The criteria for the site selection was that a negative change of forest cover was found in the RSS and that it had taken place during the period between the acquisition dates of the T2 (1991) and T3 (1998). The study areas were set to be 1 km x 1 km. Three areas were selected, and their centre points were defined as points on the boundaries of the change polygons.
Figure 1. Map of the study areas.
The fieldwork was undertaken in Petchabun, Loei and Phitsanulok Provinces of the mid-north of Thailand. For technical assistance, local knowledge and for translation, Letters of Agreement on collaboration were made with RECOFTC and the Forest Resources Assessment Division of the RFD. RFD and RECOFTC provided the aerial photos for detailed area analysis. The RFD provided analogue satellite image dated January 2000. The geo-referenced digital T3 image was purchased from the National Research Council of Thailand.
Co-ordinates were first retrieved from the geo-referenced satellite image. They were later corrected in the field after a local adjustment of the geographic registration of the image based on GPS readings. It is important to note that although the co-ordinates changed in the course of preparation due to improvement of image registration, the actual location of the sites remained fixed.
The centre points were stored in a laptop computer and in the GPS receiver as waypoints (co-ordinates that the receiver can use for navigation as destinations or part of routes). Since the projection of the image was UTM (Universal Transverse Mercator), this was also the projection used in the GPS receiver.
Table 1. List of equipment used
GPS (Garmin 12 CX) |
Camera (Kodak DC290 digital zoom) |
Analogue cameras |
Laptop with navigation software, GIS and a digital satellite image (T3) |
Aerial photographs (recent 1995/96 and historic 1975/82) |
Topographic maps (1:50,000 & 1:250,000) |
Forest survey equipment (clinometer, calliper, loggers tape) |
Before going to the sites, the closest village was identified on a 1:50.000 scale topographic map. This map was then used to reach the village. When navigating to the centre point in the study area, both GPS and aerial photographs were used. The final location of the site was determined by GPS alone.
Fifty systematic sample plots (circular, 10 m in radius representing a sampling intensity of 1,6%) were laid out for measurement at the first site. The diameter of all trees (more than 10-cm dbh), height of the tree with the largest DBH on the plot and diameters of stumps were measured. Notes were taken on each plot on availability of wood supply, the land cover class, observations on felling and removals, disturbance, species group and erosion risk (see FRA 2000 Working Paper 1 for definitions). Twenty-five of these plots were surveyed in a systematic grid covering half of the 1 km x 1-km study area (fig.2). Measurements of trees and photo-documentation were made on every second plot.
Initially, the plots were demarcated by use of measurement tape and compass. In order to speed up the survey this methodology was replaced with a combination of compass and GPS, where the GPS was used for measuring the distance from the first plot in the row.
This approach was time consuming, although part of the plot was on agricultural land. Heavy rain became a limiting factor to carry out the planned measurements. Therefore, in the following study area, fieldwork consisted of visual observations and photographic documentation.
Figure 2. Survey pattern of the study areas. Only part of the first study area (red lines) were actually surveyed, due to constraints.
In the first study area, Ban Na Por Song in Phetchabun Province, the remaining part was covered by a traverse from the southern to the northern boundary. Study area two, Ban Khok in Loei Province, was covered by going to the centre point, while in the third study area, Ban Nam Chuan in Phitsanulok Province, it was only possible to reach the southern part of the study area.
At all sites, there were villagers accompanying the field team, advising on road or path selection as well as species and forest use based on their local knowledge.
Predominantly digital, but also analogue photos were taken. For the digital photos, notes on time, GPS-position and picture-file order were made. Photos were taken in four directions, always starting heading North and turning 90 degrees clockwise. Photos were also taken in the village, during interview sessions.
In order to address the socio-economic situation affecting the management of the land and to investigate the causes of recent changes and to provide an outlook for the future, qualitative research methods were applied. Open-ended key questions were prepared for interviewing focal groups. Rapid Rural Appraisal (RRA) tools, such as resource mapping, transect walk and activity schedules were used. Key informants were identified predominantly by the village headman. The interviews tended to be in small groups. Aerial photographs were used to discuss forest resources, and a community resource map was made only at the second site. Group discussions with women were held in all the sites.
Figure 3. Villagers studying an aerial photograph (Ban Kok)
The transect-walk consisted in walking to the centre point of the study area accompanied by villagers. The walk provided a good setting for addressing issues of tenure, vegetation, agricultural practises, change over time and future prospects. Several discussions went on simultaneously; however, the information was later cross-checked.
