Form and function: structural analysis in evolutionary morphology

-A theoretical approach to the analysis of historical factors (Raup 1972) in evolutionary morphology is presented which addresses transformational hypotheses about structural systems. This (structural) approach to testing historical hypotheses about phylogenetic constraints on form and function and structural and functional versatility involves (I) the reconstruction of nested sets of structural features in monophyletic taxa, (2) the use of general or emergent organizational properties of structural and functional systems (as opposed to uniquely derived morphological features), and (3) the comparative examination of the consequences for structural and functional diversity of these general features in related monophyletic taxa. Three examples of emergent organizational properties are considered: structural complexity, repetition of parts, and the decoupling of primitively constrained systems. Two classes of hypotheses about the evolution of design are proposed. Transformational hypotheses concern historical pathways of change in form as a consequence of general organizational features which are primitive for a lineage. Relational hypotheses involve correlations between structure-function networks primitive for a clade and morphological diversity both between and within terminal taxa. To the extent that transformational and relational hypotheses about form are corroborated, they provide evidence of underlying regularity in the transformation of organic design that may be a consequence of the hierarchical organization of structural and functional patterns in organisms. George V . Lauder. Department of Anatomy, University of Chicago, 1025 East 57th Street, Chicago, Illinois 60637 Accepted: July 28, 1981 T . J. M. Schopf, acting editor for this paper. Introduction called the "historical factor" in morphologic The diversity of structure exhibited by life on analysis (Raup 1972). The constructional apearth is a recurrent theme in natural history, proach to morphology (Seilacher 1970, 1979; and it is the task of the morphologist to analyze Thomas 1979) has served to define the three and order this diversity. Darwin (1859, p. 434) main factors which must be understood in an considered morphology to be the "very soul" of attempt to explain form: fabricational, functionnatural history and cited the elucidation of "unial, and historical. Of these, only the first two ty of type" as the foremost contribution of morhave received much attention (see Gans 1960, phologists to the study of life: "What can be 1974; Hickman 1980; Seilacher 1973; Rudwick more curious than that the hand of a man, 1964, 1968). Historical factors are usually anaformed for grasping, that of a mole for digging, lyzed only in a very general way. Accretionary the leg of the horse, the paddle of the porpoise, growth may appear to constrain the range of and the wing of the bat, should all be constructpossible molluscan forms (McGhee 1980; Raup ed on the same pattern and should include the 1966, 1972) but how can such a hypothesis be same bones, in the same relative positions?" Intested? The increasing emphasis on the themes deed, the theme of unity of plan among organof limits to morphologic change, historical conisms was perhaps the cardinal principle of eightstraints, and ontogenetic constraints on possible eenth and nineteenth century anatomy and structural patterns (e.g., Alberch 1980; Fisher received its most complete expression in the 1981; Gans 1966, 1969; Gould 1980a,b; Gould work of Cuvier, Geoffroy, and Owen. and Lewontin 1979; Raup 1972; Riedl 1978; The concept of unity of type, and its phyloThomas 1976) has not focused on an important genetic implication, falls under what has been question: how can hypotheses of constraint (or 1981 The Paleontological Society. All rights reserved. 0094-8373/81/0704-0005/$1 .oo HISTORICAL ANALYSIS I N MORPHOLOGY 43 1 the converse, versatility) be tested in evolutionary morphology? In this paper I present one approach to testing hypotheses of structural and functional constraint and versatility. Four elements are basic to this analysis: (1) the recognition that hypotheses of morphological constraint and versatility are historical hypotheses and must be approached within a historical framework, (2) the importance of hierarchical organization for understanding patterns of structural and functional change through time, (3) the distinction between intrinsic explanations for form and extrinsic environmental explanations, and (4) the crucial role of corroborated phylogenetic hypotheses in permitting tests of historical hypotheses. Morphologic Analysis For the purpose of discussing the historical approach to form and function, I distinguish two classes of analyses about structural systems: equilibrium analyses (see Lewontin 1969), and transformational analyses. Equilibrium analyses focus on the relationship between the organism and the environment, and may be either (1) an examination of present-day structure environment correlations (e.g., Wiens and Rotenberry 1980) or (2) a consideration of historical patterns of covariation between environment and morphology. Explanations for structural patterns are extrinsic in that factors external to the organism (temperature, resource availability, competition) are held to be the primary determinants of structural change. Transformational hypotheses form a class of questions which have not received much attention from morphologists, and they will be the focus of most of this paper. Historical patterns of structural change are analyzed as a consequence of intrinsic organizational properties of structural systems (Whyte 1965; see below p. 434). This class of hypotheses is not the same as the "transformational approach" discussed by Eldredge (1979a) which utilizes an extrinsic explanatory framework. (I use the term structural to emphasize (1) the central role of nested sets of structural features (cladograms) in the study of form and function, (2) the intrinsic nature of explanations for historical patterns, and (3) the focus on structural transformation (Piaget 1970). This is in contrast to extrinsic explanations for structural change which dominate equilibrium analyses.) Three aspects of equilibrium analysis relate directly to the study of historical hypotheses in morphology: (1) the inference of historical selective forces to explain morphological change, (2) the use of morphological series as a reflection of the actual historical pattern of structural change, and (3) the explanation of adaptive radiations in terms of key innovations. Elucidating the selective forces that have produced structures is a widely claimed goal of evolutionary morphology (e.g., Bock 1960, 1980; Bowman 1961; Gans 1974; Lombard and Wake 1976, 1977; Simpson 1953; Williams 1966; see Cracraft 1981). Bock (1980) claimed that it is necessary to know the exact selection force because adaptation is judged with respect to the selection force, and "the selection force must be known before a feature can be considered to be an adaptation." He suggested that a detailed knowledge of the environment is necessary to specify the selective forces by which adaptations are judged. Correlations between hypothesized environmental (selective) factors and structural features of the organism can then be examined. Historical hypotheses of relationship (phylogenies) may have important consequences for the analysis of structure-environment correlations as illustrated in Fig. 1 (phylogenetic analysis is discussed in more detail below). Interpretation of the distributional pattern of taxa in the structure-environment space (Fig. 1A) depends critically on the phylogenetic hypothesis. Given hypothesis 1 (Fig. lB) , that taxa E and F are most closely related to each other, their proximity in the structure-environment space may only reflect common ancestry. The structural features shared by these taxa may reflect the environment of the common ancestor of E and F , and not the environments presently inhabited (environment four in Fig. 1A). If hypothesis 2 (Fig. 1B) is the pattern of phylogenetic relationship, then taxa E and F are only distantly related and their proximity in the structure-environment space calls for additional explanation. A phylogenetic hypothesis allows the reconstruction of the historical sequence of structural change through time and thus serves as a null hypothesis from which significant deviations may be 432 GEORGE V. LAUDER ;@ c3 1 2 3 4 . ENVIRONMENT