Trait plasticity (variation of a trait under environmental variability or gradients) and trait integration are both crucial for plant adaption to environmental change. Variations in different suite of root traits such as biomass allocation, morphology and physiology underlie diverse phosphorus (P) acquisition strategies among plants. Yet, how the intraspecific plasticity and integration of root traits influence plant adaptation to different P supply remains obscure. To characterize diverse adaptive strategies in relation to plant P acquisition, eight root traits were assessed in 20 wheat (Triticum aestivum L.) genotypes grown in a culture room with low and high P supply. High P supply increased shoot P accumulation and biomass of all wheat genotypes. The shoot P accumulation in genotypes with high P sensitivity (PS: calculated as shoot P content at low P / shoot P content at high P supply) was higher with high P supply and lower with low P supply compared with that in the genotypes with low PS. The high-PS genotypes exhibited larger variation in root length, root/shoot ratio and rhizosphere pH across P supplies than the low-PS genotypes, suggesting an integrated response at the whole-plant level. At low P supply, the high-PS genotypes had greater root length and specific root length, but lower acid phosphatase activity than the low-PS genotypes, which suggests contrasting P-acquisition strategies across the genotypes. Strong co-variation of root traits occurred across low-PS genotypes regardless of P supply; conversely, the high-PS genotypes only exhibited strong trait integration at low P supply, whereas high P supply sharply reduced root trait co-variation. Our findings suggest that P stress may strengthen root trait integration in wheat plants, and that both plasticity and integration of root traits drive plant adaptive strategies and tolerance to P-deficiency stress.

Experimental set-up

The experiment was set up as a randomized complete block design, with two factors: (1) genotypes - 30 wheat genotypes, and (2) P supplies - 0 and 200 mg P per kg soil (P was supplied as KH₂PO₄). The experiment was carried out with four replicates. Pots within each block were re-randomized weekly to further decrease effects of plant location within blocks.

Soil collection and wheat genotypes

A calcareous silt loamy, low-P soil was collected from Dongdianzhuang, Baoding (114°30′E, 39°05′N), China. Soil properties were as follows: pH 7.40 (1:5 soil:water ratio), soil bulk density 1.45 g cm^-3, total N 1.26 g kg^-1, Olsen-P 5.50 mg kg^-1, and total K 20.45 g kg^-1. The soil was sieved (2 mm) and mixed thoroughly. The sieved soil was then used to fill plastic pots of 11 cm in height and 18 cm in diameter (1 kg dry soil per pot).

Thirty wheat genotypes released from 1977 to 2016 were selected initially for this study (Table S1). They are the main winter wheat genotypes planted in Hebei Province during the relevant periods. Their P sensitivity ranged from 0.070 to 0.175, with a mean of 0.121. The classes of high/low P sensitivity (P sensitivity was calculated as shoot P content at low P / shoot P content at high P supply) genotype groups were constructed by finding a median value of P sensitivity and creating the medium-PS interval as median ± S.E of the genotype effect. Genotypes with data falling above or below that medium interval were classed as high-PS or low-PS genotypes, respectively (Malik et al. 2016; Rengel and Graham 1995). To better compare how different genotypes responded to low and high P supply, we selected 10 high-PS and 10 low-PS genotypes for further analyses without considering the medium group (Table S1).

Wheat seeds were surface sterilized (30 min in 30% v/v H₂O₂ solution), rinsed, and subsequently germinated on the wet filter paper at 25 °C for 24 h in the dark. Six uniformly germinated seeds were sown into each pot. After one week, the seedlings were thinned to four plants per pot. During the whole experimental period, soil moisture was kept at 18–20% (w/w, i.e., 70% of water holding capacity) as determined gravimetrically by weighing the pots every 2 days and adding de-ionized water when necessary. In a culture room, temperatures ranged from 10 (night) to 15 °C (day), and the average photosynthetically active radiation was 33.6 W m^-2. This was accomplished by applying 14 hours of LED light from 6:00 am to 8:00 pm. Relative air humidity was kept at 40–50%. The nutrient solutions were added to the soil as basal fertilizers at the following rates (mg kg^-1): 1687 (NH₄)₂SO₄, 335 K₂SO₄, 126 CaCl₂, 43 MgSO₄·7H₂O, 2.0 CuSO₄·5H₂O, 5.8 EDTA-FeNa, and 10 ZnSO₄·7H₂O. To achieve the same soil K condition between the two P treatments, 252 mg K as K₂SO₄ was supplied to soil in the low P treatment. The pot experiment was conducted from December 2018 to January 2019 at the eastern campus of Hebei Agricultural University (115°49′E, 38°85′N), Baoding, China.

Harvest and measurements

Plants were harvested at 37 DAS and divided into shoots and roots. Shoots were cut at the soil surface and oven-dried at 72 °C for 48 h, weighed, and ground to fine powder. Shoot P concentration was determined by the standard vanado-molybdate method (Murphy and Riley 1962) after digestion in a H₂SO₄-H₂O₂ mixture at 360 °C for 2 h.

We measured eight root traits in three categories (Bardgett et al. 2014; Wang et al. 2020a): 1) three whole-root system traits: root biomass, root/shoot ratio and root length; 2) three root morphological traits: root diameter, RTD and SRL; and 3) two root physiological traits: pH of the rhizosphere soil and phosphatase activity in the rhizosphere. A full description of every root trait is provided below.

Roots with soil adhered were shaken gently to remove bulk soil, and then the rhizosphere soil was collected by brushing it off roots. Rhizosphere soil was stored at 4 °C, and subsequently used to measure pH within 3 days. After sampling rhizosphere soil, all visible roots were sieved out. Root samples were cleaned using de-ionized water and frozen at -20 °C prior to measurement of root morphological parameters. Cleaned root samples were dispersed in water in a transparent tray (30×20×3 cm) and scanned with an EPSON scanner at a resolution of 400 dpi (Epson Expression 1600 pro, Model EU-35, Japan). The root traits such as root length and root diameter were determined by analysis of images using WinRHIZO Pro software (2009b; Regent Instruments Inc, Quebec, Canada) software. Scanned root images were shown in Figure S1. SRL (m g^-1) was assessed as the ratio of root length over dry root weight. Specific root volume (cm³ g^-1) was assessed as the ratio of root volume over dry root weight. After root samples were scanned, the roots were also oven-dried at 70 °C for 3 days and weighed as root biomass, and then root/shoot biomass ratio was calculated. In addition, we calculated RTD as root dry weight over root volume, assuming roots were perfect cylinders (Ostonen et al. 2007).

Phosphatase activity in the rhizosphere was measured according to (Alvey et al. 2001) using p-nitrophenylphosphate (p-NPP). The whole root systems with tightly adhering rhizosphere soil were transferred into 200-mL vials containing a measured amount of 0.2 mM CaCl₂ solution depending on root volume (Veneklaas et al. 2003). The pH value of Na-acetate buffer (200 mM) was adjusted to the average pH values (7.4) of the rhizosphere soil. The rhizosphere soil in the CaCl₂ suspension was separated by centrifugation for 10 min at 12,000 × g, dried at 60 °C and then weighed. The concentration of p-NPP in the supernatant was measured spectrophotometrically at 405 nm.