Many biological processes, from cellular metabolism to population dynamics,
are characterized by allometric scaling (power-law) relationships between
size and rate(1-10). An outstanding question is whether typical allometric
scaling relationships-the power-law dependence of a biological rate on body
mass-can be understood by considering the general features of branching ne
tworks serving a particular volume. Distributed networks in nature stern fr
om the need for effective connectivity(11), and occur both in biological sy
stems such as cardiovascular and respiratory networks(1-8) and plant vascul
ar and root systems(1,9,10), and in inanimate systems such as the drainage
network of river basins(12), Here we derive a general relationship between
size and flow rates in arbitrary networks with local connectivity. Our theo
ry accounts in a general way for the quarter-power allometric scaling of li
ving organisms(1-10), recently derived(8) under specific assumptions for pa
rticular network geometries. It also predicts scaling relations applicable
to all efficient transportation networks, which we verify from observationa
l data on the river drainage basins. Allometric scaling is therefore shown
to originate from the general features of networks irrespective of dynamica
l or geometric assumptions.