GROWTH RHYTHMS OF THREE Ormosia SPECIES SEEDLINGS OF DIFFERENT PROVENANCES

The superior provenance is a prerequisite for ecological restoration, and a better mastery on the growth rhythms of Ormosia species is fundamental to reforest effectively. For the selection of better provenance and the formulation of artifi cial cultivation methods, the height and the ground diameter of Ormosia hosiei Hemsl. et Wils., O. xylocarpa Chun ex L. Chen and O. henryi Prain seedlings of different provenances were monitored in the fi rst year. The results showed that their dynamic growths presented a slow-fast-slow trend that fi t a “S” growth curve by the logistic mathematical model, and the growth of Ormosia species seedlings of different provenances signifi cantly differed. The accumulated growth increment of O. xylocarpa was the largest (averagely 45.50 cm) and the accumulated growth increment of O. henryi was the least (averagely 20.33 cm). Thus, O. hosiei of Jiujiang provenance, O. xylocarpa of Liping provenance and O. henryi of Longquan provenance have a stronger adaptability for future artifi cial cultivation in Jiangxi China.


RITMOS DE CRESCIMENTO DE TRÊS ESPÉCIES DE MUDAS Ormosia DE DIFERENTES ORIGENS
. With the development of economy and society, the three Ormosia species are endangered in some regions in China due to the excessive cutting and utilization without reforestation, which is extremely affecting our national ecological safety and wood safety (Deng et al., 2011;Liu et al., 2017;Wang et al., 2015).
Most of the Ormosia species possesses a long dormancy period prior to seed germination and their ability of natural regeneration is poor because of their seeds enwrapped hard and dense seed coat with the waxy layer and poor permeability. According to the literatures, O. hosiei, O. xylocarpa and O. henryi mainly distribute in Anhui, Jiangxi, Hubei, Guangdong, Guizhou, Zhejiang, Guangxi, Hunan, Sichuan, Yunnan, Fujian, Hainan, China (Chinese ethnographic editorial board, 1994). However, the wild resources have been cut down in large quantities, and the native habitats have been destroyed seriously. Now the natural forests containing these species are tapering off (Deng et al., 2011;Liu et al., 2017;Wang et al., 2015). It is diffi cult to recover the population by natural regeneration. So appropriate provenances were introduced in the original region where Ormosia populations were disappeared, which is the important method for the Ormosia populations restoration.
The shortage of forest resources has been forcing us to explore the ways to improve the forest quality of rare tree species. However, the systematic seedling cultivation techniques was scarce. Artifi cial cultivation is one of the important ways to rebuild low yield forests and restore forest ecology (Ghosh et al., 2014). The growth rhythms and suitable provenances of seedlings are basis of forest restoration. The growth rhythm of seeding is infl uenced by the external environment, which is the embodiments of its physiological function (Chen et al., 2014;Gyllenstrand et al., 2007;Mattos et al., 2005). The growth period could be divided by observing and studying the annual growth rhythm of Ormosia seedlings. The scientifi c cultivation measures were formulated for cultivating high quality and strong seedlings according to different growth stages (Zhou et al., 2016). Therefore, understanding of growth rhythms and selection of appropriate provenances of three Ormosia species are indispensable to develop effective reforestation strategies, protect biodiversity, and meet the needs of wood-processing industry and human diversity (Cai, 2007;Ye, 2014).
At present, most researches on Ormosia species are mainly concentrated on tissue culture (He et al., 2015;Fan et al., 2011), seed reproductive ecology (Peng et al., 2009;Chen, 2016), seed germination (Vargas-Simón et al., 2017;Silva et al., 2014;Foster, 2008), seedling cultivation (Shen et al., 2009;Ye, 2014;Yang, 2011;Feng et al., 2007;Zhu et al., 2015), fertilization measures (Duan et al., 2017;Zheng, 2008) and forest afforestation techniques (Wu et al., 2009;Pan, 2004), while few studies have been performed on the growth rhythm of O. hosiei, O. xylocarpa and O. henryi from different provenances. Hence, this study is aimed to evaluate the growth rhythms and differences of the three species from different provenances for selecting the superior provenances of O. hosiei, O. xylocarpa and O. henryi for priority use and determining appropriate management technique. Our fi ndings will facilitate the cultivation of good seeds, good quality seedlings and planting, and population restoration of these species. It also provides a theoretical basis for ecological restoration and development of timber resources.

Experiment material
The ripe seeds were collected from October to November in 2014, then they were stored in nylon mesh bags in shade conditions. O. hosiei seeds were collected from Jiujiang, Jiangxi Province and Lichuan, Hubei Province, the O. xylocarpa seeds were collected from Ruyuan, Guangdong Province and Liping, Guizhou Province, and the O. henryi seeds were collected from Longquan, Wuhan, Guilin, Hangzhou and Changsha that distributed in Zhejiang, Hubei, Guangxi, Zhejiang and Hunan Provinces (Table 1).

Experiment site
This study was conducted in the greenhouse of Jiangxi Academy of Forestry, Jiangxi Province, China (115°49′E, 28°44′N). The region is dominated by humid monsoon climate of subtropics with a mean annual temperature of 18.0°C and mean annual precipitation of 1600-1700 mm. The extreme maximum temperature is 40.1°C in summer and the extreme minimum temperature is -9.7°C in winter. The annual average sunshine time is 1723-1820 hours. The rainfall periods are 147 to 157 days.

Seedling cultivation
The experimental fi eld was established in December 2014. The seedbed with a width of 100 cm and a height of 40 cm was sterilized with carbendazim diluted 500 times. In March 2015, the seeds were soaked in water at 40 °C for 6 to 8 hours and seeds displaying obvious expansion were sowed a dibble method at a spacing of 5 cm × 10 cm (Shen and Zhai, 2001). They were cover with 1 cm of yellow soil. The yellow soil is the soil without stone and roots and taken from > 20 cm depth. Then the seedbeds were covered by an arched bamboo shed (overlying 50% shade net on arched bamboo strip) to keep the seedbeds moist and cooling and shading of growing environment. Weeding, irrigation, and pest control were implemented regularly, according to normal management production processes.

Measurements
After sowing, seedlings were observed every 2 days until they were more than 90% of the fi rst leaf of seedling unfolding, which the seedling stage was determined. Thirty seedlings of each provenance in stationary plots were studied. The height of seedlings was measured with a steel tape (accuracy of 0.1 cm), every 20 days (but adjusted to 10-30 days based on weather changes) till October and the last time in the end of December since they grew slowly in winter. As the growth of the ground diameter is small and the increase is not obvious, the ground diameter of the seedling stems was measured with a digital vernier caliper (accuracy of 0.01 cm) only when the height growths were almost stopped.

Statistical analysis
The logistic curve was simulated for the dynamic growth of the seedling height from different provenances (Dong, 2007). The fi tting equation of the logistic curve is: Where y is the growth increment of the seedling height, t is the growth time, a and b are the undetermined coefficients, k is the limit value of the seedling height growth under the given conditions. (Eq. 1)

Provenance LON (E) LAT (N) ALT (m) AMT (°C) AR (mm) FFP (d) EAT (°C) EIT (°C) MAH (%) SD (h) O. hosiei
Take the derivative of formula (1) for several times, and two infl ection points of the growth rate that is the fastest can be obtained (Kuang et al., 2014): In formula (2) and (3), t1 and t2 are the dividing points from the germination to the fast-growing stage and from the fast-growing stage to the slow growth stage respectively, and the period between t1 and t2 can be regarded as the fast-growing stage. The growth stages of various provenances were divided according to the growth data obtained.
The mean height of each measurement stage, the mean ground diameter of the last measurement and the corresponding standard error were calculated. We conducted the signifi cance test of differences on the average height and ground diameter of one-year-old seedlings of O. hosiei and O. xylocarpa by means of T test and of O. henryi by means of one-way ANOVA and LSD method. All statistical analyses were performed on SPSS 20.0.

Model simulation of seedling height growth rhythm of Ormosia species
Combining the seedling height data, the seedling height growth rhythms of Ormosia species of different provenances were simulated by the logistic model. The results exhibited during the growth period of Ormosia seedlings in the fi rst year, the dynamic height growths presented a slow-fast-slow trend and fi tted a "S" growth curve by the logistic mathematical model. The achieved highly signifi cant fi tting level ranged from 0.820 to 0.994 (Table 2).

Comparison of the height growth rhythm of O. hosiei of different provenances
The seedling growth period of Jiujiang provenance was from Apr. 16 to Dec. 30 for 257 days with 46.52 cm of the accumulated growth increment, and Lichuan provenance was from Apr. 7 to Nov. 1 for 208 days with 27.90 cm of the accumulated growth increment (Table 3). Because the initial growth rate was slow, the seedling height growth of both Jiujiang and Lichuan provenances were not different obviously, while the fast-growing stages of these two provenances were quite different ( Figure 1A). For the fast-growing stage of O. hosiei seedlings, the duration of Jiujiang provenance had lasted from June 25 to October 1 for 98 days, with 28.65 cm of the net height increment concomitantly and 40.19 cm of total accumulated increment ( Table 3). The O. hosiei seedlings of Lichuan provenance entered into the fast-growing stage was mid-May, with 17.56 cm of the net height increment concomitantly and 24.10 cm of total accumulated increment ( Table 3). The O. hosiei seedlings tended to grow slowly after October ( Figure  1A; Table 3). Lichuan provenance were inclined to grow gently and stop in November while the seedlings of Jiujiang provenance hardly grew at the end of December ( Figure 1A; Table 3).

Comparison of the height growth rhythm of O. xylocarpa of different provenances
It differed greatly in the growth of O. xylocarpa seedlings between Ruyuan and Liping provenances in the fi rst year. The seedlings of Liping provenance grew much better than those of Ruyuan provenance during the whole year ( Figure 1B)  that lasting 34 days, with 26.60 cm of a net height increment (Table 3). The duration of the fast-growing stage was from Jul. 19 to Sep. 18 for two months with 34.70 cm of a net height increment, and an extreme growth occurred in the end of August which had presented nearly 9.86 cm of increment in height for 10 days, then after Sep. 18, the seedlings grew slowly and stopped ( Figure 1B). Nevertheless, contrast to the seedlings of Liping provenance, the seedlings of Ruyuan provenance got a low-growing speed at the beginning with the 7.66 cm of accumulated increment before they entered into the fast-growing stage, which was preeminently lower than those of Liping provenance ( Figure 1B). The fast-growing stage could be identifi ed from Jun. 15 to Sep. 21 for 98 days along with 14.56 cm of a height growth increment ( Figure  1B; Table 3). The seedlings of the late growth stage were late September and terminated growth in the middle of December ( Figure 1B).

Comparison of the height growth rhythm of O. henryi of different provenances
For all investigated provenances of O. henryi, the growth rate of each stage of different provenance were inequable ( Figure 1C). At the early growth stage, the seedling height of Changsha provenance was the highest with 9.39 cm, followed by the Longquan provenance, Wuhan provenance, Guilin provenance and Hangzhou provenance with the 8.96 cm, 7.19 cm, 5.23 cm and 3.22 cm of height, respectively (table 3). The seedlings of Guilin provenance were the earliest ones who entered the fast-growing stage (Apr. 12), whereas those of Hangzhou provenance were on May 19, of Wuhan provenance and Longquan provenance were in the mid and late June respectively, and of Changsha provenance were the latest (Jul. 7) ( Table 3). The net height increment of Longquan provenance was the largest (16.26 cm), followed by that of Guilin (13.03 cm), Wuhan (12.32 cm), Hangzhou (10.75 cm) and Changsha (7.93 cm) provenance (Table 3).

Comparison of one-year-old seedling growth increment for Ormosia species
Signifi cant discrepancies in the height and the ground diameter of one-year-old O. hosiei seedlings of different provenances were recognized (P < 0.05). The heights of one-year-old O. hosiei seedlings of Lichuan provenance were in the range of 9.50 -33.80 cm, averaging 28.54 ± 1.15 cm, which was 16 cm Ruyuan provenance was from Apr. 28 to Dec. 12 for 228 days with 24.01 cm of the accumulated growth increment.
For O. xylocarpa seedling of Liping provenance, the early growth stage was from mid-June to July 18 shorter than those of Jiujiang provenance. To the contrary, the average seedling ground diameters of Lichuan provenance were 0.10 cm thicker than those of Jiujiang provenance (Table 4).
The heights and the ground diameters of one-yearold O. xylocarpa seedlings of different provenances were obviously different from each other (P < 0.05). Notably, the heights of one-year-old O. xylocarpa seedlings of Ruyuan provenance were correspondingly lower than those of Liping provenance for the former were in the range of 7.00 -29.30 cm, averaging 22.90 ± 1.03 cm, whereas the latter were 64.86 ± 2.30 cm averagely with the minimum value of 22.00 cm up to 87.00 cm. Besides, the ground diameters of one-yearold O. xylocarpa seedlings of Ruyuan provenance were ranging from 0.23 cm to 0.75 cm, 0.30 ± 0.01 cm averagely, signifi cantly less than those of Liping provenance (P < 0.05) whose the maximum value was 0.75 cm and the minimum value was 0.23 cm, 0.47 ± 0.03 cm averagely (Table 4).
The heights and the ground diameters of oneyear-old O. henryi seedlings of different provenances were signifi cantly different. The seedling heights of Wuhan, Hangzhou and Changsha provenance were 12.48 ± 0.22 cm, 11.40 ± 0.50 cm and 12.73 ± 1.03 cm respectively, which were apparently lower than those of Guilin and Longquan provenance (P < 0.05). The one-year-old O. henryi seedlings of Longquan provenance were the tallest (8.60 -31.50 cm) whose average value was 21.24 ± 3.09 cm and the ground diameter was 0.32 ± 0.04 cm averagely. However, the ground diameters of one-year-old O. henryi seedlings of Guilin provenance were markedly coarser than those of other provenances (P < 0.05) because the average ground diameter was 0.36 ± 0.03 cm, the maximum value was 0.62 cm and the minimum value was 0.32 cm (Table 4).

The seedling growth rhythm of Ormosia species
Through the model simulation and the actual monitoring, the date consistent with the growth stages were determined by calculating. We found that the stages of the height growth of Ormosia seedlings could be roughly divided into fi ve stages: emergence stage (before Jun.), early growth stage (Jun.), fastgrowing stage (from Jul. to Oct.), late growth stage (from Nov. to Dec.) and stagnation stage (after Jan.).
Among the fi ve stages, the height growth of Ormosia seedlings during the fi rst year presented a slow-fast-slow trend in the light of the data of net increment. It differed greatly in the start and stop time and the duration of every growth stage of different provenances, so did the corresponding growth increment. The fast-growing stage had the largest proportion on the net growth increment and the longer duration that up to a maximum of 202 days (  (biological, ecological and forest characteristics). The time of emergence and growth duration were directly affected by species, provenances, and planting sites (Chen et al., 2014;Tang, 2013;Peng et al., 2009). The growth rhythm of plants can self-regulate with the climate change, which indirectly refl ects the adaptability of plants.
Comparing all the Ormosia species experimented, it was demonstrated that the O. xylocarpa seedlings of Liping provenance had a biggest height growth rate in the fi rst year (more than 60 cm averagely), followed by O. hosiei seedlings of Jiujiang provenance (more than 40 cm averagely) and Lichuan provenance (nearly 30 cm averagely) (Figure 1; Table 3), which may be a cause of high heritability on different species (Sudan et al., 2018) and genetic diversity on different provenances of the same species (Zhao et al., 2008;Zhang et al., 2012). Besides, O. henryi seedlings refl ected to be the shortest (most of them were between 15-30 cm) and grow the most slowly among three Ormosia species (Figure 1) which corresponds with other studies (Cai, 2007;Chen et al., 2014;Tang, 2013). Among these Ormosia species, the O. xylocarpa seedlings of Liping provenance grew the most optimally (Figure 1). O. hosiei of Jiujiang provenance and O. henryi of Longquan provenance are also the best choice for artifi cial cultivation in future. Different provenances have different physiological characteristics, which will inevitably affect their growth.
Considerable differences on the provenances of the same species accompanying the geographic and climatic variables lead to different growth states (Lukkarinen et al., 2009). Different geographical conditions have different climate, so the variation tree species have different tolerance to Jiangxi climate. Extreme climate change (for instance extreme maximum temperature showed up, air humidity increased sharply) may produce some morphological variation of forest trees (García-Plazaola and Becerril, 2000;Gratani et al., 2003;Kaleem et al., 2010), and change their photoperiod and photosynthesis (Kalita and Titlyanov, 2013;Yuan et al., 2017). It is still worth to explore further what are the main factors leading to growth differences among different provenances: genetic factors or environmental factors?

Cultivation Measures
According to their growth rhythm, it is effective to take cultivation measures timely characterized the growth of different stages (Fan et al., 2013;Santelices et al., 2015), so the cultivation procedure should be implemented before the end of Spring by adopting conservation measures such as covering the seedling shed and irrigating adequate water regularly etc. (Jiang and Xiaoli, 2016;Luis et al., 2004). The mixed fertilizer with low nitrogen and high phosphorus should be applied properly at the end of June to promote the growth of roots laying the foundation for the fastgrowing stage (Huang et al., 2015). What should be carried out in the key fast-growing stage from July to September is to spray the mixed fertilizer of high nitrogen, low potassium and low phosphorus, irrigate fully and weed comprehensively to maximize the seedling growth rates. After October, all management measures to facilitate the seedling growth should be stopped except to promote the lignifi cation and prevent frostbite as the seedling growth is stepping into the sclerosis period.

CONCLUSIONS
The dynamic height growths of one-year old Ormosia seedlings presented a slow-fast-slow trend that fi t a "S" growth curve by the logistic mathematical model (highly signifi cant fi tting level ranged from 0.820 to 0.994). Five stages were broadly divided: emergence stage (from Apr. to May.), early growth stage (Jun.), fast-growing stage (from Jul. to Oct.), late growth stage (from Nov. to Dec.) and stagnation stage (after Jan.). From the growth of the fi rst year, O. xylocarpa seedlings of Liping provenance, O. hosiei seedlings of Jiujiang provenance and O. henryi of Longquan provenance are the most suitable to be cultivated in Nanchang, Jiangxi by artifi cial afforestation. However, the growth of seedlings will be affected by many factors, such as water and fertilizer management, need to be further studied in the future.

ACKNOWLEDGMENTS
This study was supported by the Key Research Program of Jiangxi Academy of Forestry Sciences (2017511201) and the Key Research and Development Program of Jiangxi province (20161BBH80065). Thanks are due to Shaoping Hu for wholeheartedly aids in investigation process, to Jinshan Ye for valuable help of collecting data, to Huamei Lin for assistance with the nursery management and to Professor Shengquan Que for protection from diseases and insect pests.