Functional morphology in the pages of the AJPA

Abstract

“Functional morphology,” in its broadest sense, is simply the study of relationships between biological form and its functions. “Function” can include many different aspects of an organism's biological adaptations, but in this context, and as most often used in physical anthropology, I will concentrate on mechanical function. There is a very long history of interest in the mechanical significance of anatomical variation (Galilei, 1638). Engineering principles were applied to the human skeleton by anatomists and orthopedic surgeons in the 1800s, most famously by Wolff (1892), whose name became associated with the concept of bone functional adaptation (i.e., “Wolff's Law”), despite many problems with the original formulation of the concept (Ruff, Holt, & Trinkaus, 2006). Comparative anatomists and anthropologists were aware of this work (Keith, 1918; Morton, 1924). However, the actual application of mechanical principles to the interpretation of extant and fossil humans and other primates was relatively rare during the early history of physical anthropology. Today, functional morphological studies make up a significant and expanding component of research in our field, due to many conceptual and technical advances over the past 100 years, particularly during the past 50 years. In this article, I review these developments, focusing in particular on the role of the AJPA as an outlet for encouraging and presenting new research in this area. The study of morphological variability in humans and other primates and its underlying causes has been a central focus of physical anthropology since its founding as a discipline in the late 1800s (Hrdlička, 1918). In his preface to the first edition of the AJPA, Hrdlička (1918) also prominently featured “functional” relationships (in a physiological sense)—mentioned five times within the first four paragraphs and often paired with “structural”—as a major defining characteristic of the discipline. However, while the terms “functional” and “mechanical” show up sporadically in his later writings, his overall approach to morphological variability was strongly categorical and typological (Hrdlička, 1937). It is interesting that the first paper in the AJPA that could be considered to address a functional morphological issue was from overseas—an experimental study by Matiegka (1921) of the relationship between muscle and bone dimensions and the potential to generate dynamic pressure between the hands. This study reflected a life-long interest of Matiegka in environmental effects on human morphology (Skerlj & Brozek, 1952), and in some ways foreshadows important early experimental studies of bone mechanical adaptation carried out in the Czech Republic 50 years later (Hert, Liskova, & Landrgot, 1969; Liskova & Hert, 1971). Morton's early 1920s publications in the AJPA on comparative foot anatomy and function (Morton, 1922, 1924) are classics in the field and have stimulated much further research (Gebo, 1985) (also see Larson, this volume). These papers also began a string of early publications in the journal on the evolution, growth and development, functional morphology, and kinematics of the foot (Dougan, 1924; McMurrich, 1927; Patek, 1926; Straus, 1926; Weidenreich, 1923), a focus that obviously extends to the present day (Holowka, O'Neill, Thompson, & Demes, 2017; Prang, 2016; Young & Heard-Booth, 2016). As pointed out by Morton (1924), this continuing interest stems in part from the fact that the foot represents the point of contact of the body with the substrate (in bipeds), and thus should reflect, as much or more than any other structure, functional adaptations to changes in locomotion. Morton's original training was in orthopedic surgery, which he noted had stimulated his appreciation for how structural variation could affect function. As noted earlier, in his 1924 paper, he also explicitly discussed “Wolff's Law,” including the relationship between inherited and developmentally modified traits in an evolutionary context. His 1927 comprehensive summary of morphological variation among living and fossil primates is suffused with functional interpretations (Morton, 1927). The first few years of the AJPA also saw the publication of Adolf Schultz's first comparative studies of growth in humans and other primates (Schultz, 1923, 1924), an interest that he maintained throughout his long career. A number of his many comparative studies—ontogenetic and adult—were published in the AJPA (Schultz, 1949, 1953). Although largely descriptive, most of these also included discussions of the functional implications of observed variations, and in some cases were more explicitly mechanical in nature (Schultz, 1953). The latter paper, in fact, cites results of an early study by Evans and Lebow (1952) on material properties of compact bone. The important contributions of F. Gaynor Evans to physical anthropology in the middle portion of the twentieth century are discussed below. Another key functional morphological study published during the earlier years of the AJPA was that by Elftman and Manter (1935), again focusing on the foot. By comparing plantar pressure recordings of chimpanzees and humans during bipedal walking, they demonstrated significant differences in mid-tarsal mobility and weight transfer across the foot, and related this to structural (osteological) differences in the two taxa. Again, as with Morgan's anatomical analyses, this study inspired much later work and discussion regarding this issue, including its evolutionary implications (DeSilva, 2010; Gebo, 1992; Susman, 1983) (also see Larson, this volume). A very thorough survey of anatomical variation in the forelimb of primates and its functional correlates was carried out by Miller during the same decade (Miller, 1932). The early 1940s brought several calls for the incorporation of new approaches into physical anthropology, including more functionally oriented approaches. With the exception of the studies cited above and a few others, up to this time the AJPA had largely concentrated on descriptive and technique-driven investigations, in part likely due to Hrdlicka's influence, and also the (legitimate) need to amass necessary comparative collections and standardize methodology. With Hrdlicka's passing in 1939, a new generation of scholars trained as physical anthropologists, and developments in other fields such as genetics and human growth studies, a shift away from typological thinking and toward more population- and process-oriented work was advocated by at least some portions of the discipline. Goldstein's 1940 AJPA review of “Recent Trends in Physical Anthropology” (Goldstein, 1940) is a landmark in this respect. After reviewing contributions to the AJPA and the journal Human Biology since their inceptions (Human Biology in 1929), he concluded (pp. 192–193) that “An outstanding fact of the data is the consistent predominance of research of an anatomical character in the American Journal of Physical Anthropology, albeit with taxonomic or racial implications. The relative paucity of contributions of a specifically functional or physiological character is also noteworthy.” He also cited heavily from Krogman (1938) and Shapiro (1937), who both called for greater integration of physical anthropology within human biology as a whole, and use of approaches that emphasized the dynamic nature of human adaptation. It is perhaps not surprising that these recommendations came from researchers with major interests in growth and development and environmental effects on morphology. In the same year, Boyd's (1940) “Critique of Methods of Classifying Mankind” in the AJPA emphasized the importance of a population-oriented approach and the use of genetic (blood group) characteristics in defining populations and tracing human evolution. While he did not specifically discuss functional adaptation, his strong critique of the then-current methods that used morphology to classify groups was an implicit criticism of typological thinking. He particularly focused on the developmental plasticity of morphological traits (including Shapiro's (1939) study of Japanese immigrants in Hawaii). Washburn and Detwiler's (1943) paper in the AJPA, provocatively entitled “An Experiment Bearing on the Problems of Physical Anthropology,” is another landmark from this time period. In it, they critique purely comparative approaches and show how experimental developmental studies can shed light on traditional anthropological issues such as the relationship between orbital size and eyeball size (also see below on Moss' 1960 paper). The use of nonprimate models to address such issues, when appropriate, was also advocated. This theme was taken up and expanded by Washburn in his well-known paper “The New Physical Anthropology” (Washburn, 1951), which was not published in the AJPA, although comments in the proceedings of the AAPA annual meeting of that year, recorded in the AJPA (Anon., 1951), make it clear that he was expressing similar views directly to the organization. In addition to supporting the role of population genetics theory and developmental biology in the discipline, he also emphasized the importance of taking a functional approach to explaining morphological variation, in particular the concept of functional complexes, whereby functionally related traits covary and can only be understood within that context. Two other studies published in the AJPA during this general time period with a specifically mechanical focus should also be mentioned. Townsley (1948) studied ontogenetic and general variation among vertebrates in internal structure of the proximal femur in relation to Wolff's trajectorial theory (part of the “Law”). Estel and Asling (1948) carried out an experimental study on human cadavers investigating the mechanical consequences of repositioning the biceps insertion on the radius, with implications for interpretations of Neandertal morphology. Thus, despite Washburn's complaints, developmental and experimental studies relating structure to function were not entirely absent from the field. “There is still another tool available for analyzing the evidence of adaptation, which is a consideration of the functional or biomechanical change that accompanies the morphological change. The functional approach, although it has obvious limitations and is not always applicable, does possess certain merit. It introduces a dynamic factor into the evaluation of adaptive change, thereby decreasing the tendency to regard the trend as a series of more or less isolated morphological stages. It focuses attention on the possibility of continuous adaptation by inducing a more penetrating analysis of the morphological change and by suggesting the nature of unknown stages in the adaptive trend.” (p. 282) He did go on to note that very few attempts had yet been made to apply this approach (the example he gave was the evolution of the horse forelimb), in part because of the difficulty in formulating functional models for many complexes, and the incompleteness of the fossil record for many groups. At the time, the fossil record for primates was no exception, with “many serious deficiencies;” however, “The relative abundance of knowledge on recent primate morphology, plus the fact that this order includes the only vertebrate (man) that has been subjected to a reasonably complete biomechanical analysis, places the primates in a favored position for functional analysis of adaptive trends.” (p. 293) Another relevant development during this time period was the growth in interest in “applied physical anthropology” or “human engineering” (White, 1952). Although not concerned with evolutionary or “adaptive” issues in the usual sense, this work did advance the understanding of relationships between morphology and mechanical function, at least in a human-made environment. The value of training in physical anthropology for “the teaching and interpretation of functional morphology” to clinical scientists was also recognized (Krogman, 1951: 211). Evans' (1953) paper, “Methods of studying the biomechanical significance of bone form,” was a important milestone in that it clearly defined, for the first time in the journal, basic engineering principles and how they could be applied to functional analyses, particularly of long bones. He also reviewed earlier experimental and mathematical modeling studies of long bones, including the early study by Koch (1917) (published in the American Journal of Anatomy, one year before the founding of the AJPA), which was the most comprehensive analysis of this type for several decades. Evans also described the use of experimental techniques for measuring strain in bones, including strain gauges (Figure 1). This included what is apparently the first report in the literature of in-vivo strains measured using a strain gauge attached to a long bone (Figure 1b). The next study utilizing a similar technique would not appear in the AJPA for 24 years, with Hylander's (1977) investigation of in-vivo strains in the Galago mandible. Evans also described the use of the “stresscoat” technique for assessing strains in bones under mechanical loading, although he noted its limitations (only tensile strains have an effect and results are imprecise relative to those obtained with strain gauges). In a later paper, he reemphasized those limitations, noting particularly that the method could not be used to reconstruct “stress trajectories” in bones (Evans & Goff, 1957). A somewhat similar technique using “split-lines” was introduced in the journal in the same year as Evans' original article (Tappen, 1953) and applied to facial structures, with tentative interpretations regarding in-vivo stress distributions. Such interpretations were heavily criticized by Evans and Goff (1957), who argued that results using the technique were more indicative of bone histology and growth patterns and had limited mechanical significance, which led to a series of disputes (Evans, 1965; Tappen, 1964). Later studies directly comparing strain gauge to split-line results have supported Evans' contentions (Bouvier & Hylander, 1981). Both stresscoat and split-line techniques are subject to many limitations (Burr, 1980). Early experimental techniques discussed by Evans in a 1953 AJPA review article. (a) Huggouberger extensometer for measuring strains in the femoral neck and shaft under a mechanical load applied to the femoral head. (b) Oscilloscope record of strains recorded from a gauge attached in vivo to the tibia of a walking dog. Tensile strains are positive, whereas compressive strains are negative. Adapted from “Methods of studying the biomechanical significance of bone form,” by F. G. Evans, 1953, American Journal of Physical Anthropology, 11, 413–434 Scott (1957) gave an extensive review of experimental and other investigations into the relationships between muscle and bone growth, emphasizing the interaction between environmental and genetic factors. These same themes were highlighted three years later in a landmark paper by Moss and Young (1960) entitled “A Functional Approach to Craniology.” Similar to Washburn, they argued that morphology must be understood in the context of functional complexes, and stressed the importance of considering the interaction between soft tissue and bone during growth. Only in this way would the “biological meaning of an increase in cephalic index or a decrease in brow ridging,” for example, be understood (p. 281). The role of mechanical factors in this process was explicitly recognized. The dependence of bony (mandibular) morphology on muscular forces was also shown experimentally in a series of studies reported by Avis (who worked as a student with Washburn) at about the same time (Avis, 1959, 1961). The terms “biomechanics” or “biomechanical” were used in passing in the AJPA articles by Morton (1927) and White (1952) cited earlier, but the first extensive use and discussion of the terms were given in Schaeffer's (1950) and Evans' (1953) reviews. It is perhaps significant that both authors were trained outside the field of physical anthropology, that is, in zoology and vertebrate paleontology, which may have helped foster their interests in comparative mechanical approaches. Several developments in the late 1960s and 1970s combined to give new impetus to biomechanics as a field and stimulate applications within physical anthropology. The maturing of the field was reflected in the founding of the Journal of Biomechanics (co-edited by Evans) in 1968, and the appearance of textbooks such as Orthopaedic Biomechanics (Frankel & Burstein, 1970), which had a direct influence on physical anthropologists (Heiple & Lovejoy, 1971; Lovejoy, Heiple, & Burstein, 1973; also see below). Harold Frost's first paper in the AJPA, in 1968, described new methods for recording and interpreting the cellular dynamics of bone modeling and remodeling (Frost, 1968). A number of experimental investigations of bone functional adaptation to mechanical loading were carried out during this period as well (summarized in Ruff et al., 2006), including two studies published in the AJPA (Saville & Smith, 1966; Smith & Saville, 1966) (see Meade, 1989 for a comprehensive review). Improved techniques for the use of strain gauges in-vivo were developed (Lanyon & Smith, 1970). Electromyography as a technique became better known (Basmajian, 1967) and was first applied in an anthropological context (Tuttle, Basmajian, Regenos, & Shine, 1972; also see below). Early gait analyses of primates using cinematography appeared in the AJPA during this time period (Prost, 1965; Prost & Sussman, 1969). One of the first applications of cineradiography to a study of primate locomotion also appeared in the AJPA (Jouffroy, Gasc, & Oblin, 1973) (for further discussion of the history of primate locomotor studies in the AJPA, see Larson, this volume). In a series of papers published in the AJPA in the early 1970s (Heiple & Lovejoy, 1971; Lovejoy & Heiple, 1970; Lovejoy et al., 1973), Lovejoy et al. used orthopedic biomechanics principles to reconstruct and interpret lower limb anatomy in australopiths. Their 1973 paper was particularly important in that it led to a wide reassessment of the functional significance of structural variation among early hominins and stimulated new types of analyses (McHenry, 1975a, 1975b) (Figure 2). In 1976, the same research group reported in the AJPA one of the first applications of beam theory to the analysis of long bone structure in archaeological or human paleontological remains (Lovejoy, Burstein, & Heiple, 1976) (earlier analyses had been presented elsewhere by Endo & Kimura, 1970; and Kimura, 1971). This type of analysis had been explored almost 50 years earlier (Koch, 1917; also see papers republished in Pauwels, 1980), but had been largely ignored, probably in part because of the difficulty in calculating properties manually. The subsequent development of automated techniques as well as the increasing availability of computed tomography (Jungers & Minns, 1979; Nagurka & Hayes, 1980; Ruff & Leo, 1986; Sumner, Mockbee, & Morse, 1985; Tate & Cann, 1982) led to rapid expansion of this approach, including application to fossils, archaeological material, and extant comparative samples (Burr, Piotrowski, & Miller, 1981; Jungers & Minns, 1979; Lovejoy & Trinkaus, 1980; Ruff & Hayes, 1983a, 1983b; Ruff, Larsen, & Hayes, 1984). Hylander's (1975) analysis of human mandibular mechanics is another noteworthy early example of the use of beam theory to address anthropological issues. Moss' 1960 AJPA publication discussed above is cited by Hylander as an inspiration for some of these analyses. Forces contributing to mechanical load on the femoral head in a reconstruction of the Australopithecus africanus Sts 14 presented in the AJPA by Lovejoy et al. (1973). C is body weight, A is the gluteal abductor force, and B is the hip joint reaction force (reprinted with permission of C.O. Lovejoy) Advances in measuring bone tissue mechanical properties—density, strength, stiffness, etc.—were also rapid in the late 1960s and 1970s Currey, 1970; Evans, 1973; Reilly & Burstein, 1974; Yamada, 1970). Applications within physical anthropology were relatively rare (in part because of the destructive nature of most of the tests), but some early comparative studies among primates were reported in the AJPA (Burr, 1979a, 1979b). The first use of photon absorptiometry, an early technique for noninvasively assessing bone mineral content, in an archaeological sample appeared in the AJPA (Perzigian, 1973). Tuttle et al. (1972) reported the first use of EMG in a great ape in the AJPA, using the technique to explore forearm muscle activity during knuckle-walking in a juvenile gorilla. This began a series of publications in the AJPA by this group (Tuttle & Basmajian, 1974, 1978a, 1978b; Tuttle, Basmajian, & Ishida, 1978; Tuttle, Velte, & Basmajian, 1983) and others (Larson & Stern, 1985; Stern & Susman, 1981; Susman & Stern, 1979a, 1979b) employing EMG to study a range of issues in primate locomotion (see Larson, this volume). Hylander's (1977) use of strain gauges to study mandibular strains in-vivo was, as noted earlier, the first attempt to utilize this technique reported in the AJPA since Evans' preliminary report in 1953, and the first application to a nonhuman primate. Again this began a series of papers in the journal that employed this technique to investigate patterns of bone strain in the primate mandible (Hylander, 1979, 1984; Hylander, Johnson, & Crompton, 1987), with implications for interpreting early hominins (Hylander, 1988). Another influential study of craniofacial functional morphology published in the AJPA during this time period was Carlson and Van Gerven's (1977) analysis of temporal changes in Mesolithic and Neolithic Nubian crania in relation to masticatory stress. Their hypothesis that observed changes in cranial form were related to a shift to a less biomechanically demanding diet has been further explored in a number of more recent papers published in the journal (Paschetta et al., 2010; Pinhasi, Eshed, & Shaw, 2008; also see below). During the same time period that biomechanical theory and techniques were becoming better known and applied in physical anthropology, advanced numerical techniques for analyzing morphology were also developing rapidly within the field. The use of multivariate statistics for analyzing shape variation within and between populations and species became more commonplace (Howells, 1969; Oxnard, 1968). Multivariate techniques had been described in the AJPA many years previously (Bronowski & Long, 1952), but their practical application, as with many of the new methods used in biomechanics, was greatly facilitated by the development and increasing availability of digital computers. While many early studies utilizing multivariate statistics within physical anthropology were concerned mainly with classification and population (or sex) sorting (Hanna, 1962; McHenry & Giles, 1971; Rightmire, 1970), a number of attempts to apply them to analyses of functional complexes were also carried out, several of which were published in the AJPA (Andrews & Williams, 1973; Oxnard, 1967, 1969). Multivariate statistics are part of a broader set of quantitative morphometric techniques. The term “morphometrics” was coined in 1965 (Blackrith, 1965; also see Blackrith & Reyment, 1971) and came to include many different approaches to the study of biological form. It is not the intent here to review all of these (for an historical review through the end of the twentieth century, see Richtsmeier, DeLeon, & Lele, 2002) (also see Richtsmeier, this volume). Rather, I focus on one particular technique - finite element analysis (FEA)—that illustrates an interesting cross-fertilization between the fields of biomechanics and morphometrics. Finite-element analysis was developed in the 1950s and 1960s within engineering as an advanced method of structural analysis (Zienkiewicz & Cheung, 1967). It was first applied to skeletal material in 1972 (Brekelmans, Poort, & Slooff, 1972; Rybicki, Simonen, & Weis, 1972). Because it has certain advantages over traditional beam theory, including applicability to irregularly shaped structures and ability to incorporate more complex variation in material properties and loading conditions, the use of FEA in orthopedic biomechanics quickly expanded (Huiskes & Chao, 1983). In 1980, building on previous work of their own engineering team, Lewis and coworkers described an application of FEA (termed “finite element scaling”) to the analysis of form differences between two human femora (Lewis, Lew, & Zimmerman, 1980). In effect, this inverted the FEA procedure used in engineering analysis, which involves applying a mechanical force to a structure and observing the resulting deformations of elements. In contrast, in morphometric applications two forms are compared, and the “force” required to deform one structure into the other is determined (“force” here could refer to a growth trajectory, phylogenetic trend, or other underlying factor). Thus, while the basic underlying methodology is similar, the aims are reversed. The gradual maturation of the FEA technique within engineering and then orthopedic biomechanics was a prerequisite to its application within morphometrics. Finite-element scaling was first applied in an anthropological context by Cheverud and coworkers (including Lewis) in a paper published in the AJPA in 1983 (Cheverud, Lewis, Bachrach, & Lew, 1983), in which they compared cranial form in a juvenile and adult macaque. The technique was presented as a way to incorporate differences in form occurring throughout the cranium, viewed as a continuum, thus quantifying overall deformations that previously had only been qualitatively described (Thompson, 1917). The next applications of finite element scaling published in the AJPA also involved craniofacial morphology, in this case differences between normal children and those with congenital syndromes (Bookstein, 1987; Richtsmeier, 1987). Paradoxically given the origins of the method, the first application of FEA to a mechanical issue in the AJPA did not appear until later, almost 10 years after it had first been used in a morphometric context (Korioth, Romilly, & Hannam, 1992). In part, this may be due to some additional complexities inherent in applying the technique for mechanical modeling (Rayfield, 2007). The process has been streamlined to some extent in recent years and the use of FEA in mechanically oriented analyses has increased in physical anthropology, as discussed in the next section. An increasing incorporation of mechanical principles into broader comparative biological studies during this time period also had an influence on their use in physical anthropology. The book Mechanical Design in Organisms, originally published in 1976 and republished in 1982 (Wainwright, Biggs, Currey, & Gosline, 1982), was the most comprehensive survey to date of material, microstructural, and macrostructural properties as applied in biology. One of the authors later extracted and expanded much of the material on bones specifically (together with relevant soft tissue, e.g., cartilage and ligaments) into a separate text, The Mechanical Adaptations of Bones (Currey, 1984), later revised as Bones: Structure and Mechanics (Currey, 2002). The former text was reviewed in the AJPA, where it was noted that “the comparative and zoological perspectives that Currey brings to this book pique interest and stimulate thought about general properties of the skeletal system - properties that cut across taxonomic and behavioral boundaries” (Burr, 1985). McNeill Alexander's many comparative studies of animal locomotion (e.g., Alexander, 1968) and other aspects of mechanical design were also very influential. The compilation of comparative studies in Functional Vertebrate Morphology (Hildebrand, Bramble, Liem, & Wake, 1985) focused mainly on analyses of animal movement, but also included summaries of allometric principles (Alexander, 1985) and experimental studies of bone adaptation (Lanyon & Rubin, 1985). Broader comparative studies of allometry (e.g., Gould, 1966) also stimulated much interest among physical anthropologists, especially in paleontology (Pilbeam & Gould, 1974). It was increasingly recognized that many allometric problems are mechanically based, that is, that changes in proportions with a change in size are related to mechanical requirements (Alexander, Jayes, Maloiy, & Wathuta, 1979; McMahon, 1973). This general concept was incorporated into a number of papers appearing in the AJPA during this time period (Jungers, 1978; Ruff, 1984; Steudel, 1982; Sumner, 1984). The recognition of the importance of allometry in mechanical (as well as other) analyses also stimulated increased interest in accurately estimating body mass in past hominins (McHenry, 1976, 1991, 1992; Steudel, 1980). One of the most actively growing areas within functional morphology over the past three decades has been in the microstructural analysis of bone, particularly trabecular bone. Earlier comparative studies of trabecular bone structure had been limited by the need for destructive sampling, as well as manual data collection techniques (Mielke, Armelagos, & VanGerven, 1972). The development of micro-CT, first applied to skeletal material in the late 1980s (Feldkamp, Goldstein, Parfitt, Jesion, & Kleerekoper, 1989; Layton et al., 1988), greatly facilitated such analyses because it allowed nondestructive and relatively automated measurement of structural parameters (for recent reviews, see Bouxsein et al., 2010; Kivell, 2016). The first application of micro-CT in a comparative study of primate trabecular bone (in the femoral and humeral heads) was reported in the AJPA in 2001 (Fajard