The blood pressure [systolic (mmHg), diastolic (mmHg)] and heart rate of the participants were measured on the left arm using an electronic sphygmomanometer (OMRON HEM-7135) to indicate the relaxation degree of the body. The HR is considered a physiological response indicator to physiological stimuli and psychological stress (21). EEG is an oscillogram of electrophysiological signals that reflects the neurophysiological activities of brain neurons on the cerebral cortex or the scalp (22). It contains a large amount of information about brain activity that is closely related to human thinking, physical activity, mental state, and physical and mental health (23). Different wave bands can reflect different emotional states and cognitive abilities (24). The brain signals were measured in this study with a Wise Brain Instrument (Sichiray 03) at three waves (theta, alpha, and beta). Theta (θ) waves range from approximately 4–8 Hz and are typically associated with deep relaxation, meditation, and mild sleep stages such as (rapid eye movement (REM) sleep. They are often considered as the brainwaves related to creativity and subconscious thinking. At this frequency, the abilities of memory, learning, and emotional integration may be enhanced as the brain enters a state of marginal consciousness, which is conducive to the formation of new ideas and concepts. Alpha (α) waves, which range from approximately 8–12 Hz, indicate relaxation in a state of consciousness and can often get intensified while sitting down with the eyes closed. They are associated with feelings of relaxation, tranquility, and a balance of physical and mental states. Enhanced alpha wave activity suggests that an individual may be in a state of idle thinking or enhanced creativity. Alpha wave activity often increases when individuals are in contact with nature, performing deep breathing exercises, or engaging in relaxation activities. Beta (β) waves range from approximately 12–30 Hz and are associated with daily alertness, concentration, decision-making, and problem-solving activities. High-frequency beta waves are related to states of anxiety, stress, or panic. However, moderate beta wave activity is healthy, indicating that the brain is in an alert and active state, capable of effective logical thinking and critical analysis.

Scales are typically used to measure psychological restorativeness, with PRS as the most representative (Appendix A). Developed by Hartig et al. (25), the PRS contains four items and 26 projects (20). Due to the cultural differences and different language habits in the semantics of different countries, Wang (26) applied the Chinese version of PRS to the restorativeness assessment of urban park landscapes, validating the reliability of the Chinese version of PRS (27, 28). The PRS used for the psychological measurements in this study is based on the modified version of Huang et al. (29), consisting of 18 items in four dimensions (distance, extension, charisma, and compatibility), and modified according to the attention restorative theory by Kaplan et al. (27) (Appendix A).

This study further developed the PRS for low hydrodynamic landscapes (Haizi). Based on field investigations and literature reviews, the scale was ultimately determined to consist of three dimensions: the scale for the restorativeness of Haizi components; the characteristics of Haizi (Appendix B); and the elements of Haizi (Appendix C). The Haizi characteristics that vary continuously can generally be described by an interval, while the Haizi elements have discrete features. Each level is rated on a perception restorativeness scale from 1 to 7, where 7 represents the highest positive evaluation (considered the most restorative), and 1 represents the lowest evaluation (considered the least restorative).

Following the audiovisual interaction test of the low hydrodynamic landscape (Haizi), participants exhibit a significant decrease in heart rate (79.22 ± 13.30–75.59 ± 10.78, P < 0.01), systolic pressure (110.42 ± 9.22–103.81 ± 8.55, P < 0.01), and diastolic pressure (70.16 ± 9.47–67.52 ± 7.22, P < 0.05) (Figure 4). Following the pure visual test of Haizi, participants present a significant reduction in heart rate (84.50 ± 14.30–79.37 ± 12.03, P < 0.01), while systolic and diastolic blood pressure exhibit minimal changes (Figure 5).

After completing the audiovisual interaction of different hydrodynamic landscapes, the mean theta, alpha, and beta wave values are observed to increase greatly. Paired T-test results show marked differences (P < 0.05) between baseline and experimental theta, alpha, and beta wave data for the different hydrodynamic landscape tests. Moreover, the mean values of the three waves peak during the high hydrodynamic landscape (waterfall) test. However, no significant differences are observed in the corresponding ANOVA results (Table 2).

After the audiovisual interaction test (first 30-s video) of the different hydrodynamic landscapes, the mean values of the theta, alpha, and beta waves are observed to increase greatly. The three paired t-test results exhibit significant differences (P < 0.05) between the baseline data and the experimental theta, alpha, and beta wave values for the three landscape tests. Moreover, the mean values of the three waves peak during the high hydrodynamic landscape (waterfall) test. However, no significant differences are observed in the corresponding ANOVA results (Table 3).

The theta waves are observed to increase greatly for the second part of the audiovisual interaction test (second 30-s video) of the different hydrodynamic landscapes. The paired t-test results indicate significant differences (P < 0.05) between the baseline and experimental theta and alpha values for the three landscape tests, as well as between baseline and experimental beta wave values for the waterfall and combined landscape test. No significant differences are observed for Haizi. The mean values of the three waves also peak during the high hydrodynamic landscape (waterfall) test. Furthermore, there are no significant differences in the theta, alpha, and beta wave values for the three landscape tests (Table 4).

Comparing the mean values of the theta, alpha, and beta waves between the first and second 30-s tests indicate that the values of the three waves in the Haizi test decrease after the addition of the bird songs, and increase in the waterfall and combined landscape tests (Tables 3, 4).

In the tests focusing solely on the visual perception of Haizi, significant increases are observed in the mean values of the theta, alpha, and beta waves across the four Haizi landscapes (Mirror Lake, Wucaichi, Luwei Lake, and Wuhua Lake). Wucaichi exhibits the highest theta wave value, while the alpha and beta waves peak for Luwei Lake and Jing Hai, respectively. Paired t-test results reveal notable differences (P < 0.05) between the data in the baseline and experimental states for the three waves across the four Haizi landscapes. In contrast, the ANOVA results indicate no significant differences in theta, alpha, and beta waves among the four Haizi landscapes (Table 5).

According to the physiological data, all low hydrodynamic landscapes exhibit significant physiological restorative effects under pure visual conditions, with no significant differences observed in physiological restorativeness among the four Haizi landscapes. The highest values of theta, alpha, and beta brain waves are found in different Haizi. This may be attributed to both the similarities and partial differences in the features and elements of the waterscape environment of Haizi.

Table 6 reveals that all values of the four components are positive, with the plant and animal characteristics exhibiting the highest value (B = 0.362, P < 0.05). The overall restorativeness of Haizi can be calculated as −0.044 + 0.271 × [water characteristics] + 0.362 × [plant and animal characteristics] + 0.213 × [water environment] + 0.202 × [plant and animal environment]. The components of Haizi that contribute positively to overall restorativeness include animal and plant characteristics, water characteristics, water environments, and animal and plant environments, with animal and plant characteristics having the most significant influence on restorativeness. The other three components have similar effects (Table 6).

Appendix Table 7 shows that all the water characteristic non-standard coefficient values are positive, with perceived depth exhibiting the highest value (B = 0.240, P < 0.05). The restorativeness of water characteristics can be calculated as 1.978 + 0.046 × [clearness] + 0.240 × [perceived depth] + 0.178 × [openness] + 0.084 × [surface complexity] + 0.144 × [color richness]. The characteristics that contribute to restorativeness include water transparency, perceived depth, openness, surface complexity, and color richness, with perceived depth having the greatest effect (Appendix Table 7).

Appendix Table 8 indicates the plant color richness value to be negative (B = −0.163, P > 0.05), while the other values are positive. Plant density and plant level richness exhibit the highest values (B = 0.419, 0.416, P < 0.05). The restorativeness of the plant and animal characteristics can be calculated as 0.934 + 0.419 × [plant density + 0.416] × [plant level richness] + −0.163 × [plant color richness] + 0.091 × plant form richness. Plant form richness, plant density, and plant level richness exhibit positive restorativeness, with the latter two exerting the greatest effects, while plant color richness has a negative restorative effect (Appendix Table 8).

Appendix Table 9 indicates that watercolor, waterfront elements, water shape, and reflection have positive effects on restorativeness, with watercolor and reflection exerting the strongest influence. In contrast, the component individual scene has a negative effect (Appendix Table 9). The restorativeness of the water environment can be calculated as 0.739 + 0.348 × [water color] + 0.347 × [waterfront elements] + 0.052 × [water shape] + 0.373 × [water reflection] + −0.219 × [individual scene].

Further analysis was conducted on the components of water reflection, watercolor, and individual scenes. In terms of the water reflection components, Appendix Table 10 shows that clouds have the most significant effect on restorativeness (B = 0.435, P < 0.05), while plants on the water have a negative effect (Beta = −0.023, P > 0.05). The restorativeness of water reflection can be calculated as 1.101 + −0.023 × [plants on the water] + 0.219 × [waterfront plants] + 0.453 × [clouds + 0.201] × mountains (Appendix Table 10). The watercolor components are observed to have a positive effect, with green having the greatest effect (B = 0.306, P < 0.05) (Appendix Table 11). The restorativeness of watercolor can be determined as 2.238 + 0.222 × [mainly blue] + 0.306 × [mainly green] + 0.154 × [mainly yellow]. Among the individual scene components, plank roads over water exert the greatest negative effect (B = 0.381, P < 0.05) (Appendix Table 12). Note that the values in Appendix Table 12 indicate negative impacts of the individual scene components, and the higher the value, the greater the negative impact. The negative impact of the individual scene components is calculated as1.369 + 0.236 × [plank roads by the shore] + 0.381 × [plank roads over water] + 0.065 × [viewing platform with guardrails] + 0.051× [fallen trees in the water].

Analysis of the restorativeness of the plant and animal environments shows that both animal and plant species exert positive (and similar) effects (B = 0.547, 0.543, P < 0.05). The restorativeness of plant and animal environments can be calculated as −0.674 + 0.543 × [plant species] + 0.547 × [animal species] (Appendix Table 13). Through further analysis, it is revealed that submerged plants, emergent plants, wetland shrubs, and wetland trees have significant positive effects, with wetland shrubs exerting the strongest influence (B = 0.373, P < 0.05). The restorativeness of plant species can be calculated as 0.435 + 0.044 × [submerged plants] + 0.212 × [emergent plants] + 0.373 × [wetland shrubs] + 0.352 × [wetland trees] (Appendix Table 14). Terrestrial and fish are observed to have remarkable effects on restorativeness, with terrestrial demonstrating the most significant influence (B = 0.455, P < 0.05). The restorativeness of animal species can be calculated as 0.760 + 0.455 × [terrestrial] + 0.433 × [fish] (Appendix Table 15).

The components of low hydrodynamic landscapes (Haizi) in Jiuzhaigou (animal and plant characteristics, water characteristics, water environment, and animal and plant environment) have a positive impact on restorativeness, with animal and plant characteristics exerting the most significant impact. In a comparison of the restorativeness between single-form natural visual elements, Ulrich et al. (39) found that water elements have more positive effects in terms of emotion regulation compared to vegetation elements. Low hydrodynamic landscapes (Haizi) in Jiuzhaigou are natural spaces containing rich flora and fauna and large areas of water. Plants and water can enhance mental restorativeness (40) as they can provide a comfortable environment by improving the microclimate, thus positively impacting the restorativeness of human health (41, 42). Our results revealed that for low hydrodynamic landscapes (Haizi) with abundant vegetation and large-area water elements, the restorative effects of animals and plants are greater than those of water bodies. In nature, water cannot exist independently from other features, such as plants and animals. Therefore, even spaces with water as the main landscape type have differences in waterscape vision due to distinct near- and distant-view backgrounds, including color composition (43). Plants and wildlife associated with water can strongly affect individuals' positive cognition of water spaces (44), with viewers feeling a sense of belonging and attachment (45, 46), thus enhancing the attractiveness of water landscapes (47). This is in line with the Biophilia Hypothesis, which holds that humans have an inherent tendency to get close to and love natural life (plants and animals) during the long-term evolution process. This tendency is related to survival instinct. Environments with a great diversity of plants and animals by water (e.g., trees, shrubs, flowers, butterflies, and birds) have a higher level of biodiversity, which can reduce mental stress and improve the environmental restorative effects (48). Jiuzhaigou, a World Natural Heritage Site, is renowned for its rich biodiversity and the presence of rare species (49). The low hydrodynamic landscapes (Haizi) in Jiuzhaigou boast abundant animal and plant resources, a diverse ecosystem, and a superior natural environment, all of which contribute significantly to the restoration of youth with high stress levels.

The perceived depth and openness of water, as well as the density and level richness of plants, have positive effects on landscape restorativeness. In contrast, plant color richness is considered to have a negative effect on landscape restorativeness. Acquired knowledge and experience play an important role in people's understanding of the environment (50). When people can clearly perceive the depth of water, they can then determine the danger level of the water body and whether it is suitable for recreation (51). An open space is likely to provide a sense of security and control due to a wide view (52), which is in line with the need for viewing and sheltering (53). Environments with a wider view and less sheltering induce greater restorativeness than suburban park environments with a narrower view and more sheltering (54). Furthermore, the more visible trees, the greater the restorativeness (40). The density of green plants and the richness of plant layers are both related to the number of plants. Both high-density vegetation and richly layered plant communities can attract attention and evoke positive emotions (55). Our study finds that the higher the perceived plant color richness in space, the lower its perceptual restorative effects. This may be because lower plant color richness makes it easier to create harmonious spatial colors that do not induce a sense of disorder (56). It also facilitates improved coordination of plants with the surrounding features, resulting in a more orderly environment that aids in reducing the need for directed attention and promotes relaxation in both physiological and psychological states. Consequently, plant color serves as a crucial indicator in enhancing public mental health (57).

As a component of water reflection, clouds have a significant impact on restorativeness. Reflection in water is the embodiment of symmetrical beauty and form, placing minimal strain on the eyes of viewers. The comfortable and labor-saving perceptual experience can bring viewers pleasure. Moreover, due to the balanced rhythm among symmetrical objects, viewers can get a sense of order as well as a feeling of control during the perceptual experience. The pursuit of beauty in form by humans is based on a sense of order and security, which is also the most fundamental aesthetic consciousness of humans (58). For waterscapes, watercolor is a significant factor affecting the perceptual properties of water. According to neuroaesthetics, the optic nerve responds fastest to color in external stimuli, with responses reported as 30 ms faster than those to motion (59). We identified green water bodies as having a greater perceptual restorativeness for youth. Blue and green are the universal color preferences among people due to genetic codes related to blue skies and green nature in human evolution. In particular, these two colors exert a sensory calming effect (60).

There is a negative correlation between plank roads over water and restorativeness. When people perceive the environment, they are more inclined to perceive the overall picture. However, plank roads over water destroy the integrity of the visual perception of landscapes for viewers at a distance. In addition, being exposed to remote viewing and becoming a member of the observed objects can reduce individuals' sense of security, which goes against humans' “viewing–sheltering” needs. Plank roads over water may also impact the microbial environment and living habits of aquatic animals due to the materials and technology used to construct the roads (e.g., unpleasant smells). This may lower the level of biodiversity and even damage the ecological environment, further affecting the restorativeness of water spaces. Therefore, it is necessary to carefully use the facilities of key ecological environments such as the Jiuzhaigou World Natural Heritage Site to protect the long-term ecosystem value.

The species richness of animals and plants is a key indicator of general species richness. Parasympathetic nerve activity is stronger in environments with higher species richness, which can significantly reduce the anxiety level of youth (61) and improve the restorative effects for introverts (62). Enhancing the number of plant and animal species can increase human preferences and induce other positive perceptions. The greater the number of plant and animal species, the higher the perceived naturalness (56, 63). Moreover, planting more vegetation in a space can increase its influence in reducing anxiety (64). In low hydrodynamic landscapes (Haizi), trees significantly increase people's preference for interacting with nature, effectively enhancing the restorativeness of the environment. Rich plant and animal species provide rich biodiversity and are associated with positive emotional effects, suggesting that higher biodiversity can lead to greater restorativeness and lower physiological stress (65).