Annual dynamics of bird communities in urban parks in Lviv, Ukraine: preliminary analysis of diversity and composition variability

© 2019 O. Dubovyk; Published by the Ivan Franko National University of Lviv on behalf of Біологічні Cтудії / Studia Biologica. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://www.budapestopenaccessinitiative.org/ and Creative Commons Attribution 4.0 License), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. UDC: 598.2:591.5](477.83-25-021)


INTRODUCTION
The consequences of urbanization processes on wildlife are known widely: economic growth in human society could be one of the reasons of species endangerment [11], habitat loss and fragmentation are thought to be the main reasons of biodiversity declining and populations extinction [7,13]. In those conditions species have two possible options -to avoid urban habitats or to accept new conditions [16]. Accepting new conditions, or synurbization sensu M. Luniak [4,25], provides the species with some benefits such as a milder microclimate, food, places to hide etc. (as well as dangers such as heavy metal poisoning [36], stress [26] and behavioral changes [32]). Thus, this phenomenon is common for many animal species.
Bird's diversity and composition are also being affected by changes caused by urbanization [5,10,19,27]. Furthermore, considering their abundance, mobility and species richness, birds are appropriate indicators of changes in ecosystems, animal communities or landscapes [14,20,21]. This reveals why birds are common objects in ecological studies, especially in urban habitats. From the view of avifauna, it is impossible to consider the city as a single whole, and in most research the city is considered as a complexity of different habitat types [2,6,9] that makes synurbization processes easier to study in birds.
In general, habitat diversity of city urban greenery, e.g. urban parks and forests, is one of the most stable and productive in ecological meaning habitat type, which is, in fact, an "island" of wildlife in the city. It depends on origin, and can be actually a residual of native ecosystems [30]. Some authors assume that large urban parks could be a buffer for penetration of wild species into the city [37], but this question hasn't been studied enough. Nonetheless, it is known that bird diversity is high compared to other habitat types in urban parks [8,28,38,40], and changes in the most diverse communities are the most informative [35], which makes urban greenery the best habitat type to study.
While spatial distribution and dynamics of bird communities in urban parks is studied well, less attention is paid to its time dynamics [24]. Bird communities are variable because of variation in species biology features and their migration status [17]. Birdhabitat relationships are unstable during the year a breeding and non-breeding periods [29]. Also migrant species contribute to general seasonal variability and birds of some species wander during the non-breeding season which means changes in their habitat preferences during the year without visible effects on species abundance on a local scale. Annual habitat preference or migration cycles affect the composition of a bird community in places with ecological consequences; it is more important to understand it in the context of possible effects of urban heat island as a part of urbanization process on birds phenological features [41].
Summarizing, information about seasonal dynamics of bird communities supplements fields of avian biology and general ecology common with topics of urban planning [24], predicting changes of bird diversity and composition caused by anthropogenic landscape transformation, which could indicate about possible future changes in other taxonomic groups and whole ecosystems, biogeographical changes and synurbization mechanisms. Also, annual dynamics, stability of community structure and dynamics of diversity level show the stability of ecosystems and primary productivity [23], according to the species-energy hypothesis [12]. This can show possible deficiency in urban greenery management -a problem that is being acutely felt in the Ukrainian cities.
The aim of this work was to reveal general tendencies of the dynamics of bird communities in urban parks, i.e. to provide a descriptive study of this phenomenon. Primarily, it was decided not to look at avian community in a discrete way, i.e. during breeding or non-breeding season, but to concern the composition dynamics as a continual process as in [32].The objectives were (1) to reveal the species composition and community structure of birds in urban greenery during throughout the year, (2) to define the time of peaks of diversity, abundance of birds and stability of the whole community, (3) to study changes in ecological (viz. trophic) and biogeographical structure of bird communities, and (4) to reveal the abundance dynamics of most common forest bird species.

Study area.
Bird surveys were conducted in Lviv city (Western Ukraine, 49°49′N 24°00′E). Lviv is a city located on the main European watershed between the East European Plain and the North European Plain. The total population is approx. 725 thousands inhabitants and the area of administrative boundaries equals 182 km 2 . The area of urban greenery, according to the city council, is 4419 ha (24 % of total city area).
Study plots (n = 8) were chosen in order to represent the whole parks and forests on the area of the city. Linear transects (in sum, n t = 20) each of 500 m length and 100 m width were paved, the number of those per plot depended on the total area of model plot.
Holoskivsky cemetery is allocated in the northern suburbs, A = 80.3 ha, n t = 3. In fact, the cemetery is a part of the Briukhovitsky forest massif (A = 3200 ha) -typical Roztocze forest, which has undergone strong anthropogenic transformation. Pinus sylvestris, Thuja occidentalis, Betula pendula and Acer platanoides are the most common tree species on studied transects.
Yanivsky cemetery is located in the northwestern part of the city, and it is an old cemetery. A = 30 ha, n t = 2. Acer platanoides, Robinia pseudoacacia, Betula pendula, Thuja occidentalis and Fraxinus excelsior are the most common tree species.
Bilogorshcha forest is located in western suburbs near the mire and village-like part of the city. A = 195 ha, n t = 4. Alnus glutinosa, Quercus robur, Pinus sylvestris, Betula pendula and Carpinus betulus are the most common. There are one derelict pond and a stream on that area.
Lychakivsky park and Lychakivsky cemetery are located in the eastern part of city, both are old green areas. A = 10 ha and A = 34 ha, respectively; n t = 3. Acer platanoides, Fraxinus excelsior, Thuja occidentalis and Pinus sylvestris are the most common.
Pohulianka forest park is adjacent to the Lychakivsky park and stretches east. A = 108 ha, n t = 3. Fagus sylvatica, Carpinus betulus, Acer sp., Picea abies and Alnus glutinosa are the most common.
Stryisky park is an old park in the central part of city. A = 52 ha, n t = 3. Acer sp., Tilia cordata, Fraxinus excelsior, Fagus sylvatica, Carpinus betulus and some exotic species, such as Salix babylonica, Ginkgo biloba and Mahonia aquifolium, form the tree community. There is one big decorative pond "Pond with swans" on the area.

RESULTS AND DISCUSSION
79 bird species were registered during one year in urban greenery in Lviv. Those taxa are listed in the Supplement according to HBW and BirdLife International taxonomy [15]. For example, the number of species that breed in parks and forests in Lviv (but only within the ecological merges) was assessed as 61 in 1994-1995 and the number of wintering species was 38; in sum, 67 species was registered in this study period by the team of ornithologists (described in PhD thesis of A. Bokotey, 1998). There are some differences in the species list caused by strong declining of some species in the city (for example, Passer domesticus and Streptopelia turtur were observed in the 1990s), rare observations of the species (Picus viridis is rare and is being under protection of the National Red List), some species was not observed during this study due to probabilistic reasons (Athene noctua and Asio otus -because special night censuses were not conducted; Anthus trivialis, Emberiza schoeniclus). Some species that were not mentioned by Bokotey were observed in this study, mainly because it concerned the periods of migrations (but other reasons, such as changes in populations or differences of study plots, can be also considered): Accipiter gentilis, Buteo  Species richness summarized on all transects is the lowest during the non-breeding period: it changes from 38 species registered a month in July to 30 species in August and stays on this level till the end of winter (28 in September, 30 in October, 27 in November, 31 in December, 30 in January, 27 in February). The breeding season is the period of high species richness among birds in urban greenery: 39 species were registered in March, 44 -in April, 43 -in May, 45 -in June and 38 -in July (see not summarized data on Fig. 1).
Averaged by month, all three parameters that can assess diversity level -sum of individuals of all species (which does not actually mean the diversity level), species richness and Simpson diversity index -indicate that the peak of diversity level matches with the beginning of breeding season in that latitude -end of March (Σn ind. = 40.95±17.14 SD ind./5 ha, Simpson 1-D = 0.87±0.04, n sp = 12.95±2.76) and whole April (Σn ind. = 79.1±21.58, Simpson 1-D = 0.90±0.02, n sp = 16.40±2.78). It can be concluded that urban greenery is the more preferred biotope type by some bird species for breeding than for residence during the non-breeding season. I presume that this phenomenon is connected with some general biotope characteristics, i.e. abundance of safe places for nesting, than with food resource preferences, insofar as if the second option was true, the species richness would be saved during the whole vegetation season. Besides that, the peak of the sum of individuals can be explained by fledglings, but it is uncertain considering the data gathering methods.

Fig. 2.
Between-months, bird community stability was assessed with paired comparison similarity indices: species composition stability was assessed by Jaccard index and community structure stability was assessed by Bray-Curtis index (the higher it is, the more stable communities are between months) Рис. 2. Стабільність угруповань птахів між місяцями оцінено за допомогою індексів подібності попарних порівнянь (бінарних мір подібності): стабільність видового складу відображає індекс Жаккара, стабільність чисельної структури угруповань відображає індекс Брея-Кьортіса (чим більші значення індексів, тим стабільніші угруповання між роками) A similar situation was observed in community structure comparisons: significant step in community structure stability during the spring part of the non-breeding season and breeding season beginning (for BC Mar/Apr and BC Apr/May t = -6.25, p-value = 2.95·10 -7 ) is accompanied by a descending of stability in June-July (for BC May/Jun and BC Jun/Jul t = 7.52, p-value = 1.32·10 -8 ). It is important that the stability value between months has two main contributing components: real variability of the community composition and probability factor (as at one linear transect one census per month was performed, all birds which were really residents at the moment were not observed). Notable, that bymonths comparisons look smooth, i.e. there are no significant dissimilarities in community composition during long period, although there does exist strong dissimilarities between community composition in long perspective. Table 1.
Mean values of the Jaccard species composition similarity index and the Bray-Curtis community structure similarity index assessed for bird communities be tween months Similarities between communities in different months decreases with difference in time (∆τ = [0; 6] months). Both Jaccard index values (regression equation is J = 0.603-0.077·∆τ, p-value < 0.005, R 2 = 0.360, R 2 adj = 0.359) and Bray-Curtis index values (regression equation is BC = 0.647-0.078·∆τ, p-value < 0.005, R 2 = 0.353, R 2 adj = 0.359) are related with months difference. Thus, it can be assumed that dynamical similarity is a function of difference in time, inflection point is approximately in ∆τ = 3 (R 2 values for models with ∆τ> 3 omitted are R 2 = 0.493 for Jaccard index and R 2 = 0.500 for Bray-Curtis index).
European zoogeographical fauna type is most common for species of birds that are breeding and residents for urban greenery in Lviv (Table 2). Absolute number of individuals of European species varies from 18.85±10.26 ind./5 ha in October to 69.4±15.23 ind./5 ha in April; generally, European species are the most abundant during the breeding season, but their abundance is 2-4 times lower during the non-breeding season (August-March). Anyway, those species are the most abundant in all months.
Mediterranean species are common during the non-breeding season (June-March, abundance varies from 4.50±3.32 ind./5 ha in June to 23.20±22.75 ind./5 ha in December), but unreliable peak in April (from 4.60±4.62 ind./5 ha in March to 12.75±21.5 ind./5 ha in April, Welch Two sample t-test t = -0.599, p-value = 0.555) shows the breeding occasions among those species. Only three Mediterranean species were registered: not common in urban greenery in Lviv Dendrocopos syriacus, Serinus serinus and extremely distributed Columba livia f. domestica. Although abundance of Mediterranean species is high, this data is unrepresentative.  An abundance of Siberian species is larger in the autumn and winter period than in the breeding period (10.00±11.53 ind./5 ha in September and 8.57±13.21 ind./5 ha in December compared with 5.00±3.49 ind./5 ha in April -the peak value in the breeding period). However, the total occurrence of Siberian species is low both in the breeding season (20-55 %) and the non-breeding season (15-55 %).
Paying attention to occurrence distribution, European species are ubiquitous in all months, Transpalearctic species are spread too. The rarest fauna type are Mongolian (includes Phoenicurus ochruros only) and type of not defined origin (includes Streptopelia decaocto only -this species was observed in urban greenery only after the departure of migrant Columba palumbus, the species that recently became typical synanthropic in Lviv).
Invertebrate-eaters are the dominant trophic group among birds in urban greenery (see Table 3 and Fig. 3), their abundance is the largest in April (firstly, because the general birds abundance is the largest, secondly because the specific diet of the vast majority of songbirds in the period of feeding chickens). Obviously, the abundance of invertebrate-eaters is low in the winter period, only Paridae and Piciformes are common in this time. Seed-eaters are the subdominant group, the abundance of whose correlates slightly with the abundance of invertebrate-eaters (r = 0.364, p = 6.08·10 -9 ; Spearman rank correlation r = 0.371, p = 3.05·10 -9 ). Vegetative parts of plants are concerned as food resources by birds in the afterbreeding season with a strong rise in autumn (abundance goes up from 2 ind./5 ha in July to 61 ind./5 ha in November, but the occurrence of species with this diet -Anas platyrchynchos -is low, just 5-10 %; Anas platyrchynchos, obviously, was registered only on plots with ponds). Abundance of vertebrate-eaters, or raptors -Accipiter nisus, Accipiter gentilis, Buteo buteo, Strix aluco, Strix uralensis (important that owls abun-dance was not the main object of this research as this requires special censuses; registrations of those two species are accidental and their occurrence is underestimated) and partial raptor Corvus corax -is low and stable through the year: 1-2 ind./5 ha, CV of means = 0.249. Occurrence of raptors is low -10-45 % and do not depends on season (ANOVA F(2, 9) = 1.534, p-value = 0.267 for groups of March-June, August-November and December-February).
The presentation of data on trophic demands structure would be more informative in relative scale, i.e. parts of individuals of defined ecological groups' species in the community, but this meets a deficiency of available classification -groups' intersection, which means that the sum of percentages of different groups would be greater than 100 % calculated in a certain community. The fact that used classification was developed in the Siberian region also means that it could be inaccurate for individuals of the population present in Ukraine, thus, ecological classification of Ukrainian birds' species should be reviewed. Table 3. Seasons classified as winter -December, January and February, Spring -March, Breeding -April, May, June and July, autumn -August, September, October and November, affects densities of several species, as results of ANOVA tests show.

Mean abundance (n±SD, ind./5 ha) and total occurrence (%) of species of differ ent trophic groups, by food resources type Таблиця 3. Усереднена сумарна щільність населення (n±SD, ос./5 ha) та загальна час тота трапляння (%) видів різних трофічних груп за типом корму, завдяки споживанню якого задовольняються енергетичні потреби
Thus, season affects (in terms of ANOVA results) individuals density in common migrant songbirds species such as  For some species this variability can be explained by changing biotopes preferences, which should be proven by the results of special research. For example, the abundance of Turdus merula in April is 10.4±4.28 ind./ 5 ha and 11.1±4.18 ind./ 5 ha in May, but 2.69±2.06 in December; moreover, the occurrence of this species is 100 % during the breeding season, but it is 50-60 % in winter. Further statistical analysis is needed to make specific conclusions about common synurbanized forest bird species ecology dynamics.

CONCLUSIONS
79 birds' species were found to occur in urban greenery in Lviv throughout the year. The highest diversity level and abundance of birds is observed during the breeding period, especially in April, when many species begin breeding. A slight peak of abundance is observed during winter season, inasmuch as urban habitats afford special conditions, such as food resources and softer weather, to outlive this period successfully. Nonetheless, in this particular case the diversity of birds in urban greenery declines during the non-breeding season. This could mean that urban greenery is the more preferred biotope type by some bird species for breeding than for residence during the non-breeding season.
Between-months, community stability among birds in urban greenery is higher both for presence-absence species composition and numerical structure during the breeding season that shows that most of the birds are highly attached to their nesting sites, especially in April, May and June. Similarities between communities in different months decrease with difference in time.
European bird species make a nucleus of community structures in all months. Siberian, Mediterranean and Transpalearctic species are also common in urban greenery. Their role becomes more important during the non-breeding season, but separately by types all of them subdominate compared with European.
Invertebrates are the main food resources for most individual birds. Their role is especially important for birds of urban greenery while feeding nestlings; naturally, in cold months, the majority of birds predominantly use other food resources, particularly seeds. The abundance of raptor species is low and stable in all months.
Season can be considered as a factor related with the abundance of several species. Thus, bird species that are typical for urban greenery in Lviv can de divided into three groups.

ACKNOWLEDGEMENTS
Author is grateful to his father Dmytro and wants to dedicate this article to the blessed memory of him.
The support of supervisor Ihor Shydlovskyy was considerable for writing this paper. Financial support for this research was provided by the Dovkolabotanika community.
The critical reviews of Tomasz Mazgajski and anonymous peer reviewer were helpful in defining mistakes and inaccuracies in this work.
Special thanks for Benjamin Jackson for eliminating grammatical and stylistic mistakes, and Nadya Chepelevska and Vitalii Goncharenko for help with defining species composition of trees on model plots.