Trends as changes in variance: a new slant on progress and directionality in evolution

Abstract

Trends are the primary phenomenon of macroevolution, but they have often been misinterpreted because an old and deep conceptual error has induced us to misread, as anagenesis in abstracted entities, a pattern that actually records changes in variance by increase or decrease in diversity or disparity among species within clades. These patterns are actually produced by the “entity making and breaking machine” of differential species success, but we misread history as anagenesis because we focus on extreme values (though they may only represent tails in variance of a system), or on measures of central tendency that shift passively as species birth and death work in their differential way. I discuss several examples in two classes: “increase trends” mediated by growth in variance, and “decrease trends” produced by diminution of diversity or disparity. I present two examplesin extensobased on original data: threefold occurrence of Cope's rule in planktonic foraminifers as a consequence of increasing species diversity from small founding lineages (increase trend), and disappearance of 400 hitting in baseball as a decrease trend recording symmetrical decline of variance with increasing excellence of play, not the gradual extinction of a valued “thing.” A proper appreciation of trends as changes in variance flows from and into the two most important revisionary themes in modern evolutionary theory: 1) constraint and structure as an antidote to overreliance upon adaptation (questions about why founding lineages tend to be small, and why size ranges are constrained, lie primarily in structural, not adaptational, domains); 2) hierarchy (increase and decrease trends are powered by differential species sorting, not by extrapolated anagenesis of competition among organisms within populations).