Introduction

Phosphorus (P) is one of the most growth-limiting nutrients in plants (Vance et al. 2003; Vitousek et al. 2010). Studies reveal that crops take up only 10–25% of the fertilizer P applied each season (Johnston et al. 2014), and the rest of the fertilizer P is strongly bound to soil particles (Hinsinger 2001; Shen et al. 2011). Roots grown in soil with low P availability showed high plasticity to enhance the capacity of P acquisition, including morphological and physiological changes (Vance et al. 2003; Lambers et al. 2006; Shen et al. 2011). Increased photosynthates allocation to root growth, enhanced lateral root emergence, and increased production of fine roots, are the typical morphological responses that enlarge the contact area between roots and soil (Wang et al. 2010; Lambers et al. 2011; Niu et al. 2012; Postma et al. 2014; Lynch 2015). In addition, increased proton release, and higher exudation of carboxylates and phosphatases secretion into the rhizosphere may alter soil properties and release soil-bound P (Li et al. 2007; Richardson et al. 2011; Li et al. 2014; Mehra et al. 2016).

Many studies have investigated the contribution of root morphological and physiological changes to increased P acquisition in the environments with low P availability (e.g. Pang et al. 2009; Wang et al. 2010; Deng et al. 2014; Lyu et al. 2016; Li et al. 2016a). However, less attention has been paid to understanding how the aboveground light influences these adaptations in root morphology and physiology. Light intensity influence photosynthesis, the translocation of photosynthates to roots, and thus the growth and morphology of the root system (Hermans et al. 2006; Hammond and White 2008, 2011). The accumulation of root biomass and total root length are enhanced when plants are grown under high light intensity (Wissuwa et al. 2005; Cheng et al. 2014; Zhou et al. 2019). In contrast, root growth is more sensitive in reduction than shoot under low light intensity, resulting in decreased root to shoot biomass ratio, as the photosynthates are distributed preferentially to shoots during leaf extension to increase the interception of light (Hébert et al. 2001; Gommers et al. 2013; Gundel et al. 2014). The root morphology is altered by the shoot-derived carbohydrate signals that have been suggested to play a role in the systemic control of the plant P-deficiency response. Exogenous supply of sucrose and increased light intensity both stimulate the formation of cluster roots in white lupin, even when plants were under sufficient P supply (Zhou et al. 2008; Cheng et al. 2014).

Light influences the expression of genes associated with the low P responses by regulating the synthesis and transport of photosynthates (Hammond and White 2008). For example, the expression of LaPEPC3 and LaPT1, which are induced by P starvation and associated with the exudation of citrate and P uptake, respectively, is upregulated when plants grow under high light intensity, regardless P supply (Cheng et al. 2014). Moreover, the expression of three P-responsive genes (LaPT1, LaSAP1 and LaMATE) is downregulated in cluster roots of plants grown in darkness under a low P environment, indicating that the supply of photosynthates regulates the expression of these genes (Liu et al. 2005). Furthermore, the expression of Atpht1;4, which encodes a P transporter induced by P starvation, is reduced when plants are kept in the dark, but the presence of exogenous sucrose in the growth medium can sustain high expression of Atpht1;4 (Karthikeyan et al. 2007). Light not only influences root morphology by regulating the synthesis and partitioning of photosynthates but also acts as a signal to directly regulate root growth (van Gelderen et al. 2018a). For example, far-red light detection in the shoot of Arabidopsis regulates lateral root growth via the HY5 transcription factor (Chen et al. 2016; van Gelderen et al. 2018b).

Maize, as an important food and feed, is widely cultivated around the world. Maize roots respond to P starvation by exhibiting extensive morphological alterations that contribute to the efficiency of P acquisition (Zhu et al. 2005, 2010; Zhang et al. 2012; Postma et al. 2014; Miguel et al. 2015). Similarly to many species of Poaceae, maize adapts to low soil P availability by altering root morphology, rather than root physiology (Lyu et al. 2016; Wen et al. 2017). However, whether and how the light intensity affects adaptive responses of root morphological traits to low soil P in maize has not been determined, especially in natural light conditions. Notably, most studies investigating the effects of light on root growth have been conducted in laboratories (e.g., Cheng et al. 2014; van Gelderen et al. 2018a), which might not reflect natural growth conditions. For example, a light intensity of 200 μmol m−2 s−1 is typical for the cultivation of plants in the laboratory, which is considerably lower than natural light intensities found under field conditions (light intensity usually reaches more than 1000 μmol m−2 s−1 in the field).

The main purpose of this study was to investigate how the light intensity influences maize root morphology, especially under low soil P conditions. In addition, we characterized the expression of low P tolerance transcript factor (ZmPTF1), which is associated with root growth by regulating carbon metabolism in leaves and roots (Li et al. 2011). The expressions of ZmSPS1, which encodes the key enzymes in the synthesis of sucrose in leaves, and ZmSS2, which is associated with sucrose decomposition in roots (Li et al. 2011), were also evaluated. Specifically, this study aimed to (1) quantify the dynamics of maize root growth in soil with variable P supply and under natural (NL) and low light (LL, 40% of natural light intensity) conditions; and (2) characterize how maize integrates the light intensity and P supply cues to improve root growth and P uptake in soil with P availabilities ranging from deficient to optimal to excess.

Materials and methods

The experiment was conducted in a naturally-lit rain-shed at the research farm of Sichuan Agricultural University, Chengdu (30°42′ N, 103°51′ E). The height of the rain-shed was 6 m, and the top of the shed was covered with thin and transparent plastic sheets to exclude rain. There was no wall around the rain-shed to maintain the similar air temperature inside and outside.

There were nine P application rates (0, 2.5, 5, 10, 25, 50, 75, 150 and 300 mg P kg−1 soil, supplied as KH2PO4) and two light intensity treatments, natural light intensity (NL) and low light intensity (LL, 40% of natural light intensity). The crop growth was limited by light intensities lower than 50% of the natural under field conditions in the region of this study (Yang et al. 2014). Thus, a 40% natural light intensity was established as the low light intensity treatment by using shade cloth (Table S1). The shade cloth was supported by a stainless steel framework (3 × 3 × 3 m3). Under the shaded space, a time-controlled slow-speed fan was used to keep the air flow in the plant canopy. The periodic fanning was set as 30 min on and 30 min off during the day, and 30 min on and 60 min off during the night. The experiment was conducted from July 1st to August 5st, 2018. The daily light intensity and maximum and minimum air temperature during the plant growth period were obtained from the meteorological station located nearby the experimental rain-shed, as shown in Fig. 1.

Fig. 1
figure 1

Daily solar radiation (a) and minimum/maximum air temperature (b) during maize growth period, from July 1st to August 5st in 2018

Full size image

A sandy clay loam soil with low P availability was collected from a long-term fertilizer experiment in Renshou (30°06′N, 104°2′E). The soil was air-dried, sieved to 2-mm and thoroughly mixed. The soil properties were as follows: pH 6.72 (1: 2.5, soil: water), Olsen-P 2.6 mg kg−1, organic matter 8.6 g kg−1, total N 0.31 g kg−1, available K 84 mg kg−1, and available N 80 mg kg−1. Each pot (Φ 20 × 30 cm3) was filled with 5 kg of air-dried soil. To ensure the nutrient supply was adequate for plant growth, soil was fertilized with basal nutrients at the following rates (mg pot−1): Ca(NO3)2·4H2O 8435, CaCl2 630, MgSO4·7H2O 215, EDTA-FeNa 29, MnSO4·4H2O 33.5, ZnSO4·7H2O 50, CuSO4·5H2O 10.0, H3BO3 3.35, and (NH4)6Mo7O24·4H2O 1.30. The P treatments were imposed by increasing the application of KH2PO4, and the equivalent K concentrations (1675 mg K pot−1) were imposed using K2SO4. The fertilizers were manually mixed in the soil and homogenized 4 weeks before planting. After fertilization and homogenization, the soil Olsen-P concentrations were determined (Table 1).

Table 1 Shoot and root dry weight of maize plants grown with different soil Olsen-P content under natural light intensity (NL) and low light intensity (LL, 40% of natural light intensity)
Full size table

Seeds of maize (Zea mays L.) cultivar CD418 (widely cultivated in the region) were surface-sterilized in 30% v/v H2O2 for 20 min and washed with deionized water before planting. Four uniformly germinated seeds were planted per pot, and the seedlings were thinned to two plants per pot at the 2-leaf stage. Three replicates were grown for each treatment, and pots were arranged in a split-plot design. The main plots consisted of NL and LL treatment, and the subplots were nine P application rates. All pots were watered daily until harvest.

Plant harvest and measurements of root morphology

After 36 d of growth (five leaf stage), the plants were harvested. The shoot and root subsamples were frozen in liquid nitrogen and stored at −80 °C until analysis of gene expression or sucrose concentration, and same subsamples were oven-dried at 80 °C for 3 d to measure dry weight and P concentration. Root subsamples were washed with deionized water and scanned using an Epson Perfection V700 Photo scanner (Japan). Images were analyzed using WinRHIZO (WinRHIZO Pro2004, Canada) to measure root length and surface area. Fine roots were defined as roots with a diameter less than 0.2 mm (Jing et al. 2010).

Determination of net photosynthetic rate, P concentration, and sucrose concentration

Net photosynthetic rate (Pn) was measured on a youngest fully expanded leaf using a Li6400 photosynthesis system (Li-COR, Lincoln, NE, USA). Measurements were performed between 10:00 a.m. and 12:00 p.m. at the five-leaf stage.

The P concentrations were determined in leaves, stems and roots. The dry material was ground to pass through a 0.149-mm mesh sieve, a 0.3-g sample was digested with concentrated H2SO4 (5 mL) and H2O2 (30% v/v) (1 mL), and P was determined by the vanadomolybdate method (Page 1982).

The freeze-dried leaves and roots (FDU-2110, EYELA, Japan) were extracted in 80% v/v ethanol. Sucrose was measured directly in the extract, using resorcinol as the color reagent (Shi et al. 2016).

Quantitative real-time PCR analysis

Total RNA was isolated and treated with DNase I, and cDNA was synthesized to perform quantitative real time RT-PCR using the SYBR Green PCR Master Mix in the QuantStudio 6 Flex real-time PCR detection system (Thermo Fisher Scientific, USA) with appropriate primers. Relative expression of genes compared to the ACTIN was calculated using the ΔΔCt method (Schmittgen and Livak 2008).

The primers used to quantify gene expression were as follows: ZmPTF1, 5′-ACACGGAAGAATCGGAAGACAC-3′ and 5′-GGATATGGCGGAAGCAAGG-3′; ZmSPS1, 5′-ATGAGGGAGCGGCTGAGGAT-3′ and 5′-GAGTGCCTGTGACCTTGATGCT-3′; ZmSS2, 5′-CTGCCTTTACTCTGCCTGGTCT-3′ and 5′-TTGGCCTTCTCGGTATGTGG-3′; Zm38, 5′-TGGATCAACTACCTCCGGCC-3′ and 5′-GCGATGAGCGACCACTTGTT-3′; ZmACTIN, 5′-ATGTTGCTATCCAGGCTGTTCTT-3′ and 5′-TGTATCCTTCATAGATTGGAACCGT-3′.

Statistical analyses

Two-way analysis of variance was performed using the SPSS statistical software package (Version 19.0, SPSS Institute Inc., USA), and significant differences among P application rates were assessed using the LSD multiple range analysis test (p ≤ 0.05) and using t tests to assess the significant differences between NL and LL treatments (p ≤ 0.05).

The linear-plateau model was used to establish the relationship between shoot dry matter, root/shoot ratio and soil Olsen-P (Bai et al. 2013). Empirical polynomial (inverse third order) equations were used in Origin (Origin 16.0, USA) to analyze the relationship between root morphological traits (i.e., total root length, root surface area or proportion of fine roots) and soil Olsen-P (Deng et al. 2014).

Fertilizer P-use efficiency (PUE) was used to evaluate the P absorption efficiency of maize from P fertilizer applied to soil. PUE was calculated as follows:

$$ \mathrm{PUE}\ \left(\%\right)=\frac{U_f-{U}_0}{P_f}\times 100 $$

where Uf is plant total P uptake by the P-treated crops, U0 is plant total P uptake by the crops treated with zero added P, and Pf is the amount of P fertilizer applied. The gap in PUE between the NL (NLPUE) and LL treatments (LLPUE) was calculated as ΔPUE = NLPUE-LLPUE.

Results

Plant growth and biomass allocation

With the P application rate increasing from 0 to 300 mg P kg−1 soil, the soil Olsen-P increased from 2.6 to 79.0 mg kg−1 (Table 1). Compared to the zero P application, the P supply treatments increased shoot dry weight by up to 460% (NL) and 208% (LL), and increased root dry weight by up to 102% (NL) and 36.3% (LL), respectively (Table 1). Shoot and root dry weights were significantly higher in the NL compared with LL when plants were grown with P supply, but were not significantly affected by light intensity when plants were grown at 0-P addition (Table 1).

There was a strong positive correlation between shoot dry weight and soil P availability, with the soil Olsen-P increasing up to approximately 13.5 (NL) and 7.4 mg kg−1 (LL) and then leveling off (Fig. 2a). The ratio of root to shoot dry weight declined as soil Olsen-P increased to 10.8 (NL) and 7.3 mg kg−1 (LL) and then leveled off with a further increase in soil Olsen-P (Fig. 2b). Moreover, the root/shoot ratio was higher in the NL than LL treatments only when the soil Olsen-P was below 15.9 mg kg−1 (Table S2).

Fig. 2
figure 2

Effects of P supply and light intensity on shoot dry weight (DW) (a) and the ratio of root to shoot DW (b). Each symbol represents the mean (±SE) of three replicates. NL = natural light intensity, LL = low light intensity (40% of natural light intensity)

Full size image

Root morphological traits

In the NL treatment, the total root length initially increased with increasing soil Olsen-P, peaking when the soil Olsen-P was approximately 7.2–15.9 mg kg−1, and then gradually declined to a plateau (approximately 5100 cm plant−1) (Fig. 3a). In the LL treatment, the total root length initially increased with increasing soil Olsen-P up to approximately 20 mg kg−1 and then leveled off (Fig. 3a). The effects of light intensity on root growth were more pronounced when plants were grown with low, rather than adequate P supply or zero-P supply. In the treatment with soil Olsen-P of 9.7 mg kg−1, the total root length was 2-fold greater in NL than LL, but the total root length in NL and LL with no significant difference when soil Olsen-P was higher than 39.6 mg kg−1 (Table S2).

Fig. 3
figure 3

Maize root morphological traits in response to increasing soil Olsen-P and light intensity: (a) total root length, (b) fine root (diameter ≤ 0.2 mm) percentage over total root length, (c) total root surface area. Each symbol represents the mean (±SE) of three replicates. NL = natural light intensity, LL = low light intensity (40% of natural light intensity)

Full size image

The relationships between soil Olsen-P and root surface area or the fine root percentage were similar to total root length: in the NL treatment, the values of each of the root traits peaked in the Olsen-P range of 7.2–15.9 mg kg−1 followed by a decline to a plateau with a further increase in soil Olsen-P (Fig. 3b, c). In the LL treatment, root surface area initially increased with increasing soil Olsen-P up to the critical value and then leveled off (Fig. 3c). The effect of light intensity was obvious when plants were grown with a low (6.7 to 15.9 mg kg−1), rather than with an adequate P supply. However, the fine root percentage decreased with increasing soil Olsen-P under LL (Fig. 3b). Additionally, in the treatments with soil Olsen-P ranging from 6.7 to 15.9 mg kg−1, the fine root percentage was higher in NL than LL (Table S2).

Shoot P status, plant P uptake and fertilizer P use efficiency

Leaf P concentration increased with increasing soil Olsen-P up to 15.9 mg kg−1 and then leveled off regardless of light intensity (Fig. 4a). With increases in light intensity from LL to NL, leaf P concentration decreased in the treatments with relatively low soil Olsen-P, but not in the treatments with zero-P addition or soil Olsen-P higher than 20.6 mg kg−1 (Fig. 4a). Root P concentrations in the treatments were lower under NL than LL regardless of P supply (Fig. 4b). The ZmPTF1 (associated with root growth) was highly expressed under P deficiency, and the increased light intensity stimulated the expression further (Fig. 4d). In addition, the expression of Zm38 (a low P response transcription factor) shared the same trend as ZmPTF1 (Fig. 4e). Indeed, P acquisition significantly increased from LL to NL when P had been supplied (Fig. 4c, f).

Fig. 4
figure 4

Effects of P supply and light intensity on maize P status (a: leaf P concentration, b: root P concentration, c: shoot P content, f: root P content) and the expression of the low-P tolerance transcription factor ZmPTF1 (d) and low-P response gene Zm38 (e). Each symbol represents the mean (±SE) of three replicates. NL = natural light intensity, LL = low light intensity (40% of natural light intensity). For ZmPTF1 expression, data are expressed as relative values based on the expression of ZmPTF1 in roots grown with soil Olsen-P 39.6 mg kg−1 under LL referenced as 1.0. For Zm38 expression, data are expressed as relative values based on the expression of Zm38 in root grown with soil Olsen-P 39.6 mg kg−1 under LL referenced as 1.0. *and** indicated significant difference between the NL and LL treatments at a given soil Olsen-P content (t-test at p ≤ 0.05 and p ≤ 0.01, respectively)

Full size image

The fertilizer P use efficiency (PUE) decreased with increasing P fertilizer, regardless of light intensity. However, in the treatments with soil Olsen-P below 39.6 mg kg−1, the PUE was significantly higher in NL than in LL (Fig. 5a). The PUE gap between NL and LL reached a maximum (22.5%) when soil Olsen-P was 7.2 mg kg−1, but declined with increasing soil Olsen-P (Fig. 5b).

Fig. 5
figure 5

Effects of P supply and light intensity on maize fertilizer P use efficiency (a: PUE, b: the PUE gap between the treatments NL and LL). Each symbol represents the mean (±SE) of three replicates. NL = natural light intensity, LL = low light intensity (40% of natural light intensity). ** indicated significant difference between the NL and LL treatments at a given soil Olsen-P content (t-test at p ≤ 0.01). Letters are significantly different among P application rates at p ≤ 0.05

Full size image

Photosynthetic efficiency and carbohydrate accumulation in leaves and roots

In the NL treatment, net photosynthesis (Pn) increased with increasing soil Olsen-P up to 39.6 mg kg−1, and then leveled off (Fig. 6a). However, the Pn showed no response to P supply in the LL treatment (Fig. 6a). Net photosynthesis increased significantly from LL to NL regardless of P supply (Fig. 6a).

Fig. 6
figure 6

Effects of P supply and light intensity on maize net photosynthetic (Pn) rate (a), sucrose status (b: leaf sucrose concentration, c: root sucrose concentration), ratio of root to shoot sucrose (d), and the expression of the key enzymes in the sucrose synthesis (e: ZmSPS1, f: ZmSS2). Each symbol represents the mean (±SE) of three replicates. NL = natural light intensity, LL = low light intensity (40% of natural light intensity). For ZmSPS1 expression, data are expressed as relative values based on the expression of ZmSPS1 in leaves of plants grown with soil Olsen-P 2.6 mg kg−1 under LL referenced as 1.0. For ZmSS2 expression, data are expressed as relative values based on the expression of ZmSS2 in roots of plants grown with soil Olsen-P 79.0 mg kg−1 under LL referenced as 1.0. ** indicated significant difference between the NL and LL treatments at a given soil Olsen-P content (t-test at p ≤ 0.01)

Full size image

The leaf sucrose concentration generally followed similar trends as Pn under the variable soil Olsen-P and light treatments (Fig. 6b): Leaf sucrose concentration was significantly higher in the treatments under NL than LL. Moreover, leaf sucrose concentration was also increased with increasing P supply, especially under NL (Fig. 6b). The gene ZmSPS1 encoding the key enzymes participating in the synthesis of sucrose in leaves was expressed more strongly under NL than LL, and the expression was also stimulated with low-P supply (Fig. 6e).

The root sucrose concentration increased from LL to NL, regardless of P supply (Fig. 6c). Moreover, this concentration was higher in the treatments with low than sufficient P supply. The ZmSS2 (involved in sucrose decomposition in root) was highly expressed under low P conditions, and decreasing light intensity stimulated the expression of ZmSS2 further (Fig. 6f). Additionally, the ratio of root to shoot sucrose was higher in NL than LL with soil Olsen-P values ranging from 2.6 to 15.9 mg kg−1 (Fig. 6d).

Correlations between total root length and P concentration in leaves, and sucrose concentration in roots

The root sucrose concentration was inversely related to leaf P concentration: the root sucrose concentration decreased with increasing leaf P concentration regardless of light conditions (Fig. 7a). The total root length was positively related to the leaf P concentration and sucrose concentration in roots in the NL treatment, and the relationships between root length and leaf P or root sucrose were approximated to a parabola (Fig. 7b, c). In the LL treatment, total root length was negatively related to the leaf P concentration and sucrose concentration in roots (Fig. 7b, c).

Fig. 7
figure 7

Correlations between leaf P concentration and root sucrose concentration (a). Correlations between total root length and leaf P concentration (b) or root sucrose concentration (c). Data points represent individual replicates. NL = natural light intensity, LL = low light intensity (40% of natural light intensity)

Full size image

Discussion

Maize growth in response to variable soil P availability and light intensity

Maize growth first increased with an increase in soil Olsen-P and then leveled off when soil Olsen-P surpassed the critical value. For example, the critical soil Olsen-P values ranged from 13 to 15 mg kg−1 for maize and 11 to 15 mg kg−1 for wheat in long-term P application experiments in China (Tang et al. 2009). In Europe, the critical soil Olsen-P values ranged from 7 to 18 mg kg−1 (Poulton et al. 2013; Colomb et al. 2007; Bollons and Barraclough 1999). The critical soil Olsen-P value for maize (14. 6 mg kg−1) grown in northwestern China with high solar radiation was higher than that in the same region (11.1 mg kg−1) of this study with low solar radiation (Bai et al. 2013). The data reported here indicated the critical soil Olsen-P value for shoot dry weight production was higher under NL (13.5 mg kg−1) than LL (7.4 mg kg−1) (Fig. 2a). This result may indicate the promotion of high light intensity in root growth and P acquisition (Hammond and White 2008; Gao et al. 2017).

Many previous studies have reported that both P-deficiency and high light intensity increase the ratio of root to shoot dry weight, which was associated with increased carbohydrate concentration in roots (Cruz 1997; Hébert et al. 2001; Ciereszko and Barbachowska 2000; Hammond and White 2008, 2011). Similarly, in the study presented here P-deficiency increased the ratio of root to shoot dry weight and root sucrose concentration; moreover an increase in light intensity from LL to NL enhanced the root sucrose concentration further (Table 1, Figs. 2b and 6c). However, the light intensity had no significant effects on the root to shoot ratio when the soil Olsen-P surpassed the critical value (Fig. 2b). The soil Olsen-P for maize growth in the present study ranged from 2.6 to 79.0 mg kg−1, covering deficiency, optimum and excess (Table 1). As expected, the plant growth showed a positive response to increased light intensity and P supply (Table 1, Figs. 2b and 6d), suggesting that the experimental conditions were suitable for estimating the root morphological responses to the variable soil P availability and light intensity.

Root morphological response to leaf P concentration

In many species of Poaceae (such as maize, wheat, and rice), root morphological plasticity is the major response to low P stress and plays an important role in P uptake (Pearse et al. 2006, 2007; Nuruzzaman et al. 2006; Lyu et al. 2016; Wen et al. 2017). It is generally accepted that the root morphological plasticity response to P-deficiency was mainly triggered by internal P status (leaf P status) rather than external P concentration (Li et al. 2008; Lambers et al. 2011; Pang et al. 2009). For example, the leaf P status regulated the formation of cluster roots in white lupin grown with a divided root system, in which one root half was supplied with P and the other half was not (Shane et al. 2003). Moreover, leaf P status also regulated the root morphological (root/shoot ratio, fine root percentage, and specific root length) traits of maize in responses to low P stress (Wen et al. 2017). In the present study, the expressions of ZmPTF1 (associated with root growth) (Fig. 4d) and Zm38 (low P response gene) (Fig. 4e) was corresponded well with the P concentration in leaves, not in root (Fig. 4a, b), suggesting that root growth might be affected by a leaf P status.

In the present study, a wide range of root morphological changes were observed in response to altered light intensity under variable P environments (Table S2, Fig. 3). Leaf P concentration was the lowest in the treatments with soil Olsen-P of 2.6 mg kg−1, regardless of light intensity (Fig. 4a), and the total root length and root surface area were severely inhibited, with no significant difference between NL and LL (Table S2). These findings suggested leaf growth and photosynthesis were impaired at extremely low P (Mollier and Pellerin 1999; Plénet et al. 2000), which subsequently inhibited root growth and function. Maize had higher total root length (Fig. 3a), total root surface (Fig. 3c) and fine root percentage (Fig. 3b) when grown at low Olsen-P (6.7 to 15.9 mg kg−1) under NL than LL. These finding might have been due to an increase in light intensity from LL to NL increasing both Pn (Fig. 6a) and carbon fixation (Fig. 6b), followed by a growth-induced P-starvation signal (Fig. 4a) enhancing root morphological responses to low P stress. These results were consistent with findings in lupin, soybean, and maize, whereby plants with more light capture increased the photosynthesis and shoot growth, thus (1) increasing shoot P demand that might have exceeded the capaclity of roots to supply P to leaves, (2) causing growth-induced P-starvation in the shoot, and (3) producing a systemic signal to induce root morphological responses to low P stress (Cheng et al. 2014; Zhou et al. 2019). Fine roots in maize are critical for nutrient uptake (Zhang et al. 2012; Li et al. 2016b) and are regulated by the leaf P status (Wen et al. 2017). High light intensity increases the percentage of fine roots may be an efficient strategy for maize to adapt to P-deficiency (Fig. 3b), with increased total root length and surface area to improve P uptake capacity achieved at a minimal carbon cost (Zobel et al. 2007; Pang et al. 2009; Lambers et al. 2006; Lynch 2015). The low-P tolerance transcription factor ZmPTF1, which promotes root growth, was expressed in roots at a higher level under NL than LL, and corresponded well with the P concentration in leaf (Fig. 4a, d), providing additional evidence that relatively low P concentrations in leaves under NL than LL (Fig. 4a) enhanced root growth.

With soil Olsen-P above 20.6 mg kg−1, the root traits (total root length, total root surface, and fine root percentage) were relatively low and stable (Table S2, Fig. 3). Moreover, leaf P concentrations were relatively high, and expressions of both ZmPTF1 and Zm38 were relatively low; for all these variables there was no significant difference between NL and LL (Fig. 4a, d, e), suggesting that the effects of light intensity on root growth decreased at sufficient and excess P (with soil P availability not a limiting factor) compared to low P supply (see also Chiou and Lin 2011; Wen et al. 2017).

The quadratic models fit the relationship between total root length and leaf P concentration well in the NL treatment, with the low leaf P concentration (up to 2.3 mg g−1), inducing root growth (Fig. 7b). However, in the LL treatment, the total root length was not positively related to leaf P concentration (Fig. 7b), which indicated that P was not the main limiting factor under low light intensity (cf. Wissuwa et al. 2005; Cheng et al. 2014). A similar experiment (Wen et al. 2017) conducted with a wide range of soil showed that when the maize leaf P concentration ranged from 1 to 2 mg g−1, the total root length, specific root length, and fine root percentage increased with increasing leaf P; in contrast, the root morphological traits were relatively low and stable when the leaf P concentration was higher than 2.7 mg g−1. The critical leaf P values between the two experiments were different, as the various results might be due to different maize genotypes and growth conditions (i.e., temperature and light intensity) or soil properties. However, these results indicated that light intensity stimulated the root morphological traits response to P starvation by influencing leaf P status.

Root morphological responses to carbohydrate metabolism in leaves and roots

Plants with insufficient P showed reduced shoot growth rates and translocated carbohydrate in excess of their shoot growth capacity to roots, which resulted in relatively high carbohydrate concentration in roots (Rychter and Randall 1994; Hermans et al. 2006; Hammond and White 2008, 2011; Niu et al. 2012). Similarly, increasing light intensity also increased translocation of carbohydrates to roots (Cruz 1997; Hébert et al. 2001; Lemoine et al. 2013; Zhou et al. 2019). Stem girdling has been used to explore the importance of photosynthates in P-deficiency signaling. For example, in white lupin stem girdling to block phloem translocation to roots resulted in a significant reduction in the expression of the low P response genes LaPT1 and LaSAP1 (Liu et al. 2005). Similar results were also obtained in common bean (Liu et al. 2010). Taken together, the results of these studies suggest that carbohydrates (such as sucrose) serve as both a low P signal and a carbon-substrate for roots growth to increase the P foraging capacity (i.e. low P adaptations) (Hermans et al. 2006; Müller et al. 2007; Hammond and White 2008; Liu et al. 2005; Zhou et al. 2008). The findings of the study presented here were consistent with the observations that both P-deficiency and increasing light intensity increased translocation of sucrose to roots (Fig. 6d) and root sucrose concentration (Fig. 6c). However, how maize integrates sucrose- and P-based regulation of root growth with a wide range of soil P availability requires further investigation.

Increased light intensity from LL to NL increased the net photosynthetic rate of maize with soil Olsen-P covering deficiency, optimum and excess. However, net photosynthesis was impaired when maize was grown with extremely low soil P availability regardless of light intensity (Fig. 6a), which was consistent with the results obtained with other species (Rao et al. 1990; Wissuwa et al. 2005; Cheng et al. 2014). The root sucrose concentration was the highest in the treatment with soil Olsen-P of 2.6 mg kg−1, regardless of light intensity (Fig. 6c), but the root growth was severely inhibited (Table S2) due to severe root P-starvation (Fig. 4b). Similarly, in rice and white lupin under extreme P-starvation, the carbohydrates supply was not limiting root growth (Wissuwa et al. 2005; Cheng et al. 2014).

When soil Olsen-P increased from 6.7 to 15.9 mg kg−1, more photosynthates were allocated to roots as sucrose under NL than LL (Fig. 6c, d), which was associated with greater total root length and fine root percentage (Fig. 3a, b). Enhanced translocation of carbohydrates to roots and enhanced total root length were the universal responses to P deficiency in maize (Mollier and Pellerin 1999; Zhang et al. 2012; Deng et al. 2014) because P-deficient plants invested more carbon to root growth to improve the P acquisition capacity (Fernandez and Rubio 2015; Hammond and White 2008; Ruan 2014).

In excess P supply, natural light did not enhance the allocation of sucrose to the roots between the NL and LL treatments (Fig. 6d). Furthermore, there was also no significant difference in the total root length between the two light treatments (Table S2). One possible explanation might be that P-sufficient plants lacked a P-starvation signal associated with increased carbon investment to root growth and increased P acquisition capacity (Shipley and Meziane 2002; Poorter et al. 2012). Like the expression of ZmPTF1, which is associated with root growth by allocating carbon to root, was downregulated and showed no significant difference between NL and LL in the treatments with sufficient P supply (Fig. 4d). The quadratic model fit the relationship between total root length and root sucrose concentration well in NL (Fig. 7c), indicating that light intensity enhanced the root morphological adaptations response to P starvation by regulating root sucrose concentration.

Direct root morphological responses to light conditions

Stimulation of root growth by increased light intensity might not be controlled exclusively by increased allocation of sucrose to roots; instead, other systemic signals (such as phytohormone) could also contribute to root growth. For example, auxin was found to stimulate root growth in maize, particular in low P conditions (Martínez-de la Cruz et al. 2015; Li et al. 2018; Zhang et al. 2019). High light intensity enhanced sensitivity of cluster root formation to auxin by downregulating the expression of an auxin-response repressor gene (Cheng et al. 2014), whereas sucrose increased the sensitivity of lateral root development to auxin in P-deficient Arabidopsis (Jain et al. 2007). In the present study, the expression of the auxin response factor (ARF), which activates the auxin-induced gene expression response, was found to be upregulated under NL, especially under low P conditions (data not shown). In addition, light could also enhance the biosynthesis of indole-3-acetic acid (Bhalerao et al. 2002): specifically, light increases flavonol biosynthesis, which increases root auxin levels (Buer and Muday 2004; Silva-Navas et al. 2016). Except for the sensitivity and formation, light can also affects the transportation of auxin in the root by controlling the removal of PIN proteins from the plasma membrane via the process of vacuolar degradation (Korbei and Luschnig 2013). Further work is needed to elucidate a role of auxins in the light- and P-availability-dependent regulation of root growth.

Conclusion

Our results demonstrated that an increase in light intensity enhanced total root length, root surface area, and fine root percentage, underpinning increased P acquisition and P use efficiency, especially at low soil P availability (soil Olsen-P from 6.7 to 15.9 mg kg−1). These effects diminished in soil with either severely low P (soil Olsen-P 2.6 mg kg−1) or excess P (soil Olsen-P above 20.6 mg kg−1). Maize may integrate the environmental cues (P supply and light intensity) to regulate leaf P status and root sucrose concentration to determine plant growth. It is possible that photosynthesis and the leaf P status coordinate the morphological responses of roots by controlling the transfer of carbohydrates to roots.