Number and Nest-Site Selection of Breeding Black- Necked Cranes Over the Past 40 Years in the Longbao Wetland Nature Reserve, Qinghai, China

Qiufang Wei Aerospace information research institute, Chinese academy of sciences Yun Shao Chinese Academy of Sciences Aerospace Information Research Institute Chou Xie (  xiechou@aircas.ac.cn ) AIR: Chinese Academy of Sciences Aerospace Information Research Institute Baoshan Cui Beijing Normal University School of Environment Bangsen Tian Chinese Academy of Sciences Aerospace Information Research Institute Brian Brisco Natural Resources Canada Kun Li Chinese Academy of Sciences Aerospace Information Research Institute Wenjia Tang Qinghai eco-environment monitoring center


Introduction
Black-necked crane (BNC, Grus nigricollis) is the only one of the world's 15 crane species that breeds completely in the high-altitude freshwater wetlands of the Qinghai-Tibetan Plateau (Bishop 1996; Li and Bishop 1999;Li et al. 2005). The global population of this species increased from 5600-6000 to 10000-11000 from the 1990s to the present (Bishop 1996 (Dwyer et al. 1992;Cangjue et al. 2008). Earlier studies focused on Longbao, although the number is limited. In the 1980s, several scientists investigated the population numbers, geographical distribution, habitat characteristics and other ecological problems of BNC in Longbao (Guo JT 1981;Yao JC 1982;Li DH and Zhou ZJ 1985). Recently, only Farrington and Zhang surveyed this area and counted the population numbers . Researchers usually use binoculars or spotting scopes to survey areas and count the number of animals. However, studies performing qualitative in uence factor analyses (such as wetland variation and climate change) on the number of BNCs are de cient.
Only Ma researched the dynamic change in the ecological environment in the Longbao region (2009).
Understanding nest site selection characteristics is imperative for making conservation decisions on bird habitats, and managers usually lack such knowledge (Bock 1997, Caughley 1994 Song et al. 2014). According to some early surveys, the BNCs in Longbao prefer to nest in alpine meadows and marshlands with shallow water that are di cult for people to enter (Lü ZB et al. 1980;Guo JT. 1981;Wang YH et al. 1989). Wang and Chen (2002) classi ed BNC nests into ve types: island nests, mud nests, reed nests, grassy island nests, and temporary nests. Most of the nests in Longbao are island nests, and some are grass nests. Wu (2009) performed a principal component analysis (PCA) and found that the factors associated with the selection of nest sites in the Ruoergai wetland included water body size, distance to water, water depth and nest concealment level. Another study performed a factor analysis and found that nest site selection by BNC in the Yanchiwan National Nature Reserve was determined by habitat type, disturbance and water depth (Zhang and An 2017). Other researchers (Dwyer and Bishop 1992) collected nest-site data in the Tibetan autonomous region, including altitude, height above water, distance to upland and distance to disturbance. Recently, Farrington and Zhang located 29 BNC nests within the main Longbao wetland in the spring of 2011 , although the characteristics underlying the selection of nesting environment, such as the water area, vegetation area and interference condition, were not analyzed. Obtain all the nest site characteristics through observations is di cult, especially for certain nest locations that may be surrounded by water or marsh that investigators could not obtain. Compared with on-site observations, abundant nesting information can be e ciently obtained through satellite images.
Plateau wetlands provide indispensable habitats for BNCs. The dynamic changes in plateau wetlands caused by climate uctuations and human disturbance directly in uences the microhabitats and biological behaviors of BNC. Therefore, in this study, we selected the Longbao NNR as a case study to (1) collect the number of BNCs in Longbao NNR since 1978; (2) analyze the nest site selection characteristics of BNCs in Longbao wetland; and (3) analyze the in uence of land cover and climate change on the BNCs. We hope to provide a reference for decision-making regarding the protection of BNCs in the Longbao NNR wetland.

Study Area
The Longbao National Nature Reserve (NNR, 33.11°~33.27°N, 96.39°~96.69°E), which is located in Longbao Town of Yushu County, Qinghai Province, has an area of approximately 100 km 2 and an elevation of approximately 4100 ~ 4200 m. It was established as the rst NNR in 1986 to protect BNCs and its habitats in Qinghai Province. The reserve is approximately 14 km long and has a maximum width of approximately 3.5 km and a maximum depth of approximately 4 m (Fig. 1). The Longbao Lake basin is a permanent freshwater lake surrounded by mountains on three sides. The wetland is fed by groundwater, streams, precipitation, and snowmelt. It nally ows into the Tongtian River, which is the main body of the Yangtze River. The wetland is a typical swampy meadow and alpine meadow area that provides su cient food and a good ecological environment for waterfowl migratory birds. It is reputed as "the hometown of BNCs" by bird experts around the world.
The Longbao Lake basin has a typical plateau continental climate that varies signi cantly during the four seasons. Under the control of the Qinghai-Tibet high pressure, windy, dry and cold conditions occur in winter. However, under the in uence of hot and low pressure and the southwest monsoon, rainy and humid occur in summer. Because of the high altitude, the air is thin and the growth period of vegetation is short.
The entire Longbao NNR is grazed by livestock. Approximately 671 households with a population of 3263 people reside in the northwest of the reserve. The total number of livestock exceeds 28000. Large tracts of the main wetland have been divided into individual household pasture allotments, and since 1997, these single-family allotments have been fenced off. A trunk road (S308) linking Yushu and Zhiduo County was built through the reserve periphery. Animal husbandry is the leading industry and the main source of local nance.
BNCs likely arrive at Longbao at the end of March from wintering areas and depart at the end of October each year. Crane pairs prefer to nest in mounds (or small islands) in shallow-water wetlands with low vegetation and use the deep mud layers near their nest to escape from predators or to prevent the nest from being ooded. They feed mainly on wetland plants and some small animals, such as sh and shrimp. BNCs usually breed one nest and lay 1-2 eggs at the end of April to mid-June annually. After incubation for 31-33 days, the chicks break out from early June to mid-July.

Collecting BNC numbers
As the Longbao wetland is located in the remote hinterland of the plateau, observations about BNC and its habitats in Longbao are lacking or scarce. Few BNCs were observed before the establishment of the reserve. After the reserve was proposed in 1986, early statistics on the BNC population were not entirely accurate because of the di cult conditions of the reserve. Therefore, we collected the number of BNCs for 1986-2016 from the Longbao NNR and related literature.  Fig. 2(a)). We also surveyed the roosting sites in the study area from May 29 to June 1 in 2018. The nesting locations are mainly in the swamp area, which is di cult to reach. We used binoculars to observe the nesting environment of BNCs. The geographic position of the observation point and the azimuth of the nest site were recorded. We also measured the distance between the BNC nests and observation points through visual measurements. We observed 9 BNC nests in 2018 ( Fig. 2(b)).

Nest sites observation and microhabitats analysis
According to the nesting environment, the following six site characteristics were recorded to determine their in uence on nest-site selection: elevation (m), habitat type (lake island, marsh meadow or alpine meadow), distance to roads (far 1000 m and close 1000 m), shortest distance to nearest water body, nearest distance between nests, and disturbances (high: livestock, humans or dogs can approach the nest; or minor: livestock cannot approach the nest site due to a water barrier). ArcGIS software was used to locate the observation points on remote sensing images, and we measured the nesting microhabitat factors, including elevation, distance to roads, shortest distance to nearest water body and the nearest distance between nests. We obtained the six site factors of the 29 nesting locations observed by

Land cover change in the reserve
The BNCs mainly inhabit the core marsh area in the Longbao wetland. The land cover variation in the Longbao wetland represents the change in habitat for BNC. We selected and classi ed two Landsat images that covered the Longbao wetland with abundant rainfall and lush vegetation. Then, the changes in the ecological environment in the Longbao wetland since the establishment of the NNR were analyzed.
The whole Longbao NNR is divided into a core area, buffer zone and experimental area (Fig. 3), and the edge of the experimental area just overlaps with the Zhiduo highway. To comprehensively analyze the environmental changes throughout the whole area, we used the edge of the protected area as a 1 km external buffer for the study area. Through eld observations and investigations in 2018, the study areas were divided into eight categories: roads, rivers, lakes, buildings, alpine tundra, alpine meadow, marsh meadow and bare land. We interpreted roads and rivers rst as long and thin. Then, we selected the training samples and classi ed them by the support vector machine algorithm in ENVI.

Climate change and statistical analysis
Climate change affects the reproduction of BNCs and vegetation growth in the Longbao wetland. The meteorological data from 1986 to 2016 for Yushu station were obtained from the China meteorological data network, including the 20-20 hour daily precipitation and mean temperature. Then, we analyzed the change trend of meteorological factors at Yushu station. We investigated and collected the number of BNCs in the Longbao NNR over the past 40 years (Table 1). Yao (1982)  The nearest distance from the nest to the road was 400-2500 m, and the nearest distance to the water surface was no less than 800 m. The disturbance degrees ranging from strong to weak were alpine meadows, marsh meadows and lake islands.

BNC numbers
We observed 9 BNC nests with 5 island nests and 4 grass nests by telescope on May 30 th and 31 st , 2018. Three grass nests were located in the same place as in 2011. The main nest was located in the mainland of the wetland. Some nests were located in a slightly dry area in the eastern region. Island nests were distributed on islands in open waters or islands with star-like distributions. The typical features of the island nests were that they were well concealed and less disturbed nests occurred where people and animals could not easily approach. Several island nests were almost entirely round and 2-20 m from the water surface. BNCs prefer to nest on isolated islands surrounded by water. If the water level rose too much, the island nest would be submerged. This nest location is similar to the nest location in the northern Tibetan region observed by Kuang et al. (2010). The sixth island nest was black with a large radius and was slightly higher than the surrounding ground (Fig. 4). Grass nests were made with rich deposits in dry or lodging plants, were located in drier areas and had a stable shape for reuse (Fig. 5).
Nests located in the northwest of the wetland were distributed on both sides of the lake, and the nearest distances between them were maintained at 200-500 m. The nearest nest distances in the middle of the wetland were maintained at 300-1000 m. Nests in the eastern part of the wetland were distributed in the slightly dry meadows, with the nearest nest distances at 1000-2000 m. The distance between island nests was 180-550 m, and they were located near the open water; the distance between grass nests was 1000-1500 m, and they were located farther from the open water. The nearest distance from nests to the highway was 300-800 meters.

Land cover change
The classi cation results of the Longbao nature reserve are shown in Fig. 6 and Fig.7. From 1988 to 2015, marsh meadows, bare land and lakes in Longbao NNR showed a decreasing trend (Fig. 8(a)); alpine meadows, alpine tundra and rivers showed an increasing trend; and arti cial buildings, roads and houses showed an increasing trend. In particular, six new branch roads were added, which made the reserve more fragmented and disturbed the habitat of BNCs. The greatest loss in area was observed for marsh meadows, while the greatest increase in area was observed for alpine tundra. The average temperatures increased from 1978-2016, which melted the mountain snow and permafrost and washed away the mountains, thus leading to an increase in river area and widening of riverbeds. The image classi cation results on the west side of the reserve showed that the upper water area was large with less vegetation in 1988 but small with more vegetation in 2015.
In the core area of BNC habitat( Fig. 8(b)), marsh meadows, alpine meadows and bare elds decreased; alpine tundra, houses and other arti cial structures increased; and lakes remained stable. However, the number of marsh meadow patches in the core area increased mainly near the west outlet of the lake. Alpine tundra increased the most in the core area. In the buffer zone (Fig. 8(c)), alpine meadows increased the most and alpine tundra decreased the most. Marsh meadows, bare land, and lakes decreased, and rivers, roads, and buildings increased. The variation in the experimental area was consistent with that of the buffer zone. Bare land showed the maximum reduction in the 1 km buffer zone, while alpine meadows and buildings increased.

Increased in the number of BNCs
The number of BNCs has increased slowly from 24 to 216 between 1978 and 2016, with an average rate of approximately 5 per year. In addition, we performed two measurement campaigns during migration on October 28th, 2010 (118), and October 22nd, 2015 (125). We divided the BNC data into four periods (Fig. 9). During the rst ten years from 1978 to 1988, the number remained at no less than 30 and increased slowly, with an average rate of 0.4/year. In the second ten years from 1991 to 2000, the population grew from approximately 60 to 120, and the average rate was about 6/year.

Annual population changes in BNCs from April to November in Longbao
Every year in late March or early April, groups of BNCs y to Longbao from their wintering areas. In April, the population of BNCs was highest in this area. Some of the cranes occupied their territory and began to build nests, while others continued to migrate to other places as breeding habitats. As a result, the population of BNCs in the Longbao area tended to stabilize, and after nesting and breeding in May, the number of young cranes in June gradually increased and then decreased. Young cranes have to learn to nd prey, y and other skills before ying to their wintering areas within only a few months from birth to the end of October or early November. The nal survival of the crane is small due to the impact of wild dogs, prairie vultures and other natural enemies. In the breeding period of July 2011, Farrington et al.
(2013) observed a maximum of 43 young cranes and a minimum of 24 young cranes; however, only 7 young cranes ew back to Longbao in 2012. BNCs migrate to overwintering grounds in October; thus, the numbers gradually decrease. Therefore, the number of BNCs observed on October 28, 2010, and October 22, 2015, was less than that observed in other months.

Nest site microhabitats
A suitable nest site is the main factor affecting the reproductive success rate. According to the principal component analysis, two main principal components were extracted from six factors ( Table 2). The variance contribution rate of the rst component was 42%, and it mainly represented the nearest distance between nests, with more than 70% occurring within 500 m; the variance contribution rate of the second component was 27%, and it mainly represented the nest microhabitat type, with 69% built in the swamp meadow. The main factors affecting the nesting location of BNCs in the Longbao wetland were the distance between adjacent nests and habitat type.

Analysis Of Land Cover Change In Protected Areas
Although the population of BNCs has gradually increased and stabilized in the past 40 years, BNCs and their habitat are still facing severe challenges, including environmental change, human disturbance and climate uctuation.
First, the classi cation results showed that the vegetation area in the Longbao reserve increased while the grassland area was degraded. The alpine meadow and alpine tundra increased obviously, and alpine meadow patches shifted gradually into larger patches; however, marsh wetlands degenerated into alpine meadows, and marsh meadow patches became fragmented. In particular, marsh meadows and alpine meadows were reduced in the core area of the reserve that BNCs depend on for survival, which directly reduced the foraging and habitat sources of BNCs. The lake area decreased slightly, which was mainly because the lake area on the east side of the reserve decreased and became dry. Many groups of nonbreeding BNCs prefer to stay here and take breaks during migration . The amount of bare land decreased and changed to alpine tundra and meadows, especially on the southeastern slopes of the reserve, which shows that the establishment of the NNR has effectively protected the land surface and topography. The amount of water transported from the mountain snowmelt increased, which caused erosion and widening of the river channel. We found that the river ow was not large based on the eld observations at the end of May 2018, although almost all the elds beside the river were bare and a large number of yak and sheep were grazing in the area, thus leading to grassland degradation along the river bank.
Second, arti cial buildings, including buildings and roads in the reserve, increased and BNCs faced more human interference. The main reason for the increase in buildings was that Longbao Town had expanded signi cantly in the past 30 years. Moreover, victims of the Yushu earthquake in 2010 were relocated along Zhiduo Road (in the southeast with magenta color of Fig. 7), and some herdsmen settled down along Zhiduo Road and on the south bank of the reserve. The increase in buildings led to more surrounding human activities, which had a certain impact on the roosting reproduction of BNCs. The increase in road area was mainly due to the restoration and widening of S308 from Zhiduo to Yushu after the 2010 earthquake as well as the construction of new branches within the reserve, especially in Longbao town and east of the reserve. Wu (1993) found that cultivated elds increased, and woodlands decreased, from 1992 to 2006 through Landsat images in the Huize Nature Reserve, which is the southernmost reserve for wintering BNCs in China. Scott (1993) also found that habitat loss and disturbance were the main threats to the species.
Third, egg collection, feral dogs and intense grazing pressure represented important threats by Scott (1993). The Longbao reserve is nearly surrounded by pasture, and multiple fences are constructed except in the open lake area at the center; thus, grazing is extremely serious. In addition, rodent damage is serious in the reserve, and rat holes are especially dense. High voltage towers and telephone poles have also been added to the reserve. The construction of arti cial buildings has resulted in many surrounding grasslands becoming bare. We observed a Tibetan fox on the eastern branch road of the reserve and a wolf on the north side of the national highway of S308 in May. These animals are potential hazards to

Effects of climate change on the population of BNCs in Longbao
Climate change is an important factor affecting the breeding of BNCs. The average temperature of the Yushu meteorological station showed a slow rising trend since 1978 (Fig. 10). The average temperature remained at 4.56℃ during the decade from 2001 to 2010, which was 1.4℃ higher than the average temperature from 1978 to 1988 (Table 3), and then decreased slightly from 2011 to 2016. Rising temperatures lead to permafrost ablation and snow melting and affected the breeding and foraging of BNCs. There was no signi cant change in precipitation from 1978 to 2016 (Fig. 11). Precipitation showed a slow decreasing trend from 1978 to 1988 and the average precipitation was close to 500 mm (Table 3).  showed that the marsh wetland on which BNCs live decreased and the arti cial buildings, including roads and houses, increased, which affected the habitat of BNCs. The fragile plateau wetlands showed exacerbated degradation, and the connectivity of the wetland was reduced. In addition, grassland grazing and rodent damage are serious and encroach upon and destroy the grassland. Besides, the rising average temperature and precipitation over the past 40 years have also had an impact on the number of BNCs.
The carrying capacity of longbao NNR is limited. We suggest reducing grazing, removing fences and protecting wetlands to achieve a stable ecological balance. Conservation efforts should focus on the protection of existing lakes, minimizing disturbance impacts and maintaining small patches of swamp areas as important buffer zones. undertaken by Chou Xie.

Con icts of interest/Competing interests
The researcher claims no con icts of interests.
Ethics approval Not applicable.

Consent to participate
Not applicable.

Consent for publication
Not applicable.

Availability of data and material
The Landsat image can be down load from https://earthexplorer.usgs.gov/. Yushu weather station data can be downloaded from the China Meteorological data service center and the website is http://data.cma.cn/site/index.html. The other datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Code availability Not applicable.

Authors' contributions
Author contributions: Qiufang Wei, Yun Shao and Chou Xie designed the study and analysis. Baoshan Cui and Brian Brisco put forward guidance and revision opinions to the paper. Bangsen Tian and Kun Li participated in data collection and processing. Wenjia Tang observed the location of the nests. All authors contributed substantially to the data processing, writing and discussion of the paper.