Soil conservation during forestry operations is an increasing concern. This study proposes a theoretical model to forecast the depth of ruts associated with harvester trail systems and organisational boundary conditions, with the aim of reducing the harmful impacts of heavy logging machinery. The research combines empirical data on soil penetration resistance and soil water content of six soil types with theoretical rut depth models, emphasising the organisational dimensions of forest operations. Empirical data were gathered from sixteen soil sample plots in Slovenia. Based on this data and existing litterature, six models were developed to show the relationship between soil penetration resistance and soil water content. The maximum harvester trail length was then calculated for each soil type, based on harvest intensity and rut depth models. The findings indicate that modelled rut dept is less than 0.1 m if the soil can withstand a nominal ground pressure of 849 kPa in all analysed soil types. However, modelled rut depth of any soil type will consistently exceed 0.1 m when the soil cannot withstand a nominal ground pressure of 456 kPa. To avoid deeper ruts, it is advisable to plan and use harvester trail lengths of no more than 100 metres at 40% volumetric soil water content when the soil type is unknown. Incorporating soil type data into rut models could be an effective means of predicting rut depth at a given time, thereby boosting the efficiency of forestry operations.
In today's modern forestry, which is founded on sustainable management and development, it is essential to carry out the design phase of forest roads with quality, both for financial and environmental reasons. The basis for this is high-quality field data. This research tested available and potentially suitable methods for field surveys in the process of designing a forest road, including the classical survey method, GNSS and total station survey, as well as structure from motion and airbone lidar scanning survey. A total of 23 forest road designs were created through 3 design methods. Four important forest road parameters were tested: cut and fill volume, carriageway value and roadway width. No statistically significant difference was found for any of the tested parameters between the methods. Total station and ALS-based designs showed the lowest values of cut and fill volumes, 2.41 m3 and 2.54 m3 for the total station design, and 3.20 m3 and 2.47 m3 for the ALS design. Although some deviations were found between SfM designs based on different flight parameters, they were not significant. The results indicate the possibility of using all tested methods in forest areas with steep terrain, after a salvage logging has been performed. Also, by using three design methods and testing their results provides guidance how to test different measurement systems in the future when designing forest roads.
Extracting key features from remote sensing imagery for the purpose of identifying tree species holds significant importance. To assess the importance of the newly proposed polygon area index (PAI) and other features in classifying tree species, this research employed WorldView-3/2 (WV-3/2) as the experimental dataset and constructed four distinct feature sets: spectral bands, HSVs (hue, saturation, and value), textures and PAIs. Then, these feature sets and their combinations were utilized, and the random forest method was applied to classify tree species and ascertain the importance of the features involved in the classification process. The experimental results demonstrate that all texture features play a crucial role in the identification of tree species, with eight features being selected. Additionally, numerous HSV features (WV-3: 29, WV-2: 21) and PAI features (WV-3: 7, WV-2: 4) also make significant contributions. However, the spectral bands did not show a notable positive impact, as none were selected. When compared with the use of single-type features, the integration of multiple features from WorldView-3/2 resulted in more effective tree species identification, achieving an overall accuracy of 90.45% for WV-3 and 84.60% for WV-2, surpassing the highest overall accuracy achieved by single-type features, which was 76.64% for WV-3 and 78.27% for WV-2. The experimental results indicate that multi-type features should be used for tree species classification. Additionally, the newly proposed PAI feature demonstrates a positive contribution and is recommended for active use in tree species classification.
During professional work, the forestry workforce is exposed to hazardous activities and numerous risks, which may result in everyday occupational accidents or illnesses. Networking and innovative thinking with the aim of developing new legislative solutions and certification systems, e.g. chainsaw qualification standard, represents an important link in supporting accident prevention in the national, European or even global forestry industry. The starting point for the preceding ideas is systematic recording, analysis and interpretation of basic indicators of the level of occupational safety in forestry. This study is focused on analysing the different occupational safety indicators for the period 2015 to 2024, with the aim of a comprehensive analysis of occupational accidents in the forestry sector of the Republic of Croatia at the national and regional level.Data is gathered from occupational safety service databases of the Croatian Forests (CF) Ltd., which manages 2.024 million hectares of forests in the Republic of Croatia. Descriptive and inferential statistical methods were used in filtering, processing and analysing data. The inferential methods for data analyses included correlation analysis, Chi-squared test, and the Kruskal-Wallis H test to analyse the relationships between sampled variables.In the ten-year period analysed, a total of 2,390 occupational accidents were recorded, of which 9 were fatal accidents. In the same period, 157 cases of occupational illness were recorded, of which 67.5% were caused by exposure to vibrations. The annual weighted average of the total number of accidents is 239 according to Croatian Forests Ltd. database (where the company cuts and processes less than 50% of the annual felling volume), while the same indicator for Croatia according to the Eurostat report is 264. The highest average number of accidents is recorded in the process of forest harvesting (58.1%). The riskiest work-related activity is felling and wood processing where on average 48.5% of the total number of accidents occur annually. The annual statistics are 0.37 fatalities per mil. m3, one occupational accident per 10,000 m3 of felled and processed wood (1.00/10,000 m3) and 30 accidents per 1000 employees. So, safety at work in Croatian Forests Ltd. can be assessed as partially satisfactory. Accident analysis from the perspective of the time curve (month, day, hour) shows the highest proportion of accidents occurring in the third annual business quarter, on a weekly basis on Tuesdays, and on a daily basis from 9 to 11 am. The key findings regarding analysis of data at the regional level indicated that medium-size Forest Administrations (FAs) as organizational components of the company CF Ltd. had the highest risk of accidents at work (41.60 accidents per 1000 employees). The indicators of the quantity of felled and processed wood per one accident (m3/accident) revealed an unsatisfactory result achieved by small-size FAs (19,943 m3/accident), and a more satisfactory result achieved by large-size FAs (28,335 m3/accident).The range of numerical values and the weighted average value of occupational safety indicators for the sampled period did not show a decreasing trend, which indicates a similar or lower level of occupational safety in the forestry sector of the Republic of Croatia. The results obtained provide scientists and practitioners with an understanding of the level of occupational safety in Croatian forestry and form the basis for developing proposals for new measures in line with technological progress, as well as new educational programs for the development of work skills and competencies.
Efficient roadside vegetation management is essential for reducing infrastructure maintenance costs and enhancing bioenergy recovery. This study evaluates a modified excavator-based feller equipped with a specific arm to extend its operational reach. Field trials were conducted across three distinct site conditions highway embankment, overpass embankment, and riverbank to assess key performance parameters including cycle times (encompassing moving, positioning, felling, processing, and stacking), fuel consumption, and ground pressure. Compared to the conventional configuration, the arm modification yielded a 66% reduction in moving time and a 34% reduction in stacking time, albeit with a 73% increase in positioning time due to the extended reach. Overall, the modified machine demonstrated enhanced operational efficiency, reduced fuel usage by up to 7%, and decreased soil disturbance through a substantial reduction in the tracked surface area. These findings underscore the potential of telescopic boom technology to improve the versatility and sustainability of mechanised roadside biomass harvesting practices.
Chainsaws are still a commonly used tool in the tending of young forests. The small volume of felled trees makes the use of powerful multi-operational machines economically unjustifiable. Moreover, the use of traditional petrol chainsaws carries a number of problems for the operator, such as noise, vibration, and exhaust fumes. Battery-powered tools can be an alternative to traditional petrol chainsaws. The purpose of this study was to determine the possibilities of using a battery-powered chainsaw during late thinning in terms of productivity and ergonomics. For the study, an ECHO ECCS-58V battery-powered chainsaw and a DOLMAR PS 5000 petrol chainsaw were used. Measurements were conducted in a 14-year-old pine stand. All measured noise level parameters were significantly higher for the petrol chainsaw. The average operating efficiency was 0.15 ha/hfor the battery-powered chainsaw and 0.16 ha/h for the petrol chainsaw. The petrol chainsaw ran for an average of 42.0 minutes on a single tank of fuel, while the battery-powered chainsaw ran for an average of 41.0 minutes on a single battery. It was found that the battery-powered chainsaw achieved very similar productivity. With the availability of a sufficient number of batteries, it can successfully replace traditional petrol chainsaws during the tending of young forests. This will significantly reduce the negative impact of noise, vibrations, and exhaust fumes on the operator, while maintaining satisfactory productivity.
Tree felling with a chainsaw is one of the most dangerous operations, but it cannot be replaced by machines in many areas. Operators are exposed to many hazards including accidents, fatigue, fumes, dust, noise and vibration. This research focuses on vibration exposure and on how it is affected by the accuracy of saw chain maintenance. Although chain sharpening must be carried out according to the manufacturer's instructions, it is common practice to make errors in sharpening angle and chain depth gauge lowering. The aim of the study is to investigate the variations in both cutting efficiency and vibration exposure, considering three different depth gauge lowering levels (standard: 0.65 mm, over-standard: 1.0 and 1.3 mm) when cross-cutting several square beams of two species: European beech (high density) and silver fir (low density). The results show that the increase in cutting efficiency at higher depths was statistically significant for silver fir beams but limited overall (about 10%). On the other hand, the over-standard lowering of the depth gauge did not reduce the cutting time on beech. On the contrary, vibration exposure increased significantly from a minimum of 64% to a maximum of 133% compared to the standard level. These results show that the practice of lowering the depth gauge beyond the standard level is dangerous for the operator and has no real benefit in terms of cutting performance.
Black locust (Robinia pseudoacacia L.) is a key tree species globally and in Hungary, valued for its economic benefits, adaptability, and ecosystem services. Despite its invasiveness and susceptibility to frost damage, its high-quality timber and significant nectar production make it economically important. This research, conducted as a collaboration between the Hungarian Forest Research Institute and the University of Debrecen, aimed to evaluate the applicability of remote sensing technologies in supporting black locust (Robinia pseudoacacia L.) research and monitoring efforts. A clonal trial established in 2020 in eastern Hungary aimed to assess the performance of newly bred black locust clones. Tree height was measured using both conventional ground-based methods and photogrammetric analysis of unmanned aerial system (UAS) data, enabling comparison between the two approaches. Tree vitality was evaluated through UAS-based multispectral analysis using vegetation indices, including NDVI, GNDVI, NDRE, and LCI. Our findings revealed no significant differences (p>0.05) between UAS-based and traditional height measurements, confirming UAS as a reliable tool. Clones »NK2« and »PL251« showed superior growth (height of 7.6 m and 7.4 m) and health, while »Üllői« cultivar performed the weakest (5.3 m). Strong correlations were found between some vegetation indices (NDRE and LCI) and tree heights (r=0.593 and r=0.587), emphasizing the potential of remote sensing in efficient forest management. This study highlights the value of integrating UAS technology in forestry, offering cost-effective, accurate and comprehensive data for improving black locust cultivation practices.
This paper examines the workload of the choker setter in timber extraction operations using the Konrad Mounty 4000 cable yarder. A time study was conducted, measuring the choker setter's heart rate and productivity at varying extraction distances for both & raquo;uphill<< and & raquo;downhill<< extraction directions. Heart rate data collected during work was analyzed to evaluate the choker setter's workload in relation to both extraction distance and direction. A total of 257 cycles were recorded across four worksites with similar harvesting performance characteristics. This heart rate data facilitated an analysis of the choker setter's workload as influenced by field and technological factors. The results indicate that the workload comprises both physical and mental components and generally decreases with increasing extraction distance. The analysis of the measurements indicates that heart rate variability is significantly greater on & raquo;downhill<< worksites compared to & raquo;uphill<< worksites. This is a consequence of the different structure of working time at worksites with different directions of timber extraction, as well as the specific characteristics of the work process that affect the workload of the choker setter. Consequently the study has highlighted some insights that can guide the optimization of the choker setter's workload as an essential factor in designing an efficient work process.
Visual Place Recognition (VPR) enables robots to determine current location by comparing input image against previously stored reference images. It is essential in autonomous location and simultaneous localization and mapping (SLAM). A key task of VPR is evaluating similarity between images, as state-of-the-art deep learning-based approaches have achieved outstanding performance in standard indoor/outdoor scenes. However, the SOTA deep learning-based methods underperform in forestry robotic owing to two challenges, constrained computational capabilities and appearance variation due to seasonal shifts, weather/light/viewpoint variations, which substantially impair visual similarity computation. Consequently, this work proposes ForestsNet, a novel lightweight VPR network, to resolve this issue. First, a Binary Neural Network (BNN) was constructed to achieve considerable memory reduction. A novel binarization function, Leaky Sign, is proposed; it adaptively applies quantization factors to input activations, it retains richer feature information during binarization while significantly reducing accuracy degradation of place recognition. Second, Mixer Forests, a novel multi-layer perceptron-based aggregation method is introduced to integrate global context into feature maps, substantially enhancing the robustness against appearance variation. In addition, two novel evaluation metrics, Memory Allocation Efficiency and Balance Compression Recall, are designed to quantify the trade-off between memory efficiency and place recognition accuracy. Experimental results demonstrate that ForestsNet achieves substantially higher memory usage efficiency than full-precision networks. Compared to state-of-the-art BNNs, it presents superior performance in both memory efficiency and place recognition accuracy, establishing itself as a robust VPR solution for resource-constrained forestry robots.
Mechanical site preparation (MSP) is widely performed around the globe to enhance the success of forest plantations. However, MSP can cause severe soil disturbance whose magnitude depends on various factors such as soil characteristics, soil moisture conditions and the type of machinery. In an experiment conducted in Northeastern France, we analyzed the combined effects of machine size and soil water content on the soil structural quality following MSP performed with subsoilers mounted on excavators. MSP was carried out at two dates with different soil moisture contents (average soil moisture content at 10 cm depth: 40–54% during the wet period, 32–34% during the dry period), using two excavators that differed in weight (2.7 T and 9 T) and engine power, in two forest stands that differed in their soil characteristics (a sandy loam and a clay loam) and that both showed compacted soils. We used a profile wall method (i.e. a visual soil estimation method) to describe the soil structural quality, 2 to 6 months after MSP was performed. All instances of MSP resulted in an increase in the volume of soil favorable to root growth, i.e., soil with a friable structure, compared to the unprepared control (+60%). No evidence of soil degradation due to MSP (compaction, smearing, puddling, voids) was found in comparison to the unprepared control. The larger excavator prepared the soil to a greater depth due to its larger MSP tool and its higher engine capacity (depth of 40 and 53 cm, for the small and large excavator, respectively). MSP performed during the dry period resulted in a greater volume of soil favorable to root growth than MSP performed during the wet period (+60%). Our results indicate that MSP provides better results when conducted during the dry period. Therefore, it is strongly recommended that forest operators carefully consider the timing of MSP operations and prioritize dry soil conditions to execute them.
Forest road planning is a key component of sustainable forest management, as it ensures access to harvesting, transport, and protection activities while minimizing environmental impacts. Recent advances in digital terrain models (DTMs) and automated design tools have opened new possibilities for improving the efficiency of road alignment planning. This study evaluates the potential of an AI-driven corridor planning system (Path Explorer) integrated into RoadEng software compared with the traditional zero-line method in designing the forest road »Osmača–Compartment 56« in Bosnia and Herzegovina. Two DTMs with resolutions of 20×20 m and 90×90 m were used to generate alternative alignments, which were then refined in the Location module to produce preliminary projects and compared with the operational (field-designed) alignment. The results indicate that AI-driven preliminary designs achieved shorter alignments and lower estimated construction costs and earthwork volumes, suggesting potential cost savings under the tested conditions. The operational design, based on field measurements, remained the most accurate and suitable for final implementation in complex terrain. The study emphasizes that these conclusions are limited to the tested DTM resolutions and the specific case study area. Nevertheless, the integration of AI-driven corridor planning systems into early design phases can enhance the efficiency and objectivity of forest road planning, supporting more informed and sustainable engineering decisions.
A substantial percentage of global timber grows on challenging and steep terrain. This resource can be accessed with cable yarders or traction-assisted forwarders. When choosing which harvesting solution to employ for a particular logging job, its energy efficiency should also be considered, given the ever-rising energy cost. This theoretical study develops analytical models to determine and compare the fuel consumption of a cable yarder and a traction-assisted forwarder used for uphill extraction. Simulation results indicate that forwarding requires on average 69% more fuel per unit of transported payload across a range of work conditions (e.g. distance, slope gradient, payload size) and at least 47% more fuel under all simulated conditions. The efficiency advantage of cable yarding could be boosted further by transporting heavier loads. In cable yarding, a significant portion of fuel is consumed during standstill, when the engine runs idle or at low efficiency for extended periods of time. In forwarding, loading and unloading are particularly energy intensive. While cable yarders could greatly benefit from electrification, forwarders may see lower efficiency gains due to the challenges in electrifying linear hydraulic actuators in a cost-effective and robust manner.
Fully mechanized hardwood operations usually rely on manual bucking, which may decrease potential volume and value recovery as compared to automatic bucking. This study aimed at evaluating bucking done in cut-to-length (CTL) operations, hypothesizing a possible increase of value recovery by using a bucking optimizer. Evaluation of bucking was done by a comparison of actual bucking with a mathematical solution – BuckR – a bucking optimizer based on dynamic programming. Results showed that the mathematical approach significantly outperformed manual bucking when quality was not considered, and only main stem products were included. Throughout 315 study trees (Acer saccharum, Acer rubrum, and Betula alleghaniensis) located in the Acadian forests of Eastern Canada, a mean increase of 112% in value and 84% in volume recovery per tree was reached through mathematical optimization. Contributions of mathematical bucking can be summarized by 1) a significantly higher number of sawlogs and a greater mean log length of sawlogs and pulp logs, 2) a two-fold increase in processed height, and 3) a considerable reduction in minimal diameters as compared to actual bucking. Those results illustrate the possibility to increase value and volume recovery in mechanized CTL operations and therefore to utilize the wooden resource more efficiently.
In countries where private forest ownership prevails, wood mobilisation largely depends on the interest and commitment of private forest owners (PFOs) to manage their forests. In Slovenia, where 77% of forests are privately owned but insufficiently managed, this issue is particularly important. Governments not only in Slovenia, but also in other EU countries, have responded to this challenge of wood mobilisation by proposing various policy instruments, addressing the importance of PFO business cooperation both among themselves and with forest service providers. This study aims to find out: 1) whether there is a potential for business cooperation among PFOs; 2) which of the existing business organisational forms are appropriate for PFOs and 3) how property characteristics, forest management activity, forest management objectives and PFOs’ socio-demographic and economic characteristics influence PFOs readiness to engage into business cooperation. Based on the results, policy recommendations were developed to increase the share of PFO cooperating in business organisational forms in Slovenia. An online survey conducted in March and April 2022 among randomly selected Slovenian households owning forests provided data for two logistic regression models. The results show that PFOs are in general ready to engage in business cooperation; however, according to them, none of the currently existing forms of business cooperation is considered as very appropriate. The model »cooperation with other PFOs in business organisational forms« suggests that PFOs readiness to engage is influenced by forest management objectives related to carbon sequestration, wood and biomass production for personal use and for the market, as well as preserving forests as an investment for the future, PFO age and level of education. The model »cooperation with forest service providers« shows that PFOs readiness to engage is influenced by forest management objectives related to biodiversity conservation and wood and biomass production for personal use and for the market, as well as by PFO level of education. To motivate PFOs to engage into business cooperation, forest policy decision makers must take a holistic transformation of forest policy. To achieve the desired forest policy outcomes, the transformation should incorporate innovative and supportive policy instruments.
The investigation of the effects of the noise generated by harvesting equipment on the environment is one of the important topics in sustainable forestry. During timber harvesting, not only workers but also wildlife are exposed to the noise generated. Exposure to noise has both direct and indirect effects on humans and wildlife. The negative effects of noise exposure can be observed depending on its intensity and duration. Noise exposure, which has various psychological and physiological effects on humans, also negatively affects plants and animals. In this study, sound measurements of the chainsaw were conducted during thinning operations Province. The measurement area is a Turkish red pine (Pinus brutia Ten.) stand with a canopy density of 60-65%, a slope of 30-35%, and tree diameters ranging from 20 to 35 centimeters. The noise emitted by the chainsaw during production, ranging from approximately 1 meter to 200 meters, has been modeled using a feedforward Artificial Neural Network (FANN) for sound propagation. The measurement data was used 60% for training, 20% for testing, and 20% for validation. Random trees were assigned to noise attenuation effects on the sound according to the stand characteristics of the study area. Thus, it was aimed to create a realistic sound propagation model and estimation maps. The average performance metrics of the model, RMSE and R2 values, were calculated as 4.84 and 0.88, respectively. According to the sound propagation model predicted by the FANN as a static model, it is estimated that the distance at which the chainsaw could affect wildlife behavior is 400 meters or less.
In forest operations, economic advantages can be obtained by increasing the distance between the skid trails. This protects soil by reducing the compacted area, while at the same time increasing the productive timber ground area. These advantages are offset by disadvantages, as fully mechanized timber harvesting is not possible and motor-manual felling is required for the areas that cannot be reached by the harvester. This in turn reduces work safety and increases the workload and personnel requirements, possibly leading to higher timber harvesting costs. To analyze the consequences of an extended skid trail distance, a timber harvest under real conditions was carried out in north-eastern Germany in the fall of 2023. In a 72-year-old Scots pine (Pinus sylvestris, L.) stand with an area of 23.2 ha, 692 m3 of wood was harvested in a thinning operation. Three different timber harvesting methods with extended skid trail distances of approx. 40 m (ES) were investigated and compared to a fully mechanized system with conventional skid trail distances of approx. 20 m (CS) in a time study with a total of 150 recorded hours. Following the harvest, the residual stand damage was also recorded. The timber harvesting methods with ES had higher timber harvesting costs than the method with CS, although there are major differences between the three semi-mechanized timber harvesting systems: The productivity of the harvester increases as the number of passes by the harvester decreases (from 13.87 to 14.09 to 15.99 m3/PMH15). Looking at the forwarder productivity, it is higher in ES than in CS. Finally, the costs of the harvesting systems ranged between 29.18 /m3 for CS to 30.40, 32.41, 34.56 /m3, respectively, for ES. There is no significant difference in the residual stand damage across the methods. The productivity of semimechanized timber harvesting methods can be improved if the motor-manual felling is carried out before the harvester is used and if the trees are not winched with a cable tractor.
In some specific yarding situations ground contact needs to be reduced to an absolute mini-mum. In such cases, a double-hitch system can serve as an alternative to helicopter logging. In the present investigation, a cable yarding operation using double-hitch carriages for selec-tive thinning in intermediate slope was accompanied by a time and motion study to better understand the working steps and resulting productivity of cable yarding in this unconven-tional situation. The main aims were to develop a model that can be used to estimate the productivity of comparable yarding operations before their implementation, and to quantify the additional effort relative to the use of a cable yarder with a single-hitch carriage. The results indicated an overall productivity of 5.69 m3ob per productive system hour including delays of up to 15 minutes (PSH15), leading to yarding costs of 104.40 (m3ob)-1 (in addition to costs for installation and felling). Compared to a system with a common single-hitch carriage, the inclusion of a second lateral yarding stage increased time and cost, contributing 13.30 (m3ob)-1. Key factors affecting productivity were payload size, number of load pieces, and ter-rain difficulty, while yarding distances had minimal impact. Compared with traditional yard-ing, double-hitch operations were less productive and more costly, but they can still be con-sidered a suitable alternative to helicopter logging in sensitive areas.
In Bosnia and Herzegovina, black alder appears in scattered smaller forest stands, fragments and patches that are still not spatially separated and allocated in management plans, despite its high ecological importance. The objective of this study is to model a black alder ecological niche considering combined effects of climate, hydrological and air quality determinants to support decision-making of conservation and restoration activities on a local/regional level. Black alder occurrence was registered on 72 temporary sample plots representing about 1500 trees in the variables (average annual temperature, minimum temperature, maximum temperature, sum of temperature above 5 degrees, sum of precipitation, maximum precipitation), five hydrological variables (average annual flow, minimum flow, maximum flow, flow between 1961-1990 and water level) and five air quality variables (average annual concentration of air particulate matter of PM2.5 and PM10 mm, SO2, NO2, maximum CO2) were interpolated spatially on 10 m grain size based on hydro-meteorological data from 13 national stations. The MaxEnt method was used to predict spatial distribution model, where predicted occurrence probabilities are classified in habitat suitability classes. The MaxEnt model revealed high-quality spatial prediction (AUC=0.95). The most significant determinants were average annual sum of precipitation and average annual 24-hour maximum CO2 concentration (cumulative about a 72% contribution). The highest occurrence probabilities were related to areas with less than 1400 mm of annual sum precipitation and elevated CO2 linked to low NO2. The areas with high species occurrence are mainly located in continental Bosnian Internal Dinarides in the valley and partly on hilly and sub-mountainous positions overlapping pedunculated oak-hornbeam and Illyrian submountainous beech forests. Modeled ranges of precipitations and air variables concentrations indicate that black alder prefers continental low hilly and plane positions covering forest edges, although some suitable ecological niches are predicted in sub-urban and peri-urban green areas. The obtained model of species distribution determined spatially ecological niches important for conservation and restoration to maintain ecological services and biodiversity as well as aesthetic and recreational roles of black alder, which are important for local communities.
Modern forest harvesting machinery generate an abundance of underutilised data in their control systems. The Controller Area Network (CAN) bus data stream offers the opportunity to investi gate the operation of the machinery in detail while in real world harvesting scenarios. This study uses CAN data to assess a component of operator workload in forwarder operations, by introduc ing a method to interpret forwarder joystick movements. The data was captured in a clearfell logging operation case study in Canterbury, New Zealand. The joystick data was then analysed to determine the time and number of operator input movements per load cycle ("grab") total ling 418 grabs. This, combined with video analysis, identified independent variables that describe the grab cycle such as time of day, number of logs per grab, log grade, and the activities of "pencilling" (vertical drop of logs in grab to align large ends) and "dropping" (releasing logs from the grab before loading). Factors that significantly affected the operators' time taken to complete the grab and the number of required joystick movements included number of logs, pencilling and dropping. For example, the average load cycle was 18 seconds for four logs, and this increased by 6.1 seconds and 14.4 seconds per grab when pencilling or dropping, respec tively. Average total joystick movements were similar to 108 per grab. This case study demonstrated that CAN bus data can be used to improve our understanding of the operation of harvesting equipment such as forwarders. An example use of the result is to share and compare this with the harvester operator (who crosscuts and sets out the logs for forwarder collection) as the need for pencilling and dropping is a consequence of misalignment or debris caught in the piles. The method used also presents an opportunity for human factors research, particularly in operator fatigue management and training through the measurement of joystick movements with a genuine possibility of realtime performance feedback.
