Elephant grass (Pennisetum purpureum Schum.) is a fast-growing native African plant species that produces commercially useful lignocellulosic biomass. It has been used in many countries to replace wood for paper, particleboard, and fiberboard. There is a close relationship between the mechanical properties of elephant grass cell walls and the performance of its products. The objective of this research was to investigate the cell wall mechanical properties at different growth periods of five types of elephant grasses, i.e., P. americanum cv. Tift 23A×P. purpureum cv. Tift N51 (HP), P. purpureum cv. Tift N51 (N51), P. purpureum cv. Huanan (Huanan), P. purpureum cv. Sumu No.2 (Sumu-2), and (P. americanum× P. purpureum) × P. purpureum cv. Guimu No.1 (Guimu-1). The hardness and elastic modulus of the cell walls were investigated by means of nanoindentation. The results showed that the hardness and elastic modulus of these elephant grasses increased as growth period increased. However, the rate of increase varied for the different types of elephant grass, which could help guide the evaluation of the properties of this kind of bio-fiber resource for the production of high-quality biocomposite products.
Effect of Growth Period on Cell Wall Mechanical Properties of Elephant Grass
Yang Zhang,a,b Rui Yang,a Yan Wu,a,b Siqun Wang,b,* Cong Liu,a,b Xiaoxian Zhong,c and Juanzi Wu c
Elephant grass (Pennisetum purpureum Schum.) is a fast-growing native African plant species that produces commercially useful lignocellulosic biomass. It has been used in many countries to replace wood for paper, particleboard, and fiberboard. There is a close relationship between the mechanical properties of elephant grass cell walls and the performance of its products. The objective of this research was to investigate the cell wall mechanical properties at different growth periods of five types of elephant grasses, i.e., P. americanum cv. Tift 23A×P. purpureum cv. Tift N51 (HP), P. purpureumcv. Tift N51 (N51), P. purpureum cv. Huanan (Huanan), P. purpureum cv. Sumu No.2 (Sumu-2), and (P. americanum× P. purpureum) × P. purpureum cv. Guimu No.1 (Guimu-1). The hardness and elastic modulus of the cell walls were investigated by means of nanoindentation. The results showed that the hardness and elastic modulus of these elephant grasses increased as growth period increased. However, the rate of increase varied for the different types of elephant grass, which could help guide the evaluation of the properties of this kind of bio-fiber resource for the production of high-quality biocomposite products.
Keywords: Elephant grass; Growth age; Mechanical properties; Cell wall
Contact information: a: Nanjing Forestry University, Nanjing, P.R. China; b: Center for Renewable Carbon, University of Tennessee, 2506 Jacob Drive, Knoxville, TN, 37996, USA; c: Jiangsu Academy of Agricultural Sciences, Nanjing, P.R. China; *Corresponding author: firstname.lastname@example.org
Elephant grass (Pennisetum purpureum Schum.) is a plant species native to Africa that possesses high rates of growth and lignocellulosic biomass production (Aroeira et al. 1999; Xie et al. 2011). This plant was introduced to South America and Australia over a century ago (Strezov et al. 2008), and was also introduced to China in the twentieth century. Elephant grass is potentially suitable as a source of refined fuel products such as bio-oil, char, and combustible gases (Strezov et al. 2008), or for bioenergy production by direct biomass combustion (de Morais et al. 2009). In some countries, elephant grass also finds use as a food source for livestock (Mpairwe et al. 1998) and can be used to feed ruminants (Ajayi 2011).
Because of the interest in non-wood fibers as raw material for pulp and paper products, research has been conducted to evaluate elephant grass as a raw material for pulp and paper production (Madakadze et al. 2010). It has also been found to be a suitable raw material for the production of particle board, fiberboard, and composite products (Nguyen et al. 2010, 2011).
Raw material chemical composition, kraft pulp yield and properties, and fibre characteristics of elephant grass have been determined in an effort to evaluate them as raw materials for pulp and paper production. Elephant grass had α-cellulose and Klason lignin contents of 45.6 and 17.7%, respectively. Pulp yield, following a mild kraft process, was 50% for elephant grass. The weight-weighted fibre length averaged 1.32 mm. Pulp freeness was 139 mL for elephant grass. Elephant grass had a burst index above 5.85 kP.m2 g-1. These characteristics demonstrate the suitability of elephant grass for pulp production (Madakadze et al. 2010).
The pennisetum fiberboard made with urea-formaldehyde resin could be used as the core material for veneer overlaying. The mechanical property of pennisetum fiberboard was increased obviously by veneer overlaying on its surface (Nguyen et al. 2010; 2011). To utilize elephant grass more efficiently as a raw material for pulp, paper, particleboard, and fiberboard, it is important to understand the mechanical properties of the plant’s cell wall, as these properties strongly affect the quality of biocomposite products produced from such feedstock (Liao et al. 2012).
The nanoindentation technique is useful for determining the elastic modulus and hardness of wood cell walls along longitudinal direction (Gindl et al. 2002; Gacitua et al. 2007; Jakes et al. 2008) and may be used to measure the cell-wall mechanical properties of the stalks of crops including cotton, soybean, rice straw, wheat straw (Wu et al. 2010), silver grass (Liao et al. 2012), bamboo (Wang et al. 2013a), hemp stalk (Li et al. 2013), reed stalk (Wang et al. 2013b), and castor stalk (Li et al. 2014).
Nanoindentation is widely used for measuring micro scale mechanical properties of materials that are relatively isotropic in their elastic properties, and addressing whether the modulus measured in an indentation test represents that of some specific crystallographic direction or some average value is not an issue. On the other hand, some materials have complex hexagonal crystal structures, and because of this, results from these materials can be used to provide some insight into the importance of elastic anisotropy (Oliver and Pharr 1992).
Wu (2010) investigated the nano-mechanical properties of crop-stalk cell walls, i.e. those of cotton (Gossypium herbaceum) stalk, soybean (Glycine max) stalk, cassava (Manihot esculenta) stalk, rice (Oryza sativa L.) straw, and wheat (Triticum aestivum L.) straw by means of nano-indentation and atomic force microscopy (AFM), in order to evaluate their potential as materials for reinforcement. The elastic modulus of wheat straw was found to be 20.8 GPa, which was higher than that of the other four crops. The highest hardness was observed in cotton stalk at 0.85 GPa. The elastic moduli of the crop stalks were lower than those of most of the hardwood species, but higher than that of some softwoods and of lyocell fiber. The mean value of the hardness of the five crop stalks’ cell walls was higher than those of wood or lyocell fiber. Besides, nano-mechanical properties of switchgrass and cotton cellulose nanocrystals (CNCs) were measured using nano-indentation (Wu et al. 2013). Mechanical testing showed that the reduced modulus (Er) and the hardness (H) of switchgrass CNC films were higher than those of the cotton CNC films. The objective of this research was to investigate the effect of growth stage and grass type on the cell-wall mechanical properties of elephant grass using the nanoindentation technique.
Recent Chinese research into the planting and applications of elephant grass is predominantly focused on five varieties of the plant: P. americanum cv. Tift 23A×P. purpureum cv. Tift N51 (HP), P. purpureum cv. Tift N51 (N51), P. purpureum cv. Huanan (Huanan), P. purpureum cv. Sumu No.2 (Sumu-2), and (P. americanum × P. purpureum) ×P. purpureum cv. Guimu No.1 (Guimu-1). Of these, HP and Sumu-2 are grown in China across a wide range of latitudes, while Huanan and Guimu-1are cultivated in southern China because of their unsuitability to the colder climate of the north. There are plans for the cultivation of N51 in the future. Their genotypes, evaluated with a description of their origin, are listed in Table 1. These five kind of air-dried elephant grass stalks were obtained from the Garden of Jiangsu Academy of Agricultural Sciences (China). They were sampled at three different intervals during their life cycle, 3-months, 4-months, and 6-months, with the 1st sampling taking place on August 2 (a growing time of three months), the 2nd on September 2 (a growing time of four months), and the last at approximately six months, at the time of the local first frost and the cessation of growth in the plants.
Table 1. Elephant Grass Genotypes with a Description of Their Origin
Grass samples were cut into 4 to 5 blocks of 5 mm (L) by 2 mm (W) by 1 mm (T). The specimens were then sealed in FoodSaver polymer film, according to the process shown in previous research (Meng et al. 2013). Each small grass sample was placed between two films and pressed by an electric iron with the temperature set at 160 °C. The pre-sealed samples were then embedded in Spurr’s resin (ChemAce Chemical Supply, USA) and cured in an oven at 70 °C for 8 h. The surfaces of the cured specimens were prepared with three kinds of knives to obtain a very smooth surface. The smoothed specimens were conditioned at 21 °C and 60% relative humidity in the nanoindentation test room for at least 24 h before testing.
Nanoindentation tests were performed on a TriboIndenter® system manufactured by Hysitron, Inc. (Minneapolis, MN, USA). A Berkovich nanoindenter tip with a three-sided pyramidal shape and an area-to-depth function (Oliver and Pharr 1992) was used for all experiments. The single indentation procedure included four steps, as described by Liu et al. (2006). A total of 45 indentations were performed and were checked by rescanning the image (Fig. 1), as described in a previous manuscript (Zhang et al. 2012). Scanning probe microscopy was performed to measure the physical shape of the indentation and confirm the nanoindentation position.
Fig. 1. Post indentation scanning probe microscope image (Note. 1. Secondary cell walls; 2. Cell lumen; 3. indentation mark)
From the indentation showed in Fig. 1, the load-displacement curve can be attained. Consequently, hardness and reduced modulus were calculated from valid data according to Oliver and Pharr (1992) and Wu et al. (2009). The reduced modulus Er (i.e., the composite modulus for indenter and sample combination) was calculated from the nanoindentation measurements using Eq. 1 (Oliver and Pharr 1992):
where P is the indentation load, h and hc are the penetration and contact depths, respectively, and Ahcis the projected contact area, which is a function of the contact depth. The Meyer hardness (H) is given by Eq. 2,
where Pmax is the maximum indentation load.
Determination of holocellulose and lignin contents
The specimens were smashed and screened to 60 to 80 mesh. The contents of holocellulose and cellulose were determined according to the standard GB/T 2677.10 (1995), and lignin contents were determined according to the standard GB/T747 (2003).
A one-way analysis of variance (ANOVA), linear regression, and parameter estimation analysis were conducted using Statistic Analysis System (SPSS; USA) 16.0 version software. Statistical significance was at the 0.05 level.
RESULTS AND DISCUSSION
Reduced Modulus of Cell Walls for Elephant Grass at Different Growth Stages
The reduced cell wall moduli for the five kinds of elephant grass at different growth stages are shown in Fig. 2. The growth periods of elephant grass are about 6 months in most areas of China. The elephant grass grows more quickly during the first 3-months or 4-months than other growth periods. From this graph, it can be seen that the reduced modulus increased with the progression of elephant grass growth stage. When the growth period was 3 months, 4 months, and 6 months, the reduced modulus of HP were 16.9 GPa, 20.2 GPa, and 20.2 GPa, N51 were 15.4 GPa, 16.3 GPa, and 21.0 GPa, Huanan were 14.4 GPa, 15.7 GPa, and 17.9 GPa, Sumu-2 were 15.0 GPa, 18.4 GPa, and 20.7 GPa, and Guimu-1 were 9.93 GPa, 13.2 GPa, and 16.9 GPa. The differences, however, were not significant at the 0.05 level. One of the reasons for the changes was that the chemical components of elephant grass changed between the growth periods of 3 months to 6 months. A relationship exists between plant age and the chemical composition of the plant, and this composition influences the cell wall mechanical properties. The primary components of plant cell walls are holocellulose (i.e., cellulose, hemicellulose) and lignin (Thorstensson et al. 1992). It has been shown that the composition of holocellulose varies among different elephant grasses (Herrera et al. 1995). Among these components, cellulose dominates the longitudinal properties of cell walls (Bergander and Salmén 2002). For this reason, the cellulose content of elephant grass was tested to understand why the reduced moduli of cell walls varied. Table 2 lists the contents of holocellulose, cellulose, and lignin of the five elephant grasses in their different growth stages. Total content of holocellulose and lignin were not equal to 100% due to the extractives. The deduction may be made that there exists a relationship between growing time and the micro-mechanical properties of elephant grass stalks since both cellulose content and reduced cell modulus increased with plant age, although the rates at which the reduced modulus increased differed greatly among different types of plants. Table 3 shows the rate of increase of the reduced modulus for the third to the fourth month and the fourth to the sixth month of growth. The order, from the highest to lowest rate of increase, was as follows: Guimu-1 > Sumu-2 > HP >Huanan> N51 for the third to the fourth months of growth, and N51 > Guimu-1 >Huanan> Sumu-2 > HP for the fourth to the sixth months of growth. Bergander and Salmen (2002) summarized values for the longitudinal Young modulus of wood cell wall components: 2.0 GPa for lignin, 7.0 GPa for hemicelluloses, and 167.5 GPa for cellulose. Since the stiffness of cellulose is more than 80 times that of lignin, a slight increase of cellulose should noticeably alter the stiffness of the composite cell wall. This assumption is supported by the finding that changes of the elastic constants of lignin in a model for cell wall stiffness have only small effects on the overall longitudinal stiffness (Bergander and Salmen 2002). And findings (Fujino and Itoh 1998; Donaldson and Singh 1998; Hafren et al. 1999) suggested that the unlignified cell wall consists of cellulose microfibrils organized as cellulose-hemicellulose strands embedded in a hemicellulose sheath. The spaces between individual cellulose-hemicellulose strands is occasionally bridged by hemicelluloses. During lignification, the spaces between cellulose-hemicellulose strands and possibly also part of the hemicelluloses sheath covering the cellulose are filled with lignin, thus filling, sealing, and hydrophobising the structure. The filling of existing spaces with lignin certainly increases the overall stiffness of the structure.
Fig. 2. Reduced modulus of cell walls for the five kinds of elephant grass in different growth stage. Note: The error bar of the y-axis is the standard deviation.
Table 2. Holocellulose, Cellulose, and Lignin Contents of Five Types of Elephant Grass (%)
H: holocellulose; C: cellulose; L: lignin
Table 3. Increasing Rate of the Reduced Modulus for Five Types of Elephant Grass (%)
Hardness of Cell Walls in Elephant Grass at Different Growth Stages
A positive relationship was found between the hardness of plant cell walls and their lignin content. From Fig. 3, it can be seen that the hardness of cell walls for different kinds of elephant grass increased as they grew.
Fig. 3. Hardness of cell walls for the five kinds of elephant grass in different growth stage
Note: from left to right, it represents 3 months, 4 months, and 6 months separately
Comparison of Different Cell Wall Mechanical Properties
The reduced moduli of cell walls in five crop stalks (Wu et al. 2010) and the five types of elephant grass are shown in Fig. 4. Wheat straw had the highest reduced modulus (20.8 GPa) among the five crops. However, it was marginally smaller than that of N51, whose reduced modulus was 21.02 GPa. The rice straw and cassava stalk presented values of 19.4 and 19.0 GPa, respectively. The soybean stalk and cotton stalk both exhibited the lowest values of 16.3 GPa among the five crop stalks. The mean value of the five crop stalks was 18.4 GPa, which was also slightly smaller than the mean value of the five elephant grass stalks (19.4 GPa).
Fig. 4. Reduced modulus of cell wall in five crops and five elephant grass stalks with a growing time of six months (Note: the reduced modulus of cell wall in five crops was from a cited paper (Wu et al.2010), and the reduced modulus of five elephant grass stalks was from the present study)
The following conclusions can be drawn from this study. The reduced moduli and hardness of five kinds of elephant grasses increase as they mature. Different types of elephant grass showed differing rates of increase in reduced modulus and hardness. The elephant grass has similar cell wall mechanical properties by comparing to other crop stalks, which are from 16 GPa to 22 GPa. These results may prove useful as a guide by which the appropriate type and age of elephant grass may be chosen for the production of pulp, paper, particle board, and fiberboard.
This work was supported by the National Nature Science Fund of China (No. 31070492, No.31100417, and No.31302025), Jiangsu Province High-Level Creative Talent Plan, Special Fund for Independent Innovation of Agricultural Science and Technology in Jiangsu (CX (11) 2051 and CX (12) 5011), and Tennessee Experimental Station Project #TEN00422.
Ajayi, F. T. (2011). “Effects of feeding ensiled mixtures of elephant grass (Pennisetum purpureum) with three grain legume plants on digestibility and nitrogen balance of West African dwarf goats,” Livestock Sci.142(3), 80-84. DOI:10.1016/j.livsci.2011.06.020
Aroeira, L. J. M., Lopes, F. C. F., Deresz, F., Verneque, R. S., Dayrell, M. S., de Matos, L. L., Maldonado, V. H., and Vittori, A. (1999). “Pasture availability and dry matter intake of lactating crossbred cows grazing elephant grass (Pennisetum purpureum, Schum.),” Animal Feed Sci. Technol.78(3-4), 313-324.
Bergander, A., and Salmén, L. (2002). “Cell wall properties and their effects on the mechanical properties of fibers,” J. Mater. Res. 37(1), 151-156. DOI: 10.1023/A:1013115925679
de Morais, R. F., de Souza, B. J., Leite, J. M., Soares, L. H. D., Alves, B. J. R., Boddey, R. M., and Urquiaga, S. (2009). “Elephant grass genotypes for bioenergy production by direct biomass combustion,” Pesquisa Agropecuaria Brasileira 44(2), 133-140.
Donaldson, L. A., and Singh, A. P. (1998). “Bridge-like structures between cellulose microfibrils in radiata pine kraft pulp and holocellulose,” Holzforschung 52, 449-454.
Forage Product Certification Committee. (2001). “Registration of new cultivars of forage,”Acta Agrestia Sinica 9(4), 313-318.
Fujino, T., and Itoh, T. (1998). “Changes in the three dimensional architecture of the cell wall during lignifications of xylen cells in Eucalyptus regnans,” Holzforschung 52, 111-116.
GB/T 2677.10. (1995). “Fibrous raw material – Determination of holocellulose,” Standard Press of China, Beijing.
GB/T 747. (2003). “Pulps-Determination of acid-insoluble lignin,” Standard Press of China, Beijing.
Gacitua, W. E., Ballerini, A. A., Lasserre, J. P., and Bahr, D. (2007). “Nanoindentation and ultrastructure in Eucalytusnitens with micro and mesocracks,” Maderas. Ciencia y Tecnología 9(3), 259-270.
Gindl, W., Gupta, H. S., and Grünwald, C. (2002). “Lignification of spruce tracheid secondary cell walls related to longitudinal hardness and modulus of elasticity using nano-indentation,” Can. J. Bot.80(10), 1029-1033. DOI: 10.1139/B02-091
Hafren, J., Fujino, T., and Itoh, T. (1999). “Changes in cell wall architecture of differentiating tracheids of Pinus thunbergii during lignifications,” Plant Cell Physiol. 40, 532-541.
Herrera, R. S., Martinez, R. O., Cruz, R., Tuero, R., Garcia, M., Guisado, I., and Dorta, N. (1995). “Biomass production with elephant grass (Pennisetum purpureum) and sugar cane (Saccharum officinarum) for tropical livestock production. 2. Soluble and structural carbohydrates,” Cuban J. Agricult. Sci. 29(2), 239-246.
Jakes, J. E., Frihart, C. R., Beecher, J. F., Moon, R. J., and Stone, D. S. (2008). “Experimental method to account for structural compliance in nanoindentation measurements,” J. Mater. Res. 23(4), 1113-1127. DOI: 10.1557/JMR.2008.0131
Li, X., Wang, S., Du, G., Wu, Z., and Meng, Y. (2013). “Variation in physical and mechanical properties of hemp stalk fibers along height of stem,” Indust. Crops Prod. 42, 344-348. DOI: 10.1016/j.indcrop.2012.05.043
Li, X., Du, G., Wang, S., and Yu, G. (2014). “Physical and mechanical characterization of fiber cell wall in castor (Ricinus communis L.) stalk,” BioResources 9(1), 1596-1605. DOI: 10.15376/biores.9.1.1596-1605
Liao, C., Deng, Y., Wang, S., Meng, Y., Wang, X., and Wang, N. (2012). “Microstructure and mechanical properties of silvergrass fibers cell evaluated by nanoindentation,” Wood Fiber Sci. 44(1), 63-70.
Liu, C.K., Lee, S., Sung, L.P., and Nguyen, T. (2006). “Load-displacement relations for nanoindentation of viscoelastic materials,” J. Appl. Phys. 100(3), 033503. DOI: 10.1063/1.2220649
Ma, J., Zhang, J., Shan, L., and Yun, X. (2011). “Progress of the herbage variety approval and registration in China,” Acta Protaculturae Sinca. 20(1), 206-213.
Madakadze, I. C., Masamvu, T. M., Radiotis, T., Li, J., and Smith, D. L. (2010). “Evaluation of pulp and paper making characteristics of elephant grass (Pennisetum purpureum Schum) and switchgrass (Panicum virgatum L.),” African J. Environ. Sci. Technol. 4(7), 465-470.
Meng, Y., Wang, S., Cai, Z., and Young, T. M. (2013). “A novel sample preparation method to avoid influence of embedding medium during nano-indentation,” J. Appl. Phys. A 110(2), 361-369. DOI: 10.1007/s00339-012-7123-z
Mpairwe, D. R., Sabiiti, E. N., and Mugerwa, J. S. (1998). “Effects of dried Gliricidia sepium leaf supplement on feed intake, digestibility and nitrogen retention in sheep fed KW4 elephant grass (Pennisetum purpureum) ad libitum,” Agrofor. Syst. 41(6), 139-150.
Nguyen, T. K., Zhang, Y., and Li, W. (2010). “Study on the technology of particle board from Pennisetum stalk,” China Forest. Sci. Techn. 24(5), 97-99.
Nguyen, T., Zhang Y., Li, W., Zhong, X., and Wu, J. (2011). “Study on the technology of veneer overlaying pennisetum fiberboard,” Wood Process. Mach. (6), 4-7.
Oliver, W. C., and Pharr, G. M. (1992). “An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments,” J. Mater. Res.7(6), 1564-1583. DOI: http://dx.doi.org/10.1557/JMR.1992.1564
Strezov, V., Evans, T. J., and Hayman, C. (2008). “Thermal conversion of elephant grass (Pennisetum purpureum Schum) to bio-gas, bio-oil and charcoal,” Biore. Techn. 99(17), 8394-8399. DOI: 10.1016/j.biortech.2008.02.039
Thorstensson, E. M. G., Buxton, D. R., and Cherney, J. H. (1992). “Apparent inhibition to digestion by lignin in normal and brown midrib stems,” J. Sci. Food and Agricult. 59(2), 183-188.
Xie, X., Zhang, X., Dong, Z., and Guo, H. (2011). “Dynamic changes of lignin contents of MT-1 elephant grass and its closely related cultivars,” Biomass Bioenergy 35(5), 1732-1738. DOI: 10.1016/j.biombioe.2011.01.018
Wang, H., Wang, H., Li, W., Ren, D., and Yu, Y. (2013a). “Effects of moisture content on the mechanical properties of moso bamboo at the macroscopic and cellular levels,” BioResources 8(4), 5475-5484. DOI: 10.15376/biores.8.4.5475-5484
Wang, X., Deng, Y., Wang, S., Liao, C., Meng, Y., and Pham, T. (2013b). “Nanoscale characterization of reed stalk fiber cell walls,” BioResources 8(2), 1986-1996. DOI: 10.15376/biores.8.2.1986-1996
Wu, Q., Meng, Y., Concha, K., Wang, S., Li, Y., Ma, L., and Fu, S. (2013). “Influence of temperature and humidity on nano-mechanical properties of cellulose nanocrystal films made from switchgrass and cotton,” Industrial Crops and Products 48, 28-35.
Wu, Y.F. (1999). Volume of Forage Cultivars Registered in China, Agricultural University Press, Beijing. pp.103
Wu, Y., Wang, S., Zhou, D., Xing, C., Zhang, Y., and Pharr, G. M. (2009). “Use of nanoindentation and SilviScan to determine the mechanical properties of 10 hardwood species,” Wood Fiber Sci. 41(1), 64-73.
Wu, Y., Wang, S., Zhou, D., Xing, C., Zhang, Y., and Cai, Z. (2010). “Evaluation of elastic modulus and hardness of crop stalks cell walls by nanoindentation,” Bioresour. Techn. 101(8), 2867-2871. DOI: 10.1016/j.biortech.2009.10.074
Zhang, Y., Meng, Y., Wu, Y., Wang, S., Du, G., Jiang, H., and Zhou, Z. (2012). “Effect of enzyme treatment on the mechanical properties of wood cell walls by nanoindentation,” BioResources 7(2), 2249-2259. DOI: 10.15376/biores.7.2.2249-2259
Article submitted: August 20, 2014; Peer review completed: December 3, 2014; Responses to reviewer comments accepted: March 5, 2015; Revised version received: March 9, 2015; Accepted: May 10, 2015; Published: May 27, 2015.