NEST SITE SELECTION OF THE MAGPIE PICA PICA SERICEA IN BEIJING OF CHINA

Nest building and nest site selection of the Black-billed magpie Pica pica sericea were studied in Beijing, China, during 2004–2006. A total of 814 Magpie nests have been checked and measured. The results showed that the first nest building occurred at the end of December, but most pairs started in February. Mostly, a new nest was built for each nesting attempt; sometimes an old nest was renovated. Magpies mobbed potential predators very intensively during the breeding season, especially congeners. Magpies build nests in strong and tall trees (51% of all nests in Populus sp.). If there is no suitable tree for building, they also build their nests on man-made constructions, such as telegraph poles, water towers, etc. Nests were always located in the upper part of the crown, domed and with one or two entrances. Four types of habitat were identified in this study: urban residential area, urban park, suburban area, and exurb. Analysis of the factors that may affect nest site selection in different habitats indicated that the height of the nesting tree was the most important one. From the outer exurb to the city centre, with the increasing number of people and cars, the height of the trees Magpies chose for nesting increased accordingly. Magpies can find enough food and face less preda tion pressure in the urban area, which enables them to breed successfully in the urban environment of Beijing. Magpie, Pica pica , nest site selection, principal component analysis, breeding, Beijing.


INTRODUCTION
The nest site is the place where birds bring up their offspring during the breeding season. The quality of the nest site has a direct impact on the breeding activities and the fledging success, and it subsequently affects the population dynamics and the community constitution [3,4,12]. Some studies indicate that birds have behavioral mechanisms to select a specific spot as nest site. They prefer habitats that ensure a minimal different degrees of urbanization; they can be considered representative for the whole city of Beijing. The main tree species in the study area are Populus canadensis, Populus tomentosa, Salix babylonica, Paulownia tomentosa, Platanus acerifolia, Sophora japonica, Pinus tabulaeformis, Sabnia chinensis and Platycladus orientalis.

METHODS
We discovered that the timing of nest building is variable: the first activities start in December, and the latest occur in May. However, most Magpies build their nests between February and March. Therefore, we choose April and May to conduct the research on the nest sites, to begin a search for the nests. All the nests were located with GPS. Thirteen parameters were recorded: the height of the nest tree ([m], the distance from the ground to the top of the tree, eyeballing n.), the height of the nest ([m], the distance from the ground to the nest, eyeballing n.), the orientation of the nest (place of the nest in the tree), the amount of daylight (%, the percent of leaf cover over the nest, ISSN 1996-4536 • Біологічні Студії / Studia Biologica • 2010 • Том 4/№1 • С. 71-82 eyeballing n.), the species of the nesting tree, DBH (diameter at breasth hight [cm], diameter of the nest tree at the height of 1.2 m), the species of the nearest tree, the height of the nearest tree [m] (from the ground to the top of the tree), the number of trees within 10 meters of the nesting tree (counted with a tape measure), the number of other Magpie nests within 10 meters (counted as before), DBH of the nearest trees ([cm], diameter of the tree at the height of 1.2 m), the embranchment of the nest (eyeballing n.), Kf (the height of the nest/the height of the tree).
Principal Component Analysis was used to analyse the major factors that affect nest selection (Transformation with VARIMAX). We used One Way ANOVA to analyse four different types of districts and compared the differences between them. All data are analyzed by SPSS for Windows10.0.

RESULTS
A total of 814 Magpie nests have been recorded: 163 -in the urban residential area, 252 -in urban parks, 305 -in the suburbs and 94 -in the exurb. The pair members build the nest together, and usually build it in tall and strong trees (k=0.86±0.07). Sometimes they build nests on man-made constructions, such as telegraph poles, water towers, etc. More than 20 tree species were selected for nesting, Populus sp. being the most (51%) common in use ( Table 1). The nests are always in the middle of the tree and in the uppermost branches of the crown (at the very top). According to data in Table 2 and Table 3, the height of the nest trees and nests are strongly correlated to the first component, so we considered the first component to be the height factor. These data indicate that Magpies prefer to nest on high trees and particularly in the upper parts of the crown. The second component appears to be the tree factor and it contains the number of trees within 10 meters, the DBH of the nest tree and the DBH of the nearest tree. The factor of the number of trees within 10 meter is negatively related to the nest selection. The third component is named the light factor. The Magpie always build its nest on the top of the tree, where leaf cover is minimal. The fourth component is named the embranchment factor; the fifth one was called the nest density factor. It is positively related to the number of other nests within 10 meter, and negatively related to the height of the nearest tree (Table 3).  We accomplished this research in urban residential areas, urban parks, suburbs and the exurb. The difference between the exurb and the three other areas is significant.
Further analysis shows that all the factors in the exurb are significantly different from three other types of study areas, except for the height of the nearest tree ( Table 4). The difference between urban parks and the suburban area is not significant for the height of nest trees and nearest trees.

DISCUSSION
Breeding is the most important part of bird life, since a success of breeding can directly affect population dynamics and the continuity of species [63]. The research on breeding habitat selection contains three aspects: territory selection, nest site selection and fledging habitat selection. Nest site selection occurs on a small scale of vegetation structure, such as cover and height of vegetation and the field of vision [9,12,40,45]. Cody [14] suggests that microenvironmental fit, food supply and predator pressure are the most important factors that affect nest site selection in birds. The importance of each factor is determined by its effect on breeding success [14]. Bird nest site selection is optimized by building the nest in the habitiat that ensures the highest breeding success. Human population density and disturbance are also important factors that affect bird nest site selection [35,56]. Nest site selection is a protective breeding strategy during long-term evolution [2,13]; it can minimize the disturbance of congeners, predators and pests/diseases to improve breeding success [1,14,36,56].
In the study areas in Beijng, the earliest nest building started in December, while most nest building occurred between February and April, which is consistent with other Chinese records [20]. Sometimes Black-billed Magpies renovate an old nest, or build a new one on top of the old one. Some pairs build two or more nests and use one in the end. Male and female build the nest together. Magpies are territorial. They expel almost all intruding birds out of their territory; in addition, they chase raptors intensively. However they don't attack dickybirds, like Passer montanus. The reasons for that may be: 1) These birds are potential predators of eggs or chicks, thus can cause the Magpie's breeding attempt to fail; 2) Magpie's food requirements are largest when they are feeding young, and the intruders might compete for the Magpie's food; 3) Crows and some other raptors may usurp the nest 3, 15, 46, Corvids (Jackdaws also) can reduce Magpie's reproductive success [28]; 4) Raptors are potential predators to the young. In our study in Beijing, most Magpie nests were built in the middle of the tree, in the upper part of the crown (canopy). It is pantoscopic, the amount of sunshine is large. The embranchments under the nest are usually 3-5 to support the weight of nest. However, in other Magpie studies in Eurasian towns, most nests were higher still, in uppermost part of trees [5,19,29,32,39,49]. This difference may be a result of selection of other tree species, with another kind of canopy. Some authors describe a lower breeding success in lower nest sites [21,26,31]. This may explain why Magpies prefer long trees.
More than 20 tree species are commonly selected by Magpies for nesting; Populus canadensis (29.0%), Populus tomentosa (22.0%) are mostly preferred. Comparing the height and DBH of the nest tree with that of the nearest tree, the nest tree is significantly higher and thicker than the nearest one (p<0.01); the species of the nest tree and of the nearest tree do not differ significantly (p=0.056). Populus sp. was a common tree species planted in the XXth century in many towns of Eurasia. This provided many potential nesting places preferred by the Magpies. In Manchester, UK, about 38% nests were in Populus sp. [50], 44% nests in Zielona Góra, Poland, [32], and 49% in Cracow, Poland [5]. Many other authors describe that the Magpies often build nests in Populus sp.: in Bonn, Germany [57], Cuxhaven, Germany [38], Saratov, Russia [51], Voronezh, Russia [47], Berlin, Germany [37,58], Poznań, Poland [34], Cracow, Poland [27], Lublin, Poland [44] and Seoul, Korea [18].
Tall and strong buildings can be nest sites, too. It explains why the Magpies build nests on telegraph poles, water towers, etc., which was also observed in other parts of Eurasia [43,55].
The PCA-analysis of 10 factors of nest site selection shows that height is the most important component, the ratio of contribution being 25.62%. The Magpies are common in cities, but they are sensitive to human disturbance. Studies have shown that Magpies tend to build nests near the top of tall trees [19,32,49].
A second component is the tree factor. The Magpies prefer to build their nest in tall, strong trees, probably to make the nest safe and steady. A number of trees within 10 meters is negatively correlated with nest site selection; that means that Magpies like to build their nest in a separate tree. The Magpies are territorial; pairs will intensely drive conspecifics out of their territory. We observed an intruder being chased so fast that it crashed on the ground and died. Suitable nest sites are limited, so potential nest trees must be at some distance from other nesting pairs. Jerzak [32] found in Zielona Góra, Poland, a positive correlation between the distribution of the Magpie breeding pairs and tree density. Similar observations were done in Manchester, UK, by Tatner [50] and in Lvov, Ukraine by Bokotey [8]. Birkhead et al. [7] in Sheffield, UK, found also that the Magpie density was related to/depended on the distribution of suitable nesting sites (high trees). Such observations are important for landscape architectors and city government. Planting high trees in towns might favour the increase of a Magpie population. Similar observations were done by Charles [10] on Crows Corvus corone and C. cornix. Thus, we may easily facilitate a density decrease of these "conflict" species in our urban environment.
The third component is the extent of sunshine. Usually, it is high, because the nest is on the top of the tree. The nest has a roof that safely shades the eggs and chicks.
Comparing these factors within the four types of study area, the results show that the height of the nest tree and the nest gradually decrease from the urban area to the exurb. We found the lowest nest in the Fragrant Mountain, at 2 m, the highest in an urban area, at 18 m. In the urban areas, the human disturbance is higher, and to counteract this situation, the Magpies build nests on higher trees. This is consistent with the results of Górski and other investigators in Koszalin [26]. There are no significant differences between urban parks and suburban areas, but the significant difference between the exurb and the other three types of study areas is clear. This may be due to the different environment in the exurb, with less people and cars. Human influence in urban parks and suburban areas are similarly low, so the nest site selection factors are similar too.