Varietal Differences in Wet Damage of Broccoli (Brassica oleracea L. var. italica) Under Waterlogging Conditions

The pot experiment (Fig. 1) was performed in 2020 and 2021 in a rain shelter at the Experimental Farm of the Field Science Center for the Northern Biosphere at Hokkaido University in Sapporo, Hokkaido, Japan (43°07′15″, 141°33′87″). Five cultivars of broccoli (Brassica oleracea var. italica) were tested: ‘Shigemori’ (Sm), ‘Pixel’, ‘Sawayutaka’ (Sy), ‘SK9-099’, and ‘First Star’ (FS) in 2020 with a preliminary pot experiment (Hara et al. 2021). Three characteristic cultivars (tolerant Sm, sensitive Sy, and FS) were selected for the subsequent experiments.

Figure 1.

From early to late May of both years, seedlings were germinated in 128-cell plug trays filled with potting soil (nursery soil, Takii, Kyoto, Japan) that included basal fertilizer at rates of 320 mg·L⁻¹ N, 210 mg·L⁻¹ P₂O₅, and 300 mg·L⁻¹ K₂O. In late May, the seedlings were transplanted to 240-mm diameter poly-pots filled with sieved soil from the Experimental Farm consisting of andisol with a clay loam texture and mixed with 20 g per pot of chemical fertilizer (N : P : K = 8 : 8 : 8) as basal fertilizer. In addition, fertilizer was applied twice to the soil surface during the experiment: 1.5 g per pot of the same fertilizer in late June and 2.0 g per pot of chemical fertilizer (N : P : K : B = 12 : 16 : 12 : 0.2) in early July. Pesticides were also applied according to the conventional methods of the Experimental Farm.

Three types of waterlogging treatment (CONT, WET2, and WET3) were created: CONT – each pot was constantly irrigated at an amount of 4–7 mm per day using an irrigation system; WET2 – maintaining for two days under excess soil moisture, otherwise same as CONT (2021 only); WET3 – maintain three days of excess soil moisture, otherwise same as CONT.

Based on previous studies (Jitsuyama et al. 2019; Ide et al. 2022), the wet treatment consisted of maintaining 2–3 cm of water on the soil surface. Since the pots had holes in the bottom for drainage, thick vinyl bags were used to prevent drainage during the wet treatments. More specifically, the pots under cultivation were inserted into a thick plastic bag and then placed into an additional pot to hold the pots in place (Fig. 1A). The waterlogging treatments were conducted before flower bud emergence. In 2020, the treatments were conducted between June 28 and July 4 [57–63 days after sowing (das)], and in 2021 they were conducted between June 28 and July 1 (57–60 das). Each plant's fresh weights (shoots, leaves, and stems) were recorded in late July (approximately 20 days after waterlogging treatments) during the broccoli head growth stage.

The following environmental parameters were also measured: soil volumetric water content, soil dissolved oxygen concentration, air and soil temperature, relative humidity, and cumulative day radiation. A soil moisture measurement kit (SM150T, Delta-T Devices, Cambridge, UK) and soil O₂ sensors (MIJ-03, Environmental Measurement Japan, Fukuoka, Japan) were used. The thermal recorder included a relative humidity sensor and a UV illuminance recorder (TR-74Ui, T and D, Nagano, Japan). Soil moisture was measured continuously using the kit, and the other soil measurements were conducted by inserting an O₂ sensor and a thermal sensor into the potting soil to a depth of 15 cm. These measurements were taken once per day for each treatment, between 10:00 a.m. and midday. The air temperature recorder was set 30 cm above the ground, above the broccoli plants, and data were logged automatically every hour.

The average temperature during the pot experiment over approximately three months was lower in 2020 than in 2021 [20.8 °C (2020) < 22.4 °C (2021)]. However, the temperature trend was reversed for one week just after the waterlogging treatment [22.2 °C (2020) > 20.1 °C (2021)]. The average total solar radiation during the experiment was also lower in 2020 than in 2021 [18.7 MJ·m⁻² (2020) < 22.4 MJ·m⁻² (2021)].

To scrutinize the impact of waterlogging, water content and photosynthetic parameters were also measured in 2021 with the three types of waterlogging treatments listed above (CONT, WET2, and WET3).

The water content of shoots (sum of leaves and stems), leaves, and stems were also measured just after the wet treatments (65–66 das). For water content calculation, the fresh and dry weights after drying at 80 °C for 72 h with a forced-air dryer were measured. The measurements were conducted for each of the five replications.

For the calculation of photosynthetic parameters, the leaf area and the dry weight of each plant were evaluated after the wet treatments (65–66 das) and the later growth stage (79–80 das). The crop growth rate (CGR) was calculated using all dried biomass, including leaves, stems, and roots, which were washed in water. The leaf area measured by a WinRHIZO system (2004a, b, Regent Instruments, Canada) was converted to the mean leaf area index (LAI). The net assimilation rate (NAR) was calculated as the ratio of CGR to mean LAI. The plant area (450 cm² per plant) was defined from the upper side area of the polypot.

The experiment was a split-split plot design or split plot design with five replications (one plant per replication), with the main plots as experimental years (Y) or water stress treatments (T) and the subplots as water stress treatments (T) or cultivars (C), and the sub-sub plots as cultivars (C). The significance of the analysis of variance (ANOVA) was calculated as described by Little and Hills (1978). The biomass was investigated in the experiment with two wet treatments (CONT and WET3) and three cultivars (Sm, Sy, and FS) in 2020 and 2021. Measurement of water content and photosynthetic analysis were conducted with three wet treatments (CONT, WET2, and WET3) and the same three cultivars in 2021. Dunnett's and Tukey–Kramer tests were done using Statcel 4 (an add-in form in Microsoft Excel 2019 for Windows).

In 2021, the field experiment (Fig. 1) was performed in the same region but at a different site (43°07′14″, 141°34′15″) than the pot experiment (Fig. 2A), using the same three cultivars as in the 2021 pot experiment.

Figure 2.

The seeds were sown and grown as described for the pot experiment. The field was prepared with basal dressing chemical fertilizer (N : P : K = 4 : 4 : 4 kg·10a⁻¹) and plowed in with a rotary. Elongated mounds were created mechanically, 45 cm wide and 20 cm high, with 150 cm between rows. After that, 7-gallon nonwoven pots (35 cm in diameter, 30 cm deep; Root pouch, OR, USA) were buried at 40-cm intervals (Fig. 1B), and the seedlings were planted in the pots in mid-June. In mid-July, chemical fertilizer (N : P : K : B = 10 : 13.3 : 10 : 0.17 kg·10a⁻¹) was applied to the surface layer. Other cultivation management work (insect and disease control) was done using conventional methods.

Two types of waterlogging treatments were created: CONT – field with natural precipitation; WET – maintained excess moisture conditions for three days; otherwise, the same as CONT.

During the wet treatment, waterlogging was maintained to keep the water level at approximately the soil surface, and the treatment was conducted before flower bud emergence, July 12–15 (71–74 das). Because nonwoven fabrics are permeable to water, all plants were dug once with the nonwoven pots buried in the field before treatment, and only for the wet treatment were the plants covered with thick vinyl on the outside of the nonwoven pots to prevent drainage (Fig. 1B, C). The water level was checked every few hours during the day and refilled if the water level went down.

Because the maturity timing of flower buds differed among cultivars and treatments, the heads of broccoli were harvested when they reached 10 cm in diameter. The head weight was based on Hokkaido shipping standards (HRO 2019), with leaves removed from harvested heads and height trimmed to 15 cm. Heads whose diameter did not reach 10 cm due to the cultivar or waterlogging treatment were harvested, taking the moment when the flower buds began to enlarge to 2–3 mm as the limit of flower bud enlargement.

Environmental parameters (soil volumetric water content, soil dissolved oxygen concentration, air temperature, relative humidity, and cumulative day radiation) were measured as described for the pot experiment.

The average temperature in the field during the experiment of 2021 was 22.0 °C, and the average total solar radiation was 17.3 MJ·m⁻². The average temperature during the waterlogging treatment (July 12–15) was 24.9 °C, which tended to be higher than during the total duration of the experiment. However, there was no significant difference in total solar radiation during the same period compared to the entire duration of the experiment.

The experimental design was a split-plot design with five replications. The main plots and subplots were defined as the three cultivars and the two waterlogging treatments. The significance in the ANOVA was calculated as described for the pot experiment. A Student's t-test and Pearson correlation coefficient test were also done using Statcel 4.

The waterlogging treatments drastically altered the soil environments, including water content and oxygen concentration. The average soil water contents during the waterlogging treatments were significantly higher (Tukey–Kramer test, P < 0.05) than in the controls [2020: 18.9% (CONT) < 42.8% (WET3); 2021: 15.3% (CONT) < 42.9% (WET2), 42.9% (WET3)]. Similarly, the soil water content affected the average oxygen concentrations, with significantly lower oxygen concentrations in the waterlogging treatments (Tukey–Kramer test, P < 0.05) than in the controls [2020: 19.75% (CONT) > WET3: 8.69% (WET3); 2021: 18.73% (CONT) > 8.73% (WET2), 6.95% (WET3)].

In the pot experiment with three cultivars in 2020 and 2021, three days of waterlogging treatment resulted in wilting, yellowing of leaves, defoliation, and reduced growth of plants (Fig. 3, Table 1). The main effect of waterlogging treatment was significant for all parts of the plant (P < 0.001). The waterlogging treatment decreased all plant parts’ fresh weight by 25–30% of CONT at WET3. The main effects of cultivars were also significant for leaves and stems (P < 0.001). Leaf weight tended to be greater for Sm and lower for Sy, while stem weight was greater for Sy and lower for Sm and FS. There were also significant interactions between waterlogging treatment and cultivar (T × C, P < 0.001) and between year and waterlogging treatment (Y × T, P < 0.01) for all fresh weights.

Figure 3.

The effectiveness of the treatment to the shoot fresh weight for each cultivar in both years is shown in Figure 4. In WET3, the effects of waterlogging treatment were significant in all cultivars; however, the percentage of damage looked slightly different among cultivars. The Sm cultivar maintained high shoot biomass compared with the other two cultivars in the mild condition of waterlogging, WET2. The order of the wet tolerance among the three cultivars, even in different experimental years, can be arranged as Sm > Sy ≧ FS.

Figure 4.

The impact on broccoli's biomass reduction by the waterlogging treatments looked different between the two years. For example, the percentages of the decrease in shoot biomass of all cultivars in WET3 from CONT tended to be more severe in 2020 than in 2021 [Sm: 40% (2020) < 51% (2021); Sy: 16% (2020) < 37% (2021); FS: 9% (2020) < 23% (2021)].

The main effect of waterlogging treatment for water content was significant for all parts of the plant (P < 0.001) (Table 2). Three days of waterlogging treatment significantly decreased the water content in the plants. The main effects of cultivars were non-significant; however, the interactions between waterlogging treatments and cultivars were significant for all parts as well (P < 0.05). The effect of waterlogging treatment on the water content tended to be lower for Sm and greater for Sy and FS, particularly in the stem (Fig. 5A, B).

Figure 5.

The ANOVA analysis of the photosynthetic parameters CGR, mean LAI, and NAR after waterlogging treatment is shown in Table 3. The main effect of waterlogging treatment was significant for all parameters (P < 0.001). The main effect of the cultivar was significant for CGR (P < 0.05) and mean LAI (P < 0.001). Mean LAI tended to be greater for Sm and lower for FS. The interactions between waterlogging treatments and cultivars were significant for mean LAI (P < 0.05) and NAR (P < 0.05). The effect of waterlogging treatment on the mean LAI tended to be lower for Sm and Sy than FS (Fig. 6A). Additionally, the effect of waterlogging treatment on NAR was severe, particularly for FS in WET3 (Fig. 6B).

Figure 6.

As in the pot experiment, the soil moisture content and soil oxygen concentration were altered by the waterlogging treatment. The waterlogging treatment significantly (Student's t-test, P < 0.001) increased the average soil moisture content [14.0% (CONT) < 41.7% (WET)]. The average soil oxygen concentration was also significantly lowered (Student's t-test, P < 0.001) by waterlogging [18.8% (CONT) > 13.9% (WET)]. The oxygen concentration trend during the treatment decreased to 7% in the wet treatment on the third day.

As in the pot experiment, three days of waterlogging treatment caused severe wet injury, such as wilting and yellowing of the leaves (Fig. 2B–D). The sensitivity to waterlogging conditions was also demonstrated, even in the mature plantlet with a head in the field experiments (Fig. 2E–G). In the control plants, the typical head develops and can be harvested (Fig. 2E). However, the head of plants under the wet treatments retarded development and consequently wilted (Fig. 2G). The results of the relationship between head yield and shoot biomass at harvest for each cultivar in the field experiment are shown in Figure 7. The correlation coefficient of determination (R²) between shoot biomass and head yield was significant at 0.86, 0.97, and 0.89 for Sm, Sy, and FS, respectively (P < 0.001), indicating a strong positive correlation between the weight of head and shoot biomass for all cultivars.

Figure 7.

The effects of waterlogging on head diameter and head yield are shown for each cultivar (Fig. 8A, B). Waterlogging reduced the final head diameter to 71% of CONT for Sy (P < 0.001) and to 58% of CONT for FS (P < 0.01), and reduced the final head yield to 50% of CONT for Sy (P < 0.001) and to 51% of CONT for FS (P < 0.01). On the other hand, the reduction caused by waterlogging in Sm was not statistically significant in both cases for head diameter and yield.

Figure 8.