The Swedish Forests — A Biodiversity Analysis — Part 2
A data investigation using Python , Jupyter Notebook and ArcGIS
Despite being regarded as an economic success and even a model for other countries, the Swedish forestry system has brought tremendous negative effects in the landscape and biodiversity of the boreal region. Throughout this article, I will investigate the following transformations:
- Aggressive clear-cut system that has basically replaced natural forests and old-growth forests with plantations of selected tree types.
- Young and thick plantations meant to capitalize on the rapid growth rate of young trees (conifers), while eliminating unprofitable tree species (decidous trees) by selective thinning. Similar decline in other plant species.
- Various measures such as scarring the land, ditching, using pesticides and fertilizers are a regular practice meant to increase production while negatively affecting biodiversity.
- Massive fragmentation of forests and isolation of protected areas (mostly to the north of Sweden), lack of green belts that would allow forest species to maintain stable populations or re-populate exploited areas.
All of the above result in a great threat the biodiversity of this world area, as 1 in 5 forest living species are today under threat in Sweden, and even more species are registering important declines in population.
Today the Swedish ‘forest management’ model is being exported in other regions and the true danger is seeing the exact same process happening in the last remaining natural forests of our planet.
THE ROLE OF FOREST IN SWEDEN
Land Use, Geographic Distribution of Forests
Sweden has the largest forest area in Europe, followed by Finland, and is currently the number 1 producer and exporter of wood and wood-derived products in Europe.
The cool and humid climate, along with the scarcity of humans, has allowed forests to thrive. Currently, 68% of Sweden’s territory is comprised of forest, and most of it is qualified as ‘productive’, meaning it can grow and produce at least 1 cubic meter per year per hectare.
Compared to other countries in Europe, Sweden has not only the largest forest surface, but also the greatest utilization rates of forest (wood removal compared to annual growth). Combined, these two factors increase the pressure on the biodiversity of the entire boreal region.
In technical terms, forest is classified as ‘productive’ — meaning it can grow (read ‘produce’) at least 1 cubic meter of wood per year, per hectare — and ‘non-productive’. Productive forests are distributed pretty much all over Sweden’s territory, except for the northern mountainous regions.
Some important definitions:
Natural Growth Forest: Forests where the trees are not clear-cut and then replanted, but self-regenerated. A natural growth forest naturally has many different species of trees and the variety of wildlife is great. Natural forests are equally exploited in Sweden, unless they have a protected status. Today about 10% of Sweden’s forest are to be considered natural growth forests.
Virgin Forest: Forest that has never been exploited by humans. Usually a virgin forest is dominated by old trees and here we can find the greatest variety of trees species and other wildlife. Today about 0.5% of Sweden’s forests can be considered virgin forests and they are found in natural parks and reservations.
Productive Forest: Technically defined by the capacity to grow (read ‘produce’) at least 1 cubic meter of wood per year, per hectare. Most productive forests in Sweden are exploited and a small percent is protected. Exploited forests are regularly clear cut and artificially replanted with desired tree species (spruce and pine). When clear-cut, only some trees and small areas are left standing. The next generation of growing trees are regularly cleaned of ‘undesired’ tree species — which usually are locally adapted broadleaf trees. In this type of forest, most trees are of the same age and species. The variety of wildlife here is very limited.
Non-productive Forest: Forests that cannot grow more than 1 cubic meter of wood mas per year, due to unfavorable natural conditions, such as harsh climate or difficult terrain. Usually this type of forest is unexploited in Sweden because it is not economically rewarding. Many of the protected areas contain this type of forest, which is a positive factor in making this type of environment relatively favorable for wildlife.
Utilization: Growth versus Harvesting
Economically, forestry has had a great significance for Sweden. In the 1950s it represented about 15% of the national gross product, and then as new industries gained importance, gradually decreased to today’s value of 2% of GDP. As share of exports, the wood industry is still one of the most important — with about 10% of Sweden’s total exports in 2019.
To satisfy the growing demand, the industry has steadily increased the production of wood on more or less the same surface as in the 1950s .
In the last years, the utilization rate (wood removals as proportion of annual growth) has approached 100%, and some some of the reasons for this are: higher natural decay (due to drought and storms), lower natural growth (mostly due to summer draughts) and an over-optimistic prognosis on the future growth of forest.
THE CURRENT FOREST MANAGEMENT SYSTEM
How was all this achieved?
- Reclaiming land from other uses, such as pastures, agricultural land, moors and mires. Overall, about 1.1 million hectares have been reclaimed and used for forestry since 1955. One particular issue here is that these environments which were considered ‘unused’ are actually home for many species who depend on them.
- Creating thick and young plantations: overall, Sweden has increased the density of trees from 525 trees/hectare in 1955 to 700 + trees/hectare in 2017, as a consequence of the ‘rejuvenation’ of forests. As we can see in the graphs below, the number and proportion of conifers (which are the preferred product) increased sharply, while broadleaf trees and generally old thick trees are being kept at a minimum. (This data refers to trees greater the 10 cm in diamater, and clear cut surfaces are excluded).
3. By far the most important method for increasing production is the Introduction of the clear-cut system and the Shortening the forestry cycle to ensure faster returns.
The most striking feature of the Swedish forestry model is that large areas are being clear-cut yearly, and then ‘regenerated’, a process which takes in average 100 years.
In practice this means cutting old growth forests and replacing them with even aged plantations, which are harvested at a progressively younger age.
The Clear-Cut System
There is little data available before 1923, but we know that exploitation has happened since the 1600s, at a lower intensity and pace, using the ‘plucking method’ . This means picking mature trees for felling while having a continuous forest cover. In contrast to cutting large surfaces at once, this method ensures continuity of habitat for all species cohabiting there and very little damage done to the environment while extracting the trees.
Since 1955, a new method was proposed, which entails cutting of large surfaces at once, followed by land preparation, fertilizing and replanting with selected tree species. Initially it was regarded as a ‘health’ measure meant to rejuvenate the old forests of Sweden. Immense areas were cleared, burnt, fertilized and then replanted, resulting in young and thick plantations with a rapid rate of growth compared to an old forest. Today much smaller areas are chopped of every year and ‘rejuvenated’. Small and fragmented, nevertheless amounting to a huge area every year.
Overall in all Sweden, about 800 000 hectares, — a surface equal to the area deforested yearly in Amazonia — , is being cut annually using various methods, and about a third are clear-cut.
The rationality of the system
The surface that is being clear-cut every year represents aproximately 1 % of all productive forests. This means that a full growth-cycle takes about 100 years (with important regional differences), and ensures that there is a steady source of harvest-ready surfaces available (assuming growth is happening at the same rate).
It also means that apart from proteted forests, there will be scarcely any forest older than 100 years old soon. This translates differently for different regions as the south of Sweden is warmer, which makes it possible to harvest trees at a much younger age due to faster growth.
Why the forestry industry doesn’t need old forests
One of the reasons why the clear-cut system has been considered so successful is that it practically yielded more volume of wood per surface in shorter time than previously used methods.
Why? because a clear-cut area (with the right climate conditions) will have a much faster growing rate than a patch with older trees. The following graph shows the average life cycle of trees, separated by counties.
My observations:
- the most expansive age categories in terms of growth (in wood volume) are between 20 and 80 years old. Which means during this time the most volume of wood will be produced.
- The southern countries will show consistently higher growing rates due to warmer climate.
If we also take into account that the forestry industry requires high quality wood, we can start to understand why it prefers to harvest trees in their ‘prime’ years.
The Disappearance of Old Forests in Sweden
Over time, the modern forest exploitation system has resulted in the transformation of original forests to plantations. Most natural forests, and especially old-growth forests, have already been logged or are soon to disappear.
As we can see in the graph below, by adding only the surfaces which are annually ‘clear-cut’, (and ignoring the ones that are only ‘combed’ or thinned), we get to see that only about 18% of all productive forest (including protected areas) has not yet been clear cut at least once in the last 70 years. But they will be surely cut in the next years!
As a direct consequence of this system, the proportion of old-growth forests and so called ‘natural forests’ have become very small, while clear-cut areas and very young forests replaced them.
Today, the most important areas of old forests are still to be found in the north — despite experiencing the most intense exploitation since 1955. The harsh climate, along with the mountainous terrain, acted like a natural barrier protecting some of the original forests.
As per data collected by Skogstyrelsen , a complete mapping of the age of forests has been done with help from satellite imagery and is available here, unfortunately only for 2000 and 2010 the latest. Only productive forests have been mapped so far.
Using this data, along with spatial data showing the exact patches of forests that have been cut between 2010 and today, I derived a map of the current age of forests — plotted on equal cells of 10 square km. For the whole process of getting to this map, see my Jupyter Notebook here.
Accepting that the definition of ‘old forest’ is relative, the official percent of remaining old forests in Sweden are:
- 7 % out of exploited surface is older than 140 years old (varying greatly among regions). As seen in the map below, these old forests are situated mostly in the extreme north.
- 50% of the protected productive forests (in national parks and natural reservations) are older than 140 y.o. However these surfaces are very small compared to the rest.
For all regions, but more evident in the north, since the 1950s there has been a huge increase in the clear cut and young forest up to 40 y.o., from aprox. 26% to 43 % (or an increase of about 4 million hectares). That surface was basically taken from the older forest of 6o to 120 years old.
Today, the average age of exploited forests is about 50 years old in the south (where growth is more rapid) and 60 years old in the north. Let’s compare this to the same numbers at the beginning of the mechanized exploitation.
There are important regional differences. The south (Svealand, Götaland) had been exploited for far longer than the north, making the average age lower, and the forests grow faster, making it profitable to cut them at an earlier age. In Norrland on the other hand, since the intensive exploitation began in the 1950s, the old forests have been targeted for harvesting, resulting in a sharp decrease in the average age by 20 years (for example from 86 y.o. to 66 y.o. today in the most northern area).
What about statistics showing the increase of old forest in Sweden?
Official statistics often point out that old forest is increasing in Sweden, in response to the many critiques about its forest management. Even the average age is showing small signs of improvement in the last years, as shown in the graph presented above.
But an average can hide many variations, especially in a country with such climatic range as Sweden.
Two questions are on my mind:
- Is old forest truly rising in Sweden? And if yes, what age categories are rising?
- What is the geographic distribution of these changes ? Are there regions where this age increase is visible? Are there regions where the age is decreasing, thus indicating an intensified exploitaiton?
So, for the following graph, I counted the surfaces of older forests (older than 80 years old) registered every year and compared them to the previous periods.
It turns out, only since 1985, Sweden has lost a surface of almost 2 million hectares of old forest (older than 80 y.o.) which were not ‘replaced’ by younger forest fast enough. Most of them in the northern counties. Despite the fact that there were some small increases in the surface of old forest 140 y.o. plus, those pale compared to the cumulative loss in the other categories.
To get an idea of how this translates in space, the following choropleth map shows in red the counties which lost the most old forest:
- While the 140 + y.o. forest registers timidly increases in areal, the exploitation targets the 80 to 140 y.o. forests, creating a gap in the overall age structure of Sweden’s forests. This is more evident in the northern regions
- Another observation to be made is that in the ‘old forest’ category are being counted also forest edges and small areas left after clear-cutting, which do not necessarily represent a consistent ‘old-forest’ ecosystem due to their small size and fragmentation.
- Positive trends are visible for the southern areas, but remember that we have very little old forest here to begin with.
Trends At a Regional Level — 2010 to 2024
The overall age of forests is slowly increasing, as well as the surfaces of forest older than 160 y.o., and I expect to see those trends at a local level with great variations.
Since satellite data is available only from 2010 (as far as I know), this next analysis focuses on the differences trends between 2010 and the present (2023) day at a local level, more precisely:
- Are there areas where the age is decreasing constantly indicating an accelarating exploitation (let’s call them hotspots of exploitaiton) ?
- And the opposite, are there areas where exploitaiton is decreasing (coldspots) as indicated by an increasing age?
- Are those hot and cold spots significant (statistically)?
- What drives those trends — or what important variables influence them?
Old Forest versus Climate Zones in Sweden
It may seem natural that an old forest will be abundant in thick, tall trees. However, in the boreal region and in Sweden in this case, the more we go to the north, the smaller and slow-growing the trees are. The season is short and the average temperatures are so low that growth is severely affected.
If we take a look at the average tree diameter in a similar analysis, using the same standardization of 10 km by 10 km cells, we can notice that despite having a lot more old forest in the north, the trees are significantly thinner in those areas:
My observations:
- most old forest (which are concentrated in the north) are not necessarily comprised of thick trees — with an average of 25–30 cm diameter (due to slow growth)
- in the south the climate allows for more rapid growth (thick trees) despite a younger average age.
The graph below represents a summary of all forest patches (12.5 x 12.5 meters in size) with their average tree diameter, based on the same satellite mapping (data is from 2015):
Overall, the Swedish forest landscape is dominated by young trees and clear-cut areas. Old thick trees, which are especially important for providing habitat to endangered species, are becoming the exception.
The map below provides a snapshot of the current situation (2018) in terms of the percent of forests which are in fact either clear-cut areas, or in course of regeneration (defined as having an average height of less than 5 m, thus not being qualified as ‘forest’).
We can visually observe many continuous areas where the percent of clear-cut forests are between 30 and 60 %, in high contrast with areas such as Stockholm where the percent is much lower. Not accidental given the density of population here and the increasing criticism towards the current forest management in Sweden.
Biodiversity in a heavily exploited forest ecosystem
From the biodiversity point of view, old-growth forests and natural forests can support a large number of wildlife species, while clear cut areas and very young forests are practically a biodiversity desert. This is especially true for the most vulnerable (redlisted) species , which have a hard time adapting to an aggressive clear-cut system and are heavily dependent on the resources provided by natural forests.
Numerous researches point out the necessary elements for the preservation of a rich biodiversity in forests are:
According to the artfakta.se database, a total number of 1022 species are dependent of a natural growth forest environment, and 75% of them are redlisted. In comparison, only 32 species can use clear-cut areas as an important environment for their survival.
The same analysis focused on redlisted species shows the huge difference in the biodiversity capacity of natural forests versus exploited forests:
It becomes obvious that in order to protect these endangered species, the preservation of old-growth forests and natural growth forests is essential.
The fragmentation of forests
Currently, we are presented in the official statistics with an average cut surface cut of 4 per property in 2002 and 3.3 hectares in 2021, which may seem quite small. We should keep in mind that:
- about 50% of the Swedish forests are owned by private farmers with small properties, which may explain why the median and average values are small.
- Another explanation for this increased fragmentation is the public pressure. Not being able to see large clear cut surfaces with the naked eye may help ease the pressure on the industry.
I will provide one example here. Since 2000, Skogstyrelsen has made available to the public the actual plots of forest cut in the form of geodata (raster and vector files) — available here.
For this example, I analyzed an area in the county of Jämtland, where by chance there is (or was ) a higher density of old forests (> 140 y.o.) and plotted the cut forest polygons on top of the forest layer. See the actual location here.
The image below shows the progressive fragmentation of remaining old forests in this particular area between 2000 and 2020 (for which data was available):
The colors I used explained:
As we can see, small patches add up year after year and make up huge surfaces where all forest (or almost all) has been cleared with no regard for the impact on the resident wildlife.
Is cutting ‘small’ patches a practice that helps forest biodiversity in Sweden?
It is debatable that cutting relatively small surfaces is beneficial or detrimental to the biodiversity of Swedish forest. I guess the better question is What is left behind rather than what goes away. If the surface cut is surrounded by mature forests and the surface cut is sufficiently small, then the damage can be absorbed by the neighboring areas.
As we can see, small patches add up year after year and make up huge surfaces where no old forest is left. Apart from establishing protected areas (which are highly fragmented), there is no public policy towards creating green belts — continuous cover forests — which are essential for the survival of endangered species.
Back to the numbers, we can notice a slight downward trend for the median surface cut at once (per property). In this example I downloaded and analyzed data from Skogstyrelsen geodataportalen which contained all parcels cut down in Norrbotten since 2001.
However, if we take a look at the maximum surface cut yearly, we can clearly see that it’s still a current practice to cut continuous areas of hundreds of hectares, despite having a median surface cut of only 4 hectares:
Again, we can notice a slight downward trend for the maximum surface cut at once (per property).
While there are positive trends in the sense that both the maximum surfaces and median surfaces tend to decrease as seen in the graphs above, the Total surface clear cut every year continues to increase at an alarming pace. The graph below illustrates this trend in the most northern county Norrbotten:
So we can see that just in Norrbotten, more than 22,000 hectares have been clear-cut only in 2021. This is a bit more than the entire area of Stockholm.
Natural Forests versus Exploited Swedish Forests — a Numbers Perspective
To understand clearly the current state of Swedish forests, in comparison to what they once were, is almost impossible because most of the natural habitats are gone.
But we can take a look at protected forests which are virtually unexploited, and compare their main characteristics:
- Age — mean, structure
- Tree Diameter
- Tree types
- Presence of Dead Wood
Age — mean, structure
The natural, unexploited forest is dominated by old trees, which are slowly replaced by younger ones. In the exploited areas, we can see that most of the old forest is already gone, and has been replaced by clear cut areas and young forests:
A breakdown by region
The difference between exploited and natural forests becomes even more striking when looking at the different regions of Sweden. As highlighted above, the northern regions have only been ‘recently’ targeted for production, which makes even more apparent the destruction that has taken place.
The same region by region approach reveals great differences in the age structure by comparison:
In a natural/protected forest, the proportion of older trees is very high, while young trees are fewer, due to competition from old trees and established vegetation on the ground. However, trees of different ages coexist in the natural forest, ensuring a continuous forest cover, while the oldest trees offer habitat and food to various species via their slow decay.
Tree Species Diversity
Naturally, the proportion of tree species would fluctuate from year to year, depending on natural factors. In contrast, exploited forests are cut and replanted with desired tree species (mostly conifers), while broad leaf trees are thinned out while they are young.
Since the proportion of natural forests is extremely small in Sweden, the closest image we could get is that of protected forests, which are more or less shielded from human intervention. The following graph shows the proportion of the main tree types in the protected forests versus exploited areas.
My observations:
- in natural forests, the tree species proportion fluctuate from year to year, as they adapt to natural factors, such as temperature and rain.
- in exploited forests, human intervention make sure to keep the proportion of desired species constant from year to year, by using methods as clearing out broadleaf and planting conifers.
- in exploited forests, the spruce, followed by the pine tree, are preferred, and therefore planted when a forest area is being clear cut.
- in protected forests, we can notice a trend of increasing proportion of broad leaf trees, especially in the south regions. The explanation for this fact is the rising average temperatures, for which broadleaf trees are more adapted. As mentioned before, broadleaf forests have a much greater capacity to sustain wildlife.
Dead wood lying
- One of the most evident consequence of forestry is the lack of dead wood in various decomposing stages. A quick comparison between protected and exploited forests show the direct effect of forestry on this essential living substrate.
Protected Forests — getting picky with the data
What are we protecting?
Today Sweden has about 6 million hectares of protected land (7 mil. if we include marine waters), which indicates a healthy environmental policy.
However if we look closely at the nature types included in the protected areas, we can notice that half of these protected areas are actually Open Land (Non Vegetated Open Land, Open Wetland, Other Open Vegetated Land), as mapped precisely with the help of satellite data from NMD. These are mostly the tundra-like landscapes in the northern Sweden and other barren terrain which have little economic use (so far).
There is of course value in protecting those areas, however by comparison protected forests holds a disproportionate low percent with 8.5% protected out of all forest surface (productive plus non-productive). To be more exact:
- only 5.5 % for the productive forest is protected (1.3 mil. hectares out of which only 975 thousand hectares have a ‘foreskrift’ that actually forbids forest exploitation)
- In contrast, when counting protected forests, the non-productive forests — which hold no economic value - are counted in with another 1 million hectares included in protected areas which make up 25% of the nonproductive forest area in Sweden.
According to Artfakta.se, Sweden’s forests and agricultural landscapes are home to the highest number of unique species in Sweden, and many of them are already endangered or with declining populations.
Geographic Imbalance and Fragmentation
Geographically, we can observe a great concentration of protected areas in the north of Sweden, and as we move toward the south, those areas become much smaller and fragmented. The map below represents all protected areas, including water and land.
Out of all protected forests, about half lie in the northest 2 counties of Sweden, in Norra Norrland.
While here in the north we have the largest protected forests — averaging 1450 hectares per nature reserve, in the south (Götaland), the average protected forest area has only 75 hectares.
- a choropleth with average forest area per county
- a choropleth with gridcells showing protected forest as % of territory
Many redlisted species are entirely dependent on large areas of old forest and natural growth forests, and many need a warmer climate — according to artdatabanken.se 3700 redlisted species live in the south of Sweden, as compared to only 1700 in the north.
Using another way to graphically represent the protectected forests, the plot below shows all the protected areas as points, in the chronological order of their creation. The size of the data point represent the forest area of each protected polygon, and the y axis represents the latitude (south to north) of their location:
Out of a total of aprox- 5900 protected areas, 4254 of them — this represents 72% of protected areas — have an area less than 1 square kilometers. As mentioned above, the size tends to be smaller as we move towards the south. This indicates a highly fragmented landscape and calls into question the effectiveness of those small areas as a refuge for endangered species.
- the largest continous areas of protected forests in the south
- the largest areas of forests in the south are also still exploited (partially) as revealed by my analysis here —
The state of forests inside protected areas
Many newly added nature reserves have been exploited until recently, therefore the state of their forests is not great.
In the chart below we can see a lot of newer reservations having a high percent of clearcut forests. The size of the bubble represents the total area of forest inside a particular protected area.
A similar analysis reveals that many of the recently added protected areas have an Average Age that is low compared to more established protected areas such as National Parks. This suggests the newly protected areas have been or still are in exploitation, thus a reduced average tree age. It will take many decades until those areas achieve, if it is even possible, the characteristics of a true refuge for widlife species.
Why are some protected forests still cut?
Curious as it may seem, some forests inside protected areas are still being cut — see some examples here and here.
I was quite curious to see the actual areas that were cut despite having a protected status. For this exercise I used the geodata provided by Skogstyrelsen here, who used sattelite mapping to document and make public all cut areas since 2000 — in the form of geospatial data (shapefiles).
Steps in my analysis:
- Download the cut areas as a shapefile (polygons) , as well as protected areas in the same format
- Overlap these two using a geographic information system (QGIS) and Python
- Analyze the data
Results
- Since year 2000, a total of 22,000 hectares forest have been clearcut inside protected areas, most of them inside Nature Conservation Areas and Nature Reserves.
Some of the reasons behind this practice are:
- weak legal framework that still allows exploitation — for instance one particular form of protection, Nature Conservation Area, still legally allows the exploitation of forests in the same manner as it happens everywhere else. Even inside national parks and nature reserves, which are the strongest forms of protection, there is a small percent still under exploitation, as productive forests need a special document called ‘föreskrift’ to actually be exempt from it.
- maintenance cutting — some meant for preserving the health of forests as in the case of storms (notice the spike on the graph corresponding to 2005 when a severe storm felled about 75 million cubic meters of trees), or the bark beetle attack — still under debate if truly an effective measure.
- other reasons
Productive forests versus Protected Areas
Since the economic exploitation of forests is in direct conflict with the protection of forests, I expect to find a negative relationship between those 2 variables.
The maps below show these 2 measures aggregated by the Municipality level, emphasizing the presence of productive forests versus the presence of protected areas.
If only visually, it is obvious to me that a high percent of protected land is only possible where there is little productive forest.
Since a map can offer only a visual hint of the hard data behind, I looked at the correlation between these 2 measure above: is productive forest a negative predictor of the protected area in a county/municipality?
The more productive forest present in an area, the less protected it is.
The vast majority of municipalities have under 10% land protected (242 out of 290 municipalities) and most of those are in the south, where climate conditions would be much more favorable for biodiversity.
tbc