Here we present the result from a two-year field experiment in which we transplanted seedlings of 14 woody species representing three contrasting functional types(broad-leaved evergreen, deciduous and coniferous species) along a light gradient into the understory of a subtropical forest in southwestern China. We assessed mortality and measured the growth rate of all transplanted individuals at intervals. To further understand the mechanisms leading to seedling mortality and growth, we examined three functional traits of transplanted individuals (SLA, leaf mass ratio, wood density).

Fifteen shade houses were arranged in a 3 × 5 grid with a 2 m distance between shade houses. These shade houses represented three light levels with five houses as replicates for each level. The light levels were created by covering shade houses with layers of neutral shade netting with different density. The shade houses had a height of 2.2 m and an area of 4 × 5 m. Instantaneous PAR (Photosynthetically Active Radiation) readings were made both outside and inside each shade house using an SKP 215 PAR Quantum sensor. These measurements indicated that the plants were receiving 43.7±2.1%, 17.1±0.73%, 2.74±0.1% of full daylight in the high-, medium-, and low-light shade houses. The light condition gives a good simulation to the natural light gradient in the forest as 50% is usually encountered in open areas and 3% in understory of subtropical forests in the region (observations by the author). To avoid environmental heterogeneity affecting experiment results, we applied a random-block design, i.e. each of the three light levels occurred once in each block of three houses and was positioned in a random sequence. The fourth light level (full daylight) was performed outside of the shade houses. We marked 5 quadrates next to each other besides the area of the shade houses, with at least a distance of 1 m between quadrates.Air temperature and soil temperatures were measured with a Hobo U12-015 temperature data logger. We hung one thermometer in the air inside one of the shade houses and buried another thermometer 10 cm below the top soil in the same shade house. Measurements were taken every hour from day 58 until the end of the experiment, and mean daily temperatures were calculated. Seeds were collected from under parent trees in subtropical forests in the Dujiangyan region in spring 2007. All seeds were processed on the day of collection.Seeds were then germinated under a shade cloth in a nursery. In August 2007, the seedlings were individually transplanted into plastic plots (30 cm height, 30 cm diameter) filled with farmland top soil. Two seedlings from each of the 14 species were placed into each shade house and in each full-light quadrate, yielding a total of 560 pots with a single test plant each. The pots were arranged in a hexagonal pattern with a distance of 0.5 m between neighboring pots (0.8 m between the centers of neighboring pots) to avoid shading. Two weeks after moving pots to shade houses, we replaced dead seedlings with ones of the same species sown at the same time. Plants were watered every 3–4 days or when the soil was dry.The traits included the three growth components net assimilation rate (NAR), specific leaf area (SLA) and leaf mass ratio (LMR) as well as other gas-exchange traits

The experiment was carried out from August 2007 to July 2009 in an experimental garden near Dujiangyan, southwest China (31°04’ N, 103°43’ E). The climate in the region is characterized by a mean annual temperature of 15.2°C and an average July temperature of 25°C. The mean annual precipitation at Dujiangyan is 1341 mm with a dry season in winter (November–April) and with warm and rainy summers (May–October). The annual average relative humidity is above 80%. The vegetation in the region can be classified as a secondary conifer and broad-leaved mixed subtropical forest.

August 2007 to July 2009

The 14 species used in this study are all woody species commonly found in the study region. These species were selected according to their leaf habit, successional status and seedling availability. There were seven broadleaved-evergreen and seven broadleaved-deciduous species. The deciduous species are either early-successional (Alangium chinense, Castanea henryi and Rhus punjabensis) or intermediate. In evergreen species, Pyracantha fortuneana and Lindera communis are early-successional, Phoebe zhennan is late-successional and the rest are intermediate.