The relative influence of catchment and site variables on fish and macroinvertebrate richness in cerrado biome streams

Landscape and site-scale data analyses aid the interpretation of biological data and thereby help us develop more cost-effective natural resource management strategies. Our study focused on environmental influences on stream assemblages and we evaluated how three classes of environmental variables (geophysical landscape, land use and cover, and site habitat), influence fish and macroinvertebrate assemblage richness in the Brazilian Cerrado biome. We analyzed our data through use of multiple linear regression (MLR) models using the three classes of predictor variables alone and in combination. The four MLR models explained dissimilar amounts of benthic macroinvertebrate taxa richness (geophysical landscape R ² ≈ 35 %, land use and cover R ² ≈ 28 %, site habitat R ² ≈ 36 %, and combined R ² ≈ 51 %). For fish assemblages, geophysical landscape, land use and cover, site habitat, and combined models explained R ² ≈ 28 %, R ² ≈ 10 %, R ² ≈ 31 %, and R ² ≈ 47 % of the variability in fish species richness, respectively. We conclude that (1) environmental variables differed in the degree to which they explain assemblage richness, (2) the amounts of variance in assemblage richness explained by geophysical landscape and site habitat were similar, (3) the variables explained more variability in macroinvertebrate taxa richness than in fish species richness, and (4) all three classes of environmental variables studied were useful for explaining assemblage richness in Cerrado headwater streams. These results help us to understand the drivers of assemblage patterns at regional scales in tropical areas.     

Modelling Habitat Suitability for Deciduous Forest Focal Species – A Sensitivity Analysis using Different Satellite Land Cover Data

We explored the usefulness of three satellite land cover data sets available to land managers in south-central Sweden for conservation planning using four deciduous forest focal resident bird species with different habitat requirements. Habitat suitability models using empirical species-specific habitat parameters and a Geographic Information System were applied to evaluate and compare the degree of consistency among three different land cover data sets. The study area encompassed 10,000 km² in a landscape mosaic of managed boreal forests and is within the distribution range of all four focal species. Although the three land cover data sets indicated similar total amounts of deciduous forest, the habitat suitability models showed that different land cover data yielded inconsistent results regarding the amount and distribution of suitable habitat within 5×5 km grid cells. Given this sensitivity to the choice of land cover data sets, the habitat suitability models showed positive relationships among the selected focal species for each land cover data set separately. As expected, decreasing amounts of suitable habitat were identified for species with higher specialisation. Thus, because habitat suitability models are an appropriate way to gain insight into the functionality and connectivity of habitat networks, land cover data must be carefully evaluated and if necessary combined with other landscape information for effective conservation planning.     

The impacts of historical land-use and landscape variables on hollow-bearing trees along an urbanisation gradient

Hollow-bearing trees provide habitat for diverse taxonomic groups and as such they are recognised for their importance globally. There is, however scant reference to this resource relative within urban forest patches. The functional ecology of habitat remnants along an urbanisation gradient plays an important ecological, social and economic role within urban landscapes. Here we quantify the impacts of urbanisation, landscape, environmental, disturbance (past and present) and stand variables on hollow-bearing tree density within urban forest patches. This was undertaken by surveying 45 forest patches on the Gold Coast, south-east Queensland, Australia. Sites were categorised as; urban, peri-urban or rural along an urbanisation gradient, with an additional five control sites. Historical logging practices were found to be the driving factor influencing hollow-bearing tree density along the urbanisation gradient; while the impacts of urbanisation itself are not as yet discernible. These findings highlight the significance of incorporating historical land use practise into current and future urban planning, as these will have continuing impacts on remaining urban biodiversity values. These findings, will benefit natural resource managers and urban planners when making decisions about where and how best to manage for hollow-bearing trees along urbanisation gradients.     

Distribution of woodland amphibians along a forest fragmentation gradient

Understanding how changes in land-use affect the distribution and abundance of organisms is an increasingly important question in landscape ecology. Amphibians may be especially prone to local extinction resulting from human-caused transformation and fragmentation of their habitats owing to the spatially and temporally dynamic nature of their populations. In this study, distributions of five species of woodland amphibians with differing life histories were surveyed along a 10 km, spatially continuous gradient of forest fragmentation in southern Connecticut, U.S.A. Redback salamanders (Plethodon cinereus) and northern spring peepers (Pseudacris c. crucifer) occupied available habitat along the gradient's length. Wood frogs (Rana sylvatica) and spotted salamanders (Ambystoma maculatum) were absent from portions of the gradient where forest cover was reduced to below about 30%. Red-spotted newts (Notophthalmus v. viridescens) did not persist below a forest cover threshold of about 50%. Correlations between species' biological traits and their fragmentation tolerance imply that low density, population variability, and high mobility coupled with restricted habitat needs predispose woodland amphibians to local extinction caused by habitat fragmentation. These patterns are in contrast to the widely held notion that populations of the best dispersers are those most tolerant of habitat fragmentation.     

What multiscale environmental drivers can best be discriminated from a habitat index derived from a remotely sensed vegetation time series?

Understanding which environmental conditions are critical for species survival is a critical, ongoing question in ecology. These conditions can range from climate, at the broadest scale, through to elevation and other local landscape conditions, to fine scale landscape patterns of land cover and use. Remote sensing is an ideal technology to monitor and assess changes in these environmental conditions at a variety of spatial and temporal scales, with many studies focusing on the physiological state of vegetation derived from time series of satellite measurements. As vegetation occurs within specific climatic zones, over certain soil, terrain, and land cover types, it can be difficult to decipher the influence of the underlying role of climate, topography, soil, and land cover on the observed vegetation signal. In this article, we specifically addressed this problem by asking the question: what is the relative impact and importance of these different scales of environmental drivers on the temporal and spatial patterns observed on a habitat index derived from remotely sensed data? To find the solution, we utilized a SPOT VEGETATION-normalized difference vegetation index time series of Europe to create a remote-sensing-derived habitat index, which incorporates aspects of productivity, seasonality, and cover. We then compared the observed temporal and spatial variations in the index to a pan-Europe terrestrial classification system, which explicitly incorporates variations in climate, terrain, soil parent material, land cover, and use. Results indicated that the most accurate level of discrimination from the habitat index was at the broadest level of the hierarchy, climate, while the poorest degree of discrimination was associated with elevation. In terms of similarity on the index across time and space, we found that arable and forest cover classes were more similar across elevation and parent materials than across other land cover types within them. Analyzing the remote-sensing index, at multiple scales, provides significant insights into the drivers of satellite-derived greenness indices, as well as highlights the benefit and cautions associated with linking satellite-derived indirect indicators to species distribution modeling and biodiversity.     

Influence of patch and landscape characteristics on the distribution of the subterranean rodent <Emphasis Type="Italic">Ctenomys porteousi</Emphasis>

The understanding and prediction of the responses of animal populations to habitat fragmentation is a central issue in applied ecology. The identification of habitat variables associated to patch occupancy is particularly important when habitat quality is affected by human activities. Here, we analyze the influence of patch and landscape characteristics on patch occupancy by the subterranean herbivorous rodent Ctenomys porteousi. Patch occupancy was monitored in a network of 63 habitat patches identified by satellite imagery analysis which extends along almost the whole distributional range for C. porteousi. Suitable habitat for the occurrence of C. porteousi is highly fragmented and represents <10% of the total area in its distributional range. The distribution of C. porteousi in the patch network is affected not only by characteristics of the habitat patches, but also by those of the surrounding landscape matrix. Significant differences between occupied and empty patches were found in several environmental variables. Overall, occupied patches were larger, less vegetated, more connected, and had larger neighbor patches than empty patches. A stepwise procedure on a generalized linear model selected four habitat variables that explain patch occupancy in C. porteousi; it included the effects of habitat quality in the matrix surrounding the patch, average vegetation cover in the patch, minimum vegetation cover in the matrix surrounding the patch, and the area of the nearest neighbor patch. These results indicate that patch occupancy in C. porteousi is strongly influenced by the availability and quality of habitat both in the patch and in the surrounding landscape matrix.     

Influences of upland and riparian land use patterns on stream biotic integrity

We explored land use, fish assemblage structure, and stream habitat associations in 20 catchments in Opequon Creek watershed, West Virginia. The purpose was to determine the relative importance of urban and agriculture land use on stream biotic integrity, and to evaluate the spatial scale (i.e., whole-catchment vs riparian buffer) at which land use effects were most pronounced. We found that index of biological integrity (IBI) scores were strongly associated with extent of urban land use in individual catchments. Sites that received ratings of poor or very poor based on IBI scores had > 7% of urban land use in their respective catchments. Habitat correlations suggested that urban land use disrupted flow regime, reduced water quality, and altered stream channels. In contrast, we found no meaningful relationship between agricultural land use and IBI at either whole-catchment or riparian scales despite strong correlations between percent agriculture and several important stream habitat measures, including nitrate concentrations, proportion of fine sediments in riffles, and the abundance of fish cover. We also found that variation in gradient (channel slope) influenced responses of fish assemblages to land use. Urban land use was more disruptive to biological integrity in catchments with steeper channel slopes. Based on comparisons of our results in the topographically diverse Opequon Creek watershed with results from watersheds in flatter terrains, we hypothesize that the potential for riparian forests to mitigate effects of deleterious land uses in upland portions of the watershed is inversely related to gradient.This revised version was published online in May 2005 with corrections to the Cover Date.     

Riparian influences on stream fish assemblage structure in urbanizing streams

We assessed the influence of land cover at multiple spatial extents on fish assemblage integrity, and the degree to which riparian forests can mitigate the negative effects of catchment urbanization on stream fish assemblages. Riparian cover (urban, forest, and agriculture) was determined within 30 m buffers at longitudinal distances of 200 m, 1 km, and the entire network upstream of 59 non-nested fish sampling locations. Catchment and riparian land cover within the upstream network were highly correlated, so we were unable to distinguish between those variables. Most fish assemblage variables were related to % forest and % urban land cover, with the strongest relations at the largest spatial extent of land cover (catchment), followed by riparian land cover in the 1-km and 200-m reach, respectively. For fish variables related to urban land cover in the catchment, we asked whether the influence of riparian land cover on fish assemblages was dependent on the amount of urban development in the catchment. Several fish assemblage metrics (endemic richness, endemic:cosmopolitan abundance, insectivorous cyprinid richness and abundance, and fluvial specialist richness) were all best predicted by single variable models with % urban land cover. However, endemic:cosmopolitan richness, cosmopolitan abundance, and lentic tolerant abundance were related to % forest cover in the 1-km stream reach, but only in streams that had <15% catchment urban land cover. In these cases, catchment urbanization overwhelmed the potential mitigating effects of riparian forests on stream fishes. Together, these results suggest that catchment land cover is an important driver of fish assemblages in urbanizing catchments, and riparian forests are important but not sufficient for protecting stream ecosystems from the impacts of high levels of urbanization.     

Physical and land-cover variables influence ant functional groups and species diversity along elevational gradients

Of particular importance in shaping species assemblages is the spatial heterogeneity of the environment. The aim of our study was to investigate the influence of spatial heterogeneity and environmental complexity on the distribution of ant functional groups and species diversity along altitudinal gradients in a temperate ecosystem (Pyrenees Mountains). During three summers, we sampled 20 sites distributed across two Pyrenean valleys ranging in altitude from 1,009 to 2,339 m by using pitfall traps and hand collection. The environment around each sampling points was characterized by using both physical and land-cover variables. We then used a self-organizing map algorithm (SOM, neural network) to detect and characterize the relationship between the spatial distribution of ant functional groups, species diversity, and the variables measured. The use of SOM allowed us to reduce the apparent complexity of the environment to five clusters that highlighted two main gradients: an altitudinal gradient and a gradient of environmental closure. The composition of ant functional groups and species diversity changed along both of these gradients and was differently affected by environmental variables. The SOM also allowed us to validate the contours of most ant functional groups by highlighting the response of these groups to the environmental and land-cover variables.     

Historical landscape ecology of an urbanized California valley: wetlands and woodlands in the Santa Clara Valley

Historical records provide information to land managers and landscape ecologists attempting to understand current trajectories in altered landscapes. In this study, we synthesized a heterogeneous array of historical sources to reconstruct historical land cover in California’s Santa Clara Valley (a.k.a. “Silicon Valley”). To increase and assess accuracy, we used the triangulation of overlapping, independent data sources and the application of certainty level standards. The region has been subject to extensive urbanization, so we also evaluated the applicability of historical landscape reconstructions to the altered landscape. We found evidence for five major land cover types prior to significant Euro–American modification. Valley freshwater marsh, wet meadow, alkali meadow, willow grove, and valley oak savanna have all experienced extreme decline (85–100%) since Euro–American settlement. However, comparison of historical land cover patterns to contemporary land use suggested several new strategies for environmental recovery, despite the limitations of surrounding urbanization. We also observed a temporal shift in riparian habitat along the mainstem of Coyote Creek, from a relatively open mixture of riparian scrub, sycamore woodland, and unvegetated gravel bars to dense riparian forest, likely resulting from stream flow regulation. By identifying former land cover patterns we provide a basis for evaluating local landscape change and setting restoration targets, including the identification of residual features and under-recognized land cover types. These findings suggest that reliable historical landscape reconstructions can be developed in the absence of standardized historical data sources and can be of value even in highly modified regions.     

Historic landscape change and habitat loss: the case of black grouse in Lower Saxony,Germany

The declines of many specialist bird species in the agricultural landscapes of Central Europe have resulted in small and isolated populations. In the case of the black grouse, a ground-nesting bird species with large spatial requirements, empiric evidence about underlying landscape changes is scarce. In this study, we examined land cover and land cover changes in a farmland-forest mosaic in eastern Lower Saxony, Germany and how they affect occurrence and persistence of black grouse. Spatial information came from historic topographic maps from 1958 to 1975. The results show profound conversions of habitat to forest and farmland but also an increase in settlement area. Habitat conversions and suburbanization were negative correlates of black grouse persistence. Habitat models from before and after a decline period differed in some of the predictors and suggest black grouse habitat to be more diverse before the land cover changes. Our study confirms that land use factors at a landscape scale extent contribute to explain black grouse occurrence and thus can complement important small scale factors like the quality and size of individual habitat patches. Results also show that landscape factors affect black grouse distribution predominantly from an area much greater than an individual black grouse home range. Our models may be further evaluated on present-day landscapes and might be used to evaluate large-scale habitat availability for black grouse.     

Landscape influences on stream biotic integrity assessed at multiple spatial scales

The biological integrity of stream ecosystems depends critically on human activities that affect land use/cover along stream margins and possibly throughout the catchment. We evaluated stream condition using an Index of Biotic Integrity (IBI) and a habitat index (HI), and compared these measures to landscape and riparian conditions assessed at different spatial scales in a largely agricultural Midwestern watershed. Our goal was to determine whether land use/cover was an effective predictor of stream integrity, and if so, at what spatial scale. Twenty-three sites in first-through third-order headwater streams were surveyed by electrofishing and site IBIs were calculated based on ten metrics of the fish collection. Habitat features were characterized through field observation, and site HIs calculated from nine instream and bank metrics. Field surveys, aerial photograph interpretation, and geographic information system (GIS) analyses provided assessments of forested land and other vegetation covers at the local, reach, and regional (catchment) scales. The range of conditions among the 23 sites varied from poor to very good based on IBI and HI scores, and habitat and fish assemblage measures were highly correlated. Stream biotic integrity and habitat quality were negatively correlated with the extent of agriculture and positively correlated with extent of wetlands and forest. Correlations were strongest at the catchment scale (IBI with % area as agriculture, r²=0.50, HI with agriculture, r²=0.76), and tended to become weak and non-significant at local scales. Local riparian vegetation was a weak secondary predictor of stream integrity. In this watershed, regional land use is the primary determinant of stream conditions, able to overwhelm the ability of local site vegetation to support high-quality habitat and biotic communities.     

A geography of ecosystem vulnerability

Land-cover change and the subsequent potential loss of natural resources due to conversion to anthropogenic use is regarded as one of the more pervasive environmental threats. Population and road data were used to generate interpolated surfaces of land demand across a large region, the mid-Atlantic states of Pennsylvania, Delaware, Maryland, Virginia, and West Virginia. The land demand surfaces were evaluated against land-cover change, as estimated using temporal decline in Normalized Difference Vegetation Index (NDVI). In general, the interpolated surfaces exhibited a plateau along the eastern seaboard that sank to a valley in the center of the study area, and then rose again to a plateau in the west that was of overall lower height than the plateau on the eastern seaboard. The spatial pattern of land-cover change showed the same general pattern as the interpolated surfaces of land demand. Correlations were significant regardless of variations used to generate the interpolated surfaces. The results suggest that human activity is the principal agent of land-cover change at regional scales in this region, and that natural resources that change as land cover changes (e.g., water, habitat) are exposed to a gradient of vulnerability that increases from west to east.     

Spatial patterns of bird community similarity: bird responses to landscape composition and configuration in the Atlantic forest

Studies dealing with community similarity are necessary to understand large scale ecological processes causing biodiversity loss and to improve landscape and regional planning. Here, we study landscape variables influencing patterns of community similarity in fragmented and continuous forest landscapes in the Atlantic forest of South America, isolating the effects of forest loss, fragmentation and patterns of land use. Using a grid design, we surveyed birds in 41 square cells of 100 km² using the point count method. We used multivariate, regression analyses and lagged predictor autoregressive models to examine the relative influence of landscape variables on community similarity. Forest cover was the primary variable explaining patterns of bird community similarity. Similarity showed a sudden decline between 20 and 40% of forest cover. Patterns of land use had a second order effect; native bird communities were less affected by forest loss in landscapes dominated by tree plantations (the most suitable habitat for native species) than in landscapes dominated by annual crops or cattle pastures. The effects of fragmentation were inconclusive. The trade-off between local extinctions and the invasion of extra-regional species using recently created habitats is probably the mechanism generating the observed patterns of community similarity. Limiting forest loss to 30–40% of the landscape cover and improving the suitability of human-modified habitats will contribute to maintain the structure and composition of the native forest bird community in the Atlantic forest.     

Modeling relative habitat suitability of southern Florida for invasive Burmese pythons (<Emphasis Type="Italic">Python molurus bivittatus</Emphasis>)

Context

Invasive Burmese pythons are altering the ecology of southern Florida and their distribution is expanding northward. Understanding their habitat use is an important step in understanding the pathways of the invasion.

Objectives

This study identifies key landscape variables in predicting relative habitat suitability for pythons at the present stage of invasion through presence-only ecological niche modeling using geographical sampling bias correction.

Methods

We used 2014 presence-only observations from the EDDMapS database and three landscape variables to model habitat suitability: fine-scale land cover, home range-level land cover, and distance to open freshwater or wetland. Ten geographical sampling bias correction scenarios based on road presence and sampling effort were evaluated to improve the efficacy of modeling.

Results

The best performing models treated road presence as a binary factor rather than a continuous decrease in sampling effort with distance from roads. Home range-level cover contributed the most to the final prediction, followed by proximity to water and fine-scale land cover. Estuarine habitat and freshwater wetlands were the most important variables to contribute to python habitat suitability at both the home range-level and fine-scale. Suitability was highest within 30 m of open freshwater and wetlands.

Conclusions

This study provides quantifiable, predictive relationships between habitat types and python presence at the current stage of invasion. This knowledge can elucidate future targeted studies of python habitat use and behavior and help inform management efforts. Furthermore, it illustrates how estimates of relative habitat suitability derived from MaxEnt can be improved by both multi-scale perspectives on habitat and consideration of a variety of bias correction scenarios for selecting background points.

     

The relative importance of local versus landscape variables on site occupancy in bats of the Brazilian Cerrado

Context

Species site-occupancy patterns may be influenced by habitat variables at both local and landscape scales. Although local habitat variables influence whether the site is suitable for a given species, the broader landscape context can also influence site occupancy, particularly for species that are sensitive to land-use change.

Objectives

To examine the relative importance of local versus landscape variables in explaining site occupancy of eight bat species within the Brazilian Cerrado, a Neotropical savanna that is experiencing widespread habitat loss and fragmentation.

Methods

Bats were surveyed within 16 forest patches over two years. We used a multi-model information-theoretic approach, adjusted for species detection bias, to assess whether landscape variables (percent cover and number of patches of natural vegetation within a 2- and 8-km radius of each forest site) or local site variables (canopy cover, understory height, number of trees, and number of lianas) best explained site occupancy in each species.

Results

Landscape variables were among the best models (ΔAIC_c or ΔQAIC_c < 2) for four species (top-ranked model for black myotis), whereas local variables were among the best for five species (top-ranked model for vampire bats). Neither local nor landscape variables explained site occupancy in two frugivorous species.

Conclusion

Species associated with a particular habitat type will not respond similarly to the amount, distribution or relative suitability of that habitat, or even at the same scale. This reinforces the challenge of species distribution modelling, especially in the context of forecasting species’ responses to future land-use or climate-change scenarios.

     

Associations of Bird Species Richness and Community Composition with Local and Landscape-scale Environmental Factors in Borneo

A comprehensive understanding of variables associated with spatial differences in community composition is essential to explain and predict biodiversity over landscape scales. In this study, spatial patterns of bird diversity in Central Kalimantan, Indonesia, were examined and associated with local-scale (habitat structure and heterogeneity) and landscape-scale (logging, slope position and elevation) environmental variables. Within the study area (c. 196 km²) local habitat structure and heterogeneity varied considerably, largely due to logging. In total 9747 individuals of 177 bird species were recorded. Akaike's information criterion (AIC) revealed that the best explanatory models of bird community similarity and species richness included both local- and landscape-scale environmental variables. Important local-scale variables included liana abundance, fern cover, sapling density, tree density, dead wood abundance and tree architecture, while important landscape-scale variables were elevation, logging and slope position. Geographic distance between sampling sites was not significantly associated with spatial variation in either species richness or similarity. These results indicate that deterministic environmental processes, as opposed to dispersal-driven stochastic processes, primarily structure bird assemblages within the spatial scale of this study and confirm that highly variable local habitat measures can be effective means of predicting landscape-scale community patterns.     

Soil occupation efficiency and landscape conservation in four Mediterranean urban regions

This study investigates the relationship between soil sealing and landscape conservation in four Mediterranean regions (Athens, Barcelona, Lisbon, Rome) characterized by different patterns of urban expansion. Per-capita sealed land, a landscape conservation index and selected territorial variables were considered into a multivariate exploratory framework aimed at assessing the correlation between land-use efficiency (based on the degree of soil sealing per-capita) and the quality of suburban landscape. A population density gradient with intensity of sealed land decreasing with the distance from the central city was observed in compact urban regions such as Athens and Barcelona. A mixed urban gradient was observed in Rome and Lisbon. In all the considered cities the spatial distribution of per-capita sealed land was not correlated with the urban gradient indicating that land consumption follows place-specific patterns irrespective of landscape quality. These findings suggest that urban containment and landscape conservation are policy targets requiring environmental measures irrespective of the prevailing morphology of the urban region (compact vs dispersed). In this context, green infrastructure planning is a promising tool for landscape conservation and the containment of soil sealing within fragile and dynamic contexts such as the wildland-urban interface.     

Plant species response to urbanization: comparison of isolated woodland patches in two cities of North-Western France

The effect of urbanization on species distribution has been extensively documented, but a main challenge in urban ecology is to better understand the factors causing different distributions among species in response to urbanization. Hence, this paper aims to compare the effects of urbanization on woodland plant assemblages in two cities and to describe species responses by using several indicators. The study was carried out in the cities of Angers and Rennes (North-Western France) where 11 isolated woodlands were surveyed along an urban–rural gradient in each city. Abundance data of spontaneous species were collected from 220 quadrats. The effect of land cover (within a 500 m buffer around each woodland) on species assemblages was investigated by Canonical Correspondence Analysis. Buildings and pavement areas were the most significant predictors of species composition, and the effect of location in Angers or Rennes appeared on the second axis. More than 60% of the most frequent plant species were indicator of urban or rural location and their preferences were similar in the two cities. These lists of urban and rural indicator species were compared with Ellenberg’s indicator values and two other indicators specific to forest environment. The species which grow preferentially in urban woodlands are species which are already known to be associated with recent forests rather than ancient forests; with hedgerows rather than woodlands. The opposite pattern was observed concerning rural species. Moreover, urban indicator species have higher optima for soil pH and soil nitrogen content than rural indicator species. Different characteristics and history of forest habitat—continuity of the forest land cover, linearity of the habitat, change in adjacent land cover and land use—could select the same species, and the responses of the latter might involve different preferences concerning soil alkalinity and nutrient status.     

Understanding spontaneous biodiversity in informal urban green spaces: A local-landscape filtering framework with a test on wall plants

In densely urbanized areas, small pockets of vegetated areas such as street verges, vacant lots, and walls can be rich in biodiversity. In spite of their small size, these ‘informal urban greenspaces’ can provide critical ecosystem services to urban residents. Maintaining and enhancing the provisioning of ecosystem services requires a systematic understanding of biodiversity patterns and drivers in informal urban green spaces. The ‘environmental filtering’ (a process of certain species selected by specific environmental conditions) concept in community ecology theory may serve as a useful tool for this goal. We tested a multi-scale filtering framework by examining the spontaneous plant diversity patterns (from 83 surveyed sites) on the vertical surfaces of the ancient city wall of Nanjing, China. We found that the variables representing local-habitat filtering (e.g., wall substrates and aspect) and landscape filtering (including spatial configuration of urban land cover, and nighttime light intensity surrounding the local habitats) can jointly explain substantial fractions of variations in taxonomic diversity (up to ca. 60%) and functional diversity (up to ca. 40%). The explanatory power was stronger in the repaired wall habitats than in the unrepaired counterparts, in line with the prediction that environmental filtering is more pronounced during the early stages of community assembly. While the strength of landscape filtering showed clear scale-dependency, its relative importance consistently outweighs local-habitat filtering across all study scales of 200–1600 m, suggesting that configuration of neighboring landscape context can play an important role in shaping local-scale biodiversity of informal urban green spaces. Our results have useful implications for the study, design, and management of informal urban green spaces. Well-tailored multi-scale filtering frameworks may contribute to understanding urban biodiversity patterns in a systematic way.