Performance of Harvesting Residue Treatment Methods in South African Pine Plantations

Abstract

Forest harvesting generates variable amounts of residue that pose a fire hazard and a hindrance to regeneration and must be managed accordingly. In South Africa, burning is the most common residue management method, but there is interest in introducing safer and more effective techniques, such as mulching. For that reason, a productivity study was conducted in the Eastern Cape province after a mechanised cut-to-length (harvester and forwarder) harvesting operation to gather information on the productivity, cost, and work quality of the three main alternatives: manual broadcasting, manual broadcasting followed by mulching with an adapted farm tractor and mulching with a purpose-built mulcher. The stem wood residues after harvesting ranged from 12 to 14 ODT/ha. The average time consumption was 16 worker h/ha for manual broadcasting, 3.6 worker h/ha for mulching with a farm tractor, and 0.9 worker h/ha for mulching using a purpose-built mulcher (one pass). Manual broadcasting was the cheapest option, at an average cost of 400 ZAR/ha. Mulching with a farm tractor and a purpose-built mulcher incurred an average cost of 3267 ZAR/ha and 4083 ZAR/ha, respectively. Manual broadcasting achieved a minimal reduction in residue size, with 50% of the slash (branches and stem wood) having a mean length greater than 40 cm. When mulching with a farm tractor was applied, 49% of the slash (branches and stem wood) length was reduced to about 30 cm. When a purpose-built mulcher was used, only 10% of the slash elements exceeded 40 cm in length.

1. Introduction

Tree plantations are intensively managed crops designed for the sustainable and competitive supply of a whole range of wood-based products [1], such as pulp, paper, clothing, furniture, and building structures. According to [2], wood product demand will increase from 2.3 to 3.1 billion m³ of roundwood equivalent between 2020 and 2050. In plantation forestry, the management of harvesting residue offers a major challenge to replanting. From an operational point of view, residue management is a key component of soil nutrient management and forest fire prevention [3,4,5]. There are several ways in which harvesting residues can be managed, such as burning, broadcasting, mulching, or collecting for energy use. The choice of the most suitable residue management method depends on several factors, such as terrain, weather, available markets, and residue load [1,6,7].

In South Africa, out of a total annual area of 41,869 ha surveyed, burning (78%) is the most common way to manage harvesting residues, followed by broadcasting with 11% and mulching with 8% [8]. Burning is a preferred option because it is relatively cheap when compared to mechanical treatments such as mulching and chopper rolling [9]. Broadcasting is the process of distributing residues left on site after harvesting. Broadcasting offers the benefits of retaining nutrients and minimising water and soil loss [1]. Furthermore, broadcasting can improve soil structure, increase nutrient availability, and promote soil biodiversity, all of which contribute to healthy plant growth [10]. Mechanical mastication is the process of changing fuels (shrubs and tree fuels) from their natural state into a compacted layer of highly fractured woody particles on the forest floor, thus considerably reducing the severity of an eventual wildfire [6,7,11]. Mulching offers the same benefits but accrues them even faster because it accelerates residue decomposition [12]. On top of that, mulching also achieves a faster reduction in fuel loads and a more effective removal of physical hindrances to replanting [11]. There are no general work quality references for mulching, therefore operators must default to ‘company standards’ [13], which may prescribe reduction into medium- or fine-sized fragments. Mulching into fine particles can only be achieved by multiple passes, which has a strong impact on costs. The cost of mulching should be weighed against its benefits, including better accessibility, increased soil moisture, reduced nutrient loss, lower fire risks, and optimum soil cover. In particular, a mulched residue layer represents an effective barrier against soil erosion [14,15].

In South Africa, residue management is being modernised through the adoption of mulchers, with the primary objective of removing the high risks associated with burning. Unsuitable burning conditions are exacerbated by extreme and unpredictable weather events (climate change) and are made increasingly expensive due to growing labour costs. Even though broadcasting may be a viable alternative for residue management, it is labour-intensive and retains the full residue (fuel) load. Very little information is available on the use of mulchers and on their effect on fuel load reduction and distribution. The goal of this study is to compare different options for managing harvesting residues, namely: broadcasting (manual); broadcasting and mulching with a tractor-based machine (semi-mechanised); and mulching with a purpose-built machine (fully mechanised). The comparison was based on the following key performance indicators: (1) time consumption; (2) cost; and (3) work quality (i.e., size reduction effect).

2. Materials and Methods

2.1. Study Site

The study was conducted in Tsitsikamma, in the Eastern Cape province of South Africa (coordinates: −33.972340, 23.692368), with an annual precipitation of 1003 mm. The compartment consists of dark-greyish sandy loam soils. The stand was a slash pine (Pinus elliottii Engelm.) plantation established at a 3.5 × 3.5 m spacing. The trees harvested had a DBH of 28.7 cm and a height of 17.97 m. The plantation was cleared at the age of 14 and yielded 1031 m³/ha. Harvesting was carried out according to the mechanised cut-to-length method, using a harvester and a forwarder. The terrain was flat, with a slope gradient ranging from 0% to 10%. Residues were left in piles, spread across the compartment. Three methods of residue management were used, as follows: manual method (broadcasting), semi-mechanised method (manual broadcasting followed by mulching with a forestry-fitted farm tractor), and fully mechanised method (purpose-built mulcher). The three methods are depicted in Figure 1.

Table 1. Machine carrier and attachment specification.

Specifications	Farm Tractor-Based Mulcher	Purpose-Built Mulcher
Carrier
Make and model	SAME Laser 110	Tigercat M726G
Engine power	74 kW (101 hp)	275 kW (370 hp)
Weight	8280 kg	14,880 kg
Ground clearance	450 mm	635 mm
Fuel tank capacity	320 L	570 L
Cutting head
Make and model	TMF200	Tigercat 4061-25
Type	Horizontal shaft	Horizontal shaft
Width	2441 mm	3000 mm
Mulching swath	2000 mm	2500 mm
Weight	1255 kg	4625 kg

gives the specifications of the carrier and the attachments used in this study.

2.2. Experimental Design

Three (3) treatments (manual broadcasting, semi-mechanised, and fully mechanised) were observed on the site and replicated four (4) times, each time on a plot measuring 30 m × 150 m. The resulting 12 plots were distributed randomly across the same compartment, their corners being marked with white painted pegs. Plot length and width were determined with a measuring wheel.

2.3. Residue Load Assessment

Prior to treatment application, the residue load on each plot was estimated by selecting two representative residue piles, manually separating branches and stem wood, and then weighing the two fractions on a portable scale. For this purpose, residues were cut into pieces with a chainsaw so that they would fit into a bucket. The resulting weight per pile was then multiplied by the number of piles on the respective plot. The moisture content was also determined according to the gravimetric method, which involves weighing fresh stem and branch wood samples obtained from the residue piles and determining the mass of water lost as a percentage of the total mass of wet samples. The samples were dried at 103 °C with a length and width of 25 × 500 mm, then weighed at intervals of 3 h until they reached a constant weight; this process was carried out over 12 to 48 h [16]. This was performed separately for the two fractions from each sample pile.

2.4. Time Consumption Assessment

Researchers used a stopwatch to determine the time taken to complete each plot. Furthermore, elemental time studies were conducted to determine the distribution of work time among different functional tasks. Manual broadcasting was performed by a team of five workers, and, therefore, the researcher conducted the elemental time study according to the activity sampling technique. The researcher recorded at one-minute intervals what activity was conducted by each worker at that moment, among the following four options: walk, broadcast, stand, and delay. This was carried out for both the manual and semi-mechanised methods.

Given that only one mechanical unit was working at one time (a farm tractor-based mulcher or a purpose-built mulcher), the elemental time study was applied as a classic cycle study. In that case, time was measured using a Trimble handheld computer running on the dedicated UMT Manager time study software. Therefore, the time studies of those machines split the work cycle into the following work elements:

• Move: Begins when the wheels of the machine start moving and the head is not engaged and stops when the wheels stop and the head is put on the ground to start engaging.
• Mulch: Begins when the mulching head is put on the ground and the mulcher starts moving and stops when the mulcher picks up the mulching head and starts turning.
• Turn: Begins when the mulcher picks up the head and starts turning and stops when the mulching head is put on the ground and the mulcher starts moving.

Delays were excluded from the record, and, therefore, all-time consumption represents productive work. For that reason, hourly rates refer to productive work hours.

The same machine operators were used throughout the study.

2.5. Work Quality Assessment

After mulching, work quality was sampled at six (6) points per study plot. Collection points consisted of 1 × 1 m squares located 20 m from each other and placed in a zig-zag pattern. All the fragments present on each square were collected, and their length was measured with a measuring tape. Fragments were then allocated to one of the following four length classes: <20 cm; >20–50 cm; >50–100 cm; >100 cm; these classes were within the sample limits of the material measured and were derived to simplify the categorization of samples collected from the 1 × 1 m sample areas. Biomass reduction for fully mechanised mulching was further assessed using sieves to obtain the particle size distribution of the medium and fine materials.

2.6. Cost Assessment

Costs were calculated in Rands (the South African currency represented by ZAR) using an industry-accepted silviculture costing model developed by the South African Contractor’s Association Costing (SAFCA). The assumptions regarding the machine purchase price and running costs (e.g., fuel and oil consumption and maintenance were obtained from the machine suppliers and contractors. The interest rate used was 9%, as it was the prime lending rate when the study was undertaken in July 2022 [17]. General labour rates were obtained from the current government gazette wage rates in the South African forestry industry [18].

Table 2. Main cost assumptions used in the study.

Machine	Farm Tractor-Based Mulcher	Purpose-Built Mulcher
Investment (ZAR)	1,513,750	8,384,548.80
Machine life (h)	10,000	20,000
Interest rate (%)	9	9
Residual value (%)	20	20
* Insurance and Licensing (%)	2	2
Fuel cost (ZAR/L)	21.34	21.34
Fuel consumption (L/h)	7	35
Oil cost (% fuel)	5	5
Repair cost value (% dep.)	100	80
Operator	1	1
Overheads (%)	15	15
Operator cost (R/h)	40	77

* Percentage of the annual capital cost (ZAR—rands, h—productive machine hours, incl. driver).

illustrates the main cost assumptions for the machines used in the study. The currency exchange rate at the time of the study was USD 1 = R16.6 (31 July 2022).

2.7. Data Analysis

Data analysis aimed at determining and comparing the operational productivity, costs, and work quality of the three residue management methods. Descriptive statistics were used to guide detailed analysis. Given the relatively small sample size and the violation of the normality assumption, comparisons were made with non-parametric tests. In particular, the significance of any differences between three or more groups was tested using the Kruskal–Wallis test, while differences between two groups were tested using the Mann–Whitney U test (SAS Institute Inc. 1999—Cary, NC, USA). Statistical significance was assumed for α < 5%.

3. Results

3.1. Time Consumption and Costs

Test conditions (surface area per plot, number of residue piles per hectare, biomass load per hectare, and percentage of stem wood) were not significantly different among treatments (

Table 3. Stand treatment characteristics.

Treatments	Broadcasting	Farm Tractor-Based Mulcher	Purpose-Built Mulcher
Hectares (ha)	0.417	0.455	0.400
No residue piles/plot	17	16	16
Residue load ODT/ha	14	12	14
% Stem wood	61	64	59

). The mean sample plot area for all treatments was 0.424 ha, and the mean number of pine residue piles per plot was 16. The estimated residue load ranged between 12 and 14 oven-dry tonnes per hectare (ODT/ha). The stem wood proportion averaged 61% of those amounts. The average stump volume for the plots was 8.16 m³/ha.

Broadcasting only (the manual method) required 16 h/ha, while broadcasting in preparation for tractor mulching (the semi-mechanised method) required 14 h/ha. Therefore, the cost incurred was quite similar: 400 ZAR/ha for broadcasting only and 350 ZAR/ha for broadcasting in preparation for the tractor. Those differences were small and devoid of any statistical significance.

When comparing the two mulching techniques (farm tractor-based vs. purpose-built), it was found that mulching took less than half as many machine hours when applied with purpose-built equipment (1.5 h, SD = 0.67) rather than with an adapted farm tractor (3.5 h, SD = 0.09). However, the latter was much more expensive to own and operate, and the cost per hectare was significantly different between the two techniques: 4083 ZAR/ha and 3267 ZAR/ha, respectively. However, the purpose-built machine could cope with piled residues, and, therefore, its use did not require prior manual broadcasting as in the case of the farm tractor-based mulcher. The latter was less powerful and was mounted on the rear of the tractor and, therefore, needed to drive over the residues, which could not be piled too high. Working with piled residues, the purpose-built mulcher had to perform a second pass to achieve good work quality. The second pass was significantly faster than the first one as it dealt with semi-processed material (

Table 4. Time consumption and cost for the two mulching techniques on the test.

Techniques	Farm Tractor-Based Mulcher	Purpose-Built Mulcher
Mulch h/ha	3.580	1.472
Cost ZAR/ha	3267	4083
Purpose-built only (1st pass vs. 2nd pass)
Mulch in 1st pass (h/ha)	-	0.903
Mulch in 2nd pass (h/ha)	-	0.569

Note: one pass is a single, one-way move along a specific mulch swath. “Two passes” refers to two moves along the same mulch swath.

).

Therefore, if the manual method (broadcasting only) is replaced by either the semi-mechanised or the fully mechanised method (broadcasting plus mulching or mulching only), residue management costs incur an 8–10-fold increase from 400 ZAR/ha to approximately 4100 ZAR/ha for the purpose-built mulcher and 3300 ZAR/ha for the tractor, respectively. Furthermore, it was found that there is a significant difference (p-value = 0.0209) between the cost incurred by the semi-mechanised methods (3267 ZAR/ha) and the fully mechanised methods (4083 ZAR/ha).

3.2. Work Quality

The fully mechanised method (i.e., mulching residues with a purpose-built machine) achieved the strongest size reduction: median particle size was 12, 31, and 39 cm for the fully mechanised, semi-mechanised, and manual methods, respectively.

A better appreciation of work quality is achieved by examining the particle size distribution graphs for the three treatments (Figure 2). With the fully mechanised method, over 70% of the fragments were shorter than 20 cm. In contrast, most fragments (49%) produced with semi-mechanised and manual methods were included within the 20 to 50 cm length class. Overall, the fully mechanised method obtained a much better size reduction compared with the semi-mechanised method. Indeed, the semi-mechanised method yielded half as many fragments longer than 1 m compared with the manual method, that is, 5% vs. 10% of the total. However, the fully mechanised method yielded none. Figure 3 illustrates the before and after biomass distribution by treatment on the plots.

Furthermore, samples taken from the fully mechanised method were passed through sieves with widths between >5 cm and >40 cm, and the results in

Table 5. Percentage passage of biomass on sieve sizes on residue piles and in between the piles for the fully mechanised system (purpose-built).

Locations of Sample	Sieve Size	Purpose Built (%)
Residue piles	>5 cm	27
Between the residue piles	>5 cm	31
Residue piles	>40 cm	12
Between the residue piles	>40 cm	9

show that more fine materials were produced when residues were mulched in the residue piles compared to in between the residue piles. This assessment was conducted to understand the fuel load distribution after mechanical mulching in areas with residue piles versus none. Due to the residues being long and not entirely broken down in the manual (broadcasting) and semi-mechanised (broadcasting with a tractor mulcher) processes, sieving was impossible. There was no significant difference (p-value = 0.7150) between the two methods of work quality assessment for the fully mechanised method; therefore, when using the fully mechanised mulcher, the work quality produced when mulching on areas with residue piles and no piles is relatively the same.

4. Discussion

This study was conducted to determine the productivity, cost, and work quality of three residue management methods. The purpose of residue management is to reduce and redistribute the fuel load for better accessibility and a reduction in fire risks [11]. According to [19], fuel treatments modify and affect species’ structure, composition, availability, distribution, fuel moisture, and surface wind behavior. Moreover, methods used in silviculture to alter fuel load and continuity decrease the treated area’s vulnerability to fires.

4.1. Fully Mechanised Method

The fully mechanised system, applied with a purpose-built machine, was the most productive, requiring 1.47 h/ha (two passes) and costing 4083 ZAR/ha. The second pass was necessary to achieve the desired work quality, according to company standards. When observing a similar wheeled mulcher (Tigercat M726G) working in broadcast pine pulpwood harvest residues [20], we found that the time consumption was 2.90 h/ha and the cost was 6165 ZAR/ha. The cost difference between the two studies amounts to 2069 ZAR/ha. Both studies took place in stands previously planted with pine species (Pinus elliottii vs. Pinus patula), characterised by flat terrain (0%–10% vs. 0%–12%); both sites were previously harvested according to the mechanised cut-to-length method, using a harvester and a forwarder. The main difference seems to be with slash distribution, which had been piled in the case of this study and left broadcast in the case of the study by [16]. This may lead to the conclusion that a powerful frontal mulcher works best with piled residues: concentration does not exceed the available power capacity, while it avoids idle manoeuvring and piecemeal work. An additional study of mulching in Ponderosa pine (Pinus lambertiana) stands may offer further corroboration to the general validity of our figures since it estimated time consumption at 1.9 h/ha, which is well within the range of this study [21]. Regarding mulching cost, most of the existing references available on the subject come from the United States, under a very different economic environment; those studies report costs most commonly included between 5500 and 7500 ZAR/ha, or between 300 and 400 USD/acre [22,23,24].

Slash quantities, size, composition, and proximity to soil affect the speed with which residue is mineralised. Mulching speeds up the rate of decomposition and offers a safer alternative to burning [25]. Furthermore, mulching guarantees a better soil cover, which mitigates soil temperature, avoiding extremes during both summer and winter [26]. Finally, mulched residues prevent excessive evaporation and maintain favourable soil moisture conditions, thereby promoting plant health and sustained growth.

Figure 2 shows the work quality of the fully mechanised method; most fragments were reduced to a length of 12 cm after two passes. That is a good result since fragment size falls between that of medium and fine chips. Furthermore, work quality for the fully mechanised method was assessed using sieves (Table 5); the results complement each other in terms of more materials falling within the class of medium-to-fine particles. Ref. [13] indicates that mulching at a speed ranging from 1.2 to 1.6 km/h results in a fuel bed with fuels (surface and aerial) processed into chipped debris. Ref. [27] supports this by mentioning that mulching as a treatment alters the fuel load covering the forest floor. Moreover, slash management properly carried out could contribute to forest ecosystem maintenance by limiting nutrient losses and allowing nutrient release to synchronise with stand requirements [25]. This indicates that the fully mechanised system is adequately capable of performing this task.

4.2. Semi-Mechanised Method

Results for the semi-mechanised method show that there is a slight change in residue form as broadcasting is combined with tractor mulching. The median fragment length reaches 31 cm after one pass of the tractor mulcher. A study conducted by [28] compared the productivity and cost of a harvester collecting and commuting apple pruning residues with traditional mulching. The working conditions included a row spacing of 3.2 m, and it was observed on an average area of 19 m² over three to five plots. Time consumption was found to be 2.43 h/ha and cost 2858 ZAR/ha. In the current study, the semi-mechanised method incurred a time consumption of 3.58 h/ha and costs of 3267 ZAR/ha. This is expected since the mulching was not observed in similar conditions (forest compartment vs. apple orchard). The residues mulched were different in terms of thickness, length, and arrangement; the forest stand presented a higher residue load and larger slash components compared to the apple orchard. Taking that into account, however, the cost is notably different (409 ZAR/ha). The working conditions of the tractor mulcher were improved considering the initial broadcasting that took place prior to mulching. The work quality of the semi-mechanised mulching had more residue particles after mulching in the class between 20 and 50 cm.

4.3. Manual Method

Manual broadcasting incurred the lowest cost (400 ZAR/ha) among the three methods observed in this study. It also achieved the poorest work quality, as residue sizes had been minimally reduced, and the prevalent fragment length was 39 cm, with a relatively large proportion of elements longer than 1 m, which could prove to be a significant hindrance to future soil preparation and planting work. In reference [29], the author mentioned that the extensive use of fire for harvesting residue disposal contributes to nutrient loss, erosion, and leaching. Therefore, broadcasting is more efficient than burning, both in financial and environmental terms: it requires less work, maintains soil cover, and minimises nutrient leaching [30]. Broadcasting is likely more efficient than windrowing because windrow residues are slower to mineralise and do not offer the same soil cover benefits [25].

From this study, it can be gathered that mechanised mulching is the most suitable method for the reduction in and distribution of slash to reduce the potential of a fire spreading rapidly and improve accessibility on a site. Moreover, productivity for fully mechanised mulching was not adversely influenced by the amount and arrangement of the residues. The number of residue piles and their arrangement in each plot did not seem to change the mulching consistency of the fully mechanised mulcher, so the quality was similar throughout the plot. Semi-mechanised mulching productivity was much lower due to limited manoeuvrability. With this method, the mulcher was carried on the back of a farm tractor, and, therefore, it was not possible to treat piled residues. Even if the tractor had attempted to mulch the residue piles while driving in reverse gear, the mulcher would not have been very successful as it had been designed for working with a tractor in the forward gear range, straddling the residues. Therefore, slash had to be broadcast first, and mulching required travelling the whole surface of the plot, not just driving to the piles. That was far less efficient, which, combined with the lower power of the tractor, limited the overall productivity. As a result, the residue management cost was less than the purpose-built mulcher with poor work quality, while for the purpose-built mulcher, the cost was higher, with excellent work quality.

5. Conclusions

The goal of this study was to determine the performance of three methods for dealing with harvest residues on pine stands. Fuel load redistribution was best achieved using the fully mechanised method (median particle size 12 cm), which offered the highest work quality in the shortest time (1.4 h/ha) but with the highest costs (ZAR 4083/ha). On the other hand, manual broadcasting alone (no mulching) was the cheapest method (ZAR 400/ha), but it achieved the lowest work quality and required the longest time (16 h/ha). Introducing a low-investment tractor mulcher after manual broadcasting resulted in a large increment in treatment cost of ZAR 3267/ha and a modest improvement in work quality. Therefore, this solution is best avoided. Forest managers are then left to decide between a cheap solution offering the lowest work quality (the manual method) and an expensive solution offering excellent work quality (the fully mechanised method); there is no in between. Work quality cannot be clearly defined in the study as it can be subjective.

Further studies should investigate the impact of work quality on the following management steps to determine how much of the additional cost incurred with mulching is recovered through more efficient replanting. Similarly, the same comparisons could be conducted on other plantation types, such as eucalyptus- and acacia-harvested stands. In the future, an alternative method for dealing with residues that could be considered is the collection and extraction of logging residuals. This option is currently not viable due to the high costs and lack of markets.