Ecological Drivers of Natural Forest Regeneration in a Tropical Moist Forest Protected Area of Bangladesh

Technical Summary: Ecological Drivers of Natural Forest Regeneration in a Tropical Moist Forest Protected Area of Bangladesh

Problem Statement
Natural regeneration is recognized as the most cost-effective and practical method for forest recovery, essential for maintaining ecological balance and biodiversity in tropical moist forests. However, the dynamics of natural regeneration are heavily influenced by site-specific habitat factors, and the transition from seedling to sapling stages is often hindered by high mortality rates and competition. While natural regeneration is preferred over artificial planting for its ability to ensure genetic diversity and resilience, there is a scarcity of research regarding the specific ecological drivers regulating regeneration in the Tilaghor Eco-Park, a protected area in the Sylhet division of Bangladesh. This study addresses the gap in understanding how micro-environmental factors and overstory structural attributes influence the establishment and composition of regenerating vegetation in this urban-adjacent protected area.

Methodology
The study was conducted in the Tilaghor Eco-Park, a 45.3-hectare protected area within the Indo-Burma Biodiversity hotspot. The research employed a systematic sampling design involving 25 main plots (10m × 10m) spaced at least 200 meters apart. Within each main plot, four subplots (2m × 2m) were established to assess regeneration, resulting in a total of 100 regeneration plots.

Data collection focused on three distinct life stages:

• Trees: Defined as individuals >5m height and >5cm DBH.
• Saplings: Defined as 1.3m < Height ≤ 5m and 2cm < DBH ≤ 5cm.
• Seedlings: Defined as ≤1.3m height and ≤2cm DBH.

Phytosociological attributes, including species richness, abundance, density, frequency, and Important Value Index (IVI), were calculated. Environmental variables measured included elevation (via GPS), canopy coverage (via spherical crown densitometer), canopy height, tree height, and basal area.

Statistical analysis was performed using R (4.2.2), PAST 4.10, and Excel. The study utilized:

• Diversity Indices: Simpson, Shannon-Wiener, and Evenness indices to quantify species diversity.
• Non-metric Multidimensional Scaling (NMDS): To visualize community complexity and rank factors based on similarities and differences.
• Generalized Linear Models (GLM): To determine the specific effects of ecological variables on seedling richness and abundance.
• Canonical Correspondence Analysis (CCA): To correlate seedling abundance with significant habitat variables identified in previous analyses.

Key Results

• Species Composition: A total of 314 trees (21 species), 102 saplings (11 species), and 935 seedlings (10 species) were recorded. Dipterocarpus turbinatus (Garjan) was the most dominant and abundant species in both the tree and seedling strata. Conversely, Calamus longisetus (Bet) was the most dominant sapling species, accounting for 75 of the 102 recorded saplings.
• Diversity Indices: The Simpson diversity index indicated high diversity for trees (0.86), moderate diversity for seedlings (0.68), and lower diversity for saplings (0.45). The Shannon diversity index similarly reflected high tree diversity (2.27) compared to moderate seedling (1.37) and lower sapling (1.08) diversity.
• Ecological Drivers (NMDS): The analysis identified tree height and tree richness as the primary drivers of seedling community composition. Tree height showed the strongest effect ($r^2=0.30, p=0.020$), followed by tree richness ($r^2=0.29, p=0.035$). Factors such as elevation, basal area, and canopy coverage showed non-significant trends in this ordination.
• Ecological Drivers (GLM):
  • Seedling Richness: Significantly influenced by tree richness, tree abundance, sapling richness, and sapling abundance.
  • Seedling Abundance: Influenced by elevation, tree diameter, and basal area, though these relationships were marginally significant after reducing collinearity.
• Structural Collinearity: Canonical Correspondence Analysis (CCA) revealed a strong structural collinearity between richness and abundance vectors in both tree and sapling strata, suggesting that as species richness increases, individual density accumulates proportionally without shifting community composition significantly.

Significance and Claims
The paper claims that the study provides crucial insights into the current status of natural regeneration in Tilaghor Eco-Park, identifying canopy tree height and overstory species richness as the primary determinants shaping seedling community composition. The authors highlight that the extreme dominance of Dipterocarpus turbinatus in the seedling layer, combined with the monopolization of the sapling layer by Calamus longisetus, signals a potential long-term decline in species diversity and density-dependent suppression.

The study asserts that these findings are vital for guiding area-specific forest management, particularly given the park's vulnerability to urban edge effects. The authors conclude that to ensure sustainable forest recovery, management strategies must shift toward target-driven conservation, including the preservation of mature overstory trees to maintain microclimatic stability and the implementation of Assisted Natural Regeneration (ANR) to introduce under-represented indigenous species. The paper positions itself as a valuable addition to the literature on natural regeneration regulating factors in South Asian tropical moist forests, filling a knowledge gap for protected areas near urban centers.