Complexity Increases (and Decreases) in Evolutionary Sequences

Do evolution and selection inevitably increase complexity? Most of this debate focusses on biological evolution, origins of life, cosmology, and theoretical physics, but here the emphasis is on evolution of Earth surface systems (ESS)—for which, unlike Earth’s biosphere and the cosmos, N > 1. Using several approaches in entropy, information theory, and algebraic graph theory, this study addresses complexity of the overall pattern or network of historical evolutionary sequences, the variety of outcomes of evolution, and the number of selected possible states relative to the total number of potential states (functional information). Complexity of the networks of historical changes varies greatly, but the lengthening of sequences over time (or their elaboration by additional information) can only increase embedded complexity. The variety of potential outcomes and variability at the landscape scale may undergo convergent, decreasing-complexity or divergent, increasing-complexity evolution, and switches between them. Functional information may also decrease or increase over time. Changes in ESS complexity vary with the type of system considered, spatial and temporal scale, the aspect of complexity considered, and the geographical and historical context. Rather than attempting to make statements about changes in complexity of all evolving systems, it is more appropriate to address changes in complexity case-by-case.