Wood samples from five Pinus species growing in the Usutu forests were measured to determine the variation
in wood density and pulp quality across the company landholdings.
The five species included P. patula and P. elliottii, which are the most
extensively planted pines, P. taeda, which is grown in small quantities, and two species, P. kesiya and P. maximinoi which are not planted commercially but are growing in trial plots and show
A new sampling scheme was used to evaluate wood of different ages by dissecting samples from a single tree, thus
eliminating differences in wood quality that might be caused by genetic and micro site variations.
Wood densities (assessed by x-ray densitometry) and kraft pulp properties were measured for varying ages and
altitudes for each of the five species. Canonical Variance Analysis, used to group species by wood and pulp properties, indicated that the most important properties to distinguish between species were wood density,
pulp yield and burst strength. For these properties, P. taeda and P. kesiya were found to be similar, P. maximinoi and P. patula grouped together but P. elliottii stood apart from the other species.
The effects of environmental variables such as growth rate, rotation age and growing elevation were studied on pulp
properties. There were no clear trends with pulp yield, alkali consumption or burst strength but tear strength was found to increase with increasing age and decrease with increasing elevation.
Between the two non-commercial species, P. kesiya had the most promise because of
fast growth at low elevations and good inter-fibre pulp strength properties (tensile, TEA and burst). Although P. elliottii had significantly the fastest growth and the highest wood density of the three
commercial species at low altitude, the species had a low pulp yield and poor pulp properties (tearing strength and yield) at higher altitudes. P. taeda grown at 1200 m had a reasonable pulp yield and a good
balance of tear strength and tensile strength. From the results reported here, it would be worthwhile to plant P. elliottii at the lower altitudes with P. kesiya as an alternative and to plant P. taeda or P. patula at higher altitudes.
Correlations between percentage area of disc within density classes and pulp strength properties indicated that
there may be methods for indirectly assessing the properties.
The x-ray profile from pith to bark requires more in-depth analysis. In addition apparent sheet density holds some
promise for measurement of handsheet properties.
Usutu forests cover 70,000 ha (57,000 ha planted) with an annual production of
approximately 1 million tonnes of pulpwood (Evans, 1996). The main species are Pinus patula (felled at an average age of 15 to 17 years) and P. elliottii (felled at 16 to 22 years). The forests supply
a 580t per day unbleached kraft pulp mill which sells its product mainly into the sack and linerboard markets. Variation in wood quality is a problem for the mill because it adversely affects pulp quality. Wood
quality variation could be better managed by the mill and even exploited if the underlying causes of the variation were understood. The objectives of this work were therefore to examine the effects of major genetic
and environmental factors such as species, growth rate, age and elevation on wood and kraft pulp properties.
The species selected for determination of wood density and pulping characteristics were the three production
species P. patula (which is grown on about 67% of the forest area), P. elliottii (25%) and P. taeda (6%), as well as two species which are not yet grown commercially and were sampled in trial
plots, P. kesiya and P. maximinoi.
MATERIALS AND METHODS
Wood samples were collected from five Pinus species at several different
elevations and at several different ages. Details of the samples are provided in Table 1.
Table 1. Ages and elevations sampled for each of the five pine species
A breast height disc from each of these samples was used to determine the mean wood density and the distribution of density within each disc using x-ray densitometry. All samples,
except the 37 and 32 year old samples of P. elliottii and P. taeda respectively were pulped.
A new technique was used to sample at three ages within a single tree with the younger
ages obtained from growth rings cut from an older tree. The technique is described in Barnes et al. 1999 and the sample positions are shown schematically in Figure 1.
Wood of different ages can be obtained with this type of sampling from exactly the same
trees which were used for density determinations and pulping trials, eliminating variation that might be due to genetic differences, micro-site growing conditions and other possible between tree variables.
Each species/site was represented by twenty trees chosen at random and the following wood samples were collected from each tree:
i) a disc approximately 5 cm thick at breast height (1.30 m from ground) which was used to determine density;
ii) billets approximately 25 cm long at 17, 50 and 83% of the total height of the tree at
harvesting age, representing the bottom, middle and upper thirds of each tree;
iii) billets cut at 17, 50 and 83% of heights at younger ages (by counting rings at different
locations up the bole, as described above);
iv) from each billet a number of discs 18 to 20 mm thick were cut to be used in pulping trials.
Outer rings corresponding in number to the difference between the full age and the younger age were removed to leave wood formed at the younger age.
For the majority of trees the harvesting age was 23 to 26 years and the younger samples
represented the tree when it was 18 and 11 years old. Diameter at breast height over and under bark at age of felling was recorded for each tree as well as the total height at age of
felling and height at each younger age sampled.
Density was determined from the breast height disc. Two radii were selected from each disc, usually from opposite sides of the tree, but avoiding visible areas of compression wood. The
radii were machined to 5 mm square sections and resin was extracted by refluxing with ethanol-benzene solution for 24 hours. Solvents were removed by vaporisation in a vacuum
oven and discs were brought to a wood moisture content of 12% by storing them at a predetermined atmosphere (40o C, 92% RH) for a minimum of 10 days. The strips were x
-rayed and films were prepared using techniques described by Hughes and Sardinha (1975) and Plumptre (1978). X-ray densitometer measurements were made at 200 µm intervals
along the x-ray film of wood strips and at several thicknesses of a plastic for calibration. For some pieces of the lowest and highest density wood the density determined by x-ray
transmission was compared with the density determined by measuring and weighing the wood in order to prepare a correction factor for differences due to the species of wood.
Chip samples of trees at harvest age were prepared for pulping by cutting discs from 17, 50
and 83% of tree height and chipping them with a mechanical guillotine. Resultant chips were thoroughly mixed and this included 3 discs from each of 20 trees or a total of 60 discs.
The sample used for pulping was prepared by taking the same number of approximately 20
mm thick discs from all logs representing the sample. The discs were split along the grain with a mechanical guillotine to give a chip size of approximately 20 x 20 x 6 mm.
Chip samples were kraft pulped using an active alkali of 17.5%, sulphidity of 25% and a
temperature of 170 o C with 1 hour to reach temperature and 4 hours at temperature. A target kappa number similar to the one at the Usutu Pulp mill of between 35 and 40 was used
. Digestions were made in a stainless steel pressure vessel with forced circulation and an electric heat exchanger. The cooked chips were washed free of superficial black liquor and
broken up in a propeller type disintegrator to simulate the disintegration occurring during blowing in a commercial digester; the pulp was screened using a plate with 0.15 mm wide
slits to remove shives and collected on a 106 µm aperture sieve.
The screened pulp was dried in a stream of air to about 10% moisture. The total weight of
air-dry screened pulp and the moisture content of an aliquot were determined for total pulp yield calculation. Shives were also collected, dried and weighed. The chemical consumption
was determined by titrating with standard hydrochloric acid, firstly, an ashed aliquot of the black liquor to determine total alkali and, secondly, an aliquot of the black liquor from which
the reaction products of digestion had been removed by precipitation with barium chloride to determine the residual active alkali.
Unbleached pulp strengths
The kappa number, a measure of residual lignin in the pulp, was determined by TAPPI method T236. Physical characteristics of the pulp were determined by preparing sheets with an oven-dry grammage of 60 gm-2 and testing, after conditioning, according to current ISO and
British Standards (see Table 2).
Table 2. Sheet forming and pulp strength tests used for each pulp
RESULTS AND DISCUSSION
Rate of growth
Average tree heights and diameters at breast height for different ages and altitudes for each of the five species are provided in Table 3. P. elliottii, P. patula and P. taeda trees were all
tallest at the medium elevation (1200 m) which was a reflection of site quality. Even though there was a difference in site quality for the two sites of both P. kesiya and P. maximinoi there was little difference between height growth at 11 years for either species.
In terms of tree height at age 11 years, at the lower altitude P. kesiya had the fastest
growth followed by P. elliottii and P. taeda while P. patula and P. maximinoi had the slowest growth. At the highest altitude P. kesiya and P. patula grew the fastest followed by P.
maximinoi while P. elliottii had the poorest growth. At the medium altitude there was little difference between the three commercially grown species.
P. elliottii trees had the largest over-bark diameter at breast height, followed by P. taeda, P.
patula, P. maximinoi and P. kesiya. However, in terms of wood production the differences were less since P. elliottii and P. patula included 20 and 10% bark respectively.
Table 3. Tree height and diameter at different ages and altitudes for the five species of pine
Wood and pulp properties
A Canonical Variate Analysis (CVA) was performed on the sample mean values (age and altitude) for height growth at 11 years, wood density, total pulp yield, alkali consumption,
tearing strength, burst strength, tensile and tensile energy absorption. All pulp strength properties were interpolated at a freeness of 500 CSF. CVA is a non-parametric statistical
technique for grouping individuals by their properties. The growth, wood and pulp properties were re-organised into two canonical variates which between them described over 90% of
the variation between species. The weightings for the two variates are shown in Table 4. The most important properties describing the variation between the species were total pulp
yield, wood density and burst strength. The canonical variate means for each sample are shown in Figure 2. P. kesiya and P. taeda grouped together because both species have lower
densities than P. elliottii but similar pulp yields. P. patula and P. kesiya grouped together because they also had lower density than P. elliottii but, in addition, they had higher pulp
yields than the other three species. P. elliottii samples grouped separately from the other four species because they had high density but lower burst strength (particularly at the
lower altitudes) than the other pines.
Table 4. Latent vector loadings for the first two canonical variates used to group species according to growth, wood and pulp properties
The pulp yields of samples from the three ages and three altitudes of P. patula and P. elliottii have been reported and are described in Morris et al., 1997. Pulp yields for P. elliottii, P.
kesiya, P. maximinoi, P. patula and P. taeda are shown schematically in Figures 3 to 7 respectively.
Pulp yield decreased with altitude for P. elliottii and increased for P. maximinoi and P. patula. There were no clear trends for P. taeda or P. kesiya. Pulp yield increased with age for P.
elliottii at all elevations, at 800m for P. maximinoi and at 1450m for P. patula but there was no trend with age for the other two species. Overall, the species ranking for pulp yield was P
. patula, P. maximinoi, P. kesiya, P. taeda and, lastly, P. elliottii.
The value of the mean density and the proportion for each sample within each of five density classes are reported in Table 5. These values are the average for twenty trees except for P. elliottii and P. kesiya both grown at 800 m which are averages of only 16 and 15 breast height discs respectively due to mislaid samples. The values reported are higher than those usually reported in pulping trials because, firstly, the density in this study was measured on
wood at 12% moisture content rather than the normal oven dry basis and, secondly, these values were measured on breast height discs and not on the whole tree. Breast height
values are usually higher than whole tree for pines.
Density increased with age for all five species but the magnitude differed by species. There
was approximately 70 kgm-3 difference between the youngest and oldest trees of P. elliottii and P. patula at the lowest altitude but between 50 and 60 kgm-3 for the other species.
Generally, density decreased with increasing altitude but the difference between density at
low and high altitudes was more pronounced in some species than others. The largest difference was found for P. elliottii where the 18 year old samples grown at 800 m had a density of 624 kg/m-3 and those grown at 1450 m 485 kg/m-3 . For the same age and
altitude range, P. patula had densities of 523 and 494 kg/m-3 respectively. Species ranking for density was P. elliottii followed by P. taeda, P. kesiya, P. patula and P. maximinoi.
Density distribution often differed between samples with the same mean densities. For example, for a mean disc density of 500 kgm-3 P. elliottii had 2.3% below 350 kgm-3 and 8.2
% above 800 kgm-3 while, for the same mean density, P. kesiya had 0.4% and 2.4%, P. maximinoi 6.2% and 3.3% and P. patula 37.5% and 14.5% respectively. This difference in
density class distribution across the disc is probably a reflection of fibre wall thickness which is likely to have an important link to pulp strength properties but this must still be confirmed.
Most of the discs were within the 350-500 kgm-3 density class for all elevations and all species with the exception of P. patula where the discs were mostly within the <350 kgm-3
class. P. keysia was unique in having very little disc area with a density less than 350 kgm-3. Almost 80% of the breast height disc for this species was within the range of 350-650 kgm-3. The effect of increasing elevation for P. patula and P. elliottii was to decrease the
percentage area of the disc with a density greater than 800 kgm-3. Increasing age tended to decrease the percentage of disc within the lower density classes (<500 kgm-3) and
increase the disc area falling in the greater than 800 kgm-3 class.
Table 5. Mean density and proportion of each sample within five density classes by age and altitude for the five pine species
Table 6. Correlations between density, percentage of disc within density classes and pulp properties for all species across all elevations
Correlations (r values) between density, percentage of disc area within density classes and
the pulp properties are provided in Table 6. Generally, the correlations between mean density and the pulp properties were weak accounting for less than 50% of the variation.
Correlations between the pulp strength properties (TEA, stretch, tensile, tear, sheet density and burst) and the percentage of disc area with a density greater than 800 kgm-3 were
higher than with mean density. For tear and burst strength correlations were higher with the 350-500 kgm-3 class than mean density and the correlations with the greater than 800 kgm-3 class were higher again. Correlations for stretch and tensile were highest with percentage
of disc falling in the 350-500 kgm-3 class. This illustrated that the mean density of a disc may not be the best indirect measure of pulp properties and the x-ray profile from pith to
bark should be studied in more detail.
Pulp strength properties
Tear strength at the three ages and across sites for P. elliottii, P. patula and P. taeda are
provided in Figures 8 to 10 respectively. Age had a strong positive effect on tear strength for all three species with a difference of between 2 to 4 mNm2 g-1 for the 11 and 24 year
samples. The effect of elevation was pronounced in some species, such as P. elliottii where there was a difference of 4 mNm2 g-1 between tearing strength of 24 year old trees grown at 800 and 1450 m. For other species, such as P. patula, the difference in tear strength with altitude was only slight. P. patula ranked highest across age and altitude for tear strength
followed by P. taeda and P. elliottii. Out of the individual tree samples, 24 year old P. elliottii grown at 800m had the highest tearing strength (16 mNm2 g-1 ) while 11 year old P. elliottii grown at 1450m had the lowest (7.9 mNm2 g-1 ).
Burst strength at the three ages and across sites for P. elliottii, P. patula and P. taeda are provided in Figures 11 to 13 respectively. Age and elevation had little influence on burst strength. Burst strength decreased marginally with elevation for P. patula and decreased with elevation for P. elliottii. The burst strength of P. taeda was 6.5 kPam2 g-1 regardless of
age or elevation. P. taeda ranked highest overall for burst strength followed by P. patula and P. elliottii was poorest. Twenty four year old P. elliottii at 800 m had the lowest burst
strength with a value of only 5.5 kPam2 g-1 and this was increased to 5.8 kPam2 g-1 through refining to CSF.
Variation in pulp properties
The percentage of variation in pulp properties accounted for by genetic and environmental influences of species, growth rate, age and elevation are provided in Table 7. Among the environmental variates (growth rate, tree age and elevation) age by itself accounted for the most variation in pulp properties. However, the combination of age and elevation provided a
stronger influence on pulp strength properties accounting for at least 50% of the variation in tensile, tear strength and stretch. Age and elevation had a slight influence on density
accounting for 25% of the variation but otherwise the influences of environmental variates were weak. Pulp yield was influenced most by species and age (39.3% of the variance) but
alkali consumption was not significantly influenced by species or any of the environmental variates. Among the pulp strength properties burst strength was the least effected by the
environmental variates. Growth rate (height at age 11 years) had the least influence on pulp properties.
Table 7. Percentage of variation in pulp properties accounted for by combinations of genetic and environmental variables
Surprisingly, density was not strongly correlated with pulp properties with the exception of
tearing strength (r2 =0.26). However, apparent sheet density was strongly correlated with stretch, tensile, tear and burst (r2 of 0.732, 0.748, 0.758 and 0.373 respectively).
Presumably this is because sheet density is influenced by fibre dimensions such as diameter and cell wall thickness and these have a direct effect on pulp strength properties.
Wood density, pulp yield and burst strength were the pulp properties which best
distinguished between the five species of pine. P. elliottii had high density but low pulp yield while P. patula and P. maximinoi had high yield but low density and P. taeda and P. kesiya had yields in between P. elliottii and P. patula but had low density.
Between the two non-commercial species, P. kesiya had the most promise because of fast
growth at low elevations and good inter-fibre pulp strength properties (tensile, TEA and burst). P. maximinoi had reasonable burst strength and pulp yield but was not otherwise outstanding. Although P. elliottii had significantly the fastest growth and the highest wood
density of the three commercial species at low altitude, the species had a low pulp yield and poor pulp properties (tearing strength and yield) at higher altitudes. P. taeda grown at 1200
m had a reasonable pulp yield and a good balance of tear strength and tensile strength. These properties are normally inversely correlated. For example, P. patula had high tearing
strength but low tensile strength while P. elliottii had the opposite. It is highly desirable for products such as linerboard and sack kraft for wood to have both high tear and high tensile
strength. From the results reported here, it would be worthwhile to plant P. elliottii at the lower altitudes with P. kesiya as an alternative and to plant P. taeda or P. patula at higher
Age had a dramatic effect on wood density and tearing strength with older trees providing
higher values. Increasing age had a deleterious effect for tensile wood properties (burst strength and TEA) in P. elliottii at low altitude. It may be worthwhile therefore to grow P. elliottii at low altitude on a short (18 year) rotation to provide tensile strength and the other
species at higher altitudes on a longer rotation (24 years) to provide tearing strength. It would then be necessary to blend the two wood sources into the pulp mill.
Pulp properties are expensive and time consuming to measure. Correlations between
percentage area of disc within density classes and pulp strength properties indicated that there may be methods for indirectly assessing the properties. The x-ray profile from pith to
bark requires more in-depth analysis. In addition apparent sheet density holds some promise for measurement of handsheet properties.
One of the largest problems for a pulp mill is the variability introduced by the wood. The
Usutu mill requires a consistent balance of high tear and burst strength in their pulp to be successful in the linerboard market. This investigation into the wood has shown that altitude,
age and species all contribute to the variation in the strength properties of the pulp. In the past wood was supplied to the mill according to an annual plan of operations and no
consideration was given to the underlying pulp properties. It is now possible to manage the wood supply from the forest into the mill to both maximise the pulp strength properties and ensure their uniformity.
Barnes R.D., Plumtre R.A., Quilter T.K., Morris A.R., Burley J. and Palmer E.R. 1999. The use
of stem dissection to sample trees of different ages for determining pulping properties of tropical pines. IAWA Journal, Vol. 20 (1): 37 - 43
British Standards Institution. British Standards, London:BSI.
Evans J. 1996. The sustainability of wood production from plantations: evidence over three
successive rotations in the Usutu Forest, Swaziland. Commonwealth Forestry Review 75 (3): 234-239
Hughes, J.F. and Sardinha, R.M. de A. 1975. The application of optical densitometry in the
study of wood structures and properties. J. Microsoc, 104(1), 91-103.
International Organisation for Standards ISO Standard. Geneva, Switzerland ISO.
Morris A.R., Palmer E.R., Barnes R.D., Burley J., Plumtre R.A. and Quilter A. 1997. The
influence of felling age and site altitude on pulping properties of Pinus patula and Pinus elliottii. Tappi 80 (6). 133 - 138
Palmer, E.R. and Gibbs, J.A. 1976. Pulping characteristics of Pinus caribaea from Belize Report
of Tropical Products Institute L43 iii + 43 pp.18
Palmer, E.R. and Gibbs, J.A. 1977. Pulping characteristics of Pinus kesiya and Eucalyptus
saligna from Zambia. Report of the Tropical Products Institute, L47, iv + 24 pp.
Palmer, E.R., Gibbs, J.A. and Dutta, A.P. 1988. Pulping characteristics of Pinus caribaea var. hondurensis and Pinus occarpa grown in Zimbabwe. ODNRI Bulletin No 7 IV & 32.
Plumptre, R.A. 1978. Some wood properties of Pinus patula from Uganda and techniques
developed to study them. OFI Occasional Paper No 4, Oxford Forestry Institute, 55 p.
Technical Association of Pulp and Paper Industry. TAPPI Standards: official, provisional and
useful test methods. Atlanta, USA, TAPPI.
Wise, L.E., Murphy, M. and D'Addieco, A.A. 1946. Chlorate holocellulose, its fractionation and
bearing on summative wood analysis and studies on the hemicelluloses. Paper Trade Journal 122 (2), 35-13
BACK TO TOP