We worked with Populus x euramericana 'I-214' -referred to as test material (TM)- from a 20-year-old commercial plantation located in Quintanilla de Sollamas (42° 36' 00"N - 05° 49' 00" W), Spanish community of Castile-Leon. The TM was received as 50 mm x 150 mm x 3000 mm beams which were conditioned in laboratory to achieve equilibrium moisture content at 12 + 2% (determined by digital hygrometer, Testo 606-1).

Material sampling and selection was carried out following EN 350:2016 for commercial sawn timber. In accordance with the minimum recommended ra­tio by the standard, out of the total of 100 beams, 40 were selected from different positions in the lot and reprocessed to obtain samples free from defects which were then sized according to the corresponding technical specifications. Samples were extracted from the inner section (material closest to the heartwood -identified by growth rings orientation ¹ ) and outer section (material farthest from the heartwood) of each of the beams. Parameter followed for sample sectioning given that the selected beams of this commercial timber showed no evidence of duraminization and heartwood (or could not be observed/identified).

For basidiomycete fungi (brown and white rot) and termites, we used 15 x 25 x 50mm samples (standards CEN/TS 15083-1:2005 and EN-117:2012 , respectively), and for soft rot fungi, 5 x 10 x 100mm samples ( CEN/TS 15083-2:2005 ). For each section under study and type of decay, 40 repeats were carried out (320 repeats for TM).

As reference material (RM) we used Fagus sylvatica L. (beech) and Pinus sylvestris L. (pine), with similar characteristics to the TM, for fungi and termites, respectively. A total of 30 repeats for each type of decay were carried out (120 repeats for RM).

Basidiomycete fungi

Poplar resistance to xylophagous basidiomycete fungi was determined according to CEN/TS 15083-1:2005 . The fungal species used were: Coriolus versicolor (L.) Quel. [= Trametes versicolor (L.) Lloyd], strain CTB 863 A, responsible for white rot, and Coniophora puteana (Shumach.) P. Karst., strain BAM Ebw.15, responsible for brown rot. Each strain was axenically cultivated in malt agar medium (1.5% agar, 3% malt) in 400 ml culture vessels and incubated at 22 ± 2ºC and 70 ± 5% relative humidity (RH). Once the culture medium was covered by fungal tissue, the TM and RM samples were introduced into the culture vessels (2 TM or 2 RM x vessels) after determining their initial dry mass (m _i ) and having sterilized them. After 16 weeks under controlled temperature and humidity (22 ± 2ºC and 70 ± 5% RH) in the culture chamber the RM samples were withdrawn from the vessels, conditioned (superficial mycelium extraction), and oven-dried at 103 ± 2°C to determine final dry mass (m _f ). Subsequently, degradation was determined by loss in mass as a percentage of the initial dry mass (ML %), as calculated from Equation 1. Once a ML higher than 20% (established test reliability) was detected on this material, the same procedure was carried out for the TM samples (identified by sections). The criteria for determining durability classes (DC), according to ML (%), are presented in Table 1. Use class was estimated according to standard EN 335:2013 .

Where ML: mass loss, in %; m _i : initial dry mass, in g; m _f : final dry mass, in g.

In this case, CEN/TS 15083-2:2005 was followed and the microbiota of the substrate was used to inoculate the test medium. A mixture of natural soil - fertile farm soil (20:80 ratio) with water holding capacity (WHC) ranging from 25 to 60% was used as substrate. The substrate was placed in 25 ml containers, enough to cover approximately 120 mm of the depth. The sam­ples were dried in the oven and weighed to obtain their initial dry mass (m _i ) before placing them in each container (20 TM samples identified per section and 10 RM samples), separated 20 mm from each other, and buried for 20% of their length. The containers were weighed and placed in a chamber under controlled conditions at 27 ± 2°C and 70 ± 5% RH for a period of 16 weeks. The initial reference weight (indirect indicator of system dehydration throughout the experiment) was recorded. In order to maintain the right water-hol­ding capacity, each container was weighed weekly to keep the right humidity of the substrate. After the incubation period, 10 RM specimens were extracted at random in order to assess the validity of the test via ML% > 20% (Equation 1). Once the test was validated, the same procedure was repeated for the TM samples. Durability class was calculated as the quotient x (Equation 2; Table 1: EN 350:2016 ) and use class for this material was determined based on this value ( EN 335:2013 ).

The durability of wood exposed to subterranean ter­mites was assessed according to standard EN 117:2012 using the species Reticulitermes grassei (Clément) as biological material. We used colonies consisting of 250 workers (those in molting stage -with whiter abdomens-, injured or immobile re­mained behind) and several nymphs and soldiers in a ratio of 1 to 5%, with reference to the proportion found in the colony the workers were extracted from.

Each colony was placed in 400 ml glass containers with a moist sand substrate (in proportion: 1 part of water to 8-10 parts of sand) and 0.5 g of farmed timber ( Pinus sp. wood shavings) -initial food source of the colonies- evenly scattered on the approximately 40-60 mm thick sand layer. Subsequently, against one of the sides of each container (half-way up), a glass cylinder of approximately 1 mm wall thickness, height of 20 mm and diameter of 20 mm was introduced so that part of it extended out of the surface of the substrate. The containers were incubated at 28ºC and 80% RH in the culture chamber for a period of 2-4 days, until the colony was uniform and lively.

In each container, and on the glass cylinder, a wooden specimen (TM or RM) was placed. Containers were then incubated in the culture chamber for 8 weeks, periodically checking the conditions of humidity of the substrate.

Once this period elapsed, all the specimens were removed, and after cleaning, a final examination was carried out. The number of living termites was counted (workers, nymphs and soldiers) and the attack ratings were visually analyzed (EN 117:2012): (0) no attack; (1) attempted attack, (2) light attack, (3) medium attack, and (4) heavy attack. The Survival Rate -SR- (%) was estimated using the ratio of the initial number of workers to the number of survivors.

The validity of the test was confirmed in 10 RM (pine) specimens through visual evaluation (attack level 4) and surviving workers count (SR: 50%). Even if it was not prescribed by the standard, ML (%) of the specimens was also assessed (Equation 1). The criteria for determining the resistance to deterioration caused by termites, durability classes (DC), according to attack ratings, are presented in Table 1 ( EN 350:2016 ). Use class was estimated according to standard EN 335:2013 .

Anatomical studies were carried out on TM (inner section) exposed to all xylophagous agents. We selec­ted the material that showed mass loss percentages representative of the mean value obtained per type of test. The anatomical studies were carried out with an Olympus (CX31) Optical Microscope (OM) with a Lumenera's Infinity1 digital camera and a FEI (Quanta 200) Scanning Electron Microscope (SEM). For OM tests, transverse and longitudinal tissue sections (TS and LS) of the samples were stained with safranin at 1% and mounted in water temporarily. Unstained sections were observed under polarized light to detect loss of birefringence (indirect indicator of cellulolysis; diagnostic feature of brown rot). For observations with SEM, tissue sections without pretreatment were mounted on metal stubs using double-sided graphite tape and finally gold plated. Material characterization was carried out by reference to Anagnost (1998) and Schwarze (2007) .

All statistical analysis was carried out with Software R©, 3.4.4 version ( R Development Core Team, 2018 ). A total of 320 data values corresponding to each group (timber species, section under study and decay agents) were analyzed. Firstly, for all groups, statistical assumptions of independence, normality (Shapiro-Wilks test) and homoscedasticity (Bartlett test) were assessed. When the assumptions of normality and/or homoscedasticity were not achieved, the comparative analysis (ANOVA) common in classical linear statistics could not be used, then robust methods were applied to compare groups (Welch test). Bootstrap method was used to find mean and median confidence intervals and homogenous groups.

The results of macroscopic and microscopic eva­luation of wood durability against basidiomycetes, soft rot fungi xylophagous and termites are shown in Table 2 and in Figures 1 to 4.

According to the results (Table 2), both inner and outer sections of poplar wood proved to be "Not Durable" (in according to the robust median CI value ≥ 30% ML) to the xylophagous fungi employed, with statistically different, robust mean ML (%), between basidiomycete strains for each section under study (in uppercase) but without any significative difference in robust mean ML (%) between sections for both basidiomycetes and soft rot agents (in lowercase). Structural alterations, both macro and microscopic, were compatible with a low resistance to deterioration and were typical of each rot type ( Schwarse, 2007 ; Anagnost, 1998 ).

Material exposed to white rot ( T. versicolor ) had characteristic fibrillar aspect and spongy consistency and was whitish in colour. OM observation revealed wall thinning in transverse tissue sections (TS) and consequent wall deformation in vessels, rays and fibers. In some areas, wall thinning affected the integrity of the middle lamella causing cell separation (Fig. 1A). Consequently, longitudinal section samples (LS) sho­wed fibrillar aspect. SEM observation confirmed said fibrillar appearance and also revealed holes (Fig. 1B), erosion troughs (Figs. 1B y 1C) and pitting (Figs. 1B, 1D, 1E) in vessels, fibers and radial parenchyma. Vessels with abundant mycelium and evidence of fungal colonization in pitting (Fig. 1E) were also revealed.

Material exposed to brown rot ( C. puteana ) sho­wed reddish-brown colouring, cubic fracture pattern and fragile consistency (it disintegrated into dust under pressure). OM observation showed strong contraction as well as fractures in vessels (Fig. 2F), rays and fibers. There was evidence of generalized loss of birefringence (Fig. 2G). Under SEM, the features described above -deformation and fractures- were observed (Figs. 2H y 2I), mycelium in vessels and fibers (less abundant than that observed in wood exposed to white rot fungi) and fungal colonization through the pitting in vessels.

Material exposed to soft rot showed spongy con­sistency and light brown colouring. OM observation of TS tissue revealed the presence of cavities in the fiber walls (Fig. 3J), which, in LS, were found in vessels as well. In both cases, parallel to the microfibril angle (Fig. 3K). Under SEM, and in LS samples, small holes (bore holes) were observed in vessels, most under 1µm (Fig. 3L). We also observed fibers and vessels colonized by mycelium with erosion troughs of serrated edges (Figs. 3M and 3N).

With reference to durability against attack by termites, results show that poplar is "Not durable" ( EN 350:2016 ) in both sections under study (inner-outer), in accordance with what the standard EN 350:2016 describes for other species of the genus Populus . Survival rate and mass loss in both sections (inner and outer) confirmed results per attack level (only parameter for damage estimation considered by the standard).

Under SEM (Fig. 4), mechanical damage resulting from insect feeding was especially verified in fibers. This type of deterioration reduced the section of the specimens and, as a result, their quality and resistance to touch.