University campus green spaces’ potential for biodiversity conservation and ecosystem services

1 Introduction

Urbanization is widely recognized as a major driver of global biodiversity decline. With the expansion of cities, natural habitats are converted into built-up areas that lead to habitat loss, fragmentation, and disruption of ecological processes (Ren et al., 2023). This also promotes biotic homogenization, where diverse native communities are replaced by a limited set of species that are more adapted to urban conditions (Iwachido et al., 2025). Habitat fragmentation and degradation of ecological corridors further restrict species regeneration causing biodiversity loss (Beltrão et al., 2024). Urbanization also causes loss in vegetation biomass (Seto et al., 2012). In rapidly developing regions such as India, accelerated land-use change exacerbate these challenges. Therefore, understanding biodiversity patterns within urban landscapes is critical for designing sustainable cities.

Within this context, university campuses represent important ecological refuges in urban environments. Unlike many public green spaces, campuses are relatively protected areas and maintain long-term stability, allowing them to support local biodiversity (He et al., 2026; Liu et al., 2021). These landscapes not only conserve biodiversity but also serve educational, cultural, and research functions (Liu et al., 2021). Campus green spaces play a critical role in urban ecosystems by providing a wide range of ecosystem services, including climate regulation, air purification, carbon sequestration, and stormwater management (Shrimathi et al., 2023; Nguyen and Truong, 2024). These also help in mental well-being and social interaction among users (Zhang et al., 2024; Speake, 2013).

A brief rapid review using PRISMA method (Page et al., 2021) identified 48 recent studies on the role of universities in biodiversity conservation (Supplementary Appendix Figures A1, A2). The highest number of studies were reported from USA (8), followed by Italy and China (4 each). The dataset reflects global participation with uneven intensity, where influence or activity is concentrated in a few countries while many others contribute at a smaller scale. There were total 3 studies from India.

Recent literature shows an increase in campus biodiversity studies, however, there is a geographical skew in how these ecosystems are assessed (Liu et al., 2021). Developing countries often lack long-term, holistic ecological monitoring metrics that go beyond basic floral and faunal surveys. In northern India, the ecologically sensitive Aravalli landscape is subjected to severe environmental stress due to its proximity to rapidly expanding urban centers, including Delhi, Gurugram, and Jaipur (Bathla, 2024). These intense anthropogenic pressures have fragmented natural habitats, reduced vegetation cover, and disrupted local hydrological systems. Furthermore, invasive species such as Neltuma juliflora (formerly Prosopis juliflora) have aggressively colonized these degraded tracts, displacing native plant communities and diminishing overall ecosystem resilience (Sharma and Chaudhry, 2018). In this context, university campus landscapes influenced by the Aravallis represent invaluable ecological refuges.

To address the regional ecological monitoring gap, the present study in Jawaharlal Nehru University (JNU) was planned. By systematically documenting species, mapping multi-decadal vegetation cover trends, and evaluating stakeholder perceptions, this research identifies key opportunities for active ecological restoration within fragmented urban systems. The findings are expected to inform conservation planning, enhance campus management strategies, and contribute to the broader discourse on sustainable urban landscapes. It will also help to promote native species conservation, and position university landscapes as Other Effective area-based Conservation Measures (OECM) in urban cities.

2 Methodology

2.1 Study area

The study was conducted in Jawaharlal Nehru University (JNU) campus (Figure 1). The JNU campus is located within South Central Ridge of Delhi and lies in between 28°31′37″–28°33′12″ N and 77°09′08″–77°10′46″ E (Raju et al., 2014). The climate of the study area is semi-arid. June is generally the hottest month with a mean annual maximum temperature reaching 46 °C, while January is the coldest month. Monsoon period extends from July to August, and the average annual rainfall is 611 mm (Singh et al., 2017b). The JNU campus is spread over a total area of approximately 1019 acres (nearly 412 hectares), with a substantial portion of this undulated terrain remaining under forest and green cover (Kumar et al., 2025). The university is located on the Aravalli range (Srivastava et al., 2012). The Aravalli Range, one of the oldest fold mountain systems in the world, extends from Gujarat through Rajasthan and Haryana into Delhi, where it forms the Delhi Ridge and serves as an important ecological barrier against desertification (Sharma et al., 2013). The campus is characterized by extensive quartzite outcrops and buried pediments (Mukherjee, 2008). The soils of the campus have slightly acidic pH but have relatively high soil organic carbon and total nitrogen content in the forest patches (Sharma and Kulshrestha, 2018; Joshi and Garkoti, 2025).

To capture a representative socio-ecological gradient across the campus landscape, six distinct study sites were selected for parallel biophysical and social assessments: Yamuna Hostel (S1), Parthasarathy Rocks (PSR) (S2), New Transit House (S3), Chandrabhaga Lake (S4), JNU Caves (S5), and Brahmaputra-Mahanadi Hostel (S6). These locations were chosen to span a diverse range of land-use types, canopy structures, and levels of anthropogenic disturbance.

2.2 Perception survey

An online questionnaire survey using google forms was used to understand the perception of diverse stakeholders (N = 68) from the university on the benefits they derive from the green space using Likert scale ranging from 1-5 (1: not relevant to 5: highly relevant) (Carrus et al., 2015; Dallimer et al., 2012). Apart from basic questions on name and occupation, the questionnaire had 10 questions on the non-material benefits (ecosystem services) they derive from the green spaces of the campus. Three categories were listed, namely, supporting, regulating and cultural. One question was on willingness to pay, asking respondents how much they are willing to contribute each year to support conservation of the campus’ green spaces. The questionnaire is attached in Supplementary Appendix B.

2.3 Fractional vegetation index

Multi-temporal Landsat satellite imagery for the years 1992, 2002, 2012, and 2025 was used to analyze the spatio-temporal change of vegetation covers over 33 years (Supplementary Appendix Table A1). To minimize seasonal variability and cloud-related inconsistencies, average imagery from the winter months (October to December) was used for each study year. All images were atmospherically corrected and clipped to the JNU campus boundary prior to analysis. Fractional Vegetation Index (FVI), also known as Fractional Vegetation Cover, was employed in this study. FVI is a widely accepted remote sensing-derived biophysical parameter that estimates the proportion of vegetation within a pixel, with values ranging from 0 (no vegetation/bare surface) to 1 (complete vegetation cover) (Song et al., 2017). The index was computed using pixel dichotomy model based on Normalized Difference Vegetation Index (NDVI) values. NDVI was calculated using the standard Equation 1:

(1) NDVI=(NIR–Red)/(NIR+Red)

where,

NIR and Red represent the spectral reflectance values recorded in the near-infrared and red bands of the respective Landsat sensors, respectively. NDVI values range from -1.0 to +1.0, where positive values (>0.2) represent green vegetation cover, with higher values indicating increasing canopy density.

Subsequently FVI, was calculated using Equation 2:

(2) FVI=(NDVI–NDVIsoil)/(NDVIveg–NDVIsoil)

where,

• NDVI_soil (Soil Endmember) represents the NDVI of a purely bare soil pixel. For this study, NDVI_soil was set at 0.05 (Gao et al., 2020).
• NDVI_veg (Vegetation Endmember): This value represents the NDVI of a pixel with 100% vegetation cover. A value of 0.70 was utilized as the saturation threshold (Gao et al., 2020).

The analysis was performed in QGIS Desktop (3.40.11).

2.4 Woody plant diversity data collection

The study was conducted from November 2022 to March 2023. To ensure a representative sampling approach, systematic line-transect sampling with quadrats was implemented (Roleček et al., 2007; Kumar et al., 2025). Transect lines were strategically positioned at regular intervals of up to 100 m (Figure 2). To assess the vegetation systematically, quadrats measuring 10 m x 10 m (100 m²) were placed along transect lines, considering the homogeneity of the vegetation. The number of quadrats varied among the forested sites within the campus due to differences in site area, vegetation heterogeneity, and accessibility. In total, 76 quadrats were surveyed across the campus. The 76 sampling quadrats were distributed across the six designated sites to capture the unique ecological and structural traits of each zone:

• Yamuna Hostel (S1–6 quadrats): A high-use student residential and athletic zone characterized by active pedestrian traffic, managed roadside plantation.
• Parthasarathy Rocks (PSR) (S2–16 quadrats): An iconic, highly elevated quartzite outcrop of the Aravalli ridge. It represents a protected, low-disturbance semi-natural wilderness pocket dominated by tropical dry thorn forest flora.
• New Transit House (S3–6 quadrats): A transitional, moderately busy residential zone with mixed planted ornamental gardens.
• Chandrabhaga Lake (S4–16 quadrats): A low-lying wetland habitat surrounded by dense understories, representing a key focal point for both local wildlife and public recreation.
• JNU Caves (S5–16 quadrats): A relatively isolated, low-use interior woodland pocket characterized by high canopy and minimal human footprint.
• Brahmaputra-Mahanadi Hostel (S6–16 quadrats): A populated student and faculty residential complex experiencing intensive daily human activity, surrounded by dense, semi-managed green areas.

2.4.1 Plants identification

All the important parameters like species, genus, life form, etc. were recorded in field notes. Pictures of plants were also taken and longitude, latitude were recorded using handheld GPS navigator (Garmin GPS 73). The identification of plant species was supported using the citizen science platforms Pl@ntNet and iNaturalist that use extensive visual datasets for identifying the plants (Espitalier et al., 2025). The identifications were subsequently cross-verified using Flora of Delhi (Maheshwari, 1966), and Trees of Delhi: A Field Guide (Krishen, 2006) to ensure taxonomic accuracy. Furthermore, the accepted scientific names and taxonomic author citations of all documented species were verified and updated in accordance with Plants of the World Online (POWO, 2025) database. Species were classified as native, introduced, or invasive based on Reddy (2008). The floristic data of the JNU campus were compiled from field surveys and incorporated into an in-house database (https://sites.google.com/view/jnucampusflora/).

2.4.2 Plant diversity assessment

Importance Value Index (IVI), including Relative Frequency, Relative Density, and Relative Abundance, was calculated following Misra (1968). Species diversity was assessed using the Dominance, Simpson Index (Simpson, 1949), Shannon Diversity Index (also known as Shannon-Wiener index, and sometimes wrongly quoted as Shannon-Weaver index (Shannon, 1948; Spellerberg & Fedor, 2003)), Pielou’s Evenness Index (Pielou, 1966), and Menhinick Richness Index (Menhinick, 1964).

2.5 Statistical analysis

Species diversity indices (Dominance index, Simpson’s index, Shannon index H’, Pielou’s Evenness index, and Menhinick Species Richness) were calculated individually for each of the 76 sampling quadrats across the six study sites. Prior to executing comparative analyses, the distribution of each diversity index was tested for normality using the Shapiro–Wilk test, and homogeneity of variances was evaluated using Levene’s test. For indices that satisfied both standard parametric assumptions, a standard one-way Analysis of Variance (ANOVA) was performed to evaluate spatial differences across the campus sites, followed by Tukey’s Honest Significant Difference (HSD) post-hoc test for pairwise comparisons. For indices that satisfied normality but violated the assumption of equal variance (homogeneity), a robust Welch’s ANOVA was applied, followed by Games–Howell post-hoc pairwise comparisons. For indices that violated the normality assumption, spatial differences were evaluated using the non-parametric Kruskal–Wallis test, followed by Dunn’s test with Bonferroni correction for pairwise comparisons.

To evaluate the relationship between campus vegetation structure and stakeholder preferences, a Pearson correlation analysis was performed between the site-level means of the ecological diversity indices and the visiting preference (Public Perception, PP) scores. All statistical analyses and visual representations were performed in R statistical environment (v4.3.0) using the car, ggplot2, and corrplot packages. Statistical significance was evaluated at α = 0.05.

3 Results

3.1 User perceptions, ecosystem service preferences and willingness to pay

A total of 68 responses were recorded for the survey. The majority of respondents were students, accounting for 94.1% of the sample, and faculty and staff constituted the rest. In terms of residence, 64.7% of respondents reported living inside the JNU campus, whereas 35.3% resided outside the campus. The survey indicates that respondents perceive JNU’s green spaces as highly beneficial for wellbeing. Physical wellbeing recorded a high mean score of 4.51 ± 0.77, followed by emotional and mental wellbeing (4.49 ± 0.86). Both variables showed a median value of 5 while social interaction was rated comparatively lower (3.93 ± 1.05). The overall importance of campus greenery was rated high (4.67 ± 0.66), indicating strong agreement on the role of green spaces in supporting daily wellbeing. Wildlife observations highlighted the ecological richness of the campus with birds reported by 91% of respondents, followed by Nilgai (Boselaphus tragocamelus) which is Asia’s largest antelope (89.6%) (Leslie, 2008) and squirrels (88.1%). Other species such as peacocks, monkeys, and porcupines were also commonly observed. These patterns indicate frequent interaction with diverse fauna and suggest that biodiversity is an integral part of the campus experience (Figure 3). Respondents reported a wide range of activities in these green spaces with walking being universal, reported by all respondents. Nature observation was also very high (95.5%) followed by photography (68.7%) and exercise (52.2%). Activities such as birdwatching, meditation, and cycling were less frequent but still present. These results reflect both recreational and contemplative engagement with the landscape.

Respondents also positively perceived the ecosystem services provided by the campus greenery. Air quality improvement was the most valued service (95.5%), followed by biodiversity conservation (88.1%). Climate regulation (67.2%), noise reduction (61.2%), and landscape aesthetics (59.7%) were also important services perceived. This pattern highlights the dominance of regulating and cultural ecosystem services (Figure 4).

A similar trend was observed in willingness to pay per year, with the highest proportion of respondents willing to contribute INR 100 (≈1.2 USD) (24.6%), followed by INR 500 (≈6 USD) (22.3%). About 17.9% of respondents were willing to contribute INR 1000 (≈12 USD) and interestingly, an equal proportion reported no willingness to pay. Only 10.4% indicated willingness to contribute more than INR 1000 (>12 USD).

In the survey form, respondents were also asked that which sites they preferred revisiting the most in the university. Parthasarathy Rocks (PSR) area was the most liked site (S2), which is a rocky outcrop of the Aravalli hills and is a high elevated point of the campus. Other areas preferred for revisiting were the Yamuna Hostel–Stadium zone (S1), the Brahmaputra–Mahanadi hostel area (S6), the JNU caves (S5), the Chandrabhaga Lake (S4) and New Transit house (S3 in decreasing order of preference.).

3.2 Spatio-temporal analysis of vegetation density using FVI

The spatio-temporal analysis of FVI revealed an overall increase in vegetation cover across the JNU campus between 1992 and 2025 (Figure 5). The maximum FVI increased from 0.33 in 1992 to 0.47 in 2002, 0.51 in 2012, and 0.60 in 2025. The spatial distribution also indicates an expansion of areas with relatively higher FVI, particularly around the JNU Caves (S5) and Brahmaputra–Mahanadi hostel complex (S6). These results indicate an increase in the estimated fractional vegetation cover of the campus over the study period.

3.3 Species diversity in the identified sites

A total of 57 woody plant species were recorded across the six study sites, comprising 36 tree species (17 families, 32 genera) and 21 shrub species (16 families, 20 genera) (Figure 6). At site S1, 8 tree species (5 families, 8 genera) and 12 shrub species (10 families, 11 genera) were documented, with trees and shrubs contributing 56.7% and 43.2% of the total individual plant abundance, respectively. Site S2 exhibited the highest tree richness, with 28 tree species (16 families, 25 genera) and 15 shrub species (12 families, 13 genera). However, shrubs dominated the individual plant abundance (77%) here compared to trees (22.9%). At S3, 11 tree species (6 families, 11 genera) and 13 shrub species (10 families, 11 genera) were recorded. Here, shrubs accounted for 75.2% of the total individual abundance, while trees contributed 24.8%. Similarly, S4 supported 14 tree species (7 families, 14 genera) and 12 shrub species (9 families, 10 genera). In terms of individual abundance, shrubs (67%) here exceeded trees (32.9%). At S5, 11 tree species (7 families, 11 genera) and 13 shrub species (11 families, 12 genera) were observed, where shrubs constituted 74.68% and trees 25.31% of the total individual abundance. Site S6 recorded 15 tree species (10 families, 15 genera) and 16 shrub species (13 families, 15 genera), with individual abundance proportions of 37% for trees and 62.9% for shrubs. In terms of conservation status, the recorded flora comprised 48 Least Concern (LC), 7 Not Evaluated (NE), 1 Near Threatened (NT), and 1 Data Deficient (DD) species. Dominant species in each site are listed in Supplementary Appendix Tables A2, A3.

3.4 Pattern of species distribution

Species composition varied across the study sites, with differences in the relative contributions of native, introduced and invasive species. Site S2 recorded the highest total species richness, driven by a higher number of native species. In contrast, other sites such as S3 and S5 showed comparatively lower richness with a balanced distribution between native and invasive components. Among trees, Leucaena leucocephala (Subabool) emerged as the dominant species at sites S1, S2, and S3, with Importance Value Index (IVI) values of 81.62, 27.76, and 54.13, respectively. While categorized as a shrub, the prominent native species Grewia tenax (Gangeti) was also highly abundant at these sites, recording IVI values of 42.97, 21.24, and 45.90, respectively (Supplementary Appendix Tables A4, A5). At site S4, Pongamia peguensis (formerly, Millettia peguensis) (Moulmein Rosewood) was the dominant tree species with an IVI of 42.06. In contrast, Neltuma juliflora (formerly Prosopis juliflora; Vilayati kikar) dominated at sites S5 and S6, with IVI values of 73.87 and 48.29, respectively.

Additionally, Vachellia leucophloea (Ronjh) was consistently present across all study sites, exhibiting notable dominance with IVI values ranging from 13.76 to 34.21. Among the shrubs, Lantana camara (Lantana) was the most dominant species across all six sites, with IVI values of 72.45, 44.83, 37.60, 49.71, 44.46, and 45.13 for S1 through S6, respectively. Another frequently dominant shrub species was Justicia adhatoda (Adusa), recorded in sites S2 to S6 with IVI values of 34.69, 25.81, 32.40, 33.89, and 23.15, respectively (Supplementary Appendix Table A5). Other commonly occurring shrub species across the study area included Carissa spinarum, Capparis sepiaria, Abutilon indicum, and Ziziphus mauritiana, contributing to the overall floristic composition but with comparatively lower dominance.

3.5 Species diversity patterns across sites

Site S2 showed the highest overall diversity across the studied sites (Table 1). It recorded the highest Simpson’s Diversity Index (0.88), Shannon Index (2.41), Pielou’s Evenness Index (0.40), and Menhinick Richness Index (1.07). Among the remaining sites, Simpson’s Diversity Index followed the order S6 (0.84), S3 (0.83), S4 (0.80), S5 (0.79), and S1 (0.71). Shannon Index also showed a similar trend, with values ranging from 2.03 at S6 to 1.61 at S1. Evenness values varied slightly among the sites. S3 recorded the highest evenness after S2 (0.38), followed by S1 (0.33), S6 (0.32), S5 (0.30), and S4 (0.29). Richness also remained highest at S2 (1.07). It was followed by S3 (0.87), S1 (0.88), S6 (0.65), S5 (0.60), and S4 (0.52). S1, therefore showed comparatively lower diversity values across most indices.

Table 1. Diversity indices of the six study sites.

S1, S2, S3, S4, S5, S6 are the six study sites.

3.6 Open-access data integration and knowledge dissemination

To address the historical data gaps identified during the systematic review and to promote capacity building within the urban ecology community, the floristic and spatial data generated in this study were integrated into a centralized, freely accessible digital repository. This database, available at https://sites.google.com/view/jnucampusflora/, serves as an inventory for woody plant species of JNU campus. By consolidating photographs and taxonomic classifications, this platform facilitates establishing the campus as a hub of floral diversity and conservation. Furthermore, the repository also helps increase public awareness regarding Aravalli ridge’s biodiversity, shifting the research from passive documentation to a tool for environmental literacy and institutional stewardship.

3.7 Statistical comparison of diversity indices

To compare woody plant diversity across the six study sites, tests were conducted to evaluate the assumptions of normality (Shapiro-Wilk) and homogeneity of variances (Levene’s test) (Supplementary Appendix Table A6).

Diversity indices showed significant variation among the six study sites (Supplementary Appendix Table A6). For trees, Dominance and Simpson’s Diversity differed significantly among sites (Kruskal–Wallis, χ² = 27.7, p< 0.001), with significant pairwise differences between S1–S6, S4–S6, and S5–S6 based on Dunn’s test. Shannon diversity also differed significantly (χ² = 28.2, p< 0.001), with significant differences between S1–S2, S1–S6, S4–S6, and S5–S6. Evenness differed significantly among sites (Welch’s ANOVA, F(5,21.2) = 9.27, p< 0.001), with significant differences between S4–S6 and S5–S6. Menhinick Richness also showed significant variation (one-way ANOVA, F(5,70) = 6.90, p< 0.001), with S2 differing significantly from S1, S3, S4, S5, and S6.

For shrubs, Dominance and Simpson’s Diversity differed significantly among sites (Welch’s ANOVA, F(5,22.3) = 7.57, p< 0.001), with significant pairwise differences between S2–S5, S2–S6, and S4–S6. Shannon diversity also differed significantly (Kruskal–Wallis, χ² = 15.6, p = 0.008), with a significant difference only between S4–S6. Evenness showed significant site-level variation (χ² = 31.5, p< 0.001), with significant differences between S2–S6 and S4–S6. In contrast, Menhinick Richness did not differ significantly among sites (Welch’s ANOVA, F(5,22.3) = 1.91, p = 0.133), and no significant pairwise differences were detected.

3.8 Correlation between diversity and human perception

Pearson correlation analysis revealed relationships among the tree community diversity indices (Figure 7). Shannon diversity and Simpson’s Diversity exhibited a strong positive correlation (r = 0.978), while Pielou’s Evenness was also strongly positively correlated with Shannon diversity (r = 0.849). In contrast, Dominance showed a perfect negative correlation with Simpson’s Diversity (r = −1.00) and a strong negative correlation with Shannon diversity (r = −0.978).

When correlated with public preferences (PP), the ecological characteristics showed moderate-to-strong associations. Public preference exhibited a strong positive correlation with Menhinick Species Richness (r = 0.822), indicating a tendency for higher preference toward sites with greater woody plant richness. PP also showed moderate positive correlations with Shannon diversity (r = 0.615), Simpson’s Diversity (r = 0.499), and Evenness (r = 0.486), while Dominance showed a negative correlation (r = −0.499). The results indicate a biodiversity–perception association, whereby sites characterized by greater tree richness and diversity tend to receive higher visiting-preference scores, whereas sites with greater dominance by fewer taxa tended to receive lower preference.

4 Discussion

4.1 Perceptions and uses of green spaces

High ratings by respondents for physical and mental wellbeing are consistent with global studies showing that exposure to green spaces reduces stress and improves mental and physical health and social interactions (Hartig et al., 2014; Speake, 2013; Jennings and Bamkole, 2019). Similar patterns have been reported across university campuses where respondents associate green spaces with emotional stability and better quality of life (Gulwadi et al., 2019). These results indicate that in densely urbanized cities, campus greenery provides strong restorative health benefits (Kanelli et al., 2024).

University campuses have shown to be important ecosystems for flora and fauna diversity and conservation in urban landscapes (He et al., 2026; Liu et al., 2021; Singh et al., 2017b). In the Indian context, a floristic and faunal assessment conducted in and around Kuvempu University campus in Shivamogga, Karnataka documented 69 plant species belonging to 28 families and highlighted that the campus, being located near Bhadra Wildlife Sanctuary, supports rich floral and faunal diversity (Narayana et al., 2017).

Frequent observation of wildlife highlights the ecological significance of the campus. The presence of large fauna such as Nilgai (Boselaphus tragocamelus) and porcupines is relatively uncommon in urban campuses but is reported widely by the respondents in JNU. This reflects the unique ecological setting of the campus within the Aravalli landscape. Studies have shown that interaction with nature enhances wellbeing and strengthens sense of place (Soga and Gaston, 2020; Russell et al., 2013).

Activity patterns observed in this study are also consistent with global trends. Walking, nature observation, and photography were the dominant activities. Similar low-intensity and routine interactions with green spaces have been reported from campuses worldwide (Speake, 2013; Lau et al., 2014). Respondents showed positive perception for ecosystem services. Further, air quality improvement and biodiversity conservation were most positively perceived. This pattern is comparable to findings from mega cities such as Beijing, Washington, where regulating services are valued more (Elmqvist et al., 2013). The campus is also known for lower air pollution levels compared to the rest of the city (Mishra and Kulshrestha, 2017; Singh et al., 2021). The willingness to pay for biodiversity conservation further reflects a growing environmental consciousness among people, as also reported in studies from Europe and South America (Soga and Gaston, 2020).

The findings reinforce the role of university campuses as critical spaces for wellbeing, biodiversity conservation, and everyday human-nature interaction. These also highlight the need to prioritize ecological restoration, habitat connectivity, and protection of high-value green spaces in campus planning, as recommended in global urban ecology literature (Elmqvist et al., 2013).

4.2 Disturbance and spatial heterogeneity

Tree diversity varied across the sites with site S2 showing the highest diversity. This indicates that the site has better species distribution and higher richness. In contrast, sites such as S4 and S5 showed lower diversity due to dominance by specific species like Pongamia peguensis (formerly, Millettia peguensis) and Neltuma juliflora (formerly, Prosopis juliflora). Similar patterns of invasive dominance have been reported in Delhi ridge ecosystems (Naudiyal et al., 2017). Shrub diversity showed a slightly different pattern. S3 recorded higher diversity values, but overall variation among sites was limited. Statistical analysis indicated significant differences among sites for Dominance, Simpson’s Diversity, Shannon diversity, and Evenness, whereas Menhinick Richness did not differ significantly. The widespread dominance of Lantana camara likely contributes to this reduced variability by homogenizing shrub communities (Chauhan et al., 2022).

Anthropogenic disturbance can influence vegetation diversity and structure by altering habitat conditions and regeneration. Sites located away from major roads and areas with relatively lower direct human disturbance are expected to support greater richness and evenness. Previous studies have shown that disturbance can alter vegetation structure, reduce biodiversity, and affect regeneration (Chandrashekhar et al., 2003; Ram et al., 2004), while road clearings can create edge conditions that influence habitat suitability for interior species (Laurance et al., 2009). However, this expected pattern was not consistently observed across the study sites, as some road-associated locations also exhibited moderate to high diversity. This suggests that variation in urban vegetation diversity may reflect multiple interacting factors, including habitat quality, vegetation complexity, and site management, rather than road proximity or disturbance alone (Lepczyk et al., 2017; Beninde et al., 2015).

4.3 Diversity, dominance and invasive species influence

A total of 57 woody species were recorded, including 36 tree species and 21 shrub species (Table 2). Higher species richness has been reported from similar ecosystems in Delhi. For example, Asola-Bhatti Wildlife Sanctuary recorded 102 species, including multiple life forms (Sharma and Chaudhry, 2018). The vegetation structure observed was not uniform. A few species dominated across the sites, indicating uneven distribution. Leucaena leucocephala was dominant in several sites. Similar observations have been reported from the Northern Delhi ridge area, where it occurs as a highly abundant species (Kumar et al., 2022). The dominance of L. leucocephala can be explained by its ecological traits as it produces a large number of seeds, and exhibits allelopathic effects that inhibit the growth of other species (Chou and Kuo, 1986). Neltuma juliflora (formerly, Prosopis juliflora) was another dominant species. It has strong competitive ability and can suppress native vegetation and is recognized as an invasive in Delhi ridge ecosystems (Kushwaha et al., 2014).

Table 2. Trees and Shrub species identified in the study.

NE, Not Evaluated; LC, Least Concern; NT, Near Threatened; VU, Vulnerable; EN, Endangered; CR, Critically Endangered; DD, Data Deficient.

The shrubs layer showed even stronger dominance patterns. Lantana camara was dominant across all sites. L. camara forms dense thickets that suppresses native shrubs and inhibits the regeneration of native vegetation (Sharma et al., 2005). Similar dominance has been reported in JNU and other forest patches of Delhi such as Jahanpanah, Mehrauli, and Sanjay Van (Chauhan et al., 2022). Other shrub species such as Justicia adhatoda, Carissa spinarum, Capparis sepiaria, and Ziziphus mauritiana were present but less dominant. These species have also been reported from semi-arid forests of Delhi (Meena et al., 2019), indicating consistency with regional vegetation patterns.

The temporal expansion of vegetation cover, as quantified by the FVI, provides a physical dimension to these diversity patterns. The transition from an FVI peak of 0.33 in 1992 to 0.60 in 2025 (Figure 5) signifies an increase in the green cover over three decades. The temporal FVI analysis indicates a progressive increase in estimated vegetation cover across the JNU campus, with the maximum FVI increasing from 0.33 in 1992 to 0.60 in 2025. This temporal increase provides evidence of increasing vegetation cover over the years. However, higher vegetation cover does not necessarily indicate greater floristic diversity or ecological quality. The field survey revealed that several invasive species, including Neltuma juliflora, Leucaena leucocephala, and Lantana camara, were locally dominant. Thus, the increase in vegetation cover should be interpreted together with species composition, dominance, and diversity rather than as a direct indicator of ecological improvement.

Importantly, while invasive taxa accounted for 5–13% of the total species richness, they have much more significant ecological impact. Despite of this, invasive plants exhibited high Importance Value Index (IVI) scores, reflecting their abundance in the vegetation community. Their high dominance may contribute to biotic homogenization and potentially limit opportunities for native regeneration, highlighting the need for their management within the campus landscape.

4.4 Green spaces biodiversity–perception nexus

The observed association between vegetation diversity and visiting preference suggests that biodiversity may contribute to how campus green spaces are perceived and valued by users. Areas with greater woody species variety may provide greater visual heterogeneity and a more varied vegetation experience, potentially increasing their attractiveness for recreation and informal visits. This interpretation is consistent with previous studies reporting positive relationships between vegetation diversity, aesthetic appreciation, psychological restoration, and user satisfaction in urban green spaces (Simonič, 2003; Harris et al., 2018; Hartig et al., 2014; Soga and Gaston, 2020).

The relationship may also have an ecological dimension, as greater vegetation complexity can contribute to microclimatic regulation and other benefits in urban environments (Livesley et al., 2016). However, public preference is unlikely to be determined by biodiversity alone and may also reflect accessibility, site aesthetics, familiarity, recreational opportunities, and management. Moreover, the present analysis is based on six site-level observations and therefore indicates an association rather than a causal relationship. Further studies incorporating quantitative measures of site-specific perception and additional environmental attributes would help clarify the mechanisms underlying this biodiversity–perception relationship.

4.5 Policy recommendation

This interdisciplinary synthesis yields a vital directive for urban conservation. If left unmanaged, passive ecological succession on the campus will continue to increase canopy, but will simultaneously allow invasive species to erode understory evenness and species richness. Because higher species richness was associated with greater visiting preference in the present analysis, maintaining diverse native vegetation may help support both the ecological integrity and recreational value of campus green spaces. Active ecological management, specifically the targeted eradication of Lantana camara and the enrichment planting of diverse native species, is therefore ecologically justified to preserve both the ecological integrity and the cultural ecosystem services of this UGS site.

In an increasingly digital and urbanized era, such university green spaces also play an important role in enhancing the physical health, psychological well-being, and nature connectedness of younger generations. By linking biophysical monitoring with social valuation, this research provides a scalable model for ecologically sensitive campus planning in rapidly urbanizing environments.

5 Conclusion

This study provides a multidimensional assessment of the Jawaharlal Nehru University (JNU) campus as a socio-ecological system within the urban fabric of New Delhi. A key finding is the increased vegetation density/cover of the campus over the 33-year period, as reflected by the Fractional Vegetation Index (FVI). The campus landscape shifted from relatively sparse vegetation cover in 1992 (maximum FVI: 0.33) to substantially denser vegetation cover in 2025 (maximum FVI: 0.60), representing a marked increase in vegetation density. The floristic inventory comprised 57 woody species, indicating moderate woody plant diversity, with Leucaena leucocephala, Neltuma juliflora, and Lantana camara among the dominant species across the study sites.

Importantly, the findings point to a potential ‘greening paradox’ in urban ecosystems, wherein greater visible vegetation cover does not necessarily correspond to greater compositional diversity. Significant differences among sites, particularly in tree diversity and evenness, reflect spatial variation in vegetation composition beyond differences in overall greenness. The positive association between public perception and tree richness (r = 0.82) further indicates a relationship between floristic diversity and the perceived cultural value of a urban green space. Thus, ecological quality within urban forests is better represented through a combination of vegetation density, species composition, diversity, and evenness rather than vegetation cover alone. The presence of dominant and invasive species, particularly within the understory, remains an important consideration in maintaining balanced and diverse plant communities.

Finally, the establishment of the freely accessible JNU Campus Flora database (https://sites.google.com/view/jnucampusflora/) provides a permanent digital record of the campus flora beyond the research period. The database represents a resource for institutional documentation, capacity building, future biodiversity assessments, and environmental stewardship within the National Capital Territory of Delhi.

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.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fevo.2026.1886362/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: J. D. Lewis, Fordham University, United States

Reviewed by: Mohammad Zashim Uddin, University of Dhaka, Bangladesh

Abdel Aleem Bello, University of Aleppo, Syria