Xiuli Chu·Xiuhua Wang·Dongbei Zhang·Xiaolin Wu·Zhichun Zhou

Abstract Seedling quality is important for subtropical tree species endangered by the degradation of natural habitats in southern China. At present, the cultural regime for raising these seedlings involving fertilizer levels and size of container is not clear. In this study, seedlings of three endangered species, red-seed tree ( Ormosia hosiei), Zhejiang phoebe ( Phoebe chekiangensis), and Zhejiang camphor ( Cinnamomum japonicum) were evaluated along with red-bark oak ( Cyclobalanopsis gilva) as a reference, a species which is not endangered. Seedlings were raised in 2.8,3.6, 5.1, and 6.3 L containers and fertilizer applied at 1.0,2.0, 3.0 and 4.0 kg m ?3 . Seedling height and leaf biomass increased in response to higher fertilizer levels while larger containers resulted in greater stem and root biomass. Root biomass of endangered species seedlings did not respond to neither treatments. Zhejiang phoebe seedlings responded to nitrogen and phosphorus uptake but red-seed tree seedlings were unaffected by any nutrient levels. Red-bark oak seedlings had high nitrogen-use efficiency. Based on the results,it is recommended using at least 5.1 L containers to culture Zhejiang phoebe and Zhejiang camphor seedlings with fertilizer at 3.0 kg m? 3 . Red-bark oak and red-seed tree seedlings should be cultured with 2.0 kg m? 3 in smaller containers.

Keywords Anthropogenic activities·Climate change·Container type·Vulnerable species

Introduction

The global population has increased over past decades(United Nations 2015, 2017), which has led to forest degradation directly through over-exploitation and indirectly through anthropogenic-driven climate change (Frournier et al. 2017; Wan et al. 2018). Endangered species are vulnerable and their numbers and distribution have declined significantly. It is difficult for these species to persist in forest communities through natural regeneration. Therefore,considerable effort has been made to improve quality and richness of forests by restoring endangered species.

Planting species in a specific environment may achieve rapid restoration of degraded forests (Chirino et al. 2008;Pinto et al. 2011; Pokharel and Chang 2016). However,seedling quality is important for successful establishment.In subtropical regions, seedlings with vigorous growth and large size are desirable because they generally have a higher survival rate (Guo et al. 2016; Tian et al. 2017).Fertilization is a common practice during the nursery stage to improve seedling vigour. Nitrogen (N) plays a central role in protein synthesis, enzymatic activities and in photosynthetic processes (Luo et al. 2013; Hu et al. 2014,2019). Phosphorus (P) is an important element that limits physiological metabolism and reproduction during growth(Warren 2011). Therefore, N and P uptake and utilization have been identified as key issues in studies of seedling quality (Mariotti et al. 2015a; Zhu et al. 2016; Wang et al.2017; Wei et al. 2017; Zhao et al. 2017). For subtropical tree seedlings, fertilization over a range of dosages may induce different growth responses (Wei et al. 2013). Seedlings generally show nutritional symptoms of deficiency,loading, optimum uptake and excess in response to an increase of fertilizer (Salifu and Timmer 2003). Inherent nutrient reserves can be highly efficiently established in the luxury nutrient consumption status.

Increased nutrient reserves may improve seedling performance by enhancing nutrient retranslocation (Oliet et al. 2013; Pokharel and Chang 2016). In recent years,seedling quality dimensions regarding nutrient uptake and utilization have received greater attention in subtropical regions (Wang et al. 2017; Li et al. 2017, 2018). High temperatures and humidities stimulate nutrient consumption and abundant nutrient reserves promise successful transplanting. An appropriate fertilizer regime can favor nutrient uptake and utilization in some subtropical species (Li et al. 2017). Because of various strategies of carbon assimilation and allocation among different species in high temperature forests (Dos Santos et al. 2006),the response by different subtropical species to different nutrients may be species-specific, at least among species of Dalbergia odorifera (Li et al. 2018), Podocarpus macrophyllus (Thunb.) Sweet (Wang et al. 2017), and Acacia koa A. Gary (Dumroese et al. 2 011). Tree species in subtropical regions are highly variable in habitat, growth rates, and survival strategies. However, current studies are limited by the number of species that have been examined (Zhu et al. 2016; Wang et al. 2017; Li et al. 2017,2018). More research is needed to determine the optimal fertilizer regime across most subtropical endangered tree species.

Container type may determine optimum nursery stock production (Thomas et al. 2016; Tian et al. 2017). The appropriate container type would match the size of the plants growing in them (Poorter et al. 2012). Containers with larger volumes generally resulted in greater biomass of seedlings for afforestation (Pinto et al. 2011;Mariotti et al. 2015a; Thomas et al. 2 016). The container depth may contribute more to the formation of root system than the pot size. Elongated containers can promote root biomass allocation and prevent taproot deformation,enabling the production of deep-rooted stocks (Kostopoulou et al. 2011; Thomas et al. 2016; De La Fuente et al.2017). The inner diameter of a specific container had a species-specific effect on shoot growth of Mediterranean tree seedlings (Mariotti et al. 2 015a; Salto et al. 2016) but the effect of diameter on seedlings in other biomes is not clear. Generally, seedlings grown in larger containers have better transplant performance (Chirino et al. 2008; Close et al. 2010; Haywood et al. 2011; Pinto et al. 2011; Mariotti et al. 2015b). Recent evidence indicates that container type may also affect the quality of some endangered tree seedlings in the subtropics. Larger containers resulted in greater biomass and larger shoot growth (Zhu et al. 2016;Tian et al. 2017). These results suggest a possible interaction of container type and fertilizer regime to promote nutrient uptake and utilization in subtropical endangered tree seedlings but evidence is scarce.

In this study, the growth and quality of subtropical seedlings were evaluated by four concentrations of fertilizer and four sizes of container. It was hypothesized that nutrient uptake and utilization would increase with higher (1)fertilizer dosages and (2) container volumes but (3) their combined effect would be species-specific.

Materials and methods

Study site

This study was carried out in a greenhouse on the experimental forest farm (27°38′N, 119°01′E), in Qingyuan county, Zhejiang province. The average temperature is 17.6 °C, annual sunshine duration is 1700 h, annual precipitation is 1721.3 mm, and the frost-free period is 245 days per year (Qingyuan People’s Government 2018).The roof of the greenhouse was 2.2 m above the ground.Seedlings were watered by an automatic irrigation system under natural lighting and ventilation. To avoid stress from excessive illumination, seedlings were shaded by horticultural nets to provide half the intensity of sunshine.Temperatures in the greenhouse ranged between 22 and 38 °C and relative humidity was above 70% .

Plant material

Red-seed tree ( Ormosia hosiei Hemsley & E.H. Wilson),Zhejiang phoebe ( Phoebe chekiangensis Shang), and Zhejiang camphor ( Cinnamomum japonicum Sieb.) were chosen as examples of naturally occurring endangered species in the subtropics. In the IUCN Red List of Threatened Species?, natural populations of red-seed tree were assessed as near-threatened (World Conservation Monitoring Center 1998a), Zhejiang phoebe as vulnerable (World Conservation Monitoring Center 1998b), and Zhejiang camphor at a lower risk (World Conservation Monitoring Center 1998c). Red-bark oak [ Cyclobalanopsis gilva(Blume) Oerst] was not listed as an endangered species,hence it was chosen as a reference.

Seedling production

Seeds of the four species were collected in the forest farm,sown germinated and cultured for one growing season in 2014. In late April 2015, seedlings of uniform size were planted in non-woven fabric containers. Initial heights of red-bark oak, red-seed tree, Zhejiang phoebe, and Zhejiang camphor seedlings were 29.7 cm, 36.7 cm, 31.0 cm, and 34.0 cm, respectively. Initial root collar diameters, measured at 2.0 cm above the root collar, were 4.2 mm, 5.4 mm,4.6 mm, and 4.7 mm, respectively. According to local protocol of seedling culture, the container medium was peat,chaff, and soil ( total nitrogen 1.7 g kg?1, total phosphorous 0.3 g kg?1, total potassium 5.4 g kg?1, and organic matter 54.5 g kg?1) in volume ratios of 40% :30% :30% . The substrate was determined to have total N of 14.2 g kg?1, total P of 0.7 g kg?1, total potassium (K) of 2.7 g kg?1, cellulose 200.0 g kg?1, gross ash 158.0 g kg?1, organic matter 720.9 g kg?1, total humus of 381.8 g kg?1, and a pH of 6.0.

Experimental design

The experiment followed a factorial design with four fertilizer levels combined with four container types.Each combined treatment was replicated three times.The four species were four split groups and seedlings treated with control release fertilizers (,18-8-8; APEX?, J.R. Simplot Co., Lathrop, CA, USA)at 1.0 kg, 2.0 kg, 3.0 kg, and 4.0 kg per m3. Containers were columnar, non-woven fabrics pots, in four volumes:2.77 L (18 cm × 14 cm), 3.62 L (18 cm × 16 cm), 5.09 L(20 cm × 18 cm), and 6.28 L (20 cm × 20 cm). The supply of controlled release fertilizer was for seven months. In the greenhouse, there was an 11 h per day natural photoperiod over 2000 lx, considered sufficient to grow quality seedlings (Wei et al. 2013).

Seedling sampling and determination

The growth period was terminated on 2 November, 2015 to avoid possible winter frost, even in the greenhouse. Eight seedlings were randomly sampled from one replicate of the combined treatment, measured for height (root collar to the shoot tip) and RCD (diameter 1.5 cm above the root collar)and divided into leaves, stems, and roots. Excised seedling parts were oven-dried at 68 °C for 72 h. and shoot and root biomass determined, ground and passed through a 1-mm sieve. N and P concentrations were determined by digesting a 0.2 g sample in sulfuric acid and hydrogen peroxide (Li et al. 2017). Total N was determined by the Kjeldahl method and total P by the ICP-OES method.

Data calculation and analysis

Shoot biomass was the sum of biomass for leaves and stem;hence whole-plant biomass was the sum of biomass for shoot and root. Root to shoot biomass ratio (R/S) was calculated by the division of shoot biomass by root biomass. Nutrient uptake efficiency for N ( NUE) and P ( PUE) were both calculated as (Li et al. 2017):

where Bio. is the whole-plant biomass (g), % N/P is the percent of whole-plant nutrient (N or P) content per biomass,and SN/Pis the total nutrient supply for N or P (g). Nutrient utilization index for N ( NUI) and P ( PUI) were calculated as (Li et al. 2018):

where % N/PLeafis the percent of foliage nutrient (N or P)content per biomass.

All data were analyzed using SAS (ver. 9.4 64-bit, SAS Institute, Cary, NC, USA). All data were checked for normality and homogeneity of variance by the UNIVARIATE and GLM (the HOVTEST option) procedures, respectively. There was no necessity to enforce any transformation for data due to the normal distribution and homogeneity of variance. Data were analyzed by a two-way ANOVA to detect the effects of fertilizer dosage and container type on seedling parameters. When an interactive effect was detected by two-way ANOVA, combined treatments were arranged and compared by the one-way ANOVA. When the interactive effect was not significant, data were compared and arranged by two main effects of fertilizer dose or container type. The significance level was taken as the P < 0.05 according to Tukey’s honestly significant difference (HSD) post hoc test.

Results

Seedling growth

Fertilizer dosage and container type had no interactive effect on seedling height (Table 1). Fertilizer treatment alone did affect height for all species. Height increasedwith an increase of fertilizer although differences in height were not statistically different between the F3 and F4 treatments for the Zhejiang phoebe and Zhejiang camphor seedlings (Table 2). Container type significantly affected the height of Zhejiang camphor seedlings, where height was greater in the C5.1 treatment than in the C3.6 treatment (Table 2).

Table 1 ANOVA analysis of fertilization (F), container type (C),and their interaction (F × C) on growth, biomass, and nutrient concentrations in subtropical seedlings of red-bark oak ( Cyclobalanopsis gilva), red-seed tree ( Ormosia hosiei), Zhejiang phoebe ( Phoebe chekiangensis), and Zhejiang camphor ( Cinnamomum japonicum)

There was a significant interactive effect on root collar diameter (RCD) of Zhejiang phoebe seedlings (Table 1).The RCD was higher in seedlings grown in larger containers under the F3 and F4 treatments. In contrast, RCD was lower for Zhejiang phoebe seedlings in smaller containers under lower dose fertilizer treatments.

Biomass accumulation

There were no interactive effects of fertilizer and container type on biomass for any seedling organ (root, stem, leaves)of all species (Table 1). Leaf biomass was greater with higher dose fertilizer treatments F2, F3, and F4 than in the F1 treatment, but leaf biomass of Zhejiang phoebe and Zhejiang camphor were greatest with the F4 treatment (Fig. 1 c,d). Stem biomass of red-bark oak was not affected by fertilizer treatment but was greater in the other species under higher-dose treatments. In contrast, of the four species, only red-bark oak had significant root biomass response to fertilization (Fig. 1 a). In this species, root biomass declined with increasing fertilizer and was lowest in the F4 treatment.

Container type had no effect on leaf biomass in red-seed tree and Zhejiang camphor seedlings (Fig. 1 f, h). In contrast, biomass of red-bark oak and Zhejiang phoebe seedlings increased with the volume of container and was greater in the C6.3 than in the C2.8 treatment (Fig. 1 e, g). Stem biomass was greater in the C6.3 treatment than in the C2.8 and C3.6 treatments, while root biomass increased with container volume (Fig. 1, right).

Whole-plant biomass increased with fertilizer dosage for red-seed tree seedlings (26-34 g), Zhejiang phoebe(25-38 g), and Zhejiang camphor (28-38 g), but did not increase in red-bark oak seedlings (Table 1). Whole-plant biomass increased with the container size but differences between the C5.1 and C6.3 treatments were not statistically significant. Belowground to aboveground biomass ratios (R/S) of red-bark oak decreased with increasing fertilizer (0.45-0.30), red-seed tree (0.64-0.49), Zhejiang phoebe (0.52-0.33), and Zhejiang camphor (0.60-0.41), but increased with the volume of container only for red-seed tree seedlings (Table 1).

Table 2 Effects of fertilization and container type on height and root collar diameter of red-bark oak ( C. gilva), redseed tree ( O. hosiei), Zhejiang phoebe ( P. chekiangensis),and Zhejiang camphor ( C.japonicum)

N and P concentrations

Fertilizer treatments increased N levels in leaves and stems of red-bark oak seedlings, stem concentrations of Zhejiang phoebe and stem and root concentrations of Zhejiang camphor seedlings (Table 1). In red-bark oak seedlings, leaf N levels were higher in larger containers receiving lower dosages of fertilizer, and in smaller containers receiving higher fertilizer amounts (Fig. 2 a). Stem N concentrations of redbark oak seedlings were highest in the F2-C5.1 treatment(Fig. 2 d) and highest in stems of Zhejiang phoebe and Zhejiang camphor seedlings in the F3-C3.6 treatments (Fig. 2 e,f). Root N concentrations were higher fertilizer treatments in Zhejiang camphor seedlings (Fig. 2 i).

Fertilizer treatments had significant effects on leaf N levels in Zhejiang phoebe seedlings (Table 1), where concentrations increased with the amount of fertilizer and reached its highest level in the F2 treatment (Fig. 2 c). Root N concentrations increased with fertilizer levels in redbark oak and Zhejiang phoebe seedlings (Fig. 2 g and h,respectively).

For phosphorous concentrations, only Zhejiang phoebe and Zhejiang camphor seedlings responded to the interactive effects of fertilizer and container type (Table 1). Leaf P concentrations were higher with higher levels of fertilizers in these two species (Fig. 3 b, c), but root P levels were lower with higher fertilizer in Zhejiang camphor seedlings(Fig. 3 i). Leaf P concentrations were higher in the F3 and F4 treatments than in the F2 treatment for red-bark oak seedlings (Fig. 3 a). Stem P levels increased with fertilizer dosage from the F1 to the F3 treatments in red-bark oak,Zhejiang phoebe, and Zhejiang camphor (Fig. 3 d, e, f).Red-bark oak and Zhejiang phoebe seedlings had different stem P concentrations to container type. Phosphorous levels in the stems were the lowest in the smallest container for red-bark oak seedlings but highest for Zhejiang phoebe seedlings (Fig. 3 d, e). Root P concentrations varied considerably to fertilizer treatments in the different species(Table 1). In red-bark oak seedlings, root phosphorous levels increased with the volume of container from the C1 to the C3 treatments (Fig. 3 g). For Zhejiang phoebe seedlings,P concentrations in the roots decreased with increasing levels of fertilizer (Fig. 3 h).

Fig. 1 Biomass accumulation in leaves, stems, and roots of red-bark oak ( Cyclobalanopsis gilva), red-seed tree ( Ormosia hosiei), Zhejiang phoebe ( Phoebe chekiangensis), and Zhejiang camphor ( Cinnamomum japonicum) in response to fertilizer addition (left) and container size (right). Letters indicate significant differences among treatments according to Tukey test at the 0.05 level. Lower case a, b,and c for leaves, letters of α, β, and χ for stems, and letters of x, y, and z for roots

Fig. 2 Nitrogen (N) concentration in leaves, stems, and roots of redbark oak ( C. gilva), Zhejiang phoebe ( P. chekiangensis), and Zhejiang camphor ( C. japonicum) in response to the interactive effects of fertilizer and container type. Different letters indicate significant differences among treatments according to Tukey test at 0.05 level

N and P uptake and utilization efficiencies

Among the four species, only red-bark oak and Zhejiang phoebe seedlings showed significant nutrient uptake efficiency for nitrogen ( NUE) to the interactive effects of fertilizer and container type (Table 1). For these two species, NUE was higher in seedlings receiving lower amounts of fertilizer but raised in larger containers (Fig. 4 a, e). Redbark oak seedlings did not show nutrient uptake efficiency for phosphorous ( PUE) (Table 1). PUE in the other three species was higher with lower doses of fertilizers but grown in larger containers (Fig. 4 d, f, h).

Fig. 3 Phosphorus (P) concentration in leaves, stems, and roots of red-bark oak ( C. gilva), Zhejiang phoebe ( P. chekiangensis), and Zhejiang camphor ( C. japonicum) in response to the interactive effects of fertilizer and container type. Different letters indicate significant difference among treatments according to Tukey test at 0.05 level

The interactive effects of fertilizer and container type did not affect the nutrient utilization index for N ( NUI) or for phosphorous ( PUI) (Table 3). Neither the NUI nor the PUI were affected by fertilizer treatment in red-bark oak and Zhejiang camphor seedlings. For the red-seed tree seedlings, the NUI increased with fertilizer level while the PUI was highest in the F2 treatment. With the Zhejiang phoebe seedlings, both NUI and PUI increased with fertilizer dose. Container type did not affect NUI and PUI in red-seed tree and Zhejiang camphor seedlings (Table 3).For the other species, NUI increased with container type.The phosphorous utilization index was a response to container size for red-bark oak and Zhejiang phoebe seedlings. For red-bark oak seedlings, PUI increased with container volume; for Zhejiang phoebe seedlings, PUI was lowest in the C2.8 treatment.

Discussion

Growth and biomass accumulation

In this study, with increasing levels of fertilizer, seedling heights for red-bark oak and red-seed tree increased accordingly, but increment ceased in the F2 treatment. Height growth response by Zhejiang phoebe and Zhejiang camphor seedlings ceased in the F3 and F4 treatments, respectively.These results demonstrate that the four species required different levels of nutrition to support height growth. Red-bark oak and red-seed tree seedlings required 2 kg m?3controlled release fertilizer while Zhejiang phoebe and Zhejiang camphor seedlings needed 3 kg and 4 kg per m3, respectively.

With increase in fertilizer, all seedlings responded in height growth and shoot biomass without any change in higher fertilizer doses. Although root collar diameter(RCD) responded to fertilizer treatment only by Zhejiang phoebe and Zhejiang camphor seedlings, the response was in accord with that of height and shoot biomass. This means that the highest fertilizer amount resulted in some nutritional status between deficiency and loading for all four species (Salifu and Timmer 2003). The lack of root collar response to fertilizer treatments in red-bark oak and red-seed tree seedlings concur with those found with Buddhist pine [ Podocarpus macrophyllus (Thunb.) D. Don](Wei et al. 2013). Fertilization may affect height growth but not result in a diameter response for sub-tropical seedlings due to high air humidity.

Fig. 4 Nitrogen (N) and phosphorous (P) uptake efficiencies of redbark oak ( C. gilva), red-seed tree ( O. hosiei), Zhejiang phoebe ( P.chekiangensis), and Zhejiang camphor ( C. japonicum) in response to the interactive effects of fertilizer and container type. Different letters indicate significant difference among treatments according to Tukey test at 0.05 level

Table 3 Effects of fertilization and container type on nitrogen(N) and phosphorus (P)utilization of red-bark oak( C. gilva), red-seed tree ( O.hosiei), Zhejiang phoebe ( P.chekiangensis), and Zhejiang camphor ( C. japonicum)

Fertilization can impact fine root growth and plasticity(Luo et al. 2013). For some species, root growth may be too aggressive to be controlled (Haywood et al. 2011). However,root biomass of endangered species of red-seed tree, Zhejiang phoebe, and Zhejiang camphor in our study did not respond to fertilizer treatment. The null response contributed to the obstacle of successful regeneration of these seedlings.In contrast, red-bark oak, which is not endangered, had less root biomass with higher fertilizer levels. Other studies also have found the null effect of fertilization on root biomass in subtropical seedlings such as Buddhist pine (Wei et al. 2013)and fragrant rosewood ( Dalbergia odorifera T.C. Chen) (Li et al. 2017). The lack of response to fertilization by root biomass limits specific root foraging capacity of seedlings(Wei et al. 2017). This may further impact the ability for nutrient and water uptake, increasing the vulnerability of endangered seedlings.

Studies have demonstrated that large volume containers may result in larger seedlings with greater heights and diameters in Mediterranean (Mariotti et al. 2015a; Salto et al. 2016) and tropical regions (Dumroese et al. 2011).Zhu et al. ( 2016) found similar results for two subtropical species. However, for temperate and semi-arid/arid tree species, large containers did not increase shoot growth as much of the biomass was allocated to the roots (Pinto et al. 2011;De La Fuente et al. 2017). Thomas et al. ( 2016) found that different hybrid poplar clones responded to container type to different degrees. Similarly, Kostopoulou et al. ( 2011) found a species-specific response to container type. These studies suggest that container size may be the only determinative factor for the growth of tree seedlings with some ecological plasticity.

In this study, only Zhejiang camphor seedlings showed a response in height and root collar diameter to container type. With an increase in volume, growth maximized at 5 L. In addition, stem and root biomass also responded similarly to container type. These results suggest that not all larger volumes cannot promote diameter growth and biomass accumulation in the subtropical species in this study. The whole-plant biomass of Zhejiang camphor seedling species increased from 2.8 to 5.1 L as well, but the root/shoot (R/S) ratios did not change among container types. Although NeSmith and Duval ( 1998) indicated that root/shoot ratios would change with pot size,both Hess and De Kroon ( 2017) and Poorter et al. ( 2012)reported small but significant changes in root/shoot ratios.Therefore, the unchanged growth of Zhejiang camphor seedlings in the largest container is not be the result of biomass allocation. Our results do not agree with those of Pinto et al. ( 2011) and De La Fuente et al. ( 2017) where larger containers limited seedling growth by allocating biomass to the roots.

N and P concentrations in seedling tissues

Leaves play a key role in plant processes as they are the primary site of photosynthetic carbon gain and nutrient assimilation (Liao et al. 2012). An ideal fertilizer regime efficiently promotes N and P uptake and subsequently increases foliar N accumulation (Liao et al. 2012). Therefore, foliar N and P concentrations may considered as one of the primary indices to evaluate whole-plant nutrient status. None of the three endangered species in this study had significant increase in leaf N concentrations as a result of the interactive fertilizer levels and container type treatments (Fig. 2). This was because all seedlings were received the same amount of fertilizer and their biomass did not show any interactive effects.In seedlings of these three species, container type had no influence on their foliar N levels, while fertilization only affected leaf N concentration in Zhejiang phoebe seedlings.Nutrient concentrations in Pinus ponderosa Douglas ex C.Lawson (Pinto et al. 2011) and Betula pendula Roth (Aphalo and Rikala 2003) seedlings also did not respond to container type. In addition, Zhu et al. ( 2016) reported a rare nutrient leaching related to container type for Buddhist pine and Japanese maple ( Acer palmatum Thunb.) seedlings. Instead,in this study, fertilizer treatment promoted N concentration in stems and roots in Zhejiang phoebe and Zhejiang camphor seedlings in larger containers. These are evergreen-broadleaf species and may have been undergoing N retranslocation from leaves to woody tissues in the late growing season of November. Studies have shown that seedlings translocate nitrogen from leaves to woody tissues in stems and roots which may be promoted by time-related fertilization (Wei et al. 2014). For seedlings in this study, larger containers would allow additional fertilizer to increase nitrogen levels in woody tissues because seedlings invest more resources to roots in larger containers (Pinto et al. 2011; De La Fuente et al. 2017). In these two species, P concentrated in leaves as a result of fertilizer addition to larger containers. Li et al.( 2017) found that fertilizer treatment had no effect on foliar P concentrations in fragrant rosewood seedlings. Therefore,it appears that for these subtropical endangered seedlings,fertilization may result in greater phosphorous uptake in larger containers. However, nitrogen and phosphorous levels in red-seed tree seedlings did not reflect fertilizer dosage or container type, indicating a highly species-specific responsive difference.

N and P uptake and utilization

The results show that endangered seedlings have higher N and P uptake efficiencies when raised in larger containers and at lower dosages of fertilizer (Fig. 4). Larger containers permit the roots to enlarge, promoting nutrient uptake even at low nutrient levels. The lowest controlled-release fertilizer, 10 kg m ?3 resulted in the highest nutrient uptake efficiency, suggesting that any higher amount of fertilizer would be excessive. The combined effects of fertilizer supply and container type significantly affected N uptake efficiency in red-bark oak and Zhejiang phoebe seedlings, but significantly affected P uptake efficiency by all three species. Therefore, for endangered seedlings higher phosphorous availability would promote grow of mycorrhizae (De Oliveira et al. 2015) and subsequent seedling growth (Zhu et al. 2016). Nevertheless, fertilization increased the efficiency of nitrogen and phosphorous utilization by red-seed tree and Zhejiang phoebe seedlings; while larger containers resulted in higher utilization efficiencies by red-bark oak and Zhejiang phoebe seedlings (Table 3). These results were due to additional fertilizer and container type on whole-plant biomass.

Conclusions

In this study, both fertilization and container type promoted growth and biomass accumulation by endangered tree seedlings. Whole-plant biomass and seedling nitrogen and phosphorous levels increased with higher doses of fertilizer in larger containers, allowing for higher nitrogen and phosphorous utilization. However, any amount above 1 kg m?3would improve nutrient uptake efficiency. Our results were highly species-specific. Zhejiang phoebe and Zhejiang camphor seedlings responded to the interactive effects of fertilization and container type while red-seed tree seedlings did not increase nutrient uptake. As a species that is not endangered, red-bark oak seedlings had a better ability to allocate nitrogen to shoots in response to higher levels of fertilizer compared to endangered seedlings. Red-bark oak seedlings cannot efficiently absorb and utilize phosphorous for growth.Therefore, container type mainly influenced root biomass allocation to promote nutrient utilization even at low levels of fertilizer. It is recommended using 5 L containers to raise Zhejiang phoebe and Zhejiang camphor seedlings with fertilizer at 3 kg m?3. Red-bark oak and red-seed tree seedlings could be cultured 2 kg m?3of fertilizer in smaller containers.

AcknowledgementsThe authors thank Rongzhou Man and MyaRice,both of them are from Ontario Forest Research Institute in Canada, for the polishing work and suggestions for the paper.