Urban areas are not biological deserts, and what they support is a distinctive subset of species selected by specific pressures.
Urban adapters
Species that thrive in cities share characteristics.
Dietary flexibility, allowing exploitation of varied and human-derived food.
Tolerance of disturbance and of human proximity.
Use of structures as habitat, since buildings resemble cliffs for some species.
And, frequently, high reproductive rates allowing recovery from mortality.
Which selects for a recognisable set of species that appear in cities worldwide, producing biotic homogenisation.
What is excluded
Specialists with narrow requirements.
Species intolerant of disturbance or requiring large areas.
Species sensitive to light or noise.
And ground-nesting species, which face predation from domestic cats and disturbance.
Light
Artificial light at night affects a wide range of species.
Insects attracted to lights are removed from the ecosystem and killed.
Bird migration is disrupted, with collisions at lit buildings killing large numbers.
Bat activity is suppressed for light-averse species and concentrated for others.
Plant phenology is affected near lighting.
Which is addressable — shielding, reducing intensity, using longer wavelengths, and switching off when not needed all reduce effects measurably.
Noise
Interferes with acoustic communication.
Birds in noisy areas have been documented singing at higher frequencies and at different times, which has fitness consequences.
Which is harder to address than light and is affected by traffic management and by surface materials.
Green space quality
Area matters and structure matters more.
Amenity grassland mown short supports very little.
The same area with varied vegetation height, native planting, dead wood, water and undisturbed corners supports substantially more.
Which means management changes on existing green space deliver more than acquiring new space in most cities.
Private gardens
Collectively exceed public green space in area in many cities.
Which makes garden management collectively significant, and studies have documented substantial biodiversity in gardens managed for it.
Ponds are particularly effective, since standing water is scarce in urban areas and supports a disproportionate range of species.
Connectivity in cities
Railway margins, canals, cemeteries, allotments and linked gardens function as corridors.
Which is why urban ecological network mapping has become part of planning in several places.
Barriers — roads, walls, sealed surfaces — fragment these for ground-dwelling species particularly.
Human dimensions
Urban wildlife produces conflict as well as benefit.
Foxes, gulls, pigeons, rodents and deer all generate complaints, and management approaches vary from tolerance to control.
Evidence on culling urban wildlife is generally that it does not produce lasting reduction, since populations are limited by food availability rather than by mortality.
Which points at waste management as the effective intervention for several species.
Why it matters
Most people encounter nature in cities, and that encounter shapes attitudes toward conservation generally.
Which is an argument for urban nature independent of its biodiversity value, and the evidence linking access to green space with wellbeing supports it further.
Buildings as habitat
Several species nest in buildings, and modern construction and renovation exclude them.
Which has been identified as a factor in declines of several urban bird and bat species.
Integrated nest bricks and bat boxes built into new construction address this at negligible cost, and requirements have been introduced in some planning systems.
Renovation of older buildings requires checking for roosts, since disturbing them is an offence for protected species in many jurisdictions.
Green roofs and walls
Provide habitat on structures that would otherwise be sterile.
Which has documented benefits for invertebrates and some birds, with the value depending heavily on substrate depth and planting.
Shallow uniform installations support far less than varied deeper ones.
Recording urban wildlife
Urban areas are well populated with observers and produce disproportionate recording effort.
Which makes urban data unusually rich, and it means participation is genuinely useful.
Feeding wildlife
Bird feeding is widespread and has documented effects in both directions.
It supports individuals through difficult periods and demonstrably affects population distributions.
It also concentrates birds, which spreads disease, and several outbreaks have been linked to feeding stations.
Which means hygiene matters — cleaning feeders regularly and moving them periodically are the recommended practices.
Feeding mammals generally causes more problems than it solves, through habituation and dependence.
Roads
A major mortality source in urban and peri-urban areas.
Which affects some species disproportionately, particularly slow-moving ones and those with large ranges.
Underpasses and culverts designed for wildlife use have documented effectiveness where correctly placed.
Planning policy
Requirements for biodiversity net gain in development have been introduced in some jurisdictions.
Which requires developments to leave measurably more habitat than they removed, assessed with a defined metric.
Implementation quality, long-term maintenance and whether created habitat delivers what the metric assumes are the open questions.
Water
Standing water is scarce in most urban areas and supports a disproportionate range of species where it exists.
Which makes even a very small pond one of the most effective single interventions available in a garden.
Sloping sides allowing animals to exit, no fish, and native marginal planting are the features that matter.