Changing perspectives on early hominin diets

Abstract

It is axiomatic that knowledge of the diets of extinct hominin species is central to any understanding of their ecology and our evolution. The importance of diet in the paleontological realm has led to the employment of multiple approaches in its elucidation. Some of these have deep historical roots, while others are dependent upon more recent technical and methodological advances. Historically, studies of tooth size, shape, and structure have been the gold standard for reconstructing diet. They focus on species-level adaptations, and as such, they can set theoretical brackets for dietary capabilities within the context of specific evolutionary moments. Other methods (e.g., analyses of dental calculus, biogeochemistry, and dental microwear) have only been developed within the past few decades, but are now beginning to yield evidence of the actual foods consumed by individuals represented by fossil remains. Here we begin by looking at these more “direct” forms of evidence of diet before showing that, when used in conjunction with other techniques, these “multi-proxy” approaches can raise questions about traditional interpretations of early hominin diets and change the nature of paleobiological interpretations.

Public interest in paleoanthropology has always been fueled by new discoveries which have, in turn, spawned everything from minor insights to major changes in perspective. Among the latter are changes that might even be described as paradigm shifts within the field. By this, we do not mean changes in “the entire constellation of beliefs, values, and techniques shared by a majority of the members of the [paleoanthropological] community” (1) (Chamberlain and Hartwig 1999: 42), but instead revolutionary changes in specific “concrete problem solutions” within the field (2) (Kuhn, 1970: 187). While applying such a label to specific discoveries would undoubtedly generate its own discussions and debates, some possible candidates in paleoanthropology might include, for example, the fossil demonstration of the primacy of bipedalism as opposed to brain size in our earliest ancestors (3), initial radiometric dating of fossil sites in Africa and its impact on perspectives on the timing and pace of human evolution (4), and the onset of cladistic systematics (5). This is not to say that other discoveries might not also warrant such a label, but simply that, with the benefit of hindsight, each of these has caused major adjustments in paleoanthropological perspectives.

With that said, another area that’s proven of great interest is the diet of our ancestors. This is not surprising given the importance of diet in the origin, evolution, and behavior of our species—for example, the constraints it imposes on characteristics ranging from body size and life history strategies to geographic range, habitat choice, and social behavior. However, despite the effort and ingenuity that have gone into paleodietary analyses, results have, until recently, been limited by the fact that the diets of specific prehistoric individuals could never be documented, in effect leaving discussions focused on groups adapted to specific diets. This has now begun to change—to the point where we need to raise the possibility that we are now in the midst of another paradigm shift—a change of perspective that is setting off a cascade effect built upon data from multiple approaches which are impacting other approaches that will have widespread, long-term implications across paleoanthropology.

A number of different avenues of research have been used to shed light on prehistoric diets. The most common of these have focused on the fossils themselves, with traditional morphological studies generally relying on evolutionary arguments to make inferences about past functions—e.g., variations in tooth morphology being used to make inferences about the dietary capabilities of past species (6–8). Researchers have also made use of technological advances in the past few decades to develop new ways to quantify morphological similarities and differences (e.g., dental topographic analyses), yielding new arguments for the presence of specific behavior patterns in the past and their potential implications for paleodietary reconstructions (9, 10). These methods certainly give us broader perspectives—for instance, insights into species-level responses to changes in food availability associated with longer-term environmental change (11). However, by themselves, these approaches still leave us one step removed from finer intricacies of diet—with inferences of capabilities rather than evidence of performance—as none of these techniques can determine what specific individuals ate during their lifetimes. As noted elsewhere, if we rely solely on inferences of capabilities, then we run the risk of feeling so sure about what reality is that we won’t consider other possibilities (12). So, as is always the case in complex paleobiological analyses, we need to rely on every piece of evidence, every so-called dietary proxy, to gain as much insight as possible. Only with such data can we hope to meet Simpson’s challenge to consider fossil primates not as broken bits of bone and teeth, but as “flesh and blood beings” alive in the past (13).

Over the past few decades, technological advances have added immeasurably to this tool kit by beginning to yield direct evidence of past behaviors of individuals. This evidence falls into the realm of what might be called “foodprints”—traces of foods eaten by individuals preserved on or in their teeth (14). The techniques that have begun to document this evidence in fossils are analyses of phytoliths and genetic biomarkers adherent to teeth, the abrasive impact of foods and/or abrasives on teeth, and the biochemical impact of diet on the isotopic composition of tooth enamel. Each method has its own strengths and limitations, and, as a result, they can all benefit from insights generated by one another. Each, in turn, can also be used in conjunction with species-level, adaptation-scale approaches to further refine interpretations and broaden perspectives. Of course, not all approaches are equally relevant to all portions of the temporal scale of the hominin evolutionary record. Thus, some approaches, such as analysis of variation in the APOE gene, may help to document shifts in hominin diets that likely transpired after the hominin–panin divergence (15), while others, such as the retrieval of aDNA from dental calculus, are extremely exciting but may only speak to the last 40 to 50 ka under particular geographic/temperature restrictions (16, 17).

Here we provide a brief review of the approaches that have been used to provide direct evidence of the diets consumed by individual extinct hominins (so-called foodprints), i.e., analyses of 1) phytoliths and genetic biomarkers incorporated into the dental calculus of fossil teeth, 2) dental microwear analyses, and 3) stable isotopes found in fossil tooth enamel. In each case, we review recent developments that may shed light on broader aspects of early hominin paleoecology, for instance, the potential to detect seasonal and/or interannual shifts in diet, dietary fluctuations in response to fine-scale climatic shifts, habitat differences, and possible niche partitioning among extinct taxa. This is then followed by a discussion of how these results might be used with other dietary proxies to raise questions about traditional interpretations of early hominin diets and about broader topics like dental functional morphology and mechanisms of hominin evolution.

Dental Calculus

While much of the archeological record until some 6,000 y ago is concerned with diet, much of the evidence for the plant remains that were dietary staples is missing from the earlier part of the archeological record, with reasonable evidence of potential plant foods (cooked starchy rhizomes) being manifest only in the later part of the Pleistocene (ca. 110 ka) at Middle Stone Age sites such as Border Cave and Klasies River, South Africa (18). Indeed, as aptly observed by Isaac (19) (1971: 280) a half-century ago, “the archaeological study of Pleistocene diet is a little like navigating in the vicinity of an iceberg: more than four-fifths of what is of interest is not visible.”

Recent analyses of food remains embedded in the calculus on fossil teeth have begun to change this situation by giving researchers a glimpse of items brought into contact with teeth eons ago. Dental calculus results from the calcification of the bacterial component of dental plaque that forms on enamel and dentin surfaces and enables intraoral food particles to become embedded in surface pockets and covered with salivary proteins and minerals. As a result, intraoral food particles may be incorporated in its matrix during formation. These may include plant “microfossils” like starch granules and the siliceous bodies (phytoliths) that are formed by many plants, but also proteins indicative of the presence of other tissues (such as muscle versus milk) (20) or even ancient DNA yielding clues into the oral microbiome of extinct taxa (21). Because the mineralization of plaque occurs only in the presence of saliva (22), these types of materials embedded in calculus are contemporaneous bodies. Thus, they have the potential to provide a direct record of what is happening in the mouth. While these may include some non-dietary items and/or items incorporated through non-nutritive activities (23, 24), and the identification and/or extraction of these remains may prove challenging (25, 26), they should also contain remains from foods that were consumed, thus providing a potentially valuable source of information on prehistoric diets (24, 27, 28).

Although dental calculus may survive on fossils for millions of years in certain contexts (29), this is the exception rather than the rule, as it is often not present on teeth, or it has been removed taphonomically or during specimen cleaning and preparation. Or it may simply not be present in large enough quantities for some analyses (e.g., protein analysis). Still, if calculus is preserved, the evidence embedded in it will remain useful for different periods of time, with phytoliths perhaps surviving the longest (30), and ancient DNA the shortest (31). With reference to the early hominin record, phytoliths recovered from the teeth of the juvenile (MH 1) skeleton of Australopithecus sediba included those from both monocot and dicot plants (32). Interestingly, this assemblage differs considerably from that recovered from the sediments in which the hominin skeleton was buried, and the diversity of phytolith types indicates that this individual had a varied diet, including bark and woody tissues. Not surprisingly, analyses of more recent human fossils have yielded even finer-resolution results, including insights into the relative degree of meat-eating in different groups (16) and microbial coadaptations to the diets of our genus (21) through analysis of aDNA. These are merely glimpses of the possibilities that may ultimately be realized from the information stored in dental calculus, for the methods of data collection and analysis are changing at a rapid pace. Clearly none of these findings could be documented via indirect means such as functional morphology.

Dental Microwear

Because teeth come into direct contact with food, the wear they exhibit may provide one of the more direct means to ascertain the types of items that were masticated. Moreover, as teeth are the most commonly preserved elements in the hominin fossil record, this may provide a comparatively abundant form of direct evidence. Tooth wear has been considered at different scales of resolution, ranging from macroscopic to microscopic, and while each may convey some information pertinent to diet, dental microwear has the potential to yield information at a unique level of resolution.

Initial studies used qualitative assessments of tooth surfaces observed through light microscopy (33) or scanning electron microscopy (34). Subsequent quantitative analyses of scanning electron microscopic images yielded insights into differences among taxa with broad dietary differences (35) and also prehistoric species where dietary differences might have been suspected, but never proven (36, 37). However, these analyses were time-consuming and ultimately dependent on the identification of individual microwear “features” (scratches and pits), which resulted in intra- and inter-observer measurement error (38). This led to the development of automated methods of characterizing entire wear surfaces, i.e., what has now come to be known as dental microwear texture analysis (DMTA) (39, 40). DTMA has now been used on many extant and extinct clades, and it can not only differentiate among taxa with broad dietary differences (41–44), but also detect far subtler differences such as seasonal or interannual shifts in diet (45, 46).

Of course, inferences about historic and prehistoric diets require an understanding of how microscopic wear occurs, and this has been the subject of discussion and debate throughout the past 3 to 4 decades (47–50). Often, these discussions have presented microwear causation as an either-or proposition—either it’s caused by food or it’s caused by grit (51–55). As with any complex process, the truth lies somewhere in between these extremes. In other words, nanoscale deterministic models of tooth wear, like those proposed by van Casteren et al. (53), are limited by their assumptions (e.g., are the applied loads appropriate, and what are the effects of scaling on tribological processes?). Moreover, controlled-feeding experiments of groups of animals collected at specific points in time (54–56) do not follow microscopic wear processes day-to-day in their subjects, so we still do not know how the observed patterns developed. Despite these limitations, most studies of experimental animals have concluded that while exogenous mineral abrasives can certainly create dental microwear, and may be a confounding variable in interpreting dental microwear patterns, they do not overwhelm signals related to the processing of foods of different material properties (56–59). In fact, recent work seems to be leading toward a more viable vision; i.e., dental microwear is formed by a dynamic, complex process influenced by many factors including food material properties, grit load, masticatory morphology, and oral processing behaviors (56–60). In other words, it isn’t simple, and (as with all diet proxies) every bit of evidence helps.

Certainly, paleontological samples add another potential complication to the interpretation of dental microwear and that concerns the lifespan of the observed wear patterns—both in terms of day-to-day life and on a taphonomic scale. After all, dental microwear textures are formed by microscopic scratches and pits that are more ephemeral than gross wear or large-scale morphological features like cusp height and molar shearing crest length. In truth, microwear features are constantly being added and then removed from tooth surfaces, and the temporal scale of those changes depends on the types of foods and abrasives being processed (45, 61–63). For instance, under some circumstances, new features can appear within a single feeding bout (64), or in multiple days (65). That doesn’t mean that overall texture patterns will change that quickly, but it does mean that dental microwear has the potential to preserve information recorded within a relatively narrow time window. This can be a mixed blessing. On the one hand, it can certainly give insights into the diets of past individuals and shed light on, for instance, seasonal changes in diet (45), or differences in occupation time within and between archeological sites (66). On the other hand, if dental microwear only preserves dietary information for the last few days or weeks of life (a phenomenon referred to as the “Last Supper Effect”) (36), species-level inferences will certainly benefit from the use of large sample sizes.

By itself, dental microwear has been used to make inferences about the diets of a number of Pliocene and early Pleistocene hominin taxa, ranging from Ardipithecus ramidus to Homo erectus. With the exception of A. ramidus, for which there is only a qualitative assessment of scanning electron microscope images, all of the microwear evidence mentioned here derives from DMTA data generated for over 150 permanent cheek teeth that are attributable to at least eight different early hominin taxa, with the bulk of these teeth belong to Australopithecus africanus and Paranthropus robustus. However, sample sizes for Australopithecus anamensis and A. sediba are quite small (3 and 2, respectively), and we know of no published occlusal data yet for Paranthropus aethiopicus. Still, general results across the taxa for which we have data indicate the consumption of softer, tough, and perhaps fibrous foods, with scant evidence for the consumption of hard objects. The exceptions to this are one specimen of A. sediba (32), some from H. erectus (67, 68), and a number (but not all) of the P. robustus individuals (40, 69). This pattern of inter-taxa differences could not be inferred from standard morphological analyses of teeth. However, further insights from early hominins have demonstrated additional potential for DMTA. For instance, the constancy of dental microwear across time and space in Australopithecus afarensis molars suggests either that this species was able to track its preferred dietary resources in the face of changing habitats and environments, or that environmentally induced shifts in diet did not involve changes in mechanical properties of the foods typically consumed (70). In A. africanus, contra suggestions that the premolars of A. africanus were employed to crack open hard objects, while their molars were employed in the mastication of the softer kernels (71), the microwear textures of its premolars and molars do not differ from one another (72). In P. robustus, despite its higher average complexity and lower anisotropy compared with A. africanus, there is significantly more individual variation in complexity, suggesting possible similarities with extant primates, such as Sapajus apella, that have variable diets (40, 69). By contrast, the microwear fabrics of Paranthropus boisei teeth are dominated by fine striae, not the large, deep pits expected of a hard-object specialist or the uniformly large, deep, and parallel striations observed in tough food grazers (67, 73). Moreover, the complexity values for P. boisei molars fall near the bottom end of the range for extant primates examined thus far (74) and are lower than those for any early hominin with the exception of A. afarensis. Thus, while DMTA results for P. robustus seem to follow along with previous craniodental interpretations suggesting a diet including hard objects, the results for P. boisei run counter to those same interpretations. Finally, while the microwear signature for Homo habilis does not suggest the regular consumption of hard food objects or greater dietary diversity in than in the combined Australopithecus sample, the microwear signature of H. erectus suggests a broader range of foods, especially in relation to hardness, than for H. habilis and most australopiths (68). Taken together, these findings from the early hominins have begun to shed light on questions that can only be answered via evidence of individual behaviors/diets.

In more recent hominins, as the temporal resolution of archeological sites improves, even finer-resolution questions can be answered using dental microwear analyses. For example, microwear analysis suggests that in comparison with modern humans, Neanderthals had relatively stable dietary patterns for most of their existence, while the former had more variable dietary patterns linked to their rapid cultural evolution (75). In addition, there may also have been sex differences in diet or tooth use in some Neanderthal family groups (76).

Biogeochemistry

Elements in food and water are incorporated into the hard tissues of the body, permitting certain aspects of extinct diets to be discerned through chemical analyses of bones and teeth. The distinct advantage of employing biogeochemistry in paleodietary reconstruction is that various trace elements and stable isotope compositions in the hard tissues will reflect the foods that were actually eaten by an individual during the period of that tissue’s formation. With reference to the Plio-Pleistocene hominins, the chemical signals preserved in teeth can reflect the proportion of ingested plants that employ different photosynthetic pathways (77, 78), and environmental variables such as trophic level, consumption of marine versus nonmarine resources, and differences in rainwater composition (79).

Since the initial studies by DeNiro and Epstein that identified the potential of carbon isotopes to elucidate the consumption of plants that employed different photosynthetic pathways (i.e., C₃ versus C₄), this has been among the most widely utilized and best understood direct approaches to diet reconstruction. The carbon isotope (¹³C/ ¹²C) composition of tooth enamel relates largely to the consumption of plants that exhibit different photosynthetic pathways and thus differ in the enrichment of ¹³C relative to ¹²C. In the African context, this translates into trees and shrubs versus tropical grasses and sedges. It should be noted, however, that ¹³C depletion in C₃ plants can vary with the level of canopy shade cover, where a “canopy effect” in dense forests leads to even lower δ¹³C values (80). However, while previous research has suggested strong correlations between canopy cover and the δ¹³C values of mammalian herbivores (irrespective of their specific diets or gross metabolic differences) [e.g., Schoeninger (81)], more recent work has indicated that such inferences must begin with community-level, inter-taxon comparisons before broader inter-site comparisons can be made, especially in analyses of prehistoric taxa differing dramatically in sample size and diversity (82). Thus, for example, while the canopy effect might be a factor to consider in interpreting the relatively low δ¹³C values reported for A. ramidus (83) and A. sediba (32) at Aramis and Malapa, respectively, inter-taxon comparisons still show the fossil hominins to fall within the C₃ feeder range given their relatively low δ¹³C values.

An additional caveat for the interpretation of carbon isotope ratio (¹³C/ ¹²C) data is the observation that these values are not only reflected (albeit with fractionation) in the tissues of the herbivores that consume them, but also in the tissues of the carnivores that consume the herbivores (78, 84). As such, the δ¹³C values of a hominin might be indicative of its reliance on C₃ or C₄ plants and/or the meat of herbivores that consumed the plants.

Nevertheless, because of their demonstrable relationship to plant physiology and their ability to discern C₃ from C₄ plant consumption, carbon isotope data have been widely analyzed for numerous Plio–Pleistocene African hominin taxa. Indeed, Sponheimer et al. (85) compiled δ¹³C data for over 200 such specimens and to this rich abundance can be added an additional 77 specimens (11, 86). This totals over 277 fossils representing at least 11 and possibly 13 species. As discussed at length elsewhere (11, 74, 85, 87), there is a tendency for C₄ resource utilization to become more prevalent among African hominins through time, although there are differences in the timing of dietary shifts among taxa and lineages. At an even finer level of distinction, Sponheimer et al. (88) and Lee-Thorp et al. (87) have documented interannual and perhaps even inter-seasonal variation in the diet of P. robustus and A. africanus, respectively—something that, again, could never be documented with craniodental functional morphology.

While carbon can be viewed as the stable light isotope workhorse in paleodietary reconstruction, trace elements such as barium and calcium have also proven to be valuable sources of paleodietary information. For example, A. africanus is characterized by a high Sr/Ba ratio that is quite distinct from contemporaneous species in South Africa, suggesting the possibility that this species consumed items with unusually high Sr and relatively low Ba concentrations (89). Foods that could meet this requirement include USOs such as roots, rhizomes, and bulbs. Balter et al. (90) have reported Ba/Ca ratios for A. africanus, P. robustus and early Homo teeth and concluded that the former species had a more varied diet than the latter two.

Studies of calcium isotopes have demonstrated a general decrease in δ^44/42Ca ratios with increasing trophic level associated with a fairly constant offset (ca. −0.6‰) from dietary Ca to bone and tooth enamel (91). The δ^44/42Ca values of A. anamensis and Kenyanthropus platyops teeth are similar to one another and neither is notably ⁴⁴Ca depleted. Both hominins can be accommodated within the eastern African browser domain (92). On the other hand, Martin et al. (92) found that the values of geochronologically younger eastern African early Homo teeth are widely dispersed. While this may indicate greater dietary flexibility, as interpreted by the authors, it may also reflect species heterogeneity (i.e., H. habilis, H. rudolfensis, and H. erectus) in their sample. At the same time, they found P. boisei to be the most ⁴⁴Ca-enriched—not only among the hominins but also in comparison with the cercopithecoid primates—suggesting a unique dietary niche for this species that may have comprised forbs found in wetland habitats.

In sum, the five decades that have elapsed since the pioneering use of stable light isotopes to address paleodietary questions have witnessed the exploration of a wide range of chemical signals. These explorations have met with varying degrees of success, but the incredible (and seemingly ever increasing) breadth of information that can be gleaned through isotope studies [e.g., the environmental impact of late Neanderthals and Upper Pleistocene modern humans (93)] has had a profound impact in paleoanthropology in recent decades.

Discussion

Deciphering the diets of human ancestors has always been challenging given the complexity of primate diets, the idiosyncratic nature of fossil preservation, the luck and skill in finding fossils, the expertise in preparing those fossils, and the palimpsestic nature of their deposits. In fact, the task is sufficiently complex that analyses can only benefit from so-called multi-proxy approaches, with short-term dietary inferences (the so-called direct methods, or foodprints, mentioned here) providing certain types of data and long-term dietary inferences (such as morphological and biomechanical analyses) providing others (9). As the former studies provide evidence of the foods that were actually eaten by an individual during the period of time that the tissue was being deposited (in the case of biogeochemistry) or the food was being chewed (in the case of dental calculus and microwear), they can provide insights into the behavior and ecology of populations. By contrast, studies of craniodental size, shape, and structure provide insights into the evolutionary history and functional adaptations of species, in effect setting theoretical brackets for diet within the context of specific evolutionary moments (94). Of course, those contexts change through time, as do the selective forces at work on organisms and their adaptations. Moreover, the closer we get to the present day, the more similar our ancestors become to us and the tighter those brackets become, with the net effect that short-term, “direct” analyses of more recent materials are generally yielding finer and finer-resolution inferences about diet within generally accepted constraints of craniodental size, shape, and structure. In essence, finer time resolution at recent sites and more favorable preservation of fossils mean that a greater variety of short-term methods can be used. As noted by Salazar-García et al. (95), these techniques are generally complementary in that they often provide information at different time scales for the same specimens—for instance, long-term average signals for certain isotope analyses and short-term/instantaneous signals for plant micro-remains. The net effect is that they can help document the presence of meat-eating (via isotope analysis) and plant-eating (via plant micro-remains) in Neanderthals which would not be possible using either method by itself. Given the relative paucity of hominin remains in some fossil assemblages, multi-proxy analyses have even involved the associated fauna from some sites, leading to suggestions of Neanderthals using selective, seasonal hunting of specific prey (96) with the selective prey procurement being linked to Neanderthal settlement patterns (97). Still, with the better resolution time scales from more recent sites like La Ferrassie, isotopic evidence from the teeth of associated fauna is beginning to document climatic conditions at specific levels within sites that differ from average climatic conditions bracketed by the chronometric dates of archeological layers (98).

For the earliest hominins, dietary interpretations will always be limited by the depositional environments and large swaths of time bracketed by many deposits (such as those around Lake Turkana), which can rule out the use of some techniques. Still, for example, as evidenced by the discoveries at Malapa in South Africa, under the right conditions, the combination of dental calculus, biogeochemistry, and dental microwear can be used to shed light on the details of early hominin diets. This is only reinforced by perhaps the most intriguing dietary revelation to emerge from any of these studies: the wholly unexpected lack of microwear evidence of hard-object feeding in P. boisei (73) coupled with its evident dependence on C₄ resources (11, 85, 99, 100). This not only ran counter to microwear and isotope data for its South African congener, it ran counter to most functional analyses of its craniodental anatomy. As a result, it triggered an impressive array of research that is causing major adjustments in paleoanthropological and functional morphological perspectives. Some of the work has involved continuing refinements of existing techniques—for example, the use of oxygen isotopes to document seasonal variability in food and water resources in a Miocene primate (101). Some has involved new investigations of additional forms of foodprints—for instance, tooth-chipping, which is providing further evidence of the lack of hard-object feeding in P. boisei (102). However, some of the work is reexamining assumptions that have underlain functional interpretations for decades. For instance, how useful are craniofacial and mandibular morphology as indicators of diet (49, 50, 94, 103)? How might molars with low occlusal relief process tough foods (104)? Would such demands be reflected in changes in molar enamel structure, such as increased enamel thickness (105, 106), or would such changes merely represent the “path of least evolutionary resistance” (107) given existing phylogenetic constraints?

To have any hope of answering such questions, paleodietary analyses have had to move into the twenty-first century by asking far more realistic questions about diet than those tied to traditional, species-level characterizations like “fruit-eater,” “leaf-eater,” and “hard-object feeder.” Primatologists have long known that primate diets are far more complex than such categories might indicate and that temporal and geographical variations in diet can have a critical impact on survival and reproduction (108–110). New paleobiological analyses are finally allowing researchers to think in terms of more realistic levels of dietary variation—levels that are beginning to raise even more questions about the behavior and evolution of our early ancestors. In the process, these analyses are starting to break down traditional overgeneralizations about the abundance and distribution of foods and their nutritional and mechanical properties. For instance, what is the relative nutritive value of C₃ and C₄ plants and how variable are they across Sub-Saharan Africa (111, 112)? Could C₄ plants be a more stable food resource throughout the year for eastern African hominins as compared with South African hominins (112)? What is leaf toughness and how variable might it be in savanna grasses (111, 113, 114)? If all grasses are not equally tough, what dental and physiological traits would be required for early hominins to process them efficiently (115)?

With questions like these finally being asked, it is perhaps no wonder that detailed stable isotopic analyses from the Omo Valley are demonstrating evolutionary patterns that might have been puzzling 20 to 30 y ago. In other words, the shift to C₄ food consumption in early hominins occurred during the time of P. aethiopicus, after the appearance of its classic “hyper-robust” craniodental features and before the appearance of P. boisei (11). Moreover, this shift occurred simultaneously in Australopithecus and Homo as well (11). Clearly, there is no unique craniodental feature that made this dietary shift possible. Instead, we are seeing evidence of evolution operating within the framework of various constraints—essentially making use of the anatomy and physiology at-hand. This sort of discovery was unavailable to us until the development of new techniques and multi-proxy approaches to the study of paleobiology.

Conclusions

Paleoanthropologists have the opportunity to solve puzzles that stretch over dramatically different periods of time—some that allow the documentation of long-term trends in human evolution, others that allow the documentation of short-term events that underlie those long-term trends. All will ultimately allow us to reach a better understanding of “complex biological reality” (116). However, until the advent of techniques that could provide direct evidence of past events (e.g., foodprints), researchers had little hope of moving past standard evolutionary inferences into the realm of the day-to-day actions that form the foundation of evolution. Now that we have those techniques, we should stop and realize that we may be standing in the midst of a paradigm shift in paleoanthropology as we move from inferences of possibilities to evidence of behavior.

Data, Materials, and Software Availability

Previously published data were used for this work (check the bibliography).