Growth Analysis of Potted Seedlings of Satsuma Mandarin (Citrus unshiu Marc.) under Different Light Conditions and Air Temperatures

Abstract

To establish cultural practice based on a consecutive growth model for potted 1-year-old seedlings of Satsuma mandarin (Citrus unshiu Marc.), growth analysis by classical and functional approaches was conducted under different light conditions and air temperatures over 2.5 years, and the active growth of potted seedlings in the greenhouse was investigated. Under the classical approach, the general change patterns of relative growth rate (RGR) and net assimilation rate (NAR) were hard to determine because of irregularities including quiescence of vegetative growth. Under the functional approach, plant mass modeled using linear, exponential, power-law, monomolecular, three-parameter logistic, four-parameter logistic (4L), and Gompertz functions showed significant correlations with the observed plant mass. 4L was the best model because it showed the highest r, and the lowest root mean square error and Akaike Information Criterion, so RGR and NAR were estimated by 4L. Analysis of the RGR components showed significant positive correlations between RGR and NAR. Analysis of covariance indicated the NAR costs for increasing RGR were lower in the greenhouse than in open culture; this was explained by differences in specific leaf area (SLA). Therefore, in greenhouse culture, growth was primarily enhanced by NAR as net photosynthesis and underpinned by SLA as a morphological trait improvement for the relatively low light intensity compared with open culture. A multiple regression model for NAR using the pooled data (n = 60) suggested solar radiation had a positive effect (P < 0.0001) and air temperature had a negative effect (P < 0.01) on NAR.

Introduction

Satsuma mandarin (Citrus unshiu Marc.) is a major fruit crop in Japan, and the summer-harvest cropping type has been cultured in heated greenhouses. Recently, fluctuations in fuel prices have threatened heated greenhouse culture; therefore, it is important to establish energy-saving temperature management methods (Yano et al., 2014a, b, c) and to increase fruit yield by optimizing the light-intercepting characteristics of trees. In addition, most Satsuma mandarin trees planted in greenhouses are over 20 years old and have chronically low fruit yields. To overcome these problems, a new planting style is needed for Satsuma mandarin, and it is very important to shorten the vegetative growth period from planting to cropping.

All fruit farmers need cultural practice to shorten the period from planting to cropping with less labor. For example in open culture, some reports have described the cultural practice for seedlings up to 2 years old (Nakamoto et al., 2003; Shibayama and Akasaka, 2006), and a few reports have discussed the cultural practice for seedlings up to 4 years old (Iguchi et al., 1985; Yoshida and Ishida, 1982). In these reports, young seedlings were first temporarily potted or planted in a preliminary field for one or several years before planting in an orchard, although the cost of labor is lower for potted and young seedlings. Transplanting is not always done in the spring season for the summer-harvest cropping type of Satsuma mandarin because each greenhouse has a different growth cycle and yearly management schedule. Therefore, a consecutive growth model for potted seedlings over several years is essential for rational planting management (e.g. number of seedlings, environment and term of vegetative growth, and leaf area index to aim for) in the greenhouse with less labor.

Temperature is the single most important factor influencing organ growth (DeJong, 1999). Therefore, heat units or thermal time is used to measure the developmental period in fruits (DeJong and Goudriaan, 1989; Godoy et al., 2008; Wubs et al., 2012) or the growth phase (White et al., 2012). Seedling growth is also influenced by temperature, and more active and rapid seedling growth are expected in a greenhouse than in the open field. Because rapid growth is promoted by heating, it is useful for horticultural farmers to quantify growth on a thermal time basis, not a time basis.

Greenhouse horticulture sacrifices light intensity to some extent because of the covered film used to keep plants warm, and long-term cultivation of seedlings for several years faces low light intensity conditions in winter. Additionally, the cultural practice in the greenhouse tends to place seedlings in over-crowded conditions with low absolute light intensity because of space limitations, which restricts active growth, and this tendency is more apparent with high-density planting. Therefore, to establish good cultural practice, the effects of both temperature and light conditions on the growth of potted seedlings must be clarified.

Recently, to predict fruit yield and size, or for use in crop simulation models, fruit growth has often been modeled by non-linear functions (Barrera et al., 2008; Cuevas et al., 2003; Godoy et al., 2008; Tadesse et al., 2002; Wubs et al., 2012). Many studies on functional vegetative growth modeling have been reported in herbaceous and woody plants (Hunt, 1982; Nishizono, 2010; Takiya, 2014). However, to our knowledge, few studies have reported functional growth modeling of fruit tree seedlings over several years.

The goal of this study was to understand the trends of vegetative growth over consecutive seasons in potted Satsuma mandarin seedlings using environmental information. However, in the open field, seedlings are often defoliated by low temperatures and/or strong winds, which disturbs the evaluation of growth over several consecutive years. To minimize this error, in this study we examined individual plant masses and compared both classical (interval) and functional approaches (Hunt, 1982; Paine et al., 2012) for growth analysis based on thermal time.

Furthermore, to gain a universal understanding of the growth of potted Satsuma mandarin seedlings, regression analysis was conducted to quantify the relative importance of the net assimilation rate (NAR), specific leaf area (SLA), and leaf mass ratio (LMR) in determining the relative growth rate (RGR), which explains the variations in environmental factors and seedling age. In general, RGR is the key to understanding growth, NAR is a physiological component related to whole-plant daily net photosynthesis (Poorter and Van der Werf, 1998), SLA is a morphological component determined by the leaf dry matter concentration and leaf thickness (Shipley, 1995; Witkowski and Lamont, 1991), and LMR indicates the allocation of biomass to leaves vs. other plant parts. We also discuss the advantages of greenhouse culture for the growth of potted Satsuma mandarin seedlings according to the growth analysis.

Materials and Methods

Plant materials, environments, and measurements

One-year-old ‘Oita-Wase’ Satsuma mandarin seedlings grafted onto trifoliate orange [Poncirus trifoliatae (L.) Raf.] rootstock were planted in 20 L pots in March 2013, and were cultured until October 2015 in a field of the Fruit Tree Group, Oita Prefectural Agriculture, Forestry, and Fisheries Research Center (Kunisaki, Oita, Japan). Irrigation was 0.6 L of water with a drip tube every day. Each seedling was fertilized with 200 g of coated granular fertilizer (N:P:K = 14:11:13) each year. All flower buds were detached during preanthesis. The sampling intervals were 1–3 months, and five replications were performed.

Three sets of environmental conditions were used to represent the following; (1) greenhouse culture to promote the growth of seedlings, (2) greenhouse culture at low light intensity caused by shading, and (3) open culture. The covering materials were 0.15 mm polyester film for the greenhouse ceiling and knitted silver polyvinyl film for shading. The abbreviations for the environments are H: house culture (no heating), HS: house culture (no heating) + shading, and O: open culture. Air temperature (T_a), solar radiation (R_s), and soil water content (VWC) were measured with a compact water proof temperature data logger (RTR-53; T&D Co., LTD., Japan) at 10-min intervals, a pyranometer (Solar Mini PCM-01; Prede Co., LTD., Japan) at 10-min intervals, and a TDR sensor (EC-5; Decagon Devices, Inc., USA) monitored with a compact data logger (EM5b; Decagon Devices, Inc.) at 1-h intervals, respectively.

Growth analysis by the classical (interval) approach

Growth analysis by the classical approach followed Hunt (1982). The mean relative growth rate by the classical approach ($\overline{\text{RGR}}$) on a thermal time basis was calculated as   

$$\overline{\text{RGR}}=\frac{m_{P2}-m_{P1}}{{tt}_2-{tt}_1}\text{,}$$

where m_P2 and m_P1 are log-transformed plant mass at subsequent thermal times tt₂ and tt₁, respectively. Similarly, the derivative parameters net assimilation rate ($\overline{\text{NAR}}$), specific leaf area ($\overline{\text{SLA}}$), leaf mass ratio ($\overline{\text{LMR}}$), and leaf area ratio ($\overline{\text{LAR}}$) were calculated by the classical approach directly from the experimental data as   

$$\overline{\mathrm{N}\mathrm{A}\mathrm{R}}=\frac{{\text{M}}_{\text{P2}}\text{-}{\text{M}}_{\text{P1}}}{{\text{tt}}_{\text{2}}\text{-}{\text{tt}}_{\text{1}}}\text{×}\frac{{\text{log}}_{\text{e}}{\text{A}}_{\text{L2}}\text{-}{\text{log}}_{\text{e}}{\text{A}}_{\text{L1}}}{{\text{A}}_{\text{L2}}\text{-}{\text{A}}_{\text{L1}}}\text{,}$$

  

$$\overline{\mathrm{S}\mathrm{L}\mathrm{A}}=\frac{{\text{A}}_{\text{L1}}/{\text{M}}_{\text{L1}}\text{+}{\text{A}}_{\text{L2}}/{\text{M}}_{\text{L2}}}{\text{2}}\text{,}$$

  

$$\overline{\text{LMR}}\text{=}\frac{{\text{M}}_{\text{L1}}/{\text{M}}_{\text{P1}}\text{+}{\text{M}}_{\text{L2}}/{\text{M}}_{\text{P2}}}{\text{2}},$$

  

$$\overline{\text{LAR}}\text{=}\frac{{\text{A}}_{\text{L1}}/{\text{M}}_{\text{P1}}\text{+}{\text{A}}_{\text{L2}}/{\text{M}}_{\text{P2}}}{\text{2}}\text{,}$$

where M_P2 and M_P1 are non-transformed plant mass, A_L2 and A_L1 are plant leaf area, and M_L2 and M_L1 are plant leaf mass at subsequent thermal times tt₂ and tt₁, respectively.

Growth analysis by the functional approach

The function-modelled plant mass (m’, log transformed) and instantaneous RGR of plant mass derived from the models (RGR_F) were defined in a previous report (Paine et al., 2012), and are shown in Table 1. The subscript F of RGR_F contains an abbreviation of each model function (Table 1).

Table 1

Functional forms for plant growth modeling. The models are equations expressing changes in biomass (M) and relative growth rate (RGR) as a function of thermal time (tt). The parameters M₀, r, b, L, and K indicate initial biomass, the absolute increase in biomass per unit thermal time, the exponent value, the lower horizontal asymptote, and the upper horizontal asymptote, respectively.

Instantaneous NAR (NAR) was estimated by the basic equation (Hunt, 1982),   

$$\text{NAR=}\frac{{\text{RGR}}_{\text{F}}}{\text{SLA×LMR}},$$

where SLA and LMR are instantaneous SLA and LMR calculated directly from the data from each harvest, respectively.

Data analysis

Data analysis was performed using R-3.2.5 (R Core Team, 2016). Curve fitting was performed using both nls.lm (Elzhov et al., 2015) and nls (R Core Team, 2016). To compare the different functions for plant growth, the root mean square error (RMSE) between the predicted and observed values was calculated using the following equation,   

$$\mathrm{R}\mathrm{M}\mathrm{S}\mathrm{E}=\sqrt{\frac{1}{n}\sum _{i=1}^n{(P_i-O_i)}^2,}$$

where n is the number of data points, P_i is the predicted value of observation i, and O_i is the observed value of observation i. The Akaike Information Criterion (AIC) was calculated using the following equation,   

$$\text{AIC}=\mathrm{-2}\left(\text{log-likelihood}\right)+2\text{p}\text{,}$$

where p is the number of parameters. The log-likelihood was calculated using logLik (R Core Team, 2016) in R, after fitting by nls (R Core Team, 2016).

Results

Environmental conditions

Figure 1 shows the environmental data during the experimental period. The monthly mean values ± SE were as follows; air temperature: H, 19.4 ± 1.1°C, HS, 19.1 ± 1.1°C, and O, 16.9 ± 1.3°C, solar radiation: H, 9.1 ± 0.6 MJ·m⁻², HS, 3.0 ± 0.5 MJ·m⁻², and O, 14.4 ± 0.7 MJ·m⁻², soil water content: H, 0.20 ± 0.01 m³·m⁻³, HS, 0.31 ± 0.02 m³·m⁻³, and O, 0.25 ± 0.01 m³·m⁻³. HS showed the highest water content because it had the lowest level of evaporation due to shading. In general, O showed moderate water content because of natural rainfall. H showed the lowest water content except 18–24 months after planting, and at the end of experiment, the soil water content was below 0.1 m³·m⁻³, which would have been caused by a large amount of plant transpiration (Fig. 1c).

Fig. 1

Seasonal changes in the environment. (a) Monthly mean air temperature (T_a, °C). (b) Monthly mean solar radiation (R_s, MJ·m⁻²·day⁻¹). (c) Monthly mean soil water content (VWC, m³·m⁻³). ○, House; ●, House + Shading; △, Open culture.

Cumulative growth of seedlings

On a thermal time basis, M_P and A_L generally showed an exponential increase (Fig. 2a). M_P and A_L mostly showed the largest values in H among the three environments, while M_P in HS increased slowly until 9.4 × 10³ day·°C (Fig. 2a, b). The general increase pattern in HS was relatively stable (Fig. 2a). A_L in O showed some fluctuation, and an apparent reduction in A_L occurred during the second winter season, around 1.1 × 10⁴ day·°C (21–23 months after planting in Fig. 1a), after which both A_L and M_P in O were below those in HS (Fig. 2a, b).

Fig. 2

Seasonal changes in plant growth. (a) Plant mass (M_P, g). (b) Leaf area (A_L, m⁻²). ○, House; ●, House + Shading; △, Open culture.

In the comparison of photosynthetic ability between H and HS at almost equal plant ages (1.1 × 10⁴ day·°C), no significant difference was observed except under dark conditions (data not shown).

Derivatives of growth analysis by the classical approach

$\overline{\text{RGR}}$s and $\overline{\text{NAR}}$s showed huge fluctuations from 5.0–7.0 × 10³ day·°C (8–12 months after planting in Fig. 1a), and their general seasonal general tendencies were hard to determine (Fig. 3a, b).

Fig. 3

Seasonal changes in the growth components calculated by the classical (interval) approach. (a) Mean relative growth rate ($\overline{\text{RGR}}$, (10³ day·°C)⁻¹). (b) Mean net assimilation rate ($\overline{\text{NAR}}$, g·m⁻²·(10³ day·°C)⁻¹). (c) Mean specific leaf area ($\overline{\text{SLA}}$, m²·g⁻¹). (d) Mean leaf mass ratio ($\overline{\text{LMR}}$, g·g⁻¹). (e) Mean leaf area ratio ($\overline{\text{LAR}}$, m²·g⁻¹). House (H), bold gray line; House + Shading (HS), bold line; Open culture (O), narrow line.

$\overline{\text{SLA}}$s were generally affected by light conditions; for example, the lowest light conditions, which were in HS, showed the highest $\overline{\text{SLA}}$s (Fig. 3c) almost continuously. In addition, seasonal fluctuations were observed; the first peaks were observed at around 2.5 × 10³ day·°C (4 months after planting in Fig. 1a) in all treatments (H, HS, and O), and second maximum peaks were observed at around 7.5 × 10³ day·°C (14–16 months after planting in Fig. 1a) in H and HS, but not in O (Fig. 3c).

Generally, the seasonal change in $\overline{\text{LMR}}$ showed three peaks (Fig. 3d). In the first season, $\overline{\text{LMR}}$ in O showed the lowest values from 2.6 × 10³ day·°C (5 months after planting in Fig. 1a) to 6.3 × 10³ day·°C (12–14 months after planting in Fig. 1a). From the beginning of the second season, after 6.3 × 10³ day·°C, $\overline{\text{LMR}}$ in O showed a steep rise and reached the level of H and HS, and in the third season, HS showed the highest values (Fig. 3d).

$\overline{\text{LAR}}$ showed a neutral middle value compared with the models of $\overline{\text{SLA}}$ and $\overline{\text{LMR}}$, including an effect of environment similar to $\overline{\text{SLA}}$ (Fig. 3c–e). Notably, $\overline{\text{LAR}}$ showed a clear peak at around 7.5 × 10³ day·°C (14–16 months after planting in Fig. 1a) in H and HS (Fig. 3e).

Curve fitting and model selection

To understand the general trends of the RGR of seedling mass on a thermal time basis, curve fitting was conducted. In this fitting, all plant mass data were log-transformed to reduce the heteroscedasticity (Paine et al., 2012). All fitting parameters were estimated by nls.lm (R Core Team, 2016) and the AIC was calculated using nls and logLik (R Core Team, 2016) (Table 2). In all environments, the 4L function showed the highest r, and lowest RMSE and AIC values (Table 2).

Table 2

Estimated model parameters and measures of goodness-of-fit for the seasonal changes of log-transformed plant mass. The Pearson coefficient of correlation (r) and root mean square error (RMSE) are given for the difference between predicted and observed values. The Akaike information criterion (AIC) is determined by the number of parameters and the log-likelihood of the model. Function names (F) are shown in Table 1.

Sprouting times during the experiment were H: 9 times, HS: 9 times, and O: 8 times, and in general log-transformed plant mass (m_P) increased rapidly in the vigorous sprouting periods (Fig. 4a–c). From the curve fitting, the change patterns of the modeled log-transformed plant mass (m’_P) showed only small differences among non-asymptotic and asymptotic models (Fig. 4a–c). In contrast, the change patterns of the modeled RGRs were classified as three types, decreasing (Li, P, M, 3P, and Go), convex (4L), and constant (Ex) (Fig. 4d–f).

Fig. 4

Curve fitting for the seasonal changes of log-transformed plant mass (m_P), predicted plant mass (m’_P), and derived relative growth rate (RGR_F). (a–c) Log-transformed plant mass or biomass trajectories predicted by the non-asymptotic and asymptotic models shown in Table 1. The budding period is shown by an arrow in the lower panel. (d–f) RGR_F derived from the functions of Biomass M shown in Table 1. The estimated parameters M₀, r, b, L, and K are shown in Table 2.

Derivative analysis of RGR

RGR is often factored into three components (Hunt, 1982),   

$$\text{RGR}=\text{NAR}\times \text{SLA}\times \text{LMR}.$$

Therefore, once RGR, SLA, and LMR were determined, the residual NAR could be easily estimated (Fig. 5). During the first season (0.0–5.0 × 10³ day·°C), the estimated NAR was H > O > HS, similar to the RGR calculated by 4L fitting (Figs. 4d–f and 5). A large temporal fluctuation in NAR in O was observed at 5.6–6.3 × 10³ day·°C (Fig. 5), because of coincident decreases in $\overline{\text{SLA}}$, $\overline{\text{LMR}}$ (Fig. 3c, d), and A_L (Fig. 2b).

Fig. 5

Seasonal changes in the net assimilation rate (NAR) estimated using the relative growth rate (RGR) fitted with the four-parameter logistic function, instantaneous specific leaf area (SLA), and instantaneous leaf mass ratio (LMR). The basic function was RGR = NAR × SLA × LMR. ○, House; ●, House + Shading; △, Open culture.

To clarify the relative importance of the different components of RGR in determining the variation in RGR, correlation analysis was conducted. No significant correlation was observed between RGR and SLA or LMR, except between RGR and SLA in O, while NAR showed a significant positive correlation (P < 0.01) with RGR in all environments, H, HS, and O (data not shown), and the significance levels of the correlations between log-transformed NAR (nar) and log-transformed RGR (rgr) were higher (P < 0.001) than those of the non-transformed data (Fig. 6a–c). To assess the difference or correspondence of relationships between rgr and nar among H, HS, and O, the slopes and intercepts of the linear models were compared by analysis of covariance (Fig. 6a). The slopes of the three regression lines (H, HS, and O) were not significantly different (P = 0.25). H and HS had a common intercept in the regression line (P = 0.88), and O had a significantly lower intercept compared with these common intercepts (P < 0.01).

Fig. 6

Relationships between log-transformed RGR (rgr) and log-transformed NAR (nar) (a), log-transformed SLA (sla) (b), and log-transformed LMR (lmr) (c); all parameters were calculated from each harvest (n = 21). ○, House; ●, House + Shading; △, Open culture.

To determine the effect of environmental factors on NAR, multiple regression models were fitted for each environment and for the pooled data (Table 3). In H, only parameter b (coefficient of R_s) was significant. In HS, parameters a (intercept) and c (coefficient of T_a) were significant. In O, no significant parameters were observed. The estimations of all parameters were highly significant (P < 0.01) for the pooled data (n = 60), and a positive effect of R_s (P < 0.0001) and a negative effect of T_a (P < 0.01) on NAR were observed.

Table 3

Estimated parameters and their standard errors for multiple regression models for net assimilation rate (NAR) factored by solar radiation (R_s) and air temperature (T_a), fitted in each environment and the pooled data.

Discussion

To evaluate the long-term effects of light conditions on the growth of potted Satsuma mandarin seedlings, growth analysis based on thermal time was conducted using both classical (interval) and functional approaches in greenhouse and open field conditions. In the classical approach, the RGR and NAR of log-transformed seedling mass (m_P) on a thermal time basis did not show a clear general trend (Fig. 3a, b) because growth stopped at during low temperature seasons and because of defoliation in O in winter (Fig. 2b). However, in the functional approach, several functions fitted m_P significantly, and the model using the 4L function was selected as the best model because it had the highest r and lowest RMSE and AIC (Table 2).

Validity of the four-parameter logistic model

The validity of the 4L fitting model was supported by three factors. The first was the physiological interpretation. In many plants, growth is typically characterized by RGR. Comparing all the fitted curves, only RGR in 4L (RGR_4L) showed a convex curve (Fig. 4d–f). The initial low RGR_4L at the start of the experimental period is related to the transition period (20 day after budding) in young spring leaves from sink to source (Hino et al., 1974; Hunt, 1982). After the RGR_4L peak, RGR_4L decreased and fell to zero at the end of experimental period because of resource restriction; for example, self-cover shade, pot size (Poorter et al., 2012), and available soil water content (Fig. 1c). In this way, RGR_4L trends were easily explained by physiological phenomena.

The second was the significant effect of light conditions. The shape of the convex curve in RGR_4L differed by light condition; in particular, it was high in H and low in HS during the first season (Fig. 4d–f). This fitting result clearly explained the reaction of seedling growth to light intensity.

The third was the relationships among RGR and the other derivatives. The mean LAR ($\overline{\text{LAR}}$) and RGR_4L had a common feature as the peaks of $\overline{\text{LAR}}$ corresponded to those in RGR_4L that were commonly observed in H, HS, and O in the second summer, around 7.5 × 10³ day·°C (Figs. 3e and 4d–f). Many previous studies have also shown significant correlation between LAR and RGR (Lambers and Poorter, 1992; Poorter and Remkes, 1990; Reich et al., 1992, 1998; Walters et al., 1993; Wright and Westoby, 1999); therefore, these $\overline{\text{LAR}}$ peaks must contribute to RGR_4L reaching a maximum.

These results strongly support the goodness of fit and prediction using the 4L growth model based on thermal time in potted Satsuma mandarin seedlings.

Reaction of seedling growth to the environment

In this study, on the same thermal time basis, plant mass was mostly H > HS > O (Fig. 2a), while the cumulative light intensity was O > H > HS (Fig. 1b). Even if no quantitative error from defoliation was observed and plant mass was H = O, the cumulative light intensity would be H < O, and H would have a superior photosynthetic system to O despite the low light intensity. This phenomenon is very important not only for potted Satsuma mandarin seedlings, but also many horticultural crops, including fruit produced in greenhouses.

As already mentioned, growth is typically characterized by RGR; therefore, the contribution of its derivatives including NAR, SLA, and LMR has been discussed in many plants (Rees et al., 2010; Shipley, 2002, 2006). Generally, many reports on RGR including interspecific analyses of herbaceous and woody plants can be classified into roughly two groups: SLA-dominant (Lambers and Poorter, 1992; Modrzyński et al., 2015; Poorter and Remkes, 1990; Rees et al., 2010; Reich et al., 1992, 1998; Walters et al., 1993; Wright and Westoby, 1999, 2000) and NAR-dominant (Huante and Rincon, 1998; Huante et al., 1995; Poorter, 1999; Ryser and Wahl, 2001; Saverimuttu and Westoby, 1996; Shipley, 2006; Taub, 2002), and LMR hardly contributes to RGR (Shipley, 2006).

Our results showed three key points. First, the growth of potted Satsuma mandarin seedlings was NAR-dominant (Fig. 6a). Second, analysis of covariance indicated NAR costs for increasing RGR were lower in H and HS than in O, which was explained by differences in SLA (Figs. 3c and 6a). Third, high NAR was suggested to be affected by high R_s and low T_a (Table 3), as previous studies have also shown that high temperatures (23–28°C) decrease NAR through plant respiration loss (Atkin et al., 2006). In several woody plants, high SLA and LAR enhance RGR because they confer high light interception and carbon gain per unit mass invested in leaves (Lambers and Poorter, 1992; Poorter and Remkes, 1990; Reich et al., 1992, 1998; Walters et al., 1993). In this study, unlike the above cited studies, HS showed the highest initial mean SLA ($\overline{\text{SLA}}$) values but a low initial RGR_4L (Figs. 3c and 4d–f); therefore, high $\overline{\text{SLA}}$ is not always related to high productivity as low light intensity limits the absolute carbon gain.

In conclusion, the growth of potted Satsuma mandarin seedlings was best modeled using a four-parametric logistic function. In greenhouse culture, growth was primarily enhanced by NAR as net photosynthesis and underpinned by SLA as a morphological trait improvement for the relative low light intensity compared with open culture.

Literature Cited

•  Atkin,  O. K.,  B. R.  Loveys,  L. J.  Atkinson and  T. L.  Pons. 2006. Phenotypic plasticity and growth temperature: understanding interspecific variability. J. Exp. Bot. 57: 267–281.
•  Barrera,  J. A.,  M. S.  Hernández,  L.  Melgarejo,  O.  Martínez and  J. P.  Fernández-Trujillo. 2008. Physiological behavior and quality traits during fruit growth and ripening of four Amazonic hot pepper accessions. J. Sci. Food Agric. 88: 847–857.
•  Cuevas,  J.,  F. J.  Salvador-Sola,  J.  Gavilán,  N.  Lorente,  J. J.  Hueso and  C. M.  González-Padierna. 2003. Loquat fruit sink strength and growth pattern. Sci. Hortic. 98: 131–137.
•  DeJong,  T. M. 1999. Developmental and environmental control of dry-matter partitioning in peach. HortScience 34: 1037–1040.
•  DeJong,  T. M. and  J.  Goudriaan. 1989. Modeling peach fruit growth and carbohydrate requirements: reevaluation of the double-sigmoid growth pattern. J. Amer. Soc. Hort. Sci. 114: 800–804.
• Elzhov, T. V., K. M. Mullen, A.-N. Spiess and B. Bolker. 2015. minpack.lm: R interface to the Levenberg-Marquardt nonlinear least-squares algorithm found in MINPACK, plus support for bounds. R package version 1.2-0. <https://CRAN.R-project.org/package=minpack.lm>.
•  Godoy,  C.,  G.  Monterubbianesi and  J.  Tognetti. 2008. Analysis of highbush blueberry (Vaccinium corymbosum L.) fruit growth with exponential mixed models. Sci. Hortic. 115: 368–376.
•  Hino,  A.,  S.  Amano,  Y.  Sawamura,  S.  Sasaki and  T.  Kuraoka. 1974. Studies on the photosynthetic activity in several kinds of fruit trees. II. Seasonal changes in the rate of photosynthesis. J. Japan. Soc. Hort. Sci. 43: 209–214 (In Japanese with English abstract).
•  Huante,  P. and  E.  Rincon. 1998. Response to light changes in tropical deciduous woody seedlings with contrasting growth rate. Oecologia 113: 53–66.
•  Huante,  P.,  E.  Rincon and  I.  Acosta. 1995. Nutrient availability and growth rate of 34 species from a tropical deciduous forest in Mexico. Funct. Ecol. 9: 849–858.
• Hunt, R. 1982. Plant growth curves: The functional approach to plant growth analysis. Edward Arnold, London.
•  Iguchi,  I.,  T.  Suzuki,  K.  Mochizuki and  M.  Konakahara. 1985. Improvement in rapid growing methods for nursery stock of citrus. (2) Grafting time, growth environments and transplanting time. Bull. Shizuoka Citrus Exp. Stn. 21: 19–25 (In Japanese with English abstract).
•  Lambers,  H. and  H.  Poorter. 1992. Inherent variation in growth rate between higher plants: a search for physiological causes and ecological consequences. Adv. Ecol. Res. 23: 187–261.
•  Modrzyński,  J.,  D. J.  Chmura and  M. G.  Tjoelker. 2015. Seedling growth and biomass allocation in relation to leaf habit and shade tolerance among 10 temperate tree species. Tree Physiol. 35: 879–893.
•  Nakamoto,  K.,  M.  Hasegawa and  S.  Hasegawa. 2003. Effect of mixed medium of rice hulls and bark manure on the growth of nursery plant in satsuma mandarin ‘Ishiji’. J. Japan. Soc. Hort. Sci. 72: 314 (In Japanese).
•  Nishizono,  T. 2010. Effects of thinning level and site productivity on age-related change in stand volume growth can be explained by a single rescaled growth curve. For. Ecol. Manage. 259: 2276–2291.
•  Paine,  C. E. T.,  T. R.  Marthews,  D. R.  Vogt,  D.  Purves,  M.  Rees,  A.  Hector and  L. A.  Turnbull. 2012. How to fit nonlinear plant growth models and calculate growth rates: an update for ecologists. Method. Ecol. Evol. 3: 245–256.
•  Poorter,  L. 1999. Growth responses of 15 rainforest tree species to a light gradient: the relative importance of morphological and physiological traits. Funct. Ecol. 13: 396–410.
•  Poorter,  H. and  C.  Remkes. 1990. Leaf area ratio and net assimilation rate of 24 wild species differing in relative growth rate. Oecologia 83: 553–559.
• Poorter, H. and A. Van der Werf. 1998. Is inherent variation in RGR determined by LAR at low irradiance and by NAR at high irradiance? A review of herbaceous species. p. 309–336. In: H. Lambers, H. Pooter and M. M. I. Van Vuuren (eds.). Inherent variation in plant growth. physiological mechanisms and ecological consequences. Backhuys, Leiden.
•  Poorter,  H.,  J.  Bühler,  D.  Dusschoten,  J.  Climent and  J. A.  Postma. 2012. Pot size matters: a meta-analysis of the effects of rooting volume on plant growth. Funct. Plant Biol. 39: 839–850.
• R Core Team. 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. <http://www.R-project.org/>.
•  Rees,  M.,  C. P.  Osborne,  F. I.  Woodward,  S. P.  Hulme,  L. A.  Turnbull and  H.  Taylor. 2010. Partitioning the components of relative growth rate: how important is plant size variation? Am. Nat. 176: E152–E161.
•  Reich,  P. B.,  M. G.  Tjoelker,  M. B.  Walters,  D. W.  Vanderklein and  C.  Buschena. 1998. Close association of RGR, leaf and root morphology, seed mass and shade tolerance in seedling of nine boreal tree species grown in high and low light. Funct. Ecol. 12: 327–338.
•  Reich,  P. B.,  M. B.  Walters and  D. S.  Ellsworth. 1992. Leaf lifespan in relation to leaf plant and stand characteristics among diverse ecosystem. Ecol. Monogr. 62: 365–392.
•  Ryser,  P. and  S.  Wahl. 2001. Interspecific variation in RGR and underlying traits among 24 grass species grown in full daylight. Plant Biol. 3: 426–436.
•  Saverimuttu,  T. and  M.  Westoby. 1996. Components of variation in seedling relative growth rate: phylogenetically independent contrasts. Oecologia 105: 281–285.
•  Shibayama,  K. and  S.  Akasaka. 2006. Effects of pot type and soil moisture on the growth of the peach nursery stock and the yield after planting. Hort. Res. (Japan) 5: 375–379 (In Japanese with English abstract).
•  Shipley,  B. 1995. Structured interspecific determinants of specific leaf area in 34 species of herbaceous angiosperms. Funct. Ecol. 9: 312–319.
•  Shipley,  B. 2002. Trade-offs between net assimilation rate and specific leaf area in determining relative growth rate: the relationship with daily irradiance. Funct. Ecol. 16: 682–689.
•  Shipley,  B. 2006. Net assimilation rate, specific leaf area and leaf mass ratio: which is most closely correlated with relative growth rate? A meta-analysis. Funct. Ecol. 20: 565–574.
•  Tadesse,  T.,  E. W.  Hewett,  M. A.  Nichols and  K. J.  Fisher. 2002. Changes in physicochemical attributes of sweet pepper cv. Domino during fruit growth and development. Sci. Hortic. 93: 91–103.
•  Takiya,  M. 2014. Site index curve estimation on Abies sachalinensis plantation forest in Hokkaido, Japan. Bull. Hokkaido For. Res. Inst. 51: 7–11 (In Japanese with English abstract).
•  Taub,  D. R. 2002. Analysis of interspecific variation in plant growth responses to nitrogen. Can. J. Bot. 80: 34–41.
•  Walters,  M. B.,  E. L.  Kruger and  P. B.  Reich. 1993. Growth, biomass distribution and CO₂ exchange of northern hardwood seedlings in high and low light: relationships with successional status and shade tolerance. Oecologia 94: 7–16.
•  White,  S. N.,  N. S.  Boyed and  R. C.  Van Acker. 2012. Growing degree-day models for predicting lowbush blueberry (Vaccinium angustifolium Ait.) ramet emergence, tip dieback, and flowering in Nova Scotia, Canada. HortScience 47: 1014–1021.
•  Witkowski,  E. T. F. and  B. B.  Lamont. 1991. Leaf specific mass confounds leaf density and thickness. Oecologia 88: 486–493.
•  Wright,  I. J. and  M.  Westoby. 1999. Differences in seedling growth behavior among species: trait correlations across species, and trait shifts along nutrient compared to rainfall gradients. J. Ecol. 87: 85–97.
•  Wright,  I. J. and  M.  Westoby. 2000. Cross-species relationships between seedling relative growth rate, nitrogen productivity and root vs leaf function in 28 Australian woody species. Funct. Ecol. 14: 97–107.
•  Wubs,  A. M.,  Y. T.  Ma,  E.  Heuvelink and  L. F. M.  Marcelis. 2012. Model selection for nondestructive quantification of fruit growth in pepper. J. Amer. Soc. Hort. Sci. 137: 71–79.
•  Yano,  T.,  K.  Matsubara,  M.  Shimooka,  A.  Tamanoi,  M.  Narahara,  M.  Kawano,  S.  Ito,  S.  Setoyama,  E.  Yasunaga,  T.  Araki and  M.  Kitano. 2014a. Energy-saving night temperature regime for Satsuma Mandarins (Citrus unshiu Marc.) grown in a plastic house with heating. I. Effect of temperature and water condition on fruit growth and quality. Environ. Control Biol. 52: 161–166.
•  Yano,  T.,  K.  Matsubara,  M.  Shimooka,  A.  Tamanoi,  M.  Narahara,  M.  Kawano,  S.  Ito,  S.  Setoyama,  E.  Yasunaga,  T.  Araki and  M.  Kitano. 2014b. Energy saving night temperature regime for Satsuma mandarins (Citrus unshiu Marc.) grown in a plastic house with heating. II. Effect of night temperature on fruit water and carbon balances. Environ. Control Biol. 52: 167–173.
•  Yano,  T.,  K.  Matsubara,  M.  Shimooka,  A.  Tamanoi,  M.  Narahara,  M.  Kawano,  S.  Ito,  S.  Setoyama,  E.  Yasunaga,  T.  Araki and  M.  Kitano. 2014c. Energy saving night temperature regime for Satsuma mandarins (Citrus unshiu Marc.) grown in a plastic house with heating. III. Application of different night temperature patterns. Environ. Control Biol. 52: 175–181.
•  Yoshida,  S. and  T.  Ishida. 1982. Raising method of young tree of Japanese pear using the steel wire system. Bull. Chiba Pref. Agr. Exp. Stn. 23: 49–57 (In Japanese with English abstract).