Effects of gap size and within-gap position on seedling growth and biomass allocation: Is the gap partitioning hypothesis applicable to the temperate secondary forest ecosystems in Northeast China?

Forest gaps can change stand structure and affect forest regeneration dynamics. The gap partitioning hypothesis claims that tree species would regenerate along gap environmental gradients owing to their different resource demands. However, many studies that examined this hypothesis in uncontrolled gap conditions have yielded inconsistent conclusions and there are limited studies testing the hypothesis taking into account biomass allocation. In this study, we examined the gap partitioning hypothesis in a temperate secondary forest ecosystem where gap formation drives forest regeneration. We created gaps with different sizes and planted two commercially important native tree species with contrasting shade tolerance in nine positions along four cardinal directions within the gaps. We found that shade tolerance was the main factor affecting seedling regeneration performance within gaps. For the light-demanding Manchurian walnut (Juglans mandshurica Maxim.), seedling growth varied significantly along the light intensity gradients, which indicated that seedlings could greatly benefit from high light areas within gaps, especially when the gap size increased. Seedling biomass in gap centers and transitions (54.0 g) were much higher than in gap edges (13.7 g) and forest understories (8.6 g). High light environments contributed to higher biomass allocation to the leaves and accelerated carbon assimilation. Low light conditions resulted in increasing proportions of stem biomass, which might promote seedling height growth, although the promotion effects were relatively limited. However, for the shade-tolerant Korean spruce (Picea koraiensis Nakai), little evidence of seedling divergence was found within gaps. Mostly, seedlings only showed growth and biomass allocation differences between gaps and forest understories. Korean spruce showed high adaptability to various gap environments and might be a generalist species rather than being limited to small gaps. The performance of Manchurian walnut strongly supported the gap partitioning hypothesis, but Korean spruce provided little evidence for gap partitioning. Therefore, planting scenarios in silvicultural practices could be suitably designed, with Manchurian walnut in high light gap areas and Korean spruce in most positions within gaps, during the early stage after competition elimination.