Seedling regeneration patterns and resource requirements are important determinants of long term forest dynamics. We measured all native tree seedlings within 192 1-m2 seedling subplots distributed on a 20x20-m grid system throughout the 4-ha Laupahoehoe Forest Dynamics Plot on Hawaii Island in which all adult trees ≥1-cm DBH have been identified and mapped.
Less than half of the tree species found in Laupahoehoe were found in seedling plots. About 35% of the regenerating seedlings were the canopy dominant Metrosideros polymorpha (Myrtaceae) and about 33% were the subcanopy dominant, Cheirodendron trigynum (Araliaceae), but less than 1% were the other major canopy tree, Acacia koa (Fabaceae). Across the species found in seedling plots, seedling abundance was positively correlated with adult tree abundance (R2=0.85), with lower seedling than adult tree abundance for five of the eight species. For the four species with most abundant seedlings we tested for correlation of seedling with adult tree abundance across 20x20-m quadrats. For M. polymorpha and Coprosma sp. (Rubiaceae), the number of seedlings correlated with the number of adults, but for C. trigynum and Vaccinium calycinum (Ericaceae), there was no relationship between number of seedlings and adults. We also measured canopy openness and recorded the substrate occupied by each seedling to determine the importance of micro-habitat on seedling regeneration. For all species, seedling abundance was not correlated with canopy openness, possibly reflecting the relatively open canopy and high light availability in this forest; 98% of quadrats had ≥10% canopy openness. However, seedling abundance was significantly related to substrate. Most seedlings were found on organic substrates (tree fern logs, tree logs, and root mats), with only one species found more often than expected on mineral soil. The absence or rarity of seedlings of several common species points to strong recruitment limitation. Overall patterns of seedling abundance and distribution appear to be affected by complex interactions between seed and microsite availability.