Drought and Suboptimal Habitats Shape Norway Spruce Vulnerability to Bark Beetle Outbreaks in Białowieża Forest, Poland

Abstract

Norway spruce (Picea abies (L.) Karst.) is experiencing large-scale decline across Central Europe, with climate warming and bark beetle (Ips typographus L.) outbreaks as primary drivers. In lowland Białowieża Forest, Poland, spruce occupies a range of habitats that differ in their suitability for long-term persistence. We hypothesized that climate change accelerates spruce decline by reducing resilience in suboptimal habitats and increasing susceptibility to bark beetle outbreaks, with long-term persistence limited to optimal hydrological sites. To address this, we analysed spruce share from 1902–2018, its distribution across suitable versus unsuitable habitats, and long-term climate records in relation to outbreaks. Historical maps, forest site classifications, and meteorological data were used to calculate hydro-climatic indices (HTC, SPEI-12, Selyaninov), and outbreak relationships were tested using Welch’s t-test and point-biserial correlation, including lag effects. Spruce share increased from 12% in 1902 to 27% in 2015 and then declined to 9% by 2018. In 2015, 75% of spruce-dominated stands occurred in unsuitable habitats. Bark beetle outbreaks were significantly associated with drought, with outbreak years showing lower precipitation (–121 mm), reduced Selyaninov k (mean 1.40 vs. 1.61), and more negative SPEI-12 values (–0.48 vs. 0.07) compared to non-outbreak years (p < 0.05). One-year lag analysis indicated drought as both a predisposing and triggering factor. These findings highlight the interaction of habitat suitability and drought as a key driver of spruce decline, supporting adaptive management strategies that retain spruce in optimal habitats while converting suboptimal stands to more drought-tolerant species.

1. Introduction

Norway spruce (Picea abies (L.) Karst.) is one of the most important coniferous tree species in Europe, extending from boreal to temperate regions and forming both natural and managed forests [1,2]. In lowland environments, spruce provides key ecological functions by shaping forest structure, regulating microclimatic conditions, and supporting a wide range of associated biodiversity, including epiphytic lichens, bryophytes, saproxylic insects, and cavity-nesting birds [3,4]. Its dense canopy and shallow root system influence water and nutrient cycling, while its timber has historically been of high economic value [5,6]. Beyond Europe, ecologically similar conifers such as white spruce (Picea glauca (Moench) Voss) in North America and Sitka spruce (Picea sitchensis (Bong.) Carr.) in coastal ecosystems demonstrate the global significance of spruce-dominated forests as providers of carbon sequestration, habitat diversity, and wood resources [7,8]. However, in many lowland landscapes, including Central and Eastern Europe, the dominance of spruce has been shaped not only by natural processes but also by historical management and afforestation, often leading to the establishment of spruce beyond its optimal ecological range [9,10].

In northeastern Poland, Norway spruce is one of the main forest-forming species, influencing the structure and function of multiple forest ecosystems, including those of the Białowieża Forest [11,12]. The current distribution of Norway spruce in the lowland Białowieża Forest reflects not only ecological processes but also the legacy of land-use decisions. Historical management and afforestation policies promoted spruce planting for timber supply, resulting in its expansion into hydrologically suboptimal sites. These choices created a spatial mismatch between the species’ ecological optimum and its actual distribution, predisposing stands to increased vulnerability under climate stress. This is visible in recent decades, where a marked decline in spruce vitality has been observed across its natural lowland range, raising concerns about its ecological role and future persistence in this region [13,14,15,16]. A key driver of this decline is the outbreak of the European spruce bark beetle (Ips typographus L.), which has led to widespread mortality in many stands. Nonetheless, periodic outbreaks of bark beetles are considered a natural component of disturbance dynamics in spruce-dominated ecosystems [14,17,18,19,20,21,22,23,24].

Beyond insect outbreaks, climate change is increasingly recognised as a compounding stressor, especially for species at the margin of their climatic optima. Early assessments of species vulnerability in the Białowieża Forest indicated that spruce was relatively secure, but recent studies have reclassified it among the most endangered species due to its ongoing decline [25,26,27]. These findings suggest that both biotic and abiotic factors may be converging to challenge the long-term survival of spruce in lowland mixed forests. At the same time, biotic pressures, most notably mass outbreaks of the bark beetle, have escalated in frequency and severity under warmer and drier conditions. These outbreaks exploit trees weakened by water deficits and heat stress, leading to widespread mortality in spruce-dominated landscapes. Recent experiences in Central Europe, including the Czech Republic, Germany, and Austria, illustrate how climate-driven interactions between drought and bark beetle outbreaks have caused unprecedented losses in managed and semi-natural spruce forests [28,29,30,31,32].

While numerous studies have focused on the strictly protected core areas of the Białowieża National Park, the dynamics of tree species, especially spruce, in the managed portions of the forest remain less well understood. This is despite the fact that these areas make up the majority of the forest landscape and have undergone a long history of forest management, including planting, thinning, and selective harvesting. Understanding how these interventions have shaped current spruce distributions—and how well they align with ecological suitability—is crucial for evaluating past management and planning future interventions.

Moreover, although development models are frequently used to predict long-term changes in forest structure, their validity under scenarios of disturbance-driven transitions (such as bark beetle outbreaks) is questionable. Similarly, the degree to which current spruce stands occupy suitable habitat conditions—versus being legacies of past management—has not been systematically assessed.

While numerous studies highlight spruce decline in mountain and boreal environments, fewer analyses focus on lowland forests such as Białowieża, where spruce has expanded beyond its ecological optimum. We hypothesise that climate change indirectly accelerates Norway spruce decline by reducing resilience in suboptimal habitats, thereby increasing susceptibility to bark beetle outbreaks, and that long-term persistence is likely only in optimal hydrological sites.

To test this hypothesis, we pursued three objectives:

• Quantify changes in the share of spruce in the managed part of Białowieża Forest over more than a century using historical forest maps;
• Assess the distribution of spruce across habitats classified as suitable or unsuitable based on hydrological and edaphic characteristics;
• Analyse long-term climate records to identify hydro-climatic conditions associated with bark beetle outbreaks, including potential lag effects.

By linking spatial forest composition data with habitat classifications and multi-decadal climate series, this study aims to clarify the relative roles of site suitability and drought in driving spruce decline. Our results provide a basis for adaptive management strategies that balance species conservation with resilience under future climate scenarios.

2. Materials and Methods

2.1. Study Area

The Białowieża Forest is located in eastern Poland, next to the Belarusian border (Figure 1). It comprises 4 separate units: Białowieża National Park and three forest districts named Białowieża, Browsk and Hajnówka. The first unit on this list is under strict protection (IUCN level 2). Three forest districts comprise the managed part of Białowieża Forest with a combined area of 50,559.45 hectares. This was the study area in this research. The region is characterised by a temperate continental climate with mean annual precipitation of 640 mm and mean annual temperature of 6.8 °C (1950–2018), mixed lowland forests, and a mosaic of habitats ranging from dry coniferous forests to wet alder communities.

2.2. Data Sources

This study draws on three main categories of data, each aligned with a specific objective: historical cartographic records for tracking changes in Norway spruce (Picea abies) distribution over time, forest management documentation for habitat suitability assessment, and long-term climatic records for contextualising recent disturbance events, and a detailed description of each can be found below. Additionally, outbreak years (1951–2018) were identified from historical forest pest reports.

2.3. Historical Data

To reconstruct the spatial and temporal dynamics of spruce across the managed part of the Białowieża Forest, we used a series of historical and modern forest maps. These included archival forest stand maps from 1902 and 1948, as well as forest management plans and stand maps from 2015. To assess the impact of the recent bark-beetle outbreak, we have used data from over 500 circular sample plots surveyed between 2016 and 2018.

Maps were sourced from the archives of the Directorate General of State Forests and digitised using ArcGIS Pro 3.5 (ESRI Inc., Redlands, CA, USA). All maps were georeferenced and harmonised to a common projection system to allow comparative spatial analysis. Stand-level data on dominant tree species, surface area, and administrative units were extracted to assess changes in spruce-dominated area over time, classified into spruce-dominated stands (>50% spruce by volume) and non-spruce stands. The total spruce share was calculated for each year.

2.4. Habitat Classification Data

To evaluate the ecological appropriateness of current Norway spruce stands, we assessed their location relative to forest site types as defined in the forest management plans for the Białowieża, Hajnówka, and Browsk Forest Districts. These classifications follow the Polish typological system (Polish: Typ Siedliskowy Lasu), which integrates edaphic and hydrological conditions.

Given that Norway spruce is a species sensitive to both water deficits and excess soil moisture, habitat suitability was primarily determined based on groundwater level and soil water retention characteristics. Even within the same site type, the ecological suitability for spruce may vary depending on subtle hydrological differences, such as soil moisture regime, proximity to drainage features, or seasonal flooding patterns.

For the purposes of this study, we grouped all forest site types into three categories:

• Suitable habitats for spruce: Sites where spruce is ecologically well adapted and typically occurs as a dominant species under natural conditions. This includes site types where spruce is also designated as a target species in the silvicultural objectives.
• Stands lacking spruce: These were stands which were suitable for spruce, but spruce did not dominate there.
• Unsuitable habitats for spruce: All other site types, including those where spruce is present but not ecologically optimal due to shallow or fluctuating water tables, poor drainage, or frequent drought stress. This group also includes mesic and dry broadleaved sites where spruce has been introduced through silviculture but is not a naturally dominant species.

This classification enables us to assess the degree of ecological mismatch between current spruce distribution and habitat potential and to quantify the area of spruce-dominated stands growing under suboptimal or unsuitable conditions.

2.5. Climate Data

To analyse long-term climate trends relevant to spruce physiology and disturbance vulnerability, we obtained meteorological data from the Białowieża meteorological station, operated by the Institute of Meteorology and Water Management (IMGW). The dataset covers the period 1951–2018 and includes monthly and annual mean air temperatures and total monthly and annual precipitation, as well as the number of days with precipitation, snowfall, and snow cover. We used the abovementioned parameters to calculate linear trends and changes in the decades, together with a two-sided p-value based on a t-test to assess statistical significance.

We calculated (1) potential evapotranspiration (PET) using the Thornthwaite method (for latitude 52.7° N) and (2) water balance in order to calculate (3) SPEI-12 (Standardised Precipitation–Evapotranspiration Index, 12-month scale) using December values. Moreover, we calculated (4) Selyaninov hydrothermal coefficient (k) for April–September, calculated as k = 10 × P/∑T for months with mean temperature > 10 °C—an important indicator of drought stress and beetle susceptibility in conifer stands.

2.6. Statistical Analysis

We compared outbreak vs. non-outbreak years using Welch’s t-test (difference in means) and point-biserial correlation (association between binary outbreak variable and continuous climate indicators). Analyses were conducted both for the same-year climate data and with a one-year lag to test for delayed effects. All spatial analyses were performed in QGIS 3.22, and statistical analyses performed in R 4.3.

3. Results

3.1. Historical Trends in Spruce Dominance

Spatial analyses of forest maps from 1902 to 2015 revealed significant changes in the distribution and dominance of Norway spruce (Picea abies) within the managed part of the Białowieża Forest. At the beginning of the 20th century, spruce-dominated stands accounted for approximately 12% of the forested area. By 1948, this figure had doubled to about 25%, reflecting active forest management and regeneration practices favouring spruce. Throughout the latter half of the 20th century, the share of spruce remained relatively stable under continuous silvicultural intervention (Figure 2).

To supplement this long-term reconstruction, recent stand data from over 500 circular sample plots, collected between 2016 and 2018, were analysed. These plots documented the impact of the bark beetle outbreak that began in 2015. The results indicated a drastic decline in spruce dominance, with its share dropping from 27% in 2015 to approximately 9% by 2018, representing a decline of more than 66% in just three years (Figure 2).

3.2. Habitat Suitability and Spruce Distribution

In 2015, spruce-dominated stands covered a total area of 13,383.72 hectares within the study area. This area can be divided into two habitat suitability categories: 3319.24 ha (25%) of spruce-dominated stands occurred on sites classified as suitable for spruce (Figure 3, purple fill and outline), while 10,064.28 ha (75%) of spruce-dominated stands were located on sites considered unsuitable for this species (Figure 3, purple fill and red outline). This habitat mismatch is a signal for vulnerability to climatic and biotic stressors. In addition, 3348.18 ha of suitable sites were identified where spruce was not the dominant species, representing an unused potential for spruce within optimal habitat conditions.

Overall, approximately 75% of existing spruce stands in the managed part of the Białowieża Forest either occur in unsuitable habitats or are not aligned with current silvicultural goals, suggesting that their presence is primarily a legacy of past management decisions rather than ecological appropriateness.

3.3. Climatic Trends Relevant to Spruce Decline

Climate data from 1951 to 2018 showed a statistically significant increase in the mean annual air temperature, consistent with regional warming trends (Figure 4). Mean annual temperature increased by +0.208 °C per decade (p < 0.001), with the warmest year (2015) exceeding the 1961–1990 mean by +2.18 °C. Precipitation totals remained relatively stable over the same period; however, there was a notable increase in the number of rainfall days and a decrease in the number and duration of snowfall events (Figure 4). The number of precipitation days increased by +4.50 days per decade (p < 0.001). Snow-cover days showed a decreasing tendency by −2.61 days per decade (p < 0.001). This led to a shortened period of snow cover, potentially impacting overwintering spruce physiology and bark beetle dynamics.

The Selyaninov hydrothermal coefficient (k), calculated for the summer half-year, pronounced interannual variability but no long-term directional trend and showed 28 years classified as experiencing moisture deficits (Figure 5a). The mean k value hovered near the lower limit of the optimal range (1.56) for spruce, indicating increasing drought stress, particularly in recent decades. Similarly, the Standardised Precipitation–Evapotranspiration Index (SPEI-12), calculated for December to represent the cumulative water balance over the preceding 12 months, displayed marked year-to-year fluctuations without a consistent trend. Negative SPEI values, indicating below-average moisture availability, occurred in 29 years during the study period (Figure 5b). The mean annual SPEI-12 was –0.18, suggesting slightly drier-than-average conditions overall, with several pronounced drought episodes concentrated in the last three decades. These negative anomalies align with periods of increased bark beetle activity, implying that extended moisture deficits may have contributed to reduced spruce resilience.

3.4. Climate Conditions and Bark Beetle Outbreaks

Analysis of long-term meteorological data revealed that bark beetle outbreaks in Białowieża Forest were consistently linked to drier-than-average conditions. Outbreak years were characterised by markedly lower hydro-climatic balance, with the Standardized Precipitation–Evapotranspiration Index at a 12-month scale (SPEI-12, December values) averaging −0.30, compared with +0.22 in non-outbreak years (

Table 1. Comparison of mean values for climatic variables between outbreak and non-outbreak years of Ips typographus in Białowieża Forest, Poland. Values represent annual means or totals for the indicated variables. Columns show group means, Student’s t-Test (p-value), and point-biserial correlation (pbc) with corresponding significance levels.

Variable	Outbreak Mean	Non-Outbreak Mean	p-Value	pbc
Mean temp [°C]	6.7	7.1	0.1719	−0.17
Total precip [mm]	607.6	666.5	0.0317 *	−0.26 *
Days with rain	114.6	123.6	0.0572	−0.23
Days with snow	56.1	55.1	0.7664	0.04
Selyaninov	1.38	1.62	0.0087 *	−0.32 *
SPEI12	−0.30	0.22	0.0407 *	−0.24 *

* Statistically significant at p = 0.05.

). This difference was statistically significant and indicated that outbreaks tended to occur during or immediately following prolonged drought periods. Similarly, the Selyaninov hydrothermal coefficient (k) for the growing season was significantly reduced in outbreak years (mean 1.38) relative to non-outbreak years (mean 1.62), pointing to reduced moisture availability during the April–September period. Annual precipitation totals were also lower during outbreaks, declining from an average of 667 mm in non-outbreak years to 608 mm in outbreak years.

When climatic variables were lagged by one year (

Table 2. Comparison of mean values for climatic variables lagged one year between outbreak and non-outbreak years of Ips typographus in Białowieża Forest, Poland. Values represent annual means or totals for the indicated variables. Columns show group means, Student’s t-Test (p-value), and point-biserial correlation (pbc) with corresponding significance levels.

Variable	Outbreak Mean	Non-Outbreak Mean	p-Value	pbc
Mean temp [°C]	6.8	7.0	0.5118	−0.09
Total precip [mm]	591.7	679.7	0.0010 *	−0.38 *
Days with rain	116.2	122.5	0.1844	−0.16
Days with snow	52.2	58.1	0.0554	−0.23
Selyaninov	1.40	1.61	0.0223 *	−0.28 *
SPEI12	−0.45	0.35	0.0009 *	−0.38 *

* Statistically significant at p = 0.05.

), drought conditions in the preceding year still showed a measurable association with outbreak occurrence. This suggests that moisture stress may act both as a direct trigger, reducing spruce physiological resistance to bark beetle attack, and as a predisposing factor when it occurs in consecutive years.

4. Discussion

The spatial reconstruction of spruce distribution confirms that the expansion of Picea abies in the Białowieża Forest during the 20th century was largely anthropogenic. The doubling of spruce-dominated area between 1902 and 1948 coincided with periods of intense forest management, including post-war salvage logging and replanting. Our findings align with previous studies [25,33] that highlight the role of clear-cutting and artificial regeneration, often involving non-local spruce provenances, in reshaping forest composition.

This management-driven shift also resulted in the establishment of spruce stands on suboptimal or unsuitable habitats, as our habitat suitability analysis clearly demonstrates. The legacy of these practices is now manifesting as a structural vulnerability within the forest landscape.

Mismatch Between Habitat and Spruce Presence

One of the most critical findings of this study is the widespread presence of spruce on ecologically unsuitable sites. More than 70% of spruce stands occur on habitats where this species is not well-adapted, primarily due to hydrological conditions or soil constraints. These stands are likely more susceptible to drought stress and bark beetle infestations—a vulnerability already being realised in the wake of the 2015 outbreak.

At the same time, the identification of large areas with suitable habitat where spruce is absent suggests a decoupling between ecological potential and realised distribution—a mismatch shaped by management history rather than natural succession. This has significant implications for adaptive silviculture and calls for a reassessment of species–site matching in future planning.

Our analysis of long-term climate data reveals trends that are increasingly unfavourable to spruce: rising temperatures, reduced snow cover, and more frequent moisture deficits. These changes erode the competitive advantage of spruce in many lowland forest ecosystems and may shift the balance in favour of more drought-tolerant species. Importantly, such climatic stress likely interacts synergistically with biotic threats, including bark beetle outbreaks, further accelerating spruce decline—a phenomenon now documented across Central and Eastern Europe [13,14,16,23,34].

The hydrothermal index results also support the hypothesis that the Białowieża Forest is already at the lower climatic tolerance limit for spruce. Management strategies should, therefore, account for these bioclimatic constraints and avoid reinforcing maladapted stand structures.

The collapse of spruce-dominated stands is not only a loss of a single species but also a structural shift in forest communities. Spruce plays a critical role in shaping microclimates, soil conditions, and deadwood availability—elements vital for numerous forest-dependent species. The simplification of forest structure documented by Brzeziecki et al. [25] and Drozdowski et al. [35] may be a direct consequence of spruce mortality and its insufficient replacement.

Our findings support the hypothesis that climate change indirectly accelerates spruce decline in Białowieża Forest by reducing resilience in suboptimal habitats, which in turn increases susceptibility to bark beetle outbreaks [36,37]. This interaction points to the importance of habitat-based risk assessments in forest management. In habitats where spruce is currently dominant but classified as unsuitable, active conversion to more drought-tolerant native species (e.g., oak, hornbeam, pine) could reduce vulnerability to future disturbances. In contrast, spruce should be retained and promoted in ecologically optimal habitats, where its long-term persistence is more likely under projected warming.

Given that climate trends in the region already show significant warming and altered precipitation patterns, management strategies must account for the increasing frequency and severity of drought events. Integrating high-resolution habitat maps with climate risk indicators such as SPEI-12 and Selyaninov k could help prioritise adaptive interventions and improve the resilience of mixed forest landscapes.

This study provides a spatial and ecological assessment of long-term changes in Norway spruce distribution in the managed part of the Białowieża Forest, but it is subject to several limitations that should be acknowledged.

First, our habitat suitability classification simplifies a complex ecological gradient primarily based on site type and groundwater conditions. While this approach is consistent with silvicultural typology and supported by known ecological thresholds for spruce, it does not fully capture within-type microhabitat variability, particularly in transitional or mixed stands. In practice, spruce performance may also be influenced by local topography, microsite hydrology, and soil properties that are not resolved in the current dataset.

Second, while we utilised a robust series of historical and contemporary maps to reconstruct spruce dynamics, historical cartographic data are inherently limited by differences in mapping standards, species identification accuracy, and scale. Despite careful georeferencing and harmonisation, some inconsistencies in spatial precision across time periods are likely.

Third, although we analysed long-term climate data and linked them to potential stressors contributing to the bark beetle outbreak, we did not model bark beetle population dynamics or include field data on infestation severity at the stand level. This limits our ability to quantify causal links between climatic anomalies, stand conditions, and outbreak intensity. Our interpretation of climate–disturbance interactions is therefore correlative and hypothesis-generating rather than predictive.

Finally, the study does not include genetic analysis of spruce populations, which could help differentiate between local and introduced provenances [38]—a potentially important factor in understanding the differential resilience of spruce across habitat types.

Despite these limitations, the study offers a meaningful contribution by linking long-term land use, ecological suitability, and climate pressures to recent forest decline patterns. Future work should integrate high-resolution hydrological data, genetic origin studies, and process-based models to better predict stand-level outcomes under changing environmental conditions.

5. Conclusions

This study provides quantitative evidence that the large-scale Norway spruce decline in the Białowieża Forest is driven by the interaction between habitat suitability and drought-related climate extremes. By combining over a century of spruce distribution data with habitat classification and multi-decadal climate records, we demonstrate that

• A substantial share of spruce in the forest occurs in ecologically unsuitable habitats, making it inherently vulnerable to moisture deficits;
• Bark beetle outbreaks are strongly associated with multi-month drought indicators (negative SPEI-12, low Selyaninov k), confirming that hydro-climatic stress is a consistent outbreak trigger in this lowland temperate system;
• Drought in the year preceding an outbreak may act as a predisposing factor, amplifying the risk even under less extreme current-year conditions.

Our study demonstrates that the decline in Norway spruce in the Białowieża Forest is primarily driven by the interaction of habitat suitability and hydro-climatic variability, which together amplify bark beetle outbreaks. Spatial analyses revealed that nearly two-thirds of spruce-dominated stands are located on sites classified as unsuitable, while large areas of suitable habitats remain underutilised. This spatial mismatch reflects historical land-use choices that prioritised spruce afforestation beyond its ecological optimum, increasing long-term vulnerability under warming conditions.

From a land-use and sustainability perspective, these findings highlight the need for adaptive forest management. We recommend that spruce be retained only on hydrologically suitable sites, while suboptimal stands should be gradually converted to more diverse, drought-tolerant species. We propose a proactive adaptation strategy combining (i) habitat-based zoning to identify persistence areas, (ii) assisted conversion in vulnerable sites, and (iii) continuous climate–outbreak monitoring to inform flexible silvicultural responses. Such measures can reduce economic losses, preserve biodiversity, and align regional forestry with EU climate adaptation goals.