1: \begin{abstract}

2:   The spreading of evolutionary novelties across populations is the

3:   central element of adaptation.  Unless population are well-mixed

4:   (like bacteria in a shaken test tube), the spreading dynamics not

5:   only depends on fitness differences but also on the dispersal

6:   behavior of the species. Spreading at a constant speed is generally

7:   predicted when dispersal is sufficiently short-ranged. However, the

8:   case of long-range dispersal is unresolved: While it is clear that

9:   even rare long-range jumps can lead to a drastic speedup, it has

10:   been difficult to quantify the ensuing stochastic growth

11:   process. Yet such knowledge is indispensable to reveal general laws

12:   for the spread of modern human epidemics, which is greatly

13:   accelerated by aviation.  We present a simple iterative scaling

14:   approximation supported by simulations and rigorous bounds that

15:   accurately predicts evolutionary spread for broad distributions of

16:   long distance dispersal.  In contrast to the exponential laws

17:   predicted by deterministic 'mean-field' approximations, we show that the

18:   asymptotic growth is either according to a power-law or a stretched

19:   exponential, depending on the tails of the dispersal kernel. More

20:   importantly, we provide a full time-dependent description of the

21:   convergence to the asymptotic behavior which can be anomalously slow

22:   and is needed even for long times.  Our results also apply to

23:   spreading dynamics on networks with a spectrum of long-range links

24:   under certain conditions on the probabilities of long distance

25:   travel and are thus relevant for the spread of

26:   epidemics. % Due to the simplicity of our model, which does not depend on specific forms of interactions between ``individuals", some aspects of the results should be relevant to other spreading processes.

27:  %including, potentially, activity in the  brain. 

28:  \end{abstract}

29: