RETRACTED ARTICLE: Long-term dynamics and distribution of large carnivores in Poland

TL;DR

This study examines the population growth and territorial expansion of three large carnivores in Poland over several decades.

Contribution

The study provides new insights into the long-term population dynamics and distribution patterns of brown bears, lynx, and wolves in Poland.

Findings

All three carnivore populations increased, with gray wolves showing the highest growth rate.

Lynx and wolf ranges expanded westward, while the bear's range remained stable.

Multi-year data analysis improves understanding of population trends for conservation planning.

Abstract

The brown bear Ursus arctos, Eurasian lynx Lynx lynx, and gray wolf Canis lupus are Europe’s threatened large carnivores. The analyses were conducted using data on the abundance of these species in Poland, collected by the Polish Central Statistical Office (bear 1965–2023, wolf 1995–2023, and lynx 1996–2023). For the years 2000–2023, data were also available by region. We subjected these data to statistical analysis: chi-square tests, segmented regression, and principal component analysis. Biplots, charts of population dynamics, and distribution maps were created to visualize the results. In Poland in the analyzed time period, an increase in the population of all three studied carnivores was observed along with the westward expansion of the territorial range of lynx and wolf, while bear range remained unchanged. The most mean population increase was exhibited by the gray wolf (7.01%), followed by the brown bear (4.78%) and, finally, the Eurasian lynx (2.94%). The population dynamics of the carnivores showed trends over time, with a notable increase in the last decade. The use of multi-year data in modelling enables a better understanding of the mechanisms governing the abundance and distribution of populations of endangered species. This, in turn, facilitates the planning of more effective conservation measures.

Full text

Introduction

General issues

Population studies of threatened animal species are becoming increasingly necessary. Progressive changes in ecosystems, primarily caused by anthropopressure, contribute to changes in the distribution and number of animals. The shifting distribution of animals is natural to some extent, but on their migratory routes they encounter an increasing number of obstacles that results from human activity. These include the expansion of built-up areas, development of road infrastructure, establishment of agricultural crops etc. Such disruptions in natural ecosystems may contribute to the spatial isolation of populations and, consequently, to difficulties in free interbreeding unless appropriate ecological corridors are maintained. The increasing habitat loss and fragmentation create particularly serious problems for large carnivorous mammals, whose proper functioning usually necessitates an extensive home range for living, obtaining food, and reproducing^1–5^.

In the modern world, largely transformed by humans, preserving biodiversity is an extremely important issue. Therefore, a legal instrument and practical solutions were introduced to protect and monitor natural resources^5^. Three large carnivores—the brown bear Ursus arctos, the gray wolf Canis lupus, and the Eurasian lynx Lynx lynx—are protected under international law: the Berne Convention – Appendix II, the Washington Convention, CITES – Appendix II, the Habitat Directive – a priority species, Annex II, IV, and the Convention on Biological Diversity. In Poland, which is a signatory of the above-mentioned conventions and a member of the European Union, the principles of their protection are also specified in the Polish Act on Nature Conservation^6^ and the Ministry Regulation on the Protection of Animal Species^7^. All of the studied species are under strict legal protection in Poland; the brown bear since 1952, the Eurasian lynx since 1995, and the gray wolf since 1998 (since 1995 in some areas of Poland)^8^.

Brown bear Ursus arctos

The brown bear Ursus arctos is the largest terrestrial mammalian predator in Europe^5^. Historically, bears occurred throughout Europe, except on the larger islands. However, their population decreased dramatically in the 19th century due to widespread deforestation and hunting^3,9^. The current status of the brown bear on the IUCN Red List^10^ is of “least concern” and, the global and European population trends are considered “stable,” with an increasing distribution. Nevertheless, their populations are generally small and isolated from each other^3^. In Europe, the preferred habitats of the brown bear are forests, shrublands, grasslands, and wetlands^11^.

In the past, as a forest species, the brown bear occurred throughout Poland. The degradation of its habitats, in addition to its killing despite limited hunting rights, resulted in its gradual disappearance in many regions of the country. After World War II, the distribution of brown bears was limited to the Carpathian region, and their population size was estimated at 10–14 individuals^9^. Presently, the brown bear is endemic to the southern regions of Poland, inhabiting the foothills and mountains of the Carpathian Mountains. The population size of the bear in 2023 was estimated to be 390 individuals^8^.

The brown bear has been under strict protection in Poland since 1952 and is subject to international regulations outlined in agreements and legal acts ratified by Poland, among others the Carpathian Convention on the protection and sustainable development of the Carpathians 2003^9^. Based on Polish internal legal regulations^7^, the brown bear belongs to a category of strictly protected species requiring active protection. Therefore, there is a 500-m protection zone around all dens from November 1 to March 30 each year. In accordance with the Polish Act on Nature Conservation^6^, the killing of bears and other protected species of fauna; their mutilation, capture, and confinement; destruction of their burrows, and removal of cubs from these; and the storage and trade of skins and other parts of dead individuals are prohibited.

Eurasian lynx Lynx lynx

The Eurasian lynx is the largest European felid, occupying extensive forest complexes. As indicated by Sommer & Benecke^12^, the Eurasian lynx was inhabiting the Iberian Peninsula alongside the Iberian lynx (Lynx pardinus) during the late glacial period. However, it subsequently disappeared from the area by the time the Holocene epoch commenced. Over the past five centuries, the species has been subject to a marked decline in Europe, likely as a consequence of deforestation and hunting pressure on both the lynx and its prey species as well as infrastructure development resulting in habitat fragmentation. Lynxes survived in small, fragmented populations, isolated from each other^3^.

As emphasized by the aforementioned authors, the populations of Eastern Europe inhabiting the Carpathian and Baltic regions (namely the Czech Republic, Slovakia, Poland and Russia) are in the worst condition. The majority of Eurasian lynx populations in Central Europe are preserved in the territory of Slovakia and Poland^13^. Currently, the Eurasian lynx does not belong to the globally endangered species, which is reflected in its current status of “least concern” on the IUCN Red List assessment in 2023^10^. However, some lynx subspecies, e.g. Iberian lynx are still endangered^14^.

In Poland, the Eurasian lynx has been under protection since 1995. In accordance with the current Regulation on the Protection of Animal Species^7^, the Eurasian lynx is strictly protected and requires active protection, which involves periodic protection of breeding sites and of the area within a radius of up to 500 m from these sites from the beginning of April to the end of August each year.

Gray wolf Canis lupus

The gray wolf is the largest wild representative of the Canidae family in its geographic range. Currently, wolves in the wild occur mainly in North America, Europe, and Asia from approximately 75° N to 12° N^15^. The gray wolf’s original range has been reduced by approximately 30%, mainly in developed areas of Europe, Asia, the United States, and Mexico, as wolves have been killed for livestock depredation. Since the 1970s, legal protections, changes in land use, and the migration of rural dwellers to cities have halted wolf population declines and contributed to wolf recolonization and reintroductions^15^. At present, due to variation in natural conditions (i.e., climatic, topographic, vegetation) and to the degree of anthropopressure, wolf populations’ statuses vary from extinct (i.e., Ireland, United Kingdom, Japan) to relatively evenly distributed. Wolf populations are increasing in the north-central and western United States, and in much of Europe^5,15^.

The European gray wolf population is described as a metapopulation with several subpopulations, the status and abundance of which are determined by the Large Carnivore Initiative for Europe (http://www.lcie.org/). In Poland^16^, as well as in other European countries^3^, after World War II wolves were viewed as a major threat to livestock and human safety, especially in areas important for agriculture and breeding. Therefore, a variety of actions were initiated with the objective of exterminating the species, including hunting, the culling of wolf pups, and poisoning. Consequently, by the 1970s, the gray wolf had declined significantly across Europe^3,17^. However, it began to be protected in most European countries in 1979, when it was included in Appendix II of the Bern Convention. As a result of legal changes and the efforts of many people, the current status of the gray wolf is of “least concern” as confirmed by the IUCN Red List assessment in 2018^15^. This represents a positive change, as it was considered “vulnerable” for the last two decades of the 20th century. Presently, the general trend of its population is considered “stable,” while the European population has even been characterized as “increasing”^5^. On a global scale, the main threats currently affecting wolf populations are habitat fragmentation and human-wolf conflicts^5^.

The gray wolf is a native species in Poland, and the western border of the continuous range of the Central European population of gray wolf is in the eastern part of the country^18^. On the territory of Poland, the habitats preferred by the wolves include forests, meadows, pastures, wetlands, and marshes. They also sporadically appear on arable land and near human settlements^18^. In Poland, the gray wolf population had dropped to about 100 individuals by 1973^16^. However, the gray wolf has been protected in Poland since 1998 (since 1995 in some areas of Poland) and the effects of this are clearly visible in its population increase^8^. Currently, as stated by Boitani^19^, the wolf population in Poland is stable (Carpathians) or growing (Baltic and Central European areas). It is expected that the abundance and range of the wolf in Poland will continue to grow^18^.

In this study, we analyzed the dynamics of the population size and the distribution of the brown bear Ursus arctos, Eurasian lynx Lynx lynx, and gray wolf Canis lupus in Poland from the moment these data were registered by the Central Statistical Office (CSO) to the present (brown bear: 1965–2023, gray wolf: 1995–2023, Eurasian lynx: 1996–2023). For the years 2000–2023, data were also available by voivodeship. The objective of the present study was to identify trends in the population changes of carnivores in Poland as well as at the voivodeship level, based on multi-year data. The findings from the statistical modeling (e.g. segmented regression and principal component analysis) facilitate a more profound comprehension of the mechanisms governing the abundance and distribution of populations of the analyzed threatened carnivorous species. This, in turn, essential for the planning of more effective conservation measures. The results of the research are discussed in the context of threats and methods of protection of these endangered species.

Methods

Materials

The research was based on the statistical data from the Polish Central Statistical Office (CSO). The population size of the gray wolf Canis lupus in Poland has been recorded since the year 1995, that of the brown bear Ursus arctos population since 1965, and that of the Eurasian lynx Lynx lynx since 1996 (data available in the CSO office, Warsaw, Poland). For the years 2000–2023, data were also available by voivodeship (the database can be accessed via a search tool using the following address: https://bdl.stat.gov.pl/bdl/start).

Statistical analysis

Compound annual growth rate

The year-to-year ratio of population size was calculated to obtain the relative changes in the population size. The mean one-year ratio can be calculated as geometric mean of the annual ratios or by the equation:

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\rm mean\,one}{\text -}{\rm year\,ratio} = \:\sqrt[{{\rm Y}}_{{\!\text R}}]{b/a}$$\end{document}

Y_R_ is the years range, that is the last year number minus the first year number; a is the population size in the first year; b is the population size in the last year. In the event of all values existing and being strictly positive, the geometric mean and equation (Eq. 1.) generate the same result. The distinction between the two is that the equation (Eq. 1.) is robust in scenarios involving missing values. Upon reduction by 100%, this value represents the compound annual growth rate (CAGR)^20,21^.

Regression analysis

An exponential model for the population changes was assumed, Y = b·a^X^, where independent variable X is the year and dependent variable Y is the population size. The parameters a and b are estimated on the basis of an observation database. This model can be estimated by linear regression after logarithmic transformation of the population size. Because trend changes need to be taken into account, segmented regression^22–24^ was used instead of simple linear regression. In this method, the trend is constant within the time intervals and the coordinates of breakpoints were estimated (for different numbers of these points, which created candidate models). The number of justified intervals in the segmented regression (the number of intervals considered in the study was between one and five), and the determination of whether it was indeed necessary to use segmented regression, were tested by the Leave-One-Out Cross-Validation method (LOO-CV)^25,26^ with the Root Mean Squared Prediction Difference statistic (RMSPD)^27,28^ across candidate models. The RMSPD is also known as the Root Mean Squared Prediction Error (RMSPE)^29^, which is the square root of the Mean Squared Prediction Error (MSPE)^30^. This technique allows for the selection of the model with the best prediction ability according to the RMSPD statistic among models with different numbers of intervals of constant trend (Table 1).

Chi-square test

The chi-square test of independence (of two categorical variables) was used to check whether the populations had increased or decreased in the voivodeships proportionally, or whether there were areas in which population growth trends differed. The statistical decision was made with the significance level set at 0.05. The chi-square test was made based on the population distribution in voivodeships in the years 2000 (first year for which the distribution of population was available), 2010, and 2020 (10-years intervals).

Principal component analysis

Principal component analysis (PCA) is a method that allows for the generalization and description of the population distribution for a given time interval. The PCA with variables centering (the voivodeships average was shifted to value zero) but without standardization was provide in the aim to describe population allocation – to describe the number of individuals that by which the population in the voivodeships increased or decreased. For the purpose of comparison to other methods, the PCA that base on standardized population in voivodeship shows changes in percentages^31^ whereas correspondence analysis describes relative deviation from the full proportionality^32,33^. The PCA was based on the population sizes in the years 2000–2023, and in the voivodeships in which the population sizes were not very close to zero (we assumed that it should be least 30 denoted individuals in total over the years).

A polynomial regression was used to ascertain the trajectory on the PCA plot of the population allocation^33^. The polynomials of degree three were used to present the trend with the requisite degree of detail. The independent variable was the year, while the dependent variables were the coordinates derived from the principal component analysis. The curve of predictions was plotted on the PCA biplot.

Calculations

The segmented regression was calculated using the “lm” function in the R programming environment^34^ and the “segmented” function from the “segmented” package^23,35,36^ available for the R software. The chi-squared test was conducted using the ‘chisq.test’ function. The principal component analysis was performed with the “rda” function from the “vegan” package^37^. Finally, the linear regression and its prediction were carried out with the “lm” and “predict” functions contained in base R. Confidence intervals for regression function were calculated by ‘predict’ function contained in the ‘segmented’ package.

Maps

Poland is divided into 16 administrative units named “voivodeships” that constitute the highest level of the administrative divisions. Maps of the distribution of the brown bear, Eurasian lynx, and gray wolf in Poland were developed for the years 2000, 2010, 2020, and 2023, taking into account the voivodeships. The WGS84 contour map was prepared on the basis of the boundaries picture from the online platform Geoportal (https://mapy.geoportal.gov.pl/imapnext/imap/index.html? moduleId=modulGiK&mapview=51.8%2C19.77%2C4000000s) for which the data source is the State Border Register (PRG), Main Office of Geodesy and Cartography, License type: CC BY 4.0. The contour map was coloured in the R software^34^.

Results

Brown bear Ursus arctos

Population size

In 1965, the population size of the brown bear Ursus arctos in Poland was 26 individuals, while in 2023, 390 individuals were recorded, which was the smallest population size among the described carnivorous species. For the entire brown bear population in Poland in the period 1965–2023, the mean change index was 104.78%, which means that the mean increase in population size per year to the year (CAGR_1965-2023_) was 4.78%. According to the result of the segmented regression analysis (Fig. 1), the dynamic of the brown bear population in the analyzed period was not constant and, according to the cross-validation criterion, division into four periods was shown to be optimal (Table 1). The determination R^2^ coefficient was equal to 97.46% for the four periods model. In the first period, i.e., in the years 1965–1977, the regression method used showed a reduction in the number of brown bears in Poland by CAGR_1965-1977_ 3.2% per year. This was followed by an increasing trend toward 2023. The fastest population increase was 13.3% per year, CAGR_1978-1985_, and occurred in the years 1978–1985. The situation then stabilized, and the bear population increased annually by an average of about 3.6% until 2011. The population growth trend intensified again from 2012, amounting to an average of 9.5% per year, which is similar to the period from 1978 to 1985. When depicting a solitary observed value, Poland recorded the highest brown bear population growth in 1980, 2015, and 2000, when it increased, respectively, by 57%, 37%, and 36%, and the greatest reduction was in 1972, 1979, and 2009, when 31%, 25%, and 24% of the population reduced, respectively.

Table 1. The root mean square predictive differences (RMSPD) statistic for interval number in the segmented regression.species1 interval ^a^2 intervals3 intervals4 intervals5 intervals Ursus arctos 0.01230.00820.00980.0048 ^b^0.0085 Lynx lynx 0.01380.0044 ^b^0.00640.00750.0072 Canis lupus 0.02940.00330.0029 ^b^0.01900.0087^a^One interval of segmented regression means simple regression without split into the time intervals. ^b^ Lower values indicate better predictive ability of the regression model.

Distribution

In order to analyze the distribution of the brown bear in the voivodeships, a chi-square test was performed for 2000, 2010, and 2020. As the value of the P statistic was greater than 5% (p = 0.7849), the hypothesis of a uniform proportional population increase in the 10 years intervals for voivodeships was not rejected. For this reason, periodic population allocations between voivodeships (not significant) are not described here.

In the analyzed period, the population of brown bears in Poland increased, but their range did not change. Currently, in Poland, the brown bear occurs only in the southeast, in the Carpathian Mountains. Throughout the entire analyzed period, bears were recorded mainly in two voivodeships: PK, MA, and single individuals in SL and DS (Fig. 2). In 2023, as many as 322 individuals, i.e., 82.6% of the total population, occurred in the PK voivodeship, with the remaining 15.9% in the MA voivodeship.

Fig. 1. An exponential regression of the brown bear Ursus arctos population size using a segmented regression analysis in which the trend (year-to-year percent changes) is constant within the time intervals. The 95% confidence intervals were marked by green lines.

Fig. 2. The population distribution of the brown bear Ursus arctos in Poland broken down by voivodeships in the years (a) 2000, (b) 2010, (c) 2020, and (d) 2023. The WGS84 map was created using Geoportal (https://mapy.geoportal.gov.pl/) data from the State Border Register (PRG), Main Office of Geodesy and Cartography (licence: CC BY 4.0). The map was coloured in R^34^.

Eurasian lynx Lynx lynx

Population size

In 1996, 288 individuals of Eurasian lynx Lynx lynx were recorded in Poland. In 2023, the population size of the Eurasian lynx was 630 individuals. For the total lynx population in Poland in the period 1996–2023, the CAGR_1996-2023_ value was 2.94%.

The segmented regression analysis results indicate that the growth dynamic of the Eurasian lynx population in Poland was not constant (Fig. 3). The cross-validation criterion indicates that the optimal approach in this case is division into two periods (Table 1). The determination coefficient, R^2^, in the segmented regression was equal to 90.59% for the two periods model. In the years 1996–2002, the lynx population decreased by approximately 10.4% annually. Since then, the population of lynx has been increasing at a rate of 6.3% per year, as indicated by the CAGR_2003-2023_.

Fig. 3. An exponential regression of the Eurasian lynx Lynx lynx population size using a segmented regression analysis in which the trend (year-to-year percent changes) is constant within the time intervals. The 95% confidence intervals were marked by green lines.

According to source data^8^, Poland recorded the highest lynx population growth in 2010, when it increased by 34%. However, in the previous and following years the increase was small, at 4% and 2%, respectively. The largest reduction took place in 2001 and amounted to 47%, but it was also an isolated change.

Distribution

When examining the distribution of the Eurasian lynx population in voivodeships based on the results of the chi-square test for compliance of Eurasian lynx distribution in 2000, 2010, and 2020, it was found that the value of the P statistic was below 0.05 (p = 1.9·10^–18^), with a chi-square statistic value of 122.36 and 16 degrees of freedom. This indicated significant local differences in population growth rates.

Fig. 4. Principal component analysis of the Eurasian lynx Lynx lynx population distribution based on the population size in the years 2000–2023 and in voivodeships in Poland. Red arrows - voivodships, circles – years, blue curve - polynomial trend for years.

The PCA (which was only on voivodeships in which European lynx occurred not occasionally) allowed on the observation of the dynamics of the Eurasian lynx population in voivodeships (Fig. 4). The first component, PC1, described 94.9% of the variability of the lynx population in voivodeships in 2000–2023 so it is strictly dominant, and it is mainly connected with the population sizes in the PK voivodeship. In the years 2001–2007, the increase in the number of lynx occurred mainly in the voivodeships with negative PC1: PD and MZ. The total share of the lynx population increased during this time in these voivodeships from 25 to 15 individuals to 65 and 17 individuals, that is with CAGR was 14.6% and 1.8% respectively. In contrast, the lynx population in the PK voivodeship decreased from 179 to 65 individuals, by 13.5% yearly. Then, the trend reversed up to 2016 and the importance of the PD and MZ decreased with CAGR_2007-2016_ equal to 8.2% and 17.5% to 30 and 3 individuals while population in PK increased by 17.9% yearly up to 287 individuals. Until 2023, the population in ZP started occurring (from 0 individuals in 2016 to 87 individuals). Then, until 2023, the Eurasian lynx population occurring in the PK voivodeship reached 48.3% of the total population in Poland in this year. The European lynx number was almost constant in WM and WP. In summary, the largest increase in population size, visualized in the regression chart (Fig. 3), was associated mainly with the PK (48,3%) voivodeship, next the ZP (13.8%) LU (10.2%), MA (9.2%) and PD (8.3%) play jointly similar role.

In the years 2000–2023, the Eurasian lynx mainly inhabited the areas of southern and northeastern Poland, but from the 2017 it could also be found in western Poland (Figs. 4 and 5). The largest number of lynx consistently occurred in the PK voivodeship, for which 304 lynx were reported in 2023 (Fig. 5). In 2023, this animal appeared slightly more numerous in two additional voivodeships compared to 2000: ZP (87 individuals) and WP (17 individuals). Overall, the results of the analyses showed that the Eurasian lynx population in Poland in the years 1996–2023 increased in size and expanded in range.

Fig. 5. The population distribution of the Eurasian lynx Lynx lynx in Poland broken down by voivodeships in the years (a) 2000, (b) 2010, (c) 2020 and (d) 2023. The WGS84 map was created using Geoportal (https://mapy.geoportal.gov.pl/) data from the State Border Register (PRG), Main Office of Geodesy and Cartography (licence: CC BY 4.0). The map was coloured in R^34^.

Gray wolf Canis lupus

Population size

Population of the gray wolf (Canis lupus) in Poland, as well as its growth rate, exhibited the largest increase among the three analyzed predator species. Between 1995 and 2023, its population increased from 756 to 5,046 individuals. The CAGR_1995-2023_ value for the total gray wolf population in Poland in this period was higher than for the brown bear or Eurasian lynx, and amounted to 7.01%.

Fig. 6. An exponential regression of the gray wolf Canis lupus population size using a segmented regression analysis in which the trend (year-to-year percent changes) is constant within the time intervals.

The results of the segmented regression analysis indicated changes in trends in the dynamics of the gray wolf population in Poland. Using the cross-validation criterion, three periods with differing trends were distinguished (Table 1) and the determination coefficient value of R^2^ was equal to 97.70%. The first period comprised solely the years 1995 to 1998, while the subsequent period extending from 1998 to 2009 was characterized by an average annual decline of 3.2% in population size, with the number of individuals decreasing from 1066 to 696. This decline is illustrated in Fig. 6. The most pronounced decline in the population of grey wolves occurred in 2001, with an estimated reduction of up to 36%. In 2001, 2003 and 2009, the Polish population was recorded as being less than 700 individuals, which is more than seven times fewer than in 2023. Secondly, since 2010 there has been a dynamic growth in gray wolf numbers, averaging 16.3% per year. The year 2016 was characterised by the most substantial increase in the number of individuals, with a 44.14% increase in comparison with the preceding year.

Distribution

The results of the chi-square test indicated that the distribution of the gray wolf population in the voivodeships changed between the years 2000, 2010, and 2020. The value of the P statistic was below 0.05 (it was equal to zero with the computational accuracy of the R environment), with the value of the chi-square statistic equal to 1068.6, and 30 degrees of freedom. Therefore, as in the case of the Eurasian lynx, local trends could be identified.

Fig. 7. Principal component analysis of the gray wolf Canis lupus population distribution based on the population size in the years 2000–2023 and in different voivodeships in Poland. Red arrows - voivodeships, circles – years, blue curve - polynomial trend for years.

According to the PCA analysis (Fig. 7), the voivodeships exhibited similar trend in population changes. The first component was dominant (94.8% of population variability) as well as for European lynx, but all vectors of voivodeships pointed on the same direction or they were very close to 0. Nonetheless, the voivodeships were divided into groups due to the presence of disparate trends. This finding was corroborated by the chi-squared independence test. The decision was taken to divide the voivodeships into three distinct groups. The voivodeships population has increased in the PK, LB, ZP, PM and WM voivodeships since the 2001 – and the increase was small at the beginning and has accelerated from the 2015. In the other hand the population in voivodeships WP, LU, PD and DS was constant at the beginning and started increasing from the year 2018 and in this time the population grown was a little greater than for the first group. The final group was constituted by the voivodeships with particularly brief vectors – notably SL, OP and LD – for the PCA which was based on centered values of population sizes for each of the voivodeship. This indicates that the population was almost constant in such voivodeships.

Fig. 8. The population distribution of the gray wolf Canis lupus in Poland broken down by voivodeships in the years (a) 2000, (b) 2010, (c) 2020, and (d) 2023. The WGS84 map was created using Geoportal (https://mapy.geoportal.gov.pl/) data from the State Border Register (PRG), Main Office of Geodesy and Cartography (licence: CC BY 4.0). The map was coloured in R^34^.

The population increase in the first decade was observed mainly in the LB and ZP (Fig. 8). The lack of PK in this group (just as it was for PCA conclusions) was the consequence of the selection of the year 2000 – between 2000 and 2001 the recorded population size decreased drastically in PK (from 436 to 165) to grow since 2001, with accordance to PCA. During the second decade the population grown was observed especially in the PK (by 958 individuals), WP (413), LB (313) and ZP (280). The grown in the PM, LU, KP and PD was between 100 and 200 individuals and in other voivodeships even lower. During the last years of the analyzed period the population has grown mainly in WP and LU (by more than 300 individuals), in LB and PK (about 200), and in ZP, MZ, DS (about 100) – so the grown was observed in the same voivodeships (with the addition of the last two) as in the previous decade. The PK voivodeship maintained an 26.2% share in the gray wolf population, WP 15.3% and LU, LB and ZP each about 10%.

In Poland in 2023, the largest wolf populations were observed in two regions—the northwest and the east (Fig. 8d). The ZP, LB, and WP voivodeships had a total of 1,735 wolves (34.4%), while the PD, LU, and PK voivodeships provided habitat for 2,196 wolves (43.5%). The fewest wolves number were observed in the LD or OP voivodeships, with about 20 individuals in each of them.

Discussion

Population size

The results of the study indicate a rise in the population of the brown bear (1965–2023), gray wolf (1995–2023), and Eurasian lynx (1996–2023) in Poland during the specified period. The largest mean annual increase was the gray wolf population (7.01%), followed by the brown bear (4.78%) and, lastly, the Eurasian lynx population (2.94%) (Figs. 1, 3 and 6).

However, in the analyzed time periods, variable trends in the population dynamics of the studied predators were identified (Figs. 1, 3 and 6). In general, the fastest mean annual population growth occurred in the last decade and was as follows: brown bear: in the years 1978–1985 and 2012–2023, the increase was 13.3% and 9.5% per year, respectively, Eurasian lynx: in 2003–2023, amounting to 6.3% annually, gray wolf: in 2010–2023, amounting to 16.3% annually. The brown bear population has been increasing since 1978, the Eurasian lynx population since 2003, and the gray wolf population only since 2010.

The analyses carried out indicate a growing trend in the brown bear population (Fig. 1), although, its numbers still remain unsatisfactory. A natural issue that makes the rapid restoration of the population difficult is the low reproduction rate^9,11^.

Brown bears are generally solitary animals. Usually, only females with cubs can be observed to form a group. Nonetheless, bears have considerable spatial requirements and their home ranges are highly diverse. It is primarily determined by geographical location (in Europe they range in size from ca. 50 km^2^ to even 12,000 km^2^) and sex (from 58 to 225 km^2^ for females and 128 to 1,600 km^2^ for males, although single young males’ ranges may extend up to 12,000 km^2^). The ranges may also overlap, particularly among related females^9^. Consequently, the next serious threat to the bear is the habitat loss caused by human activity (habitat fragmentation, migration corridors, sports and tourism in the mountains). It is also emphasized that the high density of the human population and the number of urban settlements have a significant negative impact on the presence of bears^38,39^. The disturbance of bears in their refuges and the limited knowledge and social acceptance of this species in many European countries is also problematic^3,5^.

Brown bears are omnivorous—they can actively hunt large ungulates and often catch fish, rodents, amphibians, and insects. They my feed on carrion, while in autumn they regularly consume fruit, acorns, plant roots and tubers, and cereal grains. They can also attack farm animals and eat honey from apiaries^9,40^, which does not endear them to farm owners. There is thus a necessity to promote knowledge about this species in order to be able to take effective and beneficial actions for both animals and people (https://www.conservationevidence.com/^41^). Attempts are being made to mitigate the problem by educating the public and compensating farmers. For instance, damage caused by bears to apiaries, farm animals, orchards, and crops is currently compensated by the state treasury^6^. Despite this, in Poland, the economic importance is currently limited to damage inflicted on farm animals by wolves^42^. A more serious problem seems to be the brown bear’s progressive synanthropization, as described by Jakubiec^43^. This is especially intense in places with heavy tourist traffic, the main attractants being garbage and food discarded by humans. Synanthropized individuals provoke problematic human behavior: on the one hand, panic, and on the other, extreme disregard for danger.

The analysis carried out in this study shows the size of the Eurasian lynx population in Poland is systematically increasing (Figs. 3 and 5a-d). However, the current population size as well as growth of the Eurasian lynx in Poland is deemed unsatisfactory.

In the natural reconstruction of the population of the Eurasian lynx, the threat is mainly human activity; namely, its interference with and transformation of habitats^44–46^. One of the primary challenges in the conservation of lynx in Poland and neighbouring countries is to ensure effective gene flow among neighboring populations^47–49^. Nevertheless, the most frequently raised issues are the provision of sufficiently large home ranges with a sufficient food base and the securing of ecological corridors that enables the free interbreeding^47–49^.

Eurasian lynx is primarily nocturnal, resting and sleeping during the day^44^ and requires a substantial home range, with female territories ranging from ca. 150 to 1,500 km^2^, while males occupy ranges of 100 to 830 km^2^^50^. In the Polish Carpathians, the range is 171 km^2^ on average^51^. The main pray of the lynx in the mixed and deciduous forest zone of the Palearctic are wild ungulates, primarily roe deer Capreolus capreolus, but also red deer Cervus elaphus and chamois Rupicapra rupicapra^44^. In Poland, the red deer is the most commonly predated of the three species specified^52^ and lynxes primarily hunt young individuals of red deer^53^. The impoverishment of the species composition of forests, the removal of old and fallen trees, and the reduction of undergrowth limits the available spaces for rest, reproduction, and care of young, and negatively affects their hunting effectiveness^17^. In order to ensure the protection of the Eurasian lynx population, it is important to carry out further activities aimed at introducing them into forests, creating a network of corridors for safe movement, further monitoring their population status, and popularizing knowledge about this mammal [44, 45, 54, https://www.wwf.pl/zagrozone-gatunki/rys).

Despite existing legal regulations, lynx hunting still occurs and poachers also set traps for roe deer and red deer into which lynx fall while following the trail of these animals^44^. Another threat to the Eurasian lynx is its susceptibility to many parasites, including numerous species of nematodes, cestodes, and trematodes^55^.

In 1998, the legal situation of the gray wolf in Poland changed, and the effects of this are clearly visible in its rapid population increase since 2010 (Fig. 6). In the year 2023 there were 5,046 Gy wolves in Poland (which is about 5 times more than in 1998^8^.

Wolves are social animals and usually live in packs occupying separate territories ranging from 100 to 500 km² in Europe^3^. Jędrzejewski et al.^18^ built a habitat suitability model for wolves in Poland based on large-scale data on wolf abundance in the years 2000–2006 and using GIS tools. The authors found that habitats suitable for wolves covered at least 20%-24% of Poland’s area, while wolves occupied 16% of the country’s area at that time. Based on the empirical relationship between patch size and wolf numbers, it was estimated that the population of Polish wolves may be two-to-three times larger than estimated at that time (in these estimations: 594; according to CSO^8^ data: 715). We currently have an over seven times greater population of gray wolf than in 2006. Can we therefore assume that the wolf population in Poland will no longer increase rapidly, as it has already reached its maximum potential? In Poland, despite an increase in forest cover, the rapid economic development of the country has had negative consequences for the natural environment (habitat fragmentation, pollution etc.). It is likely that the wolf population trend will depend on many factors, but primarily on the size of suitable habitats and the existence of ecological corridors connecting them, on human attitudes, and on emerging legal conditions. However, as Gula et al.^56^ found, existing anthropogenic infrastructure does not necessarily limit the spread of wolves. The study revealed no alterations in the growing tendency were observed during the most recent period (Fig. 6).

It is also important to note that there is still an open and pertinent question regarding the nature of wolves’ interactions with prey populations^15,57^. During the analyzed period, not only did the population of predators increase in Poland, but also that of their prey^58^. In the case of wolves and lynx, these are primarily roe deer, red deer, and wild boars Sus scorfa^57,59^. The only exception to the population growth of these prey species was that of the wild boar, the abundance of which fluctuated due to African swine fever. This significantly influenced the food preferences of wolves and the number of other prey species, including livestock, as was studied in detail by Klich et al.^57^.

As the wolf population increases, the damage caused by wolves also increases, as Śmietana^42^ demonstrated for the Carpathians. Based on the research conducted in this region, the author found that the main livestock victims of wolves were sheep, and that wolves killed about 5% of them annually. According to Śmietana^42^, the growing number of wolves in this region was not the only reason for this; instead, it was largely due to the insufficient protection of livestock against predators. Based on the results, electric fences are the most effective way to limit wolf attacks on farm animals, although the author recommended their use in combination with livestock guardian dogs.

Boitani et al.^15^ also indicates many other issues that can influence wolf population sizes: wolf genetics, their behavior and diseases, and hybridization with other Canis species (including dogs). The tightening of the law had a positive impact on the development of the wolf population in Poland; however, it caused problems for livestock breeders^60^. Therefore, efforts are being made to develop an appropriate strategy that will facilitate the coexistence of humans and wolves^42^.

Distribution

During the analyzed period, the lynx and the wolf expanded their range in Poland, while the bear’s range did not change (Figs. 2, 5 and 8). Species protection and reintroduction activities resulted in the territorial expansion by the gray wolf and Eurasian lynx toward the west (Figs. 2, 5 and 8). In 2023, lynx occurred mainly in southern and eastern Poland and in north-west corner, the wolf occurred mainly in the east of Poland and in the north-west voivodeships, while the brown bear occurred exclusively on the south (Figs. 2, 5 and 8).

The study allowed for the identification of the region with the most stable population growth of bears, lynxes, and wolves. This is of particular significance given that voivodeships exhibiting stable carnivore population growth are those that reliably maintain genetic resources. The Podkarpackie voivodeship serves this purpose for the population of Eurasian lynx (with the largest population and the highest increase among voivodeships in the recent period), brown bear, and gray wolf (with the largest population and stable growth among voivodeships). The north-west voivodeships also supported wolf populations well (with the fastest recent increase and an already relatively large population among voivodeships) (Figs. 2, 5 and 8).

Although the analyses carried out indicate a growing trend in the brown bear population (Fig. 1), its distribution is still limited (Figs. 2a-d). In Poland, brown bears regularly occur only in the southeast, in five refuges in the Carpathians: Beskid Żywiecki, the Tatra Mountains, Beskid Sądecki, Beskid Niski, and the Bieszczady Mountains^39^. They constitute a small part of the Carpathian brown bear population, which occupies the territories of the Czech Republic, Slovakia, Poland, Ukraine, Romania, and Serbia (https://globeproject.pl/en/brown-bear/carpathian-population/population-numbers-and-distribution). Rarely, migrating individuals can also be found in other regions of the Carpathians, and even farther away in the lowlands^43^.

As stated by Fernandez et al.^61^, approximately 20% of Poland is suitable as bear habitat. Of this area, approximately 9% is expected to be suitable for reproduction. In addition to the Carpathian region, this includes fragments in the northwest of the country and in the northeast near Lithuania and Belarus. These areas are extensively forested, with grasslands and shrublands^8,62^, which, according to Fernandez et al.^61^, are conducive to the presence of bears. This region is characterized by a significant presence of formally protected areas^8^, although these are predominantly smaller, more fragmented areas such as nature reserves, Natura 2000 sites, or protected landscape areas, rather than larger, uniform protected areas such as national parks. Jakubiec^43^ suggested the possibility of reintroducing the bear to selected areas, including the abovementioned lowland forests of northeastern Poland and to the Sudetes. However, all these possibilities are limited due to the lack of appropriate ecological corridors connecting these regions of Poland.

New methods are constantly being sought, and attempts being made, to create the most effective brown bear protection programs in Poland^9^. In the context of bear protection, as emphasized by Jakubiec^43^, Ledger et al.^5^, and many others, the spatial isolation of this species is also a problem, including in Poland. A solution to this may be cooperation with neighboring countries^43^, because bears naturally migrate across borders, which creates an opportunity for beneficial gene exchange between subpopulations. The need for transboundary cooperation to increase the effectiveness of bear protection was also noted by, among others, Selva et al.^9^ and Bartoń et al.^63^.

The analysis conducted indicates a systematic increase in the range of the Eurasian lynx population in Poland (Figs. 5a-d). Species protection and reintroduction activities resulted in the colonization of Polish territories by Eurasian lynx toward the west. However, lynx are almost non-existent in the central part of the country (Fig. 5c, d). This is to a large extent the result of the smaller forest area and forest cover, their increasing fragmentation, and greater anthropopressure resulting from the higher population density in the central voivodeships (MZ, LD, WP) than in eastern Poland (PD, PK, MA and LU) and in the ZP voivodeship^58,62^. The relationship between forest cover and the number of lynx was previously discussed by Niedziałkowska et al.^1^ when analyzing data from the National Census of Lynx, conducted in Poland in 2000–2001, and using Geographic Information System (GIS) mapping. The authors found that the percentage of forest cover significantly influenced the occurrence of lynx, which preferred areas where forest cover exceeded 40%. In turn, the lack of corridors between forests, a dense transportation network, and human settlements negatively influenced the occurrence of lynx.

Skorupski et al.^54^ described in detail the course of the successful reintroduction of lynx in western Poland (Figs. 5a-d) (ZP voivodeship) and emphasized the importance of program continuity to maintain this positive trend. The presence of the Eurasian lynx in this part of the country is due to successive reintroduction activities carried out under the supervision of the program facilitated by the West Pomeranian Nature Society in cooperation with the Mammal Research Institute of the Polish Academy of Sciences in Białowieża and the Cultural Center in Mirosławiec, the importance of which has been emphasized by researchers^54,64^. The World Wildlife Fund Inc. Poland has also participated in lynx reintroduction activities in northern Poland since 2018 (https://www.wwf.pl/zagrozone-gatunki/rys). These activities are largely based on the “born to be free” method^65^, in which lynx kits are born in an enclosure located in the forest, which allows contact with nature from the first days of life. This method was also introduced in 2005 in the Pisz Forest in northeastern Poland^66^. A major problem related to the protection of lynx remains the lack of connectivity between the forest complexes of western and northwestern Poland and the forests of northeastern and eastern Poland, due to severe deforestation and dense human population in the central part of the country^1^.

For many years, wolves mainly inhabited the eastern areas of Poland. This was even the case during the first large-scale inventory of wolves and lynx in Poland, in 2001^52^. At that time, the gray wolf situation in Poland was generally assessed as stable. However, only 17 individuals were reported in western Poland^52^. After 2010, wolves began to inhabit new areas in the west (Figs. 8a-d). Nowak et al.^59^ described a significant population increase between 2011 and 2015 when estimating the accuracy of a habitat suitability model for wolves in western Poland. They highlighted the movement of wolf packs through an ecological corridor near the city of Suwałki (northeastern Poland) and through the Białowieża Forest to the northwest and south of Poland. The northwestern voivodeships are characterized by high forest cover, especially the LB voivodeship, where the forest cover is over 49%^62^. The density of wolf prey species is also highest in this part of the country^62^. Jędrzejewski et al.^52^ believe, however, that a better habitat for the wolf is provided by the eastern voivodeships due to their large forest cover and less-developed road infrastructure, and to the migration of wolves from Eastern Europe, where they occur in higher numbers. Despite this, the Carpathians and the PK voivodeship (southeastern part of the country) are still the largest refuge of the gray wolf and the Eurasian lynx in Poland (Figs. 5a-d and 8a-d).

Despite the diverse biological characteristics of brown bears, Eurasian lynxes and gray wolves, the results obtained and extant literature on the subject indicate that the majority of the problematic issues related to their protection can be regarded as being common to these species.

It seems to be critical to provide sufficiently large natural complexes constituting the natural habitats of these species, together with ecological corridors connecting them, in order to enable effective in situ protection. This is necessary to maintain the population of all analyzed species. In the case of lynx, and especially bears, the naturally low reproductive rate and spatial isolation of the populations are problematic, and may negatively affect their genetic structure, consequently increasing their disease susceptibility and mortality^1,3,5,9,43–45,54,55^.

In primeval ecosystems, there is a natural balance in the trophic cascade, including between predators and prey. The demographic success of the human population and the resulting fragmentation and transformation of natural ecosystems have disrupted this balance. A return to a natural state seems to be impossible; the effective protection of endangered species, including the carnivores described in this paper, is thus a demanding task.

Data and statistical challenges

Determining the population size of large carnivores such as bears, lynx, and wolves is a major methodological and organizational challenge, as has been discussed many times in the literature^52,67^. The aim of this study was to determine population trends, not precisely determine the population size of the analyzed species. And the analyses conducted allowed us to achieve this goal.

It should be assumed that data, even officially provided, may be subject to errors and should always be treated only as estimates. In the statistical field, the accuracy of estimation can be achieved through two primary approaches: the number of observations or the precision of observations. The analysis of multi-year data sets, when subjected to appropriate statistical processing, yields reliable information regarding general trends in the number and distribution of the analyzed species. The CSO data advantage was the number of observed years. The stated significance of population trends in the studied large carnivores indicates that the data from the CSO were sufficiently accurate.

In the event of populations being of a reduced size that are challenging to enumerate with precision, the use of one-year data is burdened with a clear error (population spikes of lynx and bear). However, the generalisation of this data in the form of a trend appears to be a reliable source of information.

Implementation of direct conservation measures, monitoring and accurate estimation of population sizes are vital for the effective species protection. Nevertheless, the determination and analysis of long-term trends remains important and separate issue. These methods will certainly continue to be subject to modifications resulting from new concepts and the development of, and access to, new technologies.