*3.4. Postural Risk*

Figure 3 shows a breakdown per action categories and postural risk indexes estimated at subject, observation day, location and sample level. At sample level, approximately 35% of the analyzed frames were included in the 1st action category, more than half of them were categorized in the 2nd action category and roughly 6% were interpreted as belonging to the 4th action category.

**Figure 3.** Share of the analyzed frames per action categories and postural risk indexes estimated at subject, location, observation day and sample level.

The postural risk index characterizing the job was found to be of almost 178, indicating rather the categorization of this job in the second action category which requires corrective actions to be taken in the near future. At subject level, on the other hand, distribution on action categories and the computed postural risk indexes were quite variable. Frames attributed to the 1st action category accounted for shares of 18.8 to 61%, with the latter one characterizing the postural behavior of S14; frames attributed to the 2nd action category accounted for shares in the range of 27.9–78.8%, those specific to the 3rd action category accounted for minor shares and those belonging to the 4th category accounted for shares of up to 20.1%. The postural risk indexes varied in between 151.2 (S2 × L2 × 18) and 211.9 (S12 × L1 × 13).

At the sample level (Table 7), back postures were found to be particularly uncomfortable, as in more than 55% of the cases, the subjects were found to have the back bent and twisted or bent forward and sideways. Straight postures of the back were found only in 26% of the cases. In general, the arm postures were found to be comfortable and this situation is related to the characteristics and tool use in this kind of job. Combined with poor postures of the back, legs postures codded by 4, 5 and 6 lead always to the worst postural situation which is characteristic to the fourth action category. It was not the case of the analyzed sample since these legs' postures accounted for only 7%. Therefore, from the postural analysis point of view, the main problems related to potential risks of health impairment were those specific to the back.


**Table 7.** Share of back, arms and legs postures per codes described by OWAS.

<sup>1</sup> Not applicable according to OWAS method.

#### **4. Discussion**

The main aim of this work was to characterize the difficulty and risks associated to manual cultivation operations in hybrid poplar forests. Acknowledging the limitation of the results to male subjects, as well as the fact that indirect observation may still have affected the work behavior of the observed subjects, the first thing which needs to be addressed, even in the conditions of a good utilization of available time for effective work (approximately 70%), is that relating to a particular low efficiency of such operations which was in the range of 8.6–9.0 h per double-pass operated hectare. From this point of view, the first hypothesis of this study was confirmed only from the productivity point of view. Indeed, there is limited information of operational performance metrics for this kind of jobs. Nevertheless, for something similar, de Oliveira et al. [11] found an efficiency rate of approximately 3.3 h per hectare which took 52% of the heart rate reserve during the effective work. The Romanian forestry-related rating system [31], on the other hand, indicates for the same job operational efficiencies in the range of 1.42–4.90 man-hours per 100 m2, which will probably ensure rest breaks-taking in a sustainable way. One way to improve the efficiency and to balance the effort given by workers would be that of deploying inter-row mechanized cultivation operations on two perpendicular directions since the plantation layouts would enable such an approach. In particular, this could contribute to a reduction of manually operated area to approximately one fourth compared to the current operational layouts.

In terms of physiological workload, worth mentioning that heart rate is a good estimator of the VO2 indicator [32] that is commonly used to predict the work difficulty in general ergonomic studies [21,22]. Cardiovascular workload, as found in this study indicates that this type of operation tends to overload the workers, therefore confirming the second hypothesis of the study. How the subjects experienced the workload was found to be related to their age. In average, the %HRR metric for the effective work was found to be very close to the threshold of 40% which, according to some authors [10], defines the limit between acceptable and unacceptable workloads. However, this outcome should be interpreted as indicative at least from two points of view. The first one refers to the impossibility to extend the findings to cohorts characterized by anthropometrics that are particularly contrasting to those which built the data from this study. The second one refers to the caution which should be used in the interpretation of data since the %HRR metric was based on the commonly accepted formula for estimating the maximum heart rate, which has its own limitations [33]. Also, an ambulatory trial found heart rates at rest lower when self-measured at home compared to those measured under expert observation [34]. Obviously, such an effect will probably lead to an underestimation of job's difficulty, the same way the exposure to high thermal stress will. As such, and knowing the fact that the climate change affects the health of the workers [14], the thermal stress should be considered in the improvement of work in manual cultivation operations since the climate in the studied area is generally described by hot summers.

Most probably, an increased cardiovascular activity, as found in this study, is related to the type of work, work intensity and the body parts engaged in such work since the job tasks took a great deal of using handwork which is known to affect the heart rate response and characterizes the severity of muscular work [21]. Recovery time of heart rate is dependent on the exercise intensity and may reach more than 30 min [35,36], even if most of the recovery changes may occur in the first 1–2 min [36], while the heart rate response may be sensitive to postural changes [37]. For instance, switching from lying to sitting positions was found to increase the heart rate in some subjects by approximately 10 beats per minute [38]. Therefore, it was not surprising to find that for most of the subjects observed in this study the heart rate reserve was particularly increased also during the rest pauses and during the meal taking. It is difficult to place the manual cultivation operations, in terms of difficulty and risks, amongst other forestry jobs, given the fact that heart response is dependent on many factors such as the age of subjects, gender and their operational environment. Nevertheless, in motor-manual felling, estimates from the same flat-land area and for a worker having an age close to the average of this study [17] were close to those found in this study. In steep terrain forests of Turkey, for instance, harvesting and forest nursery work was found to be difficult to moderate jobs with heart rate reserves of approximately 41 and 32%, respectively [39], while jobs such as cable work in steep terrain [40,41] and cable rigging [42] may take more effort.

Work intensity was found to be light and moderate in most of the surveyed time (more than 95%), therefore partially confirming the third hypothesis of this study. Since ENMO values of up to 0.25 g are characterizing sedentary behaviors and light work such as standing still, dusting, sweeping the floor and self-paced walking [24], this intensity threshold was used to separate light intensity work in this study. In general, vigorous activity is considered to account for more than 21 mL <sup>×</sup> kg <sup>−</sup><sup>1</sup> <sup>×</sup> min <sup>−</sup><sup>1</sup> VO2 which roughly corresponds to accelerations corrected by the mean amplitude deviation of 0.45–0.5 g [25]. However, in this study, the intensity of work was considered to be moderate when ENMO had values from 0.25 to 1.0 g, by taking into account also the cardiovascular activity and the behavior or acceleration data in effective working events versus rest pauses. It should be mentioned that even in events such as the meal pauses, the subjects were not found to sit still all the time.

Also, given the position in which the accelerometers were placed, the collected and analyzed data stands, in particular, for the activity of subjects' back. This data may be correlated also with that coming from postural analysis where the back was found to be straight in 26% of the cases and bent, twisted or both in the rest of the cases. This outcome was partially consistent with the last hypothesis of this study. In this regard, the manual cultivation operation seems to be a job that does not require immediate postural redesign since the postural risk indexes were found to be less than 200% in most of the cases. However, the main problem here is that related to the back postures assumed by subjects which were particularly uncomfortable. Working predominantly with the back bent and twisted or bent forward and sideways (56.5% of the cases) may lead to health problems related to the lower back which is a known issue of forest operations jobs [14]. From this point of view, manual cultivation is a job that is even more hazardous compared to manual harvesting operations from Nordic countries [43] and close to that of motor-manual tree felling and processing operations from the area [17]. Compared to other kind of forestry-related partly mechanized jobs such as firewood processing [30] and wood debarking [44], manual cultivation seems to be riskier with the main problems coming from the back postures assumed during the work, since the arms and legs postures were found to be comfortable in most of the cases.

#### **5. Conclusions**

The main conclusion of this study is that the manual cultivation operations in poplar forests are rather difficult and hazardous, requiring reengineering tasks from ergonomic and public health points of view. To overcome the effect of small efficiency rates found in this study, mechanization should be extended by approaching the operated plots on two perpendicular directions, limiting this way the manual job to approximately one quarter compared to actual practices. Even if not documented by the available literature, such approaches are seldom used in the Romanian practice. Obviously, this way of extending the mechanized part of operations, will reduce also the continuous physical effort of the upper limbs and back by inter-placing movements from one tree to other, therefore it will lead to an increased use of bigger muscular groups and legs, that could help in attenuating the cardiovascular activity. By such measures, the intensity of manual work will be also decreased and the frequency of poor back postures will improve. Nevertheless, in such cases in which the approach of extending the mechanized operations is not feasible, a correct training of the workers, including a redesign of rest breaks could improve the status quo. These issues may be approached by intervention programs designed to tailor the work tasks, on the one hand, and to properly train the workers, on the other hand. Such approaches are important since the results of this study clearly indicated that the job in manual cultivation operations is characterized by a dynamic work that may overload the heart, upper limbs and the back of the workers. Also, incentives to attract youth in such operations should be developed and implemented at regional and national scale.

**Author Contributions:** Conceptualization, T.M.-P. and S.A.B.; Data curation, T.M.-P., M.C. and M.V.M.; Formal analysis, M.C. and M.V.M.; Funding acquisition, T.M.-P. and C.I.D.; Investigation, T.M.-P. and M.C.; Methodology, T.M.-P. and S.A.B.; Project administration, M.V.M. and S.A.B.; Resources, C.I.D., F.I. and S.A.B.; Software, S.A.B.; Supervision, S.A.B.; Validation, T.M.-P., M.C., M.V.M., C.I.D. and A.A.B.; Writing—original draft, F.I. and S.A.B.; Writing—review & editing, F.I. and S.A.B.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors acknowledge the technical and logistic support of Department of Forest Engineering, Forest Management Planning and Terrestrial Measurements, Faculty of Silviculture and Forest Engineering, Transilvania University of Bra¸sov. The authors would like to thank the local managers of National Forest Administration for agreeing to carry on this study in the forests managed by them and to the subjects who agreed to participate in the study, for their support. This work is part of a PhD thesis developed under the supervision of Doctoral School of the Transilvania University of Bra¸sov. The authors acknowledge the support of the aforementioned organization.

**Conflicts of Interest:** The authors declare no conflict of interest.
