Ten different wood species were collected from Sawmills in Ilorin (Kwara state) and Mokwa (Niger State). These places fall within Guinea savannah regions of Nigeria. The wood Samples were first identified by the saw millers. The local identifications were further substantiated by the literature to obtain their corresponding scientific names. A list of the species used in the study with their local names is provided in Table 1. Wood samples were collected from logs whose diameter ranged from 60 - 105 centimetres. All of the wood samples were taken from mature wood. Special care was taken to ensure that species were accurately identified using both macroscopic and microscopic anatomical features such as colour, density, pores, among other.

Table 1:
Wood samples used; their botanical name, family and local names.

Fibre characterisation of the wood samples was carried out following ASTM D-1030-95 (2007) and ASTM D-1413-61 (2007). Small slivers having radial and tangential dimensions of 2 and 5 mm from each of the wood species were macerated with acetic acid and hydrogen peroxide (1:1) and boiled in a water bath at a temperature of 100°C for 10 minutes following a procedure adopted by Ogbonnaya et al. (1997). Some macerated fibres were randomly selected and mounted on slides and then examined under a Reichet Microscope. The fibre length, fibre diameter and lumen width of unbroken fibres were measured using an eye piece micrometer after calibrating with stage micrometer. Some derived values such as the cell wall thickness (Fibre diameter - Lumen width/2); Slenderness Ratio (Fibre length/Fibre diameter); Flexibility coefficient (Lumen diameter/Fibre diameter); Runkel ratio (2 x Cell wall thickness/Lumen diameter), Rigidity coefficient (2 x Cell wall thickness/Fibre diameter), Luce's Shape Factor (Fibre diameter 2 - Lumen diameter 2 /Fibre diameter 2 + Lumen diameter 2 ) and Solid Factor (Fibre diameter 2 - Lumen diameter 2 x Fibre length) were computed from the measured fibre dimensions following the method of Saikia et al. (1997), Ogbonnaya et al. (1997), Ohshima et al. (2005) and Sadiku et al. (2016). Twenty five fibres were measured from each representative sample slide.

Variation in the fibre morphology and derived values were evaluated by analysis of variance at p ≤ 0,05. Duncan Multiple Range Test was used to compare mean values for the different species. The evaluated fibre morphology and derived values were compared to conventional species and some other hardwood species used in pulp and paper making.

Fibre length and width of both woody and non-woody species vary depending on plant species and the plant part from which the fibre is derived (Illvessalo-Pfaffli 1995). The fibre morphological variations of the Nigerian guinea savannah timber species are presented in Table 2 and 3. The analysis of variance shows that there were significant differences in the fibre length, fibre diameter,

lumen width and the cell wall thickness of the timber species (Table 2). However, S. guineense had the longest fibre (2 mm) while A. zygia had the shortest fibre length of 1,05 mm. I. gabonensis had the largest fibre diameter, lumen width and cell-wall thickness. Least fibre diameter (19,11 mµ) and cell-wall thickness (4,40 mµ) were recorded in V. colorata (Table 3). S. guineense may be regarded as long fibred hardwood species with its average fibre length of 2 mm which falls in the range (2 - 5mm) of long fibre material (Hurter 1988) while the remaining nine species are short fibred hardwood. However, Metcalfe and Chalk (1983) and Anon (1984) classified fibres below 1,60mm as short while those above 1,60mm in length are said to be long. Going by this classification S. guineese, I. gabonensis and A. leiocarpa have long fibres while the other 7 species have short fibre lengths since their mean fibre lengths are all less than 1,60mm (or 1600μm) as presented in Table 3. This finding is in support of earlier reports and observations that the average length of fibres in hardwoods was only about 1mm and in coniferous wood, about 3mm (Hurter 2001). Similarly observations, Kpikpi (1992) and Uju and Ugowoke (1997) reported fibre lengths of less than 1,60mm in some Nigerian hardwood species.

Table 2:
Analysis of variance for fibre dimension of the Nigerian guinea savannah wood species.

Table 3:
Variation in the fibre morphology of the Nigerian guinea savannah wood species.

For fine papers, both long and short fibres are needed. Short fibre contributes to the properties of pulp blends, especially opacity, printability and stiffness as well as good printability. Papers from these wood species, if blended with long fibre materials, will give good paper properties. However, the fibres from these species might not be flexible enough to allow the fibres to pack and reinforce the sheets. In the same vein, lumen size and cell wall thickness affect the rigidity and strength of the papers made from the fibres. Fibres with large lumen and thin walls tend to flatten to ribbons during pulping and papermaking, giving good contact between the fibres and consequently having good strength characteristics (Wood 1981). The fibres of the species are shorter, relatively thinner and flexible. The fibres will be able to pack tightly together and thus produce smooth and dense paper (Fengel and Wegener 1989, Mc Dougall et al. 1993).

The result of ANOVA showed that there was high significant variation in all the fibre dimensions (Table 2). S. guineense had the longest fibre length (2mm) compared to other species followed by A. leiocarpa (1,75 mm) while A. zygia had the shortest fibre length of 1,05 mm (Table 3). DMR test showed that S. guineense had extremely longer length than other species. There were no significant variation in the fibre length of I. doka, K. ivorensis and A. zygia. Also, V. colorata, I. gabonensis and V. paradoxa had similar fibre length. While A. africana and L. welwitchii had comparable fibre length (Table 3).

The fibre lengths of A. africana and L. welwitchii were similar. L. welwitchii and A. africana, A. zygia, K. ivorensis and I. doka had similar fibre diameter and lumen width. The cell-wall were thickest in I. gabonensis (6,87µm) and S. guineense (6,84µm) and thinnest in K. ivorensis (5,37µm) (Table 3). It is expected that the properties of paper will greatly differ among the wood species under study. Therefore, the species which show no significant variation in their fibre length are expected to behave in similar way when they are pulped for paper making. In all, S. guineese is expected to produce better paper properties due to its exceptional fibre length compared to other species. Paper manufactured from thick walled fibres will be dense and possess coarse surface whereas paper made from thin walled fibre will be dense and well formed (Akachukwu 2001).

Similarly to the fibre morphology, all the derived morphological indices showed high significant variations from species to species (Table 4). The derived values obtained for the wood species in this study is shown in Table 5. The most important and primary parameter needed to find the suitability of any raw material for pulp and paper is the Runkel ratio. The standard for this ratio is one (1), favourable pulp strength properties are usually obtained when value is below the standard value. I. gabonensis (0,7) and S. guineense (0,75) had the best RR while V. paradoxa and A. leiocarpa had poor RR. All the species exhibited good slenderness ratio with moderate rigidity while I. gabonensis and V. paradoxa had the best felting power.

Table 4:
Analysis of variance for derived values of the wood species.

Table 5:
Variation in the derived morphological indices of the wood species.

The flexibility coefficient which is otherwise known as elasticity coefficient, or Istas coefficient, is a function of the elasticity of the wood fibres. Depending on the elasticity rate, fibres are grouped into four following Istas et al. (1954) and Bektas et al. (1999) grouping (Table 7).

Table 6:
Elasticity coefficients of the wood fibres.

According to this grouping, V. paradoxa and A. leiocarpa are included in the rigid fibres group owing to their elastic coefficient less than 50 while the other 8 wood species have their elasticity coefficient ≥ 50 and are therefore included in the elastic fibre group (Table 6). Rigid fibres do not have efficient elasticity, so they are not suitable for paper production and are used more for cardboard production and (Akgül and Tozluoğlu 2009).

The results of the ANOVA showed that there was high significant variation in all the derived values (Table 4) while the DMR test (Table 5) revealed that L. welwitchii, A. africana,K. ivorensis had similar RR while A. zygia and I. doka also had similar runkel ratio, fibre flexibility and rigidity. FC, SR and RC of L. welwitchii and K. ivorensis were similar. I. gabonensis and S. guineense had the most flexible fibres (Table 7). A. leiocarpa had the most slender fibre. V. paradoxa fibres are the most rigid (Table 7).

Table 7:
Fibre flexibility classification of the wood species.

The suitability of the wood species for pulp and paper-making is shown in Table 8. The range of RR reported for the wood species falls in the range with 0,99 reported for both Anthonatha macrophylia and Dalium guinensis (Ezeibekwe et al. 2009) and 0,75 reported by Awaku (1994) for some tropical hardwood species but higher than 0,28 - 0,68 reported for Gmelina arborea and some Ficus species respectively (Ogunkunle 2010), 059 for Leucaena leucocephala (Oluwadare and Sotannde 2007), and 0,70 for Dacryodes edulis (Ajuziogu et al. 2010). From the findings, all the species had RR ≤ 1 except V. paradoxa and A. leiocarpa which do not pass the RR < 1 acceptable value for papermaking fibre (Xu et al. 2006) (Table 8). This implies that V. paradoxa and A. leiocarpa are not suitable for pulping considering their runkel ratio as they are relative higher than other wood species. Their RR falls into the range of undesirable runkel ratio. This makes the fibres from these wood not good enough as quality raw material for quality pulp and paper. Dinwoodie (1965) stated that, the basis for establishing the suitability of raw material for pulp and paper making is that the Runkel ratio must be less than one. If the Runkel ratio is less than one, such fibre source is suitable for paper production (Volkomer 1969). In the same vain, Bektas et al. (1999) stated that higher Runkel ratio gives lower paper strength properties especially lower burst, tear and tensile indexes. Fibre with high runkel ratio value is stiff, less flexible and form bulkier paper of low bounded area than the lower ratio fibre. Therefore, it is expected that V. paradoxa and A. leiocarpa will produce poor paper.

Table 8:
Suitability for pulp and paper-making based on fibre length and runkel ratio.

The degree of fibre bonding depends largely on flexibility and compressibility of individual fibre. Other calculated wood properties of importance are flexibility ratio and rigidity coefficient. Flexibility ratio is another important criterion for evaluating fibre quality. High flexibility shows that a material may be suitable to produce paper with greater burst and tensile strength. The higher the value of fibre length to width ratio, the greater will be the fibre flexibility and the chance of forming well bonded paper. Similarly, an increase in the rigidity of fibres results in decrease in fibre bonding. The flexibility of all the wood species are in the range of 42 and 59 with I. gabonensis having the highest (59%) followed by S. guineese (57%) (Table 7). This range is almost similar to (Brindha et al. 2012) where 60% was reported. Considering the FC > 0,55 (55%) acceptable value for papermaking fibre (Bektas et al. 1999), two of the species I. gabonensis and S. guineese will be suitable. However, flexibility ratio between 50 and 70 implies that the fibers can easily be flat and give good paper with high strength properties (Brindha et al. 2012). Therefore, V. paradoxa and A. leiocarpa may not be suitable due to their low FC of 42% and 49% respectively.

It is expected that the pulp made from all the wood species except V. paradoxa and A. leiocarpa will have greater inter-fibre bond and hence greater tensile strength which favour those properties that affects printing (Ogunjobi et al. 2014). Fibre slenderness significantly influenced the breaking length, bursting, tearing and stretch of the pulp sheets (Ogunjobi et al. 2014). All the species pass the SR > 33 acceptable value for papermaking fibre according to Xu et al. (2006). However, Bektas et al. (1999) shows that if Slenderness ratio is lower than 70, it is invaluable for quality pulp and paper production (Bektas et al. 1999). But, if the Slenderness ratio is higher than 70, it can be utilized for pulp and paper production. The felting power of the wood species under study is in the range of 40 - 90 with P. elata having the highest (59) felting power followed by V. colorata (57). These values are lower than 101,50 felting coefficient of Steculia setigera reported by Oluwadare and Egbewole (2008). There were no significant variation in the felting power of L. welwitchii, I. gabonensis, K. ivorensis, I. doka. While A. africana and S. guineese had similar slenderness ratio too. Only A. zygia do not fall in the range suitable for paper making owing to very low felting power of 40%. Generally, resistance to tear increases with increasing fibre slenderness paper made from all the species is expected to have increased tear strength suitable for wrapping and packaging purposes (Sankia et al. 1997).

^♠Corresponding author: tundesalih@yahoo.com