Materials and methods Description of the study area The study was carried out in the Rincine forest, a nature reserve 40 km northeast of Florence (central Italy). The mean annual precipitation in the period 2009-2013 was 1133 mm, with the maximum in November and the minimum in July. In the same period, the mean annual temperature and the means of the coldest and warmest months (January and July) were 9.2 °C, 1.5 °C and 17.8 °C, respectively. Soil developed on Lower Miocene-Oligocene sandstone and was classified as Dystric Cambisol based on the World Reference Base for Soil Resources (IUSS Working Group 2014). The study site was a 30-years-old, evenaged plantation of Picea abies (L.) H. Karst, Pseudotsuga menziesii (Mirb.) Franco, Abies alba (Mill.), and Chamaecyparis spp. located at 900 m a.s.l. on a 5-10% slope terrain. At the beginning of the experiment, the average duff thickness was measured on 60 points by a ruler with an accuracy of ± 0.5 mm. Particle size distribution was determined using the hydrometer method, as described in Gee & Bauder (1986), on five samples of the top 30 cm of mineral
soil, randomly distributed throughout the study area. Soil organic matter content was determined by loss on ignition (Heiri et al. 2001) on fifteen samples. Two tractors were used in our study, a tracked New Holland 88-85 (3908 cm3 displacement and 62.5 kW power) with an empty mass of 4600 kg and a wheeled New Holland T 4050 (4500 cm3 displacement and 71 kW power) with an empty mass of 4145 kg, equipped with power drive forestry (11.2-24.0) tires on the front axle and agroforest 410 (420/85-30) Trelleborg on the rear axle. Both types of tires had reinforced sidewalls and a cut-resistant rubber compound. Inflation pressures were 150 kPa and 100 kPa in the rear and front tires, respectively. To measure the contact area between the tire and the ground, we pulled a rope tightly around the portion of the tire on the ground, assuming a circular contact patch (Neri et al. 2007). The average contact pressure was determined to be 46 kPa and 54 kPa in the front and rear axles, respectively; they were assumed to be uniformly distributed over every point of the four contact areas. Measurements of soil compaction and rutting were performed immediately before and after tractor passes. Since the aim was to investigate the effects of tracked and wheeled machines on soil disturbance, tractor passes were conducted without loads in order to avoid the impacts of the skidded logs on soil. Experimental design and measurements A split-plot design was applied to verify the impact of tractors on soil. Three 50 × 50 m areas were divided into four 25 × 25 m plots. Two skid trails were marked within each plot, resulting in a total of 24 trails per tractor (3 × 4 × 2 = 24). In each plot, one