Tourist trampling reduces soil hydrolytic enzymes across three vegetation types at different successional stages on Jigong Mountain

1 Introduction

Anthropogenic disturbance has become a pervasive stress in terrestrial ecosystems globally (1, 2). As economic development accelerates, tourism- induced disturbances have emerged as a significant ecological problem, particularly on soil structure and function (3). In tourism areas, soils are the primary place of human activities and are susceptible to direct impacts from trampling. Soil structural integrity and functional quality are fundamental to ecosystem stability and resilience (4, 5). The impacts of tourism on soil function occur through various mechanisms, such as visitor trampling, vehicular compression, infrastructure development, and associated pollution (6). All these processes can alter soil physicochemical properties, modify microbial community structure, and also affect soil enzyme activity (7, 8).

As key mediators of soil biochemical processes, soil enzymes represent the most labile bioactive components regulating soil nutrient cycling and energy transformations in soil subsystems (9, 10). Soil enzyme activity directly reflects microbial metabolic rates and affects fundamental nutrient cycling processes (11–13). Previous studies have found that soil enzyme diversity and activity influence ecosystem functions, including nutrient mineralization, humus formation, and organic matter transformation (14, 15). Hydrolytic enzymes in soils are capable of decomposing macromolecular organic compounds, including cellulose and starch, into soluble low-molecular-weight substances (16). This process constitutes a crucial link in soil nutrient mineralization and plays a vital role in sustaining soil functionality and promoting plant growth. The activity of soil hydrolytic enzymes exhibits high sensitivity to changing environmental factors, especially physical disturbance. For example, tourism trampling directly changes soil physical properties through compaction, thus affecting microbial communities and enzymatic processes (17). Some studies demonstrated that the response of soil enzyme activity to trampling was dependent upon trampling intensity, showing greater reduction of enzyme activities under high-intensity trampling, probably due to the marked biochemical degradation under severe disturbance (3, 18).

Tourism scenic areas are generally located in scenic landscapes with heterogeneous vegetation and plants to meet diverse aesthetic preferences (1). The distinct physiological and ecological characteristics of different vegetation types can regulate soil properties through the following aspects: (1) differential litter chemistry can regulate decomposition rate (19, 20); (2) variable litter quality can change nutrient cycling rates (21, 22); and (3) species-specific root exudates can affect soil microbial communities (23). Therefore, the differences in plants between vegetation types are regarded as a critical factor affecting soil enzyme activity. For example, deciduous broadleaf forest showed greater microbial and enzymatic activity than coniferous forest due to the differential litter decomposition rates (21). Although trampling occurs widely in different vegetation types within ecotourism hotspots, it is still unknown whether the effects of tourist trampling on soil enzymes are vegetation-dependent. Since forests have higher aboveground and belowground biomass than grasslands, making soil microorganisms more readily able to obtain carbon substrates, we hypothesize that the changes in soil microbial extracellular enzyme activities under trampling disturbance may be smaller in forest soils than in grassland soils. To address the above questions, we conducted a field investigation of trampling effects on soil hydrolytic enzymes across three vegetation types (planted grasslands, secondary deciduous forests, and secondary coniferous forests) in Jigong Mountain (Central China). The objectives were to: (1) quantify trampling-induced changes in soil physicochemical properties across three different vegetation types; (2) determine vegetation-specific response differences of soil hydrolytic enzyme activities to trampling; and (3) elucidate the relationship between soil physicochemical properties and changed soil hydrolytic enzyme activities due to trampling.

2 Materials and methods

2.1 Site description

This study was conducted in the Jigong Mountains (31°51′58″ N, 114°02′21″ E), situated in Henan Province, Central China (Figure 1). These mountains represent a western extension of the Dabie Mountains. The study area features notable topographic variation, with its highest point at 810 m a.s.l. The regional climate features a mean annual temperature of 15.2 °C and an annual precipitation ranging from 800 to 1100 mm (24). The dominant tree species include Quercus variabilis, Pinus massoniana, and Liquidambar formosana, while the predominant shrub species comprise Vitex negundo, Viburnum dilatatum, and Celtis sinensis (25, 26).

2.2 Experimental design

In 2023, experimental surveys were conducted on three vegetation types in the Jigong Mountains, namely planted grasslands (G), Pinus massoniana coniferous forests (CF), and Quercus variabilis deciduous broad-leaved forests (DF). To some extent, these three vegetation types can be considered representative of communities at different successional stages. Three vegetation types were distributed at comparable altitudes (720–730 m) and were all situated near the main tourist roads within the scenic area (Figure 1). The planted grassland was primarily established with the perennial grass species Lolium perenne. Before grassland establishment, these regions were formerly shrublands dominated by Vitex negundo and Litsea cubeba, which were converted to planted grasslands in 2018, five years before the experiment. The P. massoniana coniferous forest was a natural secondary forest, with a stand age of approximately 65 years. The Q. variabilis secondary forest has a stand age of around 70 years. Both secondary forests naturally regenerated on bare land following large-scale clear-cutting in the 1960s.

During tourism activities, visitors frequently left the boardwalk and entered these vegetation zones for sightseeing or photography. Tourist trampling was relatively common within the scenic area and occurred across different vegetation types. Although it occupied only a small proportion of the total area (about 5%), these scattered trampling paths fragment the ground into distinct patches, exerting strong negative impacts on understory plants and soil microorganisms. Some trails have been formed by tourist trampling in all three vegetation types, with widths ranging from 30 cm to 120 cm. Therefore, a paired experiment was designed in each vegetation type, with clearly demarcated non-trampling and severely trampled areas. According to the manager’s records, approximately 5–7 people traverse these trampled footpaths daily, resulting in sustained trampling impacts on the soil. In each vegetation type, six areas with distinct tourist trampling were selected as trampling plots, spaced approximately 50 m apart from each other to minimize spatial autocorrelation as much as possible (Figure 1). The primary characteristics of the trampling plots were as follows: the soil exhibited obvious compaction, there was no litter on the ground surface, and no herbaceous plants were growing. Approximately 5 m away from each trampling plot, one area with no trampling was selected and designated as the control plot.

The control plots generally exhibited no visible signs of human disturbance, with loose soil structure and a uniform layer of ground litter. In the planted grassland, the control quadrats contained almost no plants other than the sown ryegrass. In the control plots of both secondary forests, a small number of naturally occurring herbaceous plants were present, such as Carex siderosticta and Oplismenus compositus. The size of each plot was set at 1 m × 1 m. We measured soil penetration resistance (PV6.08, Eijkelkamp Giesbeek, Netherlands) of 0–20 cm to quantify soil compaction, and we found that soil penetration resistance in the untrampled plots of grassland, coniferous forest, and deciduous forest was 1.07, 1.17, and 1.14 Mpa, respectively. In contrast, soil penetration resistance in the trampled plots of grassland, coniferous forest, and deciduous forest was 1.69, 1.52, and 1.49 Mpa, respectively.

2.3 Analysis of soil physical and chemical variables

Five sampling points were randomly established in each plot for both control and trampling treatments in July 2023. No natural precipitation occurred one week before sampling. Before soil sampling, surface litter was carefully removed to minimize its influence. Soil samples from the 0–10 cm depth layer were collected from the five points and thoroughly mixed to represent each plot (27). In total, 36 soil samples were collected (2 trampling types × 3 vegetation types × 6 replications). The mixed samples were passed through a 2-mm soil sieve to eliminate plant residues and stones. Approximately 200 g of the sieved soil sample was retained and evenly divided into two parts: One portion was placed in a self-sealing bag and transferred to the laboratory, and then the soil moisture content (%) was determined using the gravimetric oven-drying method (28). Fresh soil samples were dried at 105 °C to a constant weight and then weighed (8). The remaining fresh soil was stored in a refrigerator at 4˚C to determine the activity of soil hydrolytic enzymes (29).

The activity of soil sucrase was determined by the 3,5-dinitrosalicylic acid colorimetric method (30). Soil urease activity was measured using the sodium phenolate colorimetric method (30). The activity of phosphatase was determined by the sodium phenylphosphate method (30). The other portion was naturally air-dried for 20 days and then sieved through a 100-mesh soil sieve to determine soil physical-chemical properties (8). The soil pH was determined using the electrode potential method [soil:water=1:2.5] (31). The electrical conductivity was measured with a P4 multifunctional meter. The soil organic carbon content was determined using the potassium dichromate oxidation method with external heating (32). Total nitrogen was measured via the Kjeldahl method (33). The total phosphorus content in the soil was determined using the NaOH fusion-molybdenum-antimony anti-colorimetric method (34). Available phosphorus was measured by the NaHCO₃ extraction-molybdenum-antimony anti-colorimetric method (35, 36). The ring-knife method was employed to measure soil bulk density and total porosity (%) within the 0–10 cm soil layer during soil sample collection (8). Soil microbial biomass C (MBC) was measured by the chloroform fumigation-extraction method (37). For each sample, extractable C of both filtered K₂SO₄-extracts (fumigated sample and unfumigated sample) was analyzed by a TOC analyzer (TOC-VCPH Shimadzu Corp., Japan) (38).

2.4 Statistical analysis

The Kolmogorov-Smirnov (K-S) test was performed to assess the normality of the data, and the Bartlett’s test was utilized to evaluate the homogeneity of variance. After the analysis, all data met the prerequisite conditions for variance analysis. Two-way analysis of variance (ANOVA) was used to examine the impacts of vegetation type and trampling on soil physical-chemical characteristics as well as soil hydrolytic enzyme activities. Paired-samples T-tests were then conducted to analyze the differences between the control and trampling plots within each vegetation type. Pearson correlation analysis was carried out to investigate the correlations between soil physical-chemical properties and soil hydrolytic enzyme activities. Linear regression analysis was used to test the relationships between soil enzyme activities, soil bulk density and microbial biomass carbon. Finally, multiple stepwise regression analyses were conducted to determine the main soil factors regulating the response of soil hydrolytic enzyme activities to trampling. All statistical analyses were conducted using SPSS 26.0 software (SPSS Inc., IL., USA).

3 Results

3.1 Soil physical-chemical characteristics

Vegetation type showed an effect on all soil physical and chemical properties (P < 0.05, Table 1) except for soil bulk density and total soil phosphorus. Regardless of trampling effects, deciduous forest soil exhibited the highest values of soil moisture, soil organic carbon, total nitrogen, and available phosphorus content compared to the other two vegetation types (Table 2). These values were all significantly higher than those in planted grasslands (P < 0.05, Table 2). The coniferous forest soil had the lowest average pH value. The electrical conductivity of the deciduous forest soil was significantly higher than that of the coniferous forest soil (P < 0.05, Table 2), whereas it showed no difference from the grassland soil. The soil pH of the coniferous forest soil was significantly lower than that of planted grassland soils (P < 0.05, Table 2). The soil organic carbon content of the coniferous forest was significantly greater than that of the planted grassland soil (P < 0.05, Table 2).

Table 1. The influence of tourist trampling, vegetation types, and their interaction on soil physical and chemical characteristics based on two-way ANOVA (P value).

EC, electrical conductivity; SM, soil moisture; BD, bulk density; SOC, soil organic carbon; TN, total nitrogen; TP, total phosphorus; AP, available phosphorus.

Table 2. The influence of tourist trampling on the physical and chemical properties of soil in different vegetation types.

Values marked with different lowercase letters indicate significant differences between control and trampling in the same vegetation type (P < 0.05). Different capital letters indicate significant differences among vegetation types for each variable. EC, electrical conductivity; SM, soil moisture; BD, bulk density; SOC, soil organic carbon; TN, total nitrogen; TP, total phosphorus; AP, available phosphorus; C, control; T, Trampling.

Tourist trampling showed significant effects on most soil physicochemical variables (P < 0.05, Table 1), except for soil total phosphorus and available phosphorus contents. Trampling led to substantial reductions in soil pH, soil moisture, soil organic carbon, total nitrogen, and microbial biomass carbon for all vegetation types (P < 0.05, Table 1). In contrast, it caused significant increases in soil electrical conductivity and soil bulk density (P < 0.05, Tables 1, 2). Compared to the planted grassland and coniferous forest, the deciduous forest had the greatest reduction in soil organic carbon (27.9%) but the lowest decrease in soil pH due to tourist trampling (Table 2). For the coniferous forest, tourist trampling led to the largest decline in both soil moisture (20.9%) and total nitrogen (34.8%). Tourist trampling resulted in the highest increase in soil electrical conductivity in the planted grassland. The reduction of microbial biomass carbon in grassland (11.6%) was less than that in deciduous forest (24.0%) and coniferous forest (21.4%).

3.2 Soil hydrolytic enzyme activity

Significant differences in soil hydrolytic enzyme activities were observed among different vegetation types (P < 0.05, Figure 2). Regardless of the trampling effect, deciduous forests had the highest values of soil sucrase and urease enzymes (P < 0.05, Figures 2a, b). Specifically, they were 13.0% and 7.1% higher than those in coniferous forests, and 42.3% and 42.1% higher than those in planted grasslands. In addition, soil sucrase and urease enzymes in coniferous forests were 25.6% and 33.4% higher than those in planted grasslands (P < 0.05), respectively. The alkaline phosphatase enzyme in coniferous forests was the lowest, being 15.4% lower than that in deciduous forests (P < 0.05, Figure 2c).

Tourist trampling led to differential degrees of decline (11.7-53.0%) in the soil hydrolytic enzyme activities of the three vegetation types (P < 0.05, Figure 2). Compared to deciduous forests and planted grasslands, trampling resulted in the highest decline in soil sucrase (32.6%) and urease enzymes (50.9%) in coniferous forests (P < 0.05, Figures 2a, b). The reduction of soil urease enzyme in deciduous forests due to trampling was greater than that in coniferous forests and planted grasslands. The decline in the three soil hydrolytic enzymes in planted grasslands caused by trampling was lower than that in coniferous forests and deciduous forests (Figure 2). Furthermore, trampling had diminished the differences in soil hydrolytic enzymes among the three vegetation types.

3.3 Relationships between soil hydrolytic enzyme activity and physicochemical characteristics

In planted grasslands, soil sucrase and alkaline phosphatase enzymes showed significant positive correlations with soil organic carbon, total nitrogen, and pH, whereas they exhibited negative correlations with soil electrical conductivity and bulk density (P < 0.05, Figure 3A). Soil urease enzyme was significantly and positively correlated with soil moisture and soil organic carbon (P < 0.05, Figure 3A). In coniferous forests, all three soil hydrolytic enzymes were substantially positively related to soil total nitrogen, soil moisture, and pH; however, they were negatively correlated with electrical conductivity and soil bulk density (P < 0.05, Figure 3B). In deciduous forests, soil sucrase enzyme was positively correlated with soil organic carbon and total nitrogen (P < 0.05, Figure 3C). Soil urease and alkaline phosphatase enzymes were significantly positively correlated with soil organic carbon and pH, while being negatively related to soil electrical conductivity and bulk density (Figure 3C).

The linear regression analysis between soil hydrolytic enzyme activity and soil bulk density showed that the regression slope in coniferous forest was higher than that in planted grassland and deciduous forest (Figures 4A–C). The regression slope of grassland was lower than the other two types (Figure 4). In addition, the regression analysis between soil hydrolytic enzyme activity and microbial biomass carbon showed that secondary forest had greater regression slopes than grassland (Figures 4D–F).

Based on the multiple stepwise regression analyses, we found that in planted grasslands, soil total nitrogen and moisture were the main factors influencing soil hydrolytic enzymes (P < 0.05). Soil total nitrogen and bulk density explained 94%-99% of the variations of soil hydrolytic enzymes in coniferous forests, while soil total nitrogen and electrical conductivity were the main factors regulating soil hydrolytic enzymes in deciduous forests (P < 0.05, Table 3).

Table 3. The results of multiple stepwise regression analyses of soil enzyme activities and soil physical and chemical properties under different vegetation types.

TN, total nitrogen; SM, soil moisture; BD, bulk density; SOC, soil organic carbon; EC, electrical conductivity.

4 Discussion

4.1 Effects of vegetation types

Our study found significant differences in soil hydrolytic enzyme activities among three vegetation types, showing higher soil hydrolytic enzyme activity in deciduous forests than in coniferous forests and planted grasslands (Figure 2). This finding was consistent with previous research conducted in forest and grassland ecosystems (39–41). This could be ascribed to the following two aspects. First, compared to planted grasslands, both deciduous broad-leaved and coniferous forests had a continuous input of substantial amounts of litter into the soil each year (23). In planted grasslands, however, the quantity of litter is relatively scarce, resulting in a relatively low contribution to soil nutrient replenishment (42, 43). Plant species in the grassland have a greater cellulose/lignin content in their residual roots and above-ground litter (44), which may result in a slow decomposition rate. In addition, the grassland had lower soil moisture (Table 2), which was also unfavorable for microbial activity. Second, the woody plants in secondary forests possess deep and extensive root systems (45, 46). Notably, the high turnover rate of fine roots not only directly secretes organic matter but also serves as a crucial source of organic matter upon their turnover. In addition, these roots can markedly improve soil structure, forming stable aggregates and increasing soil porosity (47). A well-structured soil environment facilitates water infiltration and gas exchange, and thus favors root and microbial activities (48), all of which are fundamental to high soil hydrolytic enzyme activity. In contrast, although grassland plants have dense fibrous root systems, their roots are predominantly distributed in the surface layer of the soil (49). Moreover, most of these roots are fine roots with a short lifespan and rapid turnover.

Deciduous broad-leaved forests exhibit the highest hydrolytic enzyme activity, which is mainly associated with soil total nitrogen content (Table 3). On one hand, the litter in deciduous forests generally contains higher values of nutrients (including N and P), showing a lower C/N ratio, and less lignin compared to the litter in coniferous forests (50), which is helpful for soil nitrogen improvement. Moreover, litter from deciduous trees generally has a relatively simpler chemical composition. These properties make the deciduous litter easier to decompose and utilize by soil microorganisms. As a result, it contributes to an increase in the biomass and activity of microorganisms (Table 2). Since microorganisms are the primary producers of soil hydrolytic enzymes, the enhanced microbial activity directly leads to augmented synthesis and secretion of soil hydrolytic enzymes. On the other hand, the decomposition of needle litter from coniferous forest typically generates a relatively large amount of organic acids, which causes the soil to become acidic (lower pH value, Table 2). In contrast, the decomposition of deciduous broad-leaved forest litter usually produces a smaller quantity of organic acids, and the soil pH is closer to neutral or mildly acidic (51). The majority of soil hydrolytic enzymes, especially hydrolases, show peak activity in a neutral to mildly acidic environment. The strongly acidic soil environment in coniferous forests (Table 2) inhibits the activity of numerous key soil hydrolytic enzymes. Additionally, the acidic environment can also exert an impact on the composition and activity of the microbial community.

4.2 Effects of trampling

Our findings were in line with the results of previous studies conducted in an alpine meadow (52), which also reported a decreased soil hydrolytic enzyme activity under trampling. These negative effects of trampling may be attributed to the following two reasons. The first explanation is the substrate removal hypothesis. Trampling activities can result in the disappearance of the soil humus layer, a reduction in soil thickness, and an increased soil bulk density (8). Soil organic matter gradually declined, mainly attributed to the low input of litter and the continuous depletion of readily decomposable components. As a consequence, soil microorganisms can only utilize a common pool of soil organic matter for their survival. This could lead to a decline in both microbial abundance and activity, thereby exerting a negative influence on soil hydrolytic enzyme activity. Another reason is the physical filtering hypothesis. Soil compaction from trampling decreases spatial heterogeneity in water and oxygen availability, creating a physically uniform environment that supersedes biological drivers of enzyme synthesis. This physical dominance ultimately alters soil enzyme activities, decoupling them from their typical biotic determinants and aligning them instead with the homogenized soil physical conditions.

It was noteworthy that soil hydrolytic enzymes in the three vegetation types showed varying degrees of response to trampling. The negative response of soil hydrolytic enzyme activity in planted grasslands to trampling was weaker compared to that in coniferous forests and deciduous forests. This could be ascribed to the following two aspects. First, the soil underwent extensive land preparation when the planted grassland was established, which could reduce the baseline of soil microbial activity (53). Moreover, tourist trampling during the grassland establishment phase also had a direct impact on the soil structure. Second, herbaceous plant roots are generally shallower than those of trees, making them more sensitive to disturbance from trampling (17, 46, 54).

Another interesting finding of this study was that the differences in soil hydrolytic enzymes among the three vegetation types tended to diminish after trampling (Figure 2). This implies that despite significant differences occurring among different vegetation types, the values were relatively close after trampling. It implies that the spatial heterogeneity of soil microbes was reduced under trampling disturbance. These could be attributed to the following reasons. First, trampling substantially increases soil bulk density and reduces soil porosity (Table 2), which collapses the unique water–oxygen gradients that normally distinguish surface soil in deciduous forests from that in coniferous forests. This homogenization of physical microenvironments suppresses specialist microbial taxa that depend on site‐specific conditions, while favoring stress‐tolerant generalists. Second, trampling interrupts the vertical transfer of fresh litter inputs, thereby decreasing the supply of vegetation‐specific organic substrates. Due to the decreased and diverse carbon sources from different plants, microbial communities increasingly rely on older, recalcitrant soil organic matter, which induces a narrower and more uniform set of hydrolytic enzymes across all three vegetation types.

Pearson correlation analysis showed that both soil organic carbon and total nitrogen contents were positively correlated with soil enzyme activities. However, soil organic carbon and microbial biomass carbon were excluded from the models in the multiple stepwise regression analysis, while total nitrogen plays a consistently important role across different vegetation types within the study area. In addition, neither total phosphorus nor available phosphorus exhibited a direct relationship with soil enzyme activities in the correlation analysis, implying that soil phosphorus is not a limiting factor for soil microorganisms in this region. Compared with grassland, soil microbial carbon was more closely associated with enzyme activities in coniferous and deciduous forests, indicating that trampling-induced changes in soil microbial biomass exert a more pronounced influence on enzyme activities in forest ecosystems.

4.3 Limitation of the study

It is worth noting that there remain certain limitations in the experimental design of this field survey. First, the three vegetation types selected in this study differ in terms of litter production, fine root biomass, and litter decomposition rate. This study used soil organic carbon, total nitrogen, and phosphorus contents as indicator factors and did not directly evaluate the impact of vegetation differences. Second, while the trampling treatment was derived from visitor-induced trampled zones rather than a standardized experimental design, which may have introduced limitations in the comparison between different vegetation types. Third, several important soil indicators, such as soil temperature and trace elements, were not monitored in this experiment. For example, grasslands typically exhibit higher soil temperatures than adjacent forested sites in the same region due to the absence of a canopy cover, which would presumably support greater soil microbial activity. However, in this study, grassland soils consistently showed the lowest microbial biomass and hydrolytic enzyme activities (Table 2; Figure 2), regardless of trampling treatment. This suggests that substrate availability, driven primarily by the specific vegetation type, may play a more dominant role in regulating enzyme activity than microclimatic factors such as temperature.

5 Conclusions

This study evaluated the effects of tourist trampling on soil enzymatic activities across three vegetation types with different succession stages in the Jigong Mountains. Our results showed that trampling suppressed soil hydrolytic enzyme activities across all vegetation types. Specifically, soil hydrolytic enzymes in planted grasslands exhibited greater resistance to trampling compared to secondary forest ecosystems, likely attributable to their lower fine root biomass and soil total nitrogen content. A key finding was that the trampling-induced homogenization of soil hydrolytic enzyme activity patterns among vegetation types reflected a weakened spatial heterogeneity in belowground biochemical processes. This suggests that chronic recreational pressure may reduce the intrinsic functional diversity of soil microbial communities across different vegetation types. Importantly, trampling-induced alterations in soil nitrogen content showed strong correlations with soil hydrolytic enzymatic responses, highlighting the tight coupling between soil nitrogen and microbial metabolic function in disturbed soils.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Rodrigo Gouvea Taketani, Rothamsted Research, United Kingdom

Reviewed by: Zifeng Huang, Fujian Agriculture and Forestry University, China

Shun Liu, Chinese Academy of Forestry, China