ANALYSIS OF THE CURRENT STATUS OF AVIFAUNA IN KOSTOMUKSHA STATE NATURE RESERVE AND KALEVALA NATIONAL PARK (NORTH-WEST RUSSIA), TAKING INTO ACCOUNT INFLUENCE FROM ADJACENT AREAS

The paper offers an assessment of the current status of the avifauna in Kostomuksha State Nature Reserve (KNR) and Kalevala National Park (KNP). They are the two largest Protected Areas (PAs) in Russian Karelia, having a key importance for the conservation of pristine boreal ecosystems in the Green Belt of Fennoscandia. The species composition and abundance parameters of birds in these PAs are described on the basis of data from 2015–2019 and compared with data gathered in 1980–1990. We specifically address the influence of adjacent areas on the bird fauna of the PAs. First of all, it was the potential effect of recently logged areas in the vicinity of the protected territories on avifauna in their periphery. The studies were carried out as transect counts and point counts with the resultant data processing in R environment. Comparison with data obtained in 1980–1990 corroborates the hypothesis about a species turnover of boreal ecosystems in the direction of increasing the ratio of southern species. The observed species turnover in the PAs is consistent with the ideas about ongoing global species turnover due to climate change. As for the anthropogenic load on ecosystems, until now, no critical ef� fect of logged areas adjoining KNR and KNP on the avifauna in these PAs has been detected. Presumably, the pressure on the ecosystems is alleviated by the size of the PAs and the scope of their conservation. Arguably, if the key current parameters of the PAs are maintained, in KNR and KNP, the natural bird fauna would be pre� served in spite of the increased anthropogenic pressure in adjacent areas. However, it cannot be ruled out, that as the logged area around the PAs expands, and the number of species associated with open habitats increases accordingly, some of them may disperse into the transformed habitats available in the PAs. Hence, monitoring of the avifauna in these PAs should be continued, especially in their periphery, considering the tendencies for an increase in the concentration of southern bird species and widespread species near the disturbed areas. The high� est threat contained in continuing logging around the PAs is a farther spread and a rise in the numbers of species not typical of boreal ecosystems. Thus, logging can contribute to the species turnover induced by climate change and exacerbate the situation. Such a shift in the ratio of northern and southern birds may, in turn, jeopardise the balance in boreal communities and thereby increase their vulnerability.

Despite the mismatch between species range shifts and the stable location of PAs (Heller & Za� valeta, 2009;, they are still considered to reduce the negative effects of cli� mate change on communities and contribute to spe� cies conservation (Thomas et al., 2012;Virkkala et al., , 2018Gillingham et al., 2015;Santangeli et al., 2017;Lehikoin� en et al., 2018). Thus, in the light of climate change and other anthropogenic loads on ecosystems as well as the significant role of PAs in bird preserva� tion, the assessment of the current state of their avi� fauna is becoming especially relevant.
The study of northern habitats occupies an im� portant place among such research. In the northern taiga regions, the loss of pristine habitats, espe� cially forests, due to commercial use, is one of the main problems in relation to local fauna preserva� tion. Besides the loss of forests itself, a shift in the ratio of mature forests and young age habitats are increasing. These problems can be the causes of changes in bird communities, and they are relevant to northern taiga regions (e.g. Fraixedas, 2017;Mäntyranta, 2019). Another type of highly threat� ). Another type of highly threat� ened ecosystems there is peatland habitats (Fraixe� das et al., 2017). Despite this fact, there is a defi� cit of research on the state of bird communities in peatlands .
Addressing these problems, in northern regions, PAs with pristine forests and peatlands can play the main role in bird species preservation. Indeed, there is some evidence that PAs contribute to maintain� ing the community of northern species (e.g. Virk� kala et al., 2014Virk� kala et al., , 2018Santangeli et al., 2017), and it requires investigations of the bird status in such PAs. Moreover, the Arctic Ocean is a natural barrier to species' northward range shifts (Virkkala et al., 2008), which increases the need to study bird com� munities and their dynamics in the northern boreal and Arctic PAs (Virkkala et al., 2018).
Russian ornithologists also surveyed some ar� eas inside KNR in the spring and summer seasons in 2000 and 2009 (Sazonov, 2015). On the basis of a summary of all the collected material, conclu� sions were drawn about the zoogeographical status of the area, and the PA�averaged number of breed� ing pairs was estimated for various species. These results were successfully used for evaluating the importance of the PA for birds. But they are hardly applicable for interpreting local breeding densities, and therefore have little potential of being used for comparisons in biogeographical papers. Further� more, these data may be no longer valid for the current situation in KNR avifauna because a ma� jority of them was gathered over 20 years ago, and the most recent ones being 10 years old.
Also, some data from winter counts in KNR, especially on Galliformes, were published (Ka� shevarov, 1998;Kashevarov & Heikkilä, 2003;Bologov & Sikkilä, 2013;Preobrazhenskaya, 2015). These data are of high value for the analysis of the wintering avifauna in KNR but cannot be extrapolated to breeding densities in the PA.
Importantly, forests around KNR and KNP have been actively logged in the past decade (Hans� en et al., 2013;Global Forest Change, 2019). Now these PAs are almost surrounded by logged areas spreading towards their borders. Thus, in a situa� tion where we lack recent datasets on the status of avifauna in KNR and KNP, and where the fauna is influenced by the climate change, succession and impact of adjacent logged areas, there is a high de� mand for new ornithological surveys both in their central parts, away from the transformed habitats, and in the periphery bordering the disturbed areas. This is needed to enable the analysis of the current state of these PAs.
Our studies aimed to analyse the current status of avifauna in KNR and KNP using the material gathered in the field in 2015-2019. The first results of these studies (covering 2015-2016) have been published (Simonov & Matantseva, 2017). Data collection in the following seasons and expansion of the area covered by surveys to include the sur� roundings of the PAs were designed to update the information on the species composition of birds in these PAs, their status and abundance, as well as to assess the effect of recently logged areas near the PAs on their avifauna. The working hypothesis implied a multifaceted remote detrimental effect of habitat disturbance on the abundance of birds representing the boreal avifauna, a neutral effect on the numbers of background species (see section Material and Methods), and a positive effect on the numbers of open habitat species in the territories bordering the logged areas.
Being situated at the Finnish�Russian border and possessing large areas of intact old�growth bo� real forests, these PAs are of global value as key components of conserved ecosystems native to the Green Belt of Fennoscandia (Sazonov, 1997(Sazonov, , 2015. KNR is a part of the Finnish�Russian Friendship Nature Reserve. Both PAs are incorporated in the UNESCO Metsola Biosphere Reserve (UNESCO, 2020b). Metsola was established in 2017 within a project of the UNESCO Man and the Biosphere Programme. According to that, a biosphere reserve is a �Science for Sustainability support site� be� �Science for Sustainability support site� be� Science for Sustainability support site� be� � be� ing a special place for testing interdisciplinary ap� a special place for testing interdisciplinary ap� proaches to understanding and managing changes and interactions between social and ecological sys� tems, including conflict prevention and manage� ment of biodiversity (UNESCO, 2020a).
Field data were gathered in 2015-2019 (June 2015; July 2016, 2018, 2019; August 2017) by transect (Sazonov, 1997) and point (Bogolyubov, 1996) count techniques. The transect method was applied to enable a comparison between our data and the data collected in 1980-1990s by Sazo� nov (1997), whose publication is very informa� tive and also presents the exact routs of surveys which could be repeated. Sazonov (1997) used the transect method with routes covered all types of habitats. Minimal data processing excluded any species-specific coefficients, corrections or any other manipulations, and all available data repre� sented �raw� material. Our transect counts, as well as Sazonov's (1997) counts, were conducted by the �route accounting unlimited bandwidth detection�. Five transects copied the routes of Sazonov (1997) and 15 transects were freely chosen. We repeated counts in transects 2-3 times per season.
The average population density was calcu� lated according to the average detection range of the birds. The abundance limits in the Electronic Supplement were selected from the density of every species along every route inside the PAs. The popu� The popu� lation density of every species was estimated with the following formula (Romanov & Maltsev, 2005): where N -the population density, individuals/ km 2 ; X -the average number of individuals; hthe average detection range, km; L -the length of the route, km.
Point counts were used for a more precise as� sessment of changes in bird numbers depending on the distance to the source of impact (in our case, the edge of an adjacent logged area) with the cho� sen models (see below). We used two parameters of bird numbers: (1) the number of individuals and (2) the number of species. Model circular transects were marked out within KNR and KNP, and in ad� jacent areas recently logged through checker-wise series of fellings. The points for the counts were evenly distributed along each transect, with 250-300 m spacing. Observation at each point lasted 5 min. Each transect (both in transect counts as such and in point counts) was 5-10 km long. To analyse how recently logged areas adjoin� ing the PAs might have affected the avifauna of KNR and KNP we estimated the frequency of oc� currence of birds of various species along transects at various distances from the PA�bordering logged areas. In our study, the model type of �recently logged areas� was a site harvested within the past five years. Surveys spanned 0 to 12 km from the nearest edge of the model�type logged area closest to the PA border and directly into the PA. Bird en� counters were referenced to the geographic co�or� dinate system. The smallest distances from obser� vation points to the nearest edge of a logged area were calculated by using geographic co�ordinates and the haversine formula.
Species names and taxonomic affiliations are cited according to the Bird Checklists of the World (Avibase, 2019). Their classifica� tion into faunistic groups is present according to Sazonov (2012). We used the classification into faunistic groups distinguished by Sazonov (2012) to enable a comparison between our data and the data collected by the researchers who have worked in KNR and the planned KNP area in the past (Sazonov, 1997). Three faunistic groups (northern species, widespread species and southern species) were selected according to the origin of the species (Sazonov, 2012). The other three groups (target groups) were se� lected to verify the hypothesis about a logging influence on the birds in the studied ecosys� tems. Two of these groups (boreal species and open�site species) conform to the groups dis� tinguished by Sazonov (2012), and one group (background species) includes the most numer� ous taxa. Background bird species are all bird species with a density more than one individual per unit (1 km 2 ) (Tsybulin, 2009).
Statistical analysis was done in R environ� ment v. 3.6.1 (R Core Team, 2019) and using Mi� crosoft Excel software. We compared the level of diversity and evenness in the surveyed communi� ties utilising the Shannon and Simpson indices. Data distribution within groups was tested for normality by the Anderson�Darling test (package �nortest� v. 1.0�4 (Ligges & Gross, 2015)). The statistical significance of average diversity dif� ferences between communities was measured by the Student's t�test (Microsoft Excel software). The similarity between the communities was es� timated by the Jaccard index. Pairwise correla� tions were calculated using the Spearman rank correlation coefficient (package �stats�, part of R v.3.6.2, by R Core Team, 2019).

Species composition and abundance parameters of birds in Kostomuksha State Nature Reserve and Kalevala National Park
Our surveys inside KNR and in its adjacent ar� eas yielded records 1279 individuals of 114 bird species (Electronic Supplement) belonging to 38 families and 12 orders. Among these, 109 species (106 breeding or allegedly breeding) were encoun� tered only inside KNR, and two species (both of them are breeding birds: Motacilla flava Linnaeus, 1758 and Oenanthe oenanthe (Linnaeus, 1758)) were found only in nearby logged areas. Three other breeding species (Columba livia Gmelin, 1789, Phylloscopus sibilatrix (Bechstein, 1793), and Passer domesticus (Linnaeus, 1758)) were re� ported from Kostomuksha town suburbs adminis� tratively belonging to the Protected Area.
Columba livia and Passer domesticus are typ� ically synanthropic birds. Their presence within KNR is unlikely, although not ruled out, given the vicinity of the town. Phylloscopus sibilatrix forms a low�density (not more than 1-2 pairs/ km 2 ) population in Kostomuksha's green belt, in the best�suited habitats in the study period. Being in the northern periphery of the species' range, the area contains low numbers of these birds, fluctu� ating substantially through the years. Phylloscopus sibilatrix breeds in sparse and tall pine�birch (Pinus sylvestris, Betula sp.) forests with scarce undergrowth which occur in the town's green belt but not in the KNR main area. It is probably be� area. It is probably be� . It is probably be� cause of the lack of good breeding habitats that P. sibilatrix was not encountered inside KNR.
Encounters of species such as Anas platyrhynchos Linnaeus, 1758, Chroicocephalus ridibundus Linnaeus, 1766, and Larus canus Lin� naeus, 1758 inside KNR were singular, but their numbers were much higher in the town. Up to several broods of A. platyrhynchos aggregate near a small lake, where local people come to feed them. Chroicocephalus ridibundus and L. canus form larger aggregations of up to several dozens of birds on Lake Kontokkijarvi. Sightings of Sturnus vulgaris Linnaeus, 1758 are rare both in the town and in KNR.
Surveys in KNP and its surroundings in 2015-2019 yielded records 982 individuals of 87 bird species (Electronic Supplement) belong� Electronic Supplement) belong� ) belong� ing to 31 families and 11 orders. In the KNP, the breeding was confirmed or alleged for all these species, excluding Stercorarius pomarinus (Tem� minck, 1815), a bird vagrant in KNP. Of these, 85 species were encountered inside the PA, and two species, as in the case of KNR, were sighted only in logged areas adjoining the PA: Motacilla flava and Oenanthe oenanthe.
The Electronic Supplement reports data from the surveys carried out in 2015-2019 similarly to Sazonov (1997) to enable a comparison between new records and the data gathered more than 20 years ago. Similarly to Sazonov's studies, the bird population densities were averaged over the entire landscape (the type of terrain), including mires and small water bodies.
Surveys in the logged areas closest to the PAs detected the breeding of birds typical for open habitats, such as Oenanthe oenanthe, Saxicola rubetra (Linnaeus, 1758), Motacilla alba Linnaeus, 1758, and Motacilla flava. Oenanthe oenanthe and M. flava have not been encountered inside the PAs. But near the KNR and KNP their breeding sug� gests that these birds may be visiting the PAs, too.
In 2015-2019, some species were not observed in KNR and KNP, first of all, some semiaquatic birds, as well as some diurnal and nocturnal raptors, which, judging by earlier reports (Danilov et al., 1977;Adri� anova et al., 1990;Rajasärkkä & Virolainen, 1995;Sazonov, 1997Sazonov, , 2015Zimin & Sazonov, 1997;Gromtsev, 2002;Rajasärkkä, 2004;Preobrazhens� kaya, 2015), were rare or very rare breeders in and around the PAs. The status of the previously observed species has not necessarily changed. Possibly, these birds did not occur in our surveys because of their low numbers and irregular breeding. Further surveys are needed to get updates on the status of such species and their breeding number in the PAs.

Rare bird species in Kostomuksha State Nature Reserve and Kalevala National Park
By surveying the PAs in 2015-2019, we re� corded ten species (subspecies) listed in the Red Data Book of the Republic of Karelia (2007) It is also interesting to look at some other spe� cies that are not included in the Red Data Books of either Russia or Karelia but are rare and in need of monitoring and protection. For instance Cygnus cygnus Linnaeus, 1758 is supposed to breed in KNR and KNP. Several times in both PAs we observed C. cygnus with young individu� als. Besides, KNR is a breeding site for Anser fabalis fabalis (Latham, 1787), a subspecies whose population has been declined in the past several decades throughout its range (Marjakangas et al., 2015). In 2015, in late June, we observed here a brood of Anser fabalis fabalis with goslings about half the size of an adult. Since the PAs are part of the Russian�Finnish Nature Reserve, it would be relevant to mention that we registered there 44 species included in the 2019 Red List of Finn� ish Species (Hyvärinen et al., 2019) (Electronic Supplement). Rare species, especially special� ). Rare species, especially special� ists, deserve particular attention because whereas generalist species can be tolerated to transformed habitats and even benefit from them. Rare spe� cialist species can be especially vulnerable to habitat loss or fragmentation (e.g. Davies et al., 2004;Laurance & Vasconcelos, 2009;Lebbin et al., 2010;Morante�Filho et al., 2015).

Comparative analysis of avifauna in Kostomuksha State Nature Reserve and Kalevala National Park at present and in the past
The species composition of bird populations was similar between KNR and KNP (Electronic Supplement), i.e. the Jaccard index was 0.76. The similarity in the species composition and number of birds between these PAs was corroborated also by the Shannon Diversity Index (SHDI), which was 3.70 for KNR and 3.39 for KNP (t = 1.19, t tab. = 1.97, t < t tab., i.e. differences are insignifi� cant). Similar values were also obtained for the Simpson Dominance Index (D), which measures community evenness numerically (D = 0.03 for KNR and D = 0.04 for KNP).
Thus, data from 2015-2019 give no rea� data from 2015-2019 give no rea� from 2015-2019 give no rea� from 2015-2019 give no rea� 2015-2019 give no rea� give no rea� no rea� no rea� rea� rea� son to speak of any significant distinctions in the composition of avifauna between KNR and KNP previously reported by Sazonov (2015). Analysing the composition of faunistic groups in the PAs, Sazonov (2015) noted that KNP has a higher proportion of northern species and a lower proportion of southern species than KNR. These differences, however, were relatively mi� nor: 43 northern species (39.1% of the avifauna) and 23 southern species (22.3%) in KNR, and 46 northern species (41.1%) and 21 southern spe� cies (18.7%) in KNP.
Our data on the distribution of species in the PAs among the faunistic groups distinguished by Sazonov (2012Sazonov ( , 2015 are given in Table 1. Among the species breeding or allegedly breeding in the PAs, northern species account for 31-35%, widespread species for 42%, and southern species for 24-27%. A similar ratio is observed for regu� larly breeding species taken alone, too. This dis� tribution pattern is generally typical in boreal bird faunas (Sazonov, 2015). It would be premature to proclaim there is a significant difference between the PAs on this parameter since a more complete faunogenetic analysis requires a clarification of the status of the species that are now only alleged to breed in the conservation areas. Note: 1 -breeding and allegedly breeding species (Sazonov, 1997), 2 -breeding and allegedly breeding species (our data from 2015-2019). By judging the data of 1980(Sazonov, 1997, in the PAs in the latest decades, a tendency has been for a decline in the proportion of north� ern species and a rise in the proportion of southern species (Table 1). Also, earlier surveys in KNR re� corded higher numbers of boreal birds (e.g. Tetrao urogallus Linnaeus, 1758; Dryocopus martius (Lin� naeus, 1758), Picoides tridactylus (Linnaeus, 1758), Perisoreus infaustus Linnaeus, 1758, Poecile cinctus Boddaert, 1783, Bombycilla garrulus (Linnaeus, 1758), Emberiza rustica Pallas, 1776). In the earlier surveys, higher breeding densities were observed even among such vulnerable boreal species as Gavia arctica (Linnaeus, 1758), Cygnus cygnus, Anser fabalis, Bucephala clangula (Linnaeus, 1758), and Mergus merganser Linnaeus, 1758. None of these species had high numbers in our surveys (Electronic Supplement). On the contrary, almost all of them demonstrated a low population density. In KNP, some increase in breeding density was detected in 2019 for the boreal species Fringilla montifringilla Linnaeus, 1758 and Emberiza rustica, i.e. 5.1 pairs/ km 2 and 2.7 pairs/km 2 , respectively. This is notably higher than their annual breeding density averaged over our study period (1.5 pairs/km 2 and 0.2 pairs/ km 2 , respectively); but it is still lower than Sazo� nov (1997) reported: 13.5-41.7 pairs/km 2 and 2.8-12.5 pairs/km 2 , respectively.
Thus, our data corroborate the conclusions about an ongoing species turnover. Researchers working in boreal landscapes of Northwest Russia have noted a reduction in the abundance of northern species, in� cluding boreal specialists, and their replacement by birds of southern latitudes. The latter lack the full set of adaptations to living in the north, first of all, to the reduced duration of the season suitable for breeding, creating risks for the stability of northern communi� ties (Sazonov et al., 2002;Hokhlova & Artemiev, 2007;Danilov, 2010;Khokhlova & Artemiev, 2011).
Importantly, the decrease in number of northern species and increase in number of southern species associated with northward range shifts and density shifts in boreal landscapes have been also described in different northern regions (Lindström et al., 2013;Lehikoinen & Virkkala, 2016;Fraixedas et al., 2017;Virkkala et al., 2018). For instance, in Finland latitudinal northward shifts of the northern range borders in both southern and northern species were found. And for southern species, it was more pronounced and confirmed (Brommer et al., 2012). Later, climate�induced den� sity northward shifts of species were described, al� al� though such shifts were faster in northern birds than in southern ones . The researchers also expressed concern that a high bird species turnover observed in northern Europe may af� fect the functional diversity of species communities (e.g. Virkkala & Lehikoinen, 2017).
Considering that the same process is happening in many boreal landscapes, it may partially reflect the global species abundance trends (EBCC, 2019) in� duced by a combined action of many factors not only in breeding areas, but also in overwintering grounds and along flyways. Indeed, similar processes are ob� served in many studied communities all around the world. And the main reason for them considers to be the ongoing climate warming (Parmesan, 2006;De� victor et al., 2008De� victor et al., , 2012Jiguet et al., 2010;Chen et al., 2011;Lindström et al., 2013;Roth et al., 2014;Lehikoinen & Virkkala, 2016;Virkkala & Lehikoin� en, 2017;Brotons et al., 2019;Yang et al., 2020).
The additional reason of such changes might be the increasing human pressure on ecosystems, such as agricultural land�use, habitat disturbance, shrinking of old�growth forests, and expansion of logged areas, habitat fragmentation, and habitat loss (e.g. Bowler et al., 2018;Yalcin & Leroux, 2018;Flesch, 2019). Particularly, in northern re� Particularly, in northern re� gions, disturbed habitats may not simply lose the properties required for them to be inhabited by typical boreal species, but also turn into ecological channels for the northwards expansion of species associated with broadleaved forests (Zimin, 2001).
In this light, the role of PAs may be even more appreciated. Obviously, the influence of climate change and its possible effects on the bird fauna spreads even on large pristine nature reserves (Langdon & Lawler, 2015;Virkkala et al., 2018). Although in boreal PAs birds also shift northwards in the warming climate, and PAs can be not so ef� ficient in preventing the climate-induced decline of migrants, the PAs are still successful in preserving resident bird species (Virkkala et al., 2018). More� over, the size of PAs had a positive effect on the trends of residents (Virkkala et al., 2018), that in� creases the value of PAs like KNR and KNP. There is some evidence that PAs are really important for supporting the communities of northern species. Nevertheless, these communities are still shifting both inside and outside of PAs that requires proper conservation measures (Santangeli et al., 2017).

Assessment of nearby logged area's effects on avifauna in the Protected Areas
The most typical species inhabiting logged areas near KNR and KNP were birds associated with open habitats: Saxicola rubetra, Motacilla flava, and Oenanthe oenanthe. Saxicola rubetra and O. oenanthe bred in nearly every large logged area in the study area. Motacilla flava was not so frequent. As a rule, only one pair of this species was spotted in the surroundings of each PA per season. As mentioned above, M. flava and O. oenanthe were encountered in logged areas only, with no records from inside the PAs. Saxicola rubetra had a higher density in logged areas but reached also into the PAs periphery.
The cases of an increase in bird species di� cases of an increase in bird species di� increase in bird species di� versity immediately after some anthropogenic im� pact, including habitat transformation and habitat fragmentation, have been noticed under different conditions and in different countries (e.g. Azeve� do�Ramos et al., 2006;Kurhinen et al., 2009;Ma� et al., 2006;Kurhinen et al., 2009;Ma� Kurhinen et al., 2009;Ma� karova & Manukov, 2016;Lapshin et al., 2010). At the same time, it should be taken into account that this may be a result of so�called pseudo�en� richment of species diversity, which, as noted by Kurhinen et al. (2009), corresponds to the �inter� mediate violation hypothesis� (Connell, 1977). According to this hypothesis, the violation of the biocenosis at the initial stages may contribute to a temporal increase in its species diversity but subse� quently threatens to reduce the abundance of spe� cialised species. Similarly, when taiga ecosystems change against the background of an increase in species diversity due to species of open habitats, a decrease in the proportion of typical boreal species may occur. This again brings us to the problem of the species turnover in taiga fauna.
Besides that, the initial increase in diversity after anthropogenic impact can be followed by a contraction, because the remaining habitats may become insufficient for the increased abundance or an increased number of species. Moreover, in� vasion of non�forest bird species may increase competition for resources as well as parasite load, which later may reduce the reproductive success and viability of populations (Barrantes et al., 2016).
In the case of our research, calculations with the total data pool revealed no significant cor� relation between the number of species (Spear� man rank correlation: S = 1129900, rho = 0.072, p = 0.32) or total bird population density (S = 1225800, rho = �0.007, p = 0.92) and the distance to the edge of the nearest logged area. However, if the bird number of different faunistic groups (Electronic Supplement) was analysed separate� Electronic Supplement) was analysed separate� was analysed separate� ly, an inverse correlation was found between the abundance and the number of southern species and the distance to the edge of the nearest logged area (Table 2). In other words, both the number of southern species and their total abundance were higher closer to logged areas. The effect of the vicinity of the logged areas on the number of spe� cies and abundance of northern birds and wide� spread species was insignificant (Table 2).
By verifying the working hypothesis using the survey data, we did not find the expected negative effect of a logged area's vicinity on the numbers of boreal species (Table 2). Its effect on the num� ber of background species (Anthus trivialis (Lin� naeus, 1758), Regulus regulus (Linnaeus, 1758), Poecile montanus Conrad von Baldenstein, 1827, Fringilla coelebs Linnaeus, 1758, Fringilla montifringilla) was supposed to be neutral. But a weak negative correlation was detected between their abundance and distance to the logged area ( Table  2). The assumption about a positive effect of the logged area's vicinity on the number of birds asso� ciated with open habitats was confirmed (Table 2). Thus, judging by our data, boreal birds inhabiting the PAs are tolerant to the presence of logged areas around the PAs. The most probable explanations for that are the large size of the PAs, substantial areas of habitats suitable for living and breeding, and low human pressure inside the PAs. The negative correlation between the abundance of background species and distance to a logged area is also quite explicable, given that inner parts of the PAs mainly have climax taiga communities with less diverse habitats. Compared to these monotonous cli� Compared to these monotonous cli� Compared to these monotonous cli� max habitats, the more diverse sites in the PA periph� ery, with a larger proportion of habitats not yet in the final stages of the succession, and with a higher con� tribution of edge habitats, attract many bird species, including the most abundant ones.
The positive effect of logging on the abundance of birds associated with open habitats is not surprising either. Moreover, some species appeared around the PAs exactly owing to the presence of logged areas. Apparently, the lack of open habitats inside the PAs prevents to dispersing of open site species into inner parts of them. However, there are several demolished houses and engineering constructions inside both PAs, but habitats in their surroundings are suboptimal for open site species. So, it cannot be ruled out, that as the logged area around the PAs expands, and the number of species associated with open habitats in� creases accordingly, some of them may disperse into such transformed habitats in the PAs. Whether or not this increase would induce critical changes in the na� tive avifauna is still a question. As now the threat ap� pears minor, further monitoring of the bird population around the PAs is needed to be able to predict the situ� ation for the PA fauna and secure timely response to potential risks. In the light of this fact, it is even more important, that an increase in number of birds affi li-of birds affi li-birds affili� ated with open habitats might be also partially caused by global warming. The climate change�induced shifts in bird communities must be timely monitored inside PAs as well as outside them.
For example, taking a pair of typical boreal birds belonging to the same family (Tetrastes bonasia (Lin� naeus, 1758) and Tetrao urogallus) and differing in resistance to human pressure, we estimated the small� est distance to a recently logged area where these birds were encountered in the PAs. The occurrence of Tetrastes bonasia in the PAs started 270 m away from recently logged areas, whereas T. urogallus did not occur until 2 km away from them. However, in our study, no further correlation was detected between the abundance of these birds and distance to the logged area (Spearman rank correlation: S = 5920200, rho = 0.104, p = 0.06, and S = 6236100, rho = 0.056, p = 0.30 for T. bonasia and T. urogallus, respectively). Nonetheless, in northern taiga forests of Karelia, the distribution of T. urogallus has been previously dem� onstrated its correlation with the proportion of forest cover and the distribution of mature and over�mature coniferous stands (Danilov, 2010). Here, the species biology, specific movements of individuals among habitats, should be taken into account. In particular, when the mating period ends, T. urogallus retreats to �hideaways� in the forests. By the late summer, its females with broods occupy the most food�rich lo� cations, and the distribution of birds among habitats evens out (Danilov, 2010). A more comprehensive analysis of the remote effect of logging on T. urogallus abundance in undisturbed areas certainly requires gathering more data for different stages in the annual cycle of these bird species.
Among Galliformes, the highest tolerance to logged areas in the surroundings was demonstrated by Lyrurus tetrix Linnaeus, 1758, a widespread spe� cies. Individuals of this species were encountered at the very edge of logged areas. The association of these birds with disturbed areas was highlighted in previous studies as well. In Karelia, the abundance of L. tetrix positively correlated with the proportion of logged areas and young forest stands, while it cor� related negatively with the proportion of old�growth forests (Danilov, 2010).
Thus logged areas are a factor attracting southern birds and widespread species to the PAs periphery. On the other hand, the extensiveness of KNR and KNP areas mitigates the effects of nearby logged areas on the abundance of northern species, which would have otherwise been more pronounced.

Conclusions
In 2015-2019 surveys, the species composition of the bird population between KNR and KNP was similar and typical for boreal bird faunas. The trend observed lately in the studied PAs is a decline in the proportion of northern species and an increase in the proportion of southern species. Presumably, the most probable reason of such shifts in boreal ecosystems is global warming. Against this background, the an� Against this background, the an� an� thropogenic transformation of habitats may intensify changes in the species composition. Considering that similar processes are happening in many boreal land� scapes, including PAs, they may partially reflect the global species abundance trends induced by the cli� mate change, human impact and combined action of other factors not only in breeding areas but also along flyways and in wintering sites. These issues need to be studied in more detail.
Assessing the effect of logged areas adjoining to PAs, we observed that among the three bird spe� cies dwelling in logged areas around KNR and KNP (Saxicola rubetra, Motacilla flava, and Oenanthe oenanthe), only one (S. rubetra) was encountered in peripheral areas within the PAs. However, it cannot be ruled out, that as the logged area around the PAs expands, the number of species associated with open habitats also increases. Some of them may disperse into the transformed habitats available in the PAs. Whether or not this would induce critical changes in the native avifauna is still a question. The threat ap� pears minor so far, but in the light of ongoing climate change and habitat transformation, further monitor� ing of the bird communities around the PAs is needed to be able to predict the situation for the PA fauna and secure timely response if risks are identified.
At present, the effect of logged area's vicinity is most pronounced for the abundance and composition of southern species, which usually favour disturbed habitats and avoid the �wildest� parts of the taiga. A similar tendency among widespread species is evidenced by the negative correlation between their abundance and distance to the logged area's edge. It would be logical to assume that continuing logging near the PAs would enlarge the proportion of wide� proportion of wide� of wide� spread species and promote the species turnover in� side the PAs as well, at least in their marginal parts.
Analysis of data obtained in 2015-2019 revealed no negative impact of logged areas' vicinity on the abundance of boreal bird species inside the surveyed PAs. The resistance of the conservation areas to log� ging around them is most probably due to their sub� stantial size, and strict limitations on human activities in the area. The latter proves once again how impor� tant it is to preserve large pristine forest stands and PAs Santangeli et al., 2017;Lehikoinen et al., 2018), that should be prioritised by the nations possessing such natural areas (Gorshkov & Makarieva, 1998;Makarieva & Gorshkov, 2012).
In conclusion, we can state that both climate change and logging change the boreal bird com� munities towards domination of southern species. Thus, anthropogenic pressures are shaping bird communities, which can influence both PAs and areas outside of them.

Supporting Information
The full dataset with 118 bird species (Elec� tronic Supplement: Bird species, faunistic and target groups, conservation status, and abundance param� eters averaged over the entire area of the surveyed type of landscape in Kostomuksha State Nature Re� serve (KNR) and Kalevala National Park (KNP)), may be found in the Supporting Information here.