Tolerance to low phosphorus in rice varieties is conferred by regulation of root growth

Yaping Deng, Chuanbao Men, Shengfeng Qiao, Wenjie Wang, Junfei Gu, Lijun Liu,Zujian Zhang, Hao Zhang, Zhiqin Wang, Jianchang Yang

Jiangsu Key Laboratory of Crop Genetics and Physiology/Jiangsu Key Laboratory of Crop Cultivation and Physiology, Jiangsu Co-Innovation Center for Modern Production Technology of Grain Crops,Yangzhou University,Yangzhou 225009, Jiangsu,China

Keywords: Rice (Oryza sativa L.) Tolerance Phosphorus use efficiency Root traits

ABSTRACT Phosphorus use efficiency (PUE) can be improved through cultivation techniques and breeding. However, little is known about rice (Oryza sativa L.) agronomic and physiological traits associated with high PUE.We characterized the agronomic and physiological traits of rice varieties with different tolerances to low phosphorus in nutrient solution.Two varieties with strong tolerance to low phosphorus(STVs)and two with weak tolerance(WTVs)were grown at normal (NP, control) and low phosphorus (LP, 1/20 of NP) concentrations. Plants grown at LP produced significantly lower grain yield than those grown at NP. WTVs yields were lower than STVs yields. Compared to NP, LP significantly increased phosphorus translocation efficiency(PTE),internal phosphorus efficiency(IPE)and phosphorus harvest index(PHI).Under the LP condition,PTE and IPE were higher for STVs than for WTVs.LP also reduced tiller number, shoot biomass, leaf area index (LAI), leaf photosynthetic rate, and mean root diameter of both kinds of varieties at the main growth stages, but to a lower extent in STVs. LP significantly increased the number of productive tillers, root biomass,root-shoot ratio, root bleeding rate, and root acid phosphatase (RAP) activity. Total root length, root oxidation activity (ROA), and root total and active absorbing surface areas for STVs were significantly increased under LP,whereas the opposite responses were observed for WTVs.Total root length,ROA,root bleeding rate,root active absorbing surface area,and RAP activity were positively and significantly correlated with grain yield, PTE, and IPE.These results suggest that the tolerance of rice varieties to a low-phosphorus growth condition is closely associated with root growth with higher biomass and activity.

1. Introduction

Rice (Oryza sativa L.), a staple food crop for ＞3.5 billion people worldwide,demands large amounts of nitrogen(N),phosphorus(P)and potassium(K)[1-3].Heavy application of inorganic fertilizer supported increased rice yields in the last half century [4-6]. P, as an essential plant macronutrient, plays a key role in regulating plant life processes including nucleic acid, membrane lipid, and protein synthesis and energy metabolism[7].P availability in soil is low owing to its binding to organic compounds and minerals, and a shortage of available P often limits growth development and formation of yield and quality in crops unless P is supplied as fertilizer[8,9]. Although phosphate fertilization has lessened the problem to date, it is not sustainable, as phosphate rock used in fertilizer production is a finite resource [10]. In agricultural production, phosphate fertilizer consumption reached 43.8 million tons in 2015 and may increase to 52.9 million tons by 2030[11].However,only approximately 20-30% of applied P can be used by cultivated plants,and the rest is lost by leaching into ground and surface water[9,12].To reduce dependence on high phosphate input,efforts should be made to breed and identify rice varieties with high phosphorus use efficiency(PUE).

Wide differences in P uptake and utilization among rice genotypes [13-16] suggest the potential for identifying rice genotypes with high efficiency in P uptake or utilization.Generally, hydroponics culture, soil culture and sand culture are used for coarse screening at the seedling stage, and the genotypes preliminarily identified are re-screened and tested at the seedling or maturity in the field. Of these cultivation methods, hydroponic culture is widely used for convenience in controlling P concentration[17,18].A shortage of available P leads to decreased tiller number, dry matter accumulation and distribution, and leaf photosynthetic capacity, finally reducing rice grain yield.Compared to the growth of varieties with weak tolerance to low phosphorus, the aboveground growth of varieties with strong tolerance was less affected[19].However,most studies of the response of rice varieties to low phosphorus have focused mainly on the seedling stage or a defined growth stage, and little is known about changes in rice agronomic and physiological traits throughout the growth period under a low-phosphorus condition. It has thus been difficult to draw a direct relationship between tolerance to low phosphorus and grain yield.

The root system is the main organ for nutrient acquisition, and changes in root morphology and physiology are crucial for P effective acquisition and utilization under a low-phosphorus condition [20,21]. Root morphology and physiology are closely associated with the growth and development of aboveground parts [22-24]. High root biomass and root oxidation activity in roots are required for achieving high panicle number, spikelets per panicle,grain-filling percentage, and grain yield [25]. One means of improving crop growth and PUE under low phosphorus is to increase P acquisition efficiency by improving root traits[26,27]. Plants often improve their P acquisition in soil by changing their root morphology, including by root elongation, root axis thinning, and increases in the number and density of root hairs [28,29]. In many studies, root weight and total root length, diameter, quantity, and density were positively correlated with P uptake and accumulation.Plants could adapt to low-P environments by adjusting their root-shoot ratio [30,31]. Root physiological activity,including root oxidation activity (ROA), root absorbing surface area, and root acid phosphatase (RAP) activity, also plays a key role in P availability in soil and P uptake by roots[32,33]. Plant growth and development and yield formation are closely associated with root morphology and physiology[23,25]. However, little is known about the effect of low phosphorus on root morphological and physiological traits during the full growth period and their relationships with grain yield and PUE.

The objective of this study was to test the hypothesis that an improved root system contributes to high PUE and grain yield of a rice variety with strong tolerance to low phosphorus.Changes in morphological and physiological traits of rice shoots and roots during the growing season were observed,and correlations of root traits with grain yield and PUE were determined.

2. Materials and methods

2.1.Plant materials and growing conditions

Hydroponic experiments were conducted in 2017 and repeated in 2018 at an experimental farm in Yangzhou University, Jiangsu, China (32°30′N, 119°25′E) during the rice-growing season (May to October). Two varieties with strong tolerance to low phosphorus (STVs), Yanjing 2 (YJ2,japonica rice) and Yongyou 2640 (YY2640, indica-japonica hybrid rice) and two with weak tolerance (WTVs), Zhendao 88 (ZD88, japonica rice) and Liming (LM, japonica rice) were grown using the water culture method described by Mae and Ohira [34] (Table 1). The low-phosphorus tolerance index([grain yield of tested variety under low phosphorus×(grain yield of tested variety under low phosphorus/grain yield of tested variety in control)] / mean grain yield of all tested varieties under low phosphorus) and dry matter index (shoot dry weight of tested variety under low phosphorus/shoot dry weight of tested variety in control)were chosen as indexes of tolerance to low phosphorus by rice varieties in our earlier work [35]. The low-phosphorus tolerance index and dry matter index were compared with the evaluation indexes of drought resistant varieties [36].All tested varieties were classified into three categories based on the above two indexes: strong tolerance to low phosphorus (both low-phosphorus tolerance index and dry matter index ＞1), medium tolerance to low phosphorus(both low-phosphorus tolerance index and dry matter index＞0.6 and either low-phosphorus tolerance index or dry matter index ＜1), and weak tolerance to low phosphorus(both low-phosphorus tolerance index and dry matter index≤0.6) [35]. In this study, two groups of varieties, STVs and WTVs,were selected for comparison of their agronomic and physiological traits.In both years,seedlings were sown in a seedbed on May 12 and transplanted on June 10 into concrete tanks located in a farm field. The hill spacing was 0.20 m × 0.16 m with two seedlings per hill. The volume of each tank was 6.4 m3(8 m long, 2 m wide, 0.4 m high). To ensure uniform nutrition in the tanks,the nutrient solution was circulated with a pump from transplanting until harvest. The dates of mid-tillering, panicle initiation, and heading were recorded (Table S1). All plots were harvested 55-60 days after heading date.

2.2.Treatment

Two phosphorus levels, the normal phosphorus concentration (NP, nutrient solution formulation described above,control) and low phosphorus concentration (LP, 1/20 of the NP),were tested.In earlier testing of 1/4, 1/8, 1/16, 1/20, and 1/30 of NP, tested rice varieties showed significant differences under 1/20 NP and the low-phosphorus tolerance index and dry matter index were chosen to evaluate tolerance to low phosphorus in rice cultivars at the same P level. The treatments were laid out in a split-plot design,with each of the treatments applied to two plots (tanks) as replicates. H3PO4was used to control the supply of phosphorus. The nutrient solution formulation was otherwise identical in LP and NP. The solution was changed once weekly and its pH was adjusted daily to 5.0 with 1 mol L−1HCl or 1 mol L−1NaOH.

Table 1-Hydroponic culture method.

2.3. Sampling and measurement

After the mean stem number in each plot was recorded,plants of 10 hills were sampled for measurements of shoot and root biomass, total root length, mean root diameter,ROA, root total and active absorbing surface area, root bleeding and root acid phosphatase (RAP) activity at midtillering,panicle initiation,heading time,and maturity.

Plants from two hills were used for measurements of total root length and mean root diameter. Plants from two hills were used for measurements of ROA and root total and active absorbing surface area. A root sample from one hill was stored at −80 °C for measurements of RAP activity. The rest of the roots and the root to be used for measurement of total root length and mean root diameter were dried in an oven at 70 °C to constant weight and weighed. Root length and mean diameter were measured following Zhang et al.[37], as follows: roots were floated on shallow water in a glass tray (30 cm × 30 cm) and scanned (Epson Expression 1680 Scanner, Seiko Epson Corp., Tokyo, Japan) and their length and diameter determined with WinRHIZO Root Analyzer System (Regent Instruments Inc., Quebec, Canada).ROA was determined by measuring oxidation of alphanaphthylamine following Ramasamy et al. [38]. Root total and active absorbing surface area were determined by the methylthionine chloride dipping method[39].

Acid phosphatase (EC3.1.3.2) (APase) activity was measured following Tabatabai and Bremmer [40]. Frozen root tissue was ground to powder under liquid nitrogen and 0.2 g was homogenized with 8 mL extraction buffer (0.2 mol L−1sodium acetate buffer, pH 5.8). The homogenate was centrifuged at 12,000×g for 20 min at 4°C.The resulting supernatant containing intracellular proteins was used for the enzyme assay[41].A 0.1 mL aliquot of the supernatant was incubated for 30 min with 8 mL p-nitrophenyl phosphate(p-NPP) as the substrate for enzyme activity. The reaction was terminated with 6 mol L−1NaOH, and absorbance was measured spectroscopically at 405 nm. Controls without tissue powder were processed in parallel to correct for background coloration. RAP activity was expressed as μmol p-NPP g−1tissue min−1.

Plants from three hills in each plot were sampled and cut 12 cm above the soil level at 6:00 PM, and an absorbent cotton pad of known weight was placed on top of each decapitated stem and covered with a polyethylene sheet.At 6:00 AM the following morning, the absorbent cotton was weighed.Root bleeding was calculated from the increase of cotton weight.

Before root sampling, aboveground plant tissues were sampled and separated into leaves,stems(culms+sheaths)and panicles(at heading time and maturity).The dry weight of each component was determined after drying at 70°C for 72 h.After leaves were removed from the stem,leaf area was immediately measured with an area meter (LI-3000C, LICOR,Lincoln,NE,USA).Tissue P content was determined by inductively coupled plasma (ICP) emission spectrometry(iCAP6300, Thermo Fisher Scientific, Waltham, MA, USA).PTE, IPE, and PHI were calculated using the following formulas:

where Phand Pmare the P contents of stem,sheath,and leaf at respectively heading time and maturity (g per plant); Y is grain yield (g m−2); and Pgand Ptmare the P contents of respectively grain and plants at maturity(g per plant).

Photosynthetic rate was measured at mid-tillering,panicle initiation, heading time, and maturity. The photosynthetic rate of leaves was measured with a gas exchange analyzer(Li-Cor 6400 portable photosynthesis measurement system, LICOR, Lincoln, NE, USA). The measurement was made from 9:00 AM to 11:00 AM, when photosynthetic active radiation above the canopy was 1300 to 1500 μmol m−2s−1. Ten leaves were used for each treatment.

2.4. Final harvest

Number of panicles per square meter, percentage of filled kernels, and grain weight were determined from 20 plants(excluding border ones) sampled randomly from each plot.Grain yield was determined from all plants from a 1 m2area(except border plants)in each plot and adjusted to a moisture content of 0.14 g H2O g−1fresh weight. Definition and calculation of percentage of filled kernels and number of spikelets per panicle followed Zhang et al.[34].

2.5.Statistical analysis

Analyses of variance were fitted using the SAS/STAT statistical analysis package (version 9.2, SAS Institute, Cary, NC,USA). Plots were generated with SigmaPlot 10.0 (Systat Software Inc., Chicago, IL, USA). Data from each sampling date were analyzed separately,and the resulting means were tested by the least significant difference at P ＜0.05 (LSD0.05).The ANOVA for all the root traits showed significant differences in root traits between or among treatments and varieties and an interaction of treatment × variety, but not significant differences between years or the interaction of year × treatment or year × variety. No significant differences in root-shoot ratio, mean root diameter, or root total absorbing surface area were observed among the interaction of year×treatment ×variety(Table S2).

Table 2-Grain yield and its components in rice varieties under two phosphorus levels in 2017 and 2018.

3. Results

3.1.Grain yield and its components

Compared to NP,LP significantly decreased grain yield means by 2.72% for YY2640,11.07% for YJ2,14.50% for LM,and 26.24% for ZD88(Table 2).Compared to NP,LP decreased the number of panicles and spikelets per panicle and increased the percentage of filled kernels and grain weight of rice varieties.The decreased grain yield under LP was due mainly to a smaller sink size(total number of spikelets per square meter).The reduction in grain yield of STVs (YY2640 and YJ2) was smaller under LP than that of WTVs(LM and ZD88).The trend was the same in both years(Table 2).

3.2. Tiller number, leaf area index (LAI), and leaf photosynthetic rate

The number of tillers increased from mid-tillering and peaked at the panicle initiation stage, then declined and became stable (Table 3). The number of tillers varied with variety and growth stage. Compared to NP, LP significantly reduced the number of tillers of the four rice varieties at the main growth stages. Compared to NP, LP significantly increased the ratio of productive tillers by 4.52% for WTVs and 6.88% for STVs(Table 3).LAI increased from mid-tillering,peaked at heading time, and then declined (Table 4).Compared to NP, LP significantly decreased LAI of the four rice varieties at main growth stages. The LAI of STVs was higher than that of WTVs under LP (Table 4). Leaf photosynthetic rate was increased from mid-tillering to heading time,peaked at heading time, and declined thereafter (Fig. 1).Compared to NP, LP significantly decreased leaf photosynthetic rate of the four rice varieties at main growth stages.The 14.72% reduction in leaf photosynthetic rate in STVs was smaller than the 38.84% reduction in WTVs under LP. The trend was the same in both years(Fig.1).

Table 3-Number of tillers and the ratio of productive tillers of rice varieties under two phosphorus levels in 2017 and 2018.

3.3. Phosphorus translocation efficiency (PTE), internal phosphorus efficiency (IPE), phosphorus harvest index (PHI), and total phosphorus accumulation

Compared to NP,LP significantly increased PTE,IPE and PHI by 23.57%, 97.29%, and 67.01% for WTVs and by 21.86%,150.55%, and 32.63% for STVs, respectively (Table 5). Under NP, PTE, IPE and PHI for STVs were higher than those of WTVs. Under LP, PTE, and IPE for STVs were higher than those of WTVs, PHI for STVs were lower than that of WTVs(Table 5). Total P accumulation was increased from midtillering to maturity (Fig. 2). Compared to NP, LP significantly decreased total P accumulation of the four rice varieties at main growth stages. The 62.34% reduction in total P accumulation of STVs was smaller than 67.75% reduction of WTVs under LP(Fig.2).

3.4.Shoot dry weight and root morphology traits

Compared to NP, LP significantly decreased shoot biomass means by 17.85% for STVs and 30.52% for WTVs at main growth stages. Compared to NP, root biomass was decreased only under LP at mid-tillering, whereas it was significantly increased by 14.14% for STVs and 17.25% for WTVs under LP from panicle initiation to maturity(Fig.3-AD).The ratio of root to shoot decreased gradually.Compared to NP,LP significantly increased the ratio of root to shoot by 29.23% for STVs and 44.21% for WTVs at main growth stages.The trend was the same in two years(Fig.3-E,F).Compared to NP, total root length was significantly decreased by 19.31% for WTVs, and increased by 19.41% for STVs under LP at main growth stages (Fig. 4-A, B). Compared to NP,mean root diameter was significantly decreased by 16.74% for WTVs and 10.17% for STVs under LP at main growth stages(Fig.4-C,D).

Table 4-Leaf area index(LAI) of rice varieties under two phosphorus levels in 2017 and 2018.

Fig.1-Leaf photosynthetic rate(A,B)of rice varieties under two phosphorus levels in 2017 and 2018.NP,normal phosphorus level;LP,low phosphorus level;MT,mid-tillering;PI,panicle initiation;HT,heading time;MA,maturity.Vertical bars represent±standard error of the mean(n=20)where these exceed the size of the symbol.Different letters above bars indicate statistical significance at P ＜0.05 within the same measurement stage.

Table 5-Phosphorus translocation efficiency (PTE), internal phosphorus efficiency (IPE), and phosphorus harvest index(PHI)of rice varieties under two phosphorus levels in 2017 and 2018.

3.5. Physiological traits of roots

Compared to NP, ROA was significantly decreased by 14.72% for WTVs and increased by 15.17% for STVs under LP at main growth stages (Fig. 5A, B). Compared to NP, root bleeding rate was significantly increased by 20.41% for WTVs and 30.63% for STVs under LP at main growth stages(Fig. 5C, D).

Very similar results were observed for root total absorbing surface area and active absorbing surface area throughout the growth season (Fig. 6). Compared to NP, LP significantly decreased total absorbing surface area and active absorbing surface area for WTVs at main growth stages, whereas the opposite trend was observed for STVs(Fig. 6).

Compared to NP,LP significantly increased RAP activity by 33.45% for STVs and 16.54% for WTVs at main growth stages(Fig.7).Under either LP or NP,RAP activity of STVs was higher than that of WTVs(Fig.7).

3.6.Correlations of roots traits with yield and PUE

Root morphological and physiological traits (root dry weight,total root length, average mean diameter, ROA, root bleeding rate, root total and active absorbing surface area, and RAP activity at the main growth stages were positively and significantly or very significantly correlated with grain yield(Table 6). Root-shoot ratio, total root length, ROA, root bleeding rate, root active absorbing surface area and RAP activity at the main growth stages were positively and significantly or very significantly correlated with PTE and IPE. PHI was positively and very significantly correlated with root-shoot ratio and negatively and very significantly correlated with mean root diameter (Table 6).

Fig.4-Total root length(A,B)and mean root diameter(C,D)of rice varieties under two phosphorus levels in 2017 and 2018.NP,normal phosphorus level;LP,low phosphorus level;MT, mid-tillering;PI,panicle initiation; HT,heading time; MA,maturity.Vertical bars represent±standard error of the mean(n=4)where these exceed the size of the symbol.Different letters above bars indicate statistical significance at P ＜0.05 within the same measurement stage.

4. Discussion

Compared to NP, LP significantly decreased grain yield for STVs (YY2640 and YJ2) and WTVs (LM and ZD88). The WTVs yield were more affected under LP than the STVs The decreased yield was due mainly to the reduction in sink size(the number of panicles × spikelets per panicle) although the percentage of filled kernels and grain weight were increased.However,the increase in percentage of filled kernels and grain weight could not compensate the decrease in sink size. A smaller reduction in the yield for STVs may be attributed mainly to a smaller decrease in sink size under LP(Table 2).

Because previous studies on rice growth in LP were performed mainly with rice seedlings or certain growth stage, little is known about changes in agronomic and physiological traits in rice during the full growing season[42-44]. We observed marked differences in agronomic and physiological traits of rice shoots between the two types of rice varieties throughout the growing season.Compared with WTVs, STVs showed the following characteristics under low phosphorus:1)a higher proportion of productive tillers(Table 3). This characteristic will produce more effective panicles and a larger panicle and result in larger sink size and higher yield [45]; 2) higher LAI (Table 4). Keeping a higher leaf area under LP will promote crop growth, increase photosynthesis and dry matter production.A larger sink size of STVs under LP also maintains the balance between source and sink, which influences crop yields [46]; 3) higher leaf photosynthesis capacity (Fig. 1). Leaf photosynthesis in rice during grain filling contributes 60%-100% of carbon content in the grain[47].The increase in source activity(photosynthesis)under LP of STVs compared to WTVs could contribute to higher grain yield; and 4) stronger dry-matter production capacity from heading to maturity (Fig. 3). Dry matter production of rice from heading to maturity accounts for about 90% of grain weight. Increasing dry matter accumulation during this period increases yield [48]. Thus, the greater dry-matter production capacity from heading to maturity partially accounts for the higher grain yield of STVs than WTVs under LP. We speculate that varieties tolerant to low-phosphorus with a large sink size and strong tillering ability and photosynthetic productivity during the full growing season will show lower yield loss in low-P environments.

In low-P environments,plants have had to evolve complex strategies to adapt to low-P conditions, including modifications of root architecture, morphology and physiology[9,30,49]. However, previous studies [50,51] of the response of rice root traits to low phosphorus focused mainly on the seedling stage or certain growth stage. Over the full rice growing season, compared to NP, LP significantly decreased shoot biomass and root-shoot ratio for STVs and WTVs at main growth stages.Compared to NP,the root biomass under LP decreased only at mid-tillering, whereas it increased significantly for STVs and WTVs from panicle initiation to maturity(Fig.3A-D).P deficiency typically leads to an increase in root-shoot ratio, and while this increase may involve an absolute increase in root biomass under mild P deficiency,more severe P typically leads to an absolute decrease in root biomass[13,52].Wissuwa et al.[53]observed increasing starch concentrations with increasing P deficiency in roots but much less in shoots and increasing root-shoot ratio.Dissanayaka et al. [54] also proposed that a larger root system potentially promotes P acquisition under low P availability during vegetative and reproductive growth and acts as a pool of P for remobilization to the shoot.Those results suggest that the assimilates were preferentially distributed to roots to maintain normal growth. We also observed that LP significantly increased the root bleeding rate and RAP activity of the two types of varieties at the main growth stages compared with NP and that ROA and root total and active absorbing surface area for STVs were significantly increased under LP, whereas the opposite results were observed for WTVs (Figs. 3-7). The response of root physiology in STVs to low phosphorus was an increase in P absorption capacity resulting from increased root activity (ROA, root bleeding rate, root total and active absorbing surface area, and RAP activity). These results suggest that STVs could regulate root growth with stronger vigor to adapt to a low-phosphorus condition.

Fig.5-Root oxidation activity(A,B)and root bleeding rate(C,D)of rice varieties under two phosphorus levels in 2017 and 2018.NP,normal phosphorus level;LP, low phosphorus level;MT,mid-tillering; PI,panicle initiation;HT,heading time;MA,maturity. Vertical bars represent±standard error of the mean(n =4) where these exceed the size of the symbol.Different letters above bars indicate statistical significance at P ＜0.05 within the same measurement stage.

Phosphorus use efficiency (PUE) is commonly defined as the biomass or grain yield produced per unit of P absorbed by a crop.P-efficient crops can maintain high biomass and grain yield under low-P conditions. High P acquisition efficiency and utilization efficiency are two key traits conferring high P efficiency in crops[26,55,56].Compared to NP,LP significantly decreased total P accumulation of the four rice varieties at main growth stages.The reduction in total P accumulation of STVs was smaller than that of WTVs under LP (Fig. 2). This finding indicated that STVs could absorb more P than WTVs to maintain plant growth in the low-P environment. LP significantly increased PTE, IPE, and PHI of rice varieties compared to NP (Table 5), indicating that a low-phosphorus condition could promote P transfer to grain. IPE here was defined as the grain yield produced per unit of P absorbed by rice. The observation that IPE was higher for STVs than for WTVs under LP suggests that A STV also has a high PUE.

Fig.7-Root acid phosphatase(RAP)activity(A,B)of rice varieties under two phosphorus levels in 2017 and 2018.NP,normal phosphorus level;LP,low phosphorus level;MT,mid-tillering;PI,panicle initiation;HT,heading time;MA,maturity.Vertical bars represent± standard error of the mean(n =4) where these exceed the size of the symbol.Different letters above bars indicate statistical significance at P ＜0.05 within the same measurement stage.

Table 6-Correlations of roots traits with yield and PUE of rice.

Roots are involved in acquisition of nutrients and water,synthesis of plant hormones,organic acids and amino acids,and anchorage of plants. Root morphology and physiology play an important role in the growth and development of aboveground organs and yield formation [57,58]. Higher root biomass and root oxidation activity in roots are required for achieving more panicle number, more spikelets per panicle,greater grain-filling percentage, and higher grain yield [25].Roots and shoots are interdependent. When root growth is limited by abiotic stress, sufficient carbohydrates are allocated to the root system to develop and maintain root functions[59-61].Root biomass and ROA are regarded as the two most important traits in root morphology and physiology because root biomass is closely associated with root nutrientand water-absorbing ability and aboveground biomass,and a higher ROA is necessary to maintain root biomass, root and shoot growth,and ion uptake[38,59,62,63].In this study,root biomass was positively highly correlated with grain yield,and total root length, ROA, root bleeding rate, root active absorbing surface area, and RAP activity were positively correlated with grain yield, PTE, and IPE (Table 6). We conclude that greater root biomass and higher root activity increased P acquisition efficiency and contributed to higher grain yield and PUE of STVs in comparison with WTVs. Root improvement offers high potential for breeding rice varieties with tolerance to low phosphorus and high PUE. The rootshoot interactions in STVs under low-phosphorus conditions await further study.

5. Conclusions

Although LP limited the aboveground growth of rice(expressed as number of tillers, shoot biomass, LAI, and leaf photosynthetic rate),STVs adapted to a low-phosphorus environment by increasing root growth. Compared to NP, LP significantly increased PUE (PTE, IPE, and PHI) in the two types of rice varieties, especially STVs, and yield reduction was lower in STVs than in WTVs.The increased PUE and smaller reduction of grain yield in STVs were attributed mainly to improvement in root morpho-physiological traits including total root length,ROA,root bleeding rate,root active absorbing surface area,and RAP activity.The mechanism underlying root-shoot interaction and correlations between PUE and roots for high grain yield and high PUE in STVs merits further investigation.

Acknowledgments

This work was supported by the National Key Research and Development Program (2016YFD0300206-4; 2018YFD0300800),the National Natural Science Foundation of China(31461143015, 31771710, 31871559), Young Elite Scientists Sponsorship Program by CAST (2016QNRC001), Six Talent Peaks Project in Jiangsu Province (SWYY-151), the Jiangsu Provincial Key Research and Development Program (Modern Agriculture) (BE2015320), the Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions, and the Top Talent Supporting Program of Yangzhou University (2015-01)

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.01.002.