The experimental materials consisted of one-year-old C. illinoinensis seedlings, including two major cultivars from the Yangtze River Delta region: ‘Mahan’ and ‘Pawnee’. All seedlings were sourced from the American Pecan Experimental Base in Zhangmiao Village, Houbai Town, Jurong City, Jiangsu Province. The potted seedling containers were black gallon pots (16 cm upper diameter × 12 cm lower diameter × 30 cm high), each of which was filled with soil up to 2 cm from the upper rim of the pot. The culture medium consisted of garden soil, earthworm soil, and organic fertilizer, which were uniformly mixed at a 3:6:1 ratio. The soil pH was 5.5, the field water capacity was 31.05%, the proportion of organic matter was 2.83%, and the contents of available N, P, and K were 121.25, 51.18, and 93.29 mg kg⁻¹, respectively.

In June 2021, uniformly growing (~40 cm high) seedlings were selected and acclimatized for 15 days in the greenhouse with shading function of Nanjing Forestry University’s Xiashu Forestry (N32.07, E119.12). The temperature of the greenhouse ranged from 23.5 to 34.0°C, and the relative humidity ranged from 63.42 to 84.97% throughout the experiment. The experimental design was in randomized blocks in a 3 × 2 factorial scheme, consisting of three treatments (field capacity 75%, high-oxygen flooding, and low-oxygen flooding) and two cultivars (‘Mahan’ and ‘Pawnee’). Each of the six treatment combinations was replicated three times (n = 3), with each replicate consisting of an experimental unit containing three seedlings. Therefore, each treatment combination included 9 seedlings (3 replicates × 3 seedlings), resulting in a total of 54 seedlings.

In the period before the beginning of the experiment, all seedlings were watered until percolation commenced, then allowed to drain. Once drainage ceased, indicating that the soil had reached field capacity, each seedling in the control group (CK) was immediately weighed to establish a baseline, and the flooding treatments were initiated. The CK group was maintained at 75% field capacity through daily weighing and supplemental watering. Both the high-oxygen (HO) and low-oxygen (LO) flooding treatments were conducted in low-density polyethylene resin tanks (56 cm × 56 cm × 80 cm), in which the water level was maintained 30 cm above the root collar. No water changes were performed throughout the experiment. In the HO group, continuous aeration was provided using an air pump coupled with nanoporous aeration stones to maintain a high dissolved oxygen environment. In contrast, the LO group relied solely on static flooding via water-air exchange to simulate natural hypoxic flooding conditions (Figure 1).

The primary difference between the HO treatment and the LO treatment lies in DO concentrations. The HO treatment maintained an average dissolved oxygen concentration of 7.11 ± 0.10 mg L^-¹, while the LO treatment averaged 1.88 ± 0.09 mg L^-¹. All other water quality parameters (temperature, pH, turbidity) remained consistent between treatments, continuously monitored using a Hydrolab multiparameter water quality analyzer (Table 1).

Morphological indicators, including leaf numbers, injury numbers, injury grades, injury rate and injury index, were recorded on days 15, 30, 45, and 60, respectively.

The leaf numbers and injury numbers of the seedlings were recorded by manual counting. The classification of injury grades was based on the method described by Zhou et al. (2019) with some adjustments. Grade 1: leaves appeared curling but remained green; Grade 2: leaves showed yellowing; Grade3: leaves showed dark brown patches; Grade 4: leaves were scorched with partial shedding.

Injury rate and injury index were obtained from Equation 1, 2, respectively:

Calculation example: On the 60th day of flooding, for the ‘Mahan’ cultivar under the HO treatment, a sample of 9 seedlings was assessed: 4 seedlings were classified as Grade 1, 2 as Grade 2, 1 as Grade 3, and 1 as Grade 4. Therefore, the maximum injury grade was 4. The injury index was then calculated as follows: Injury Index= [(1×4)+(2×2)+(3×1)+(4×1)]/(9×4) = 0.417.

The seedling height was determined with a measuring tape, while the ground diameter was measured using vernier calipers. Akin to the morphological indicators, the heights and ground diameters of the seedlings were recorded on days 0, 15, 30, 45, and 60, respectively. Thus, the growth rates of height and ground diameter were calculated according to Equation 3:

where, R is the rate of height or ground diameter; Y is the amount of change in the mean value of height or ground diameter; and X is the corresponding time variation.

All seedlings were extracted to determine biomass at the conclusion of the experiment (day 60). For the dry mass quantification, Caray illinoinensis seedlings were dug out whole and divided into three portions (roots, stems, and leaves) and then washed and dried at 80 °C to a constant mass, whereafter the dry masses of the three portions were determined separately. The root-to-shoot ratio (RSR) was obtained using Equation 4:

Mature leaves from the middle sections of the seedlings were harvested on days 15, 30, 45, and 60. Some leaves were taken for the determination of plasma membrane permeability, and the rest were stored in ultra-low temperature refrigerator (-80 °C) for the determination of other physiological indicators.

The malondialdehyde (MDA) content was extracted using a thiobarbituric acid (TBA) reagent and boiled at 100 °C for 20 min according to the method of Wassie et al. (2019). Once the solution was cooled to room temperature, it was centrifuged at 3000 rpm for 10 min. The absorbance of the supernatant was then measured using a spectrophotometer (Shimadzu UV-2450, Shimadzu Corporation, Kyoto, Japan) at 450, 532, and 600 nm. The MDA concentration was estimated using Equation 5:

where A₄₅₀, A₅₃₂, and A₆₀₀ are the absorbance of the supernatant at 450, 532, and 600 nm, respectively.

The soluble protein (SP) content was determined via a Coomassie brilliant blue G-250 staining method. A 0.1 g volume of leaves was ground using a mortar with a 1 mL phosphate buffer solution (PBS, pH = 7.8), and then transferred to a 10 mL centrifuge tube. The material was then washed twice with 2 mL PBS and transferred to a 10 mL centrifuge tube and centrifuged at 10,000 rpm for 15 min at 4 °C. After centrifugation, the 0.1 mL supernatant was transferred into a test tube, to which 5 mL of Coomassie brilliant blue G-250 was added and thoroughly mixed. Two min later, the absorbance of the resulting solution was measured at 595 nm using a spectrophotometer. The soluble protein was calculated according to Equation 6:

where C is the protein content from a standard curve; V_T is the total volume of extracting solution; W is the sample weight; and V_S is the sampling volume of determination.

The superoxide dismutase (SOD) activity was measured through a nitro blue tetrazolium (NBT) reduction method. To prepare the enzyme solution: 0.5 g of leaves were ground into a homogenate with a 10 mL of phosphate buffer (PBS, pH = 7.8), transferred to a centrifuge tube, and centrifuged at 12,000 rpm for 20 min at 4 °C. The supernatant was a SOD crude enzyme solution. The reaction solution consisted of a 14.5 mM methionine (Met) solution, 3 mM ethylenediaminetetraacetic acid disodium salt (EDTA-Na2) solution, 60 μM riboflavin solution, and 2.25 mM NBT solution. Subsequently, 2.9 mL of the reaction mixture and 0.1 mL of the enzyme solution were extracted via testing tubes. Two control tubes were prepared simultaneously, one with 2.9 mL of a reaction mixture and 0.1 mL of PBS (without the enzyme solution) as the maximum light reduction tube, and the other with 2.9 mL of the reaction mixture and 0.1 mL of PBS while wrapped in tin foil to block the light. The tubes were then placed in a light incubator at 4000 lux for 20 min at 25 °C and then in darkness to stop the reaction. Following the reaction, the spectrophotometer was zeroed with the control tube without illumination, and the absorbance of each tube was measured separately at 560 nm. One unit of SOD activity was defined as the amount of enzyme that inhibited the NBT reduction by 50%. The SOD activity was calculated according to Equation 7:

where A₀ is the absorbance of control; A_S is the absorbance of the sample; V_T is the total volume of the extracting solution; W is the sample weight; and V₁ is the sampling volume of determination.

The catalase (CAT) activity was determined using a UV absorption method. For enzyme extraction, 0.5 g of leaves was weighed in a mortar, and 2–3 mL of a pre-cooled phosphate buffer (PBS, pH = 7.8) was added at 4 °C with a small amount of quartz sand and ground into a homogenate. This material was transferred to a 25 mL volumetric flask, and the mortar was rinsed several times with buffer solution to fix the volume on the scale. Subsequently, the volumetric flask was placed in a refrigerator at 5 °C for 10 min, and the upper clarified liquid was centrifuged at 4000 rpm for 15 min. The supernatant was the CAT crude enzyme. For the determination of enzyme activity, three 10 mL test tubes were used (two for sample measurement and one as a blank) and designated as S0, S1, and S2 respectively, to which 0.2 mL of the crude enzyme solution, 1.5 mL of a phosphate buffer (PBS, pH = 7.8), and 1.0 mL of distilled water were added, respectively. The test tube S0 was heated in a boiling water bath for 1 min to deactivate the enzyme solution and cooled. All of the test tubes were then preheated at 25 °C to which 0.3 mL of 0.1 mol/L H₂O₂ was added one by one. Each tube was quickly timed and quickly poured into a quartz cuvette, and the absorbance was measured at 240 nm with a 1 min reading at each interval for 4 min. Once all three tubes were measured, the enzyme activity was calculated. A unit of CAT activity was defined as the changes in absorbance at 240 nm per min. The CAT activity was calculated using Equations 8 and 9:

where ΔA240 is the absorbance (control minus the assay); V_T is the total volume of extracting solution; V_S is the sampling volume of determination; t is the time from the beginning of the addition of H₂O₂ to the last reading; and W is the sample weight.

The experimental data were subjected to preliminary statistical analysis and multivariate analysis of variance (MANOVA) for between-subject effects using SPSS 20.0 software. Post hoc comparisons were carried out with Tukey’s HSD test at a 95% confidence level, with statistical significance defined as p< 0.05. The data were expressed as mean ± standard deviation (SD).

Comprehensive evaluation of flooding stress tolerance was performed using the membership function method described by Liang et al. (2025). This analysis was based on the key morphological, growth, and physiological indicators measured after 60 days of flooding stress. To eliminate the influence of inherent trait differences between cultivars, the resistance coefficient (X) for each indicator was first calculated separately for each cultivar using Equation 10:

The membership function value (U) was then calculated for each indicator of each cultivar. The average of these values (D) was used for comparison, where a higher D value indicates stronger comprehensive stress resistance of the cultivar.

Where X_ij denotes the measured value of indicator j for cultivar i, while X_jmax and X_jmin represent the maximum and minimum values of indicator j across all cultivars, respectively. Equation 11 was applied when indicator j was positively correlated with stress resistance; otherwise, Equation 12 was used.

The comprehensive evaluation value (D-value) for each cultivar under each treatment was calculated as the mean of all membership function values (U) according to Equation 13:

where n is the number of indicators included in the analysis. A higher D-value indicates stronger overall flooding stress tolerance. The D-values were finally used to rank all treatment–cultivar combinations.

The comprehensive stress tolerance of the two cultivars under different flooding treatments was evaluated using the membership function method. As summarized in Table 5, the average membership function value (D-value) for ‘Mahan’ under HO and LO treatments was 0.56 and 0.47, respectively, while that for ‘Pawnee’ was 0.43 and 0.35. Based on these D-values, the stress resistance of the treatment–cultivar combinations followed the rank order: ‘Mahan’–HO > ‘Mahan’–LO > ‘Pawnee’–HO > ‘Pawnee’–LO. (Table 5).

Morphological structure serves as a direct indicator of plant damage under abiotic stress (Zahedi et al., 2025; Fang et al., 2025). Under flooding conditions, plants often develop symptoms such as leaf yellowing, desiccation, and abscission (Bispo and Vieira, 2022; Wiström et al., 2023). To systematically evaluate the degree of leaf damage across different treatments, this study quantitatively analyzed injury rate and injury index. Results showed that compared to CK, both cultivars demonstrated increased injury indicators (IR and II) under HO and LO treatments (Table 2). This is similar to the results reported by Zhang et al. (2025), which indicate that flooding induces root hypoxia, inhibits photosynthesis, and drives plants toward anaerobic respiration. Anaerobic respiration not only leads to insufficient energy supply but also accumulates toxic substances such as alcohol, lactic acid, and acetaldehyde, which are harmful to plant. Notably, under the HO treatment, both cultivars exhibited lower injury indicators than under LO treatment (Table 2), suggesting that increasing environmental DO effectively mitigates leaf damage caused by flooding (Du et al., 2016). This may be attributed to the oxygenation treatment directly improving oxygen levels in the root zone, delaying metabolic disorders caused by anaerobic respiration, and maintaining normal root physiological functions, thereby enhancing seedling tolerance to flooding (Li et al., 2025). Furthermore, we observed differences in injury indicators between cultivars (Table 2). The ‘Mahan’ cultivar exhibited a slower increase in injury rate under both treatments, reflecting stronger stress tolerance. In contrast, the ‘Pawnee’ cultivar not only showed a more rapid rise in injury rate but also reached 100% injury earlier under LO conditions, indicating higher sensitivity to root-zone hypoxia.

This study demonstrates that both HO and LO treatments significantly inhibited GRH in both cultivars (Table 4). This observation is consistent with the Low Oxygen Quiescence Syndrome (LOQS) strategy proposed by Voesenek and Bailey-Serres (2015), in which plants actively slow their growth under flooding to minimize energy consumption and prolong survival under stress (Fan et al., 2018; Sun H, et al., 2020). Notably, compared to the LO treatment, the HO treatment exhibited a reduced inhibitory effect on plant height growth (Table 4). This indicates that increasing DO levels can partially alleviate the suppression of plant height elongation caused by flooding stress. Similar phenomena have been reported in tomatoes, where supplemental aeration significantly improved plant height growth oxygenation treatments effectively promote plant height growth under hypoxic conditions (Zheng et al., 2017). In contrast, both cultivars showed a stimulatory effect on GRD under flooding, with values following the order CK< HO< LO (Table 4). This morphological response can be attributed to the formation of hypertrophic lenticels at the stem base. Research indicates that hypertrophic lenticels is the signature morphological response of flood-tolerant woody plants to flooding (Phukan et al., 2016; Argus et al., 2015). Plants adapted by increasing the total area of O2-absorbing tissues exposed to the air (Li et al., 2023), thereby enhancing the ventilation of flooded organs to avoid energy crises caused by hypoxia (Fukao et al., 2019).

The allocation of biomass between aboveground and belowground parts represents an important adaptive strategy of plants to abiotic stress (Poorter et al., 2012). In this study, the ‘Pawnee’ cultivar exhibited a significantly lower RSR under LO treatment than under the CK conditions (Figure 4), reflecting a marked reduction in the proportion of photosynthetic products allocated to the root system. This shift in resource distribution likely reduces the metabolic demand and oxygen consumption of roots under hypoxic stress, thereby temporarily alleviating oxygen deficiency in the root zone (Zhao et al., 2024). At the same time, increased resource allocation to aboveground tissues helps sustain leaf expansion and photosynthetic function, maintains carbon assimilation capacity, and prepares the plant for rapid recovery once the stress is relieved (Sanchez et al., 2024). Furthermore, both ‘Mahan’ and ‘Pawnee’ cultivars reached their maximum TB and RSR under HO treatment (Figure 4), which aligns with previous oxygenation studies on young peach trees conducted by Xiao et al. (2014). This consistency confirms that oxygen enrichment effectively promotes overall dry matter accumulation in woody plants (Zheng et al., 2017). Mechanistically, improved oxygen availability enhances organic matter accumulation through two synergistic pathways. First, it boosts root vitality and aerobic respiration efficiency (Li et al., 2024; Ayi et al., 2019), thereby improving the plant’s capacity for water and nutrient uptake. Second, it improves leaf photosynthetic performance, which facilitates the synthesis and translocation of photosynthetic products to various organs (Sun et al., 2008).

Flooding and anoxic conditions disrupt the equilibrium between reactive oxygen species (ROS) production and scavenging in plants, thereby inducing peroxidative damage to cellular membranes. Malondialdehyde (MDA), a product of membrane peroxidation, significantly increases under anoxic conditions, causing structural and functional damage to plant cell membranes (Yu et al., 2015; Bai et al., 2024). Zhao et al. (2023) reported substantial MDA accumulation in Chionanthus virginicus leaves after 35 days of flooding treatments, respectively. This accelerated membrane peroxidation, leading to structural disruption and extensive seedling mortality. In this study, MDA content in both cultivars showed a continuous increase throughout the experimental period under HO and LO treatments, reaching peak levels by day 60 (Figure 5a). This progressive accumulation indicates severe damage of cell membrane structure during the late stages of flooding stress. Notably, the MDA content under HO treatment was significantly lower than under LO treatment on day 15 and day 60 for ‘Mahan’, and on days 15, 30, and 60 for ‘Pawnee’. These results indicate that oxygen enrichment through aeration effectively mitigates membrane lipid peroxidation and reduces damage to the membrane system (Zheng et al., 2019).

The accumulation of soluble protein (SP) content and the elevation activities of superoxide dismutase (SOD) and catalase (CAT) collectively reflect the osmotic regulation and antioxidant defense mechanisms initiated by plants under stress conditions (Li et al., 2021). Under flooding stress, plants synthesize the osmotic regulator SP to maintain a certain turgor pressure, thereby safeguarding fundamental physiological processes such as cell growth, stomatal opening, and photosynthesis, and enhancing overall stress resistance (Wang et al., 2015). Simultaneously, plants effectively scavenge excess ROS and mitigate oxidative damage by elevating antioxidant enzyme activities such as SOD and CAT (Noctor and Foyer, 1998). In this study, SP content peaked at day 60 across most treatments, with the exception of the ‘Pawnee’ cultivar under LO treatment, which reached its maximum at day 45 (Figure 5b). Similarly, the SOD and CAT activities in both cultivars under HO and LO treatments also attained their highest levels at the end of the 60-day experimental period (Figures 5c, d). This indicates that C. illinoinensis seedlings progressively enhance both osmotic regulation and antioxidant capacity during the late flooding stress, forming a synergistic mechanism that collectively mitigates physiological damage induced by flooding (Yan et al., 2019). Notably, after 60 days of flooding, both cultivars exhibited a consistent gradient in SP content and the activities of SOD and CAT, following the order LO > HO > CK (Figures 5b–d). This pattern clearly indicates that the activation level of the plant’s antioxidant defense system is positively correlated with the intensity of the flooding stress (Chen et al., 2025).

Overall, increasing the DO concentration from 1.88 to 7.11 mg L^-¹ significantly enhanced the flood tolerance of C. illinoinensis (Table 5). Specifically, oxygenation reduced leaf damage, decreased key antioxidant enzyme activities, and increased TB accumulation (Figure 6). These findings demonstrate that elevated DO is an effective strategy for mitigating flooding stress in plants. This conclusion aligns with previous studies on other mesophytes, including tomato (Li et al., 2025), Cymbidium (Song et al., 2024), and rice (Liu et al., 2024). Future research should investigate the effects of different dissolved oxygen gradients on plant flooding tolerance and extend this work to additional mesophytic species. This systematic approach will help elucidate the physiological thresholds and adaptive mechanisms of plants in response to DO fluctuations.