## Introduction

The State Timber Corporation (STC) of Sri Lanka has classified valuable timber species such as Tectona grandis (Teak), Swietenia macrophylla (Mahogany), Artocarpus heterophyllus (Jack) into super-luxurious and luxurious classes1. Because of the high demand, rareness, and prohibitions of the transportation of logs of the luxurious timber species, they are highly vulnerable for adulterations. Although the timber adulteration is a well-known issue in the country, no survey has ever conducted to document the intensity of the problem.

The most common method of timber adulteration involves substituting original timber with lower grade wood sources2. The common adulterant timber species are Persea americana (Avocado), Mangifera indica (Mango), Magnolia champaca (Vern. Sapu), Samanea saman (Vern. Mara) and Toona ciliata (Vern. Toona), and many others which are abundantly grown in Sri Lanka. The timber of the adulterant species is applied with various dyes to conjure the characteristic color patterns. The products of adulterated wood will not be able to address consumer demand. The timber adulterations result in the declination of the sales of the timber products and the violation of consumer rights. Moreover, diminishing timber sales would negatively affect the Sri Lankan economy. According to the Central Bank of Sri Lanka, wood and wood products account for the total value of USD nine million in 20133. Hence authentication of timber products, which involves the accurate identification of timber species, is vital to gain the end-user trust and sustain the timber industry. However, there is no precise tool available in Sri Lanka for tracking timber adulterations.

The most commonly practiced method of timber identification is the observation of wood appearance, density, and cell anatomy. The timber inspectors use physical features such as color and grain patterns, hardness, weight, odor, luster, and texture to identify the species of a particular timber source4. The anatomical features such as vessel diameter, fiber length, fiber diameter, ray height, number of rays, and ray width can be measured to distinguish timber species5,6. However, inferences made using anatomical features are not highly accurate as morphological parameters can be varied due to the age of the tree and the environmental conditions such as water availability, nutrient content of the soil, wind, and temperature7. Furthermore, related timber species may contain similar morphological features, obscuring the discrimination8. The microscopic assessment of the timber anatomy is also time-consuming, and expensive9,10. The microscopic assessment of 10 timber samples requires two-days of working on replicates which costs approximately USD 275/=. The DNA barcoding of two loci of the 10 timber samples requires only a day of working which costs USD 150/= (Pers. Com).

The DNA barcoding can be used in place of physical methods for detecting timber adulterations9,10 serving as a rapid, accurate and automatable technique and the best way for species identification11. The plant DNA barcoding is done based on specific loci in the plastid, and nuclear genomes12. For examples, chloroplast genes matK-trnT and atpB-rbcL have been reported as the robust markers for the identification of land plants13,14. Because of the improvement of molecular biology and usage of automated machines that reduces the labor and the requirement of skills, DNA barcoding has become inexpensive15, making it more suitable for rapid identification of species. Although DNA barcoding can provide unique identity and implement in many fields in an inexpensive manner, it is not famous as a tracking method of timber adulterations. However, DNA barcoding is proven capable of detecting timber adulterations as reported for sandalwood15 and monitor illegal logging of endangered species16. However, the challenging step is the extraction of good quality DNA from a small amount of dried or processed wood. In many cases, the samples have to be obtained without harming the timber product. As wood is dead plant material, it only contains trace amounts of mostly degraded DNA17. This sort of DNA is sufficient as the template for PCR; however, the DNA quality could be low due to the presence of PCR inhibitors such as polyphenols, tannins, and resins18,19. However, improved methods to extract DNA from the plant tissues such as wood and seeds are currently available20.

In the present study, we aimed to report the status of timber adulterations in Sri Lanka using a survey, detect the typical scenarios of adulterations and reveal the inability of morphometric methods in detecting adulterations. We also targeted to demonstrate that DNA barcoding could accurately identify the timber adulterations.

## Materials and Methods

### Questionnaire guided survey

We formulated four questionnaires to collect information from patrons (i.e., customers), manufacturers, carpenters, and regulators to gather knowledge about the timber adulterations in Sri Lanka. We obtained informed consent from the participants to use the data only for the present research project. The ethical clearance for the survey was obtained from the Ethics Committee of the Postgraduate Institute of Science, University of Peradeniya, Sri Lanka. Table 1 summarizes the demographic variables and specific questions included in the four questionnaires. We collected data by interviewing randomly selected 40 patrons, 30 manufacturers/traders, 30 carpenters, and 15 regulators. The adulteration scenarios (i.e., species combinations) were selected from the survey information for morphometric and DNA analyses. All the methods employed in this study performed in compliance with the relevant guidelines and regulations imposed by the Ethics Committee of the Postgraduate Institute of Science, University of Peradeniya, Sri Lanka.

### Sample collection

We collected leaf and wood samples from eight species, T. grandis, S. macrophylla, A. heterophyllus, P. americana, M. indica, M. champaca, S. saman, and T. ciliata. The leaf samples were stored at −80 °C and wood samples at room temperature. We also photographed the representative trees, cross sections of the matured logs, twigs and leaves of all the species for the illustration purposes.

### Creation of adulteration scenarios under workshop setting

We smoothened the collected wood samples using sand papers to obtain furniture quality wood surfaces and photographed wood sections before polishing. A professional carpenter was hired to stain and polish the wood sections. A set of horizontally and vertically cut wood pieces were polished using a commercially available formula of natural color enhancer (i.e., top-coat or polish) without any staining for the comparison purposes. We used the commercially available T. grandis colorant to dye S. saman and M. champaca wood to create adulteration scenario of T. grandis. Furthermore, we used the commercially available colorant of A. heterophyllus to obtain the adulteration scenario of A. heterophyllus by dyeing P. americana and M. indica wood. Similarly, the adulteration scenario of S. macrophylla was created by coloring P. americana, M. champaca, and T. ciliata wood with the commercially available colorant of S. macrophylla. The top-coat was applied to all the stained wood surfaces. Finally, we compared the non-polished, polished, and stained and polished sections of each adulteration scenarios by taking photographs.

### Morphometric data collection

We obtained wood density measurements by cutting the wood samples into square shaped pieces and recording the length (m), height (m) and width (m) and weight (kg). We calculated the density of each wood sample using the following equation.

$${\rm{Density}}\,({\rm{kg}}\,{{\rm{m}}}^{-3})={\rm{Weight}}\,({\rm{kg}})/{\rm{Volume}}\,({{\rm{m}}}^{3})$$

Simultaneously, to compare the microscopic anatomy of wood; we obtained the xylem vessel and fiber diameter measurements. We cut very thin (10–15 µm) cross sections from each wood piece using a microtome. The sections were then kept in distilled water for five mins and transferred through a series of ethanol (25%, 50%, and 70%). Then we stained the sections with safranin for one min. The stained section was mounted on a glass slide with 50% glycerin, topped with a coverslip, and sealed with a colorless topcoat. We observed the prepared slides using a light microscope (Carl Zeiss Microscopy GmbH, SN 3150000610) and captured the microscopic images at low, medium and high powers using AxioCam ICc 5 camera and Zen lite 2.1 software (Carl Zeiss Microscopy GmbH). The xylem vessel and fiber diameters were recorded from 15 different xylem vessels and 30 different fibers using a calibrated eyepiece graticule. The vertical diameter and horizontal diameter were measured for each cell and the mean diameters were calculated for analysis.

### DNA based detection of timber adulterations

#### DNA extraction from leaves

We isolated the genomic DNA from leaves using a modified CTAB plant DNA extraction protocol according to the instructions given in Porebski et al.21 and stored the extracted DNA at −20 °C.

### DNA extraction from wood

We collected approximately 30 mg of wood dust from each timber sample into an eppendorf tube and mixed with 600 µL of extraction buffer (8 g of CTAB, 10 mL of 2 M Tris-HCl [pH 8.0], 8 mL 0.5 M EDTA, 16.36 g of NaCl and 2 g of PVP in 200 mL of extraction buffer). We incubated the tubes at 60 °C in a water bath for 30 mins with periodic mixing by gently inverting the tubes in every ten mins. After that, we added 600 µL of chloroform/isoamyl alcohol (24:1) and mixed by inverting the tubes several times. Then we centrifuged the tubes at 12,000 rpm for ten mins and transferred 500 µL of the supernatant into a sterile Eppendorf tube, and added with 300 µL of isopropanol and 50 µL of 3 M Na-acetate. The contents were mixed by inverting the tubes several times and then incubated on ice for 30 mins. We spun the mixture at 12,000 rpm for ten mins and discarded the supernatant. The DNA pellet was washed by adding 70% ethanol and inverting the tube a few times. Finally, we air dried the pellet entirely and then added 30 µL of TE + RNase (10 mg RNase in 1 mL of TE buffer). We incubated the mixture at 60 °C to dissolve DNA22. The quality and quantity of the extracted DNA were assessed using the absorbance ratio of 260/280 and 1% agarose gel electrophoresis.

### PCR amplification and DNA sequencing

We selected the DNA barcoding markers matK-trnT (5′-3′GCATAAATATAYTCCYGAAARATAAGTGG/TGGGTTGCTAACTCAATGG) and atpB-rbcL (5′-3′GAAGTAGTAGGATTGATTCTC/TACAGTTGTCCATGTACCAG) for the present study considering the availability of comparison sequences in public domain. The DNA samples extracted from each species were amplified using the two markers. We carried out PCR in a thermal cycler (Takara, Otsu Shiga, Japan). The PCR of 15 µL volume contained 1 × GoTaq Green Master Mix (Promega Corporation, Madison, Wisconsin, USA), 0.3 pmol of reverse and forward primers and 50 ng of template DNA. The PCR cycle consisted of the initial denaturation at 95 °C for 1.5 min, followed by 35 cycles of denaturation at 95 °C for 30 secs, annealing at 48 °C for 1 min, initial extension at 68 °C for 2 min and a final extension at 68 °C for 20 min for matK-trnT primers23 and for atpB-rbcL primer pair with initial denaturation at 94 °C for 4 min followed by 40 cycles of denaturation at 94 °C for 30 sec, annealing of primers at 45 °C for 30 sec, initial extension at 72 °C for 2 min and the final extension at 72 °C for 5 min24. We size-separated the amplified products in 2.5% agarose gel electrophoresis. We purified the PCR products using QIAquick PCR Purification Kit (Catalog No: 28104, Qiagen, Hilden, Germany) and cycle sequenced (3×) using the Genetic Analyzer ABI 3500 (Applied Bio Systems®).

### Data analysis

We summarized the responses gathered from patrons, manufacturers, carpenters, and regulators. We used the cross-tab procedure in the Statistical Package SPSS 16, SPSS Statistics Software, IBM ® SPSS Predictive Analytics (IBM.com) to assess the associations among variables. We calculated the percentages of respondents considering all patrons, manufacturers, carpenters, and regulators together to identify the adulterant species to the standard timber species. Based on these details, we selected the adulteration scenarios for the morphometric and DNA barcoding assessments. We subjected the wood density, xylem vessel diameter, and xylem fiber diameter data to General Linear Model (GLM) procedure in the Statistical Package SAS 9.4 (SAS Institute, Cary, NC, USA). We examined the matk-trnT and atpB-rbcL sequence data for matching lengths with bands detected in the agarose gel electrophoresis for PCR fragments generated by DNA samples extracted from leaves and wood. We aligned the forward and reverse sequences of each species using CLUSTALW alignment tool embedded in sequence alignment software MEGA 7.025 to obtain the consensus sequences. We performed BLAST (https://www.ncbi.nlm.nih.gov/BLAST/) searches using the generated DNA sequences to confirm the species identity. We submitted the generated sequences to NCBI- GeneBank. During the submission process, we trimmed the sequence-edges according to the obligatory GenBank requirements (https://www.ncbi.nlm.nih.gov). Finally, we created the DNA barcodes using the barcode generating tool available at Bio-Rad (http://biorad-ads.com/DNABarcodeWeb/).

## Results

### Adulterations and relevant factors in the timber market

Table 2 displays the demographic variability of the respondents. The majority of the respondents were from the age group of 40–60 years. All patrons, manufacturers, carpenters, and regulators were mostly well experienced in their roles of the timber industry. All most all patrons looked for the species when buying timber products mainly because of the strength and durability. A total of 70.0% of the manufacturers and 86.7% of the carpenters said that coloring of low-quality timber with available dyes in local timber shops to have the look of intended luxurious timber is the most frequent form of adulteration. We got to know from 73.3% of the carpenters that the manufacturers always instructed the carpenters to adulterate timber for better profits.

### Perceptions of the patrons

We identified two groups of patrons according to the purchasing interests; 72.5% of them were interested in buying multiple timber products (furniture, rafters, ceiling boards, door/window frames, and other timber-based items) and 27.5% were interested only in purchasing furniture. A total of 97.5% of the patrons paid attention to the timber species when buying the products (Table 2). Accordingly, 82.5% of the patrons mentioned T. grandis as their preferred species. The 62.5% and 40.0% of the patrons preferred A. heterophyllus and S. macrophylla respectively (Table 3). Out of the patrons, 92.5% preferred T. grandis, A. heterophyllus and S. macrophylla due to their strength and durability; whereas 97.5% indicated that quality standards of the timber products are prerequisites making the purchasing decisions (Table 2).

Moreover, 95.0% of the patrons looked for high-quality original timber. However, in a separate question, 10.0% of the patrons said that they might look for the products made out of cheap timber having the look of original timber (Table 3). The patrons mentioned color and grain patterns of the wood, weight-felt, and hardness as the parameters to check the timber quality. When questioned about the timber adulterations, astonishingly 92.5% of the patrons said that they know that timber adulterations are going on in the Sri Lankan timber market. However, 47.5% of the patrons said that they could not identify the timber adulterations in the finished products. In summary, timber adulteration is a serious concern according to the opinion of the patrons as 90.0% of them said it is a violation of consumer rights. Also, 95.0% of the patrons said that if an accurate method is available to detect the adulterations, it would build up the confidence and trust, and timber industry would get uplifted from its current status (Table 3).

The association analysis between the parameters assessed with respect to the opinions of the patrons revealed that the preference to purchase T. grandis products was significantly associated with the class of profession of the patrons. The choice for high-quality timber and the strength and durability of the timber product were also significantly associated. The opinion on the adulteration was significantly associated with the patrons’ concern on timber quality, awareness on the occurrence of timber adulterations and the preference for strength and durability of the timber products (P < 0.05) (Table 4).

### Perceptions of the manufacturers

Out of the manufacturers interviewed, 96.7% of them said that T. grandis is the most preferred and frequently used timber species. The 50.0% and 16.7% of the manufacturers also preferred S. macrophylla and A. heterophyllus respectively. The high quality of the timber and low price with the original look were essential when purchasing as revealed by 50.0% and 46.5% manufacturers respectively. Only 77.3% of the manufacturers expressed that they aware about the timber adulteration is going on in Sri Lanka and the respondents also independently agreed with the opinion of patrons stating that it is a violation of consumer rights and an accurate system for timber adulteration is necessary (Table 3).

According to the opinion of the manufacturers, the association between patrons’ inquiries about the timber quality and the preference for T. grandis was significant. The association between the experience of the manufacturers in the timber industry and methods used in adulterating procedure was also significant. The manufacturer’s perception on the preference of the patrons and the degree of experience of the manufacturers in timber industry were also significantly associated (P < 0.05) (Table 4).

### Perceptions of the carpenters

For carpenters also, T. grandis was the most popular timber species as revealed by 90.0% of them. Also, 83.3% and 86.7% of the carpenters respectively said that S. macrophylla and A. heterophyllus are popular timber species. A total of 86.7% of the carpenters said that they use color and grain patterns to check the quality of the timber and adulterants commonly used in Sri Lanka. Moreover, 63.3% of the carpenters said once adulterated; the products were finished by coloring and polishing; they cannot be differentiated from the non-adulterated products. The 73.3% of the carpenters also agreed that the timber adulterations violate the consumer rights and 86.7% said that a system is required to detect the timber adulterations to uplift the timber industry (Table 3). Interestingly, according to the opinion of the carpenters, the degree of the timber adulterations in the market and the patron preference/search for the high-quality final product were significantly associated (Table 4).

### Perceptions of the regulators

Two third of the regulators were aged between 40–60 years, and 60.0% of them had more than 10-years of experience in the timber trade industry. The regulators must recognize the type of timber during the transportation of logs, timber, and furniture or during any inspection event. All the regulators said that they use color and grain patterns for differentiation of the timber species. Moreover, 40.0% and 26.7% of the regulators employed the hardness plus texture of grains and the density respectively for the identification of the timber species. A total of 66.7% of the regulators mentioned that only partial identification accuracy is possible when inspecting adulterated timber products. Only, 26.7% of the regulators claimed that they got skills to identify the species origin of the timber samples (Table 2).

According to the opinion of regulators, the majority (86.7%) stated that timber adulteration violates consumer rights and it has to be stopped. The 73.3% of the regulators claimed that timber adulteration would lead to less durability of the product. The 46.7% and 40.0% of the regulators said that the timber adulterations would lead to less attractiveness of timber products and reduction of the demand respectively. Two third of the regulators also felt the essential need of a novel tracking method for timber identification. Totals of 66.7% and 33.3% of the regulators expressed that a novel approach of detecting timber adulteration would help to identify illegal timber and safeguard the consumer rights respectively.

Out of the parameters assessed, the experience of the regulators was significantly associated with their ability to identify adulterations and the timber species in a finished product. Also, the opinion of the regulators on timber adulteration and their perceived understanding of the durability of the timber product were significantly associated (P < 0.05) (Table 4).

### Standard and adulterant timber species

The survey respondents expressed the names of the standard species and the adulterant species according to their experience. The respondents nominated many standard and adulterated species. During the survey, the respondents had recognized eight standard high-quality timber species in Sri Lanka currently undergone adulterations. We labeled the combinations of standard species and its nominated adulterations as an adulteration scenario (AS). There is a total of eight AS identified. In AS1, T. grandis is adulterated with 12 species in which M. champaca (25.4% respondents said), S. saman (14.9%) and ten other species. Similarly, in AS2, A. heterophyllus is adulterated with P. americana (16.5%) and M. indica (13.4%) and eight more species. In AS3, S. macrophylla is adulterated with P. americana (16.4%), M. champaca (10.4%) and T. ciliata (10.4%) and five others. The full list of standard species and their adulterant species are given with the percentage respondents in Table 5.

Based on the quantity of usage in the timber industry of Sri Lanka, we selected of T. grandis (AS1), A. heterophyllus (AS2) and S. macrophylla (AS3) and their two major adulterants for the morphometric and DNA barcoding analyses. Because of the two matching numbers for the second adulterant species in S. macrophylla of AS3, we had to select another species for further assessment with morphometry and DNA barcoding (Table 5). The representative images of the adaxial and abaxial surfaces of the leaves, branches and fully grown trees are given in Fig. S1.

Figure 1 displays the cross sections of the matured logs with prominent secondary growth for the species selected for AS1, AS2, and AS3. A set of individual photos for the cross sections of the matured logs are given in Figs. S2–S9 for clear illustration purpose. Figure 2 summarizes the processed timber pieces (cut and smoothened with sandpapers) of the original wood of all the species, polished pieces of the three standard pieces and dyed and polished adulterant species. Accordingly, in their original forms, logs and wood sections displayed visual differences, especially in their colors. However, after application of respective dyes/stains and polishing to give them a finishing, it was evident that we could not distinguish species of the same AS by visual differences. Also, strikingly, the wood sections of P. americana and M. champaca which naturally got light colors could be successfully stained with more than one type of dyes (Figs. 1 and 2). Thus it can be shown that even using one adulterant species; two timber species could be mimicked easily. A patron with a general idea of timber would never be able to differentiate an adulterated timber.

#### Variation of the wood density

Table 6 shows the mean wood density of the original species and the adulterants. In AS1, although S. saman got a low mean wood density (443.67 kg/m3), T. grandis and M. champaca could not be discriminated as they got significantly similar mean density values. Similarly in AS2, although P. americana got the significantly least mean wood density (180.10 kg/m3), A. heterophyllus and M. indica were not significantly different. The same situation was observed in AS3 where P. americana got the significantly least mean wood density, however, all the other three species got significantly similar means (P < 0.05).

#### Variation of the xylem vessel and fiber diameters

Table 6 also shows the mean xylem vessel diameter (XVD) of the original species and the adulterants. In AS1, although S. saman got significantly higher XVD (0.210 mm), T. grandis and M. champaca could not be discriminated as they got significantly similar mean XVD values. In AS2, P. americana got the significantly least mean XVD (0.106 mm) while A. heterophyllus got the mean XVD of 0.180 mm which was significantly less than the XVD of M. indica (0.277 mm). The same situation was observed in AS3 where P. americana got the significantly least mean wood density, however, all the other three species got significantly similar mean XVDs (P < 0.05). In AS3, the mean XVDs of S. macrophylla and T. ciliata were significantly similar and collectively higher than those of P. americana and M. champaca (P < 0.05).

Moreover, Table 6 depicts xylem fiber diameter (XFD) measurements which support the failure to distinguish species in AS1 as suggested by XVD measurements. Accordingly, T. grandis and M. champaca could not be distinguished from one another as they possess the same mean XFD value of 0.006 mm. The vessels of T. grandis and M. champaca were arranged in a solitary manner, while vessels of A. heterophyllus, P. americana, M. indica and S. macrophylla were partly solitary, where both single and multiple vessels were present. T. ciliata wood contained deviated microscopic characters from above two categories where vessels were present solitarily in low numbers and contained a high amount of fibers. Contrastingly, S. saman had a special arrangement where vessels were clustered together (Fig. S10).

#### Polymorphism of PCR bands and DNA barcodes

Figures 3A and S11 depict the length polymorphism of the matK-trnT locus for leaf and wood DNA arranged into three adulteration scenarios (AS1, AS2, and AS3) respectively. In AS1, T. grandis had a band of ~800 bp in length while its adulterants S. saman and M. champaca had bands of ~850 bp in length. In AS2, A. heterophyllus provided a band of ~925 bp in length which ran behind the two bands provided by its adulterants P. americana (~820 bp) and M. indica (~890 bp). S. macrophylla of AS3 amplified a region of ~900 bp in length whose band was longer than that of its adulterants P. americana, M. champaca and T. ciliata (~890 bp).

Figures 3B and S12 show the length polymorphism of the atpB-rbcL locus for leaf and wood DNA categorized into AS1, AS2, and AS3 respectively. Similar to that of matK-trnT, we detected length polymorphisms in atpB-rbcL in all three adulteration scenario classes revealing the applicability of using atpB-rbcL locus to differentiate adulterants from their original timber species. Accordingly, in AS1, S. saman and M. champaca had shorter bands of ~930 bp in length compared to that of T. grandis (~1000 bp). In AS2, the longest-amplified band of ~1050 bp corresponded to A. heterophyllus while its two adulterants P. americana and M. indica provided bands of ~920 bp and ~950 bp respectively. In AS3, S. macrophylla amplified a region of ~950 bp in length whose band was longer than those of its adulterants P. americana, M. champaca and shorter than that of T. ciliata (~960 bp) (Figs. 3, Figs.S11 and S12). The DNA barcodes also provided a distinct species specific variation with length polymorphism enabling the use of DNA based detection methodology as the routine platform for detection of timber adulterants (Fig.S13). The generated DNA sequences were deposited in GenBank under the Accession Numbers MK264363-MK264378.