A new genus, Tupadesmus (Polydesmida, Chelodesmidae), and a new species, T. muyrakyta (type species for the genus) are herein described from the Amazon rainforest of northern Brazil. Tupadesmus pugiunculus (Schubart, 1946), new combination, is placed in the new genus. We compared the new genus with other Chelodesminae taxa, in particular with Eucampesmella, Brachyurodesmus and Arthrosolaenomeridini. Males of the genus Tupadesmus differ from those of other Chelodesmidae genera by the presence of well-defined striations surrounding incisura lateralis and by the gonopodal prefemoral process divided into two thin, elongate, subequal and acuminate branches. An updated checklist of the Chelodesmidae of Pará State, Brazil, is also provided.

Smicridea McLachlan, 1871, is the richest genus of Hydropsychidae in the New World, with its subgenus Smicridea (Smicridea) being the most diverse, comprising 145 species. However, immature stages of only 13% of these species have been described. In Brazil, only five of 29 species of S. (Smicridea) have documented immature stages. To fill this gap, we describe and illustrate the larva and pupa of S. (Smicridea) erecta Flint, 1974. Using the metamorphotype method, we associated adults with their immature stages. Larvae of S. erecta are characterized by ten dark brown oval muscular scars on the posterior region of the frontoclypeus, arranged in a boomerang shape, and a convex and entirely crenulate anterior margin. The pupa exhibits pairs of hook plates exclusively on the abdominal terga of segments II–VI. This study adds to the understanding of Smicridea (Smicridea), making Smicridea (Smicridea) erecta the sixth species in Brazil and the twentieth globally with documented immature stages.

A new species of the snail-case caddisfly, Helicopsyche (Feropsyche) imperial Silva-Pereira, Desidério, Pereira & Hamada subgenus Helicopsyche (Feropsyche), is described and illustrated based on adult males from a protected and preserved area in the Cerrado biome in Federal District, Central West Region, Brazil. The new species can be recognized mainly by the morphology of tergum X in the dorsal view and inferior appendages in the ventral view. In addition, new records are provided for H. (Feropsyche) vergelana and H. (Cochliopsyche) opalescens. The discovery of the new species and the new record of H. vergelana are the first records of H. (Feropsyche) from the Central West Region of Brazil.

Many associations with microbial species significantly affect the biology, ecology and evolution of the host. Yet, our understanding of the species composition of the gut microbiota remains limited for many host species. Here, we provide a new step towards filling this gap, and characterize the bacterial microbiota of 60 specimens of Lithobius forficatus, a brown stone centipede commonly found in Finland. Many specimens analysed in this study were found to have a very species-rich bacterial community, while others hosted communities clearly dominated by one bacterial species. The most abundant phylotypes included some potential pathogens such as Borrelia and Pseudomonas, a honeybee gut symbiont Gilliamella and some maternally inherited symbiotic bacteria, including Wolbachia and Rickettsiaceae. While females and males were found to carry similar bacterial communities, population had a significant effect on the bacterial community composition. Bacterial species richness did not differ between sexes or between populations in Lithobious forficatus.

The thriving school of palaeontological research that Björn Kurtén surprisingly established at Helsinki came about through a unique combination of Swedish and Finnish traditions that were merged in Kurtén's personal history, inclination and personality. Late in his life, the authors experienced him as a teacher and Ph.D. adviser in the many-splendoured scientific and cultural circumstances that his international recognition as scientist and writer had generated. Here, we share recollections of those days and look back, trying to understand how the seemingly unexpected establishment of successful palaeontology in a nearly fossil-free environment grew through the choices and activities of a very special individual.

Björn Kurtén's novel Dance of the Tiger has some vivid scenes of interaction amongst Neanderthals, in which great care is taken in language to show courtesy and avoid conflict. It was a creative expression developing from feelings that had developed in his mind over a lifetime of studying the remnants of ancient societies. Supporting evidence for this comes from social anthropology, where structures in the society of archaic cultures today seem to concur with the picture formed by Björn Kurtén. This also agrees with Gregory Bateson's concept of schismogenesis, where he looked at patterns of interaction in human society which can escalate into conflict, and at deeply rooted methods of damping them down before they develop.

I was one of Björn Kurtén's last students and came to spend much time with him during the years that we overlapped at the University of Helsinki, which I entered in 1973. Those were marvellous, unforgettable and deeply formative years, rich with science, art and people of all sorts (see Fortelius & Werdelin 2024). While recently sorting through my old papers I stumbled on a forgotten diary from 1979, which contains, among other things, the description of a journey in deep winter from Helsinki to Marseille and back. The main goal was the Ph.D. defence of Evelyne Crégut¹ and a seminar to be held there at the same time. Björn was on the Ph.D. jury and had invited me to come along, at my own expense, of course. I think this was because he thought my experience of excavations at the Petralona Cave in Greece two years before might be of interest to the local scientists, whose research was focused on the Caune de l'Arago Cave in Tautavel, in French Catalonia. Both these caves had delivered fossil human cranial and postcranial remains, which at the time were among the oldest evidence for a human presence in Europe. I believe Björn travelled by air. I travelled separately by train, stopping at several points to see friends, researchers and collections; a holdover from my Interrail-and-hitch-hiker period a few years earlier. In the excerpt below I mainly describe events that occurred during an excursion to the Caune de l'Arago Cave and Prehistory Museum in Tautavel, the day after the defence and seminar (see Figs. 1–3). Despite some moments of transient chaos, it shows Björn as I remember him: kind, enthusiastic, open minded, determined, generous, sometimes rowdy — and with a wonderful sense of humour, even in moments of adversity.

The text has been Google-translated from re-typed, handwritten diary notes in Swedish, with (only) the worst errors fixed manually; ‘/.../’ indicates omitted parts.

Herein, I review the Finnish palaeontologist Björn Kurtén's (1924–1988) views on human evolution, as he presented them in his books. In particular, I discuss Kurtén's controversial hypothesis of an ancient, Oligocene separation of the human and great ape evolutionary lineages. This hypothesis, which he argued for strongly in the early 1970s, was based mainly on the idea that the human dentition could not have evolved from an ape-like dentition. Kurtén thought that some of the fossil primates known at the time, especially Propliopithecus and Ramapithecus from the Oligocene and the Miocene, respectively, were plausible human ancestors because of their supposedly human-like dentitions. New lines of scientific evidence forced Kurtén to abandon his ‘not from the apes’ hypothesis in the 1980s. However, he continued to speak favourably of other minority views within palaeoanthropology, such as the so-called aquatic ape hypothesis, until the end of his life.

This is a story about a 39 000-year-old reindeer antler found in Tornio, southern Finnish Lapland, nearly 60 years ago. It is based on personal experiences and memories that involve Björn Kurtén. It begins with the discovery of the antler and the observations and interpretations made in the 20th century, and is followed by a revision of those interpretations in the light of old and new archaeological and paleoenvironmental data. These suggest that during Lapland's MIS 3 ice-free conditions the antler was shed by a reindeer bull somewhere in the Tornio river valley and was subsequently transported down to its discovery site. Finally, there is a reflection on the effect that the fossil reindeer antler may have had on Kurtén's literary production.

While there is a rich Miocene fossil record from Europe, it is heavily biased to areas south of 50°N, and Miocene records from northern Europe are very scarce. Especially Fennoscandia has traditionally been considered to lack practically all evidence from the Neogene mainly due to heavy glacial erosion during the Pleistocene. However, during recent years, a few isolated, re-deposited finds of Miocene mammal fossils from Finland and Sweden have somewhat changed the situation. While the original provenance of these finds is uncertain, they most likely originate from eroded Miocene sediments within the Fennoscandian shield itself, in any case broadly representing northern Europe. A proboscidean humerus fragment from Suomusjärvi, Finland, is considered Miocene in age because of microfossil contents of the attached sediment remnant, and it has been tentatively identified as Deinotherium sp. Microfossils from the Suomusjärvi specimen indicate a warm, humid freshwater shore environment, broadly similar to the swamp-forest dominated plant communities from Miocene deposits of Denmark and Iceland. Molars of a tetraconodontine suid and a gomphothere from Sweden add to the Miocene finds from Fennoscandia, although the provenance of the latter is still uncertain.

Giant rhinoceroids were once named as the largest land mammals ever imagined. They lived in Eurasia during the Eocene and the Oligocene, and they could have potentially lived in the area of present-day Helsinki, Finland. We, however, may never know this with certainty because not only have no fossil remains of those animals been found in what is now Finland, but nearly no terrestrial sediments from the Eocene or the Oligocene are known to have been preserved in the area. Paleogene sediments, which formed in southern Finland, were destroyed by the local river systems in the Neogene and by the subsequent glaciations in the Late Pleistocene. Here, we present an analytical argument about whether giant rhinoceroses, known as indricotheres, could potentially have lived in Helsinki. From the continental palaeogeography and palaeoclimate estimates concerning the palaeoecology of indricotheres, we conclude that they could potentially have lived here in the Oligocene, even if climatically, they would have been on the margin of their environmental tolerances. Considering their metabolism, the Helsinki area could have accommodated 3–21 indricotheres. From a climatic perspective, perhaps closer to three individuals could have lived here at a time.

Life reconstructions are a popular way for natural history museums to inspire and educate the public about the prehistoric world. Palaeontological display models combine the results of scientific research and apply this knowledge to a three-dimensional piece. No single model-making technique is optimal for all palaeontological subjects. We compare here traditional taxidermy techniques, sculpting with polymer clay, and 3D printing with resin or a filament printer to produce animal reconstructions. Using four extinct animals, the Eurasian cave lion (Panthera spelaea), a theropod dinosaur (Carnotaurus sastrei), a lobe-finned fish (Gyroptychius) and a sea scorpion (Eurypterus remipes) as examples, we study the advantages and limitations of each method and document the process of translating the results of palaeontological studies into accurate scientific models.

Prompted by our previous findings of several subtle but distinct differences in some anatomical aspects of the Saimaa (Pusa hispida saimensis) and Baltic ringed seals (Pusa hispida botnica), particularly in the head region, we investigated the osteological and arterial architecture of the nasal conchae and the ophthalmic rete of through computer tomography, magnetic resonance imaging, arterial silicone injections, dissection, histological examination, and comparison of skull specimens. The conchal and ophthalmic retia, involved in the thermoregulation of mammals, were similar in their anatomical structure to those reported from other pinnipeds and did not differ between the ringed seal subspecies. The endangered status of the Saimaa ringed seal prevented invasive investigation methods helpful for linking anatomical findings to functional properties. However, with silicone models, we were able to confirm the existence of delicate structures such as a separate vibrissal arterial circulation and the lateral and medial posterior ciliary arteries.

Functional traits of herbivorous mammals provide tools for reconstructing past environments. In 1952, Björn Kurtén used distribution of ecomorphological features in fossil herbivorous mammal communities from Late Miocene “Hipparion faunas” of Eurasia to characterize their paleoenvironments as “steppe”, “forest” and “mixed” types. We tested Kurtén's results with a revised set of ecometric methods. We used dental ecometric estimates of mean annual temperature and precipitation, net primary productivity, and normalized difference vegetation index to compare Miocene localities with modern biomes, and dental mesowear to estimate woody and grass cover in the paleoenvironments. Our results agree with Kurtén's, indicating steppe-edge environments in northern China, wooded paleoenvironments in Pikermi, Greece, and central Europe, and open woodland-grassland environment in Maragheh, Iran. Our analyses indicate the presence of wooded grassland savanna in Lothagam and tropical forest in Lukeino in East Africa, further demonstrating paleoenvironmental variation and ecological diversity within later Late Miocene “Hipparion faunas”.

The Late Miocene vertebrate localities in the Axios Valley (Macedonia, Greece) were discovered at the beginning of the last century by the French geologist Camille Arambourg, who made a large fossil collection and carried it back to France. During the last 50 years, extensive fieldwork have been carried out by a Greek–French team, which has yielded a large number of fossils. Important data on fauna, stratigraphy, biochronology, and palaeoecology were obtained from the study of this collection. The fauna of the Axios Valley proved diverse, comprising numerous groups of mammals and number of new taxa. It is divided into five distinct faunal assemblages, which cover chronologically the entire Late Miocene period. These faunas have been systematically studied, dated, and correlated with the European MN (mammal Neogene) biozones. Their analysis has been a major contribution to the understanding of palaeoenvironmental conditions and climatic changes during the Late Miocene in Greece and the wider Eastern Mediterranean region. The faunas of the Axios Valley have a well-established geographic, stratigraphic, systematic, and chronological background. They were the first to appear in the Eastern Mediterranean and can be used as a comparative fauna for the Late Miocene. Subsequent works in Greece and Türkiye added more data, further improving our understanding of the Neogene period in the Eastern Mediterranean. This paper is an attempt to summarise the work done on the Axios Valley Late Miocene mammal fauna in the last 50 years and present the results.

Paleoecological and ecometric studies of animal communities' functional composition require quantification of whether fossil faunas reliably record trait distributions of living communities. We analyzed whether body size and dietary biases exist within Pleistocene Eurasian sites included in the NOW database, compared with modern communities. Based on mass distributions, we discriminated small mammal sites, large mammal sites, and mixed sites in this record. Large mammals made up 50% of occurring genera and, on average, 74% of genera occurring within sites. Mixed sites with more generic occurrences (17% of sites) fell within modern communities' variation in their proportions of body size and dietary categories. Both large and small mammal sites fell outside the range of modern variation. Whereas most modern communities included 50%–60% herbivorous genera, small and large mammal fossil sites record more herbivorous genera (average 70%). Overall, large mammal sites predominated in this record, and herbivores were over-represented relative to non-herbivores.

The “grazing horses″ (Equinae) were successful large herbivores that arrived from North America to Eurasia at the Plio–Pleistocene boundary. The adaptiveness of high-crowned (hypsodont) Equus spp. allowed them to conquer environments from open to forested, resulting in different body sizes. Few large-scale studies on horse body mass in northern Eurasia exist. Despite controversial systematics of Equidae, their ecomorphological features are quantifiable. The Pleistocene habitat and Equus spp. body mass changes were studied across northern Eurasia. Mean non-equid ungulate hypsodonty demonstrated progressive opening of landscapes throughout the Pleistocene, returning towards more humid and closed environments (taiga) below 70°N by the Holocene. Landscape heterogeneity across northern Eurasia indicated different zoogeographic provinces. Mean Equus spp. body mass decreased during the Middle Pleistocene to resemble the Holocene estimates. The Holocene domestication might have prevented small-sized and grazing-adapted horse lineages from extinction when taiga forests spread across northern Eurasia.

We studied the Vallesian records from Türkiye and Iran, and uncovered a diverse record of Cormohipparion species from Yeniyaylacık and Sinap, Türkiye, and their apparent descendant, Hipparion gettyi from Maragheh, Iran. This prompted us to examine the early evolutionary diversification, biogeography and paleoclimatology of western Eurasian hipparions. Our study also included the early hipparion records from Austria, Germany and the Iberian Peninsula. Whereas the Vienna Basin, Austria, has the oldest recorded occurrence of Hippotherium sp. at 11.4–11.0 Ma, our investigation revealed that the most primitive Old World hipparions belonged to the North American genus Cormohipparion. We integrated the data on crania, dentitions, 3rd metacarpals and 3rd metatarsals to further evaluate the systematics of Vallesian hipparions (11.2–9.6 Ma). Sinap, Türkiye, has the most primitive Cormohipparion in Eurasia, Co. sinapensis, but based on postcranial anatomy, we found that also Co. kecigibi is a second Cormohipparion species of Sinap's early Vallesian. The locality of Yeniyaylacık, Türkiye, dated ca. 9.1 Ma, has the latest occurrence of Eurasian Cormohipparion, Co. cappadocium. Our analyses also revealed that the genus Hipparion sensu stricto was probably directly derived from Subparatethyan (including Türkiye) Cormohipparion, with the most primitive taxa being Hipparion gettyi and Hipparion prostylum. Hipparion gettyi first occurred at Maragheh, Iran, ca. 9.0 Ma at the end of MN 10. Hipparion sensu stricto includes H. gettyi, H. prostylum, H. dietrichi and H. hippidiodus and had a range that extended from France through Greece and Türkiye to China.

A tiny feline mandibular corpus fragment from the late Middle Pleistocene fossil hominin locality of Hualongdong is reported here as a new species Prionailurus kurteni. This species represents the smallest known fossil member of the family Felidae to date, comparable in size to two of the smallest modern cats, P. rubiginosus and Felis nigripes. It is unique in having a small p4 mesial accessory cuspid, very weak p4 distal cingulid, relatively deep mandibular corpus under m1, and a distally located anterior border of the masseteric fossa. A notable synapomorphy observed in the lineage including Prionailurus, Felis and Otocolobus, is the mesially inclined m1. The identification of P. kurteni suggests a potentially high diversity of Prionailurus during the prehistorical time, underscoring the significance of revisiting the taxonomy of small Felidae to gain a better understanding of the evolution and diversification of this family.

Since the discovery of Venta Micena in 1976 until now, the Orce archaeopalaeontological sites (Guadix Baza basin) in the northern part of Granada have contributed significantly to the body of knowledge on Early Pleistocene ecosystem dynamics. The exceptional fossil accumulation at Venta Micena stands out, housing important examples of early Pleistocene European faunas. Additionally, the discovery of new sites such as Barranco León and Fuente Nueva 3 has yielded evidence of some of the oldest human presence in western Eurasia, alongside bones bearing cut and percussion marks made by stone tools. These discoveries were made by using novel techniques and methodology, allowing for new interpretations of the fossil record. Integration of artificial intelligence and geometric morphometrics applied to fossil studies contributed to better understanding of the genesis of the sites, and to unravelling the role of humans and other mammals in creating fossil accumulations at the Orce sites. At Barranco León, Canis mosbachensis was found to be the most active carnivore and not, as earlier thought, the large hyaena Pachycrocuta brevirostris. At Fuente Nueva 3, large saber-toothed cats seem to be the top consumers of carcasses. This evidence suggests the existence of more complex associations between humans and various carnivorous taxa.

Polar bears (Ursus maritimus) diverged from brown bears (U. arctos) in the last one million years. Polar bears have a strikingly different external appearance because of their white pelts and large size, but osteological differences are thought to be limited to flatter crania and altered dentition, with little difference in the postcrania despite swimming frequently and living almost entirely on icy substrates. This paper shows that polar bears have substantial differences in their tarsal form and function from other bears, including their close relatives, the brown bears. Ankle gear ratio records the major functional transition from semi-cursorial hemicyonines to the more plantigrade locomotion of crown ursines. Analysis of tarsal morphology among seven extant ursines show that the arboreal species Helarctos malayanus, Melursus ursinus, Tremarctos ornatus, U. thibetanus, and U. americanus have morphologies that permit greater movement at the transverse tarsal and lower ankle joints, especially broader and more gently curved astragalocalcaneal and sustentacular facets on the calcaneum and their articular equivalents on the astragalus, as well as broader and more gently curved navicular facets on the astragalar head. U. arctos and U. maritimus, which are strongly terrestrial, have smaller sustentacular facets and pronounced interlocking between astragalus and calcaneum at the sustentaculum. Polar bears, however, differ from brown bears in that this interlocking is less tight and thus permits more movement at the lower ankle joint. The phylogenetic comparative analysis of shape shows that the divergence in ankle morphology of the polar bear from the brown is one the most rapid bursts of tarsal evolution in ursines.

Mammals have evolved a broad variety of dental morphologies. Nevertheless, the development of the mammalian dentition is considered highly conserved. Molar size proportions exemplify this as a system where small changes in shared developmental mechanisms yield a defined range of morphological outcomes. The Inhibitory Cascade (IC) model states that as molars develop in a sequence, the first developing anterior molars inhibit the development of subsequent posterior ones. The IC model thus predicts a trend of linear tooth size change along the molar row, as has been observed in a wide range of mammalian taxa with otherwise differing dental morphologies. Perhaps the starkest exceptions to the IC rule are bears, in which the second molar is the largest and the third one is disproportionally small. Here we sought to calculate when and how during development, the bear dentition moves away from the IC prediction. We examined molar proportions in eight bear species, and estimated tooth sizes during development. The results indicate that development of bear molars already deviates from IC expectation during patterning. However, during the earlier cap stage, size proportions of bear molars still seem to adhere to the IC model predictions. Overall, these analyses suggest that irrespective of the final outcome, the process of initial splitting of the molar-forming region into individual teeth is conserved and follows the IC rule.

The brown bear, Ursus arctos, has one of the widest geographical distributions of any carnivoran in the world and it also has a rich fossil record. However, in Scotland extensive glaciation and acidic soils have resulted in few fossil or archaeological remains of Ursus spp. being discovered. Here, the palaeontological and archaeological records of bears in Scotland are reviewed. The results of recent analytical methods including radiocarbon dating, genetic and stable isotope analyses on many of these specimens are presented and discussed in relation to long-term climatic and ecological changes in Scotland over the last 50 000 years.

Hyaenidae is a carnivore family with only four extant species, yet a diverse and abundant presence in the fossil record. Today's generalized antagonistic attitudes towards hyenas may be the product of entangled hyena–hominin geographies since the origin of humanity. These shaped ecocultural relationships and may well have affected the convergent evolution of salient traits like complex social systems. In fact, spotted hyenas have been proposed as models of early human evolution. Björn Kurtén had a great interest in fossil hyenas as study objects, having written several publications on this group, including pioneering and comprehensive studies that laid the foundation for major works to follow. Thus, in tribute to Björn Kurtén, we provide an encapsulated overview of his contributions and review recent developments in the understanding of Hyaenidae evolution and ecologies. We conclude with a synopsis of hyenid and hominid interactions from ecological, evolutionary and cultural perspectives.

Dental ecometric traits in large herbivores have been used to reconstruct palaeoenvironments, given the known relationships that these traits have to modern environments (such as the negative correlation between hypsodonty and precipitation). These techniques have largely focused on environments in North America and particularly both Eurasia and Africa, and consequently ecometric models have been trained on groups of herbivores that are most significant in those regions today (Artiodactyla, Perissodactyla, Primates and Proboscidea). The extent to which these relationships are persistent in communities with other dominant herbivores has never been tested. Because South America has been isolated for most of its history, fossil assemblages in the continent contain other clades, which likely have different trait–environment relationships due to their evolutionary history. Quantitative testing of these relationships and tailored regional models for South America, particularly those incorporating dietary information, will improve palaeoenvironmental reconstructions in the continent using mammal communities.

In phylogenetic analysis using morphological characters, probability-based methods are increasingly employed compared to the parsimony methods. Due to the high diversity of morphological data, however, realistic models for among-states transitional rates have not been found, unlike in the case of molecular sequences. Parsimony, as a classical philosophical thinking framework, remains a powerful tool for inferring phylogenies using diversified data. In parsimony-based phylogenetic analysis, how to deal with character conflicts is essential. The long-standing numerical method, searching for the tree(s) with the overall fewest evolutionary steps, however, brings in the issue of weighting characters in an unjustifiable way. By standardizing step number changing ranges among different characters, I proposed the corrected parsimony approach and restate it here. How to properly weight characters a priori is a separate issue that requires carefully scrutinizing the biological evidence case by case, as previously emphasized by numbers of scholars, explicitly or implicitly, including Hennig himself.

Many calicioid fungi accumulate ascospores into an adhesive mass, called the mazaedium, at tips of stipitate apothecia. Fossil specimens from European ambers demonstrate that this morphology had evolved by the Paleogene and has since remained unchanged. The conserved maintenance of a spore-saving strategy is probably linked to animal-vectored dispersal, but experimental evidence confirming this is lacking. Here, we approached the question with a series of experiments, in which ascomata of three distantly related calicioid species were exposed to living individuals of four insect species. The results confirmed that calicioid ascospores are readily attached to any insects that touch the mazaedial spore mass. Adhered ascospores could be recovered from insect surfaces with ultrasonic cleaning. We found no significant differences in the amounts of attached ascospores of different fungal species. We discuss the new findings in the context of previous observations supporting the ecological and evolutionary role of animal-vectored dispersal in calicioid fungi.

Methods to estimate withers height, croup height and other body dimensions from skeletal dimensions of Equus should be developed because these dimensions, generally taken from living domestic Equus, are needed in combination with body mass to determine locomotor ability and thermoregulation. In this study, I applied the so-called hybrid approach to height estimation of past and present Equus because there are only few Equus for which heights and skeletal dimensions are available for extant and extinct specimens. First, I performed anatomical reconstructions of shoulder joint, hip joint, croup height, and withers heights of extant Equus representing all extant species. Second, I generated single-predictor regression equations by regressing these four heights against metapodial lengths and two-predictor equations by regressing them against metapodial lengths and distal-breadth–length ratios. Because two-predictor equations outperformed single-predictor ones in samples of extant Equus, I used only two-predictor equations to estimate heights of extinct Equus.

It is a common observation that most larger (over around 5–10 kg) herbivorous mammals have longer faces than carnivorous ones. A horse has a relatively longer face than a lion, or even than a wolf. This difference in face length is not the case for smaller mammals: rabbits and ferrets have similarly short faces. A shorter face bestows a mechanical advantage at the front of the jaw, advantageous for cropping forage as well as killing prey. Why is this ability sacrificed in long-faced larger herbivores? We propose that longer faces relate to the ability to use the forelimbs in feeding; for reasons of allometric scaling and locomotor specialization, forelimb mobility becomes restricted in larger animals, and a longer face becomes important for food prehension. We test this hypothesis with a large dataset of face lengths and anatomical assessment of forelimb mobility in extant mammals (including a few extinct equids).

Based on his analyses of lynx and brown bear teeth, Björn Kurtén coined the concept of ‘realisation threshold’, the smallest size at which a tooth can form and erupt properly. Kurtén found the smallest sizes of the studied lynx and bear teeth to be 2.5 mm and 3.5 mm, respectively, much greater than for example the smallest rodent teeth known to readily erupt. A recent study comparing developing teeth from shrews to elephants suggested a relatively unchanged theoretical minimum tooth size for mammalian teeth. Together, these studies have left open the question of whether realisation thresholds of teeth are larger in larger mammals than in small ones. Here, we followed Kurtén’s line of thought and compared the sizes of teeth that are variably present in dentitions, and therefore likely cross the realisation threshold only occasionally. First, we show using published reports that variably present teeth are relatively small in large mammals, but larger than the previously suggested theoretical minimum tooth size. Next, we examine the canines of mares, which are known to be variably present. We report one canine that, compared with information found in the literature, is by far the smallest compared to the body size. In conclusion, whereas the variably present teeth tend to be larger in large mammals, there may be overlooked potential for large mammals to develop very small teeth. This information can be helpful in extrapolating findings from common small model organisms, such as mice, to larger mammals, including humans.

Upper molar crown types, characterized by the number, location, and shape of the main cusps alongside the presence and the orientation of cutting edges, facilitate rapid classification of basic morphotypes. This enables broad taxonomic and temporal sampling efficiently. Our research extends the application of crown types to lower molars and premolars of Paleogene primates in North America to categorize the variety of lower premolars and molars that arose among early primates and that likely played a key role in their diversification. We further took advantage of the natural division of lower cheek teeth into a higher (i.e., earlier developing) trigonid and a lower (i.e., later developing) talonid to test evolutionary hypotheses arising from differences in developmental timing. We tested whether the talonid evolved greater diversity in shape than the trigonid, and we assessed the relative contributions of the trigonid and the talonid to the temporal pattern of dental diversification in early primates. In our data, crown type richness generally varied with species sampling. Disparity measures, such as city block distance or total range, may be more independent of species richness, but values were also largely uniform through the p3–m2 series, showed little difference between trigonids and talonids, and were sensitive to the effects of an unusually diverse structure restricted to a single family (i.e., the multi-cusped, bladed premolars of carpolestid plesiadapiforms). Remaining comparisons therefore focused on diversity, rather than disparity of crown types. In p3, the number of talonid crown types outnumbered trigonid crown types by more than two to one. In p4–m2, the number of trigonid and talonid crown types are similar, but species distributions across crown types differed markedly. Species were distributed highly unevenly across trigonid crown type such that at each tooth locus one trigonid crown type tended to dominate whereas others were represented by just one or a few species. Species were somewhat more evenly distributed across talonid crown types. Temporal trends during the Paleogene primate radiation revealed that, overall, crown type richness paralleled species richness. However, separating the trends into trigonid and talonid components revealed a distinct temporal difference between talonid and trigonid diversification. Initially, during a Paleocene phase of the radiation, talonid crown type richness exceeded trigonid crown type richness in p3 and p4, and peaked earlier than trigonid crown type richness in m1 and m2. Later, during the Eocene, trigonid crown type richness either met (p3) or somewhat exceeded (p4–m2) talonid crown type richness. Talonid evolution probably played a greater role than trigonid evolution during the Paleocene phase of the primate radiation, which occurred among plesiadapiform lineages starting from an ancestral condition with a much lower talonid than trigonid. Talonid and trigonid richness trends were more similar during the Eocene phase, which occurred among euprimate and some plesiadapiform lineages where the talonid approached the trigonid in height. This finding underscores the prominent role that highly variable structures play during the early phase of an adaptive radiation.