Back to index of forest structure and botany studies

Back to index of 2002 research reports

THE EFFECTS OF A ROAD ON MICROCLIMATIC AND VEGETATION STRUCTURE INDICES IN KAKENAUWE FOREST

 

Aim  

To establish the effect that the presence of the road bordering Kakenauwe forest has had on the forest structure and various microclimate elements. 

 

Objectives

  1. To measure the relative humidity; soil, ground and air temperature; wind speed and to assess the ‘gappiness’ of the canopy with increasing distance from the road.
  1. Record the presence and absence of light and moisture indicator species with distance from the road.
  1. To assess any change in forest structure with distance from the road.
  1. To compare the results of the Kakenauwe observations to the control, undisturbed forest, Lambusango.

Introduction

The underlying bedrock of the Kakenauwe reserve is limestone karst.  The rocks are jagged and pitted.  The soil is thin, probably due to the steep topography.  The soil on the plateau appears to be nutrient poor, it may experience a lot of leaching due to the well drained nature of the bedrock.  The soil is clayey so should have a good water retention capacity. 

The forest is a lowland forest.  The typical characteristics of a lowland forest are three layers to the structure, an emergent layer (40-50m high), a second layer and a third, shade tolerant layer.  The plateau displays similar characteristics to this but does not have three layers to its structure.  The trees are straight and there is a stable crown structure.  Typically, a high altitude forest has a more open canopy, with just 2 layers.  Most of the Kakenauwe forest has a structure resembling that of a high altitude forest.  This is probably due to the thin layer of soil and the disturbance that the forest has experienced.

Several hypotheses are put forward; 

At Kakenauwe NR;

1)   Wind speed, soil, ground and air temperature and gappiness will decrease with distance from the road.  Relative humidity will increase.

2)      Light indicating species will be more frequent closer to the road.  Moisture indicating species will be more frequent with increasing distance from the road.

3)      The structure of the forest should make a transition from gappy canopy with dense understorey to a more typical three layered rainforest structure with a low ground cover, a sub-canopy layer and an emergent layer.

At Lambusango

4)      The forest microclimate and forest structure will remain more or less constant with distance from the same elevation.

 

Methodology 

Both the Kakenauwe and Lambusango rainforests have a 1000m2 grid set up with grid lines spaced at 1 every 100m, dividing the grid up into squares of approximately 100m2.  

Three transects were randomly selected in Kakenauwe, W1, W2 and W4.  The three transects at Lambusango were selected by assessing the topography along each transect and choosing those most similar to the transects in Kakenauwe.  It is for this reason that the Kakenauwe transects go north and the Lambusango transects go west.  Along each transect the following microclimate measurements were made at intervals of 0, 5, 10, 15, 20, 25, 30, 40, 50, 70, 100m and then every 50m up to 900m from an elevation of   (the altitude of the road). 

Using a whirling hygrometer held at 1.5m, the relative humidity was determined using dry and wet bulb temperatures and a slide rule.  The hygrometer was whirled for one minute before the reading was taken.  The wind speed was measured at 0.5m and 1.5m simultaneously using anemometers. The anemometers were held at arm’s length facing into the wind. The wind direction was also noted using a compass. The ground and air temperatures were measured by placing a digital probe thermometer on the ground and at 1.5m. 

Along each transect, quadrats of 10m2 were set out using a tape measure.  The bottom left hand corner of the quadrat was placed 5m in from the trails north-east of the trail intersections.  The quadrats were placed at 5, 50, 100, and then every 100 metres up to 900 m from elevation. In each quadrat, cover values were estimated for plants <1m, 1-5m, 5-20m, 20+ using the Braun Blanquet scale. (see table 1)  The presence of pioneer species such as Macaranga was noted.  The presence or absence of the following light indicators: grasses <1m, ginger >2m, ferns >1m, lianes and exotics and moisture indicators; filmy ferns, epiphylls, the number of higher plant canopy epiphytes present, the moss line (height to which moss occurred on tree boles) and the height of bole climbers on trees >30cm circumference was also noted.  The circumference of all trees >30cm were measured at breast height.  The height of the tallest tree was also estimated.  The disturbance level of the forest was measured by noting if rattan, fallen trees, stumps, paths, dykes were present.  The level of disturbance depended on how many of the above features were present.

 

Table 1. The Braun-Blanquet Scale

Class = %cover

 

1=<5%

2= 5-25%

3= 25-50%

5= 75-100%

4= 50-75%

 

A soil sample was also taken from the centre of each 10m2 quadrat and the temperature of the soil was taken using the digital thermometer (at a depth of approximately 5cm).  A soil core was taken using a crude metal corer.  The soil was weighed on the day of collection to determine its bulk density.  The soil was then tested for its water content by weighing out approximately 100g of soil, placing it in an oven and reweighing the soil at intervals until there was no longer a change in weight.  Marked changes in slope, such as a cliff, were present along transects W1, W4 (Kakenauwe) and N7.(Lambusanga).  Microclimate measurements were made at the foot of the breaks in slope and at the top.

 

Initial results

Overall, it was clear that the road has a discernible impact on forest structure at Kakenauwe NR, with decreased humidity, increased temperatures and disturbed forest canopy close to the road. The effect appeared to stretch approximately 400m into the forest. No such obvious trends were observed at Lambusango and other trends were due to topography only.

 

Microclimate 

Ø      wind direction, wind speed

There was little wind movement during the sampling period but where wind was measurable, it tended to be evident at cliff faces or steep rises in topography at both sites. This has implications for natural windfall of trees which is very common at both sites. 

Ø      relative humidity

Relative humidity at Kakenauwe increased away from the road and varied from 25% close to the road to 50% within the forest. At Lambusango, RH was much higher overall, ranging from 40% to 80% and showed no discernible trends except to reduce slightly with altitude within the grid. 

Ø      temperature, soil, ground and air

At Kakenauwe, all temperature measurements decreased away from the road, while there were no discernible differences at Lambusango. 

Ø      Soil moisture

Ø      Soil moisture increased with distance from the road, reflecting RH values at Kakenauwe. No trends were observed at Lambusango.

 

Vegetation Cover 

Due to disturbance and increased light levels close to the road, the cover of shrubs was higher there than inside the forest. Canopy closure was greater in the interior shown by the decrease in gappiness and sky cover away from the road and increased shade. This led to lower light levels and less scrub development and better forest ground flora development. No trends were observed at Lambusanga, although light levels are higher there naturally.

 

Moisture indicators

Also notable was the increase in all the moisture indicators above 400m from the road at Kakenauwe. At Lambusanga the moisture indicators were evenly distributed throughout the sample area.

 

Report

A dissertation entitled The effects of the road on the Kakenauwe Nature Reserve, Buton Island, South East Sulawesi, with reference to microclimatic and vegetation indices will be produced by Sheona McKay, University of Sheffield by May 2003.