The cuttings of plants with attached roots and rhizomes were gently collected individually by hand from bottom sediments of a 100 × 100 m donor bed in Swan Lake on the eastern coast of Shandong Peninsula, China (37.3382°–37.3588°N, 122.5551°–122.5793°E) in June 2019. Intact shoots were selected and then rinsed thoroughly in seawater to remove sediments, shells and other debris. The shoots were immediately placed in plastic boxes and transported to aquaculture farm within 4 h.

For shoot acclimatization, shoots were held in the farm for 7 days in a 12.5 m × 1.5 m × 1.0 m (length × width × height) tank with flowing seawater at temperature (20 ± 1.2°C), natural light condition and seawater salinity (30 PSU) until prepared for the experiment.

In this experiment, four different prolonged light durations 2, 4, 6, and 8 h d⁻¹ were produced by using light-emitting diode (LED) plant lamps. Natural light duration (average 14 h d⁻¹) was used as reference prolonged light duration (0 h d⁻¹).

A total of 360 cuttings were randomly selected from those that had been acclimated to laboratory conditions over 7 days, and then divided into 20 groups of 18 cuttings each. Each treatment was replicated four times (four groups). All cuttings used in this study were carefully selected to obtain the similar morphological characteristics with rhizomes of 5–7 cm, height of 25 ± 2 cm, and numbers of leaves (three to four). All selected cuttings were marked for leaf and rhizome growth estimations using the leaf punching method and rhizome binding method (Short and Duarte, 2001).

A 20.5 cm × 14.3 cm × 9 cm plastic box, filled with sediments, was used as planting container for each group. Filling sediments were collected at the beach just outside the experimental site. The mixture of 25% fine sand (sediment particle size between 63 and 250 μm) and 75% silt and clay (sediment particle size <63 μm) in weight percentages of sediments were obtained using a standard series of sieves according to Zhang et al. (2015). Each box was filled with the sediment at an 8 cm thick layer and then was kept submerged in sea water for at least 48 h until the shoots were planted.

The 18 cuttings in each group were evenly anchored by carefully inserting the rhizome meristem into the sediment at a depth of 4 cm. Then the four boxes for each treatment were placed at random in a 90 × 60 × 75 cm polyethylene tank with an effective water volume of 300 L. The experiment was conducted in a circulating-water system, which consisted of five polyethylene tanks (experimental tanks), a transport tank with a 200 W aquarium pump for the recycling of the water at an exchange rate of 900 L h⁻¹ and a programmed temperature controller with a heater and a refrigerator for the holding of the water temperature (Figure 1). The position of the four boxes in the same tank was changed randomly every day to reduce the error in light intensity. The experiment lasted for 42 days.

LED plant lamps (red: blue = 7 : 3) were used as the supplementary light source to keep the light intensity over 2000 lx. Light control switch and time control switch were used to produced different prolonged light duration. Enclosure barriers were used in the treatments of prolonged light durations to separate the treatments each other and to avoid interaction effects between different treatments (Figure 1).

During the course of the experiment, water temperature and salinity of the water in the tank were maintained at 20.0 ± 1.0°C, 30.0 ± 1.2 PSU, respectively. Temperature and salinity were checked daily using a mercury thermometer and a portable refractometer (Atago). The water in the tank was renewed daily at a rate of 30%.

Statistical analyses of all data were performed using the SPSS Windows Program (Release 25.0, SPSS Inc.). All data were reported as means ± SD, and were analysed by one-way analysis of variance (ANOVA) to test for the effect of different light durations on the responses of Z. marina plants. All data were tested for normality and homogeneity of variance to meet the assumptions of parametric statistical analysis. Post hoc mean comparisons (Student-Newman-Keuls; S-N-K) were performed to identify the significance of differences among treatments when the one-way ANOVAs were significant at α = 0.05.

Principal component analysis (PCA) was conducted using the FactoMine package in R software (version 4.1.1) to reveal the relationship between the variables of Z. marina plants. The result of PCA showed the explanation and contribution of ADP, BDP, IER and RF to each variable.

The quadratic function curves were fitted according to the relationship between the prolonged light durations and SR, ADP, BDP, IER, and RF, and the optimal suitable growth and propagation periods were obtained according to the fitting relationship.

Pearson and Mantel correlation analysis were conducted using the ggcor package in R software (version 4.1.1) to identify the degree of potential correlations between the variables, including survival, growth, propagation and physiological of Z. marina plants.

The survival rate of Z. marina plants exposed to different prolonged light durations increased firstly from 0 h d⁻¹–4 h d⁻¹ up to 97.2% and then decreased to 93.1% at 8 h d⁻¹ (Figure 2A). The result of one-way ANOVA indicated that the prolonged light duration had a significant effect on the survival of Z. marina plants (p < 0.05; Table 1). The survival attained under the 4 h d⁻¹ treatment was 1.1 and 1.0 times significantly higher than those attained under control and 8 h d⁻¹ treatments. The relationship between survival rate (S) and prolonged light duration (PLD) could be expressed as quadratic function through curve estimation. The estimated maximum survival rate (S_max) was achieved at 96.9%, i.e., PLD_opt= 4.1 h d⁻¹ and S_max= 96.9%, and the range of prolonged light duration within which survival rate is >90% of the maximum were estimated between 3.7 and 4.6 h d⁻¹.

The PCA using the seven indices of the growth and propagation of Z. marina plants accounted for 70% of the variance in the data with the first two dimensions, and allowed for clear separation of the treatment groups (Figure 3A; Table 2). PC1 accounted for 59.8% of the total variance, and was defined primarily by ADP, BDP, RF and IER, which had the highest positive loadings, respectively. PC2 explained 15.0% of the variance in Z. marina growth and propagation, and was primarily defined by LA (highest positive loadings). The treatment of 4 h d⁻¹ produced more ramet (RF) and faster growth (ADP, BDP and IER) than other treatments.

The ADP, BDP, IER and RF of Z. marina plants exposed to different prolonged light durations increased firstly from 0 h d⁻¹–4 h d⁻¹ and then decreased at 8 h d⁻¹ (Figures 2B–E). The result of one-way ANOVA indicated that the prolonged light duration had a significant effect on the ADP, BDP, IER and RF of Z. marina plants (p < 0.05; Table 1). The greatest ADP, BDP, IER and RF attained were 9.7 mg DW shoot⁻¹ day⁻¹, 5.6 mg DW shoot⁻¹ day⁻¹, 2.5 mm/day and 41.2% under the 4 h d⁻¹ treatment, which was 1.5, 1.6, 2.1, and 1.5 times significantly higher than those attained under control. The curve fitting showed that there was a significant quadratic relation between the growth (ADP, BDP and IER) and propagation (RF) and the light prolonged durations (p < 0.01).

The range of prolonged light duration within which growth and propagation are >90% of the maximum were estimated, and the lower limit of IER and the upper limit of RF or SR were selected as the optimum durations range, that is, 3.9–4.6 h d⁻¹ (Figure 4).

The PCA using the eight indices of the physiological of Z. marina plants accounted for 90.7% of the variance in the data with the first two dimensions, and allowed for clear separation of the treatment groups (Figure 3B; Table 3). PC1 accounted for 82.6% of the total variance, and was defined primarily by aSS, bSS, bS and Chl a, which had the highest positive loadings, respectively. PC2 explained 8.1% of the variance in physiology of Z. marina. The content of chlorophyll and non-structural carbohydrate of Z. marina at 4 h d⁻¹ were higher than those at other treatments.

One-way ANOVAs indicated that prolonged light durations significantly affected the aSS, bSS, bS and Chl a of Z. marina plants (p < 0.05; Figure 5; Table 1). The aSS, bSS, bS and Chl a varied with durations of prolonged light, which were all the lowest at control (0 h d⁻¹). The highest values were observed at 4 h d⁻¹, which were 1.6, 1.5, 1.8, and 1.3 times higher than the average contents (p < 0.05).

Pearson correlation analysis showed a significantly positive correlation between the physiological indicators of Z. marina plants exposed to different prolonged light durations (Figure 6; Table 4). Mantel analysis showed that the survival rate and content of chlorophyll and starch of leaves were significantly affected by soluble sugar and starch of belowground tissues of Z. marina plants (p < 0.05).

Our observation results show that the promotion effect of seagrass growth and propagation is more significant under the condition of 4 h d⁻¹. Especially, the greatest internode elongation rate attained was 2.5 mm day⁻¹ at 4 h d⁻¹, which was 2.1 times significantly higher than that attained under control. Survival analysis combined with growth assessment indicated that the optimum range of prolonged light durations for the establishment of Z. marina plants is 3.9–4.6 h d⁻¹. Especially, the propagation coefficient of Z. marina plants exposed to the optimum range of prolonged light durations were 1.9–2.1 times higher than that under the control. Some studies on terrestrial vegetations demonstrated that prolonged light duration had a positive impact on the growth of plants. For example, leaves elongation and biomass of Nicotiana tabacum increased with the increase of prolonged light durations based on the natural light durations, in which the leaf length and plant height of plants exposed to 2 h d⁻¹ increased 7.2 and 5.5% compare to the control (0 h d⁻¹), respectively (Xu et al., 2019). Li et al. (2009) indicated that the seedlings of Cucurbita pepo obtained a higher dry weight growth rate (6.6 mg shoot⁻¹) under full sunlight (24 h d⁻¹), which was 2.0 times higher than that of plants under the treatment of 14 h d⁻¹.

Similar results are also found on submerged plants. For example, Dennison and Alberte (1985) conducted a field experiment and demonstrated that the leave biomass of Z. marina plants under prolonged light duration of 3.0–5.0 h d⁻¹ was 1.1 times higher than that of plants under the prolonged light duration of 4.0–6.0 h d⁻¹. These results suggest that appropriate prolonged light duration has an important role in promoting the growth and development of plants, yet overlong duration of prolonged light has no obvious effect on the growth of plants, even limits the plant growth. Our results found that no marked differences in survival rate, internode elongation rate, ramet frequency and productivity were found between the treatments of 8 h d⁻¹ and 0 h d⁻¹.

Some studies have found that the photosynthetic function, photosynthetic rate and dry matter synthesis of terrestrial plants reduced due to the light suppression under the excessive light duration, mostly limiting the extension of underground tissues (Digby and Dyson, 1973; Zhang et al., 2012; Nitschke et al., 2017). For example, root of Vitis champinii exhibited higher growth rate when they grew under 10 h d⁻¹ (up to 7.2 mm day⁻¹) than those under the 14 h d⁻¹ (Mahmud et al., 2018). The stem diameter of Vigna angularis exposed to the 12 h d⁻¹ was up to 1.8 cm, which was lower than that under natural light duration (Dong et al., 2015).

In addition, these results also found that the optimum range of prolonged light duration range of seagrasses was larger than that of terrestrial plants. To adapt the marine low light environment, seagrasses have lower light exploitation rate (3–10 mg O₂ DW⁻¹ h⁻¹; Vermaat et al., 1997), lower light requirement (no less than 10%; Duarte, 1991) and lower light compensation point compare with terrestrial plants (Jiang et al., 2013). Moreover, seagrasses can reach net photosynthetic production at lower light compensation point and make them grow normally (Vermaat et al., 1997; Alcoverro et al., 2001). Appropriate prolonged light durations could significantly promote the growth and development of plants when incident light intensity is above the light compensation point of plants (Birch et al., 1998; Searle and Coupland, 2014; Liu et al., 2021).

The results of our study revealed that an increase in the content of non-structural carbohydrates (starch and soluble sugar) and chlorophyll was found in the response of Z. marina plants to suitable prolonged light duration, which promoted the survival, growth and propagation. Especially, the aboveground soluble sugar and chlorophyll a content were 1.4–1.6 and 1.2–1.3 times higher than those under the control. Some studies on terrestrial vegetation demonstrated that non-structural carbohydrates, especially soluble sugars, not only exist widely in plants as signal molecules and regulate physiological processes such as growth and development of plants (Dennison and Alberte, 1985; Zhang et al., 2021), but also can be used as energy supply to promote growth and development of plants (Oberhuber et al., 1993; Hao and Papadopoulos, 1999). And it also provides a carbon skeleton for the synthesis of organic matter such as plant tissue proteins (Oliveira and Coruzzi, 1999). For example, these studies on the Cucumis sativus and Lycopersicon esculentum have proved that the increase of photosynthetic efficiency and the accumulation of starch and sugar promoted the growth of plants (Digby and Dyson, 1973; Demers et al., 1998; Pettersen, et al., 2010; Doğru, 2020). The soluble sugar of Helianthus annuus irradiated by red light for 20 h d⁻¹ increased by 38.6% compared with 12 h d⁻¹, indicating that the sugar content of plant leaves affected starch synthesis by regulating the activity of caramel acidification enzyme (AGPase) to regulate the growth of plants (Zhang et al., 2012).

Similar results are also found on submerged plants. For example, the longer photoperiod in summer could accumulate higher soluble sugar (as higher as 96.2 mg g⁻¹) for Thalassia hemprichii, which allowed it to adapt quickly to the environment (Jiang et al., 2013). Dennison and Alberte (1985) indicated that increased prolonged light durations had a positive impact on the maximum net photosynthetic rate (P_max) and chlorophyll content of Z. marina leaves under 3.0–5.0 h d⁻¹, and had a negative effect on P_max under 4.0–6.0 h d⁻¹. Consequently, the appropriate durations of prolonged light can improve the accumulation of soluble sugar in submerged plants and promote significantly their growth and respiration.

However, excessive prolonged light duration resulted in a lower photosynthetic efficiency and amount of soluble sugar synthesis of plants (Li et al., 2021b). The leaves of plants expressed photoinhibition during the excessive prolonged light durations with lower accumulation of photosynthetic pigments and synthesis of dry matter such as starch, resulting in slow growth and reducing ramet frequency of plants (Ruiz and Romero, 2003; Campbell et al., 2012). For example, the plants of Nicotiana tabacum exposed to the light duration of 3 h d⁻¹ indicated an obvious photoinhibition, in which the net photosynthetic rate was significantly lower than that of plants under the 2 h d⁻¹ (Xu et al., 2019).