Browsing by Author "Gaudel, Rabins"

Canopy gaps and their characteristic features (e.g. area and shape) influence the availability of nutrients, moisture and light in a forest ecosystem, and consequently affect the regeneration process and species composition in the forest. Most of the earlier research on canopy gap used field measurement and conventional remote sensing to quantify gap and these methods have limitations and accuracy problems. However, the development in Light Detecting and Ranging (LiDAR) technology has been effective in overcoming limitations and challenges associated with conventional remote sensing. The ability of LiDAR to represent the three-dimensional structure of the canopies and the sub-canopy resulting in high-resolution topographic maps, highly accurate estimated of vegetation height, cover and canopy structure makes it suitable technology for gap studies. LiDAR-based digital surface model (DSM) and digital elevation model (DEM) were used to quantify the canopy gaps over 5124ha of University of Tokyo Chichibu Forests (UTCF) consisting of three forest-types; primary, secondary and plantation forest. Disturbance driven canopy gaps might have spatial and characteristic variation due to differences in disturbance history, nature, frequency and intensity in different forest and land-types. Quantifying gap characteristics and studying variation and size distribution in different forest types and topography help to understand the different gap dynamics and their ecological perspectives. In this study, a gap was defined as an opening with a maximum height of 2m and minimum area threshold of 10m2. The minimum area threshold, which represents the gap area created by the death of at least a single tree, was determined through a random sampling of 100 tree crowns at UTCF using high resolution aerial photographs. Gap size distribution was analyzed in different forest types and land types. Spatial autocorrelation of gap occurrence was studied using semivariance analysis and distance to the nearest gap (DNG), which is the distance to the nearest gap for an individual gap. Canopy gap size frequency distribution in different forest-types was investigated using power-law. The negative exponent (α), which is also the scaling component of the power-law distribution, was compared between forest-types. Altogether, 6179 gaps with area 10-11603 m2 were found. Gap size distribution in UTCF showed skewness with a high frequency of smaller gaps and a few large gaps. Half of the gaps were smaller than 19 m2 and less than one percent of gaps (0.73 %) were larger than 400 m2. Primary forest contained high gap density (1.85 gaps per ha), shortest mean-DNG (22m) and second-largest gap-area fraction (0.72 %) after plantation forest area (0.76 %). Secondary forest had the lowest gap density (1.03 gaps per hectare) but had the larger mean gap-area (43 m2) than in primary forest (39 m2). The Kolmogorov–Smirnov test showed differences (p<0.05) in gap size distribution between primary and secondary forest. However, the gap size distribution in primary forest show similarity (p=0.59) with plantation forest area. In primary and plantation forest there was a high frequency of small gaps and few very large gaps (2000-10500 m2), whereas very large gaps (>2400 m2) were absent in the secondary forest. Gap size frequency distribution followed a power-law distribution only in plantation forest area (p>0.1, α =2.27). The scaling parameter in the primary and secondary forest was 2.56 (p=0.01) and 2.20 (p=0.02), respectively. Gap distribution showed some spatial autocorrelation in primary and secondary forest at least with distance up to 1300m. Most of the gaps in the primary forest were concentrated in the valley and middle slope, whereas the upper and middle slope had fewest gaps.