Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2025 Nov 3;35(7):e70128. doi: 10.1002/eap.70128

Bug roads: Modeling the green space connectivity and pollinator habitat in a large city using open GIS data and tools

Matthew J Lundquist 1,✉, Pamela C Lovejoy 2, Brianna G Fay 1, Juliet E Hernandez 1, Martha Madrid 1
PMCID: PMC12582645  PMID: 41183563

Abstract

The conservation of native bees and other pollinators is an important consideration for the future of urban sustainability. Parks, urban gardens, cemeteries, and other green spaces can provide habitat space for both native and non‐native pollinators in cities. These publicly managed green spaces are not evenly distributed across otherwise inhospitable urban landscapes. Buildings and other human‐made structures could act as barriers to the movement of pollinators, especially in highly built‐up cities. Little is known about how bees navigate cities, and finding suitable habitat in urban ecosystems may be particularly difficult for native solitary bees, which have small foraging ranges. In this study, we utilized open GIS data as well as open‐source software (Quantum GIS and Python) to model the shortest flight paths between parks and other public green spaces in New York City, New York, USA. We also used open light detection and ranging (LiDAR) data to assess plausible pollinator habitat in New York City parks. We found that the majority of straight‐line (Euclidean) paths between parks intersected at least one building and that shortest paths around buildings were generally 20% longer than their Euclidean equivalent. We found that most managed properties alone, or within connected clusters, did not have sufficient plausible pollinator habitat to support pollinators with medium foraging distances, which include most solitary native bees. Our findings suggest limited connectivity and potential barriers between managed properties in New York City. Increasing pollinator habitat within smaller managed properties and building green roofs on shorter buildings and establishing stepping stone habitats like tree pits and vacant lots could increase overall green space connectivity. This technique for assessing connectivity between green spaces utilizes open data and tools that can be used by conservationists, planners, and policymakers to explore questions related to supporting pollinators or other species of interest in urban landscapes.

Keywords: GIS, native bees, open data, pathfinding, pollinators, urban landscapes

INTRODUCTION

Native bees are important pollinators and are essential for the maintenance of local native floral biodiversity (Tanda, 2023), but their populations are declining in habitats globally (Potts et al., 2010). North American native bees include solitary and social bees that nest either in soil or cavities in rocks or wood (Pindar & Raine, 2023). Urban environments have been shown to have both positive and negative effects on native bee biodiversity. The introduction of non‐native plants, both ornamental or unintentional, in parks and gardens can provide pollen resources as well as habitat space (Lowenstein et al., 2019; Weber et al., 2023). Conversely, the fragmentation within the urban landscape and overall loss of natural habitat can have a net negative effect on bee richness and abundance (Remmers & Frantzeskaki, 2024).

Green spaces and green infrastructure within cities may have beneficial direct or indirect effects on bees. For example, green roofs, both with and without intentional floral plantings have been shown to support native and non‐native bees (Braaker et al., 2017; Tonietto et al., 2011). Urban parks and gardens are also important for sustaining local bee populations since they are typically the largest green spaces within the urban matrix and may support a large diversity of flowering plants and habitat types (Banaszak‐Cibicka et al., 2018), and even small parks can support pollinator biodiversity (Egerer et al., 2024). Parks, gardens, green roofs, and other green spaces are not evenly distributed within the largely inhospitable urban matrix, making conservation of native bees, many of which are solitary and have small foraging ranges, challenging in cities (Ayers & Rehan, 2021). This is made more complex by the heterogeneity of habitat type and quality within parks and other green spaces (Harrison & Winfree, 2015) as well as potential barriers to movement (e.g., tall buildings). Therefore, understanding the potential pathways in which native bees may be able to move about in cities is extremely important for the success of conservation efforts.

The movement of bees within natural and agricultural environments has been well studied in the eusocial European honey bee, Apis mellifera (Apidae), which uses complex navigational cues to locate floral resources and communicate the location and quality of the resource to other members of their hive (Menzel et al., 2005; Ratnieks & Shackleton, 2015). The movement of solitary bees is less understood, and current knowledge is generally limited to the maximum foraging distances of particular species (as reviewed by Zurbuchen, Bachofen, et al., 2010, Zurbuchen, Cheesman, et al., 2010, Zurbuchen, Landert, et al., 2010). Nevertheless, it has been postulated that limited foraging distance in solitary bees is a major contributor to native bee population declines in cities (Ayers & Rehan, 2021). Studies have found a correlation between body size, foraging distance, and eusociality (Grüter & Hayes, 2022; Kendall et al., 2022), with larger bees like A. mellifera being able to travel far from their nest sites (Greenleaf et al., 2007) while many small, solitary bees are only able to forage a maximum of 1000 m or less, with many having ranges much less than that (Gathmann & Tscharntke, 2002; Zurbuchen, Cheesman, et al., 2010; Zurbuchen, Landert, et al., 2010). A limitation to these maximum foraging distances is that they were measured in natural or agricultural settings using Euclidean (straight‐line) distance (Zurbuchen, Landert, et al., 2010). Urban landscapes are heterogeneous and are dominated by human‐made structures, which may preclude straight‐line flight (Zurbuchen, Bachofen, et al., 2010). Therefore, it may be difficult for urban bees to find new sources of floral resources because of the increased complexity of navigating within cities, including the choice of flying around or over obstacles like buildings (Johansson et al., 2018).

There is evidence that bees will fly upward to visit urban green roofs (Passaseo et al., 2021; Tonietto et al., 2011); however, there is evidence that roof height (e.g., 24 m) limits visitation, even if the green roof has flowers in bloom (Underwood et al., 2025). While there is no comprehensive list of green roofs in New York City and most buildings do not have green roofs, the largest green roof in New York City is on top of the Jacob K. Javits Convention Center in Manhattan (Smalls‐Mantey & Montalto, 2021). It is approximately 25 m above ground level, is primarily planted with sedum, and houses A. mellifera hives. It is likely that those bees forage both on the roof and in nearby parks at ground level. In contrast, there is almost no information about whether and, if so, how high bees will fly over buildings without green roofs or other floral resources (e.g., window boxes, patio gardens). Indeed, buildings with no floral or nesting resources, regardless of height, could act as barriers to the movement of bees in urban environments. Instead, bees, particularly those with shorter foraging ranges, may need to navigate around buildings to search for floral resources or nesting habitat.

In this study, we modeled the connectivity between urban parks and community gardens, and other green spaces within the five boroughs of New York City, NY, USA, by determining the shortest path (“bug road”) between each green space using Euclidean distance (i.e., over buildings), as well as Dijkstra's pathfinding algorithm (i.e., around buildings). Connectivity between multiple green spaces across New York City was determined by mapping Girvan–Newman networks. We assessed the plausible habitat space in each green space and Girvan–Newman networks by analyzing light detection and ranging (LiDAR) vegetation point clouds. We postulated that most straight‐line paths would be blocked by tall buildings and that there would be low‐level connectivity for species with small to medium maximum foraging ranges (e.g., solitary bees) and that habitat space would be heterogeneous across managed green spaces in New York City.

METHODS

Study area

The study was conducted within the boundaries of New York City, NY, USA, which includes five boroughs: Bronx, Manhattan, Brooklyn, Queens, and Staten Island (Figure 1). Each borough has a mix of commercial and residential buildings as well as mixed‐use parks and other green spaces (Table 1). Publicly managed green spaces in the city include parks managed by New York City Parks (Parks), and others are comanaged by independent conservancies (e.g., The Central Park Conservancy) or other city entities (e.g., the New York City Department of Transportation). Community gardens in New York City are under the parks department's “GreenThumb” program but are each managed by their community gardeners. In addition to these city parks and community gardens, New York City has some large cemeteries, New York State parks, US National Parks Department sites, private land, and vacant open space that all contribute to overall urban green space and may provide some level of pollinator habitat (Sallay et al., 2023).

FIGURE 1.

FIGURE 1

Open in a new tab

Map of New York City including building outlines (gray) and managed properties (green) with each borough labeled. The map is projected to NAD83/UTM zone 18N.

TABLE 1.

Land cover analysis of New York City using LiDAR (Garner, 2021; New York State, 2021) and open GIS data (NYC Office of Technology and Innovation, 2024). Building heights are represented as mean roof heights from ground level ± SE.

Borough Land cover
High vegetation (km2) Medium vegetation (km2) Low vegetation (km2) Building (km2) Ground (km2) Building height (m) No. managed properties
Citywide 135.6 66.21 33.04 168.36 462.25 8.44 ± 0.01 4328
Bronx 22.5 9.99 4.59 22.54 64.82 8.91 ± 0.02 824
Brooklyn 25.07 13.47 6.82 50.24 97.27 8.63 ± 0.01 1069
Manhattan 10.61 4.23 1.69 19.18 27.64 23.06 ± 9.10 715
Queens 39.93 21.64 11.23 57.56 176.33 7.01 ± 0.01 1204
Staten Island 37.49 16.88 8.71 18.84 96.19 7.63 ± 0.01 516
Open in a new tab

Map preparation

Open‐access GIS shape files of New York City borders, building footprints, publicly managed properties, and open and vacant land were acquired from NYC Open Data (NYC Office of Technology and Innovation, 2024), NYS GIS Data (New York State, 2021), and United States National Parks (United States National Parks Service, 2020) and loaded into Quantum GIS 3.36.0 (QGIS.org, 2024). Publicly managed properties within these datasets include parks, community gardens, playgrounds, and recreation facilities managed either by New York City, the State of New York, or the United States Government. Cemeteries were also added from the NYC Open Data repository. All layers were reprojected to NAD83/UTM zone 18N, which uses metric units, before analysis.

A grid of 50 m × 50 m polygons was produced over the entire extent of New York City, including the nearby water bodies (approximately 1346 km2 total) to act as a network for shortest path analysis. Building footprints were then merged with the grid to act as network edges (Figure 2). Connected buildings were merged to ensure that only open paths were available for analysis.

FIGURE 2.

FIGURE 2

Open in a new tab

Example of shortest path analysis for “bug roads” using Dijkstra pathfinding algorithm including a generalized workflow.

The park and cemetery shapes were then overlaid on top of the network (Figure 2). A total of 1992 public lands and cemeteries, hereafter referred to as “managed properties,” were included in the analysis. Many of these sites were included as multiple (multipart) and/or noncontiguous patches (e.g., Broadway Malls) in the dataset and were split and each part was considered an independent managed property, resulting in a total of 4405 managed property patches for shortest path analysis.

Shortest path analysis

Euclidean (straight line) distance between managed properties was calculated using the shapely library in Python (version 3.12.4). To assess the potential effects of buildings on straight‐line flight, the number and height (in meters) of each building intersected by each straight‐line path were also collected using the shapely library.

The shortest (i.e., best possible) paths between managed properties were determined using Dijkstra's pathfinding algorithm provided by the networkx library in Python. Dijkstra's algorithm works iteratively along the network, preferring edges with smaller weights (i.e., shorter) to find the shortest overall path (Fan & Shi, 2010). The polygons of the 50 m × 50 m foraging network grid as well as the outlines of buildings acted as edges for this analysis (Figure 2). A 30‐m topographical buffer was also introduced to allow for modeling of more realistic movement across the urban landscape. We also calculated sinuosity (path length/straight line length) as a measure of path complexity.

The start and end points for both Euclidean distance and Dijkstra pathfinding analyses were determined by identifying the closest edges between each pair of managed properties, then mapping these points to the nearest foraging network nodes. This ensured proper path detection by the pathfinding algorithms. The closest points between park boundaries were identified by sampling points at intervals along each park's boundary and calculating pairwise distances to determine the minimum separation. Preliminary analysis indicated that there was a high degree of connectivity between parks when considering paths above 1000 m. Most solitary bees have maximum foraging ranges ≤1000 m (reviewed by Zurbuchen, Landert, et al., 2010). To reduce complexity and computational load, only managed properties with a Euclidean distance ≤1000 m between their closest edges were considered. Paths between managed properties that were geographically adjacent (i.e., Euclidean distance = 0 m) were also excluded from analysis.

Community cluster analysis

Community cluster analysis was performed using the Girvan–Newman algorithm in the networkx library in Python. The Girvan–Newman algorithm works by iteratively removing edges from the graph to reveal the underlying community structure (Despalatović et al., 2014). In this case, parks that are connected within a particular path distance to each other were found. Community clusters were determined from both Euclidean and Dijkstra‐derived paths of ≤100 m, ≤250 m, ≤500 m, and ≤1000 m based on the range maximum foraging distances of solitary bees previously reported (reviewed by Zurbuchen, Landert, et al., 2010).

Habitat area estimation

Landscape heterogeneity within green spaces can promote flower diversity as well as provide nesting resources for urban bees (Ayers & Rehan, 2021); however, there is no exhaustive study of flowering plant resources in New York City. Therefore, we instead utilized high‐resolution LiDAR data (12 pixels m−2) collected in 2021 (Garner, 2021; New York State, 2021) to estimate plausible habitat space (i.e., vegetation is more likely to have habitat space of some kind than open ground). The total area of high vegetation (>5 m), medium vegetation (1–5 m), and low vegetation (0.3–1 m), ground (lawns and, bare ground), and buildings for each managed property was determined by counting the number of pixels of each cover type within the shapes of each managed property in Python (Figure 3). There is likely a high level of overlap between layers within individual parks and some medium and low vegetation may be obfuscated by tall vegetation, so we considered them all together as “plausible pollinator habitat.”

FIGURE 3.

FIGURE 3

Open in a new tab

Example LiDAR data (New York State, 2021) for high vegetation (A), medium vegetation (B), and low vegetation (C). Pixels with the land cover types of interest are represented in gray. Yellow outlines represent parks' property borders.

It has recently been suggested that conserving 11.6%–16.7% of habitat space is needed to support wild bee biodiversity (Pindar & Raine, 2023). We investigated habitat suitability within individual parks and within clusters based on whether they had enough plausible habitat to account for ≥11.6% of circular foraging areas with radii of 100, 250, 500, and 1000 m. For example, an individual managed property or a cluster of parks would need to have approximately 21,900 m2 of combined plausible habitat space (about the size of four American football fields) to support bees with foraging ranges of 250 m or less.

We also conducted an analysis of street trees intersecting flight paths. Street trees may act as habitat (Lundquist et al., 2022) or stepping stones for flying insects (Płaskonka et al., 2024). Street tree locations were obtained from the NYC Open Data Street Tree Census (NYC Office of Technology and Innovation, 2024). We counted the number of street trees intersected by each Dijkstra‐derived path using the geopandas library in Python.

RESULTS

Shortest path analysis

Euclidean distance

A total of 74,276 straight‐line paths ≤1000 m were determined among the managed property patches (N = 4405) across the five New York City boroughs. Of those paths, 86% intersected at least one building and had an average distance of 550.24 ± 1.107 m. Of the paths that did intersect buildings, they intersected between 1 and 47 buildings, with a maximum building height of 429.27 m (average, 16.31 ± 0.02 m; Figure 4).

FIGURE 4.

FIGURE 4

Open in a new tab

Summary plots of building intersections by Euclidean‐derived paths in New York City as well as each borough separately. The solid black line represents the average building height for paths with n = i number of buildings within the path. Light blue bars represent minimum and maximum height of buildings within paths with n = i number of buildings. Heat maps represent the number of paths n = i buildings, with darker shades representing a higher number of paths. A key with the number of paths represented by the different shades are included to the right of the heat maps. There were a total of 10,539 paths that had no building intersections.

Dijkstra's algorithm

A total of 64,118 paths ≤1000 m were determined among the managed property patches using Dijkstra's pathfinding algorithm which had an average length of 551.79 ± 1.14 m. The average path segment length was 36.60 ± 0.03 m, and each path had an average of 4.85 ± 0.02 turns. Average sinuosity was 1.20 ± 0.04, meaning that Dijkstra‐derived paths were 20% longer than their comparable Euclidean path. Results of Dijkstra‐derived paths at 100‐m, 250‐m, 500‐m, and 1000‐m cutoffs can be found in Appendix S1: Figures S1–S4.

Community cluster analysis

In general, there was an increase in the average number of managed property sites per cluster between path distances of 100 m and 1000 m for both Euclidean and Dijkstra‐derived paths. The total number of clusters generally varied from 100 to 500 m and was the lowest at 1000 m for both Euclidean‐derived and Dijkstra‐derived paths (Table 2).

TABLE 2.

Community cluster analysis for managed properties connected by Euclidean or Dijkstra‐derived paths at 100‐, 250‐, 500‐, and 1000‐m foraging distance classes.

Path and borough 100 m 250 m 500 m 1000 m
No. sites ± SE Total clusters No. sites ± SE Total clusters No. sites ± SE Total clusters No. sites ± SE Total clusters
Euclidean
Citywide 3.7 ± 0.14 632 7.3 ± 0.58 487 20.3 ± 7.41 206 231.2 ± 189.61 19
Bronx 3.6 ± 0.31 118 8.2 ± 1.36 84 41.3 ± 28.40 19 403.0 ± 399.00 2
Manhattan 4.7 ± 0.52 97 10.3 ± 2.47 64 65.7 ± 44.59 11 364.0 ± 361.00 2
Brooklyn 3.4 ± 0.24 163 6.8 ± 1.23 131 23.0 ± 11.08 45 270.8 ± 264.75 4
Queens 3.6 ± 0.27 176 6.7 ± 0.84 132 12.7 ± 2.72 87 172.1 ± 139.49 7
Staten Island 3.4 ± 0.29 72 5.8 ± 0.78 66 10.8 ± 2.40 43 72.0 ± 41.09 7
Multi‐borough 2.5 ± 0.50 2 21.8 ± 14.21 4 323.8 ± 288.58 5 1219.7 ± 1205.67 3
Dijkstra
Citywide 9.2 ± 1.49 339 8.4 ± 2.59 241 17.8 ± 5.47 220 47.8 ± 30.87 90
Bronx 9.2 ± 3.89 65 16.8 ± 11.33 20 30.7 ± 21.59 24 88.6 ± 85.31 9
Manhattan 5.0 ± 1.02 47 19.0 ± 10.19 20 29.2 ± 17.49 24 103.9 ± 72.87 7
Brooklyn 8.8 ± 2.60 81 7.9 ± 2.43 64 13.3 ± 6.94 71 44.2 ± 40.09 24
Queens 11.6 ± 4.25 73 6.1 ± 2.39 53 13.9 ± 5.07 73 33.1 ± 16.54 35
Staten Island 8.4 ± 2.15 47 7.0 ± 3.10 21 16.4 ± 8.55 28 29.3 ± 21.78 17
Multi‐borough 59.5 ± 57.50 2 222.2 ± 123.33 4 393.0 ± 232.59 4 1426.5 ± 1311.50 2
Open in a new tab

Note: Values are the mean number of managed sites (initial N = 3576) within clusters ± SE, and the total number of clusters detected at each distance class. Clusters include at least two unique managed properties.

Plausible habitat space assessment

There was a wide range of property sizes (Figure 5). Small properties had a higher proportion of buildings, medium‐sized properties had the largest proportion of low and medium vegetation, and larger properties had the highest relative proportion of ground (i.e., lawn and bare ground). The proportion of high vegetation cover was relatively consistent across park sizes. (Figure 6).

FIGURE 5.

FIGURE 5

Open in a new tab

Box and whisker plots of managed property size across New York City (citywide) as well as in the five boroughs. Boxes represent the interquartile range (IQR), the line within the box represents the median, and the whiskers represent 1.5× the IQR. Points outside the whiskers represent outliers. Area (in square meters) of each managed property has been log10 transformed for readability because the distribution of managed property areas is highly right skewed.

FIGURE 6.

FIGURE 6

Open in a new tab

Stacked area plots of landscape cover determined from light detection and ranging (LiDAR) imaging for managed park and open space properties in New York City. The land cover type area (in square meters) of each managed property has been log10 transformed for readability because the distribution of land cover type area is highly right skewed.

If we consider all vegetative cover types (low, medium, and high) together as plausible pollinator habitat, 24.09% of managed properties have enough vegetative cover to potentially support pollinators with 100‐m foraging ranges in accordance with the ≥11.6% of habitat space recommendations by Pindar and Raine (2023). This decreases to 0.64% for those with 1000 m (Table 3). When considering neighborhood clusters, those based on either Euclidean distance or Dijkstra‐derived paths, the largest percentage of clusters support 100‐ and 1000‐m foraging ranges. The percentage of clusters is markedly lower for 250 m and 500 m (Table 3). On average, we found that Dijkstra‐derived paths intersected 33.91 ± 0.13 street trees.

TABLE 3.

Analysis of vegetative cover in the whole (i.e., not subdivided) managed properties (N = 1992) and neighborhood clusters derived from pathfinding analysis (initial sites = 4405).

100 m 250 m 500 m 1000 m
Sufficient support (%) Total clusters Sufficient support (%) Total clusters Sufficient support (%) Total clusters Sufficient support (%) Total clusters
Managed properties 24.09 6.26 2.25 0.64
Euclidean‐derived clusters 41.92 2631 26.95 1258 24.06 399 37.50 24
Dijkstra‐derived clusters 30.40 1156 13.02 676 11.49 174 50.00 4
Open in a new tab

Note: For managed properties, values are in percentage of the area of each property that has sufficient vegetation cover (high, medium, and low vegetation combined) to account for ≥11.6% of the habitat space required to support pollinator biodiversity in each maximum foraging distance class (Pindar & Raine, 2023). For clusters, the values are the percentage of clusters that have sufficient combined managed area to support ≥11.6% of the habitat space required for each maximum foraging distance class and the total number of clusters within that foraging distance class.

DISCUSSION

Our results suggest that there is a lack of connectivity between managed properties within the highly built‐up New York City urban matrix, particularly for solitary bees which have small to medium maximum foraging ranges. We also found that buildings may pose a significant barrier to pollinator movement by increasing the complexity of paths between managed properties. Our analysis of plausible habitat space in managed properties showed that most individual properties did not have enough vegetative cover to support even short foraging ranges. This was also true when considering parks as connected clusters. It is therefore likely that pollinators may find it difficult to find suitable foraging or habitat patches within the urban landscape. However, we have also identified avenues for increasing connectivity and promoting native pollinator conservation in New York City as well as in urban ecosystems in general.

Pollinator habitat connectivity within New York City

The landscape of New York City is highly fragmented by urban structures, with the majority of the Euclidean paths intersecting buildings. The average distance of the Dijkstra‐determined shortest paths between managed properties was 20% longer than their associated straight‐line distance, suggesting that moving around buildings added a biologically significant heightened level of navigational complexity. While the maximum foraging distance for a particular species may be between 500 and 1000 m (Zurbuchen, Landert, et al., 2010), it is likely that typical foraging distances for individuals of those species will be much lower (Wolf & Moritz, 2008).

Plausible pollinator habitat, as measured by overall vegetative cover within managed properties, was also generally low. Pindar and Raine (2023) suggest that 11.6%–16.7% of conserved habitat space is needed to support bee biodiversity. In general, just under a quarter of managed properties had enough vegetative cover to support ≥11.6% of a 100‐m foraging range, and that decreased as maximum foraging distances increased (Table 3). When considering parks as connected clusters serving a range of maximum foraging distances, plausible habitat availability increased substantially at 1000 m. For intermediate distances (250 and 500 m), <30% of the clusters have enough plausible pollinator habitat to support at least 11.6%, regardless of considering buildings as a barrier to movement (Table 3). This suggests that pollinator conservation efforts in cities should pair improving the connectivity between managed properties (Graffigna et al., 2023) with increasing habitat within managed properties, particularly if the focus is on species with shorter maximum foraging distances.

Effect of cities on pollinator movement

Pathfinding in bees has been best studied in A. mellifera (Menzel et al., 2005) and Bombus (Hagen et al., 2011), and little is known about pathfinding in solitary bees, and no direct studies of movement have been done in cities (but see MacIvor et al., 2014). Briefly, in eusocial bees, foragers will perform an orientation flight and then will fly away from the hive in straight‐line paths in search of suitable foraging patches (Menzel et al., 2005). Solitary bee foraging is understudied, but they may have similar pathfinding behaviors. Bee foraging behavior has been shown to be impacted by linear landscape features (e.g., hedgerows) in rural environments (Klaus et al., 2015; Krewenka et al., 2011), and while their effects have been little studied in pollinators directly (but see Zurbuchen, Bachofen, et al., 2010), buildings, bridges, and other urban structures have been shown to disrupt the movement of flying insects (Blakely et al., 2006; Kriska et al., 2008; Málnás et al., 2011).

The Euclidean‐derived paths assume that no buildings in New York City block pollinator movement. Conversely, the Dijkstra‐derived paths assume that all buildings in New York City act as barriers to pollinator movement. It is likely that shorter buildings, particularly those that have green roofs or gardens, may increase connectivity instead of hindering it (MacIvor, 2016). Indeed, on average 95.55% ± 10.40% of the buildings within a 1000‐m radius of each managed property in our study are ≤25 m tall (i.e., the height of the Jacob K. Javits Center green roof). However, most of the current buildings in New York City do not have green roofs, and whether bees will fly over shorter buildings versus around them is an open question. Recent work by Underwood et al. (2025), found that insect biodiversity and visitation rates to green roofs with floral resources decreased as roof height increased, with no insects detected at their highest green roof (24 m). Another recent study found that buildings impacted pollen movement by large bees like Bombus sp. and Apis mellifera (Roper & Youngsteadt, 2025). It is reasonable to predict that bee foraging movement is reduced and pathfinding strategies are modulated by buildings and other urban landscape features (i.e., more like those predicted by Dijkstra's algorithm than those measured by Euclidean distance), but further field studies are needed to determine to what extent.

Increased complexity and habitat fragmentation due to urban landscape barriers and distance between managed properties may be especially impactful on solitary bees, many of which have a maximum foraging distance well below 1000 m (Zurbuchen, Landert, et al., 2010). Conversely, foraging by eusocial bees like A. mellifera and Bombus sp. may be less impeded by urban landscape. Indeed, it has been suggested that cities may provide refuge for A. mellifera due to increased flower diversity (Fox et al., 2022) and lower exposure to pathogens than in rural environments (Samuelson et al., 2020). While solitary bees may receive similar benefits, the success of A. mellifera in urban environments may also pose a problem for native bee conservation due to competition for the limited pool of foraging resources (Prendergast et al., 2021).

Potential solutions

Green roofs

Tall buildings likely act as barriers to movement for pollinators and other animals in cities, but short‐ or medium‐sized buildings may not. While not studied directly in cities, solitary, wood‐nesting bees have been shown to forage and nest within the canopy of trees ~25 m above the forest floor in both temperate (Urban‐Mead et al., 2021) and tropical forests (Oliveira et al., 2024). Green roofs are an increasingly popular method for supporting pollinator biodiversity, among other benefits. While there is little published information about pollinator communities in New York City green roofs, green roofs in other cities have been shown to support diverse native pollinator communities (Passaseo et al., 2021; Tonietto et al., 2011). The heights of these green roofs are rarely reported, but there is evidence that the diversity of bees and other pollinators decreases with increased roof height (MacIvor, 2016). While there is little information about changes in altitude during foraging flights, it is reasonable to predict that it is more likely that bees will fly around buildings instead of over them unless there is some floral resource on that building. Green roofs are likely to play a critical role in facilitating the movement of pollinators between larger, ground‐level green spaces, and the establishment of new green roofs will be essential for urban pollinator conservation into the future (Susca et al., 2011).

Improving small parks

Most of the managed properties in New York City are small to medium sized (Figure 5) and have the greatest relative proportion of low, medium, and high vegetation cover (Figure 6). Small improvements, including the planting of pollinator‐friendly flowers in small patches can have a positive impact on pollinator communities (Daniels et al., 2020). This may be made even more impactful by considering park management within neighborhood clusters, particularly focused on pollinators with small foraging ranges.

Improving connectivity to large parks

Large managed properties (e.g., flagship parks like Central Park in Manhattan) are most likely to support larger patches of floral resources and habitat space (Banaszak‐Cibicka et al., 2018). Improving connectivity between smaller managed properties and larger ones could help both with the movement of bees between habitats and increase the resilience of pollinator populations across urban environments.

While building new parks and other green spaces would likely improve connectivity, a less costly solution could be to improve floral resources in street tree pits (Fryd et al., 2012; Reid et al., 2017). Indeed, our exploratory analysis of our Dijkstra‐derived paths intersected an average of 33.91 ± 0.13 street trees. While it has been demonstrated that street trees in New York City support a diverse insect community (Lundquist et al., 2022), there is little known about how pollinators currently utilize them but the planting of pollinator‐attracting flowers within street tree beds could act as stepping stones between parks (Płaskonka et al., 2024), creating a corridor for the movement of pollinators throughout the city. Other, non‐publicly managed open spaces (e.g., abandoned lots) may also serve as pollinator habitat or could be improved to support pollinators (Turo et al., 2021). Currently, reported additional open space totals 14.2 km2 citywide (sites 994 ± 58.4 m2 on average); however, only around 13% of that is combined low, medium, and high vegetation, and 23% is ground, which includes lawn and bare ground.

Utilization of this method in other cities

All the pathfinding modeling in this study utilized entirely open data provided by The City of New York (NYC Office of Technology and Innovation, 2024) and the United States Geological Survey (Garner, 2021). The analysis itself was done using the open‐source programs (QGIS) and Python libraries (e.g., shapely and networkx). Therefore, if local municipalities provide the necessary geographical data freely, anyone with a reasonably performant computer can utilize our methods of mapping the connectivity between parks or other green spaces. Furthermore, the use of this methodology is not limited to pollinators and could be used by conservationists, planners, and policymakers to explore questions related to conservation and habitat connectivity in highly urban landscapes.

CONCLUSIONS

The conservation of native pollinators in cities is an important component for urban sustainability. However, while parks and other green spaces may be able to support insect pollinators, we have found that the movement of bees around cities is likely hindered by buildings and other structures within the urban landscape. We have also demonstrated that in New York City, urban park size and vegetative cover are heterogeneous and a combination of improving connectivity as well as increasing habitat space would likely benefit urban pollinators. We know very little about the dynamics of pollinator movement and dispersal within urban landscapes and modeling these “bug roads” within cities lays the groundwork for future elucidation of the effects of urbanization on these ecologically important animals.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Supporting information

Appendix S1.

EAP-35-e70128-s001.pdf (1.1MB, pdf)

ACKNOWLEDGMENTS

We would like to acknowledge the assistance of Marymount Manhattan Lab Supervisor Paul Della‐Rocca. Funding for this project was from Marymount Manhattan College Faculty Scholarship Award and the Robert W. Ligon and Evelyn M. Ligon Memorial Fund.

Lundquist, Matthew J. , Lovejoy Pamela C., Fay Brianna G., Hernandez Juliet E., and Madrid Martha. 2025. “Bug Roads: Modeling the Green Space Connectivity and Pollinator Habitat in a Large City Using Open GIS Data and Tools.” Ecological Applications 35(7): e70128. 10.1002/eap.70128

Handling Editor: Juan C. Corley

DATA AVAILABILITY STATEMENT

All GIS files produced for this publication were derived from publicly accessible, open databases (NYC Open Data, New York State GIS Clearinghouse, Data.gov) and are available as GeoPackage (.gpkg) files in Lundquist (2025a) on the Open Science Framework at https://osf.io/9j6cq/. LiDAR data used for plausible pollinator habitat determination can be downloaded from the NYS GIS Clearinghouse LiDAR database (https://gis.ny.gov/lidar; direct download via FTP client: ftp://ftp.gis.ny.gov/elevation/LIDAR/NYC_2021). All data derived from analysis, including distances between parks and analysis of park clusters using Euclidean distance and Dijkstra's pathfinding algorithm, are available in Lundquist (2025a) on the Open Science Framework at https://osf.io/9j6cq/. Functions for shortest path analysis, network analysis, and LiDAR analysis (Lundquist, 2025b) are available in Zenodo at https://doi.org/10.5281/zenodo.17051114.

REFERENCES

  1. Ayers, A. C. , and Rehan S. M.. 2021. “Supporting Bees in Cities: How Bees Are Influenced by Local and Landscape Features.” Insects 12(2): 128. 10.3390/insects12020128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Banaszak‐Cibicka, W. , Twerd L., Fliszkiewicz M., Giejdasz K., and Langowska A.. 2018. “City Parks vs. Natural Areas – Is It Possible to Preserve a Natural Level of Bee Richness and Abundance in a City Park?” Urban Ecosystems 21(4): 599–613. 10.1007/s11252-018-0756-8. [DOI] [Google Scholar]
  3. Blakely, T. J. , Harding J. S., Mcintosh A. R., and Winterbourn M. J.. 2006. “Barriers to the Recovery of Aquatic Insect Communities in Urban Streams.” Freshwater Biology 51(9): 1634–1645. 10.1111/j.1365-2427.2006.01601.x. [DOI] [Google Scholar]
  4. Braaker, S. , Obrist M. K., Ghazoul J., and Moretti M.. 2017. “Habitat Connectivity and Local Conditions Shape Taxonomic and Functional Diversity of Arthropods on Green Roofs.” Journal of Animal Ecology 86(3): 521–531. 10.1111/1365-2656.12648. [DOI] [PubMed] [Google Scholar]
  5. Daniels, B. , Jedamski J., Ottermanns R., and Ross‐Nickoll M.. 2020. “A ‘Plan Bee’ for Cities: Pollinator Diversity and Plant‐Pollinator Interactions in Urban Green Spaces.” PLoS One 15(7): e0235492. 10.1371/journal.pone.0235492. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Despalatović, L. , Vojković T., and Vukic̆ević D.. 2014. “Community Structure in Networks: Girvan‐Newman Algorithm Improvement.” In 2014 37th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO), 997–1002. 10.1109/MIPRO.2014.6859714 [DOI]
  7. Egerer, M. , Annighöfer P., Arzberger S., Burger S., Hecher Y., Knill V., Probst B., and Suda M.. 2024. “Urban Oases: The Social‐Ecological Importance of Small Urban Green Spaces.” Ecosystems and People 20(1): 2315991. 10.1080/26395916.2024.2315991. [DOI] [Google Scholar]
  8. Fan, D. K. , and Shi P.. 2010. “Improvement of Dijkstra's Algorithm and Its Application in Route Planning.” In 2010 Seventh International Conference on Fuzzy Systems and Knowledge Discovery, 4:1901–1904. 10.1109/FSKD.2010.5569452. [DOI]
  9. Fox, G. , Vellaniparambil L. R., Ros L., Sammy J., Preziosi R. F., and Rowntree J. K.. 2022. “Complex Urban Environments Provide Apis mellifera with a Richer Plant Forage than Suburban and more Rural Landscapes.” Ecology and Evolution 12(11): e9490. 10.1002/ece3.9490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Fryd, O. , Pauleit S., and Bühler O.. 2012. “The Role of Urban Green Space and Trees in Relation to Climate Change.” Cabi Reviews 2011(January): 1–18. 10.1079/PAVSNNR20116053. [DOI] [Google Scholar]
  11. Garner, J. 2021. “Lidar Mapping Report: Acquisition, Processing, and Delivery of Airborne LiDAR Elevation Data. Conducted by Eagleview Advanced Remote Sensing.” Eagleview Advanced Remote Sensing.
  12. Gathmann, A. , and Tscharntke T.. 2002. “Foraging Ranges of Solitary Bees.” Journal of Animal Ecology 71(5): 757–764. 10.1046/j.1365-2656.2002.00641.x. [DOI] [Google Scholar]
  13. Graffigna, S. , Gonzalez‐Vaquero R., Torretta J., and Marrero H.. 2023. “Importance of Urban Green Areas' Connectivity for the Conservation of Pollinators.” Urban Ecosystems 27 (November):3: 417–426. 10.1007/s11252-023-01457-2. [DOI] [Google Scholar]
  14. Greenleaf, S. S. , Williams N. M., Winfree R., and Kremen C.. 2007. “Bee Foraging Ranges and their Relationship to Body Size.” Oecologia 153(3): 589–596. [DOI] [PubMed] [Google Scholar]
  15. Grüter, C. , and Hayes L.. 2022. “Sociality Is a Key Driver of Foraging Ranges in Bees.” Current Biology 32(24): 5390–5397.e3. 10.1016/j.cub.2022.10.064. [DOI] [PubMed] [Google Scholar]
  16. Hagen, M. , Wikelski M., and Kissling W. D.. 2011. “Space Use of Bumblebees (Bombus spp.) Revealed by Radio‐Tracking.” PLoS One 6(5): e19997. 10.1371/journal.pone.0019997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Harrison, T. , and Winfree R.. 2015. “Urban Drivers of Plant‐Pollinator Interactions.” Functional Ecology 29(7): 879–888. 10.1111/1365-2435.12486. [DOI] [Google Scholar]
  18. Johansson, V. , Koffman A., Hedbolm M., Deboni G., and Andersson P.. 2018. “Estimates of Accessible Food Resources for Pollinators in Urban Landscapes Should Take Landscape Friction into Account.” Ecosphere 9(10): e02486. 10.1002/ecs2.2486. [DOI] [Google Scholar]
  19. Kendall, L. K. , Mola J. M., Portman Z. M., Cariveau D. P., Smith H. G., and Bartomeus I.. 2022. “The Potential and Realized Foraging Movements of Bees Are Differentially Determined by Body Size and Sociality.” Ecology 103(11): e3809. 10.1002/ecy.3809. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Klaus, F. , Bass J., Marholt L., Müller B., Klatt B., and Kormann U.. 2015. “Hedgerows Have a Barrier Effect and Channel Pollinator Movement in the Agricultural Landscape.” Journal of Landscape Ecology 8(1): 22–31. 10.1515/jlecol-2015-0001. [DOI] [Google Scholar]
  21. Krewenka, K. M. , Holzschuh A., Tscharntke T., and Dormann C. F.. 2011. “Landscape Elements as Potential Barriers and Corridors for Bees, Wasps and Parasitoids.” Biological Conservation 144(6): 1816–1825. 10.1016/j.biocon.2011.03.014. [DOI] [Google Scholar]
  22. Kriska, G. , Malik P., Szivák I., and Horváth G.. 2008. “Glass Buildings on River Banks as ‘Polarized Light Traps’ for Mass‐Swarming Polarotactic Caddis Flies.” Naturwissenschaften 95(5): 461–467. 10.1007/s00114-008-0345-4. [DOI] [PubMed] [Google Scholar]
  23. Lowenstein, D. M. , Matteson K. C., and Minor E. S.. 2019. “Evaluating the Dependence of Urban Pollinators on Ornamental, Non‐Native, and ‘Weedy’ Floral Resources.” Urban Ecosystems 22(2): 293–302. 10.1007/s11252-018-0817-z. [DOI] [Google Scholar]
  24. Lundquist, M. J. 2025a. “Bug Roads: Modeling the Green Space Connectivity and Pollinator Habitat in a Large City Using Open GIS Data and Tools.” Open Science Framework. 10.17605/OSF.IO/9J6CQ. [DOI] [PubMed]
  25. Lundquist, M. J. 2025b. “Source Code for Bug Roads, Ecological Applications.” Zenodo. 10.5281/zenodo.17051114 [DOI] [PubMed]
  26. Lundquist, M. J. , Weisend M. R., and Kenmore H. H.. 2022. “Insect Biodiversity in Urban Tree Pit Habitats.” Urban Forestry & Urban Greening 78: 127788. 10.1016/j.ufug.2022.127788. [DOI] [Google Scholar]
  27. MacIvor, J. S. 2016. “Building Height Matters: Nesting Activity of Bees and Wasps on Vegetated Roofs.” Israel Journal of Ecology & Evolution 62(1–2): 88–96. 10.1080/15659801.2015.1052635. [DOI] [Google Scholar]
  28. MacIvor, J. S. , Cabral J. M., and Packer L.. 2014. “Pollen Specialization by Solitary Bees in an Urban Landscape.” Urban Ecosystems 17(1): 139–147. 10.1007/s11252-013-0321-4. [DOI] [Google Scholar]
  29. Málnás, K. , Polyák L., Prill É., Hegedüs R., Kriska G., Dévai G., Horváth G., and Lengyel S.. 2011. “Bridges as Optical Barriers and Population Disruptors for the Mayfly Palingenia longicauda: An Overlooked Threat to Freshwater Biodiversity?” Journal of Insect Conservation 15(6): 823–832. 10.1007/s10841-011-9380-0. [DOI] [Google Scholar]
  30. Menzel, R. , Greggers U., Smith A., Berger S., Brandt R., Brunke S., Bundrock G., et al. 2005. “Honey Bees Navigate According to a Map‐like Spatial Memory.” Proceedings of the National Academy of Sciences 102(8): 3040–3045. 10.1073/pnas.0408550102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. New York State Office of Parks, Recreation, and Historic Preservation . 2021. NY State Parks Property. Albany, NY: NYS GIS Clearinghouse. https://data.gis.ny.gov/datasets/nysparks::ny-state-parks-property/explore?location=40.836535%2C-73.892906%2C11.81. [Google Scholar]
  32. NYC Office of Technology and Innovation . 2024. “NYC Open Data.” https://opendata.cityofnewyork.us/
  33. Oliveira, H. K. L. G. , Miranda P. N., Ortega J. C. G., and Morato E. F.. 2024. “Vertical Stratification of Solitary Bees and Wasps in an Urban Forest from the Brazilian Amazon.” Neotropical Entomology 53(3): 552–567. 10.1007/s13744-024-01142-9. [DOI] [PubMed] [Google Scholar]
  34. Passaseo, A. , Rochefort S., Pétremand G., and Castella E.. 2021. “Pollinators on Green Roofs: Diversity and Trait Analysis of Wild Bees (Hymenoptera: Anthophila) and Hoverflies (Diptera: Syrphidae) in an Urban Area (Geneva, Switzerland).” Cities and the Environment (CATE) 14(2): Article 1.  10.15365/cate.2021.140201. [DOI] [Google Scholar]
  35. Pindar, A. , and Raine N. E.. 2023. “Safeguarding Pollinators Requires Specific Habitat Prescriptions and Substantially More Land Area than Suggested by Current Policy.” Scientific Reports 13(1): 1040. 10.1038/s41598-022-26872-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Płaskonka, B. , Zych M., Mazurkiewicz M., Skłodowski M., and Roguz K.. 2024. “Pollinator‐Mediated Connectivity in Fragmented Urban Green Spaces—Tracking Pollen Grain Movements in the City Center.” Acta Oecologica 123(May): 103985. 10.1016/j.actao.2024.103985. [DOI] [Google Scholar]
  37. Potts, S. G. , Biesmeijer J. C., Kremen C., Neumann P., Schweiger O., and Kunin W. E.. 2010. “Global Pollinator Declines: Trends, Impacts and Drivers.” Trends in Ecology & Evolution 25(6): 345–353. 10.1016/j.tree.2010.01.007. [DOI] [PubMed] [Google Scholar]
  38. Prendergast, K. S. , Dixon K. W., and Bateman P. W.. 2021. “Interactions between the Introduced European Honey Bee and Native Bees in Urban Areas Varies by Year, Habitat Type and Native Bee Guild.” Biological Journal of the Linnean Society 133(3): 725–743. 10.1093/biolinnean/blab024. [DOI] [Google Scholar]
  39. QGIS.org . 2024. “QGIS Geographic Information System.” QGIS Association. https://www.qgis.org
  40. Ratnieks, F. L. W. , and Shackleton K.. 2015. “Does the Waggle Dance Help Honey Bees to Forage at Greater Distances than Expected for Their Body Size?” Frontiers in Ecology and Evolution 3‐2015. 10.3389/fevo.2015.00031. [DOI] [Google Scholar]
  41. Reid, C. E. , Clougherty J. E., Shmool J. L. C., and Kubzansky L. D.. 2017. “Is All Urban Green Space the Same? A Comparison of the Health Benefits of Trees and Grass in New York City.” International Journal of Environmental Research and Public Health 14(11): 1411. 10.3390/ijerph14111411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Remmers, R. , and Frantzeskaki N.. 2024. “Bees in the City: Findings from a Scoping Review and Recommendations for Urban Planning.” Ambio 53: 1281–1295. 10.1007/s13280-024-02028-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Roper, O. I. , and Youngsteadt E.. 2025. “Bee‐Mediated Pollen Transport across Five Urban Landscape Features: Buildings Are Important Barriers.” Ecology and Evolution 15: e71339. 10.1002/ece3.71339. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Sallay, Á. , Gecséné Tar I., Mikházi Z., Takács K., Furlan C., and Krippner U.. 2023. “The Role of Urban Cemeteries in Ecosystem Services and Habitat Protection.” Plants 12(6): 1269. 10.3390/plants12061269. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Samuelson, A. E. , Gill R. J., and Leadbeater E.. 2020. “Urbanisation Is Associated with Reduced Nosema Sp. Infection, Higher Colony Strength and Higher Richness of Foraged Pollen in Honeybees.” Apidologie 51(5): 746–762. 10.1007/s13592-020-00758-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Smalls‐Mantey, L. , and Montalto F.. 2021. “The Seasonal Microclimate Trends of a Large Scale Extensive Green Roof.” Building and Environment 197(June): 107792. 10.1016/j.buildenv.2021.107792. [DOI] [Google Scholar]
  47. Susca, T. , Gaffin S. R., and Dell'Osso G. R.. 2011. “Positive Effects of Vegetation: Urban Heat Island and Green Roofs.” Environmental Pollution 159(8–9): 2119–2126. 10.1016/j.envpol.2011.03.007. [DOI] [PubMed] [Google Scholar]
  48. Tanda, A. S. 2023. “Advances In Insect Pollination Technology in Sustainable Agriculture.” I K International Pvt Ltd
  49. Tonietto, R. , Fant J., Ascher J., Ellis K., and Larkin D.. 2011. “A Comparison of Bee Communities of Chicago Green Roofs, Parks and Prairies.” Landscape and Urban Planning 103(1): 102–108. 10.1016/j.landurbplan.2011.07.004. [DOI] [Google Scholar]
  50. Turo, K. J. , Spring M. L. R., Sivakoff F. S., de la Flor Y. A. D., and Gardiner M. M.. 2021. “Conservation in Post‐Industrial Cities: How Does Vacant Land Management and Landscape Configuration Influence Urban Bees?” Journal of Applied Ecology 58(1): 58–69. 10.1111/1365-2664.13773. [DOI] [Google Scholar]
  51. Underwood, S. M. , Shookhan N., Hung K.‐L. J., and MacIvor J. S.. 2025. “Pollinator Visits Increase with Bloom Amount but Decline with Building Height on Extensive Green Roofs.” Insect Conservation and Diversity 18(3): 438–445. 10.1111/icad.12806. [DOI] [Google Scholar]
  52. United States National Park Service . 2020. “National Parks Boundaries GIS Data.” U.S. Department of the Interior. https://catalog.data.gov/dataset/national-park-boundaries.
  53. Urban‐Mead, K. R. , Muñiz P., Gillung J., Espinoza A., Fordyce R., van Dyke M., McArt S. H., and Danforth B. N.. 2021. “Bees in the Trees: Diverse Spring Fauna in Temperate Forest Edge Canopies.” Forest Ecology and Management 482(February): 118903. 10.1016/j.foreco.2020.118903. [DOI] [Google Scholar]
  54. Weber, M. , Diekötter T., Dietzsch A. C., Erler S., Greil H., Jütte T., Krahner A., and Pistorius J.. 2023. “Urban Wild Bees Benefit from Flower‐Rich Anthropogenic Land Use Depending on Bee Trait and Scale.” Landscape Ecology 38(11): 2981–2999. 10.1007/s10980-023-01755-2. [DOI] [Google Scholar]
  55. Wolf, S. , and Moritz R. F. A.. 2008. “Foraging Distance in Bombus terrestris L. (Hymenoptera: Apidae).” Apidologie 39(4): 419–427. 10.1051/apido:2008020. [DOI] [Google Scholar]
  56. Zurbuchen, A. , Bachofen C., Müller A., Hein S., and Dorn S.. 2010. “Are Landscape Structures Insurmountable Barriers for Foraging Bees? A Mark‐Recapture Study with Two Solitary Pollen Specialist Species.” Apidologie 41(4): 497–508. 10.1051/apido/2009084. [DOI] [Google Scholar]
  57. Zurbuchen, A. , Cheesman S., Klaiber J., Müller A., Hein S., and Dorn S.. 2010. “Long Foraging Distances Impose High Costs on Offspring Production in Solitary Bees.” Journal of Animal Ecology 79(3): 674–681. 10.1111/j.1365-2656.2010.01675.x. [DOI] [PubMed] [Google Scholar]
  58. Zurbuchen, A. , Landert L., Klaiber J., Müller A., Hein S., and Dorn S.. 2010. “Maximum Foraging Ranges in Solitary Bees: Only Few Individuals Have the Capability to Cover Long Foraging Distances.” Biological Conservation 143(3): 669–676. 10.1016/j.biocon.2009.12.003. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix S1.

EAP-35-e70128-s001.pdf (1.1MB, pdf)

Data Availability Statement

All GIS files produced for this publication were derived from publicly accessible, open databases (NYC Open Data, New York State GIS Clearinghouse, Data.gov) and are available as GeoPackage (.gpkg) files in Lundquist (2025a) on the Open Science Framework at https://osf.io/9j6cq/. LiDAR data used for plausible pollinator habitat determination can be downloaded from the NYS GIS Clearinghouse LiDAR database (https://gis.ny.gov/lidar; direct download via FTP client: ftp://ftp.gis.ny.gov/elevation/LIDAR/NYC_2021). All data derived from analysis, including distances between parks and analysis of park clusters using Euclidean distance and Dijkstra's pathfinding algorithm, are available in Lundquist (2025a) on the Open Science Framework at https://osf.io/9j6cq/. Functions for shortest path analysis, network analysis, and LiDAR analysis (Lundquist, 2025b) are available in Zenodo at https://doi.org/10.5281/zenodo.17051114.

Articles from Ecological Applications are provided here courtesy of Wiley

RESOURCES