Urban avifauna distribution explained by road noise in an Eastern European city

Alfred-Ștefn Ccort-Luccu,Hnem-Ver Kesht,Pul-Vnd Popovc,Dvd Munk′cs,Ile-Cˇtˇln Telcen,Crmen Gche

a Department of Biology,Faculty of Informatics and Sciences,University of Oradea,1 Universitˇații Str.,410087,Oradea,Bihor County,Romania

b Independent Researcher,2 Alexandru Roman Str.,415200,Beiuș,Bihor County,Romania

c“înv.G^algˇau Iosif”Secondary School,61 Pagaia,417063,Boianu Mare,Bihor County,Romania

d“Miskolczy Karoly”Secondary School,197 Mișca,417183,Chișlaz,Bihor County,Romania

e“Kazinczy Ferenc”Secondary School,602 Simian,417545,Simian,Bihor County,Romania

f“Toldy”Secondary School,26 S^annicolau de Munte,417441,Sˇacuieni,Bihor County,Romania

g“Mihail Eminescu”National College,18 Roman Ciorogariu Str.,410017,Oradea,Bihor County,Romania

h Technological High School No.1,48 de Jos Str.,Cadea,417436,Sˇacuieni,Bihor County,Romania

i Department of Biology,Faculty of Biology,“Alexandru Ioan Cuza”University,11A Bd.Carol I Str.,700505,Iași,Iași County,Romania

Keywords:Road noise Urban green spaces Urbanization gradient

ABSTRACT In Eastern Europe,efforts are being made for road infrastructure development,as this area has not yet reached the level of that in the West.Road noise negatively affects birds,especially in urban habitats.Therefore,seven green spaces from an Eastern European city were monitored,which were selected according to the urbanization gradient and the hydrographic peculiarities of the area.The correlation between species richness and human population density,urban mobility,or road noise was verified.These three factors together explained over 99%of species richness distribution in habitats and during study periods.Road noise was the strongest predictor and the most negative influencer of bird species.These correlations were higher when we also considered the variability of species in habitats over time,instead of the total species richness of a habitat,as it expresses more accurately the impact of noise on the presence of species.The presence of species has decreased from one habitat to another along the urbanization gradient,from the edge of the city to the center,both in the case of terrestrial and aquatic green spaces.The most central park differed significantly from the other terrestrial green spaces in the city.At the same time,no significant differences were registered between the aquatic habitats or the rest of the terrestrial habitats.The ideal mitigation measure in this case would be that there are no roads in the immediate vicinity of green spaces.Still,if this is not possible(as in the case of all the existing green spaces in a city),we discuss/recommend placing noise barriers at the edge of green spaces or high-speed roads.

1.Introduction

Land use change is one of the most important factors that negatively affect global biodiversity and is expected to become the main problem for terrestrial ecosystems in the future(Sala et al.,2000).Urbanization is considered the most radical way of land use change(Deng et al.,2009;Vasenev et al.,2019).Starting with land cover change,climatic(Dale,1997;Kalnay and Cai,2003)or hydrological(Carlson and Arthur,2000;Jeong,2001)imbalances are generated,which are often felt at the regional level,not just locally(Berry,2008).Studies on the urban-rural gradient analyze the changes of biota along a transect from the city center to the surrounding,less disturbed ecosystems,following either a habitat-loss gradient or various physical changes closely related to natural habitat losses(McKinney,2002).

For cities,parks represent true biodiversity hotspots,where the relatively large variety of habitats and availability of resources increase birds’diversity(Fernndez-Juricic and Kokimaki,2001).However,birds are negatively affected by several human interventions in urban green spaces(Luniak,2019;Bernat-Ponce et al.,2020;Tashakkor et al.,2020).Additionally,the general particularities of urban habitats influence urban birds,such as the structure of built surfaces(Shimazaki et al.,2016;Batisteli et al.,2021),noise(Merrall and Evans,2020),night light(Aulsebrook et al.,2020;Jiang et al.,2020)or other threats,which often lead to the death of some individuals(Francis et al.,2020;Lao et al.,2020).Also,we must consider the very high level of human presence specific to city environments,which every wild animal in urban habitats is forced to live with.For many birds,the direct presence of man is a strong disruptive factor,especially since some urban green spaces,such as parks,are specifically designed for human recreational activities(Fern′andez-Juricic and Kokimaki,2001).This is one of the main reasons why many bird species have developed different adaptive strategies to human presence,proving high behavioral plasticity,such as the decrease of alert distance,flight-initiation distance,or the distance at which birds fled(Stansell et al.,2022;de Carvalho and de Toledo,2021),forage in interspecific flocks(de Carvalho and de Toledo,2021),higher propensity to innovate(Ducatez et al.,2020)and others.Some birds can even distinguish human clothing,which they associate with various characteristic human activities and implicitly with varying degrees of danger(Feng and Liang,2020).

Human population size negatively affects urban birds(Leveau,2021).However,exceptionally,human population density can be positively related to some bird species highly specialized for urban environments(Jokim¨aki and Suhonen,1998).Unlike this,on a small spatial scale,when comparing different areas of the same city,human population density is not related to bird species richness(de Camargo-Barbosa et al.,2020).Noise levels and tree species richness explained better the variation in bird species richness than other environmental variables,such as human population density,the size of green spaces,and distance from a natural forest or a body of water(de Camargo-Barbosa et al.,2020;da Silva et al.,2021).

The road infrastructure in Eastern Europe has not yet reached the level of development like in the rest of Europe(Stelder,2014;Strano et al.,2017).Solutions are being sought for Romania in this regard(Lucian,2015),and some are being implemented(Man et al.,2015;Ib′a~nez and Rotoli,2017).For the modernization of the urban road network,the administrations of big cities carry out studies on urban traffic,considering sustainable urban mobility(e.g.,INCERTRANS,2016).Usually,the plans for these studies are published online and provide detailed information on human population density in each city area and residents’travel from one part of the city to another.At the same time,local administrations make road noise maps according to the regulations of the Environmental Noise Directive(EP and CEP,2002),which are later considered in urban planning.Based on such documents and considering that birds are an essential component of urban biodiversity(McKinney,2002)and that they are affected by human presence and urban noise,we decided to verify if there is a correlation between human population density,urban mobility or road noise and bird species richness in an Eastern European city(Oradea,Romania).Also,we wanted to estimate to what extent all these factors can be predictors for urban birds.This information could be important for wildlife conservation strategies within cities.

2.Methods

2.1.Study area

The city of Oradea is located between the Pannonian Plain and the hills bordering the Apuseni Mountains,and it is crossed by the Crișul Repede River(Ilieșet al.,2007).The region was once covered by forest-steppe(Erd″os et al.,2017)but has lost over 96% of the forested area in the last 300 years(Pop,1968).One of the remaining forests here is the Nojorid Forest(Mˇahˇara,2010),the only urban forest within Oradea.According to the European classification,it is a Danubian-Balkanic Turkey Oak(Quercus cerris)forest with Festuca heterophylla(RNP-ICAS,1997;according to Donițˇa et al.,2005).The management of the forest over time has led to a radical change in its composition from pure Turkey Oak(Q.cerris)forest into a forest with over 86%Common Oak(Q.robur)(RNP-ICAS,1997;OMNI,2014;VI(V^arful^Inalt,2017).

Another essential feature of the area is the Pețea Creek(Ilieșet al.,2015),which never freezes due to its geothermal springs(Cosma,1977).However,the thermal groundwater overexploitation has led to a gradual anthropogenic degradation,leading to severe alteration of the entire ecosystem,which has resulted in the extinction of some endemic species and the endangerment of the priority habitat 31A0*Transylvanian hot-spring lotus beds(Cohut and Pa′al,2017;Linc et al.,2019).The Pet¸ea Lake,from which the Pețea brook springs,dried up in 2012(Telcean and Cups¸a,2013;Vicașet al.,2013),but the thermal character of the creek has not been completely lost due to the discharges of the thermal water obtained through drilling from the hotels and climatic hospitals in the area(Cohut and Pa′al,2017).

Until 1989 in Oradea,there was a continuum of woody vegetation,corresponding quite well to the land-sharing development strategy in the literature(Ib′a~nez-′Alamo et al.,2020;Jokim¨aki et al.,2020).The city's former green spaces have been severely degraded and reduced in the last decades.The small green spaces between the buildings were turned into parking lots,and the percentage of the city's built area has increased,an urban phenomenon referred to in other studies as buildings'“crowding policy”(Coroiu and Radulovici,2002).

2.2.Data on urban avifauna

During the years 2018–2020,we investigated seven urban green spaces in the city of Oradea,western Romania.Each habitat was investigated once every two weeks;thus,in a monitoring interval of one year,we made 26 field trips to each habitat.Each green space was investigated in the morning for an hour and a half.During the field observations,we only noted the identified species,without the number of individuals of each species,as our interest was to verify the distribution of the species by habitats and periods.

When choosing the green spaces for the study,we considered as follows:1)Identification and delimitation of the area mostly affected by urbanization due to the lack of vegetation or pedestrian and road congestion;in our case,it was the city center;2)Delimitation of the zone on the city's outskirts where the surfaces are covered by constructions(residential or industrial)alternating with agricultural land;3)Identification of the largest green spaces in the city,if possible,an arboreal habitat and an aquatic habitat(riverine or lacustrine).In our case,these were the Nojorid Forest on the outskirts of the city and the green corridor of the Crișul Repede river that crosses the city(see Fig.1);4)Choosing an imaginary line which connects the city center with the main green spaces of the city and its outskirt;5)Identifying the green spaces on the established alignment and choosing only one green space from each habitat type(forest or aquatic)in each of the three zones of the city(hyperurbanized,urban,and suburban);6)In the case of riverine habitats,which cross all three city zones,one river sector was chosen in each zone,both upstream and downstream of the hyperurbanized zone;7)Urban habitats with no free access inside or in their immediate vicinity by road and pedestrian traffic were not studied.This was the case of the Crișul Repede river,which in the suburban zone has up to 200–300 m shore with restricted access due to the upstream city water supply system and the downstream wastewater treatment system.

Each studied arboreal green space has a small aquatic habitat,represented by artificial or semi-natural pools.Also,along each studied river sector,there are more or less natural riparian forest buffers and a park.Both banks of the smaller river(Pețea Creek)were monitored simultaneously.This was not possible in the case of the larger river(Crișul Repede);therefore,the observations were made only on the right bank and on the water surface.Even though the surfaces of the seven green spaces are different,the total length of the transects made in each of them was about 3 km.

Domestic birds such as the Domestic Pigeon(Columba livia domestica,which had a constant presence in all the studied habitats)or the Domestic Duck(Anas platyrhynchos domesticus,which escaped from the households and sometimes accompanied wild specimens of the same species)were not included in the list of the species.On the banks of the Pet¸a Creek and at the edge of the forest,birds from neighboring households,such as chickens,ducks,or geese were often encountered.

Fig.1.The areas of Oradea city by the predominant types of buildings(Z1:the city's downtown zone with historic buildings that have a height regime of 1–3 levels and an institutional or residential destination.Z2:the urban zone with houses of 1–2 levels and blocks of 4–10 floors that generally have a residential role.Z3:the suburban zone with houses of 1–2 levels and agricultural lands.S1–S7:the studied urban green spaces.S1:Nojorid Forest;S2:Rulikowski Cemetery;S3:Pet¸a Creek(R2);S4:1 Decembrie Park;S5–S7:Crișul Repede River(R1).BG:Republic of Bulgaria;BS:Black Sea;HU:Hungary;MD:Republic of Moldova;RO:Romania;RS:Republic of Serbia;UA:Ukraine).

The bird species have been classified in the literature according to their degree of tolerance to urbanization in several ways.An important classification of birds by their adaptability to urbanization was based on the average daily densities of species(Blair,1996).An alternative classification was based on the birds'diet(Jokim¨aki and Suhonen,1998).Later McKinney(2002)extended Blair's classification to plants and other groups of animals and tried to correlate it to the ecological peculiarities of the species,using in the case of birds their diet,as Jokim¨aki and Suhonen(1998).However,Adams(2016)showed that the bird species in Blair's categories do not have a fixed status,as we would expect due to their diet,because they sometimes jump from one category to another due to their different densities from one case to another.In our case,because we did not have quantitative data for density calculations,we applied the principle stated by Blair(1996),namely that the abundance and distribution of birds are related to habitat changes due to urbanization.Therefore,we classified bird species based on qualitative data obtained from species distribution.Thus,the species common for all studied green spaces were considered to have a high degree of tolerance;with a low degree of tolerance were the species identified only in the forest on the outskirts of the city or those identified only once in the city,in any of the other habitats.The rest of the species were considered to have an average degree of tolerance.

2.3.Data on the human factor

Information on urban density and mobility was obtained from the Sustainable Urban Mobility Plan,developed by the Transport Research Institute INCERTRANS for the local government of Oradea City(INCERTRANS,2016).The plan shows the city's number of residents per hectare in each of its 30 neighborhoods.In the same plan,the city was divided into 60 territorial subunits,according to the criteria of the arterial road,to achieve two origin-destination matrices,corresponding to the two time-intervals of the day,when the road traffic in the city is the busiest.The method they used follows the methodology established by the Center for Road Technical Studies and Informatics CESTRIN,which belongs to the National Company of Road Infrastructure Management(CNADNR,2014).The matrices provide for each territorial subunit several trips declared by the city's inhabitants from the other territorial subunits.The statements were taken during home surveys and contain details of each trip,such as origin,destination,duration,purpose,mode of transport used,etc.

The two data sets were converted to be statistically processed in the present study.When a studied green space was located at the border between several neighborhoods,we considered important to take into account data from all the surrounding neighborhoods,even if each studied urban green space belongs territorially to only one of the neighborhoods.In such cases,we measured from the perimeter of the green space the segment corresponding to each of the surrounding neighborhoods,then calculated a weighted average number of residents for each of them(see“residents”in Appendix 1).Of the two origindestination matrices,we used only the one corresponding to the morning hours because our observations of the birds were made in the first part of the day,not in the evening.The morning matrix sums up to 58,507 trips,which we considered travelers because for birds in urban green spaces,if two trips are made by the same person or by two distinct people,is irrelevant.The number of travelers of each territorial subunit was reported to the surface for establishing a density as in the case of the residents of the neighborhoods.In the case of urban green spaces bordered by several territorial subunits,we used the same weighted average method(see“travelers”in Appendix 1).

Road noise levels were taken from the Road Noise Map of Oradea(EC,2017).This was made according to the Environmental Noise Europen Directive(EP and CEP,2002)and Guide regarding the creation,analysis,and evaluation of strategic noise maps,published by the Ministry of the Environment and Sustainable Development(MMDD,2007).The main type of map is the one that shows the values of the Lden(day-evening--night level)parameter,which expresses the average value of the noise during a whole day(24 h).The sound measurements were made 4 m above the ground,and the grid was 10 m(EC,2017).For each urban green space,we summed up the noise intervals corresponding to the colors on the map of that space.The resulting mid-range noise was introduced into the statistical calculations(see“noise”in Appendix 1).In the case of the studied river sectors,we did not consider the noise from the bridges,which is punctual and usually higher,because we considered the bridges a habitat disruption.

2.4.Statistical data processing

For statistical calculations,the species richness(the number of species)and the species presence records(the number of presence records,which is the absolute frequency of the species)were used.The diversity was expressed using the Shannon Index(H),which was calculated by considering presence records as abundance.The Jaccard Index(J)was used to estimate the similarity,which only considers the presence or the absence of species.The significance of the differences between all habitats was calculated with the Kruskal-Wallis test,and that of the pairing habitats with the Mann-Whitney test(U).

The link between the urban factors(residents,travelers,and road noise)and birds(species richness and species presence records)was verified using the Pearson correlation coefficient(r)and the coefficient of determination(r2).The multiple variants of the two coefficients(multiple R and R2)were used to estimate the connection between the three independent variables with urban birds.Adjusted R2was used instead of R2to penalize the points that did not fit the model.All statistical processing and graphical representations were performed using the Past 3.26 software(Hammer et al.,2001).

Because only the species identified in all studied urban green spaces were considered to have a high degree of urban tolerance,the variable of the species richness has the same value for all green spaces(see“species,high tolerance”in Appendix 1).This fact has led to the use of the species presence records for calculating the two indices above.

3.Results

3.1.Analysis of dependent variables along the urbanization gradient

In the investigated urban green spaces,we identified 113 bird species(Appendix 2),for which 2694 presence records were made.The species richness(NS)and presence records(F)have decreased from one habitat to another along the urbanization gradient,from the edge of the city to its center,both in the case of terrestrial and aquatic green spaces.The S3 brook sector(see Fig.1)had the species richness and presence records(NS=49,F=303)much smaller than the S5(NS=57,F=338)and S7 river sectors(NS=58,F=409),which are located in areas with the same degree of urbanization.The species richness and the presence records in S5(NS=57,F=338),located at the limit between the hills and the plain,with diverse geomorphology and habitats,were smaller compared to S7(NS=58,F=409),located in the plain,with lower habitat diversity and more water surface.

The diversity of the species(H)also decreased along the urbanization gradient,as did the species richness and the presence records(Fig.2).The S3 creek sector had lower diversity than the other studied river sectors.Along the river,the diversity of birds from S5(which has more diverse neighboring habitats)was greater than in S7.This situation is exactly the opposite of that of the species richness and the presence records.The total diversity of birds in the terrestrial habitats(H=3.84,NS=80)was higher than the total diversity in the aquatic habitats(H=3.73),even if the latter have a higher species richness(NS=89),due to a large number of species with very low frequency.

The Kruskal-Wallis test indicated significant differences between the studied urban green spaces(p=0.0002).By comparing them two by two,the Mann-Whitney test showed significant differences between some of the terrestrial habitats,namely between S4 and S1(p＜0.0001),respectively between S4 and S2(p=0.0003),indicating that the park located in the central area is the most different habitat.There were no significant differences between the aquatic habitats.If we compare the terrestrial habitats with the aquatic ones,in the Z2 zone with an average degree of urbanization there are significant differences between S2 and S3(p=0.01),whereas,in the Z1 zone,which is the most affected by urbanization,the differences are not significant between S4 and S6(p=0.29).

Fig.2.Shannon Diversity Index for the bird fauna on the territory of Oradea city(S1–S7:the studied urban green spaces according to Fig.1;arrow:sense of urbanization gradient).

The Jaccard similarity graph shows a decreased similarity between terrestrial and aquatic habitats(Fig.3).Among terrestrial habitats,the greatest similarity was registered between S1 and S2(J=0.63),and in the case of aquatic habitats,between S5 and S6(J=0.62).The S3 creek sector has the greatest similarity with S7(J=0.53).

3.2.Relation between the urban factors and the dependent variables

In the Multiple Linear Regression,the multiple adjusted R2showed that 87.11%of the variance of the species richness(multiple R2=0.87,p=0.02)and 99.2% of the variance of the species presence records(multiple R2=0.99,p=0.0004)are explained by the three independent variables together.Multiple R shows a very strong correlation between the three urban factors and the species richness(multiple R=0.96)or the species presence records(multiple R=0.99).

The Bivariate Linear Regression showed that the travelers explain the variance of each dependent variable in rather small percentages(30–40%)and the residents in very small percentages(both below 1%).The road noise is the only one that explains the very high percentage of the variance of dependent variables(Fig.4),more precisely it explains 72.54%of the species richness variance(r2=0.72,p=0.01)and 89.49%of the presence records variance(r2=0.89,p=0.001).The Pearson's coefficient for the bivariate correlations showed that birds are negatively affected,especially by road noise,both in species richness(r=-0.85)and species presence records(r=-0.94).For travelers,a lower negative correlation was obtained for both dependent variables(r=-0.64,respectively r=-0.54).The correlation between residents and birds was almost zero.

Fig.3.Jaccard Similarity Index for the bird fauna in Oradea city(S1–S7:the studied urban green spaces according to Fig.1).

Fig.4.Bivariate Linear Regression Model for the bird fauna in Oradea city.

Applying the same analysis for each bird category by their tolerance to urbanization,the highest values were obtained between species presence records and road noise.The strongest negative correlation was obtained by birds with a high degree of tolerance to urbanization(r=-0.97),followed by those with a medium degree(r=-0.84),and the weakest by those with a low degree(r=-0.71).Similarly,noise was the best predictor for the species presence records in birds with high degree of tolerance(r2=0.95,p=0.0001),then in those with a medium degree(r2=0.70,p=0.01),and the weakest predictor for low degree of tolerance(r2=0.50,p=0.07).

4.Discussion

The first reports on birds from the Oradea area date back to the second half of the nineteenth century(Mayer,1861;Kert′esz,1890).A century later,the avifauna studies conducted in the area focused on the riparian habitats that cross the city,either on the Crișul Repede River(Kov′ats,1973,1974,1976,1977a)or on the Pețea Creek(B′eczy et al.,1969;Kov′ats,1977b).However,very few records were made inside the city,most of which were near the flowing waters of the periurban area.Another study reports birds that have wintered in Romania in an unusual way,with observations made in several parts of the country,including Oradea(Mˇatieșand Kohl,1965).A total of 212 species were identified in all these studies.To this number,another 11 species identified in two fossiliferous sites near the city can be added;these are not extinct species but have not yet been confirmed alive in the area(Kessler,1974,1975).The species identified in the present study were mostly found in previous studies,only seven being new to the area(Aix galericulata,Aythya fuligula,

Circus pygargus,Hieraaetus pennatus,Iduna pallida,Mergus merganser,and Serinus serinus).Therefore,the total number of species identified so far in Oradea sums up to 230,and the species richness we identified in the city's green spaces is almost half of this value.This low percentage is not only caused by the limitations currently imposed on birds by the degree of urbanization of the habitats but also by the extinction of some species in the country's or region's fauna.Among the extinct species,there are mainly large birds,such as Aegypius monachus(Kert′esz,1890),Gyps fulvus(Kert′esz,1890),and Otis tarda(B′eczy et al.,1969).

In cities,the bird species richness generally decreases with the urbanization gradient due to the reduction of available habitats and the increase of the percentage of built areas(Mardiastuti et al.,2020).This usually leads to an increase in the number of species from the city center to the periphery(Gustin,2003;Rheinwald and Kelcey,2005).However,there may be exceptions to this general rule.If a city has a higher diversity of vegetation in its center than in other areas,then the bird species richness will be higher in the center(e.g.,Magre et al.,2019).If the study grouped the bird species according to certain ecological criteria,it was found that species richness in some groups increases with the gradient of urbanization if they have a low degree of specialization for the forest habitats(Sandstrom et al.,2006)or are omnivorous(M′ath′e and Bat′ary,2015).

Anthropogenic noise alters the biophony in urban green spaces,both in insects and in different groups of vertebrates,including birds(Lengagne,2008;Santos et al.,2021).In the case of the last one,the noise primarily affects their communication through song,whether it is about short or long distances between individuals(Grabarczyk and Gill,2019).Therefore,in the cities,birds sing louder and sometimes with higher pitch(Nemeth and Brumm,2010),and some species even change the structure of their song(Phillips and Katti,2020).However,urban noise not only makes communication more difficult for birds but also negatively affects their distribution,reproduction,or feeding,regardless of their tolerance to urbanization(Slabbekoorn,2013;Merrall and Evans,2020).It has been experimentally proven that loud noises can even cause damage to the auditory system of birds(Dooling and Popper,2007).Therefore,urban noise is considered a strong limiting factor,which often negatively affects urban birds much more than any other environmental factor(de Camargo-Barbosa et al.,2020;da Silva et al.,2021).An important recommendation from this point of view is that there should be no roads in the immediate vicinity of urban green spaces.When this is not possible,then high and dense vegetation at the edge of the green spaces would be recommended,which should act as a noise barrier(Tashakkor et al.,2020).

The roads affect birds in many ways,not only through the noise(Kociolek et al.,2011).The phantom road method was used to confirm the impact of road noise on birds;this involved simulating road noise with loudspeakers,eventually leading to a reduction in species richness and abundance(McClure et al.,2013).Studies of this type of impact have shown that noise intensity matters,larger and noisier vehicles making stronger disturbances than the smaller ones(Zhang et al.,2012).Also,the negative impact of a road is greater as its traffic volume increases,first affecting the species’reproduction and then their presence(Forman et al.,2002).

Even if there is a common European methodology for creating Lden noise maps,each state sets the allowed limit values of road noise so that the principle of prevention can be respected to preserve quiet areas in agglomerations(EP and CEP,2002).In Romania,the limit value established for streets,roads,and highways is 70 dB(A),and for public green spaces,it is 55 dB(A)(MMDD et al.,2008).These values are above those mentioned in the scientific literature about birds(Pat′on et al.,2012;Zhang et al.,2012;Wiącek et al.,2015).However,in Oradea,the entire ring road and several streets crossing the city and its center have values above 75 dB(A).Moreover,in five of the seven green spaces studied,the noise values exceed the allowed limit(see“noise”in Appendix 1).Currently,there is an action plan meant to manage the problems and effects caused by noise in Oradea,but this,unfortunately does not address the noise problem in the city's green spaces(EC and CPML,2018).For example,installing noise barriers could be a solution to reduce the disturbing effects of the road on birds(Zhang et al.,2012).However,the action plan mentions installating such barriers to protect only a few residential constructions on a very short distance of 1.8 km of the approximately 25 km of the ring road(EC and CPML,2018).On the other hand,in the city's most central park(1 Decembrie Park,marked S4 in figures),the city administration cut down the hedge that once surrounded this green space.The lowest bird diversity,the highest value of road noise,and the highest number of travelers was registered here.Such action plans target not only road noise but also railway,aviation,or industrial noises.The cumulative effect of several measures implemented to reduce urban noise can lead to real success,which has been achieved elsewhere(e.g.,Bento Coelho et al.,2005).In the case of the city of Oradea,we will see the results only after the new noise maps are made.

Birds respond differently to urbanization,from high sensitivity to human persecution and habitat disturbances to high(often totally)dependence on human resources(McKinney,2002).The difference between these two kinds of responses is caused by the tolerance to urbanization,which is determined by appropriate adaptations for exploiting resources and avoiding the risks of the urban environment(Sol et al.,2014).The consequence is that species with reduced sensitivity especially those dependent on the urban environment are found in most or even all green spaces of the city.Species with high sensitivity that do not depend on the urban environment are found very rarely in the city's green spaces,usually accidentally in one of them or possibly more frequently in well-preserved habitats.Therefore,species with a very high presence(present in all green spaces of the city)are more subject to the urban stress caused by road noise than those with a lower presence(present only in some of the green spaces).These considerations were used to establish the tolerance categories in the present study.The high tolerance species we identified had the frequency over time of the recordings(F)more controlled by the disturbing factor(road noise),and implicitly the two correlation indices(r2and r)had higher values for this category of tolerance.Similarly,species with lower tolerance had lower correlation indices values.According to data from the literature,birds have a different tolerance from one species to another(Pat′on et al.,2012).An important role in this is played by the frequency of each species'vocalizations since low-frequency vocalizations can be better masked by urban noise,which is more intense at low frequencies(Hu and Cardoso,2009;Cardoso et al.,2020).The effect will decrease the abundance of susceptible species(Wilson et al.,2021)and even species richness(Mena and Garcia,2018).Therefore,it was expected that in the present study,the highest correlations would have been identified in species with a lower tolerance,not those with high tolerance.But road noise in most of the green spaces we studied had values far above the legal limit(MMDD et al.,2008)or the acceptable level in other European studies(Pat′on et al.,2012;Wiącek et al.,2015).Probably the loud noise exceeded the tolerance threshold of many of the species known in the area,which partially explains why we identified only 113 of the 230 species ever listed in the area.Species with a low tolerance for urbanization identified in the city have a limited distribution not only because of the noise but also because of other reasons,such as their specific ecological preferences and the corresponding ecological particularities of the habitats in which they have been identified.

Increasing human population size and density is one of the validated extinction threats to birds(McKee et al.,2013).However,at a global(Luck,2007)or sometimes continental and national scales(Evans et al.,2007;Lazarina et al.,2020),a significant positive relationship has been found between bird species richness and human population density.The cause is the spatial congruence between humans and birds due to primary productivity,to which both human population density and bird species richness respond positively(Chown et al.,2003;Luck et al.,2010).But this positive relationship decreases with the sampling unit used(Luck,2007).That is why,in other studies on a national(Koh et al.,2006;Wilson et al.,2021)and especially on a local scale(Fontana et al.,2011),human population density has a weak negative relationship with species richness.And in such cases,one of the main causes is primary productivity,which is affected by increasing human population density(Koh et al.,2006).In other words,man either co-occurs with birds if there is sufficient primary productivity in their shared space(which puts birds in an indirect positive relationship with human)or eliminates birds if primary productivity is reduced by urbanization(thus causing an indirect negative relationship between the two parties).Land use change during urbanization participates in these consequences.If we assume that in our study primary production plays an important role in determining weak negative correlation between human population density and birds,we can affirm that urban planning did not take into account a direct proportionality between human population density and primary productivity in each territorial subunit.But on such a small spatial scale,many other particular issues in addition to primary productivity can have an influence,such as the size,structure,location,or connectivity of green spaces,which in turn have led to many other variables proposed and verified in urban green space statistics(Qian et al.,2009;Kang et al.,2015;Shimazaki et al.,2016;de Camargo-Barbosa et al.,2020;Pirzio--Biroli et al.,2020;Campos-Silva and Piratelli,2021;da Silva et al.,2021).In contrast to road noise(which affected birds strongly and directly)or human population density(which affected weakly and indirectly),travelers identified according to territorial subunits had an average correlation with birds.Their relationship with birds was both a direct one,in the case of at least some of the pedestrians(Fern′andez-Juricic,2000)and an indirect one,in the case of trips with public transport or personal vehicles,due to the noise produced during transport(Perillo et al.,2017).

5.Conclusions

Seven representative urban green spaces were studied,in which 113 species of birds out of the 230 known in the area were identified.The species richness and diversity have declined along the urbanization gradient.The effects of urbanization are more obvious in terrestrial habitats compared to aquatic ones.Significant differences between terrestrial and aquatic habitats become insignificant along the urbanization gradient.

Human density slightly negatively affects the remaining bird species in the city after urbanization.Still,the mobility of people and especially the disruptive factors generated by this mobility,such as road noise,have a huge impact.The most sensitive to road noise are the species with the highest degree of tolerance to urbanization.Road noise is a better predictor of species presence records than species richness.The road noise explained in a very high percentage the species presence records.In comparison,urban mobility explained much less the species presence records,and the density of the human population almost nothing.The noise,urban mobility,and population density explained over 99%of the species presence records.

Authors’contributions

AȘCL designed the study.HVK,PVP,DM and AȘCL collected the field data and sometimes took pictures.CG identified the birds from the pictures.ASCL wrote the manuscript.CG and ICT revised the manuscript.All authors read and approved the final manuscript.

Funding

The authors declare that the study was not funded.

Ethics statement

Ethical approval is not required for the study because data aqusition of our research was based on an observational way and no animals were manipulated.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

We are grateful to Sara Ferenți for language edits and to the two anonymous reviewers for their comments,which highly improved the manuscript.

Appendix A.Supplementary data

Supplementary data to this article can be found online at https://do i.org/10.1016/j.avrs.2022.100067.