Habitat Connectivity Index Calculator

Connectivity analysis treats a fragmented landscape as a network: each habitat patch is a node and each feasible movement route is a link. With N patches the maximum possible number of undirected links is N(N-1)/2, so 12 patches allow 66 links, and dividing observed links by that maximum gives a connectance value between 0 and 1. Structure alone is not enough, because whether an organism crosses a gap depends on its dispersal ability. That is handled by a negative exponential kernel, P = exp(-d / alpha), where d is the mean edge-to-edge distance and alpha the species mean dispersal distance. When the gap equals the dispersal distance the crossing probability is 0.368; at half that distance it is 0.607, and at three times it falls to 0.050. Multiplying structural connectance by this probability yields the composite index reported here. Realistic alpha values span orders of magnitude, from under 200 m for amphibians and small mammals to several kilometres for woodland passerines and tens of kilometres for large carnivores, so the same landscape can be well connected for one species and impassable for another. Three limitations matter. Connectance falls mechanically as patch count rises, because real networks never approach N(N-1)/2 links, making comparisons between landscapes of different patch counts unreliable. The mesh figure here is simply mean patch area times patch count, that is total habitat area, not the effective mesh size defined by Jaeger as the sum of squared patch areas divided by total area. And the model ignores matrix resistance, treating a motorway and a hedgerow as equivalent gaps.

The conceptual groundwork was laid by Robert MacArthur and Edward O. Wilson, whose 1967 theory of island biogeography linked species number to island area and isolation, and by Richard Levins, who introduced the metapopulation idea in 1969 as a set of local populations linked by colonisation and extinction. Gray Merriam coined the term landscape connectivity in 1984 at a European landscape ecology congress, and Philip Taylor, Lenore Fahrig, Kringen Henein and Merriam gave it its standard definition in Oikos in 1993 as the degree to which the landscape facilitates or impedes movement among resource patches. Ilkka Hanski turned the idea into quantitative prediction with his incidence function model in 1994, calibrated on Glanville fritillary butterflies in the Aland Islands of Finland. The graph-theoretic formulation used here dates to Dean Urban and Timothy Keitt, whose 2001 paper in Ecology showed that decades of network mathematics could be applied directly to habitat mosaics and could identify which individual patches were critical stepping stones. Santiago Saura and Lucia Pascual-Hortal introduced the integral index of connectivity and the probability of connectivity index in 2006 and 2007, distributed through the free Conefor software, which became standard in reserve planning. Brad McRae proposed circuit theory as an alternative in 2006, modelling the landscape as a resistive surface where current density represents movement probability across all possible routes at once; the resulting Circuitscape software has since been used for continental corridor mapping and road crossing placement.