In 2000, the forest cover in the world is estimated at 29.6% of the total land area or 3,869 million hectares and divided into four broad categories mainly temperate, boreal, subtropical and tropical forests. Temperate, boreal and subtropical forests occupy about 1.86 billion hectares and tropical forests occupy 2.01 billion hectares (Food and Agriculture Organization [FAO], 2005). Twenty nine countries have more than half of their land covered with forest. Malaysia is one of the countries having more than 50% of the land area covered with forest. The Malaysian tropical rain forests are one of the twelve mega-biodiverse countries in the world. The forest contains 14,500 species of flowering plants and trees, 600 species of birds, 286 species of mammals, 140 species of snakes and 80 species of lizards (Forestry Department Peninsular Malaysia [FDPM], 2004).
In 2006, the total forested land in Malaysia is 19.52 million hectares (59.50% of total land area), with 5.88 (44. 7 %) million hectares in Peninsular Malaysia, 4.40 (59.7%) million hectares in Sabah and 9.24 (75.1%) million hectares in Sarawak. From the 19.52 million hectares of forested land, 17.40 million hectares are dry inland forest or dipterocarp forest while the remaining 1.54 million hectares and 0.58 million hectares are swamp forest and mangrove forest, respectively. In order to preserve the forest resources as a renewable resource, 14.39 million ha or 73 % of the forest area have been gazetted as a Permanent Reserved Forest (PRF) for sustainable management. The area of PRF in Peninsular Malaysia, Sabah and Sarawak are 4.70 million hectares, 3.59 million hectares and 6.10 million hectares, respectively (Abdul Rahman, 2008).
The PRF in Malaysia has been systematically managed since the beginning of the century. The forests are managed into three broad forest types such as dipterocarp forest, swamp forest and mangrove forest. There are two concepts in managing the dipterocarp forest in Peninsular Malaysia, namely Modified Malayan Uniform System (MUS) and Selective Management System (SMS). The aim of MUS and SMS is to manage the forest on a sustainable basis. However, the principle behind the system is different. Under the selective logging system there is an intention to harvest the timber at the end of the 30 years cutting cycle while under the uniform system one has to wait a complete rotation of 55 years or more until the regeneration has grown into harvestable trees (FDPM, 2003b).
The MUS consists of removing the matured crop in a single felling of all trees below 45 cm diameter at breast height (dbh) for all species while SMS entails selection of optimum management regimes based on pre-felling inventory data with the minimum cutting regime of 50 cm for dipterocarps, except Chengal (Neobalanocarpus heimii) where the cutting limit would be above 60 cm dbh and 45 cm for non-dipterocarps (FDPM, 2003b).
The total forested areas to be managed under SMS and MUS are estimated at 1.3 million hectares and 1.54 million hectares respectively or 2.84 hectares of productive forests of the PFR (FDPM, 2005a). To ensure that the forest is better conserved and sustained, the annual coupe in Peninsular Malaysia has been scaled down under the 6th Malaysia Plan (1991-1995), 7th Malaysia Plan (1996-2000) and 8th Malaysia Plan (2001-2005) to 52,250 ha, 46,040 ha and 42,870 ha per annum, respectively (Azmi and Yap, 2004).
1. 2 Problem Statement and Objectives of the study
The forestry sector has significantly contributed to the socio-economic development of the country. In 2005, the forestry sector and wood products (including wood and rattan furniture) contributed about RM21.5 billion of the Malaysian Gross Domestic Product. This accounted for 4.0% of the total gross export receipts of Malaysia at f.o.b of RM533.8 billion (FDPM, 2005a). However, harvesting in the hill forests draws the public’s attention and has been blamed for environmental problems especially land slide, soil erosion, flooding and degradation of water quality. Currently, most of the harvesting is in the hill forests because of the reduction in areas of the lowland forest mainly cleared for agriculture, industrial and urban expansion since the1960s (Daniel and Kulaisingham, 1974; Wan Razali and Abdul Rahim, 1987). The hill forests are sources of water supply in the country where most of the river systems are located. According to Department of Drainage and Irrigation (2004) Malaysia has 189 river systems of which 89 are in Peninsular Malaysia, 22 in Sarawak and 78 in Sabah. According to Zul Mukhshar (2000), 97% percent of the water for agriculture, industries and domestic use is derived from streams and rivers that flow from the forest.
Forest harvesting in the hill forest in Peninsular Malaysia is mainly a ground-based system with the combination of crawler tractors and winch. Many studies recognized that forest operations using heavy machine in the hill forests affect the physical soil properties through soil compaction, increased run-off and erosion and degradation of river water quality (Borhan et al., 1987; Zulkifli, 1990; Zulkifli et al., 1987,1993; Kamaruzaman and Nik Muhamad, 1992; Baharuddin, 1995; Cornish, 2001; Mohd Kamil et al., 2001; Laffan et al., 2001; Rab, 2004).
In an attempt to reduce soil disturbance and degradation of river water quality, Rimbaka Timber Harvester (Rimbaka) machine was introduced as an alternative to the crawler tractor. Rimbaka is a modified cabling system to winch the felled logged for minimized movement of heavy machines. A study of forest harvesting using Rimbaka technique to extract logs in the hill forest is important as an alternative to the ground-based system. The Rimbaka selected in this study can be used as a basis for comparing the impact of Rimbaka and ground-based logging on physical soil properties and river water quality.
The general objective of the study is to quantify the effects of forest harvesting using Rimbaka Timber Harvester (Rimbaka) system on the river water quality and selected physical soil properties. The specific objectives are to:
a) assess selected river water quality parameters and Water Quality Index (DOE-WQI) at different forest operation stages.
b) determine the relationship between turbidity and total suspended solids (TSS) before, during and after harvesting.
c) evaluate the change in physical soil properties at different forest operations.
1. 3 Scope and Limitations of the Study
The duration of the study was 11 months which conducted before, during and after forest harvesting. The studies on physical soil properties are limited to five parameters which are texture, particle density, bulk density, porosity, and soil moisture content. The water quality index parameters measured are dissolved oxygen (DO), chemical oxygen demand (COD), biochemical oxygen demand (BOD), total suspended solids (TSS), ammoniacal nitrogen (AN) and pH. Other parameter such as turbidity is also being measured. For this study, water quality index and measured parameters such as turbidity and total suspended solid are used as indicator to evaluate the changes of water quality before, during and after harvesting. The comparisons also involved conditions such as dry and wet season. The undisturbed forest, logging area and road construction across the stream are also being considered. This study was conducted from November 2005 to September 2006 at Compartment 35, Gunung Benum Forest Reserved, Raub, Pahang.
2.1 Reduced Impact Logging
Forest harvesting operation involves three main activities mainly tree felling and log cutting, extracting of logs to temporary log landing and transportation of logs to primary log landing. In the conventional method, trees were felled using axe or hand sawing, buck into suitable length and size before being pulled out by animals (FAO, 1973). The technique of forest harvesting has changed over time. Chainsaws are used to fell the trees and heavy machines are introduced to replace animals for log extraction. The method of harvesting depends on many aspects such as physical characteristics of the environment, infrastructure, species composition, labor availability, cost and technology. There are many techniques of timber extraction which were introduced such as manual labor, animals, trucks, farm tractors, wheel skidders, tracked skidders, bulldozers, railways, cable systems (portable winch, gravity cable ways, high lead yarding and skyline system), helicopters, balloons, airships and loaders (Borhan et al., 1987; Jonsson and Lindgren, 1990; Kamaruzaman and Nik Muhamad, 1992; Shamsudin et al., 1999 and Danny, 2001).
In Peninsular Malaysia, timber harvesting using ground-based system with the combination of crawler tractors and winch lorries has been the common practice. This technique was introduced in the 1960s to replace the extraction of logs using buffalos and elephants (FAO, 1973). The combination of crawler tractor and winch lorries was proven successful in gentle terrain of lowland forest. At present, most of the forest harvesting areas are in hilly areas and are still using a combination of crawler tractors and winch lorries because of the greater production per working day (Saharudin et al., 1999).
Many studies in Malaysia and other countries reported that forest harvesting using heavy machines can cause significant and wide spread soil disturbance (Borhan et al, 1987; Kamaruzaman and Nik Muhamad, 1986,1992; Rab, 1994, 1996; Laffan et al., 2001). Forest harvesting and associated operations have also affected stream water quality through erosion and sedimentation (Baharuddin, 1988, 1995; Grayson et al., 1993; Cornish, 2001; Mohd Kamil et al., 2001; Inthavy, 2005).
In general, the effects of forest harvesting can be reduced by the adoption of Reduced Impact Logging (RIL) technique (Anon., 2003b). The term “RIL” was first used by Pinard et al. in 1995. It stands for efficient timber harvesting technique that minimizes the damage to the forest ecosystem (Jonkers, 2001). According to Richard (1999), RIL is defined as the proper implementation of a set of harvesting techniques which results in lower level of incidental damage to both the residual stand and soil, reduced soil erosion and improved water quality, so that the productive and functional capacity of the residual after logging is sustained. There are three main components of RIL namely pre-harvest operation, during harvesting operation and post-harvest operation.
Prior to harvesting, a pre-felling forest inventory is normally conducted to provide information on species distribution, timber stock, geology status, elevation and terrain. The information on forest stocking is to determine the minimum cutting regime for tree marking. Tree marking is to identify and to mark all trees for harvesting. In tree marking, all the information on the site such as river buffer zone, forest roads such as feeder road and skid trails are marked on the ground and also on the map. The activities are usually performed during forest harvesting, feeder road construction, followed by skid trail construction, log landing construction, felling the trees and cutting the trees to the log sizes. The logs were extracted from the stump to the temporary log landing followed by transportation using winch lorries to the primary log landing (FDPM, 2003). The activities after forest harvesting involve the rehabilitation of feeder roads, skid trail, log landings and logging camps. The rehabilitation activities are carried out to ensure environmental stability of the logging area after the completion of the harvesting operation.
2.1.1 Forest Road
Forest road is an important part of forest harvesting operation to ensure a smooth flow of logs from the forest. Before forest harvesting is done, road planning must be completed and subsequently constructed according to the specifications in the logging permit and proper forest engineering standards to control density and to minimize soil disturbances and stream sedimentation (FDPM, 1999).
Gullison and Hardner (1993) reported that knowledge on road design ia an important component of forest management. Their studies and computer simulation showed how road designs can be used to minimize damage to the forest during logging. In order to minimize the impact of compaction, the forest manager should minimize the area of skid trail and rehabilitate the areas (Rab, 1996). Poorly located, constructed or maintained road can cause major water degradation with serious and long term social implications to the affected communities (Wells, 2001). According to Jonsson and Lindgren (1990) the construction of forest roads is blamed for the negative impacts of forestry activities such as erosion and other damaging environmental consequences. Yet some of these effects can be controlled by proper planning.
There are four classes of logging roads in Peninsular Malaysia mainly primary road, secondary road, feeder road and skid trail. The main and secondary roads are long haul transportation systems connected to the public road to harvesting area while feeder road and skid trail are short haul transportation in the harvesting area (FDPM, 1999). The types of the forest road in Peninsular Malaysia are shown in figure 2.1.
2.1.2 Log Landing
Log landing is an area where logs are assembled before loaded into lorries to be transported to sawmills. The proper construction of log landing is to minimize its size in order to reduce the loss of harvestable forest area, as well as to prevent sedimentation of watercourse. The log-landing density and surface area can be minimized by proper planning of roads and landings prior to logging. Heninrich (1995) suggested that log landing area should be located on slight sloping ground-based. The best sites for log landing are in the open areas away from streams or normally at least 30 meters radius from any running stream. In the United States, the minimum radius of log landing is around 30-40 m but it depends on the size of logging setting. The log landing is important to minimize the effects of soil disturbance without undue adverse effect on the efficiency of yarding and operation (Chua, 1986). Rab (1996) reported that log landing in Australia occupies 3% - 4% of the harvesting area. In Peninsular Malaysia, the areas of log landing do not exceed 0.5 ha for every 100 ha of the area to be harvested. The log landing should be at least 50 meters away from permanent watercourse (FDPM, 2003b).
2.1.3 Buffer Zone
A buffer zone is defined as an area of land along a stream, left undisturbed during forest operations. Both sides of the river are sensitive areas especially for the movement of heavy machines and felled trees. The purpose of the buffer zone is to provide an undisturbed area on both sides of the stream to protect the stream from soil erosion and as a filter to the stream to maintain water quality (Bren, 1998).
The concept of buffer zone has been used in the forestry sector for a long time and widely adopted in many countries including Australia, Canada, New Zealand and Malaysia and United States (Pearce and Hamilton, 1986; O’ Loughlin et al., 1990; California Department of Forestry, 1992 as cited in Abdul Rahim et al., 1993). Gilmour (1977) recommended that the width of the river buffer zone should be at least 20 meters for both sides of the stream. The most commonly recommended width ranges from 20 to 30 meters on both sides of the stream (Abdul Rahim et al., 1993). The minimum width of river buffer zone in Peninsular Malaysia is 20 meters but this depends on slope and calculated as: Buffer width (m) = (7.6 + 0.6 x slope %). Table 2.1 shows the width of the buffer zone in Peninsular Malaysia.
Protection areas-harvesting is not allowed in this area
Sources (FDPM, 1999)
Effects of forest harvesting on soil properties
According to Abdul Rahim (2004) the soil is in the state of equilibrium. Acceleration occurs when the soil is disturbed by harvesting activities. Soil disturbance by the forest harvesting operations depends on various factors such as inherent soil properties, soil moisture content, topography, technique of harvesting, and forest conservation practices (Laffan et al., 2001). Various techniques of forest harvesting have been studied to determine the extent of soil disturbance for ground-based, cable logging and helicopter logging. The comparison of soil disturbance between cable logging and conventional logging in Tasmania (Williamson, 1990) and in Victoria (Rab, 1994, 1996) showed that cable-logging has significantly reduced surface soil disturbance. Approximately about 11% of the cable-logged coupe has the topsoil disturbed compared to between 60% and 80% of the coupes in the Victorian central highlands where conventional logging was practised .
Rab (1994) studied the changes in physical soil properties associated with clear felling of Eucalyptus regnans forest in Southern Australia using mechanical skid trail of log to a landing using tyre skidder and flexible steel track skidder. It was found that 87% of the area had some soil disturbance. About two thirds of the logging areas were disturbed by the logging activities. Skid trail and log landing contribute to the soil disturbance of about 25% and 5%, respectively. The most common disturbance was in the topsoil (65%).
Borhan et al. (1987) compared soil surface damages in Tekam Forest Reserve, Jengka, Pahang by highlead yarding system with a crawler tractor logging. They reported that the total area of surface damaged by highlead yarding system was higher (23%) compared to tractor logging (18%). A comparative study of helicopter harvesting and tractor harvesting by Danny (2001) reported that helicopter harvesting created canopy opening and skid trail of 4% - 11% compared to tractor harvesting (15.91%). A higher canopy allows more direct rainfall to the soil surface hence resulting in higher erosion.
Kamaruzaman (1992) reported a major ground disturbance on crawler tractor-logged areas in Tekam Forest Reserve, Pahang. Based on his observation, the total percentage of areas disturbed increased by the logging operation during wet season on skid trail caused by skidding. The moving logged being pulled by the tractor tore out vegetation, displaced litter and dug into the soil along the path. This problem became more serious on steep topography. Skid trail occupies the largest percentage ranging from 34.2 to 52.1% of the total disturbed area.
According to Baharuddin (1995) the effects of crawler tractor logging in Jengka Experimental Basin in Tekam Forest Reserve, Jerantut, Pahang significantly changed the soil physical properties. The infiltration rate was higher in undisturbed forest soil compared to the skid trail and logging road. The higher the infiltration rate the lower is the surface runoff resulting in reduced erosion.
Severe erosion occurred during the first year after logging and was less severe in the second year. Slope is one of the factors influencing soil erosion from the skid trail and logging road during the first year after logging operations. The amount of soil erosion can be reduced if the gradient of the skid trail and logging roads does not exceed 20 percent.
Shukri et al. (1999) studied the impact of forest harvesting using Long Haulage Ground Cable System (LHGCS) and tractor logging on soil properties. No skid trails were constructed in the harvesting area using LHGCS. They found that the LHGCS had less soil compaction and greater soil moisture compared to the conventional method. These results clearly showed that the movement of heavy machine causes severe soil compaction.
The effects of various techniques of forest harvesting on soil disturbance also influence the physical soil properties. Soil physical properties mainly texture, particle density, bulk density, porosity and soil moisture content are dominant factors which affect plant growth. These properties determine the availability of oxygen in soil, the mobility of water into or through the soil profile, and the ease of the root penetration (Duane and Raymond, 2004).
2.2.1 Soil Texture
Soil consists of soil particles of varying sizes and can be separated into sand, silt and clay. Sand, silt and clay contents are the most fundamental soil physical properties and are widely used in soil classification. Naime et al. (2001) classified soil particles, according to their size such as clay (<2 µm), silt (from 2 to 50 µm) and sand (from 50 µm to 2 mm). The knowledge of soil particle size distribution provides information to predict water movement in the soil and its vulnerability to erosion.
Baharuddin (1995) reported that silt and clay in the logging road were significantly higher than that in the skid trail and undisturbed forest. The logging road and skid trail were located on exposed area without vegetative cover to protect soil surface from the impact of rain drops compared to undisturbed forest. Rab (1996) studied the effects of physical soil properties on logging and slash burning in the Eucalyptus forest of southeastern Australia, He reported that in the burnt areas, there is no difference in particle size distribution of topsoil profiles compared to the undisturbed forest.
Makineci et al. (2006) studied the effects on soil texture in the long-term harvesting on skid trail in fir (Abis bornmulleeriana Matt) plantation forest in Belgrade forest, Turkey. They reported that the composition of sand, silt and clay in the skid trail is 52.13%, 27.55% and 20.32% respectively compared to undisturbed area with the composition of 48.84%, 21.99% and 29.17%, respectively. The reason for the difference in the silt and clay might be due to the surface flow and soil erosion.
2.2.2 Bulk Density and Porosity
Bulk density is the parameter used to measure soil compaction. It refers to the mass of soil and pore space per unit volume of soil on oven dry basis. The higher the bulk density the more compact the soil. Soil compaction reduces soil pore space, soil water content and infiltration. The excess water will form a puddle or flow as a surface runoff (Anna, 2005). Rajaran and Erbach (1998) reported that the soil bulk density does not change significantly with the drying stress. Froehlich, (1983) as cited by Jonsson and Lindgren (1990) pointed that soil compaction depends on soil texture, moisture, machine pressure (machine size, load, support area, weight, distribution) and number of machine passes.
Many studies reported that the application of heavy machine in forest operations generally increased soil bulk density. Crucial changes may occur on the physical properties particularly the higher bulk density, lower porosity, organic matter content and water holding capacity of the soil because of the compaction (Kamaruzaman and Nik Muhamad, 1992; Rab, 1994, 1996; 2004; Grigal and Batest, 1997; Laffan ,2001; Makinechi et al., 2006). Barber and Romero (1994) found that the use of bulldozers and straight blade for clearing subtropical forest in eastern Bolivia caused significant soil degradation as manifested by 10% to 20% increase in bulk density and 6.2% loss of total porosity at 0-20 cm depth.
Rab (1994) reported that the bulk density in the top and sub soil of disturbed areas were significantly greater than the undisturbed areas. The mean bulk density in the top and sub soil of disturbed areas increased about 33% and 65% respectively compared to the undisturbed areas. Soil compaction would also significantly increased bulk density in the skid trail and the log landing. Laffan (2001) reported that bulk density is 20% higher in the disturbed sites compared to undisturbed sites which used cable logging on steep slopes in Tasmania, Australia.
In Malaysia, Kamaruzaman and Nik Muhamad (1992) reported that the range of bulk density under the five tier classification of soil disturbance in the hill forest is from 1.01 -1.55 g/cm3. Nussbaum et al., (1994) found that the removal of top soil during logging activities in Sabah can lead to soil compaction and low infiltration of water resulting in high bulk density of 1.44 g/cm3 as compared to the bulk density of the top soil of only 0.52 g/cm3. Baharuddin (1995) reported that the bulk density in undisturbed forest soil ranged from 1.08 to 1.18 g/cm3. The bulk density values for soils on the skid trail and logging road were 1.49 g/cm3 and 1.60 g/cm3, respectively. The increase on the skid trail and logging road are 32% and 42%, respectively.
On the other hand, a study by Shukri et al. (1999) reported that forest harvesting using Long Haulage Ground Cable System Technique (LHGCS) resulted in less soil compaction. The average bulk density is 1.1 g/cm3 compared to 1.5 g/cm3 in the tractor logging method.
Kazuhito et al., (2004) reported that the bulk density on steep slope at a Bukit Tarik experimental Watershed, Selangor ranged from 0.56 to 0.63 g/cm3 before logging and 0.72 -0.77 g/cm3 after logging. However, on the gentle slope the bulk density before and after logging is 0.54-0.57 g/m3 and 0.71 to 1.07 g/cm3, respectively.
2.2.3 Soil Moisture Content
Kamaruzaman and Nik Muhamad (1992) found that soil moisture content from logging operation in hill forest for five soil disturbance classes ranged from 12.28-29.65%. Undisturbed Class had the highest soil moisture content of 29.65% followed by Litter Layer Disturbed Class with 23.50%, Top Soil Exposed Class had 19.54%, Subsoil Exposed Class 16.14% and Subsoil Exposed and Compacted or Puddled Class 12.28%. Soil moisture content showed a decreasing trend from undisturbed to compacted or puddled soils.
Baharuddin (1995) reported that the mean available water in an undisturbed forest soil was 151.99 mm/m as compared to 69.4 mm/ m for the skid trail and 63.6 mm/m for logging road. A higher amount of water in the undisturbed area evidently shows that the litter layer retains more water. The presence of a humus layer prevents crust formation and probably reduces evaporation rate, improving soil structure and increase infiltration rate.
A study by Shukri et al.,(1999) reported that forest harvesting using Long Haulage Ground Cable System Technique (LHGCS) had 8% greater soil moisture compared to the tractor logging. This could be due to more canopies opening in the tractor logging as a result of large number of residuals which were removed along the skid trail and logging road. Consequently, there was more direct sunlight reaching the forest floor and promoted a greater rate of evaporation.
2.3 Degradation of River Water Quality
Rivers are important water resource and considered polluted when there are changes in their chemical and physical characteristics that make them unsuitable for consumption. One of the indicators for river health is water quality. According to Mohd Kamil and Abdul Rashid (1999), water quality is defined as the suitability of water in a water body especially a river for a specific use.
Many studies reported that forest harvesting and associated operations have been shown to affect the river water quality in many countries, principally through erosion, runoff and sedimentation (Zulkifli et al.,1987; Zulkifli, 1990; Baharuddin, 1995; Cornish, 2001; Laffan et al., 2001; Inthavy, 2005). Removal of trees, construction of logging roads and log landing exposing the soil to the driving force of rainfall, resulting in acceleration in soil erosion and surface runoff affected the river water quality.
According to Urdabe (1992) soil erosion is minimal when the soil surface is protected by the forest and this will increase slightly when the trees are removed. This is supported by Baharuddin (1995) who reported that soil loss was 13,300 kg/ha/year and 10,100 kg/ha/year for logging road and skid trail compared to 500 kg/ha/year for the undisturbed forest.
Zulkifli et al., (1987) studied the effects of logging on stream water quality. They compared changes in pH, color, conductivity, turbidity and suspended solids in discharge of three watersheds in the hill forest areas of Negeri Sembilan. They found an increase in turbidity and suspended solid concentration as a result of conventional logging during wet season. Hence they concluded that harvesting using tractor logging seriously deteriorated stream water quality parameters.
However, Grayson et al., (1993) reported that forest harvesting does not have a major impact on river water quality. There was an increase in turbidity and suspended solids at baseflow, but these were small in absolute terms and of similar magnitude to the measurement error. The use of gravel reduce sediment production, provided a sufficient depth of material used. By increasing the level of road maintenance, this will increase the traffic load control sediment production rates. But by not doing so, it will cause an approximately 40% of sediment increase. The result indicates that by identifying the areas that produce runoff, it is possible to prevent contaminated runoff reaching the streams. Roads, on the other hand, produce large quantities of sediments, even when well maintained. Carefull consideration of their placement and management is therefore paramount.
Baharuddin et al., (1999) studied the impact of Long Haulage Ground Cable System (LHGCS) on sediment load and water quality in Jengai Forest Reserve, Dungun, Terengganu. There are two catchments area at Jengai Forest Reserve. One catchment area undergo logging activities, while the other catchment functions as a control. Based on 26 months of measurements they found that the mean monthly suspended sediment yield increased 34% after logging but no changes were observed in physical stream water quality except for a slight increase in turbidity of about 10%. However, stream water chemistry changed before and after logging.
Cornish (2001) studied the effects of harvesting, logging road construction and forest regeneration on stream turbidity level in a moist eucalypt forest, Australia. He reported that the construction of permanent road increased the turbidity level, but the increase only persisted in catchment area containing a number of stream crossing. The construction and subsequent use and presence of permanent road provide a source of additional sediments in the streams. The road that generates sediments appeared to reach the stream at stream crossing. A direct connectivity between road and stream is a factor in the supply of additional sediments to drainage.
Mohd Kamil et al., (2001) studied the effects of landuse activities in Pasoh, Negeri Sembilan based on the computation of DOE-WQI and Harkin’s Index. They compared four different types of land use: undisturbed forest, swamp, degraded logged-over forest and agriculture. The water quality chosen for the study included dissolve oxygen (DO), biochemical oxygen demand (BOD), chemical oxygen demand (COD), total suspended solid(TSS), ammoniacal nitrogen(AN), pH, temperature, electric conductivity and turbidity. They found that all the areas of the stream were under class II according to the Department of Environment- water Quality Index (DOE-WQI). Class II means that the water quality only need conventional treatment for water supply compared to no treatment for class I. They concluded that the water quality was of high quality in the undisturbed forest area.
Samin (2002) also studied the effects of logging on stream water quality before, during and after forest harvesting. He assessed the water quality parameters such as pH, electrical conductivity, temperature, dissolve oxygen, color, turbidity and suspended solids in the discharge of three watersheds in hill forest areas in Deramakot Forest Reserve, Sabah. He observed that after the logging road construction and harvesting had commenced the values of pH, electrical conductivity, temperature, color, turbidity and suspended solids increased. Four months after logging activities stopped the concentration of suspended solid, turbidity, pH, electric conductivity and colour reverted back to normal conditions. This was attributed to the rapid spread of ground vegetation thereby protecting the soil from erosion. He also found that the buffer zone of 5-30 meter width on both sides of the stream was not able to completely control the movement of the sediments into the stream.
There are many parameters in determining river water quality. The parameters normally used depend on the objectives of the study. The Department of Environmental, Malaysia uses six parameters to determine water quality index mainly dissolved oxygen (DO), biochemical oxygen demand (BOD), chemical oxygen demand (COD), total suspended solid(TSS), ammoniacal nitrogen (AN) and pH. Turbidity is one of the parameters to determine the degradation of water quality by visual observation.
2.3.1 Dissolved Oxygen (DO)
Oxygen is a necessary element to all forms of life. According to Lopes et al., (2005) the concentration of oxygen is one of the main criteria for the assessment of water quality. Oxygen is involved in the main processes occurring in the water column, photosynthesis, organic matter degradation and bacterial nitrification, as well as surface reaeration. The available oxygen in the river body is the main factor for aquatic life in rivers. Several studies reported that the concentration of DO depends on the velocity of flowing river. DO concentrations will increase with increasing velocity of the river flow (Bardalo et al., 2001; Mohd Kamil et al., 2001; Samin, 2002; Jasrul, 2004). In this context, rainfall is an important factor influencing the concentration of DO.
Shahril (2001) conducted a study on river classification in Sungai Tembeling, Sungai Pergau and Sungai Keniang Kecil. He found out that DO concentration is higher at upstream compared to the middle stream and downstream. Jasrul Nizam (2004) studied the effects of highway construction on DO in several rivers along East Coast Highway from Karak to Kuantan. He reported that the mean concentration of DO is 7.66 mg/L with the minimum value of 1.87 mg/L.
Danny (1989) studied the effects of logging road construction on water quality in Sungai Menuap Watershed, Sarawak and reported that the mean concentration of DO in undisturbed forest is 7.94 g/mL and increased to 8.28 mg/L during logging road construction. However, Samin (2001) in his study on the effects of logging road construction and timber harvesting activities in Deramakot Forest Reserve, Sabah reported that the mean DO concentration decreased during and after logging with the concentrations of 4.98 mg/L and 3.07 mg/L respectively compared to the undisturbed forest with a value of 6.51 mg/L.
Mohd Kamil et al., (2001) studied the effects of landuse patterns on the stream water quality in Pasoh, Negeri Sembilan and reported that the mean DO concentration in undisturbed forest, swamp forest, logged-over forest and agriculture area are 6.84 mg/L, 5.98 mg/L, 6.38 mg/L and 6.94 mg/L, respectively. The concentration of DO in undisturbed forest is high because the water body is still clean while the process of photosynthesis by aquatic plants maintains the oxygen content in the water.
2.3.2 Biochemical Oxygen Demand (BOD)
The concentration of Biochemical Oxygen Demand (BOD) indicates the extent of pollution by biodegradable organic matter present in the water. The BOD test is an important environmental index to determine the relative oxygen requirements of wastewater, effluents and polluted waters. It measures the molecular oxygen utilized during a specified incubation period for the biochemical degradation of organic materials and the oxygen used to oxidize inorganic materials such as sulphides and ferrous ions. It can also be an indicator to measure oxygen used to oxidize reduced forms of nitrogen (nitrogenous demand), unless their oxidation is prevented by an inhibitor (American Public Health Association [APHA], 1999). The BOD concentration depends on the river inputs, season, and temperature (Bardalo et al., 2001; Mohd Kamil et al., 2002; Jasrul, 2004; Lopes et al., 2005).
Jonnalagadda and Mhere (2001) studied water quality of Odzi river in eastern highlands of Zimbabwe. They found that BOD value was low during the dry season (1.5 mg/L) compared to the wet season (2.4 mg/L). Bardalo et al. (2001) in his studies in Bangpakong River, eastern Thailand found that the BOD increased in the dry season (1.36 mg/L) compared to the wet season (1.24 mg/L).
Mohd Kamil et al. (2001) reported that BOD values in undisturbed forest, swamp forest, logged over forest and agriculture area are 0.55 mg/L, 0.96 mg/L, 0.73 mg/L and 61.03 mg/L, respectively. The value of BOD in undisturbed forest is lower because the water body is still clean and will remain clean unless there is increasing organic matter in the water.
2.3.3 Chemical Oxygen Demand (COD)
Chemical Oxygen Demand (COD) is defined as the measure of the oxygen equivalent to the organic matter content of a sample that is susceptible to oxidation by a strong chemical oxidant (APHA, 1989). Until now, COD which can indicate the level of pollution for water contaminated by reductive pollutants is the main determinant used to assess organic pollution in aqueous systems and is one of the most important parameters in water monitoring (Yonggang and Zeyu, 2003). Bardalo et al. (2001) in their studies in Banggpakong River, eastern Thailand reported that the COD increased dramatically in the dry season from 27.5 mg/L to 68.5 mg/L.
Mohd Kamil et al. (2001) reported that the values of COD are higher in agriculture area (18.0 mg/L), followed by logged-over forest (13.8mg/L), swamp forest (10.5 mg/L), and undisturbed forest (5.83 mg/L). They also reported that the values of COD increased during the dry season because of the low flow and reduced dilution of chemicals in the water. According to Jasrul (2004), the concentration of COD in the construction of east coast highway from Karak to Kuantan was between 2.60 – 16.20 mg/L. Generally the value of COD increased because of decomposition of organic matter. In undisturbed forest, the COD is lower because the water is still clean and not polluted.
2.3.4 Ammoniacal Nitrogen
Nitrogen is an essential nutrient for plant growth, normally comprising about 12-14% of the mass of the cell protein. Its availability to any particular type of organism, however, depends on the chemical form in which it is present. The form of nitrogen in the environment ranges from organic and inorganic such as ammonium nitrogen, nitrogen gas, nitrites and nitrates. Ammonia may be produced from the organically bound nitrogen by extra-cellular biochemical action on dead plant and animal tissues and animal fecal matter, by endogenous respiration of living bacteria cell, and from dead and lysed cells. The general term, ammonification, has been applied to the release of ammonia from organic matter (Barnes and Bliss, 1983).
Jasrul (2004) reported that the concentration of ammoniacal nitrogen in the river along the east coast highway is from 0.04 to 0.5 mg/L. Mohd Kamil et al., (2001) found that the concentration of ammoniacal nitrogen is higher in swamp forest with the concentration of 0.17 mg/L compared to 0.03, 0.02, 0.01 for agriculture area, logged-over forest and undisturbed forest, respectively. The decomposition of leaves and other plant debris by microorganisms are believed to have contributed more nitrogen to the water. Sources of nitrogen include the fixation of nitrogen gas by certain bacteria and plants, the addition of organic matter to water bodies, and small amounts from the weathering of nitrogen rocks. Organic N breaks down into ammonia which eventually becomes oxidized to nitrate nitrogen (Barnes and Bliss, 1983).
The pH of the water serves as an index to pollution which shows the degree of acid or base forming chemicals. However, pH by itself is not an absolute measure of pollution or contamination by natural solutes (Samin, 2001). The pH value of acidic water varies from 0 to 7 and that of alkaline water between 7 to14, while neutral water has a value of 7. The measurement of pH is one of the most important and frequently used test in water chemistry. It is the symbol of a scale, numbered from 0 to 14, that rates water solutions according to their acidity or alkalinity (Elvina and Mohd Kamil, 2004). According to Jonnalagadda and Mhere, (2001) unpolluted streams normally show a near neutral or slightly alkaline pH. The pH is slightly alkaline due to effluents from the fish. The pH value less than 7 is due to the leachates and runoffs.
Danny (1989) reported that the mean pH value in undisturbed forest is 5.37 and increased to 5.65 during the construction of logging road. Zulkifli et al.(1987) also reported that the value of pH in undisturbed forest is less than the disturbed forest. The mean pH value in undisturbed forest is 6.1 compared to 6.7 in disturbed forest. During the wet season the pH value is 6.0 and increased during the dry season. However, Mohd Kamil et al. (2001) reported that the pH value in undisturbed forests is 6.0 and higher than in the logged-over forest, swamp forest and agriculture area.
2.3.6 Total Suspended Solids (TSS)
Suspended solids are the solids that can be filtered by a 1.2 µm membrane filter. Any particle passing the filter is considered dissolved, and any particle retained is considered suspended. The sum of the dissolved and suspended solid is the total soil sediment. Total Suspended solid (TSS) is the sum of the suspended solid (Tchobanoglous and Schroeder, 1987). TSS may include both organic and inorganic materials. The inorganic fraction is made up largely of silt, sand and clay soil particles, while the organic fraction may be either living or dead organic matter (Darrell, 1971).
Season influences the concentration of TSS. Jonnalagadda and Mhere (2001) reported that the TSS increased in wet season and decreased in the dry season. The highest value of TSS in a wet season is 107 mg/L and the lowest is 40 mg/L in the dry season. In Bukit Berembun Forest Reserve, Zulkifli et al., (1987) studied the effects of selective logging on physical stream water quality between three watershed areas in a hill tropical forest. They reported that the TSS was lower during the dry season and higher in the wet season. The mean TSS during the dry season is 14.29 mg/L and dramatically increased to 73.29 mg/L during the wet season.
The comparative study between undisturbed forest, swamp forest, logged over forest and agriculture area by Mohd Kamil et al. (2001) shows that the concentration of TSS is 1.42, 4.08, 2.50 and 6.42 mg/L, respectively. In undisturbed forest and logged over forest, the concentration of TSS is lower because the soil surface is protected by vegetative cover. TSS values are higher in downstream compared to the upper stream because all the sediments produced at the upper stream accumulate at the downstream (Mohd Kamil et al., 2001).
Turbidity is an indication of apparent color of water due to the presence of suspended inorganic matter such as silt and clay (Husain, 1990). Turbidity reduces light penetration caused by particulate solids (Darrell, 1971). According to Bardalo et al. (2001), turbidity level is high due to the runoff from surrounding agriculture and other farming activities. The water transparency decreased significantly during the dry season with a mean turbidity of 69.9 NTU become 30.9 NTU in the wet season. Cornish (2001) studied the effects of roading, harvesting and forest regeneration on stream turbidity level in a moist eucalyptus forest, New South Wales, Australia. He reported that turbidity level was higher in forest harvesting areas either without regeneration or with regeneration compared to the undisturbed areas.
Danny (1989) reported that the turbidity mean level in undisturbed forest was 8.7 NTU and increased to 60.1 NTU during construction of logging road. He also compared the effects of tractor harvesting on stream turbidity between dry season and wet season and found that the stream turbidity in the dry season (35 NTU) had increased during the wet season (287 NTU).
Samin (2002) compared the turbidity between the dry and wet seasons. He reported that the levels of turbidity during the dry season are 7.02, 8.73 and 6.53 NTU for before, during and after logging, respectively. The level of turbidity increased during the wet season with means of 88.51, 174.31 and 110.21 NTU.
Turbidity is higher during the wet season and lower during the dry season because of the rainfall influencing runoff, erosion and sedimentation. In undisturbed forest, the values of turbidity is low compared to disturbed forest due to the protection of soil surface by the vegetative cover and only a small amount of runoff and erosion flow into the streams.
2.3.8 Water Quality Index
The Water Quality Index (WQI) is a mathematical instrument used to transform large quantities of water quality data into a single number which represents the water quality level (Nives, 1999; Jonnalagadda and Mhere, 2001). WQI provides a simple and understandable tool for managers and decision makers on the quality and possible uses of a given water body .The first WQI was developed in the United States by Horton (1965) and applied in Europe since the 1970s (Bardalo et al., 2001).
The parameters selected for the calculation of WQI is through the combined judgement of a panel of water quality experts. Water quality evaluation (WQE) is calculated by summing up individual quality ratings and weighting these parameters in total quality evaluation. The highest value of water quality is 100 score (Bordalo et al., 2001; Nives, 1999; DOE, 2004).
Different countries use different parameters. For example, Scottish WQI use nine parameters such as biochemical oxygen demand, dissolved oxygen, fecal coliform bacteria, pH, temperature, total nitrate, total phosphorus, total solids and turbidity (Nives, 1999). Bordalo et al. (2001) used eight parameters to determine the WQI in Bangpakong River, Thailand. These are biochemical oxygen demand, dissolved oxygen, fecal coliform bacteria, pH, temperature, total phosphorus, total solids and turbidity.
There are two methods to determine water quality index in Malaysia. These are the Department of Environment- Water Quality Index (DOE-WQI) and Harkin’s Index. The DOE – WQI is an opinion-poll formula. A panel of experts was consulted on the choice of parameters and on the weightage to be assigned to each parameter. Harkin’s WQI Index is an objective method based on Kendall’s nonparametric multivariate ranking procedure (Harkins, 1974). This index follows a statistical approach for analyzing water quality based on the rank order of observations compared to a set of “control values”, which are usually a set of water quality standards or recommended limits. The index does not set parameter requirements and is therefore very flexible in application since only those parameters of interest to the index that user needs to be included in the computations (Mohd Kamil et al., 2001).
The calculations of DOE-WQI are performed not on the parameters themselves but on their subindices where the values were obtained from a series of equations shown in Appendix 1. These are best-fit equations obtained from rating curves. The subindices for the chosen parameters are named SIDO, SIBOD, SICOD, SIAN, SISS and SIPH, and the formula used in the calculation is:
WQI = 0.22* SIDO + 0.19* SIBOD + 0.16* SICOD + 0.15* SIAN
+ 0.16* SISS + 0.12* SIPH
The classification of water quality according to DOE-WQI is divided into six classes according to the usage of water using Interim National Water Quality Standards as shown in Table 2.2.
Mohd Kamil et al. (2001) reported that the DOE-WQI in an undisturbed forest, swamp forest, logged over forest and agriculture are classified as Class II where the water supply required conventional treatment before used. However, the values of DOE-WQI are higher in undisturbed forest followed by disturbed forest, swamp forest and agriculture area. They also reported that the stream water quality depends on the season. Rainfall affects water quality by decreasing infiltration and increasing runoff and soil erosion.
The main species dominating the area are Seraya (Shorea curtisii), Meranti Melantai (Shorea macroptera Dyer.), Meranti Tembaga (Shorea leprosula Mig.), Meranti Kepong (Shorea ovalis Korth), Meranti Rambai Daun (Shorea accuminata Dyer) and Meranti Sarang Punai (Shorea parvifolia Dyer). The main species other than Meranti are Mersawa (Anisoptera spp), merawan (Hopea spp) and Keruing (Dipterocarpus spp). The main non–Dipterocarp species comprise of Kelat (Syzygium spp), Merpauh (Swintonia spp), Penarahan (Myristicaea spp) and Medang ( Lauraceae spp ).
One of the main reasons that caused changes in soil texture might be due to soil surface erosion. The process of soil erosion is unavoidable because it happens naturally. This was reported by Baharuddin (1995) and Hartanto et al. (2003) where soil erosion still happens either in undisturbed forest or logging areas. Soil erosion process will remove clay, silt and sand through surface run-off thus change the soil texture. The effects of forest harvesting on soil texture was also reported by other studies. According to Makineci et al.,(2006) in his study in Belgrade, Turkey, the values of sand, silt and clay in the skid trail are 52.13, 25.55 and 20.32 % as compared to the undisturbed area with values of 48.84, 21.99 and 29.17%, respectively. Baharuddin (1995) reported that the soil loss was 13,300 kg/ha/year and 10,100 kg/ha/year for the logging road and skid trail compared to 500 kg/ha /year for the undisturbed area. However, studies in Sabah showed that surface erosion losses from harvesting areas are quite low at less than 200 kg/ha/year (Malmer and Grip,1990). According to Hartanto et al., (2003) the difference between logging system can only be significant for soil loss but not to runoff. This may due to the fact that runoff is strongly determined by rainfall.
The conditions of soil surface before harvesting phase are sheltered by tree canopy and litter layer. The tree canopy and litter on the soil surface acted as a buffer or filter to the direct rainfall. During forest harvesting, various activities which include feeder road and skid trail construction, cutting down trees and log disposal will initiate canopy opening and soil surface disturbance and this leads to soil erosion. Sidle et al., (2006) reported that logging road activities contributed the highest erosion compared to other land use. The higher percentage of sand during and after logging was probably due to the removal of fine fraction (silt and clay). Baharuddin (1995) believed that soil loss is highest on the logging road where the soil surface was exposed. This is because of the direct rainfall to the soil surface on the logging road without a buffer or filter compared to skid trail or undisturbed forest. Higher soil loss was detected during harvesting and it was reduced due to re-establishment of ground cover.
From t-test analysis, as shown in Table 4.1 also indicated that there was no significant difference detected among the sand, silt and clay between the two altitudes. This concludes that the elevation did not significantly affect soil texture for each harvesting phase. Figure 4.1 shows the soil texture in relation to elevation and harvesting phase.
The effects of forest harvesting on soil bulk density has already been reported by various other studies. Rab (1994) in his study in Eucalptus regnan forest, Southeastern Australia reported that bulk density increased by 33% to 65% compared to that in undisturbed forest. Merino et al. (1998) indicated that bulk density increased by 17% after whole tree harvesting in his study of Pinus radiate plantation in Northern Spain. In Malaysia, Baharuddin (1995) reported that forest harvesting that applied conventional logging in Tekam Forest Reserved Pahang produced a higher value in bulk density especially when involved in using skid trail and logging road. The values were reported to be 1.49 and 1.60 g/cm3, respectively.
Regarding the elevation analysis using t- test, the impact of forest harvesting using RIMBAKA technique on bulk density was quite similar. Both altitudinal levels showed no significant difference in bulk density value in each harvesting phase (Table 4.3). This is because the four main elements (slope class, the number of trees harvested, soil texture and method of harvesting) have a similar pattern for both altitudes. This scenario also occurred in other places where many researchers found that forest harvesting from different altitudes has no impact on bulk density. The increasing value in bulk density after forest harvesting very much depends on many factors such as soil type, soil movement, amount of timber removed, type of heavy machine used, movement of heavy machine in the logging areas, slope and site characteristics (Kamaruzaman, 1991; Baharuddin, 1995; Merino et al., 1998; Shukri et al., 1999; Andrian et al., 2005; Demir et al., 2006).
Water Quality Index
The site condition is an important factor that relates to river water quality status and the results are showed in Table 4.9. The statistical analysis showed that the undisturbed forest has better river water quality status compared to the logging and road crossing areas either during or after forest operations (P≤0.01). At the same time the weather conditions also play an important role in influencing the river water quality status for the three areas (undisturbed, logging and road crossing) (P≤0.01).
Based on the WQI values, the worst river water quality status was recorded at the road crossing area compared to the undisturbed and logging areas due to active usage of the area because the road has to be regularly maintained through surface grading and the surface is almost 100% exposed to direct rainfall.
Figure 4.6 shows the profile of mean WQI for different site conditions recorded from 21st November 2005 to 19th September 2006. The figure shows that the rivers in the logged and road crossing areas has lower quality status compared to undisturbed areas and the profile changes with seasons as indicated along x-axis.
According to Baharuddin (1995) surface runoff from forest road was nine times higher than the undisturbed forest and six times higher than the skid road. Besides that, rainfall distribution also influenced the turbidity value between those three locations and between phases of forest harvesting.
Figures 4.7 and 4.8 illustrate the trend of turbidity and TSS between phases of forest harvesting among the undisturbed forest, logging areas and road crossing. The value of turbidity were found higher at the road crossing area followed by the logging areas and undisturbed forest. The values of turbidity were low before harvesting but higher during harvesting. However, the values of turbidity decreased to normal value after harvesting. These results are similar with the studies conducted by Gomi et al. (2006) where the sediment in the river was related to land disturbance for the road and skid trail.