Evaluation of a new pruning and tending system for young stands of Douglas fir

In Germany, management restrictions for Norway spruce (Picea abies (L.) H. Karst.) due to climate change lead to increasing interest in Douglas fir (Pseudotsuga menziesii (Mirb.) Franco) as a potential substituting species. However, Douglas fir requires cost-intensive silvicultural treatments, such as periodic thinnings and, in particular, pruning. In order to improve the efficiency of such treatments, a new tending system with an adapted two-step work system was analyzed. The new system, using electric pruning shears and the backpack clearing saw Husqvarna 535FBX ‘Spacer’, was compared to the conventional three-step work system, using handsaw and chainsaw and characterized by tree selection previously conducted as an independent work step. Time and motion studies to determine productivity and costs, as well as ergonomic analysis through heart rate measurements and posture analysis were conducted. Overall, the new system was found to be more productive and to have lower costs, with 8.9 trees per scheduled system hour (4.17 € tree–1), compared to the conventional system with 8.1 trees per scheduled system hour (4.44 € tree–1). Ergonomic improvements with the new system could be mainly observed during the felling of competing trees, when the level of heart rate reserve was reduced by 9.3 percent points, compared to the conventional system. However, significant advantages in reducing unfavorable body postures expected for the ‘Spacer’ could not be confirmed. Since time savings within the new system were mainly attributed to the adaptation of workflow and the use of the electric shears during pruning, it should be considered to replace the ‘Spacer’ within the new system by light chainsaws for best results under the conditions investigated.

and the thinning by using a chainsaw (third step). The tree selection within the first step is commonly done by a forester with academic silvicultural education, leading to a timely deferred scheduling of the consecutive work steps. However, due to today's more comprehensive training of professional forest workers, they are technically capable of selecting future crop trees themselves. Therefore, the separation of the three work steps does not seem expedient anymore, and a re-organization of the work offers potential efficiency gains.
Within the present study, the conventional three-step work system is compared with a recently developed two-step work system. Apart of work re-organization, the new system also includes the use of new tending equipment, to replace the commonly used handsaw and chainsaw, which are known to entail cost-intensive and tedious work (Ligné et al. 2005). The study aims to compare the two work systems in respect to: (i) work productivity and associated costs, and (ii) physical workload. Integrated into real-world operations, the study does not aim to investigate the technical properties of the used equipment per se, but rather focuses on the holistic system approach of the whole work process through operational case studies.

Study sites
In-field operational case studies were conducted in July and October 2018, and in September and October 2019, in the German states of North-Rhine Westphalia (sites 'Hachen' and 'Mellen') and Lower Saxony (sites 'Göttingerode' and 'Hahnenklee'), respectively. Different young stands of Douglas fir due for tending were selected, and divided into blocks for comparisons between the surveyed work systems (Fig. 1). Priority was given to dense stands, with trees having dead branches in the lower stem compartment. The selected stands were established after the storm Kyrill in 2007, with an initial spacing of approximately 2.5-3.0 m (Table 1). Overall, average dbh of trees selected for pruning was 12.3 ± 2.3 cm, measured with a diameter tape. Tree height averaged at 11.1 ± 3.0 m, measured on a subsample (consisting of 63 selected trees) using a Vertex IV (Haglöf Sweden AB, Sweden).

Work systems and equipment used
Three professional forest workers (Table 2), being proficient users of the investigated equipment, were monitored when applying both work systems in alternation. For this purpose, the operators were assigned to select approximately 80 future crop trees per hectare at the study sites. The selected trees should receive a first pruning lift until a height of 2.3-2.5 m, as feasible to reach for Table 1. Characteristics of pruned Pseudotsuga menziesii trees in the study, on four different stands (site). For pruning, a conventional three-step system (hand-held pruning saw and chainsaw) and a new two-step system (electric pruning shears and 'Spacer') were applied. With number of observations (n, trees), mean values, standard deviation (SD), minimum and maximum of diameter at breast height (dbh), tree height and number of branches per tree. an operator standing close to the tree without the support of a ladder. Depending on the individual situation and operators' appraisal, up to three competing trees to the future crop trees were removed if necessary. Within the study, two systems for fulfilling this assignment were compared through a block design (Fig. 1). Each operator worked with both systems, changing at hourly levels, allocated throughout the various periods of the work day.
The two investigated systems are referred to as: (I) the new system, developed by the Forest Education Center FBZ in Arnsberg/Germany and (II) the conventional system, which consists of three separately conducted steps, each of them completed for the entire operation before commencing the subsequent step: (1) Future crop trees and adjacent competing trees are selected and marked by a forest worker.
(In practice, this work step is usually conducted by a forester with academic education. Within the present study, the forest worker selected the future crop trees.) (2) Marked future crop trees are searched in the stand, pruned by a forest worker using a conventional hand-held pruning saw (Table 3). (3) The marked and pruned trees are searched and individual competing trees are removed by a forest worker using a chainsaw (Table 4).  In contrast, within the new system, steps (1) and (2) of the conventional system are jointly conducted, leading to a system comprising two steps: (1) At the same time, the forest worker selects, prunes and permanently marks future crop trees, as well as marking adjacent competing trees, in a continuous work process -tree by tree. The pruning is conducted using electric pruning shears (Figs. 2 A,a and Table 3).
(2) Marked competing trees, adjacent to the selected future crop trees are searched and removed by the forest worker, using a 'Spacer' (Figs. 2 B,b and Table 4).
The new system is characterized by a re-organized workflow, and additionally by the employment of novel equipment (Fig. 2). The electric pruning shears STIHL ASA 85 (Andreas Stihl AG & Co. KG, Waiblingen/Germany) and INFACO F3015 (INFACO s.a.s. Cahuzac-sur-Vère/ France) were used at the study sites in North-Rhine Westphalia and Lower Saxony, respectively. Both models are renowned for use in forestry, and comparable in both their technical specifications (Table 3) and their handling. Therefore, an effect on the outcomes of this study from the usage of one or the other of these two tools is not assumed. However, it should be noted that the INFACO F3015 was equipped with an optional safety glove for the hand not operating the shears that interrupts the cutting action if the shears touch a conductive fabric. This extra safety feature is an asset when pruning while standing on a ladder, with vision obstructed by the tree trunk, but will not be of influence in the set-up of this study, where the worker can walk around the trunk to reach all branches to be cut.
During removal activities, conventional brushcutters allow a fast work progress within early stand tending, at much better ergonomics compared to chainsaws. But the circular saw cutting unit of conventional brushcutters cannot cope with stump diameters exceeding 80% of the blade radius (approximately 7 cm) (Wagner and Nolte, 2014). As an alternative for both conventional brushcutters and chainsaws, the new work system is complemented by the Husqvarna 535FBX 'Spacer' (Husqvarna AB, Stockholm/Sweden, Table 4). This power tool is combining a backpack brushcutter with a pole-saw-type chainsaw, complemented with a push-and-pull blade, supporting clearing work in young and dense stands. In order to reduce the static strain on the operator's arms caused by the weight of the cutting unit, a lightweight aluminum arch is added to the harness (Figs. 2 B,b). The cutting unit is suspended from this arch by an adjustable flexible synthetic rope, allowing the release of both hands if needed for other intermittent work tasks.

Time and motion study
An elemental time study, following the REFA (1998) method with the continuous time measurement, was conducted. Thereby, the treatment of each future crop tree was defined as one work cycle, which was further split into two subcycles; pruning and removal ( Table 5). Within the subcycles, work elements were defined and classified according to the REFA (1998) timing system, with the corresponding delay factors ( Table 5). The recording of five discrete work elements was conducted in a net-time approach, excluding delay times and additional supportive work time. Preparatory time and the time for searching the first tree at each block were excluded as well. This approach allowed a work cycle modelling based on main work time (MWt) and complementary work time (CWt) only, and led to PSH 0 (productive system hours, excluding all delays). For estimations of overall costs and gross-productivity of the systems, scheduled system hours (SSH) are required. Therefore, element-specific delay factors were applied onto PSH 0 times, according to Hittenbeck (2015).

Ergonomic analysis
In order to rate the physiological workload of the monitored forest workers (Table 2) when applying the two work systems, they carried a holter ECG (electrocardiogram) for heart rate (HR) monitoring throughout the entire work shifts, which were synchronized with the elemental time studies. This holter ECG is recording cardiac activities with a lightweight chest belt, transmitting wirelessly to the ECG unit in form of a wrist watch with storing device (custo med GmbH, Ottobrunn/ Germany). In order to compare the HR measurements at work (HR w ) between work systems, HR was transferred to an individual percentage of heart rate reserve (%HRR) according to Karvonen et al. (1957)  where: HR w = heart rate at work, HR r = heart rate at rest, HR max = maximum heart rate. For the determination of the individual's heart rate at rest (HR r , Table 2), the person's HR was consecutively measured, at least for 10 minutes without major fluctuations, in a seated and relaxed position at the beginning of the work day. The maximum HR (HR max , Table 2) was determined by the common approach, given as 220 minus age of the monitored operator (Robergs and Landwehr 2002). Unfortunately, due to the limited budget within the study, a full performance diagnostic, including a ramp test to determine more precise HR max figures, could not be realized.
Principally, HR measurements, also with limited precision levels, can generate valuable information regarding cardiovascular workload. However, a comprehensive workload assessment should include postural workload, too. Therefore, the time studies were combined with an analysis of postural workload according to the Ovako Working Posture Assessment System (OWAS) (Karhu et al. 1977). The OWAS is an observational technique, where body postures causing discomfort and fatigue are rated according to a code system and their frequency of occurrence during the execution of specific work tasks. The code system covers three body sections -upper limbs, lower limbs and back -creating, together with the weight to handle, a four-digit code. The resulting codes are used for rating the postural strain in four action classes, based on the increasingly urgent need for remedial intervention, as follows: class 1 = normal postures, which do not need any special attention; class 2 = slightly harmful postures that must be considered during the next regular check of work methods; class 3 = distinctly harmful postures that need consideration in the near future; class 4 = extremely harmful postures that need immediate consideration. For that purpose, video footage of the operations was analyzed by an additional time study, using a frequency method with 30-second intervals. This interval length was reasonable for this analysis, since 80% of all captured work elements exceeded this margin. In order to compare and rate the overall difference in work-posture strain between the two work systems, the OWAS-based Lundqvist-index (Lundqvist 1988) was calculated. This index is calculated from the occurrence-weighted frequency of the individual codes during the time of observations.

Statistical analysis
Analysis were conducted using the programming language R (R Core Team 2020) together with RStudio (RStudio Team 2020). Linear mixed-effect models (package 'lme4', Bates et al. 2015) were used to compare net time demands between the fixed effect of system, considering the study site as random effect. Contrast-sums were set to zero for included fixed effects, motivated by unbalanced groups and significant interactions. Subsequently, type-III analyses of variance (Fox and Weisberg 2019) and post-hoc analyses (Tukey HSD tests, Lenth et al. 2019 andHothorn et al. 2008) were applied to compare mean values per element, work system, subcycles and sites. Residuals of all linear models were checked for normal distribution via QQ-plots and histograms. The occurrences of action class 4 according to OWAS was compared between the systems by use of a Poisson test. All statistical tests were carried out on net-time values and %HRR averaged for each element and cycle. The significance level used for all tests was α = 0.05. Reference values are given as mean ± standard deviation (SD).

Total time demand per future crop tree
Total required net work time (MWt + CWt, Table 5) per selected and pruned future crop tree, consisting of pruning and removal of competing trees if necessary, was found to be 6.09 ± 2.37 min tree -1 for the conventional system ( Table 6). The time study provided clear evidence that total work time could be reduced through the application of the new system by 0.62 min tree -1 (p = 0.002, Table 7), resulting in an average total time demand per treated future crop tree of 5.48 ± 2.68 min tree -1 ( Table 6). Substantial differences were observed between study sites, during both subcycles (Fig. 3). Table 7. Summary of coefficients of the linear mixed effects model fitted to values of: (Model 1) total time demand per tree, (Model 2) total time demand for pruning activities of future crop trees and (Model 3) total time demand for removal of competing trees to adjacent, pruned future crop trees in young stands of Pseudotsuga menziesii for two systems: the conventional and the new one, using hand-held pruning saws and a chainsaw or electric pruning shears and the 'Spacer', respectively.

Total time demand for the subcycle pruning
Total time demand per treated tree, defined as work cycle, was divided into two subcycles (Table 5): One represents pruning (consisting of the work elements search, cutting and marking), where either the electric pruning shears or handsaw was used. The second subcycle represents removal of competing trees (consisting search and saw work), where either a chainsaw or the 'Spacer' was used. As given in Table 6, total time demand for pruning was considerably higher when the operator followed the conventional system. Through the application of the new system, time demand for pruning could be reduced by almost one minute per tree (p < 0.001). The marked time savings were realized through a higher efficiency when using the electric pruning shears (see Supplementary file S1 for details, available at https://doi.org/10.14214/sf.10447). In particular, when the electric pruning shears were used, cutting was on average 0.35 min tree -1 (Suppl. file S2) faster (p < 0.001, Suppl. file S1), compared to cutting with a hand saw. As can be seen in Fig. 4, this was statistically significant for all sites, except for 'Mellen', where required time for cutting was equal for both systems.
On average, 82 branches were removed from each pruned tree. Time demand for cutting clearly responded to the number of branches (p < 0.001). It was estimated, that one additional branch would result in an increase of 1.77 seconds, required for the element cutting. Still, this increase was significantly reduced when working with the electric pruning shears (p = 0.007). Besides changes within the element cutting, time savings within this subcycle also occurred during the search of future crop trees ( Fig. 4 and Suppl. file S1), which was usually more time demanding when done separately (three-step system), following the conventional system.

Total time demand for the subcycle removal
While time savings occurred during pruning when applying the new system, drawbacks on efficiency appeared during the subcycle removal (Fig. 4). This subcycle covers the search for future crop trees and saw work on usually one competing tree, except for the site 'Mellen', where more than two competing trees were removed on average (data not shown). Still, no correlation between the quantity of removed trees and total required work time could be observed. Total required work time for the subcycle removal was significantly higher during the application of the new system (p < 0.001), resulting in 1.84 ± 1.67 min tree -1 , compared to 1.38 ± 0.97 min tree -1 ( Table 6) when using the chainsaw (conventional system). But, the significant differences within the model applied are mainly driven by values gathered at 'Hahnenklee', since pairwise comparisons of mean time values of all remaining study sites did not reveal any differences between both systems (Fig. 4).
Accordingly, efficiency is hampered by higher time demand for the element search, when the forest workers were equipped with the 'Spacer' (p = 0.049, Suppl. file S1). In addition to search, also the element saw work on competing trees had a considerably higher time demand (p < 0.001, Suppl. file S1) by 0.28 min -1 (Suppl. file S2) and averaged 1.10 ± 1.30 min per released crop tree when following the new system, compared to 0.82 ± 0.71 min tree -1 required when the conventional system was applied ( Fig. 4 and Suppl. file S2).

Heart rate measurements
The new system significantly reduced the operators' cardiovascular workload, as indicated through the %HRR, which was derived from heart rate measurements. During the subcycle pruning, the overall difference was 4.40 %HRR (p < 0.001), even though no clear differences could be observed during the elements search and marking (Fig. 5 and Suppl. file S3). Subsequently, the reduction of %HRR during pruning was predominantly driven through the decreased cardiovascular workload during cutting of branches (p < 0.001, Suppl. file S3), whereas extents of these differed between sites (Fig. 5). Overall, during cutting of branches for pruning the future crop trees with electric pruning shears, %HRR was reduced by as much as 7.54 percent points.
Maximum values of %HRR reached during the study, 86.5 %HRR, were detected for saw work with a chainsaw during the conventional system. Likewise, mean values for this element were 9.34 %HRR higher compared to the new system (p < 0.001, estimated, Suppl. file S3). When competing trees were removed, lower values of %HRR appeared during both elements included (search and saw work, p < 0.001, Suppl. file S3), as shown in Fig. 5. Considering the differences in both elements, physiological workload during removal activities averaged at 50 ± 17 %HRR and 44 ± 18 %HRR for the conventional and new system, respectively.

Postural analysis
During both work systems, and among the subcycles, the frequency occurrence of OWAS action class 1 (normal) was predominant, with 76% for pruning, and 83% for removal during the application of the new system. Application of the conventional system lead to 79% observations allocated to class 1 during pruning and 81% during removal. The remaining observations could mainly be allocated to class 2 (slightly harmful postures), with both systems causing higher, and similar, postural workload during pruning (21% of observations for the conventional and 22% for the new system), compared to removal, where the posture workload with the 'Spacer' was lower (16% of observations), compared to the use of a chainsaw (19% of observations). Critical postures of class 3 (distinctly harmful) and 4 (extremely harmful) occurred only to a minor extent for both systems. Yet, a slightly negative tendency with the new system was observed in a direct comparison. Both pruning and removal accounted for 1.4% of class 3 observations with the new system each, in contrast to 0.63% (pruning) and 0.0% (removal) for the conventional system. Postures of class 4 were almost not observable, and limited to the pruning process for both systems (0.34% for the new system; 0.06% for the conventional system), only.
With respect to reduced strain on the muscle-skeleton system, no obvious differences emerged between the two work systems, with the Lundqvist index being 122 and 124 for the conventional and the new system, respectively. When separating according to the subcycles, a slightly more favorable Lundqvist index is generated during removal, resulting in 119 for both systems, compared to 122 (conventional) and 126 (new) during pruning.

Discussion
The increasing need for tending of young Douglas fir stands requires consistently improved work systems for efficient operations. Within this study, the conventional practice of pruning by handsaw and removal of competing trees by chainsaw was compared with a new system in field trials. The new system did not only replace the commonly used equipment (i.e. pruning saw and chainsaw), by electric pruning shears and the 'Spacer' clearing saw, but also adapted the workflow and organization. Primary aim with this new work system was to increase productivity at reduced costs, and to reduce work-related fatigue and occupational health risks by improved ergonomics.
Overall, the new system resulted in time savings compared to the conventional system during the observed tending operations, with a net time demand of 5.48 ± 2.68 min tree -1 versus 6.09 ± 2.37 min tree -1 (or 10.9 versus 9.9 trees PSH 0 -1 ) ( Table 6), reducing the work time demand by 10%. By adding the work-element-specific delay factors (Table 5), gross productivity can be determined at 8.9 trees SSH -1 (6.7 min tree -1 ) for the new system, and 8.1 trees SSH -1 (7.4 min tree -1 ) for the conventional system. Although with the new system costs per SSH were increased by 1.20 €, the higher productivity of this system reduced the costs per tree from 4.44 € tree -1 (for the conventional system) to 4.17 € tree -1 , a cost reduction of 6%. Most noticeable improvements of productivity were caused by the less complex two-step work system and the use of the electric pruning shears (Suppl. file S1). Whereas the pruning shears appeared to work within its optimal branch diameter capacity range, being superior to the handsaw, the stand conditions apparently did not match the requirements of the 'Spacer', which demanded more work time compared to the use of a chainsaw (Suppl. file S3). This was even more pronounced when tree dimensions required generally higher work time input, as it was the case at the 'Hahnenklee' site (Fig. 4).
For the operators, it was difficult to maneuver through the stands with the bulky arch and long pole of the 'Spacer' (Fig. 2), since all stands were very dense. In contrast, the chainsaw operator had a clear advantage with its compact equipment, allowing faster movements through the stand. This resulted in a lower time demand during the element search within the subcycle removal. Further, the larger tree dimensions at the 'Hahnenklee' site (Table 1) did not allow a simple oblique cut, but rather required a multiple side-cut, or even the formation of a felling notch for directional felling, causing significantly higher saw work time (Fig. 4). Generally, the element saw work was more time consuming with the 'Spacer' compared to the use of a chainsaw, probably since the 'Spacer' is difficult to guide for precise multiple cuts in various angles. In contrast, chainsaws are directly guided at the stump and with handles specifically designed for multiple cutting angles.
The re-organization of the work within the new system led to a decreased time demand for search (Fig. 4), since the work steps required for pruning and cutting were merged. Hence, the operator selected a suitable tree in an appropriate spacing to the previously pruned one and could directly commence the cutting of branches. In contrast, the two separated steps in the conventional system resulted in a higher time demand (Table 6), since it is necessary to cruise twice through the stand and to retrieve the previously marked tree before commencing the second step cutting. Further, it has to be considered that if tree marking is conducted during a different time period, way in advance to cutting, as it is for example common when a professional forester does the tree selection, additional costs will arise from extra travel efforts and preparatory times, not contained in the presented figures. However, although not explicitly captured within this study either, it should be noted that extra preparatory time is required with the equipment used in the new system. This is mainly related to the slightly higher effort of the operator to assemble the electric pruning shears and the 'Spacer', compared to the handsaw and the chainsaw.
Besides time and cost savings, the new system considerably reduced the cardiovascular workload during both subcycles (Suppl. file S4). According to Garber et al. (2011), exercise intensity can be classified as: very light (%HRR < 30), light (30 < %HRR < 40), moderate (40 < %HRR < 60), vigorous (60 < %HRR < 90) and near to maximum (%HRR ≥ 90). The reduced %HRR during cutting with the electric pruning shears is classified with a mean %HRR of 31.3 as 'light work', whereas the use of the handsaws can be classified as moderate work already (mean %HRR of 41.0, Suppl. file S4). The positive effect on task demand by using electric pruning shears, instead of handsaws, was already assumed by Nutto et al. (2013), while investigating pruning in Eucalyptus plantations, but could not be confirmed.
The lower cardiovascular workload of the new system during the element search, within the subcycle removal ( Fig. 5 and Suppl. file S3), may be attributed to the obstructing arch of the 'Spacer', forcing the operator to move slower through the dense stands. Although rated as negative for the work time demand, it has to be seen as positive from an ergonomic point of view. During this element, lower mean values of %HRR were measured for the new system (43.3 %HRR) compared to the conventional system (47.3 %HRR), yet both classified as moderate work (Fig. 5 and Suppl. file S4). It can be assumed, that slower but also constant pace allowed the operator to recover during search from the more demanding work element cutting, whereas the faster movement with the chainsaw, holding it in a one-sided static posture, is physically more demanding. Since with the 'Spacer' a different group of muscles is used with light loads, it can be considered as a disguised pause (Gallis, 2013). The relevance of such small recovery periods that avoid fatigue and allow for maintaining a constant performance level throughout the day was found already when studying e.g. cable skidding work (Berendt et al. 2020).
During the actual saw work, the 'Spacer' showed ergonomic advantages compared to the chainsaw through a lower %HRR, too. The mean %HRR measurements showed considerable differences between the use of the 'Spacer' (45.4 %HRR) or a chainsaw (53.2 %HRR, Suppl. file S4). Both values are marking saw work as the most labor-intensive work element of the investigated systems. However, the use of both the 'Spacer' and a chainsaw remain in average below a 59 %HRR, indicating still a moderate work level (Garber et al. 2011). Surprisingly, the OWAS assessment exposed a nearly similar Lundqvist index for the use of both tools, not confirming any advantage with the use of the 'Spacer' (purposely designed to ergonomic construction principles) with respect to avoiding unfavorable forced body postures. In fact, for the felling cuts close to the tree base the operator had to bend forward when using either the 'Spacer' or the chainsaw, leading to similar body postures. An advantage of the 'Spacer' can be pinpointed only to the fact that the machine weight (although higher in total) is causing less concentrated static forces due to its weight distribution over the whole body and suspension through the harness and arch. In contrast, the chainsaw's weight needs to be fully suspended by the upper limbs.
However, it should be considered that the chainsaws used have been generally high in engine output (>2.3 kW) and weight (>4.8 kg), with respect to the conducted operation. A lighter chainsaw (~1.8 kW, ~4.0 kg), would have probably been sufficient too, bearing a further positive effect on the experienced workload in favor of the conventional system.
A comprehensive rating of the chainsaw vs. the 'Spacer' cannot be conducted within this study due to the limited scope of the trials, where in particular operator and site effects were not the main focus of our investigations. The practice-oriented character of our study clearly limits the interpretation of the results, although they give legitimated indications. For instance, test person P1 working at the two sites 'Hachen' and 'Mellen' was directly exposed to differing site conditions. The 'Hachen' site, where warm temperatures prevailed during data collection, is further characterized by steeper terrain in contrast to the flat 'Mellen' site which also had smaller tree dimensions. These easier work conditions in return reduced the overall work time during removal, although at small magnitude, with an advantage to the use of the chainsaw on both sites (Fig. 4). In contrast, the labor effort in terms of %HRR differed noticeably between the two sites during removal, obviously related to the site conditions, but still at both sites in favor of the 'Spacer' (Fig. 5). In general, the inclusion of site as random-effect within the linear model applied allows a comparison between both systems, and a more general evaluation of the suitability for the investigated systems in the defined operational range. Nevertheless, in further investigations more focus should be given to varying operational conditions in terms of stand structure (i.e. diameter and height range, especially of removed competing trees, thickness of branches, terrain and understory vegetation), but also worker age and qualification. This would enable to clearly define best application areas for the two individual systems and allow to set realistic operator rates for best practices.
With respect to the current investigation, improvements through the application of the new system are primarily seen in the modified work organization. In addition to time and cost savings during the investigated execution of work, considerable savings due to obsolete preceding site visits for tree selection by a worker or professional forester can be expected, although this was not quantified in the experiment. Although still practiced in many regions, the overall need for a professional forester for tree selection is on general debate. Forest workers are better qualified these days and experienced workers and contractors proved to be qualified to conduct targetorientated tree selection in complex silvicultural systems to a satisfying level (e.g. Spinelli et al. 2016). Therefore, the trend of leaving tree selection to the appraisal of the designated worker will continue. However, scheduling this task ahead and completely separating it from the actual tending operation as in the conventional system clearly proves to be a restriction for efficient operations.
Besides work organization, the introduction of new equipment did not attain any specific cost savings, nor ergonomic improvements of the workflow. While the use of the electric pruning shears can be rated as more advantageous compared to the conventional handsaw, the 'Spacer' did not show the expected major advantages compared to the use of a chainsaw for the removal of competing trees, at least for the investigated stand conditions. In fact, with respect to productivity, the chainsaw can be rated as superior, and the ergonomic improvements by the use of the 'Spacer' neither considerably changed the labor effort required for motor-manual tending operations, nor resulted in the expected relief of the muscle-skeleton system. A dense stand structure, tree heights far above the operator and low branched trees seem to limit efficient work progress with the 'Spacer' due to the bulky construction, less convenient cutting guidance and limited engine power. In this respect, the development stage of the selected stands was out of its optimal working range, limiting the achievement potential of the 'Spacer'. Subsequently, the application of the 'Spacer' seems to be more appropriate in much earlier pre-commercial thinning stages, with a focus on schematic stem number reduction, only. If such stands comprise multiple species and a selective approach needs to be followed, the 'Spacer' proved already to be indeed an efficient tool to be considered as an alternative to the conventional brushcutter (Wagner and Nolte 2014;Suchomel and Jaeger 2017).

Conclusion
With respect to the limited operational range of the 'Spacer', a modification of the new system can provide potential advantages. While the subcycle pruning of the new system clearly outperformed the conventional one, drawbacks towards efficiency arose during removal. Therefore, as indicated by the results, the chainsaw seems to be the equipment of choice for removal of competing trees within tending operations. However, the use of smaller chainsaws should be considered. Due to the fact that the costs of a hand tool are nearly neglectable, no major cost savings from introducing the new system should be expected. However, further contributions to reducing work fatigue due to lower weight and lower coupling forces can be assumed. In this respect, the use of the currently upcoming professional-grade battery saws can provide further ergonomic benefits such as reduced noise and vibrations and the elimination of exhaust fumes. In combination with the electric pruning shears, an efficient battery-based work system with numerous benefits for the forest worker could be established.

Disclosure statement
No potential conflict of interest was reported by the authors.