Multifractality in the Tree of Life: A Branching-Process RIFS Proof
Kevin Hudnall
TL;DR
The paper addresses why the tree of life exhibits multifractal scaling by introducing a branching-process random iterated function system (RIFS) that couples Galton–Watson branching with recursive contractions. It develops a rigorous multifractal formalism for this genealogically dynamic fractal, proving the existence and properties of an $L^q$-spectrum $ au(q)$ and a Legendre-derived spectrum $f( obreak obreak ( obreak))$ under mild non-degeneracy assumptions. It analyzes two variants: a non-anchored case with a Cantor-like attractor and an anchored case where the invariant set collapses to a point, yet tangent measures recover the same multifractal law, highlighting mortality as a biologically natural constraint. The results provide a principled explanation for the ubiquity of multifractal signatures in biological data and extend the multiplicative-cascade framework to a genealogically recursive setting, with potential broad implications for modeling evolution and scaling in biological systems.
Abstract
We study a branching-process random iterated function system (RIFS) that formalizes the foundational principles of nestedness, duality, and randomness in the living tree of life (Hudnall & D'Souza, 2025). In this construction, each leaf of a branching process generates a subtree at a strictly smaller contraction scale, thereby unifying classical branching processes and random IFS theory in a single framework. We prove rigorously that this branching-process RIFS is multifractal under explicit, mild assumptions. Two variants are analyzed: a non-anchored case with a nontrivial compact attractor, and a biologically motivated anchored case in which the invariant set collapses to a point while tangent measures obey the same multifractal law. Thus, multifractality emerges as a necessary mathematical consequence of nestedness, duality, and randomness, yielding a minimal-condition theorem that explains the ubiquity of multifractal signatures in biological data.
