Economic losses caused by butt rot in Norway spruce trees in Norway

ABSTRACT Butt rot is a main defect in Norway spruce (Picea abies (L.) Karst.) trees and causes large economic losses for forest owners. However, little empirical research has been done on the effects of butt rot on harvested roundwood and the magnitude of the resulting economic losses. The main objective of this study was to characterize the direct economic losses caused by butt rot in Norway spruce trees for Norwegian forest owners. We used data obtained from seven cut-to-length harvesters, comprising ∼400,000 trees (∼140,000 m3) with corresponding stem profiles and wood grade information. We quantified the economic losses due to butt rot using bucking simulations, for which in a first case, defects caused by butt rot were included, and in a second case, all trees were assumed to be free of butt rot. 16% of trees were affected by butt rot, whereby butt rot tended to occur in larger trees. When butt rot was present in a tree, the saw log volume was reduced by 48%. Proportions of roundwood volume affected by butt rot varied considerably across harvested stands. Our results suggest that butt rot causes economic losses upwards of 7% of wood revenues, corresponding to € 18.5 million annually in Norway.


Introduction
Butt rot is one of the most significant defects in Norway spruce (Picea abies (L.) Karst.)roundwood and has a considerable economic impact on forest owners in the northern hemisphere (Asiegbu et al. 2005).The main causes of butt rot are the fungi Heterobasidion parviporum (Fr.)Niemelä & Korhonen and Heterobasidion annosum (Fr.)Bref (Woodward et al. 1998;Pukkala et al. 2005).Other fungi, such as Armillaria spp.(Keča et al. 2009;Brglez and Ogris 2019) and Phellinus spp.(Hansen and Goheen 2000) can be prevalent, but are considered to have a smaller economic impact (Asiegbu et al. 2005).The fungi spread between trees either through root contact (Piri and Korhonen 2008) or by releasing basidiospores into the air, which can then infect fresh wood of trees through stumps and wounds, for example after thinning or clearcutting (Rishbeth 1951;Redfern 1998).After harvesting a diseased stand, the fungi can survive in dead stumps for decades, and infect the next forest generation (Piri 1996).Besides downgrading the wood quality, butt rot leads to yield losses due to increment reduction and increased mortality (Hellgren and Stenlid 1995) and increases a stands' vulnerability to wind damage (Honkaniemi et al. 2017).
Butt rot primarily affects the bottom-most part of the tree stem, which is the most valuable section for wood production (Kärhä et al. 2019).Even when considerable butt rot is observed, tree crowns typically show no symptoms of yellowing or defoliation (Szczepkowski et al. 2022).Norway spruce makes up around 73% of roundwood volume harvested in Norway (Statistics Norway 2022), highlighting its importance for the forest industry.In a study assessing the frequency of butt rot in Norway spruce, rot columns were found to reach heights upwards of 10 m (Stenlid and Wästerlund 1986).Sawn wood is required to be free of butt rot, and only limited proportions of rot are accepted in pulpwood assortments (Norsk virkesmåling 2015).Therefore, butt rot is widely considered the main cause of downgraded wood of Norway spruce trees.
At European level, annual economic losses caused by butt rot have been approximated at € 800 million (Woodward et al. 1998), of which the Norwegian share has been suggested to comprise around € 10 million (Solheim 2010).However, these reported losses may be considered generalized approximations based only on national statistics on roundwood removals and simplified assumptions.Little is known about how butt rot affects different tree dimensions, and by how much saw log volumes are reduced.Although butt rot is considered a main cause of value reduction in Norway spruce roundwood, a detailed analysis of the magnitude of the consequent economic losses is missing to date.
In Scandinavia as well as other parts of the world, the vast majority of roundwood is harvested by cut-to-length harvesters, which record large amounts of standardized production data during operation.Sensors mounted on the harvester head record log dimensions and lengths (Nordström and Hemmingsson 2018), and each cross cut is allocated a time stamp.The operator selects and records tree species and wood grades, providing information on the characteristics and quality of the harvested logs.Consequently, harvester data are a key source of information in the fields of bucking optimization, harvester productivity assessment, forest inventory and operation management (Kemmerer and Labelle 2020).
Tree bucking, i.e., the process of segmenting a tree into logs, is a primary task in cut-to-length harvesting and has great impact on the value of the produced wood.Therefore, bucking optimization has long been a central task in forest engineering (Olsen et al. 1991;Bowers 1998).Besides providing harvester operators with bucking suggestings, bucking algorithms have found widespread application in economic simulations (Gobakken 2000;Malinen et al., 2007;Nybakk et al. 2007;Puumalainen 1998).Although butt rot has been identified as a key challenge in bucking optimization (Rapoport 1998), bucking simulations have yet to be used to assess how butt rot affects bucking patterns.Such analyses can provide detailed insights into the economic impact of butt rot on harvested trees, as well as approximating the magnitude of the economic losses caused by butt rot for forest owners.
The main objective of this study was to characterize the direct economic losses experienced by Norwegian forest owners due to butt rot in Norway spruce trees.We used production data obtained from cut-to-length harvesters, with corresponding stem profiles and wood grade information as input in bucking simulations.Specifically, we investigated: 1. What proportion of harvested Norway spruce roundwood is affected by butt rot? 2. How are the occurrence of butt rot and resulting value losses distributed across tree dimensions? 3.By how much are saw log volumes, lengths and values reduced when affected by butt rot? 4. How large is the direct economic impact of butt rot for Norwegian forest owners?

Harvester data
We used harvester data collected by seven contractors operating in different parts of Norway (Figure 1).The data comprised 399,207 Norway spruce trees (142,149 m 3 ) harvested in 149 clear-felled stands, in the period January 2020-August 2021.A subset of the harvester dataset was used in a previous study (Räty et al. 2021).The harvested stands were dominated by Norway spruce, however small proportions of other tree species were present, mainly Scots pine (Pinus sylvestris L.) and birch (Betula pubescens Ehrh.).We only included Norway spruce trees in the current study, as butt rot mainly affects Norway spruce and data on butt rot were collected specifically for this species.In total, 13,441 of the harvested Norway spruce trees did not contain any commercial log and we therefore discarded these trees from the analysis.For most of those trees, the harvester operator selected a wood grade indicating waste, and some of the trees did not meet the dimension requirements of the assortments used in the bucking.The resulting number of trees used in this study was 385,766 which corresponded to 141,776 m 3 .Harvester data were stored following a unified data format, i.e., the standard for forest machine data and communication (StanForD, Arlinger et al. 2012).All data were recorded as harvester production report (HPR) files (Arlinger et al. 2012), as well as additional stem files (STM) in cases where stem profiles were not included in the HPR files.From the HPR and STM files, we extracted data on harvested trees (tree ID, stand ID, diameter at breast height (DBH), wood grades, diameters at 10 cm intervals along the tree stem) and produced logs (log ID, tree ID, stand ID, start position, end position, assortment, log dimensions and volumes over and under bark).An overview of the obtained harvester data is shown in Table 1.

Butt rot registration
In addition to their regular work related to harvesting and processing trees, the harvester operators recorded the presence of butt rot by selecting assortments which we added to the cutting instructions.The operators classified the presence or absence of butt rot based on visual assessments of produced logs.Logs classified as pulpwood and waste were labeled according to whether the wood had <50% or >50% butt rot, respectively (% observed on the bottom cross section of the log).In our analysis, we assumed the operators' assessments of butt rot to be free of error.

Wood grade data and log assortments
Because wood grades, log assortments and assortment requirements for dimensions and wood grades varied across contractors and harvested stands, we applied a common classification of wood grades.For each log segment corresponding to a given log or waste segment, we determined the wood grade based on the recorded assortment and the log dimensions.The wood grades indicated the best possible assortment out of five assortments (Table 2) for a given log segment.We rounded starting and ending positions of logs to the nearest dm.
In determining the wood grades, we first classified all saw log segments as saw logs, assumed to have no defects.Second, we classified log segments without rot and with dimensions that did not meet the requirements of saw logs as pulp.Third, we classified log segments with <50% rot as rotten pulp.Fourth, we classified all energy wood segments as energy wood with >50% rot, because energy wood almost exclusively comprises rotten logs in Norway and typically, pulp wood would have been selected if rot were absent.Finally, we classified log segments which the operator determined to be waste as either rotten waste, according to the wood grade selected by the operator, or waste without rot which we assumed to have defects other than butt rot.
We retrieved assortment prices from national statistics for 2021 (Statistics Norway 2022).We rounded the mean prices down to the nearest 50 Norwegian krone (NOK) because wood prices in 2021 were substantially greater than previous years, and wood prices decreased in 2022.The assortment prices were then converted from NOK to Euro (€) using an average conversion rate for 2021 (0.096, www.x-rates.com/average).We assumed dimension requirements as shown in Table 2, corresponding to typical length and diameter requirements in Norway, although dimensions can vary slightly across sawmills.We further assumed a fixed price for all assortments, regardless of log dimensions.

Bucking algorithm
Bucking optimization has commonly been based on the theory of dynamic programming (Bellman 1954;Faaland and Briggs 1984), whereby the goal is usually to maximize the total tree value according to the decision makers' conditions.In dynamic programming, the optimization problem is solved computationally in multiple stages in a forward-reaching manner.The stages are starting points in cross cutting positions between log segments.At each stage, the optimal cut is selected given the conditions and optimal cross cuts in the processed part of the tree stem.
We used the R package optBuck (Noordermeer 2022) for the bucking simulations.OptBuck is a bucking algorithm that applies dynamic programming to find the bucking pattern that maximizes the value of a given tree.The algorithm starts at the butt end of the tree, the position of which is the first stage.All possible log lengths are tested, given the dimension restraints provided for the potential assortments.Each potential log is then designated the most valuable assortment based on its dimensions and wood grades.The segment of the tree with the poorest wood grade typically defines the assortment.If, for example, the wood grade for most of a given log allows sawn wood, and only a small portion of poorer wood grade exists at one  end of the log, the poorer wood grade will be the limiting factor.
Once the most valuable potential logs for the first stage are selected, the procedure starts at the next stage, which is the top end of the shortest log, and the procedure is repeated iteratively.For each stage, the most valuable log is selected, and its value is added to the accumulated value of the preceding optimum bucking pattern which ended at that stage.The algorithm proceeded along the tree in 10 cm sections, corresponding to the stem profile increments recorded in the harvester production files.For the first 100 cm of the tree stem, an assortment "waste" was enabled with a value of 0, for cases in which wood defects such as butt rot were present.In such cases, it is often beneficial to disregard the first segment of the tree as waste to ensure potentially longer and more valuable logs in subsequent tree stem segments.The algorithm terminated at the top end of the tree and selected the bucking pattern that maximized the total value.The output of the algorithm contained the selected bucking positions along the tree stem, the assortments of the produced logs and the corresponding volumes and values.
Diameters under bark were computed following the Stanford 2010 "Skogforsk 2004, Norway spruce" bark function (Arlinger et al. 2020).Log volumes under bark were computed differently depending on the log assortment.For all assortments except sawn wood, log volumes under bark were calculated using the Stanford 2010 'All diameters (solid volume)' method.For saw logs, log volumes under bark were calculated using the Stanford 2010 'top' method.

Bucking simulations
We assessed the effects of butt rot on roundwood values using two bucking simulations.In the first simulation, we included defects caused by butt rot, as obtained from the harvester data.In the second simulation, we assumed all trees to be free of butt rot, whereby we substituted wood grades of segments which were affected by butt rot with healthy wood.For both cases, we used the harvester data as input for the bucking algorithm and computed the optimal bucking pattern and the corresponding tree value.

Data analyses
We assessed the effects of butt rot on roundwood values at tree level, stand level and country level.At tree level, we assessed what proportion of trees was affected by butt rot and which volume this corresponded to, as well as the DBH distributions of affected and unaffected trees.Further, we assessed the consequent economic losses for individual trees, assessed how the losses related to DBH, and compared the losses across contractors to identify potential differences between regions in which the contractors operated.Finally, we assessed the value losses, saw lengths and number of saw logs from trees in five DBH classes: 0-100 mm, 100-200 mm, 200-300 mm, 300-400 mm and > 400 mm.
For the stand level analyses, we computed the stand-wise mean roundwood volumes affected by butt rot, the resulting economic losses, and values of DBH.We assessed how the obtained proportions of trees affected by butt rot varied across contractors, i.e., across different regions within Norway.Further, we assessed the relationship between values of mean DBH and mean value losses, and the relationship between the proportion of roundwood affected by butt rot and the mean value losses.
Lastly, we approximated the direct economic impact of butt rot at country level.To achieve this, we gathered country-level statistics on commercial roundwood removals of Norway spruce for the year 2021 (Statistics Norway 2022).We then multiplied the mean value loss (in €/m 3 ) obtained from the bucking simulations with the total harvested Norway spruce roundwood volume.

Tree level analysis
Of the 385,766 trees used in the analysis, 62,763 trees were affected by butt rot, corresponding to 16% of the harvested trees.Butt rot affected 15,005 m 3 of the roundwood volume, corresponding to 11% of the harvested volume.Relative shares of log assortments obtained from the first simulation, in which rot was included in the data, and the second simulation, in which we assumed trees to be free of rot, are shown in Table 3.For the trees affected by butt rot, the mean height along the stem affected by butt rot was 4.73 m.The length of The DBH distribution of trees without rot and trees with rot is shown in Figure 2.
The value losses obtained for trees affected by butt rot are plotted against corresponding values of DBH in Figure 3. Value losses of trees affected by butt rot increased with DBH, whereby losses started to increase when DBHs exceeded 150 mm.The largest obtained tree value loss was € 64.05 with a DBH of 515 mm.However, there was substantial variation between similar trees with regard to DBH, for example another tree with a DBH of 516 mm had a value loss of only € 0.67.
The distributions of value losses obtained for the seven contractors are shown in Figure 4 and ranged considerably from 0 to 75% and with a mean of 32% (Figure 4).For a total of 12 trees, the presence of butt rot did not decrease the tree value, i.e., the value of the tree obtained for the first and second simulations were identical.In all those cases, the bucking outcome of the first simulation comprised a single pulp log, and the bucking outcome in the second simulation comprised an energy wood log and a pulp log.Table 4 shows that mean value losses increased with DBH, whereby the largest value loss was obtained for the diameter class of >400 mm.Saw log lengths were substantially greater in the second simulation, in which all trees were assumed to be free of rot.Similarly, the mean number of saw logs was greater for all but the first diameter class, for which almost no saw logs were included in the bucking outcomes.Distributions of saw log volume losses are shown in Figure 5 and show that butt rot caused a mean reduction in saw log volume of 48%, corresponding to 0.15 m 3 .

Stand level analysis
Stand-wise proportions of roundwood volume affected by butt rot varied from 0 to 44%, with a mean of 13% and a standard deviation of 10.5% (Figure 6).As expected, the mean value losses increased with the proportion of roundwood volume affected by butt rot in the stand.The stand-wise value losses varied according to DBH from 0-15% and increased with the proportion of roundwood volume affected by butt rot.There was a positive correlation between the mean DBH in the stand and the mean value losses.

Country level analysis
Based on the results obtained from the bucking simulations, we approximated the economic impact of butt rot on Norwegian forest owners.In the first simulation, where butt rot was included, the mean wood value in harvested trees was 31.02€/m 3 .In the second simulation, for which we assumed trees to be free of rot, the mean wood value was 33.25 €/m 3 .According to the presented log assortment values, the average monetary loss caused by butt rot was 2.22 €/m 3 , corresponding to approximately 7%.Commercial roundwood removals of Norway spruce comprised 8.335 million m 3 in 2021 (Statistics Norway 2022).For that specific year, the mean loss obtained in this study would imply a total loss of approximately €18.5 million annually for Norway spruce caused by butt rot.This figure signifies the direct monetary losses for forest owners, not including indirect losses due to, e.g., increased mortality, increment reduction and increased wind damage.

Discussion
Evaluating the impacts of butt rot in Norway spruce trees is essential, given current trends in decreased tree vitality worldwide, driven by environmental stressors such as climate extremes (Hartmann et al. 2022) and pathogenic diseases (Ghelardini et al. 2022).We characterized the economic losses caused by butt rot in final fellings of Norway spruce in Norway, using production data obtained from harvesters as input in bucking simulations.In our analysis, we assessed the effects of butt rot on roundwood values at the tree level, stand level and at country level.Although butt rot has  been assumed to cause large economic losses, only few previous studies had assessed how butt rot affects roundwood values.
Our results confirmed that butt rot causes substantial economic losses to Norwegian forest owners.In our dataset, 16% of trees were affected by butt rot, in contrast to earlier findings presented by Granhus and Hylen (2016) who found butt rot in 10% of trees in productive Norway spruce forests in Norway.The mentioned study was based on core samples from temporary plots measured as part of the national forest inventory in the period 1986-2004.The sample plots largely covered the same geographic extent of our data, however, data collected in young forest were also included in their analysis, whereas we primarily used data from stands that were ready for harvest.This may explain the larger proportion of trees affected by butt rot in our study.
In another study based on core samples taken from national forest inventory data in the period 1964-1976, Huse (1983) ) found that 8% of Norway spruce trees had butt rot.However, when the presence of butt rot is assessed by means of core samples, its presence will not   be detected in a substantial part of sampled trees, say, 20-50% of cases, as shown by Stenlid (1992) and Kallio and Tamminen (1974).Errors of omission are particularly likely to occur when butt rot is present at stump height but not at breast height, or when the rot column is not in the center of the tree stem.
Butt rot is more likely to be detected on stumps, after a tree is felled, than in core samples.Indeed, in another nationwide study assessing the presence of butt rot on stumps of harvested Norway spruce trees in Norway, butt rot was detected on 27% of stumps (Huse et al. 1994).The mentioned study was based on questionnaires filled out by approximately 5000 forest owners, who assessed the presence of butt rot on subjectively placed sample plots in clear-felled stands of Norway spruce.The sample plots contained a minimum of 30 stumps, whereas we used data from entire stands.Although a comparison with that study is not straightforward due to different methods being used, the results point to a possible decrease in butt rot frequency since the 1990s.
Our results indicate that for trees affected by butt rot, the saw log volume can be expected to be reduced by 48%.This is smaller than a result obtained in a Finnish study, where Mattila and Nuutinen (2007) reported a mean loss of 60%.In the Finnish study, the minimum saw log dimensions were larger, i.e., a mean of 16 cm in diameter over bark, with 400 cm in length in comparison to the 12 and 370 cm used in this study, respectively.Greater minimum log dimensions lead to fewer potential bucking outcomes, particularly so for trees affected by butt rot, which may explain the difference in results.Nevertheless, both studies agree on butt rot leading to substantial losses in saw log volume.Gonthier et al. (2012) estimated the economic losses caused by butt rot based on wood values and corresponding hypothetical values if butt rot were absent.Their study site was the Aosta Valley in the Western Italian Alps, and they obtained economic losses of 18-34%, depending on the assumed damage scenario.In our study, the presence of butt rot did not cause major value losses in all trees, as for many trees, we obtained only minimal losses <10 €/m 3 and in a few cases even 0 €/m 3 .Yet, butt rot caused a mean loss of 7% when all harvested trees were considered, i.e., when trees without rot were included in the assessment.This was substantially smaller than the losses obtained by Gonthier et al. (2012).The difference in the magnitude of the losses may be explained by the fact that their sampled stands had relatively large proportions of wood affected by butt rot.In addition, their assumed wood prices ranged from 60-137 €/m 3 , i.e., much greater than the prices used in this study which ranged from 14.40-52.80€/m 3 .Nevertheless, the loss of 7% obtained in this study will have considerable impacts on a forest owner's earnings when aggregated to, for example, the logging operation level or property level.
Stand-wise proportions of wood volume affected by butt rot varied considerably from 0-44%, with a mean of 13% and a standard deviation of 10.5%.In a previous study, Hylen and Granhus (2018) modeled the occurrence of butt rot using the variables: DBH, stand age, elevation, growing season temperature sum and vegetation type, and found that all the studied variables affected the occurrence of butt rot significantly.Other factors, such as the stand management history and the number of previous Norway spruce rotations have also been found to influence butt rot presence (Woodward et al. 1998).Data on most of these variables were unavailable for this study, however we also found that the occurrence of butt rot, as well as the resultant economic losses, increased with DBH.Correspondingly, Gonthier et al. (2012) found that DBH and stump diameter were good predictors of the occurrence of butt rot.Our dataset covered a wide geographical range within the area of Norway in which most roundwood is harvested, and included a range of elevations and climatic conditions, which may explain the large variation in stand-wise proportions of roundwood volume affected by butt rot.
We approximated the economic losses of forest owners due to butt rot to be €18.5 million at country level.Though this figure is substantially greater than a previous approximation (€ 10 million; Solheim 2010), it is somewhat aligned with approximations in other Nordic countries.In Finland, both the approximated annual losses due to butt rot (€ 50 million; Piri et al. 2019) and total amount of harvested Norway spruce (26 million m 3; Luke 2022) were slightly smaller than threefold the corresponding figures in our analysis.In a Swedish study assessing the economic impact of butt rot, Stenlid (1992) approximated the annual economic losses at roughly €25 million.The study suggested that 15-20% of harvested roundwood volume can be expected to be affected by butt rot, corresponding to the 16% obtained in our study.Stenlid (1992) assumed 6% of logs to be rotten in Sweden, which with today's logging volume would imply 2.6 million m 3 of roundwood affected by butt rot (SLU 2022).A subsequent study reported that five million m 3 of roundwood is downgraded annually due to butt rot in Sweden, and about 675,000 m 3 is lost due to reduced growth (Normark and Fries 2018).
Although the harvester data covered a wide geographical range, the data were not distributed evenly across productive Norway spruce forests in Norway.For example, no data were collected in the north of Norway, nor in the counties Vestland, Rogaland and Agder.However, only a relatively small proportion of the roundwood harvested annually in Norway is harvested in these regions (approximately 15%; Svensson and Dalen 2021).Further, the harvester data used in this study were not collected according to a probability-based sampling design, and therefore, we were unable to provide a statistically rigorous estimate of the total economic loss at country level with a corresponding confidence interval.It is however challenging to obtain harvester datasets that are spatially distributed evenly, at least currently.Ongoing efforts to centralize the flow of harvester data in online databases (Berg et al. 2019) may however improve the prospect of massive and georeferenced data collection by harvesters.
Two simplifications in this study may have undervalued the economic losses due to butt rot.First, for simplicity, we assumed fixed wood prices for the assortments, which we kept constant across harvesting operations.However, assortment prices will have varied over time, as well as for different log diameter and length combinations.Saw logs with greater diameters generally have greater value and are also more likely to be affected by butt rot.Similarly, log lengths that are preferred by the industry are typically priced accordingly in price matrices, the effect of which was not considered in this study as no price matrices were available.When a tree is affected by butt rot, chances of a greater, more valuable, log length being selected will tend to decrease, further reducing the obtained value.Second, we assumed the operators' assessments of butt rot presence to be free of error.The operators will likely have overlooked some of the rot cases, the losses of which will not have been included in our analysis, but which likely will have been detected further on in the production chain, causing additional economic losses to the forest owners.Due to these reasons, the losses presented here should be considered a lower bound of the economic losses caused by butt rot.
We only considered the direct effects of butt rot on the bucking outcome of trees.Indirect effects, such as growth reduction and increased risk of wind damage, will have resulted in additional economic losses.This highlights the importance of understanding the mechanisms governing infection initiation and progression, particularly because they can differ among fungi species.In the case of Heterobasidion spp., secondary infections occur mainly through root contacts (Swedjemark and Stenlid 2001), while Armillaria spp.infections are commonly triggered by fungal rhizomorphs (Devkota and Hammerschmidt 2020).Additional research is needed to quantify indirect effects of butt rot, as well as to explore potential strategies for controlling the spread of different fungi species.

Conclusions
Our results indicate that butt rot causes substantial economic losses to forest owners.Of the 385,766 trees harvested in different regions of Norway and assessed by seven harvester operators, 16% were affected by butt rot, whereby butt rot had an increased tendency to occur in larger trees.The occurrence of butt rot caused a mean reduction in saw log volume of 48%.Proportions of roundwood volume affected by butt rot varied considerably across harvested stands.On average, roundwood values were reduced by 2.22 €/m 3 , corresponding to a value reduction of 7%.Both saw log lengths and the number of saw logs were reduced due to butt rot, whereby the losses increased with DBH.Our results suggest that butt rot causes direct economic losses of around €18.5 million annually to forest owners in Norway.

Figure 1 .
Figure 1.Map showing locations of the assessed stands in Norway, colors indicating the seven contractors that provided the harvester data.

Figure 2 .
Figure 2. Density plot of diameters at breast height (DBH) of trees without butt rot and trees with butt rot.Rotten and non-rotten are here considered different populations.

Figure 3 .
Figure 3. Diameter at breast height (DBH) of harvested trees plotted against value losses.

Figure 4 .
Figure 4. Violin plot of value losses of trees affected by rot.Dots indicate mean values calculated for contractors and whiskers indicate corresponding standard deviations.

Figure 5 .
Figure 5. Density plots of saw log volume losses in percent of total tree volume (left panel) and in m 3 (right panel), obtained for trees affected by butt rot.Dashed lines indicate mean values.

Figure 6 .
Figure 6.Stand-wise mean value losses plotted against mean proportions of butt rot, mean diameter at breast height (DBH) and mean loss in saw log length obtained for the harvested stands.

Table 1 .
Summary statistics of diameters at breast height (DBH) and tree volumes under bark (including cut-offs/waste) recorded by the harvesters, and tree values according to the assumed assortment prices (Table2) for the harvested Norway spruce trees (n = 385,766).

Table 2 .
Norway spruce log assortments used in the bucking simulations, assumed prices (€/m 3 under bark), and permitted log dimensions (over bark).

Table 3 .
Proportions of log assortments obtained from the bucking simulations.

Table 4 .
Number of trees, mean value loss per tree, mean saw log length and no. of saw logs per tree obtained in the simulations, calculated for five diameter at breast height (DBH) classes.