Changes in Forest Cover of Municipalities in Poland in 1990–2018

Abstract

The study is part of the trend of searching for research methods to demonstrate changes in forest cover at the level of basic units of public administration with greater precision and accuracy. The purpose of the article is to present, for the first time, changes in forest cover of municipalities in Poland from 1990 to 2018 using CORINE Land Cover (CLC) data. The contributions of this study are threefold. Firstly, using GIS and CLC data (3.1. Forests), multivariate analyses of forest cover changes were carried out for 2481 municipalities for the CLC data collection years (1990, 2000, 2006, 2012, 2018), which showed the temporal and spatial dynamics of changes, with a predominance of deforestation in 1990–2000 and 2012–2018, and afforestation in 2000–2006 and 2006–2012. Secondly, the formal, legal and financial rationale for these changes was indicated. The increase in afforestation was a result of financial incentives under the National Program for Increasing Forest Cover and the EU’s Common Agricultural Policy (under Rural Development Programs—RDPs). Deforestation was related to a decrease in the supply of land for afforestation, the competitiveness of subsidies implemented under RDPs, and statutory liberalization of logging. Thirdly, the main discrepancies between the data obtained from CLC and from the public data collected by Statistics Poland (GUS) and the State Forests were indicated, which ranged from −32 kha to +310 kha, corresponding to percentage differences of 2.3% and 1.8%, respectively. This was mainly influenced by the differences in the complexity and updating of data collected for state and private forests, as well as delays in introducing changes to the land register by the public administration. This work contributes significantly to our understanding of the dynamics of forest cover changes in relation to the actual degree of forestation and deforestation, and the determinants of forest transformation in Polish municipalities, as well as demonstrating the new applicability of CLC data and their limitations related mainly to the generalization of forest cover area.

1. Introduction

Forests are crucial to life on Earth [1]. They perform key ecological functions by stabilizing water in nature, protecting soils from erosion, shaping the global and local climate and preserving plant and animal habitats [2,3,4]. In addition, forests have great socio-economic value due to the improvement of health conditions, diversification of the labor market and industrial production, development of tourism, etc. [5,6,7]. Therefore, forests are crucial resources for implementing sustainable development principles related to sustainable production and consumption, poverty alleviation, food security, protection of biodiversity and climate change [8]. In the global context, this is reflected in the Sustainable Development Goals (SDGs), most notably Goal 15: Life on land, which, in particular, puts forests at the center of the sustainability of terrestrial ecosystems, aiming to “protect, restore and promote sustainable use of terrestrial ecosystems, sustainably manage forests, combat desertification, and halt and reverse land degradation and halt biodiversity loss” (www.globalgoals.org/goals/15-life-on-land/, accessed on 16 February 2024). With regard to the main objective of the 15th SDG and its specific objectives, the protection and restoration of forest resources, cessation of deforestation and increase in reforestation, on the one hand, and balanced management, on the other, come to the fore [4,8].

For forest conservation efforts, the main challenge is to contain or mitigate forest disturbances, which includes both deforestation defined as “the conversion of forest to other land use independently of whether human-induced or not” [9], and forest degradation involving “the long-term reduction in the overall supply of benefits from forests, including timber, biodiversity and other products and services” [9]. Both lead to a number of environmental consequences, such as 6–17% of global CO₂ emissions [10,11,12], soil erosion, wildlife habit loss, reduction in biodiversity, etc. [1,13]. Global Forest Watch estimates that 459 Mha of forest was deforested worldwide between 2001 and 2022, equivalent to a 12% decline in tree cover since 2000 (www.globalforestwatch.org, accessed on 24 February 2024). Although recent studies have found that the global rate of deforestation has decreased [8] (according to the 2020 Global Forest Resources Assessment (FRA), in 2015–2020, to 10 Mha per year [8]), deforestation continues to occur at alarming rates and forest degradation has increased in many regions [14,15]. Alternatively, the forest area can increase when trees are planted on land that was not previously forested (“afforestation”) or when trees grow back on abandoned agricultural or other land (“natural forest expansion”) [8]. Afforestation provides a number of environmental [2] as well as socio-economic benefits [5].

Regarding adequate management and stewardship of forest resources in a sustainable manner, monitoring the condition of the forest and the dynamics of change is an indispensable condition for the rational formulation, implementation and evaluation of the performance of the forest policy of countries and regions by decision-making bodies, but also a guarantee for the preservation of forests for future generations [16]. Successful forest governance requires (1) transparent legislation, (2) effective tools for supervision and control of forest management and (3) sources of information and databases based on reliable and unified reporting [4,6,7]. This is especially relevant for the aggregation and processing of data on changes in forest area due to afforestation, deforestation or conversion of forest land to non-forest land and, in practice, faces a number of limitations [2,3,4,12,15,16,17]. These are caused, among other things, by the complex ownership situation [6], changes in reference units [14], variations in measurement methods [18,19], legal and regulatory changes [20,21] and the implementation of various mechanisms for financing measures [17,22]. This makes the data on forest cover changes heterogeneous and incomparable, which poses a huge challenge for sustainable management of such a valuable resource as forests.

One country where these problems are accumulating is Poland, which ranks among the top European countries (fifth behind Sweden, Finland, Spain and France) with the greatest loss of tree cover. From 2000 to 2020, Poland experienced a net change of 407 kha (3.8%) in tree cover, including stable forest 9.11 Mha, gain 892 kha, loss 485 kha and disturbed 975 kha (www.globalforestwatch.org, accessed on 22 February 2024). This is due to complex property relations sanctioned under the statutory dichotomous division of forests into those owned by the State Treasury (most under the management of the National Forest Holding “the State Forests”—hereafter SF) and those not owned by the State Treasury (mainly owned by individuals), and the related separation of entities that manage and supervise forests, which are, moreover, subject to different levels of public administration and various institutions. In addition, the situation is complicated by changes in legislation and financing conditions for afforestation and conversion of forest land to non-forest, which is mainly related to the use of EU funds after Poland’s accession to the EU in May 2004. All this means that the data collected so far on forests in Poland differ in timeliness, scope and accuracy, primarily depending on the forms of forest ownership, and are often not comparable [19,20].

Therefore, the search for methods and new tools that can be used in measuring changes in forest cover is still timely and desirable. In recent years, image data acquired through sensors on satellites, aircraft and, more recently, drones have become increasingly popular [18]. One such tool is CORINE (CO-oRdination of INformation on Environment). This is a program of the European Community, undertaken in 1985, which was intended to unify agricultural and environmental policies in EU countries. Its objectives were (1) to collect harmonized information on the state of the environment for all Community countries; (2) to coordinate data collection efforts; (3) to ensure data compatibility. One of the program’s components was land cover and land use mapping, implemented as CORINE Land Cover (CLC). The thematic scope of the cyclically collected data and its level 3 detail were adopted from Heymann et al. [23] and embrace all land cover forms found on the European continent, including forests. The land cover data were collected cyclically, which makes it possible to analyze changes that occurred between successive cycles. Poland, along with other Central European countries, joined its implementation in 1993. The area of Poland was covered by all the surveys (1990, 2000, 2006, 2012 and 2018). The universality of the approach and the detail of the databases provide a strong foundation for comparing changes in forest cover dynamically and for any reference units.

The aim of this article is to present changes in forest cover of municipalities in Poland in 1990–2018 based on data from CORINE Land Cover (CLC). The specific objectives are: (1) to demonstrate the dynamics of changes in forest cover in the periods 1990–2000, 2000–2006, 2006–2012 and 2012–2018 using CLC data; (2) to indicate their formal, legal and financial conditions; (3) to compare the results of research from CLC with data from public statistics. The crosstabulation function in ArcGIS Pro 3.2 software was used to calculate the forest area in each of the 2481 municipalities in the analyzed periods, for which CLC data are available (1990, 2000, 2006, 2012, 2018). This work provides insight into the dynamics of forest cover changes, indicating both deforestation and afforestation, as well as net changes in forest area (CLC 3.1 forests) and its share in the total area of each municipality in Poland.

2. Changes in Forest Cover in Poland—Formal, Legal and Spatial Aspects

2.1. Changes in Forest Cover in Poland until 1990

Forests dominated the Polish landscape for many centuries [24,25,26,27]. The area of forested land has decreased significantly, mainly as a result of the development of settlements, which has been associated with deforestation of fertile soils and expansion of land used for agriculture, as well as overexploitation of forests due to the high demand for timber [28]. Poland’s forest cover at the end of the 18th century stood at about 37% and declined to about 23.1% in 1918, and to 22.2% in 1939. The devastation of forests during World War II caused their area to shrink to 6.5 Mha in 1946, when the forest cover reached 20.7%. This process was reversed in the second half of the 20th century, when afforestation of arable land with the most vulnerable soils and wasteland began. The increase in forest area occurred mainly as a result of successive afforestation of non-forest land and, in the case of the State Forests, was aimed at reducing the so-called ‘property checkerboard’ and equalizing field–forest boundaries. The decrease in forest area occurred at a lower rate and generally resulted from the allocation of forest land for non-forest purposes, as well as timber harvesting for the needs of the country’s economy and exports, and various natural disasters [29].

Hence, in 1947–1990, Poland’s forest area increased to 8.8 Mha, and its forest cover rate to 27.8%. However, this was clearly lower than the average forest cover of Europe (32%), as well as the forest cover of other European countries with similar natural conditions (e.g., Germany—30%, Austria—46%) [29]. Therefore, despite the intensive afforestation work carried out in the post-war period, the problem of Poland’s forest cover has not lost its relevance [24,30]. Since the 1990s, increasing Poland’s forest cover has become one of the key points of the state’s environmental and forestry policy [31]. According to the National Program for Increasing Forest Cover (NPIFC) (1995), in 2020, Poland’s forests were to constitute 30% of the country’s area, and, in 2050, they will constitute 33% [32].

In 1990, the forest area totaled 8694 kha (27.8%); by 2000, the increase reached 2%, and, in subsequent periods, it decreased, achieving only 1.0% in 2018 compared to 2012. Forest cover in 1990–2018 increased from 27.8% to 29.6% (1.8%) (

Table 1. Forest land property structure in Poland in 1990–2018.

	Year	Area in kha	% of Area	Changes in kha	Changes in %
	1990	8694	100	n/a	n/a
	2000	8865	100	171.0	2.0
total forests	2006	9026	100	161.4	1.8
	2012	9164	100	137.4	1.5
	2018	9255	100	91.2	1.0
	1990	7219	83.0	n/a	n/a
total	2000	7341	82.8	122.3	1.7
public	2006	7419	82.2	77.9	1.1
forests	2012	7439	81.2	20.2	0.3
	2018	7467	80.7	27.2	0.4
	1990	7174	82.5	n/a	n/a
public	2000	7262	81.9	87.7	1.2
state	2006	7337	81.3	75.5	1.0
forests	2012	7355	80.3	18.0	0.2
	2018	7361	79.5	5.9	0.1
	1990	6805	78.3	n/a	n/a
managed	2000	6953	78.4	148.1	2.2
by	2006	7053	78.2	100.1	1.4
State Forests	2012	7079	77.3	26.2	0.4
	2018	7115	76.9	35.3	0.5
	1990	113	1.3	n/a	n/a
managed	2000	181	2.0	68.0	60.2
by the	2006	183	2.0	2.0	1.1
national parks	2012	185	2.0	2.0	1.1
	2018	186	2.0	0.7	0.4
	1990	1475	17.0	n/a	n/a
	2000	1524	17.2	49.0	3.3
private forests	2006	1606	17.8	82.0	5.4
	2012	1725	18.8	119.0	7.4
	2018	1748	18.9	23.0	1.3
	1990	n/a	n/a	27.8	n/a
	2000	n/a	n/a	28.4	0.6
forest cover in %	2006	n/a	n/a	28.9	0.5
	2012	n/a	n/a	29.3	0.4
	2018	n/a	n/a	29.6	0.3

n/a—not available; kha—thousands of hectares. Data of the GUS derived from https://bdl.stat.gov.pl/bdl/dane/podgrup/tablica, accessed on 21 February 2024.

). These are lower values than for other EU countries. Over a comparable period, forest area in the EU (EU28) increased from 148 Mha (1990) to 162.4 Mha (2020), and the average forest cover from 34.9% to 38.3%, an increase of 3.4% [8]. In 1990, the area of state forests in Poland amounted to 7219 kha (83%) (of which SF covered 6805 kha—78.3%), and private forests 1475 kha (17%). In 2018, public forests occupied 7467 kha (80.7%), of which SF covered 7115 kha (76.9%) and private forests 1748 kha (18.9%). Thus, there was a significant increase in the share of private forests and a decrease in the area of SF, which was the result of the main transition of forests from SF to national parks in 1900–2000 (from 1.3% to 2%) (Table 1). The areas covered by forests in Poland by voivodeship are presented in Figure 1.

2.2. Formal and Legal Framework, Management and Financing

The legal basis for forest management in Poland is the 1991 Forest Act (hereinafter: FA) [33], and in the case of a change in land use, the 1995 Act on the Protection of Agricultural and Forest Land (hereinafter: APAFL) [34].

Forests owned by the state are under the management of the National Forest Holding “State Forests” (SF; abbreviated to “Lasy Państwowe” in Polish), national parks, the resources of the Agricultural Property of the State Treasury (hereinafter: APST) and municipalities. They are supervised by the Ministry of Climate and Environment. The main player is the SF, which are a kind of state within a state, whose management model is based on three levels: (1) the Director General of the State Forests and the General Directorate of the State Forests—the central unit; (2) 17 Regional Directorates of the State Forests (RDSFs); (3) 429 forest districts. Forest districts carry out forest management on the basis of a forest management plan, which is mandatorily prepared every 10 years or more frequently (e.g., in the event of a natural disaster) (Article 18(1,2) [33]). These plans are not based on public consultation. On their basis, logging and afforestation, among other things, are carried out. The supervisory authority for national parks is a director. Management of national parks, as well as other forms of nature protection, is conducted on the basis of conservation plans.

Forests not owned by the State Treasury include those owned by individuals (94.4% in 2018), and the remaining 5.6% are forests of land communities, cooperatives, churches, religious associations, social organizations, companies and other private entities. In this case, forest management is supervised by a representative of the county (poviat)—starosta (Article 5(1)(2) [33]), i.e., the body authorized to make changes in the Land and Property Registry (hereinafter: LPR) concerning, among other things, changes in the use of forest land for non-forest purposes and vice versa. In private forests, forest management is carried out under the control of county administration with the institutional support of forest districts. The bases for afforestation and timber harvesting are simplified forest management plans (hereinafter: SFMPs) (Article 19(2) [33]), or the starost decision based on the inventories of forest condition (hereinafter: IFC), which are prepared for fragmented forests up to 10 ha in area (Article 19(3) [33]). Harvesting of timber not in accordance with the SFMP or IFC is possible in random cases based on decisions issued by the starosta (Article 23(4) [33]). However, the situation is more complex, as the owner of a private forest can cut down trees on their own, then file a notification with the forest district; legalization is carried out by the forest service, issuing a certificate. Therefore, the actual private forest area should be verified during the development or updating of the SFMPs or IFCs once every 10 years (this provision was introduced in 2012), by preparing a list of discrepancies between the registration data relating to the land covered by the SFMP or IFC and the actual state of the land (Articles 7(4.c), 8 (2.c) [35]).

The provisions of the FMP and SFMP regarding the boundaries and area of the forest are included in the LPR (Article 20(2) [33]). In the case of afforestation carried out under EU CAP, the reclassification of land to forestry is carried out after an assessment of the success of forest crops no later than the fifth year after afforestation of the land (Article 13 (7) [33]). Forest owners (private, public) are required to restore forest vegetation (silviculture) in the forests within 5 years of clearing the stand (Article 13 (1.2) [33]). After 1990, three systems of afforestation support were formed. The first system is afforestation carried out within the framework of the National Program for Increasing Forest Cover [32], implemented since 1995. The second is support for afforestation under the 2001 Act on the Allocation of Agricultural Land for Afforestation (hereinafter: Afforestation Act) [36]. The third system is related to funding provided by the European Union (EU) under the Rural Development Plan (2004–2006), and then under the Rural Development Programs (RDPs) 2007–2013 and 2014–2020 [28]. In 2004–2020, the following RDPs were implemented in Poland: (1) afforestation measures (RDP 2004–2006, RDP 2007–2013 and RDP 2014–2020); (2) restoration of forest production potential destroyed by disasters and introduction of preventive instruments (RDP 2007–2013); support for investments increasing the resilience of forest ecosystems and their value for the environment (RDP 2014–2020) [22].

2.3. Databases on Forests in Poland

Polish statistics divide forests into “public” forests, which include forests owned by the State Treasury, and “private” forests, i.e., those not owned by the State Treasury. Already, this dichotomy makes it extremely challenging to obtain complete insight into forest cover changes [20]. In addition, Poland has several units responsible for collecting and processing data and information on forests. Statistics Poland (GUS) (the country’s statistical office) is responsible for public statistics, providing data on forests and forest management within the Local Data Bank (bdl.stat.gov.pl, accessed on 14 February 2024) under the “Forestry and Hunting” section. Depending on the content presented, the data are made available in different breakdowns (NUTS 1—macroregions, NUTS 2—regions (voivodeships), NUTS 3—subregions, NUTS 4—counties (poviats) and cities with county (poviat) rights, NUTS 5—municipalities) and for different periods (from 1998, 2001, 2010, 2022, 2023) and ownership forms. At the same time, the scope and detail in the case of private forests are smaller compared to the characteristics describing public sector forests [18]. In addition, the GUS annually publishes compilation studies entitled “Forestry” and “Statistical Yearbook of Forestry”.

The State Forests—in accordance with the Forest Law—develop a “Report on the State of Forests” annually. The data are collected from the forest districts and compiled by the responsible unit in the structure of the SF by region and for the districts. The State Forests also prepare an annual study, “State Forests in Numbers”. As part of the cooperation between the SF and the Bureau of Forest Management and Geodesy (abbreviated to BULiGL in Polish), a Forest Data Bank portal (bdl.lasy.gov.pl, accessed on 14 February 2024) has been established, with the main goal of providing information on forest management and changes in forest area of all forms of ownership.

2.4. Problems in Data Acquisition, Collation and Comparability

Poland is obliged to report its forest area to the Climate Convention (UNFCCC Kyoto Protocol) and the Food and Agriculture Organization of the United Nations (FAO/UN). Forest area in Poland is annually evaluated as a part of public statistics research. The State Forests reporting is based on a division into 17 RDSFs and 429 forest districts (by 2020—430). Meanwhile, public statistics by the GUS are based on the following divisions: 16 voivodeships (from 2018—17 NUTS 2), 73 sub-regions (NUTS 3), 308 counties (poviats) and 66 cities with poviat rights (NUTS 4), and 2481 municipalities (communes) (NUTS 5). The boundaries of reference units for public statistics and SF mostly fail to overlap. In addition, in 1998, there was a change in the territorial division of the country from a two-tier (voivodeships, municipality) to a three-tier (voivodeships, poviat, municipality) division [37], which makes it even more difficult to aggregate reproducible data. Therefore, in the compilation of forest statistics, there is a migration of data between the General Directorate of the State Forests, the GUS, the BULiGL and the Ministry of Climate and Environment.

Differences between national and international forest definitions lead to differences between actual and reported forest area [38]. According to Art. 3. of the FA [33], a forest is land of compact area of at least 0.10 ha, covered with forest vegetation (forest crops) (or temporarily deprived of it but intended for forest production), covered by forms of nature protection or entered into the register of historical monuments, and moreover related to forest management and occupied by infrastructure used for the purposes of forest management. This differs from the definition adopted in the EU by EUROSTAT, based on the FAO classification system, in which “forest” means land with a stand density (or equivalent degree of forest cover) of more than 10% and an area of more than 0.5 ha, as well as trees that should be able to reach a height of at least 5 m at maturity in situ. Therefore, the total forest cover areas reported in EUROSTAT databases (https://ec.europa.eu/eurostat/web/forestry/database, accessed on 16 February 2024) are 2% higher than the data provided by Statistics Poland (Table 1), which are based on the records of the Land and Property Registry, including land of all forms of ownership with an area of 0.1 ha and, in addition, with a canopy cover of more than 10% and a minimum width of more than 10 m.

The updating of data on forest area and timber resources is carried out by the BULiGL, which was initially for forests under SF management, and, since 2015, has also been for private forests, using data from the Forest Data Bank. It should be noted, however, that in the case of private forests, the updates are conducted on the basis of incomplete data both in terms of the description of the state of the forest and the main use implemented, which are in part taken from already outdated forest management plans, often of unverified quality [39]. This is because SFMPs and IFCs are not drawn up and updated (mandatory updates every 10 years were introduced only in 2012) in counties (poviats) on a regular basis [6,39], which makes actual assessment of forest cover problematic. Although the situation is improving (in 2008, 53% of forests were covered by management documentation; in 2018, 83%), this still applies to more than 20% of private forest areas [6]. In SF, updates to forest areas are conducted in much more detail as part of a supplementary survey before the revision of FMPs [28].

The methods of data updating, as well as the entities responsible for data entry, are changing. In 1983–2001, the updates were carried out under the performance of large-scale inventories of forest health and sanitary condition, and since 2005, a national forest inventory (NFI; abbreviated to WISL in Polish) has been introduced in 4 × 4 km grids with clusters consisting of five plots. Until 2014, measurements were made only on these sample plots, which were located in forest according to LPR records. The third NFI cycle (2015–2019) also included areas subject to the definition of forest but located on non-forest land, referring to instances of LPR “off-register forest area”, i.e., land excluded from agricultural production for many years, on which, thanks to natural succession, forest communities composed of forest-forming light tree species have appeared. The inclusion of areas covered with forest vegetation but which are not forests within the meaning of the LPR and, at the same time, not designated for non-forest purposes aligned with the definition of forest reported by Poland to the Kyoto Protocol, i.e., including areas with at least 10% canopy cover [40]. Until 1999, the entity responsible for entering the source data was BULiGL, and since 2000, the data have been collected by employees of forest districts, mainly using the State Forest Information System (SFIS; abbreviated to SILP in Polish).

Data on forests for the purposes of public statistics are based on the Land and Property Register. It should also be noted that the actual size of the forested area is slightly larger than reported in statistical studies, as reporting is delayed. This is because 5 years after the plantation of forest crops on former farmland and wasteland, after the commission recognizes them as successful, they are entered into the LPR and counted as forests. This also applies to afforestation resulting from natural succession on non-forest land. Delays in the reclassification of afforested land, as well as the natural expansion of trees on abandoned agricultural lands, observed over the last decades, have meant that information from the LPR becomes unreliable [29,40].

All this means that the body of knowledge about forests in Poland is limited and insufficient for the effective shaping of the state’s forest policy, or for programming or evaluating the effectiveness of financial support under the RDPs. Therefore, it is necessary to explore new tools that can make these statistics more detailed.

3. Materials and Methods

Studies of land cover changes, including the coverage of forested areas, are currently carried out mainly using data acquired by remote sensing methods, among which satellite imagery, e.g., Landsat, is very common [41,42,43]. Their high information abundance makes it possible to perform their own land cover classification and a range of analyses based on widely accepted indicators derived from them, e.g., NDVI [41,42]. Meanwhile, the CORINE CLC database, which is a ready-to-use, standardized product, makes it possible to perform both spatial and temporal comparisons in any area of Europe [44] or for a chosen country [45].

This study was based on data obtained from the CLC database compiled through visual interpretation of aerial and satellite imagery (Landsat 4/5. Landsat 7, SPOT 4/5, IRS P6 LISS III, RapidEye, Sentinel 2A/b and Landsat 8), with a detail equivalent to a scale of 1:100,000. Topographic maps (50 and 100k), city plans and, most recently, the Topographic Object Database (BDoT10k) are also used in developing the CLC database. We decided to use the CLC because it is the only database containing complete spatial information on forest cover and its changes in Poland for the period 1990–2018, collected and organized in a standardized manner. Other databases have a lower (300 m) spatial resolution (e.g., CCI-LC dataset) (ESA) or cover a short (2020–2021) period of time (ESA WorldCover) (https://esa-worldcover.org/en, accessed on 10 January 2024).

According to CORINE technical guide [23], forests are “areas occupied by forests and woodlands with a vegetation pattern composed of native or exotic coniferous and/or broad-leaved trees and which can be used for the production of timber or other forest products. The forest trees are under normal climatic conditions higher than 5 m with a canopy closure of 30% at least”. The CLC database includes separated areas of no less than 25 ha, with a width of more than 100 m. In contrast, the change database includes, since 2000, polygons of 5 ha and more. The database is provided by the Chief Inspectorate of Environmental Protection, as two separate shapefiles for each cycle, containing land cover data and the change base. This study used CLC databases from 1990, 2000, 2006, 2012 and 2018, as well as land cover change databases from the 1990–2000, 2006–2012 and 2012–2018 cycles in a basic grid (25 ha) [46]. The second element was a layer with the boundaries of municipalities in Poland provided by the Central Office of Geodesy and Cartography with the national geoportal (www.geoportal.pl, accessed on 13 October 2016), on which 2481 units were separated. As an auxiliary, databases of forest district boundaries obtained from the national geoportal were also used.

The work was divided into three stages. In the first stage, forest plots were extracted from the CLC databases (3.1. Forests) for the individual years 1990, 2000, 2006, 2012 and 2018, for which, according to CORINE guidelines, “The predominant classifying parameter is a crown cover density of >30% or a minimum 500 subjects/ha density, with broad-leaved trees representing >75% of the formation. The minimum tree height is 5 m” [23]. Then, using the crosstabulation tool (crosstabulation, ArcGIS Pro), the total forest area (ha) in each municipality in each analyzed year was calculated separately, and we then referred to the area of municipalities to calculate the share of forest areas in municipalities. In the last stage, changes in forested areas (ha) and their shares (%) in municipalities were calculated in successive cycles (1990–2000, 2006–2012 and 2012–2018) and in the entire analyzed period (1990–2018) by subtracting the values from the end of the cycle from the values from the beginning. The results of these are visualized on maps by municipality, i.e., the smallest units in the administrative division of Poland (Figure 2).

The results obtained from the CLC database were compared with statistical data obtained from studies and databases of the GUS [47,48,49,50,51,52,53,54,55,56] and the SF [57,58,59]. The source data from the Central Statistical Office are L-01—report on public forests (excluding municipal forests and those included in the APST), L-03—report on private forests (individuals and legal entities) and SG-01—municipal statistics: forestry and environmental protection. Data from the databases of the State Forests were obtained from the collection, ‘data on the area of forest land and land intended for afforestation’. Sets from the section ‘forest area’ were used in the following cross-sections: country (NUTS 1), voivodeships (NUTS 2), poviats and cities with poviat rights (NUTS 4), municipalities (NUTS 5) and forms of forest ownership. The data extracted from the CLC database were compared with the data extracted from public statistics compiled in tables.

4. Results

4.1. 1990–2000

The period 1990–2000 was characterized by a slight predominance of deforestation, which covered 1276 municipalities. Deforestation was noted in municipalities located in the mountainous and upland areas of southern and eastern Poland (Dolnośląskie, Opolskie, Śląskie, Małopolskie, Podkarpackie, Świętorzyskie, Lubelskie, Łódzkie) (Figure 3). The deforestation covered the main areas of dense forest complexes with economically valuable forest communities located in the Sudeten and Carpathian Mountains and uplands (Krakowsko-Częstochowska, Radomska, Kielce with the Świętokrzyskie Mountains and Lublin with the Roztocze). A decrease in the share of forests in the municipalities was evident in cities and suburban zones. Afforestation was recorded mainly in the areas of western and northern Poland. Only in single municipalities did the increase or loss of forest area exceed 5%. The most significant changes were recorded in the municipalities of Kuźnia Raciborska (−26.66%) and Bierawa (−24.21%) located on the border of the Śląskie and Opolskie Voivodeships in the Rudy Raciborskie and Rudzieniec Forest Districts, as well as in the Lubliniec Forest District. This was the result of 9000 ha of forest fires in 1992 [60]. In 1205 municipalities, increments of up to 5% were recorded (the largest was 5.08% in the municipality of Drawno in Zachodniopomorskie Voivodeship).

4.2. 2000–2006

The situation changed dramatically between 2000 and 2006 (Figure 4). Most of the country’s municipalities recorded an increase in the share of forests (1964, including 48 with more than 5%). Deforestation covered a total of 519 municipalities. These included individual municipalities from the Warmińsko-Mazurskie, Podlaskie, Pomorskie, Kujawsko-Pomorskie, Zachodniopomorskie, Mazowieckie, Dolnośląskie and Śląskie Voivodeships. The deforestation covered a total of 519 municipalities. These were individual municipalities from the Warmińsko-Mazurskie, Podlaskie, Pomorskie, Kujawsko-Pomorskie, Zachodniopomorskie, Mazowieckie, Dolnośląskie and Śląskie Voivodeships. After EU accession, large-scale cutting down of trees in urban forests was significantly reduced—the exceptions were the largest cities, such as Warsaw, Wrocław, Poznań, Olsztyn, Gorzów Wielkopolski, Kielce and Lublin, as well as some neighboring municipalities where intensive suburbanization processes were observed. The largest increases (more than 10%) were recorded in municipalities located in the RDSF Katowice (in the Forest Districts of Lubliniec, Zawadzkie, Opole, Turawa and Strzelce Opolskie) in the Opolskie Voivodeship. On the other hand, the largest losses were in the municipalities Pisz and Turośl located in the RDSF Białystok (Pisz, Drygały and Rajgród Forest Districts) in the Podlaskie and Warmińsko-Mazurskie Voivodeships.

4.3. 2006–2012

The 2006–2012 period was also dominated by afforestation, which was recorded in a total of 1890 municipalities, 107 of which exceeded 5%, including the municipality of Kuźnia Raciborska, which was deforested the most in the previous period (Figure 5). Deforestation was also stopped in the largest cities (exceptions: Gdańsk, Białystok, Bydgoszcz, Wrocław) and generally also in their suburban zones (exceptions: Poznan, Łódź, Szczecin). Deforestation covered a total of 595 municipalities, mainly from the Mazowieckie, Podlaskie and Łódzkie Voivodeships, where there was a high percentage of private forests. The largest increases in forest shares (more than 20%) were recorded south of Cracow (Myślenice, Sucha, Andrychów), south-west of Rzeszów (Strzyżów Forest Districts), west of Wroclaw (Bolesławiec, Świętoszów) and in the Pisz Forest District (deforested in the previous period), as well as in municipalities located in the highest parts of the Sudeten Mountains (Świeradów, Szklarska Poręba, Śnieżka Forest Districts), while the largest deforestation covered the areas of Łódzkie (Kutno, Brzeziny, Spała, Skierniewice Forest Districts), Mazowieckie (Łąck, Gostynin Radziwiłłów, Grójec Forest Districts), Podlaskie (Rudka, Łomża, Waliły) and Podkarpackie (Jarosław, Krasiczyn, Oleszyce Forest Districts). An interesting example is the area of Beskid Żywiecki mountain (Śląskie Voivodeship), where extreme values were observed—the largest increments in the Jeleśnia Forest District, and the largest losses in the neighboring Ujsoły, Wisła and Węgierska Górka Forest Districts. This was related to the planned reconstruction by the Katowice RDSF of artificial spruce monocultures that were not adapted to the habitat conditions (in 2006–2007, a prolonged drought caused the death of spruce trees, which were cut down and replaced by other species) [61].

4.4. 2012–2018

In the 2012–2018 period, the largest decreases in the share of forests in the municipalities occurred (Figure 6). Deforestation was recorded in 1723 municipalities, while afforestation was recorded in only 758, which mainly affected Lubuskie, the southern part of Lubelskie and some municipalities in the Podlaskie, Warmińsko-Mazurskie and Zachodniopomorskie Voivodeships. Deforestation also affected most of the largest cities (except Warsaw, Gorzow Wielkopolski and Wrocław). Large losses also affected municipalities located in the Czersk, Przymuszewo and Lipusz Forest Districts in the Pomorskie Voivodeship (over 20%) were as a result of the storms that passed through the area in 2017 (in the Lipusz FD alone, 4.3 kha of forest was completely destroyed), and those in the Białowieża and Hajnówka Forest Districts (Białowieża National Park) were as a consequence of indiscriminate logging in 2017 (logging was halted only after a ruling by the EU Court of Justice). Losses in the share of forests were also recorded in the communes of Kościelisko and Zakopane located in the Tatra Mountains, while the largest increases were in the highest parts of the Sudeten Mountains (Szklarska Poręba and Świeradów Forest Districts).

4.5. 1990–2018

A compilation of data from 1990 to 2018 shows the prevalence of afforestation over deforestation (Figure 7). Overall, increases in the share of forest area affected 1758 municipalities with changes of up to 5% and 208 above 5.01%, for a total of 1966 municipalities (

Table 2. The number of municipalities in selected class intervals of changes in the share of forest area in the municipalities in Poland in 1990–2018.

Interval %	1990–2000	2000–2006	2006–2012	2012–2018	1990–2018
5.01–25.0	1	48	107	5	208
0.01–5.00	1204	1914	1783	753	1758
−4.99–0.00	1249	517	572	1708	484
−9.99–−5.00	21	1	17	8	21
<−10.00	6	1	6	7	10

). On the other hand, a decrease in the share of forests was recorded in a total of 515 municipalities, of which 484 stated changes of over 5% of the municipal area, 21 stated −9.99–−5% and 10 stated more than −10%. The largest values of increases in the share of forest area were observed in municipalities located in the Podkarpackie (Strzyżów Forest District), Małopolskie (Myślenice and Sucha Forest Districts) and Dolnośląskie (Szklarska Poręba, Bolesławiec, Świętoszów) Voivodeships, mainly due to afforestation in 2006–2012. In contrast, the largest percentage losses in forest area were recorded in the municipalities of the Pomorskie Voivodeship (located in the Lipusz, Przymuszewo, Czersk, Rytel Forest Districts affected by the storm in 2017),and in the municipality of Białowieża (caused by the felling of trees in the Białowieża Forest in 2017).

4.6. Changes in Forest Cover in 1990–2018 in the Context of Public Statistics and CLC Data

A comparison of source data from the GUS, SF and CLC databases for the analyzed period of 1990–2018 demonstrates significant discrepancies (

Table 3. Comparison of forest land area in Poland in 1990–2018, derived from GUS and CORINE Land Cover datasets.

	GUS			CORINE Land Cover
Year	Area in kha	Changes in kha	Changes in %	Area in kha	Changes in kha	Changes in %	% of Difference
1990	8694	n/a	n/a	9220	n/a	n/a	−5.7
2000	8865	171.0	+2.0	9188	−32	−0.3	−3.5
2006	9026	161.4	+1.8	9395	207	+2.3	−3.9
2012	9164	137.4	+1.5	9705	310	+3.3	−5.6
2018	9255	91.2	+1.0	9699	−6	−0.06	−4.6

n/a—not available; kha—thousands of hectares.

).

The CLC-estimated amounts of forested area are larger in subsequent periods than those reported by Statistics Poland (Table 3). The differences range from 323 to 541 kha. The underestimation of the CLC statistics expressed as a percentage ranges from 3.5 to 5.7%. This is due to the aforementioned method of land classification and the associated delays in introducing changes to the LPR.

5. Discussion

5.1. 1990–2000

The transformation period saw the shift from a centrally planned economy to a market economy, accompanied by the reconstruction and rebuilding of the industry’s branch structure, an increase in market-based trade, the revival of private ownership and intensification of housing construction, especially in suburban zones. Forests played an important role in this process, on the one hand, as a resource for production in the timber, paper and furniture industries and in the construction industry, and on the other, as a reserve of investment land [62,63]. This was facilitated by the liberalization of the legislation on the protection of agricultural and forest land (in 1990 and 1991, the obligation to an agricultural use of agricultural land was repealed), which, on the one hand, influenced the expansion of the area of set-aside and fallow land and, on the other hand, the suspension of the area of land excluded for non-agricultural and non-forest purposes [21,64]. This trend was interrupted by the coming into force of the 1995 APAFL [34], which introduced clear provisions on restrictions on the development of forest land and the conversion of forest land to non-forest land, as well as the conduct of afforestation, indicating additional entities responsible for these measures. In contrast, the total area of clear-cutting in state forests declined between 1990 and 1996 (from 34 kha in 1990 to 29.7 kha in 2000), but still remained at a high level of 30 kha per year (an average of 0.5% in SF) [29,30]. This was mainly due to the growing needs of the dynamically developing timber industry [62].

After 1995, forest policy program documents were formulated and institutional and financial mechanisms to support afforestation began to be implemented. The first was the National Program for Increasing Forest Cover [32], approved in 1995, which assumed extending Poland’s forest cover to 30% in 2020 (about 700 kha needed to be afforested) and to 33% in 2050 (about 1.5 Mha). Initially, the program was implemented within the framework of the 1991 Forest Act [33] regulation defining the land classes indicated for afforestation (IV permanent grassland and V–VI), as well as the principles of financing afforestation for particular types of properties. In order to intensify the afforestation process, further statutory support mechanisms began to be implemented from 1997 (including the obligation to provide private owners with tree seedlings free of charge, purchased from the forest fund of the State Forests). Afforestation was mainly carried out on fallow and set-aside land. During Phase I of the NPIFC implementation (1995–2000), a total of 111.3 kha of land was afforested (11.3 kha more than planned), including 70.1 kha of land owned by the State Treasury (that is, 1% of public state forests) and 41.2 kha of land not owned by the State Treasury (2.7% of private forests) [65].

According to the GUS, afforestation of the first stage of the NPIFC covered 1.3% of the forest area, but according to CLC data, this was only 1.2%. The implementation of this mechanism, according to the GUS/SF, offset the scale of deforestation (an increase in forest cover of 2% (+171 kha), while according to the CLC, there was a decrease in forest cover of −0.3% (32 kha), i.e., logging dominated (Table 3), as confirmed by the CLC data for municipalities (Figure 3).

5.2. 2000–2006

Between 2000 and 2006, an increase in the share of forest area was recorded in 80% of the municipalities in Poland (Figure 4). In the case of private forests, this was influenced by legal changes resulting from the implementation of the 2001 Afforestation Act [36] (in force from 2001 to 2003), which offered further institutional and financial support mechanisms to owners of private farms, and from the diversification of afforestation financing sources associated with Poland’s accession to the EU in 2004. This was reflected in the implementation of Phase II of the NPIFC (2001–2005). In 2001–2003, the area of afforestation exceeded 20 kha per year, which was mainly influenced by afforestation on private land (only in 2002–2003, a total of 19.6 kha). After Poland’s accession to the EU in 2004, under RDP 2004–2006, a very attractive system of subsidies for afforestation was introduced, which allowed the establishment of afforestation, with a minimum area of 0.3 ha (there were no upper limits), on areas of arable land, orchards, meadows and wasteland, and additionally assumed the maintenance of the support system (afforestation bonus) for the next 20 years [65].

Changes in forest cover in public forests were mainly associated with the implementation of the community’s agricultural policy. The significant restriction introduced in 2005 on the transfer of land for afforestation from the State Treasury resources to the SF, together with the restrictions arising from the establishment of Natura 2000 areas (a requirement to comply with EU forestry policies) [29], was reflected in a noticeable reduction after 2004 in annual afforestation under the NPIFC in the SF to an area of less than 13 kha [65]. As a result, out of 120 kha of land planned for afforestation in Phase II of the NPIFC, a total of 95.4 kha of land was afforested (GUS 1.1%; CLC 1%), including 46.3 kha of public (according to the GUS 0.6%; a definite decrease compared to the previous period) and 49.1 kha of private land (GUS 3.1%; an increase compared to the previous period was recorded).

Overall, in 2000–2006, the forest area in Poland increased by 1.8% according to the GUS/SP and by 2.2% according to the CLC (Table 3). The differences between GUS/SP and CLC data were due to five-year (under the FA) delays in entering data in the LPR, mainly in the case of afforestation on private properties. In contrast, the GUS’s statistics include most afforestation under the 2001 Afforestation Act [36], according to which the reclassification of afforested land as forest land occurred in the earliest two years after afforestation.

5.3. 2006–2012

In 2006–2012, an increase in the share of forest area was recorded in 76.3% of the municipalities in Poland (Figure 5). In the analyzed period, the Agricultural Property Agency continued to limit the transfer of land for afforestation to the SF (in 2011, only 61.3 ha), which, combined with the intensification of the implementation of Natura 2000 protected areas, created real difficulties in the implementation of the assumptions of Phase III of the NPIFC (2006–2010), which assumed afforestation of 260 kha (160 kha on public land; 100 kha on private land). Of the 40 kha planned to be afforested (8 kha each year) on the SF lands, a total of 709.3 ha was afforested in 2011, and 552.6 ha in 2012 [65].

On land that was not owned by the State Treasury after 2005, practically all afforestation was carried out with the support of EU funds from the RDPs under the measure “Afforestation of agricultural and non-agricultural land” (in 2008, the possibility was additionally introduced of subsidizing afforestation on land where natural succession of woody vegetation has occurred as a result of abandonment of agricultural production), effectively leading to a decrease in interest in afforestation carried out under the 1991 Forest Act [33] (under NPIFC). These declines were taken into account in the afforestation projections for Phase IV of the NPIFC (2011–2014), which were expected to cover 4.4 kha on public land and 20 kha on private land [65]. Approximately 80 kha of private land was afforested under RDP 2007–2013 [66]. However, the high price competitiveness from direct payments for agricultural crops, the inability to carry out afforestation on permanent grassland (since 2008), the increase in the minimum area of land for afforestation (from 0.3 ha in 2004–2006 to 0.5 ha in 2007–2013), the imposition of a limit of 20 ha per year per owner, the shortening of the period of payment of the afforestation bonus (from 20 years in 2004–2006 to 15 years in 2007–2013), the restrictions arising from the establishment of the Natura 2000 protected areas and the complicated procedures and legal consequences involved in the reclassification of agricultural land into forest land resulted in a steady decrease in the number of farmers interested in afforestation, thus decreasing the area of afforestation. In addition, land reserves for afforestation were slowly being depleted, and the price of land began to rise. This was reflected in the data. In RDP 2004–2006, the average annual afforestation area was about 13 kha (the highest in 2006—14.3 kha and 2007—18 kha, and then there was a decline), in RDP 2007–2013, to about 4.6 kha [66].

According to GUS data, during this period, the forest area growth was 1.5% and, according to CLC data, it was as much as 3.3%. These differences were caused by the fact that the areas of afforestation carried out within the framework of Phase III of the NPIFC and RDPs (especially in 2006 and 2007) had not yet been entered into the LPR and indicated in the GUS/FS data.

5.4. 2012–2018

In 2012–2018, a significant decrease in the share of forest area was recorded in 70% of the municipalities in Poland (Figure 6). The recorded changes reflected a progressive decrease in afforestation, especially on private land financed under RDP (the average annual area of afforestation under RDP 2014–2020 was only about 800 ha) [66]. The lower interest in afforestation was mainly due to a significant decrease in the supply of land for afforestation and an increase in its prices as a result of the entry into force of the amendment to the 2003 Act on the Formation of the Agricultural System (hereinafter: LFAS) [67], which introduced significant restrictions on the trade in agricultural land, as well as the competitiveness of other subsidies implemented under the CAP, mainly direct payments for agricultural production. In addition, the period of payment of the afforestation bonus was shortened (from 15 to 12 years), and its rates were significantly lowered compared to earlier programming periods; moreover, the subsidy (after the amendment of the LFAS) could only be received by a landowner who had the status of a farmer (before that, the owner of land for afforestation did not have to be a farmer). In order to reverse this trend, afforestation support rates were increased by more than 67% from 2019 onwards compared to existing rates [22,66].

Another important factor for the increase in deforestation was the entry into force of the 2016 amendment to the 2004 Nature Conservation Act [68] (so-called ‘lex Szyszko’ after the name of the minister who forced this law, which entered into force on 1.01.2017), according to which a property owner did not have to obtain any consent or register the felling of a tree (regardless of its circumference) if the felling was not related to business activity. This resulted in massive tree cutting on private land, whose owners had not previously obtained official permission to cut down trees, and in non-public forests, which should have an established forest management plan. An estimated 1.5 million trees were cut down in the first two months of the law, of which about 80% came from private land. This was most evident in large cities [69]. Many of the trees were cut down “just in time” to come in before the stricter law (which went into force in mid-2017 and concerned the need to report the logging to the municipality).

The five-fold amendment of the provisions of the Law on Nature Protection (between 2015 and 2018), which regulated the logging of trees and shrubs, exacerbated the collapse and complete lack of control by the responsible public administration bodies, as confirmed by the report of the Supreme Audit Office (SAO; abbreviated to NIK in Polish). Between 2015 and 2018, the reports on tree and shrub logging prepared in municipalities were unreliable and riddled with errors, making them unreliable data for public statistics. Natural compensation for tree and shrub logging was also inadequate [70]. In turn, devastation of forests began in the State Forests due to politically driven requirements to maximize profits from timber sales. This was manifested by a very significant increase in clear fillings (2016—29 kha, 2017—36 kha, 2018—41 kha, i.e., exceeding the post-war maximum of 36 kha reached in 1985) [59]. All these factors meant that the period after 2017 can be described as a massive logging of forests in Poland. Due to the fact that most of these felling were not recorded in any way, the available SF statistics did not overlap with the GUS/SP reports, which indicated different numbers—the situation was completely out of the control of both the SF and GUS. However, this was not reflected in the total balance for the country by the GUS/SP (Table 3), which showed a 1% increase (91.2 kha) in forested areas, which was mainly due to the entry of 2006–2007 afforestation in the LPR. In contrast, the CLC recorded a decrease in forest cover of 0.06% (−6 kha). In this case, the differences were due to the way changes are detected. CLC’s algorithms take into account changes occurring in basic survey fields of 25 ha (and 5 ha from the CLC change base); hence, not all minor changes were included in the overall balance for the period, while, in practice, they were much higher (Figure 6).

5.5. Problems in Data Comparability

The delays demonstrated in this study, which were mainly due to the aforementioned statutory 5–6-year grace periods in entering data on new afforestation areas into the LPR, caused a number of discrepancies, especially in the case of private forests. In the “Inventory of the country’s actual forest cover using existing photogrammetric data”, prepared by the Institute of Geodesy and Cartography, the layers of actual forest cover in 2015 showed that Poland’s forest area was 9,997,698 ha, accounting for 32.0% of the country’s area. In comparison, the forest area according to the GUS in 2014 was 9,197,879 ha, which accounted for 29.4% of the country. The difference in actual forest cover of the country according to the GUS and NFI was 799,819 ha. NFI and GUS data also diverged. The NFI for 2015–2019, which also took into account “non-registered forests”, indicated that the area of total forests in Poland was much larger than that reported by the GUS. The first data showing the scale of this phenomenon established on the basis of research by the Forest Research Institute indicated that the area of “forests outside the register” in 2006 in Poland could have been as much as 300–400 kha. The results of the NFI verified this figure, and today it is considered that it may be as much as 750 kha, of which more than 80% of the area is on private land [19]. Jabłoński et al. [40] indicated that forest cover in Poland is in the range of 32.8−33.0%, depending on the forest definition used (under the LPR or reported by Poland to the Kyoto Protocol), and is almost 3% higher than official LPR data (30.1%). The standard error of forest cover based on two years of NFI data is 0.44. Thus, with 95% probability, the true value of this parameter lies between 31.9 and 33.7% while country thresholds of forest definition are used. The discrepancies highlighted in this study between CLC data and the GUS/SF are mainly due to delays in entering data into public statistics. Thus, the CLC data are more current and robust. This was confirmed in a study by Hościło et al. [71] based on spatial databases, which showed that Poland’s forest cover was already 32% in 2014, and the actual forest area was almost 800 kha larger than that published by Statistics Poland. A study by Jablonski et al. [72] conducted in the Podlaskie Voivodeship indicate similar discrepancies (about 2.4 percentage points more than the public data). The main reasons for the mentioned differences are the spontaneous appearance of trees on abandoned agricultural land [73] and, especially, the aforementioned delays in the reclassification of afforestation into forest land.

5.6. Limitations and Future Research

The search for new methods and data sources for assessing various aspects of forest condition and functioning has intensified in recent years. Studies discussed by Holzwarth et al. [18] demonstrate that methods using high-resolution Earth observations are preferred for more accurate analyses of forest area separations.

This study has referred to changes in forest cover in municipalities in Poland based on data from the entire period implemented by the CORINE project, an approach that has not been taken before. It has also demonstrated discrepancies with data from public statistics. These discrepancies are mainly due to the fact that the minimum forested area delineated in CORINE is 25 ha and its width exceeds 100 m, with a canopy closure of 30% at least. Thus, CORINE ignores smaller forest areas unless the forested parcels are located next to each other and their total area exceeds the mentioned 25 ha or the change in forest area size exceeds 5 ha, which is taken into account by the CLC change base. As a result, the CLC does not record small-area and scattered changes (this applies mainly to private forests), which are recorded by public statistics based on reports from the LPR, which takes into account 0.1 ha areas and 10% canopy cover (CLC 25 ha and 30%). These variations are most evident in the 2000–2006 and especially 2012–2018 periods, when there was an increase in small-area forest clearing [69].

Therefore, future research should focus on analyzing changes in small-area deforestation. In order to demonstrate the validity or limitations of the method used, studies should cover selected test fields in private forests (19% of forested areas in Poland, a problem area in terms of obtaining comprehensive data on forest cover changes) and relate them to data from public statistics in order to check the scale of real discrepancies and demonstrate the opportunity or limitations of using the CLC in this regard. This research should focus especially on the 2012–2018 period when most logging took place in small parcels or in scattered locations.

6. Conclusions

This study is part of the trend of searching for research methods to demonstrate changes in forest cover at the level of basic units of public administration with greater precision and accuracy. This is particularly important in the case of complex forest ownership systems associated with the legally sanctioned division between public and private forests, which results in different authorities responsible for policy, management, control and reporting. Due to the lack of consistent statistical and spatial data on the forest cover in Poland, this study used analysis at the level of municipalities with the application of CLC data, which provided an opportunity to capture actual changes in forest cover at the local level, eliminating any formal and legal limitations of public statistics. The contributions of the work are threefold.

Firstly, the multidimensional and multifaceted analyses of changes in forest cover highlighted their temporal and spatial variability. Two periods showed a predominance of deforestation (1990–2000 and 2012–2018), which intensified in 2012–2018 (70% of municipalities). In the periods 2000–2006 and 2006–2012, there was an increase in the share of forest area in most municipalities in Poland (80% and 76.3%, respectively).

Secondly, the determinants of these changes in each of the analyzed periods were presented. These fluctuations were mainly influenced by formal and legal conditions related to changes in national legislation and adaptation to the requirements of EU policies (mainly CAP) after Poland’s accession to the EU in 2004, as well as the implementation of financial support mechanisms for afforestation from national and EU funds. The most significant impact on increasing the forested area was exerted by the implementation of the NPIFC and the statutory sanctioned by 2001 Afforestation Act institutional support of private entities in this regard. After Poland’s accession to the EU, afforestation was financed by subsidies under RDPs (especially under the formally and financially favorable requirements for 2004–2006). This was reflected in an increase in forest area in most municipalities in Poland between 2000 and 2012. The main factors of the decrease in the dynamics of afforestation included the exclusion of permanent grassland afforestation from RDP support, an increase in the minimum area of an afforested plot to 0.5 ha (2007–2013), shortening of afforestation bonus payments (from 20 to 12 years) and underestimation of its rates, and gradual depletion of land reserves for afforestation. After 2016, a decrease in the supply of land for afforestation and an increase in its prices caused by restrictions on land trade and restrictions on buyers (farmers only) introduced by the amendment of the AFAS and the competitiveness of direct payments to agricultural production implemented under the EU CAP came to the fore. After 2017, as a result of the statutory liberalization of the logging of trees and shrubs under ‘Lex Szyszko’, an uncontrolled process of massive logging of forests began in Poland, which was reflected in decreases in the share of forest area in 70% of municipalities in 2012–2018.

Thirdly, the analysis of changes in forest cover in Poland’s municipalities between 1990 and 2018 revealed significant differences between GUS and State Forests public statistics data compared to CLC data obtained from satellite sources, which ranged from −32 kha to +310 kha, corresponding to percentage differences of 2.3% and 1.8%, respectively. In addition, this study identified their main determinants, including the differences in the complexity and updating of data collected for state and private forests, as well as the 5–6-year delay in introducing changes to the LPR, which constitutes the basis for public statistics.

This study, based on independent CLC data, contributes significantly to our understanding of the dynamics of forest cover changes in relation to the actual degree of deforestation and afforestation processes and the determinants of forest transformation in Polish municipalities. In addition, this study provides an important contribution to the comparability of CLC-based survey methodologies and demonstrates new opportunities to apply the CLC to estimate forest area changes and their limitations, which include, first and foremost, the generalization of forest area coverage. This guides future research, which should focus on analyzing changes within small areas, especially in private forests.