Abstract:Seedling recruitment is important for forest regeneration. During the seedling phase, individual plants are sensitive to a wide variety of habitat conditions. Seedling establishment depends not only on the specific species involved, but also on the site specific conditions where the plants grow. However, few studies have examined the ways in which habitat influences the survival and growth of seedlings. In 2003, a single 5 hm2 (250 m × 200 m) permanent plot was established in a subtropical evergreen broad-leaved forest in the Baishanzu mountainous region, Zhejiang Province, China (119°3'53"E, 27°40'54"N) to monitor the dynamics of the forest. To explore the dynamics of seedling survival as well as the correlation between the seedling mortality/germination and the habitats of various seedlings in the plot, 50 seed-seedling monitoring stations were established along the inner path of the plot in October 2008. Each seed-seedling monitor station consists of a seed collector (0.75 m × 0.75 m) and 3 seedling quadrats (1 m ×1 m), which were placed 2 m away from the seed collector. The quadrats were surveyed three times each year to tag, identify and quantify the emergence of new germinating seedlings and to track seedling mortality in these quadrats during May, August and October, from 2008 to 2011. During data analysis, we first calculated the germination rate, death rate and survival rate (the difference between the germination and death rates) of all seedlings present during the period. Then, the numbers of seedlings, newly germinated seedlings, and dead seedlings were classified based on slope position, slope degree, soil texture, soil moisture, and depth of leaf litter in the seedling stations by partial correlation. The results show that: 1) the seedlings that geminated between October, 2008 and October, 2011 in the Baishanzu plot belonged to 53 species, 40 genera, 26 families; 2) different species germinated at different times and the duration of the germination period during a year varied among these species with 11 species germinating during spring to autumn, five species germinating in spring and summer, five species germinating in summer and autumn, 13 species germinating only in spring, eight species germinating only in summer, one species germinating only in autumn, while the remaining ten species did not germinate during the study period; 3) the total number of seedlings declined from 428 in October 2008 to 364 in October 2011; also, the number of species present in the quadrats declined from 53 to 45 during this study; 4) the seedling survival rates were 7.7%, -20.8%, and -0.3% in 2009, 2010 and 2011, respectively; 5) dominant species in living seedlings present in the plot were Symplocos lucida, Camellia cuspidata, Neolitsea aurata var. undulatula, Smilax arisanensis and Lithocarpus brevicaudatus; these five species included > 50% of the total number of plants present in the quadrats; 6) the number of germinating seedlings in a particular seedling quadrat was correlated with the habitat type, which was itself determined by water availability at the site; 7) the number of surviving seedlings in a quadrat was significantly correlated to slope position, moisture content, and depth of leaf litter of the seedling quadrat. These results suggest that species identity, water availability and light were the major factors affecting seedling germination and survival. Species exhibit differences in regeneration niche during the seedling stage of growth.