Protecting habitat is not simply a matter of total area. How the area is arranged determines what it can support, and the effects of fragmentation are substantial.
Edge effects
The boundary between habitat types has different conditions from the interior.
More light, more wind, different temperature and humidity, and greater exposure to species from the adjacent habitat.
Which means a fragment consists largely of edge, and interior-dependent species have nowhere to live even where the total area seems adequate.
The effect penetrates a considerable distance, which means small fragments can be entirely edge.
Population viability
Small isolated populations face several compounding risks.
Random variation in births and deaths can eliminate a small population by chance alone.
Environmental variation — a bad year, a disease outbreak, a fire — affects an entire small population simultaneously.
And genetic effects accumulate, with inbreeding reducing fitness and reduced variation limiting adaptation.
Which means a population below a certain size is at risk regardless of the habitat quality.
The rescue effect
Connected populations exchange individuals, which addresses all three risks.
Immigration replenishes numbers after local declines.
Gene flow maintains genetic variation.
And recolonisation restores populations after local extinction.
Which means a network of connected fragments functions very differently from the same fragments isolated.
Corridors
Strips of habitat connecting larger areas.
Evidence that they increase movement between patches is reasonably strong from experimental studies.
The design questions — width, habitat quality, whether they must be continuous — are more contested.
Criticisms include that corridors can also transmit disease and invasive species, and that the resources might achieve more if spent on habitat area.
Which is a genuine debate rather than a settled question, and the answer varies by species and context.
Crossing structures
Overpasses and underpasses allowing wildlife to cross roads.
Monitoring has demonstrated substantial use by a range of species and measurable reductions in vehicle collisions.
Which is among the better-evidenced interventions in this area, with cost-benefit analyses that include the human safety benefit.
Design matters — different species require different structures, and fencing directing animals toward crossings is generally necessary for effectiveness.
The matrix
What lies between habitat patches determines whether animals can move at all.
Which means agricultural practices, urban form and land use in the surrounding area affect connectivity as much as formal corridors do.
Wildlife-friendly farming, hedgerows, field margins and permeable urban landscapes all contribute, and they cover far more area than protected sites.
Planning at scale
Conservation planning has moved toward landscape-scale approaches rather than site-by-site protection.
Which requires coordination across landowners and jurisdictions, and it is administratively considerably harder than designating a reserve.
Where it has been done, the results have generally supported the approach.
Climate adds urgency
Species ranges are shifting in response to changing conditions.
Which means connectivity is required not only to maintain populations but to allow them to move as their suitable climate does.
A protected area that becomes climatically unsuitable protects nothing if the species cannot reach the area that has become suitable.
Measuring connectivity
Structural connectivity describes the physical arrangement of habitat.
Functional connectivity describes whether organisms actually move, which depends on the species.
Which means a landscape can be structurally connected and functionally isolated for a species unwilling to cross the intervening habitat.
Genetic analysis measures functional connectivity directly, by assessing whether populations are exchanging genes, and it has become the standard evidence.
Stepping stones
Small patches between larger ones, which can facilitate movement without forming a continuous corridor.
Which is frequently more achievable in developed landscapes than continuous corridors and has evidence supporting it for mobile species.
Ponds, hedgerow trees, small woodlands and urban green spaces all function this way.
Rivers
Natural corridors that have been extensively fragmented by barriers.
Which blocks migratory fish entirely, and barrier removal has produced rapid documented recovery in several rivers.
Fish passes provide partial mitigation with variable effectiveness depending on design and species.
Urban connectivity
Cities are not uniformly hostile, and green corridors within them support movement for some species.
Railway margins, canal banks, cemeteries and linked gardens all function as connective habitat.
Which is why urban planning increasingly incorporates green infrastructure networks rather than isolated parks.
Legal mechanisms
Connectivity is harder to protect legally than sites, since it involves land that is not itself notable.
Which is why some jurisdictions have introduced ecological network designations and requirements for development to demonstrate net gain.
Implementation quality varies considerably and the direction is established.
Small actions that connect
Gaps at the base of garden fences allow movement of ground-dwelling mammals through otherwise sealed suburban blocks.
Which is a coordinated action requiring neighbours, and schemes organising exactly this have been established in several places.
Ponds, native planting and reduced lighting all extend the usable habitat within a connected network.
Which means connectivity is one of the few conservation objectives that individuals can contribute to directly on land they control.