, C. columbianus could still be considered to be much more sensitive to changes than C. cygnus. Analysis of the various records that indirectly identified the cause underlying the change in wintering population showed the damage to food supply at the estuary to be the most relevant.
The Nakdonggang estuary has allowed large-scale growth of Scirpus planiculmis in the past, and Cygnus spp. did not fly to the surrounding area for food. However, as the area of distribution of S. planiculmis decreased sharply in 2009 (Yiet al., 2011), the wintering population of Cygnus spp. decreased extraordinarily. While the relationship between reduction of wetland plant area and decrease of the wintering population of Cygnus spp. cannot be clearly stated statistically, decrease in the distribution area of seagrass and Schenoplectus trqueter, an important food of Cygnus spp., may have affected the wintering population (Kim et al., 2015).
Wintering populations from 2005 to 2019 were analyzed monthly from November to February (Figs. 4 and 5). Analyzing the year-by-year population changes, a very important aspect of the change in wintering population of Cygnus spp. was observed. The total wintering population did not drop sharply until 2019, rather it had been declining since 2006. For three years from 2006, the initial wintering population remained intact; however, since January, the population started reducing rapidly. This may be attributed to the fact that the birds moved due to a lack of food in the middle of the wintering period. This phenomenon was consistent in 2010, 2011, 2013, and 2014. The early wintering population had disappeared.
Considering their wintering pattern, Cygnus spp. seemed to continue visiting the area for several years, even after the winter habitat was damaged; it took years of repeated learning to infer that they were abandoning their habitat. C. cygnus, whose trend of population decline was relatively unclear, tended to maintain a certain population continuously, especially toward the second half of the winter.
Considering the wintering pattern of these two species, differences between them in terms of wintering site selection factors may be considered significant. Therefore, further studies on the key factors affecting wintering of the species would be required in future.
The general trend of local wintering movements during a single wintering period was the same as in previous studies.
For an analysis of the behavior of C. columbianus in Huangpi lake, located in the mid-to-lower floodplain area of the Yangtze River, Wang et al. (2021) explored the behavioral patterns from early November to late February and found them moving to alternative habitats when the environment in a particular area deteriorated. Hong (2004; 2009) conducted spatially separated surveys at the Nakdonggang River mouth and reported C. cygnus, which visited this place, to move locally according to environmental changes after the initial winter. Hahm and Kim (2001) inferred, from 1989 to 1999, that Cygnus spp. populations that was initially wintering at Junam Reservoir had moved to the mouth of Nakdonggang (Riv.) due to habitat instability. However, this pattern was not entrenched by habitat disturbances in a single year; rather, changes in the initial selection of wintering sites occurred only after learning experiences over several years.
3.2. Change of land-cover types according to time-series analysis
In the late 1980s, the Cygnus spp. concentrated in the wetland protected area of the Nakdonggang estuary for wintering, as is evident from a time-series analysis of the changes in surrounding land coverage by the end of 2010 (Table 1, Fig. 6). Of the seven types of land coverage, including urbanized area, farmland, forest, grassland, wetland (waterside vegetation), bare land, and water area, there was a sharp change in the ratio of more than 5% compared to that at the end of the 1980s.
Considering the distance, within 2 km of the protected boundary the urbanized area increased by 19.6%, from 9.1% in the late 1980s to 28.7% in the late 2010s, and farmland decreased by 17.8%, from 21.6–3.9%. The water area also plunged 7.6%, from 52.1–44.5%. If the distance was extended to 4 km, the increase in urbanized area was 15.3%, which was slightly lower than that of the area within 2 km. The rate of increase of urbanized areas was slightly low, although the changes were similar enough to threaten the habitat. However, the decline in farmland was the opposite of that of the urbanized area, which implies that the farmland mainly changed into an urbanized area. The rate of decrease of water area was also evident, identified by the urban development through reclamation.
Busan is the second-largest city in Korea. In the past, there was very little urban development around the Nakdonggang River, since it was a floodplain; however, urbanization of this area progressed quickly due to the expansion of the city and creation of embankments to prevent flooding.
The wetland protected area at the mouth of Nakdonggang was not set up with a buffer zone; therefore, the urbanized area expanded rapidly near the wintering site of Cygnus spp. Moreover, the sharp decline in farmland, including paddy fields (Shimada, 2014, Watanabe et al., 2008), which are used as major feeding and resting areas by Cygnus spp., was notable. It failed to prevent a sharp decline in the rice paddies, which provides food for the Cygnus spp., and showed why the characteristics of wintering population changes in the two species were different. Fujioka et al. (2010) had reported that C. columbianus was more dependent on rice paddies than C. cygnus, as per feeding activity analysis of both species during the wintering period. However, C. cygnus had the advantage of circumventing the food shortage (due to a decrease in rice fields) with other sources of food.
Kim et al. (2005) had reported that reclamation of the estuary mudflats of Nakdonggang could have caused a significant change in the distribution of seagrass and S. trqueter, the main food plants for Cygnus spp. C. columbianus, which are approximately 20 cm smaller in length than C. cygnus, may have been more affected due to the reduction of food in the water.
Table 1
Area ratio (%) of land cover types according to the time-series.
Buffer
|
Year
|
Urbanized area
|
Farmland
(mostly paddy field)
|
Forest
|
Grassland
|
Wetland
(Shore vegetation)
|
Bare land
|
Water
|
2 km
|
1980s
|
9.07
|
21.60
|
10.60
|
1.02
|
0.32
|
5.30
|
52.08
|
1990s
|
17.69
|
13.91
|
7.96
|
4.32
|
2.14
|
6.29
|
47.69
|
2000s
|
20.33
|
12.42
|
7.44
|
3.41
|
1.94
|
6.81
|
47.63
|
2010s
|
28.68
|
3.85
|
6.19
|
2.71
|
7.23
|
6.88
|
44.46
|
4 km
|
1980s
|
7.69
|
16.85
|
16.71
|
1.37
|
0.24
|
3.11
|
54.02
|
1990s
|
14.28
|
12.77
|
13.75
|
4.01
|
1.34
|
4.76
|
49.08
|
2000s
|
17.46
|
11.88
|
13.72
|
2.86
|
1.24
|
3.82
|
49.02
|
2010s
|
22.94
|
4.54
|
12.74
|
2.34
|
3.96
|
5.71
|
47.78
|
6 km
|
1980s
|
8.39
|
17.53
|
20.88
|
2.51
|
0.23
|
2.23
|
48.23
|
1990s
|
13.63
|
14.71
|
18.22
|
4.38
|
1.20
|
4.11
|
43.76
|
2000s
|
17.07
|
13.86
|
18.58
|
3.14
|
1.12
|
3.71
|
42.51
|
2010s
|
22.96
|
6.38
|
17.04
|
2.60
|
3.15
|
6.13
|
41.74
|
8 km
|
1980s
|
9.46
|
17.60
|
23.69
|
2.84
|
0.19
|
1.92
|
44.30
|
1990s
|
14.00
|
14.82
|
20.63
|
4.79
|
0.90
|
3.90
|
40.95
|
2000s
|
17.35
|
13.84
|
21.23
|
3.44
|
0.84
|
4.15
|
39.14
|
2010s
|
23.52
|
6.82
|
19.76
|
3.03
|
2.59
|
6.15
|
38.13
|
10 km
|
1980s
|
10.17
|
17.16
|
24.89
|
3.06
|
0.15
|
1.80
|
42.76
|
1990s
|
14.16
|
14.68
|
21.78
|
4.91
|
0.75
|
3.60
|
40.12
|
2000s
|
17.31
|
13.83
|
22.21
|
3.42
|
0.69
|
4.21
|
38.33
|
2010s
|
23.07
|
7.66
|
20.69
|
3.07
|
2.26
|
5.94
|
37.31
|
As a result of analyzing the trend of changes in paddy fields, which are spaces where Cygnus spp. intensively consume food during wintering (Fig. 7, Fig. 8), 28% of paddy filed with a total area of 16㎢ decreased from 2006 to 2019. During this period, the annual increase or decrease in the number of C. cygnus was somewhat severe, but the overall decline was identified. In the case of C. columbianus, the trend of decline was clearer, with a large decrease since 2007 and no recovery. In line with analysis of the time-series land cover change at the Nakdonggang estuary for about 40 years, the relationship of the population changes of C. cygnus and C. columbianus according to the trend of farmland change was identified, especially the rapid decrease of C. columbianus is strongly related to the the decrease in the area ratio of paddy fields.