A Combination of Traditional and Mechanized Logging for Protected Areas

Abstract

Teaming draught animals with modern forest machines may offer an innovative low-impact solution to biomass harvesting in protected areas. Machine traffic only occurs on pre-designated access corridors set 50 m apart, while trees are cut with chainsaws and dragged to the corridor’s edge by draught horses. The operation presented in this study included one chainsaw operator, two draught horses with their driver, an excavator-based processor with its driver and a helper equipped with a chainsaw for knocking off forks and large branches, and a light forwarder (7 t) with his driver. Researchers assessed work productivity and harvesting cost through a time study repeated on 20 sample plots. Descriptive statistics were used to estimate productivity and cost benchmark figures, which were matched against the existing references for the traditional alternatives. The new system achieved a productivity in excess of 4 m³ over bark per scheduled hour (including delays). Harvesting cost averaged EUR 53 m⁻³, which was between 15% and 30% cheaper than the traditional alternatives. What is more, the new system increased labor and horse productivity by a factor of 2 and 7, respectively, which can effectively counteract the increasingly severe shortage of men and animals.

1. Introduction

Urban economy and culture dominate all aspects of life in industrialized countries [1]. That has radically changed the role of the surrounding agrarian landscape, which has turned from a main supplier of food and raw materials into a provider of amenity and other intangible benefits [2,3]. The land anxiety of urban dwellers and the loss of economic value of agricultural land have resulted in the rapid expansion of protected areas, where economic activities are restricted.

In Italy alone, protected areas occupy over 3 million hectares, accounting for almost 10% of the Italian territory [4]. The distribution of protected areas across the National territory is quite heterogenous and depends on several factors, not least the intrinsic naturalistic value of the sites and the alternatives for economic development. Out of 22 administrative regions, Abruzzo tops the list: it hosts 109 protected areas, including 3 large National Parks, for a total protected area exceeding 390,000 ha, or 37% of the regional land surface [5]. Protection status can be decreed by different levels in the administration (e.g., national, regional, etc.) and it may conform to different rules, but it always involves restrictions to use motivated by public concern [6]. That may represent an obstacle to the financial sustainability of forest management, further motivating abandonment and possible degradation.

In the protected areas of Abruzzo, a main obstacle to cost-effective forest management is the requirement of animal logging, often prescribed by the forest authorities. Such prescription reflects the traditional use of pack animals, which have been widely available in the region until recent times [7]. Unfavorable demographics and lifestyle changes have much reduced the number of animal loggers, and the few remaining contractors cannot match the increased demand for their services, despite a general rise in contracting rates. The latter is not high enough to stimulate recruitment among animal loggers, but it often exceeds the buying power of forest owners who struggle with the increased management cost. Owners would like to mechanize wood extraction, replacing traditional pack mules with their mechanical substitute, namely, farm tractors equipped with loading platforms fore and aft, which is a popular solution for extracting 1 m long firewood logs in most of central and southern Italy [8].

Both the authorities and the forest owners are right: the authorities fear that the indiscriminate movement of mechanical equipment in the forest will cause serious damage to the soil and the residual stand [9]; the owners demand the mechanization of forest operations, without which, no economic and social sustainability can be achieved [10].

Both are also wrong, however: pack animals are not the only solution capable of preventing forest damage, while “tractor-based pack logging” is certainly not the most rational system for improving the economic sustainability of forest operations—and it is a complete failure when it comes to social sustainability [11].

To help break the current deadlock, the National Council for Research (CNR) was tasked with developing a compromise solution, which may respond to the needs of both parties. In turn, the CNR designed, tested, and demonstrated an entirely new system, as described in this study. The goal of this study was to assess system performance, in terms of work productivity and harvesting cost, and produce benchmark figures that can be used to promote a new way to modernization, which harmonizes technical development with environmental protection and the conservation of cultural heritage [12]. The new integrated harvesting system fulfills the request of regional authorities for using work animals: in addition to its environmental benefits, animal logging has historical and cultural value, which should be preserved by all available means [13]. The best way to preserve animal logging is to make it competitive on the market; pack logging is not, which explains its rapid decline [14]. The proposed new working method features a targeted integration of animals and machines, which can make animal work faster, easier, and more competitive [15]. In turn, any increase in team productivity allows introducing animal logging to more harvesting sites than the current limited availability of working animal permits. At the same time, the new technique results in a strong mitigation of the environmental impacts, compared with conventional modern technology [16].

2. Materials and Methods

2.1. The Harvesting System

Under the new system, harvesting is organized in 50 m-wide swathes, each cut in the middle by a 4 m wide corridor. Trees are cut with chainsaws, then dragged to the edge of the corridors by draught horses. Since the corridors are set at a 50 m spacing, horses drag their load only very short distances, reaching a very high productivity [17]. The typical team consists of one chainsaw operator, one horse driver, and two horses, so that the animals can work and rest in turns. Once at the corridor’s edge, trees are picked up by an excavator-based processor and turned into 2.3 m logs. Finally, a light forwarder (7 t capacity) collects the logs and moves them to the roadside landing. Simple, fast, and painless (Figure 1).

All operators were skilled professionals and were familiar with their respective equipment (or animals). All operators were inducted about the goal of the study and its methods, and they all agreed to participate. Due compensation was offered to all entrepreneurs for the disturbance caused by this study; in fact, they were all interested in learning the new work method, which was conducive to fruitful cooperation.

There is nothing particularly innovative about the animals and machines employed at the new site, but the organization of work represents a clear break with tradition in terms of at least two key elements.

First, the planning of in-stand traffic, which is permitted only on the corridors. To foster precision forestry, and to ease the operational work on-site, the corridors were first designed digitally using QGIS 1.1 and its Forest Road Network Plugin 1.6 [18]. The latter is a tool for sketching an optimized forest road network based on the topographic characteristics of the site. The fundamental idea consists of developing the road network out of a cost map and following the principle of the least cost. The software calculates the cost to cross each pixel on the cost map and selects the path of total minimum cost. This work requires a digital terrain model, which is freely available on the web for the entire Abruzzo region at a resolution of 10 × 10 m² [19]. For the cost calculation, the following parameters were integrated: (i) a layer of existing roads (where the crossing cost equals zero), (ii) the gradient of each pixel, and (iii) the topographical conditions within each pixel, as represented by the length of the contour lines. For simplicity, the resulting costs were ranked in five classes. It is important to note that the evaluated costs represent relative values. Since the cost-structure of our operational zone is very homogeneous, the first draft of the corridors was not conducted by the plugin-specific algorithm, but visually and by hand, using the previously generated cost map as a reference. The main criterion when drafting the corridors manually (apart from selecting the paths with least costs) was to arrange them at an angle to the main forest road, to facilitate turning and minimize visual impact, which is especially critical in a National Park. Then, all mapped corridors were checked in the field, together with the driver of the processor, which was the least mobile unit in the fleet. Finally, the final approved corridors were flagged with colored ribbons, so that the felling team would know in which direction to fell the trees and where to pull them.

The second strategic innovation of the proposed technique is the lengthening of the assortments: we dropped the traditional 1 m long firewood specification (essentially designed to facilitate manual handling) and processed all trees into 2.3 m long logs, which is the minimum length for effective mechanical handling. What is more, lengthening the assortments favors better use of that share of valuable timber, which may be turned into better products than humble firewood.

2.2. The Test Site

System performance was tested in a 4-day trial organized in the week of 16–20 October 2023, under good weather conditions, warm and dry all along. The trial was conducted in the “Macchia Grande” forest, at the border of the Gran Sasso National Park, just below the valley station of the scenic cableway, at an elevation of about 1000 m a.s.l. (coordinates: 42° 25′45.22” N, 13° 30′52.37” E; Figure 2). The forest was a converted coppice dominated by turkey oak (Quercus cerris L.), with downy oak (Q. pubescens L.) and English oak (Q. petraea L.) as secondary species, and a dense understory of maple and ash. The silvicultural prescription was a light thinning from below, aimed at removing about 20 percent of the standing mass, equal to approximately 30 m³ ha⁻¹ (

Table 1. Description of the test site.

Municipality		Assergi
Province		Aquila
Surface area	hectares	13
Altitude	m a.s.l.	1030–1090
Slope gradient	%	20
Trail gradient	%	15
Species		Quercus cerris L., Q. pubescens Willd., Fraxinus ornus L.
Management		Converted coppice
Treatment		Thinning
Age	Years	~40
Removal	m3 ha−1	27.8
Removal	trees ha−1	120
Removal	% total	20–25
Tree DBH	m	0.19
Stem height	m	15.8
Stem volume	m3	0. 232

).

2.3. The Time Study

System performance was recorded for each of the three main tasks (e.g., felling and pre-skidding, processing, and extraction) using the classic time-and-motion study technique [20]. For that purpose, 20 plots were marked inside the forest, placed at different depths within four 50 m work bands (swathes). For each plot, researchers determined the number and the diameter at breast height (DBH) of all trees being cut, the time to fell and pre-skid them, and the time to process them with the excavator-based processor. Since the volume accumulated on a single plot would not match the forwarder payload exactly, the extraction productivity was determined with a cycle-level time-and-motion study, rather than a block-level one, as for the previous two work steps. In both cases, delay time was recorded separately and added to the productive work time recorded for each block or cycle as a delay factor, calculated as the ratio of all delay time to all productive time for that team, which served to smooth down the impact of erratic delay occurrence [21]. Product volume was estimated based on DBH values, as follows: a DBH-to-height curve was developed using 30 sample trees, distributed across the whole diameter range; then, the measured DBH and the estimated height of each tree in each plot was entered into the dedicated volume functions developed for the main tree species of the Italian forests by Tabacchi et al. [22]. As a further check, the volume calculated for each plot was matched against the volume measured by the processor on-board computer for that same plot, which had been carefully recorded upon completion of the plot. The few additional trees cut by the processor were also measured, but they were counted separately (i.e., attributed to the processor only).

2.4. Costing

Machine costs were calculated with the method developed within the scope of COST Action FP0902 [23]. Costing assumptions (investment cost, service life, fuel use, etc.) were obtained directly from each individual machine owner, or from the manufacturers. Labor cost was assumed to be EUR 15 per scheduled machine hour (SMH), according to the approved regional price lists (

Table 2. Costing: assumptions, estimates, and total cost.

Team		Chainsaw	Horses	Processor	Forwarder	Mules	Tractor
Make		Stihl	(n = 2) animals	Doosan	Novotny	(n = 6) animals	Valtra
Model		MS 261	AITPR	DX 140	LVS 720	Mules	6050
Power	kW	3	0.7	86	100	3	75
Investment	Euro	1200	10,000	210,000	200,000	22,500	65,000
Resale	Euro	0	2000	63,000	60,000	4500	19,500
Service life	Years	2	15	8	8	15	8
Utilization	h year−1	1200	1200	1200	1200	1200	1200
Interest rate	%	3	3	3	3	3	3
Depreciation	Euro year−1	420	530	18,375	17,500	1200	5690
Interests	Euro year−1	30	188	4370	4160	420	1350
Insurance	Euro year−1	1000	1000	2500	2500	1000	2500
Fodder and care	Euro year−1	0	11,600	0	0	26,100	0
Fuel and lube	Euro year−1	1800	0	20,590	19,050	0	11,880
Repairs	Euro year−1	250	0	9190	8750	0	5690
Total	Euro h−1	2.9	11.1	45.9	43.3	23.9	22.6
Crew	n.	1	1	2	1	2	2
Labor	Euro h−1	15.0	15.0	30.0	15.0	30.0	30.0
Overheads	Euro h−1	3.6	5.2	15.2	11.7	10.8	10.5
Total rate	Euro h−1	21.5	31.3	91.0	70.0	64.7	63.1

Notes: h = scheduled machine hour, including delay time; AITPR = Italian fast-draught horse breed; the Doosan DX140 excavator was fitted with a Keto 150 harvester head; a second worker was teamed with the processor, equipped with a chainsaw and tasked with knocking down the largest branches, which would have hindered smooth processing; all cost in Euro (EUR) as on 23 January 2024.

).

2.5. Comparison with Traditional Systems

Limited time and finances excluded a proper comparative study, which would have been the best and most direct way to check whether the new proposed system would increase profitability, compared with the traditional alternatives, animal or mechanized. As the second-best option, we estimated the productivity of those alternatives based on the existing literature. Several studies have documented the productivity achieved when felling and processing hardwood trees into 1 m logs with chainsaws [24,25,26,27], and when extracting those logs with pack mules or tractors equipped with front and rear containers [8,11,24,26,27,28,29,30,31]. Those studies were carefully reviewed, screened for their compatibility with the work conditions met at “Macchia Grande”, and eventually gathered in a database for extracting average productivity figures. Cost was recalculated for 2024 (several studies were >20 years old) using the same method described above.

2.6. Data Analysis

Data analysis consisted of the extraction of simple descriptive statistics aimed at providing solid indicators for centrality and variability (i.e., means, standard deviation, interquartile range, etc.). Those are the main qualities used for assessing the value and reliability of the general benchmarks we wanted to obtain. Bivariate plots and regression analysis were also attempted with the sole purpose of showing general trends and assessing if those trends were reasonable or would denounce gross errors in the dataset, or in the experiment. In fact, this study was not designed for productivity modeling, and therefore the trends described in the regression graphs should not be used for making detailed predictions.

3. Results

The forwarding corridors drafted with the help of the QGIS tool are shown in Figure 3. The paths were laid at an angle to the main forest road in order to facilitate turning into the road and to mitigate visual impact for the many visitors of the National Park. The corridors tend to turn around the patches with a higher cost (i.e., orange-colored pixels), accordant with the principle of least cost design.

The new harvesting system performed well, achieving an average productivity slightly above 4 m³ per scheduled machine hour, or over one full load (30 t) per day. The data in

Table 3. Productivity and cost summary table.

Felling–Pre-Skidding (n = 20)		Mean	Std. Dev.	Median	LQ	UQ
Productivity	m3 h−1	4.4	1.3	4.4	3.5	5.3
Cost	EUR m−3	13.0	3.9	12.2	10.2	15.8
Tree size	m3	0.192	0.03	0.193	0.177	0.207
Processing (n = 16)
Productivity	m3 h−1	4.2	1.0	4.1	3.5	5.0
Cost	EUR m−3	22.8	5.3	22.4	18.7	26.9
Tree size	m3	0.186	0.036	0.19	0.161	0.212
Extraction (n = 15)
Productivity	m3 h−1	4.1	0.9	4.0	3.4	4.8
Cost	EUR m−3	18.0	3.9	17.6	14.9	21.2
Load size	m3	5.2	-	-	-	-

Notes: n = number of observations (i.e., blocks or work cycles); m³ = solid volume over bark, including branches to a minimum diameter of 5 cm; h = scheduled machine hour, including delay time; LQ = lower quartile; UQ = upper quartile; no statistics are available for the forwarder load size, since load size was not measured individually for each cycle, but was taken as the grand mean (i.e., total m3 divided number of cycles).

show a remarkable balance between the three teams engaged with the main tasks, namely, felling and pre-skidding, processing, and extraction. That facilitates operational planning, avoiding the need for complex scheduling. A second element emerging from a close look at the table is the relative stability of the work routines, which is demonstrated by the relatively low variability in the dataset. Under similar work conditions, one obtains similar performance levels, which is a witness to smooth operation. That is further confirmed by the incidence of delays, which is well within the bounds of operational norms, and amounts to 22%, 30%, and 23% of total worksite time, respectively, for the felling and pre-skidding team, the processor, and the forwarder. In fact, the incidence of delays could be further reduced for the processor, which should be introduced later than it was in this trial, to increase the distance from the felling and pre-skidding team. The organizational constraints of this study and the need to run a public demonstration on the 5th day did not permit waiting for one additional day before introducing the processor to the site, and therefore the machine had to be paused occasionally, if it got too close to the felling team.

Of course, productivity will change with working conditions, and especially with stem size and extraction distance; the former will affect mostly felling and processing, the latter pre-skidding and forwarding. The impact of stem size on the productivity of both felling and processing is easily detected just by observing the scatterplots of our data (Figure 4), even if the study was not specifically designed to determine such an effect. However, the graphs clearly demonstrate that changing working conditions may affect system balance, thus requiring precise operational management. That would be the case of stands offering larger stems and longer extraction distances; the larger stems will increase felling and processing productivity, while the longer extraction distances will decrease forwarding productivity, so that one will need to schedule fewer felling and processing days, but more forwarding days. The productivity and cost achieved with the new proposed system compare quite favorably with those of the two traditional alternatives, pack mules and pack tractors (

Table 4. Comparison with the two traditional options—animal and mechanized.

		Manual	Mechanized	Integrated
Felling-processing	Equipment	Chainsaw	Chainsaw	Chainsaw, Horses, Processor
Crew	n°	1	1	4
Cost	EUR h−1	21.5	21.5	143.8
Productivity	m3 h−1	0.56	0.56	4.02
Cost	EUR m−3	38.4	38.4	35.8
Labor	man h m−3	1.79	1.79	1.00
Extraction	Equipment	Mules (6)	Tractor	Forwarder
Crew	n°	2	2	1
Cost	EUR h−1	64.7	63.1	70.0
Productivity	m3 h−1	1.8	2.8	4.1
Cost	EUR m−3	35.9	22.5	17.1
Labor	man h m−3	1.11	0.71	0.24
Total cost	EUR m−3	74.3	60.9	52.9
Total labor	man h m−3	2.90	2.50	1.24

Notes: h = scheduled machine hour (or worker), including delay time; m³ = solid volume over bark, including branches to a minimum diameter of 5 cm.

). At EUR 53 m⁻³ the new system is 30% cheaper than the old traditional system based on pack mules, and 15% cheaper than the new traditional system based on pack tractors. What is more, the new system requires a much lower manual input: half as much as for the more mechanized between the two traditional systems (i.e., pack tractors), and even less for the other one (i.e., pack mules). The dramatic reduction in manual input is the indicator of a less tiresome and uncomfortable job, which may become a strategic asset for attracting new recruits to a dwindling loggers’ community.

If the new system dramatically boosts labor efficiency, it boosts horse efficiency even more. One horse used for pre-skidding on short distances achieves the same productivity as seven horses used for pack hauling on medium to long distances (300–600 m). In other words, the same number of animals can cover 7 times the surface they could cover if they were used according to the old system. Therefore, smart mechanization becomes the way to introduce animal logging to more settings, not fewer.

4. Discussion and Conclusions

Before endeavoring into a proper discussion of our results, we need to make sure readers are aware of the main limitations of this study, namely, the absence of a proper side-to-side comparison and the limited time left to the test team for getting acquainted with the new system. The indirect comparison between our results and those presented in the literature for the alternatives is approximate by its own nature; therefore, its results must be taken as suggestive, not conclusive. In fact, the value of any such comparison is made weaker by the fact that our test team had very little time to hone the new work technique. For that reason, our results must be considered as a baseline that can be further improved with a little more experience. However, the individual tasks assigned to the crew members were the same as they would normally perform every day; what changed was the settings and the general organization of the workflow. So, if something needs to be improved, that is the interaction between teams, not their work routines. For that reason, one may expect that the margin for improvement is incremental, not radical. Although relatively small, such margin for improvement tends to corroborate the results of our preliminary comparison, which points at the new integrated system as a most likely winner. One may not quote as exact references the 15% or 30% margins estimated with this study, but certainly all indicators point at the better performance of the new system, despite its potential for further improvement, which—once released—would make it an even stronger contestant.

Both traditional systems are very weak competitors because they rely on motor–manual processing. Delimbing and crosscutting are the two most time-consuming tasks in all tree harvesting systems, and especially in small-tree operations [32]. For that reason, a system that relies on mechanized delimbing and crosscutting will always enjoy a very large advantage over any other system relying on manual delimbing and crosscutting. Our study just confirms that, while suggesting that the margin of the new mechanized system is likely to vary from 15 to over 30%, under the conditions of the test. Those conditions include a relatively low labor cost (EUR 15 h⁻¹), which reflects current local rate, and there, “current” is the crucial attribute. For most industries and in most of Italy, labor cost is higher than assumed in this study, and it is bound to increase in Abruzzo as well. The eventual rise of labor cost will further increase the edge of the more mechanized system, thus reinforcing the findings of this study.

Ultimately, this study may not offer exact figures for the cost advantage enjoyed by the new system, but it does confirm it is the winner, and its better performance is not limited to a lower harvesting cost, either. Compared with the traditional systems, the new proposed system grants at least three further advantages: better product value, higher labor efficiency, and stricter control of in-stand traffic.

First, by producing longer logs, the new system offers a wider range of opportunities for adding value to the harvest. One can now think of manufacturing posts, poles, and even timber from the occasional English oaks that populate the transitional high forest. Given that one of the targets of coppice conversion is to increase the proportion of noble hardwoods in the mix, post-conversion treatments will increasingly yield a certain amount of those species. If those stems are crosscut into 1 m logs, they can only be used as firewood, which is a poor use of their value potential.

Second, by doubling labor productivity, the new system allows coping with the growing lack of forest workers, which is a generalized global trend and has motivated countless efforts to increase labor productivity even when engaged with simple manual tasks. In fact, such a large productivity increase would justify a general wage increase that would help in recruiting new workers and retaining old ones.

Third, confining machine access within a proper design corridor network is the best guarantee for a successful transition to low-impact mechanized harvesting [33]. The fear for high site impact is the main reason why authorities often prohibit machine access. In fact, the term “access” needs to be qualified. So far, it has been interpreted as “unrestricted access” by both parties. If so, the potential for damage is high and the concerns expressed by the authorities are justified [34]. Regulated access is the most sensible solution, which is routinely adopted by experienced forest managers [35]. The system we propose follows that philosophy, which is one of its integral components. Machine traffic is concentrated on straight corridors, where impact is minimized by avoiding sharp turns, by handling pre-bunched stems (or logs), and by using relatively compact and agile forest machines (a 14 t excavator and a 7 t 8-wheeled forwarder). The result is a strong mitigation of all site impacts, which should reassure forest owners and authorities [36,37].

The new system is particularly well suited to light harvests, as is generally the case of many of the protected forests in Central and Southern Italy. Those forests derive from coppice stands, which are being restored into uneven aged high forests. Ecosystem restoration work is conducted through repeated light thinning operations, which are especially constraining for highly productive modern machines. Satisfactory performance is only achieved if the harvest density reaches a certain critical level, which is hard to do when the removal is in the order of 30 m³ ha⁻¹. By concentrating the wood on the access corridors, the new system achieves that critical density that supports effective mechanical processing and extraction.

Finally, the proposed new system offers a unique opportunity to preserve animal logging as a modern work technique rather than a relic of the past, artificially kept alive by subsidies that—incidentally—have never been released in its favor within the territory considered for this study.

One may wonder why such an effective system has not been designed and tested before, if it offers so many benefits. In fact, it has. The integration of animal and mechanized logging has been documented for Germany, Hungary, Italy, Poland, and Romania—as far as Europe is concerned [38,39,40,41]. However, the few scientific papers that analyze in detail the efficiency of animal–machine integration concern cases where animals are used alongside sub-standard mechanization and help compensate for its weaknesses [42]. In the present study, we have built and tested an optimized system, using state-of-the-art equipment, not obsolescent makeshift solutions, so common in logging operations. In fact, the idea of limiting machine traffic to widely spaced designated trails and pre-skidding whole trees to the trails is not new, either. That is, rather, becoming the norm in Germany, where regulations severely constrain in-stand machine traffic [43,44]. For that reason, manufacturers have developed a wide variety of compact tractors, specifically designed for pre-skidding on short distances [45]. Those machines are generally tracked, remote controlled, and far more expensive than a horse team (e.g., EUR 50–90,000 vs. EUR 10,000). Of course, that is the only solution once horse teams have become extinct in a region; that is not the case of Abruzzo, where reconversion to the new integrated system would make lots of sense and should be done as soon as possible, before horse loggers disappear there, too [14]. A better use of the animal pool still available in the region would prevent its eventual extinction, while accruing technical, financial, and cultural benefits [46]. Of course, the implementation of the new system requires some adjustments, including the re-training of existing animal logging teams. Skidding is not the same as hauling on pack saddles, and it requires different skills and equipment. Horses are better at skidding than mules, but the latter can be used, too; they just need to be re-trained, and so do their drivers. Efficiency increase and re-training may not be enough, however. The reason why loggers may prefer to spend EUR 100,000 on a mini forestry crawler (MFC) rather than EUR 10,000 on a pair of draught horses is that the latter require constant care, whereas the MFC can be stored in a garage when not in use. Therefore, one needs to devise new solutions for managing one’s horses, so that new owners can still enjoy their modern lifestyle without excessive sacrifice. Among such new solutions, one may consider the establishment of cooperation agreements with other horse loggers, or with conventional animal farms—the latter still widespread in all rural settings.