An Open-Source Digital Street Tree Inventory for Neighborhood-Scale Assessment in Rome

Abstract

Systematic, spatially explicit tree inventories are increasingly implemented in cities worldwide, as they are crucial for evidence-based green infrastructure planning. Currently, different approaches are adopted, which differ in methodological framework and parameter standardization, limiting comparative assessments and coordinated monitoring. This study presents a replicable protocol for a field-based digital street tree census, applied in a densely built central area and in a low-density suburban area of Rome. Field surveys documented a set of 15 parameters, including species identity, dendrometric and tree pit parameters, acquired using open-source QGIS/QField tools. Subsequent analysis evaluated floristic diversity, population structure, and climate suitability at the neighborhood scale, enabling the identification of context-specific vulnerabilities. The testing of the methodology shown in this work involved 13,017 georeferenced tree pits, pointing out substantial pit restoration needs and insufficient soil conditions in the most densely urbanized area, whereas the suburban area shows optimal conditions with extensive road verge green spaces. Joint interpretation of the considered parameters reveals that high floristic diversity alone does not guarantee climate resilience: high-diversity neighborhoods can exhibit substantial non-climate-resilient species and limited alignment with local species recommendations, demonstrating that comprehensive evaluation of street tree populations requires integrated analysis. The operationalized protocol establishes a replicable, municipally scalable methodological framework, providing policymakers with fine-scale, actionable insights enabling differentiated urban forestry strategies addressing both infrastructure deficits and long-term species climate suitability.

1. Introduction

The challenges imposed by the combined pressures of settlement dynamics and soil sealing have exacerbated the effects of climate crisis in urban environments, underscoring the centrality of street trees and urban green infrastructure in promoting environmental health, social well-being, and climate resilience in cities [1,2,3]. Urban tree canopies represent an essential component of green infrastructure, capable of delivering multiple ecosystem services [4,5,6,7], with positive impacts on public health and social equity, establishing urban green management and protection as a fundamental priority for local administrations and European Union environmental policies [8,9].

Bringing nature back into cities is a central objective in several European thematic strategies and policies, requiring the definition of appropriate monitoring tools to support the management and planning of urban green infrastructure and tree cover. The EU Biodiversity Strategy 2030 [10] encourages European cities with populations exceeding 20,000 inhabitants to develop Urban Nature Plans (UNP), which recommend equipping municipalities with decision-support tools, such as urban tree inventories [11]. Furthermore, the recent Nature Restoration Regulation (NRR), has recognized tree inventories as an effective tool for monitoring the state of urban tree cover and defining appropriate restoration measures within National Restoration Plans (NRP) [12].

Even at an international level, the System of Environmental Economic Accounting–Ecosystem Accounting (SEEA-EA) [13,14] statistical framework adopted by the United Nations gives great attention to the issue of urban greenery and urban street trees, due to their ability to provide measurable and essential ecosystem services in highly anthropized environments (such as local climate regulation, carbon sequestration, air filtration or amenity services).

At the national level, the mapping and monitoring of urban green spaces and urban tree cover is required by Law 10/2013 [15], which mandates the assessment and publication of the status of the urban forest heritage of municipalities with more than 15,000 inhabitants. Additionally, with the adoption of Ministerial Decree No. 63 of 10 March 2020, setting out the Minimum Environmental Criteria (CAM) [16], urban tree inventories have taken on an even more important role, becoming a mandatory technical requirement in contracts for the management of public green spaces.

Different methodological approaches exist for urban tree assessment, each presenting distinct trade-offs in accuracy, cost, and operational feasibility. Field-based censuses are more accurate, robust and complete, involving direct examination of individual trees, nevertheless, they become expensive and labor-intensive for large-scale urban inventories [17]. Remote sensing technologies combined with deep learning models offer cost-reduction potential for large-scale urban tree assessments but generally achieve lower species-level accuracy than field-based surveys, particularly in complex urban environments [18]. Recent evidence on urban forest monitoring demonstrates that comprehensive assessment tools requiring tree-level data (e.g., trunk diameter and species) depend on field-based baseline inventories, as remote sensing cannot accurately measure these data; however, remote sensing technologies can subsequently enable cost-effective updating of structural attributes without repeating full field surveys, supporting scalable hybrid inventory systems [19].

In Europe, several cities have already adopted integrated digital systems for urban green infrastructure management, based on common operational elements, i.e., GIS integration, dynamic data updating, open-data publication, and structured citizen participation. Paris has managed a patrimony of approximately 500,000 trees since 2014 through the “ARBRES” GIS platform, featuring open-data publication and initiatives for improving knowledge of privately owned trees [20,21,22]. Barcelona has operated the GAVI system since 2008 for street tree inventory and management, receiving an average of 5000 annual citizen-generated communications regarding tree maintenance [23,24].

In Italy, the 2023 data on urban greenery [25] of the National Institute of Statistics (ISTAT) reveal significant disparities in tree inventory implementation: while 52.7% of provincial capitals have adopted georeferenced inventories, only 24.5% cover their entire municipal territories [26]. This fragmented information on urban forest resources is driven by the lack of a shared methodological framework and by operational constraints, including budget limitations, city size, and the availability of trained staff. These factors hinder the generation of comparable data, complicating the definition of best practices and effective governance guidelines. The Italian Institute for Environmental Protection and Research (ISPRA) also emphasizes the need to move from isolated municipal censuses to coordinated monitoring based on transparent geographical platforms that are also accessible to citizens who can be involved in data collection and validation [27]. Milan currently represents one of the most advanced implementations of such systems, providing open-access georeferenced data for over 240,000 public trees through its municipal geoportal [28]. Also, Florence renovated its SiVeP system in 2018 and currently manages dynamic data for over 79,000 tree-related sites and associated urban green elements through open-access platforms and interactive cartographic interfaces [29].

In Rome, urban street tree management falls under the centralized authority of the Department for Environmental Protection (DTA) and is carried out in compliance with the Capitoline Regulation on Public and Private Greenery. This framework mandates the creation of a comprehensive Green Cadastre and defines specific strategies for maintenance. While it establishes species recommendations and planting standards aligned with climate adaptation objectives, it simultaneously enforces strict heritage conservation measures, prioritizing native species and mandatory replacement [30,31]. Furthermore, Rome’s Strategic Guidelines for Urban Green [32] acknowledge urban tree inventory as an indispensable foundation for planning and managing the city’s extensive green patrimony. Despite this administrative framework, Rome lacks a comprehensive publicly available tree inventory that would enable low-cost, continuous updating. In response to this imperative, the Municipality of Rome commissioned the Council for Agricultural Research and Economics (CREA) to conduct a science-based strategic study for assessing and enhancing Rome’s street tree population, establishing initial criteria for evidence-based species selection that support both adaptation to future climate conditions and the preservation of Rome’s historical urban identity [33,34,35]. This partnership aims to support the implementation of the Green Cadastre mandate by establishing a scientifically validated census protocol. The strategic study produced a first street tree field survey, recording basic information for each tree surveyed on a sample of approximately 200 streets distributed across all municipia. The research presented in this paper builds on the sampling scheme proposed by the strategic study undertaken by CREA [33] and proposes an expanded census protocol that enhances its recommended parameters while incorporating additional variables for street tree population analysis, related to dendrometric and tree pit parameters, in terms of condition, size and available soil.

In detail, the general objective of this research is to define a set of indicators in compliance with the requirements of the main regulatory instruments in force at the international, European and Italian levels, which support the species-level mapping and monitoring of urban street trees, allowing the identification of context-specific vulnerabilities and orienting evidence-based intervention priorities at the neighborhood scale. Furthermore, ensuring the replicability, scalability and comparability of the methodology also with respect to different territorial contexts and favoring the use of free and open-source software.

To this end, as a specific objective, a first detailed census of two pilot areas within the municipality of Rome was conducted, helping to fill the knowledge gap resulting from the lack of a comprehensive municipal street tree inventory. This initial census activity covered study areas with significant differences in settlement and territorial characteristics, with a view to testing the effectiveness of the proposed set of indicators across different urban contexts.

The research findings are part of the activities of the CREA working group that is developing the official census of street trees and the future Street Tree Master Plan [33,34], representing a first attempt to develop a strategy to refine and thicken the neighborhood-scale monitoring network in critical areas. This is with a view to support interventions as consistent as possible with the climate suitability and characteristics (in terms of sustainability/structure/quality) of the existing street tree population. Furthermore, the data collected in this research contributed to the tree census activities carried out by the Casal Palocco Consortium.

2. Materials and Methods

2.1. Overview

This research presents a comprehensive street tree census protocol within the municipal road network across two study areas in Rome. The activity was based on open-source GIS methodology for the development of an indicator-based assessment integrating dendrometric and tree pit parameters, dimensional development classes, floristic diversity indices, local regulatory framework compliance, and climate suitability classification. The census was also designed to document all discrete tree pits regardless of occupancy status and generate species-level data suitable for identifying context-specific vulnerabilities and supporting evidence-based intervention priorities at the neighborhood-scale.

2.2. Study Area

The municipality of Rome extends across a territory of 1286.8 km², making it the largest municipality in the European Union. The administrative structure consists of 15 municipia, each with autonomous management, financial, and administrative authority. From a historical urban geography perspective, the city is divided into 327 neighborhoods, 22 districts (the historic centre) and 104 functional zones. The management and maintenance of green spaces smaller than 20,000 m² is under the responsibility of the municipia, while Rome’s street tree populations remain under the centralized authority of the Department for Environmental Protection [31].

The two areas chosen as pilot cases for testing the set of indicators were selected for their environmental and infrastructural characteristics, which make them complementary and representative of the urbanization gradients of the Roman territory.

The first study area (Figure 1) extends over 500 hectares, mainly within municipium II, in an area between the two consular roads of Via Nomentana and Via Tiburtina. This area represents a strategic hub for services and infrastructure, hosting facilities such as the main campus of La Sapienza University, the Umberto I Polyclinic, and several ministries. Administratively composed of four neighborhoods, two functional zones, and one district, the area functions as a highly active urban node; within this context, San Lorenzo and Italia neighborhoods represent the focal points where social life and the local economy are closely linked to the presence of the university. The urban fabric is a clear example of a compact city, characterized by a high density of built-up areas and resident population.

Figure 1. Urban context (A), distribution of artificial surfaces (B), and tree cover density (C) within municipium II, highlighting the surveyed census area (yellow boundary).

Analysing the National Land Consumption Map [2], nearly 90% of the area is covered by artificial surfaces, while, according to the 2023 version of the Copernicus High Resolution Layer Tree Cover Density (TCD) data [36], tree cover is limited to just 27.3 hectares, relegated to a few historic parks or small internal courtyards. The scarcity of green infrastructure appears even more critical considering the high population density (over 9000 inhabitants/km²) and the vulnerable population (under 14 and over 65), which is about 37% of the total residents (18,149 inhabitants) [37].

The second study area extends for approximately 270 hectares and is located within municipium X, including a large part of Casal Palocco neighborhood. The neighborhood was created in 1960 along Via Cristoforo Colombo, the main road connecting the center of Rome to the sea, and is surrounded by important nature reserves, including Castelporziano, Castel Fusano, and the Riserva del Litorale Romano (Figure 2). Unlike the urban fabric of municipium II, the Casal Palocco settlement follows an extensive urban logic, inspired by the Anglo-Saxon garden city model (isolated buildings, such as villas and terraced houses, in a continuous system of greenery). In this area, the building and residential density are much lower than those of municipium II (45% artificial surfaces and just under 3500 inhabitants/km²). The neighborhood’s green spaces are managed by the Casal Palocco Consortium (and not by the municipality), which conducted an initial tree inventory in 2013. The activities conducted as part of this research contributed to the inventory’s update to 2025.

Figure 2. Urban context (A), distribution of artificial surfaces (B), and tree cover density (C) within the surveyed census area (yellow boundary) in municipium X.

2.3. Field Survey Design and Data Collection

This study proposes an open-source census methodology focused on collecting dendrometric and tree pit parameters, through these sequential steps:

• Database schema design using QGIS 3.40.9, defining dendrometric and tree pit parameters to be collected during field surveys;
• Field survey, using QFieldSync 4.18.0 tool to upload the complete dataset with attribute tables to QField Cloud, enabling download to mobile devices for field data collection;
• Post-field analysis of collected data using QGIS, with supplementary elaborations for parameters not directly recorded in the field.

2.3.1. Database Schema Design and Parameter Selection

In the database schema design, the parameter selection was based on five criteria:

• Consistency with established urban tree inventory methodologies [38,39,40,41] and ecosystem service assessment models [42,43,44];
• Adherence to global statistical standards, structuring the inventory to comply with the UN SEEA-EA framework for urban areas [14];
• Alignment with EU restoration targets, structuring the inventory to support the longitudinal monitoring required by the NRR [12];
• Support for UNP, designing the schema to serve as the baseline assessment for the EU Biodiversity Strategy 2030 [10,11];
• Inclusion of tree pit and soil parameters, rarely assessed in standardized field protocols [45,46,47].

These parameters were organized into four thematic groups to facilitate systematic data collection and analysis: tree and pit status characteristics (Table 1) (Figure 3), taxonomy, dendrometric parameters and census identification attributes (Table 2).

Table 1. Recorded parameters for tree and pit status characteristics.

Group	Parameter	Parameter Description	Class	Class Description
Tree and pit status characteristics	Tree status	Assesses the presence/absence of tree and its vitality condition	Stump	Tree pit contains a stump
			Dead tree	Tree pit contains a dead tree
			Live tree	Tree pit contains a living tree
			Vacant	Tree pit contains no tree
	Pit condition	Assesses the condition of the individual tree pit	Optimal	Pit shows no damage and is clearly visible
			Requires restoration	Pit is damaged, absent, not visible, or no longer suitable for tree growth
	Pit size	Assesses, through visual estimation, the surface area of the tree pit	Small	Pit surface area is less than a 55 cm × 55 cm square
			Medium	Pit surface area ranges between a 55 cm × 55 cm and 105 cm × 105 cm square
			Large	Pit surface area exceeds a 105 cm × 105 cm square
			Planting strip	Pit extends linearly in length
			Road verge green space	Extensive green areas not classified as discrete tree pits (e.g., planted roundabouts, planted median strips)
	Available soil	Assesses, through visual estimation, the amount of soil accessible to the tree within the tree pit	Not visible	Visible soil surface area is absent
			Insufficient	Visible soil surface area is insufficient for current tree growth
			Sufficient	Visible soil surface area is sufficient for current tree growth
			Abundant	Visible soil surface area is abundant and permits both current and future tree growth
	Spontaneous growth	Assesses whether tree growth is spontaneous or planted	Yes	Species is not planted in a discrete pit and displays spontaneous characteristics (roadside colonization, ability to establish in harsh environments)
			No (default)	Species is planted within a discrete tree pit

Figure 3. Field examples of tree pit and related parameters: (A) “Requires restoration” pit condition and “Medium” pit size with “Sufficient” soil; (B) “Optimal” pit condition and “Medium” pit size with “Not visible” soil; (C) “Optimal” pit condition and “Small” pit size with “Insufficient” soil; (D) damaged pit with “Sufficient” soil.

Table 2. Recorded parameters for taxonomy, dendrometric parameters, and census identification attributes.

Group	Parameter	Parameter Description
Taxonomy	Common name	Common name by which the species is recognized
	Genus	Genus of the tree
	Species	Species of the tree
	Variety	Variety of the tree (if identifiable)
Dendrometric parameters	Circumference at 1.3 m	Precise measurement of circumference at 1.3 m above ground, made using a 150 cm flexible tape
	Diameter at 1.3 m (DBH)	Diameter calculated from the circumference measured at 1.3 m above ground
	Total height	Estimated total height using handheld laser distance meters
	Crown insertion height	Estimated height to the base of the crown using handheld laser distance meters
	Crown diameter	Visual estimation of crown diameter
	Crown health	Visual estimation of the percentage of missing crown compared to total possible crown, using regular 10% intervals up to 90%
Census identification attributes	Street	Location where the tree is surveyed (street, square, piazza, etc.)
	Survey date	Date the tree is surveyed (DD/MM/YYYY)
	Unique identification code	Unique identification code assigned to the individual tree or tree pit
	Notes	Additional descriptions or observations about the tree or tree pit

2.3.2. Field Survey and Species Identification

Species identification was conducted through visual assessment based on knowledge from the literature and environmental expertise, with the Pl@ntNet 3.23.4 mobile application employed as a supplementary tool where necessary [48]. All tree species present within tree pits were inventoried, including palms and Yucca species, as these are widely utilized as street trees despite not meeting conventional arboricultural definitions [49]. For these non-conventional taxa, all parameters were recorded except trunk circumference because standard DBH protocols are designed for cylindrical woody stems and, in these taxa, diameter is not a reliable proxy for size or crown development [50].

In the field survey, geospatial referencing of each tree pit was accomplished using the GPS receiver integrated within the mobile device, supplemented by the visual verification of georeferenced Google Satellite imagery overlaid within the project via the QGIS plugin QuickMapServices. Every identifiable tree pit was surveyed as a discrete unit. However, where extensive road verges or planting strips were present, independent pit units could not be distinguished; in these cases, only the specific locations occupied by existing trees were mapped. The unplanted soil areas were not recorded as “Vacant” pits, leading to an underestimation of the potential planting capacity. Additionally, the census scope was limited to public street trees, excluding private green spaces. Following field data collection conducted between November 2024 and June 2025, all records were aggregated into a comprehensive, georeferenced street tree inventory. While seasonality may affect “Crown health” estimates in deciduous species, vitality assessment remained robust by relying on structural indicators identifiable during dormancy (e.g., twig dieback, bark condition), minimizing phenological bias.

2.3.3. Post-Field Analysis and Indicator-Based Assessment

In the post-field analysis, an indicator-based assessment was conducted at two spatial scales. At the municipal level, descriptive statistics were computed to characterize overall dendrometric structure, tree pit infrastructural conditions, and floristic diversity metrics. Floristic diversity was evaluated using the Shannon–Wiener index (H′) and Simpson’s diversity index (D), both well-established metrics for characterizing urban tree species diversity [51,52]. The integrated application of both indices enables identification not only of neighborhoods exhibiting low overall diversity, but also of neighborhoods where diversity is dominated by few species.

At the neighborhood scale, analyses were carried out by integrating dimensional development classes, local regulatory framework compliance and climate suitability classification. This approach enabled the evaluation of street tree populations and supported the identification of context-specific vulnerabilities and restoration priorities.

Dimensional development classes were assigned based on circumference and height measurements, classifying trees into three distinct categories representing different life stages: juvenile (third strength), mature (second strength), and senescent (first strength) [33,53]. This classification reveals the age structure and regeneration potential of street tree populations.

To identify potential conflicts between conservation obligations and future resilience, species were classified using two frameworks:

• Regulatory compliance, based on the Capitoline Regulation [30], which prioritizes native species and the strict protection of historical taxa, while incorporating climate adaptation criteria for new plantings;
• Climate suitability, based on Rome’s strategic study [33], which applies the Esperon-Rodriguez et al. [54] methodology under SSP1-2.6 climate projections (2041–2080) and five bioclimatic variables (temperature and precipitation parameters). Accordingly, species were categorized into climate-resilient (taxa with high adaptive capacity to heat and drought stress) and high-risk (taxa vulnerable to future stressors or biotic threats, requiring monitoring or replacement).

3. Results

The implementation of the described methodology enabled a comprehensive inventory of street trees within sampled areas of Rome’s municipium II and municipium X. In total, 13,017 individual point features were surveyed and georeferenced along the municipal road network (Figure 4).

Figure 4. Spatial distribution of surveyed tree pits within a sampled section of municipium II, classified by the “Tree status” parameter.

3.1. Field Survey Results and Species Composition

Within the surveyed samples, municipium II contains 6463 live trees (99 species, 71 genera) and municipium X contains 5551 live trees (80 species, 56 genera). In both areas, the ten most abundant species account for more than two-thirds of all live trees (Table 3, Figure 5). Species occurring at frequencies below 1% constitute about 13–15% of total live trees in both study areas, represented by 65 species in municipium X and 79 species in municipium II.

Table 3. Live trees in the two study areas: species dominance, frequency and diversity indices.

Study Area	Live Trees [n°]	Dominant Species	Dominant Species Frequency [%]	Total Frequency of Top Ten Species [%]	Simpson [D]	Shannon–Wiener [H′]
Municipium II	6463	Platanus × hispanica	9.60	67.85	0.06	3.23
Municipium X	5551	Quercus ilex	28.34	79.68	0.15	2.62

Figure 5. Distribution of the ten most abundant species across study areas, expressed as a percentage of total live trees surveyed.

3.2. Dendrometric Characteristics and Structural Patterns

The mean dendrometric parameters reveal distinct structural differences between the study areas (Table 4). Mean DBH and height are lower in municipium II than in municipium X, and diameter-class distribution shows a higher proportion of small trees in municipium II and of medium to large trees in municipium X (Figure 6). Overall phytosanitary condition assessment indicates a pronounced prevalence of living trees in both study areas, with percentages exceeding 85%.

Table 4. Dendrometric parameters and tree and pit status in the two study areas.

Study Area	DBH [cm]		Height [m]		Pit Condition [%]		Tree Status [%]
	Mean	Max	Mean	Max	Requires Restoration	Optimal	Live Tree	Vacant	Dead Tree	Stump
Municipium II	24.87	142	7.56	28	45.94	54.06	86.47	7.07	6.13	0.32
Municipium X	40.10	172	11.12	28	2.51	97.49	99.46	0.09	0.36	0.09

Figure 6. Diameter at breast height and height classes of trees surveyed in the two study areas.

Among the critical infrastructure constraints identified, 42.44% of live trees in municipium II experience insufficient soil availability, including 4.03% with no visible soil surface. By contrast, the surveyed area in municipium X demonstrates markedly different conditions, with approximately 80% of live trees located in contexts with abundant soil availability, consistent with the predominance of road verge green spaces (77.40% of total surveyed tree pits). Quantitative assessment of pit size classes in relation to soil availability reveals pronounced disparities in growth potential. In municipium II, the medium-sized pit class, representing 57.79% of all surveyed trees, rarely exhibits optimal soil conditions (3.47%), instead concentrating in sufficient (29.00%) and insufficient (22.94%) soil availability categories. By contrast, municipium X demonstrates positive conditions, with 74.04% of trees located within road verge green spaces exhibiting optimal growth conditions with abundant soil availability (Table 5).

Table 5. Percentage of pit size classes in relation to soil availability in the two surveyed study areas.

Pit Size	Available Soil [%]	Study Area
		Municipium II	Municipium X
Road verge green space	Abundant	4.71	74.04
	Sufficient	0.61	3.34
	Insufficient	0.02	0.02
	Total	5.34	77.40
Planting strip	Abundant	2.75	4.32
	Sufficient	1.00	4.52
	Insufficient	0.08	0.83
	Total	3.83	9.67
Large	Abundant	8.97	1.18
	Sufficient	5.85	1.60
	Insufficient	7.95	0.29
	Not visible	0.62	0.00
	Total	23.39	3.07
Medium	Abundant	3.47	0.02
	Sufficient	29.00	3.78
	Insufficient	22.94	5.70
	Not visible	2.38	0.02
	Total	57.79	9.52
Small	Sufficient	1.21	0.00
	Insufficient	7.42	0.32
	Not visible	1.03	0.02
	Total	9.66	0.34

Based on Rome’s latest administrative classification into neighborhoods, functional zones, and districts, parameters were analyzed to obtain a local-scale overview of surveyed street tree populations. The complete set of maps illustrating the spatial distribution of the analyzed parameters for each neighborhood, functional zone and district is provided in Appendix A.

The results shown in Table 6 reveal neighborhood-scale differences in street tree population characteristics. Mean DBH ranges from a minimum of 19.76 cm in the Italia neighborhood to a maximum of 44.98 cm in the Scalo San Lorenzo functional zone. Excluding the La Sapienza University and Scalo San Lorenzo functional zones, the highest value of live trees per inhabitant occurs in the Casal Palocco neighborhood, while the lowest values occur in the Castro Pretorio district and San Lorenzo neighborhood. Also, in terms of live trees per street area, the Casal Palocco neighborhood shows the maximum density, followed by the Italia neighborhood, declining progressively to the Castro Pretorio district with 10.76 live trees per hectare of street area.

Table 6. Dendrometric, density, and diversity metrics of street tree populations categorized by neighborhoods, functional zones and districts. * Surveyed area represents a subset of the total extent. ** Based on Rome’s administrative classification.

Neighborhood/Functional Zone/District	Live Trees	Vacant, Stump and Dead Trees	Mean DBH [cm]	Mean Height [m]	Surface Area ** [ha]	Live Trees/Inhabitant	Live Trees/Area [ha]	Street Area [ha]	Live Trees/Street Area [ha]	Simpson [D]	Shannon–Wiener [H′]
Castro Pretorio *	303	16	27.06	8.08	64.73	0.11	4.68	28.17	10.76	0.23	1.82
Italia	3193	570	19.76	6.30	153.64	0.18	20.78	44.93	71.07	0.06	3.18
La Sapienza University	562	77	34.00	10.00	84.83	1.25	6.62	16.10	34.91	0.14	2.25
Lanciani *	658	90	23.89	7.00	34.23	0.16	19.22	10.83	60.76	0.11	2.68
Nomentano	547	108	33.03	9.61	79.09	0.25	6.92	18.27	29.94	0.20	2.21
S. Lorenzo	732	92	24.56	8.35	67.55	0.11	10.84	19.75	37.06	0.11	2.62
Scalo S. Lorenzo	70	18	44.98	10.37	16.39	0.41	4.27	4.06	17.24	0.26	1.58
Casal Palocco *	5382	7	40.32	11.22	273.03	1.23	19.71	53.02	101.51	0.15	2.61

Based on Simpson’s dominance index, maximum species diversity is achieved in the Italia neighborhood, while minimum diversity, excluding the Scalo San Lorenzo functional zone, occurs in the Castro Pretorio district. Identical extremes are observed for the Shannon–Wiener richness index, with values ranging between 2.21 and 2.68 for the remaining analyzed neighborhoods. The table provided in Appendix B reports the percentage frequencies of the ten most abundant species for each analyzed area.

The dimensional development classification, based on the circumference and height measurements of individual surveyed trees, highlights differences in population structure among neighborhoods. Figure 7 reveals that in 6 of the 7 analyzed areas (in this and the next analysis San Lorenzo neighborhood and Scalo San Lorenzo functional zone were combined), the third development class, representing juvenile trees destined to progressively replace mature canopy trees, constitutes the minority component.

Figure 7. Percentage of dimensional development classes in the analyzed areas. * Surveyed area represents a subset of the total extent.

Figure 8 compares tree species regulatory compliance (Figure 8A) with climate suitability profiles (Figure 8B). Results reveal strong regulatory compliance, with values exceeding 50% in 4 out of the 7 areas. Regarding climate adaptation, while a significant portion of the population does not meet the climate resilient criteria (exceeding 43% in 4 areas), the fraction of species classified as high risk remains relatively low (generally below 13%). The combined assessment highlights Italia as the most critical area, exhibiting the highest incidence of high-risk species alongside low regulatory compliance. Castro Pretorio presents a divergent pattern, recording the highest regulatory compliance but the lowest climate resilience. Finally, Casal Palocco demonstrates the most favorable profile, showing high regulatory adherence and climate suitability.

Figure 8. Proportion of street tree species classified, as recommended by Rome’s Capitoline Regulation (A), as climate-resilient and high risk (B) for each analyzed area. * Surveyed area represents a subset of the total extent.

The integrated indicator-based assessment of all surveyed areas is summarized in Figure 9, revealing distinct neighborhood profiles defined by varying combinations of dendrometric parameters, population structure, floristic diversity, regulatory compliance and climate suitability. For comparative visualization in the radar chart, all selected indicators were rescaled to a 0–1 range to enable direct comparison across neighborhoods. Percentage-based metrics were already expressed as proportions. Simpson’s diversity index (range 0–0.30), Shannon’s index (0–3.50), mean height (5–12 m), and mean DBH (15–45 cm) were normalized using min–max values derived from the observed dataset in the surveyed areas.

Figure 9. Radar chart representing selected parameters for each surveyed area. * Surveyed area represents a subset of the total extent. ** Mean dendrometric parameters are calculated on the total area of San Lorenzo and Scalo San Lorenzo.

4. Discussion

The results of this study allow for two types of considerations regarding the methodological framework and the specific results obtained from its application in the study areas.

4.1. General Considerations on the Proposed Monitoring Scheme

General considerations concern, first of all, the choice of the parameter set, which follows the following main criteria: the versatility of the methodology in terms of replicability, modularity and comparability; the cost-effectiveness of its implementation; and the ability to support the monitoring activities introduced by the main regulatory instruments in force at the national, international and European levels. Regarding replicability, in defining the set of indicators, priority was given to parameters that are well-known in the literature and therefore characterised by consolidated and recognised calculation methods, which make their use more versatile and shared [38,39,40,41]. In terms of comparability, from a qualitative perspective, the results obtained allowed for a clear and unambiguous characterization of the study areas, both in terms of structural characteristics and critical issues, providing initial guidance for management, restoration, and regeneration interventions. A quantitative comparison is planned following the expansion of the number of surveyed areas, in order to have a more robust sample. From a modular perspective, the set of indicators was designed to maximize the number of parameters detected while keeping the cost in terms of collection time to a minimum, and above all, to maximize the construction of a set of indicators capable of adequately supporting the monitoring of the obligations imposed by current regulations. The methodology also lends itself to the integration of additional parameters, such as visual tree assessment, or additional tree pit information, such as pH and soil compaction. The introduction of additional parameters, which increase the time and therefore the cost of surveys, can be considered during the preliminary study of future sample areas, assessing the main critical issues identified in the analysis of parameters already collected in other areas with similar urban and settlement characteristics (e.g., in densely urbanized areas, pit-related parameters will be more important than in areas with a low degree of artificialization, making it useful to integrate additional parameters related to monitoring these aspects).

It is worth noting that all survey phases were conducted using open-source tools and software, as well as mobile applications with user-friendly interfaces. This contributes to the methodology’s cost-effectiveness and replicability, while also ensuring its openness to the involvement of additional stakeholders in monitoring activities. This may be beneficial for the higher-level involvement of expert stakeholders, such as universities, research centers, and freelancers, as well as for citizen engagement, especially when reporting critical issues and emergencies, such as trees that are unsafe, dead, or causing problems for the community.

To ensure that the parameters would meet the monitoring needs required to comply with the obligations introduced by the main existing regulatory instruments, the selection was based on the following criteria:

• Consistency of dendrometric and taxonomic parameters with the input requirements of i-Tree Eco and AIRTREE for ecosystem service quantification and adherence to global statistical standards of the UN SEEA-EA framework for urban areas. Specifically, the SEEA-EA schema adopts the recommended individual asset approach, which prioritizes the granular tracking of specific assets (e.g., street trees) over coarse landscape-level assessments. This approach is critical for revealing intra-urban disparities hidden by municipal averages, enabling the disaggregated reporting required by SEEA to support targeted policies at the sub-district or neighborhood-scale;
• Alignment with EU restoration targets required by the NRR; specifically, the inclusion of canopy dimensions and geolocation enables the tracking of urban green space and urban tree canopy cover trends to verify compliance with the 2030 no net loss goals;
• Support for UNP, designing the schema to serve as the baseline assessment for the EU Biodiversity Strategy 2030; by capturing granular data on species diversity and ecosystem condition, the inventory facilitates the calculation of the core indicators needed to spatially prioritize interventions and address environmental inequalities;
• The inclusion of tree pit and soil parameters aims to operationalize CAM requirements for assessing rooting space availability [16] and aligns with ISPRA recommendations [27] to move towards a qualitative assessment of station conditions, recognizing that restricted pit volumes, often overlooked in canopy-focused inventories, are critical constraints for tree health, stability and ecosystem service provision.

4.2. Analysis of the Total Surveyed Areas

The overall results confirm Rome’s high street tree biodiversity (130 species), aligning with previous studies [49,55]. However, this richness is unevenly distributed, reflecting distinct street tree patterns. Municipium II area exhibits high diversity driven by the selection of small to medium sized ornamental species, typical of residential and commercial areas where spatial constraints necessitate the selection of smaller species. This is consistent with the historical shift in urban tree species selection, from large-statured trees to smaller-sized species [56]. Conversely, municipium X area shows greater structural maturity but relies heavily on three dominant species, exposing the stock to concentration-related phytosanitary risks [57].

This divergence extends to tree pit infrastructural conditions. In the central area, the prevalence of small, highly sealed tree pits limits root expansion and stability [58], whereas the extensive road verge green spaces in the suburban context support larger, healthier trees with lower mortality rates [59]. These findings underscore that soil volume and pit design are critical determinants of tree structure, indicating that infrastructural interventions must be prioritized in compact urban fabrics, where high tree density coexists with severe tree pit infrastructural constraints.

4.3. Neighborhood-Scale Analysis

The transition from aggregate municipal data to a neighborhood-scale indicator-based assessment proved essential for revealing critical vulnerabilities that would remain hidden at the municipal level [60]. While the georeferenced database supports even finer-scale analysis (individual streets or trees), this resolution aims to generate actionable decision-support data for strategic planning. At the neighborhood scale, the indicator-based assessment distinguishes different street tree population profiles, defined by trade-offs between dendrometric parameters, diversity, population structure, regulatory compliance, and climate suitability.

In the Italia neighborhood, the combination of high planting density and severely constrained tree pits is associated with a prevalence of small-statured trees, resulting in high diversity and limited dominance. However, the indicator-based assessment reveals a critical fragility, as a substantial share of the population relies on species with low regulatory compliance and higher climate vulnerability. Furthermore, the low percentage of juvenile component suggests limited renewal, potentially leading to a senescent population without adequate replacement. Overall, this indicates that optimal diversity indices alone may not be sufficient proxies for functional resilience when species composition and renewal dynamics are not aligned with long-term climate suitability. Conversely, Castro Pretorio and Nomentano show stronger regulatory compliance, reflecting a commitment to maintaining the city’s historic vegetative identity through monospecific rows. This approach leads to a negative synergy where high dominance and low diversity pair with limited climate resilience. The divergence between the two profiles lies in their population structure:

• Castro Pretorio, despite an active turnover, appears to replicate existing vulnerabilities because replacement plantings remain concentrated on the same species, prioritizing landscape continuity at the expense of adaptive diversification and increasing exposure to phytosanitary risks [57];
• Nomentano, in contrast, faces a regeneration gap characterized by a dominant mature population and minimal turnover, increasing the likelihood of abrupt canopy loss in the near future.

In contrast, the Casal Palocco suburban context allows for large-statured trees species that are simultaneously regulation-compliant and climate-resilient. However, this structure relies on a simplified floristic composition, where a few dominant species account for the majority of the population. While currently stable, this low diversity, combined with an aging population structure and low juvenile component, exposes the neighborhood to species-specific risks (e.g., pests) and simultaneous senescence, requiring diversification strategies in future renewal plans.

The indicator-based assessment demonstrates that conformity to conventional diversity thresholds (e.g., Santamour’s empirical 10/20/30 rule [61]), while necessary, is not sufficient to guarantee long-term sustainability. The analysis reveals vulnerabilities linked to age structure, climate unsuitability and tree pit constraints in compact urban fabrics. Consequently, future planting strategies must integrate diversity targets with a multi-dimensional assessment of tree pit parameters, demographic structure, and climate suitability. Another aspect emerging from the analysis is that high floristic diversity does not always ensure resilience to future climate scenarios [54,62].

5. Conclusions

This study presents a methodology for the systematic inventory of urban street trees in a georeferenced database through standardized field surveys.

Overall, the research has led to the development of a set of easily replicable parameters, optimized in terms of time and cost of collection and capable of providing useful information to support the mapping and monitoring activities required by the main current regulations.

The results obtained are encouraging, as the analysis of the set of indicators enabled the characterization of the study areas in terms of composition and main critical issues. For example, in the two study areas, the indicator-based neighborhood assessment revealed vulnerabilities, such as diversity–resilience mismatches or regeneration deficits, that dendrometric parameters and diversity indices alone would mask, underscoring the necessity for integrated frameworks in street tree assessment.

The results demonstrate that high floristic diversity is not always an indicator of climate resilience. The optimal profile is not maximum absolute diversity but rather maximum diversity of climate-resilient species. This constraint necessitates a shift in management perspective: future diversification strategies must systematically integrate climate suitability assessment alongside traditional species selection criteria. In the context of accelerating climate challenges, systematic and continuously updated inventories of urban street trees assume crucial strategic value for building resilient cities.

The introduction of parameters related to the tree pit characteristics is a strong point of the sampling scheme, and the analysis of the results shows wide-spread structural limitations, as soil volume and spatial constraints often represent the primary bottleneck for canopy expansion. This limitation operates not only by physiologically restricting growth but also by necessitating the selection of smaller-statured species to comply with regulatory distances, avoid infrastructure conflicts, or maintain aesthetic continuity. Consequently, physical site conditions effectively cap potential ecosystem services regardless of biological diversification, underscoring that resilient urban forestry requires integrating species selection with rigorous site suitability assessment.

The current inventory, when extended to municipal scale, could also support more accurate estimation of urban tree canopy cover by providing field-based information that may be combined with Copernicus products for NRR assessments, although specific approaches for integrating datasets with different spatial resolutions still need to be investigated.

The census itself remains time- and resource-intensive, particularly in densely built urban contexts. Progressive extension to municipal scale could be costly but necessary to provide a robust baseline. Once this baseline is established, continuous monitoring could be maintained more cost-effectively through integration with remote sensing data and citizen science, representing a potential future development that could update the baseline providing essential tree status changes. The modular database structure and accessibility of open-source tools make such continuous updating feasible.

The inventory excludes private green space, a relevant component particularly in low-density residential areas. Participatory census protocols for private green space represent an emerging research priority, with initial approaches underway in European cities such as Paris [20]; this extension would increase the representativeness of urban tree canopy coverage assessments.

At present, although the results obtained are encouraging in terms of capturing the characteristics of the surveyed areas, statistical analyses are limited by the small sample size. The current monitored areas nevertheless represent an important step forward in the mapping of Rome’s urban street trees, both contributing to the census conducted by the working group developing the official street tree census of Rome and supporting the tree census activities carried out by the Casal Palocco Consortium. Therefore, continuing the monitoring activity and the resulting expansion of the sample of points and surveyed areas will allow us to define a set of indicators for statistically characterizing and comparing the areas, as well as to develop a multicriteria analysis methodology based on the integration of the considered parameters.

Appendix A.1. Municipium X Area

Figure A1. Casal Palocco neighborhood. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A1. Casal Palocco neighborhood. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Appendix A.2. Municipium II Area

Figure A2. Castro Pretorio district. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A2. Castro Pretorio district. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A3. Italia neighborhood. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A3. Italia neighborhood. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A4. Lanciani neighborhood. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A4. Lanciani neighborhood. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A5. Nomentano neighborhood. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A5. Nomentano neighborhood. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A6. San Lorenzo neighborhood and Scalo San Lorenzo functional zone. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A6. San Lorenzo neighborhood and Scalo San Lorenzo functional zone. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A7. La Sapienza University functional zone. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.

Figure A7. La Sapienza University functional zone. (A) Spatial distribution of tree species. Point size is proportional to DBH. (B) Dimensional development classes. Point size is proportional to DBH.