INTRODUCTION
Mammals and dinosaurs compared
Since their earliest descriptions in the mid-19th Century, dinosaurs were compared by natural historians with Cenozoic mammals in terms of anatomy, ecology, and possibly phylogeny. Meyer (1832, p. 201) observed that Megalosaurus and Iguanodon differed from all other known reptiles as “Saurier mit Gliedmassen ähnlich denen der schweren Landsäugthiere” (“saurian with locomotive extremities like those of the bulky terrestrial mammals” – translation from Owen, 1842). Owen (1842), in the document naming “Dinosauria”, affirmed this observation and further added that dinosaurs were the reptiles “which in structure most nearly approached Mammalia” (p. 203) and speculated that they possessed a physiology “more nearly approaching that which now characterizes the warm-blooded Vertebrata” (p. 204). The subsequent centuries of study revealed copious aspects of Mesozoic dinosaurs which were more broadly similar to Cenozoic mammals than to other clades of sauropsids: e.g., parasagittal stance with digitigrade posture; complex feeding adaptions in many clades; wide latitudinal distributions; evidence for elevated thermometabolism; and nearly exclusive occupation of the medium-to-large size niches in terrestrial communities in their respective time periods of ecological significance (Jurassic and Cretaceous for dinosaurs; Cenozoic for mammals) (Bakker, 1975; Ostrom, 1980; Benson, 2018; Benton, 2021; Mannion et al., 2024; Upchurch & Chiarenza, 2024; Baumgart et al., 2025; Falkingham, 2025).
Famously, one key difference between Mesozoic dinosaurs and Cenozoic terrestrial mammals known since the earliest discovery of the former was the ability for dinosaurs to achieve enormous size: indeed, the trait recognised by Owen in applying the Greek root deinos (“fearfully great”) when naming the group. O’Gorman & Hone (2012), among others, have shown that average Mesozoic dinosaur size is higher than the average Cenozoic mammal size, and that unlike other vertebrate clades the frequency distribution of Mesozoic dinosaur sizes are strongly negatively skewed (that is, larger taxa are more common than small and medium forms).
Another key difference between these clades is their mode of reproduction and its effect on the ecology of dinosaurs and mammals. The evidence supports oviparity for all dinosaurs (extant birds and the various extinct clades of dinosaurs), in contrast with viviparity for therian mammals (Chappelle et al., 2025). Structural limitations of external eggs limits their maximum size (Horner, 2000), so that the size disparity between hatchling large dinosaurs and their adults is orders of magnitude greater than in large mammals. For instance, adult titanosaurs might be 2500 times more massive than their hatchlings, whereas adult Elephas and Loxodonta are only 22-75 times the size of their newborns (Codron et al., 2012). As Janis & Carrano (1992) and Werner & Griebeler (2011) have shown, litter size in therian mammals decreases with body size, such that only a single offspring at a time is the norm for mammals of 1 tonne or greater; in contrast, dinosaurian clutch size remains high (1-2 dozen or more) even in the largest dinosaurs.
In what ways might the size distribution and mode of reproduction differences between Mesozoic dinosaurs and Cenozoic mammals result in different community structures? This study is a preliminary work incorporating size, ontogeny, and degree of parental care in examining differences between the terrestrial ecosystems of the Jurassic and Cretaceous on the one hand, and the Cenozoic on the other. This set of analyses examines how size distributions compare between dinosaurian and mammalian faunal communities looking at adult body size only, as well as incorporating the sizes of juveniles. Additionally, it compares species richness (number of identified species in a community) and how incorporating the likelihood that juvenile dinosaurs lived separately from and were functionally distinct ecologically from their adults resulted in a different community profile than if dinosaur species were represented only by their somatically mature forms.
Dinosaurian ontogenetic niche shifts and mammalian extended parental care
In extant dinosaurs (crown birds, Aves or Neornithes), young birds are dependent on one or more parents for provisioning and protection, typically past the onset of flight (fledging) and up to the point the offspring are of nearly adult size, morphology, and behavior (e.g., Naef-Daenzer & Grüebler 2016). Consequently, crown birds are not free-living until they are effectively adult data, and hence in terms of trophic ecology each bird species exists essentially only as the adult form.
A very different life history strategy is present in Crocodylia, the extant sister taxon to dinosaurs. Crocodilians engage in nest attendance and extended periods of parental care (including provisioning of the offspring) for several months to over a year and a half (Brazaitis & Watanabe, 2011; Carl & Darlington, 2017; Murray et al., 2020). Free-living juvenile crocodilians engage in substantial ontogenetic niche shifts once they are no longer under the protection and provisioning of an adult. For instance, Alligator mississippiensis eat mostly insects until reaching about 60 cm length, shifting to crustacean and fish prey until about 120 cm long, then incorporating mammals and turtles after that (Platt et al., 1990; Delany et al., 1999). More distantly related to dinosaurs, other large reptiles lacking parental care, such as the large varanid Varanus komodoensis, show rapid transitions in prey size, foraging strategies, and home range during ontogeny (Puruwandana et al., 2016). Such marked ontogenetic niche shifts would be expected in animals with large disparities between free-living juveniles and full adults, such that the mobility, food accessibility, and dangers from various predators would be markedly different at each growth stage.
The shared presence of nest attendance and parental care of at least several months in both crown-birds and crown-crocodilians strongly supports the hypothesis that this state was ancestral for all stem-birds (including all non-avian dinosaurs) and stem-crocodilians (non-crocodilian pseudosuchians). The primary distinction between the life histories of the living taxa is that at the end of the parental care phase birds are essentially fully grown and occupying the same ecological niche as their parents, while crocodilians still have many years of growth before achieving the size and ecological role of the adults.
What was the condition in Mesozoic dinosaurs? Substantial evidence of many decades (reviewed in Erickson, 2014) demonstrate that dinosaurs took years to a few decades to achieve full body size, and that non-neornithine Mesozoic birds such as Archaeopteryx and Confuciusornis retained this pattern, requiring multiple years to achieve full body size (Erickson et al., 2009). But what degree of parental care was present, and for how long were young dinosaurs cared for? A variety of evidence supports at least some parental care including provisioning for hatchlings of many dinosaurian clades (Horner, 2000; Chapelle et al., 2024), including at least basal sauropodomorphs (Reisz et al., 2024).
Varricchio (2011) proposed that most non-avian dinosaurs had a life history distinct from crown birds, crocodilians, and mammals. Based on the relatively common occurrence of monospecific clusters of juveniles of the same ontogenetic status (e.g., Jerzykiewicz et al., 1993; Zhao et al., 2007; Varrichio et al., 2008, Myers & Fiorillo, 2009), he suggests that in many dinosaur species there were free-living aggregations of similar aged individuals (presumably nestmates) which lived and traveled together independent of parents or other adults. Subsequent evidence including trackways (Fiorillo et al., 2014) and additional bonebeds (Joubarne et al., 2024) is consistent with small individuals being absent in at least some gregarious dinosaur herds. Under this scenario pods of young dinosaurs would fend for themselves after the end of their parental care for at least some period of time; in gregarious species, it would appear they then became associated with a group of larger individuals at some point (perhaps when their size allowed them to keep up with the speed of adult animals.) One alternative possibility that should be considered, however, is that the absence of large adults with aggregations of juveniles or of tracks of small young on tracksites with medium-and-large individuals might be due to taphonomy. For instance, it may be that the conditions under which with younger skeletons were buried was insufficient to kill or bury adult individuals (Zhao et al., 2013), or the substrate was not compliant enough to record tracks of very small young.
A consequence of free-living dinosaur juveniles smaller than the adults is that many, if not all, non-avian dinosaurs likely experienced ontogenetic niche shifts comparable to what is seen in crocodilians and Komodo dragons (Platt et al., 1990; Delany et al., 1999; Puruwandana et al., 2016). Palaeoecological, morphological, biomechanical, isotopic, and theoretical studies are consistent with both carnivorous and herbivorous dinosaurs experiencing different trophic modes and roles, different locomotory capabilities, and different ecological interactions at different growth stages (e.g., Brett-Surman, 1989, 1997; Codron et al., 2012, 2013; Woodruff et al., 2018; Snively et al., 2019; Frederickson et al., 2020; Holtz, 2021; Schroeder et al., 2021; Therrien et al., 2021; Wyenberg-Henzler et al., 2022a, b), even in non-neornithine birds (Marugán-Lobón & Chiappe 2022). Following these observations, any individual free-living dinosaur represents functionally a different “species” at different stages of growth from the time they are living independently of their parents through the time they are fully grown.
Extant mammalian life histories differ from those of both extant and extinct dinosaurian species. Like extinct dinosaurs but unlike crown birds, medium-to-large sized mammals required multiple years to achieve full body size; like birds but unlike extinct dinosaurs, mammals typically have extended parental care until the offspring are almost fully grown (Varricchio, 2011). The extended parental care (more specifically, maternal care) in mammals involves lactation for early stages of growth, and provisioning and/or protection until near-adult size. Indeed, in some mammals the mothers are responsible for provisioning offspring up to the point where the juveniles are as large or larger than the mother (Khaewphakdee et al., 2020). Even in post-weaning herbivorous gregarious mammals such as macropod marsupials and ungulate placentals (where the mother is not directly responsible for capturing food items for the young), maternal vigilance remains important until late in the life history of the offspring (Fisher et al., 2002).
The result of this prolonged period of maternal care is that there is no expectation of ontogenetic niche shifts in mammals. Young mammals represent functionally the same species as their mothers in terms of their feeding ecology until they are fully-grown.
MATERIALS AND METHODS
Fossil communities
Following earlier work (Holtz, 2021), data were collected on the species list of a sample of fossil communities from the Jurassic and Cretaceous to represent dinosaurian faunas, and the Eocene through Pleistocene to represent mammalian ones. Triassic communities were excluded; although dinosaurs were present in many of these, pseudosuchian archosaurs were ecologically and taxonomically more abundant (Brusatte et al., 2008; Zanno et al., 2015). Paleocene faunas were excluded as mammals crossing the 1000 kg threshold used in this study had yet to evolve (Alroy, 1999). Where possible, assemblages were chosen that represented instantaneous diversity: that is, communities where the taxa in question cooccurred in time and space. Consequently, these are typically communities within a particular member or biozone of a formation, rather than the lumping together of taxa from across formation deposited over multi-million-year intervals.
The faunal assemblages used here are not intended to be a comprehensive overview of all such palaeocommunities but instead were selected to reflect several criteria: varying palaeogeographic settings; a diversity of component clades represented; relatively secure age dates; and relatively secure taxonomic assignments of the relevant taxa. To keep communities broadly comparable, only those with at least one taxon massing 1000 kg or more must be present. Formations with unusual taphonomic filters (such as the various Jehol biota communities (Pan et al., 2013) or the Messel Formation of the Eocene of Germany (Smith et al., 2024) were excluded: such units often produce excellent samples of the smaller bodied members of ecosystems, but due to their sedimentology often preferentially exclude the larger animals of interest here.
In several cases a series of formations representing a sequence of assemblages from the same depositional basin over an extended period of time were included: for instance, the Late Jurassic Lourinhã Formation of Portugal (Mateus et al., 2017), the Cretaceous Cedar Mountain Formation of Utah, USA (Kirkland et al., 2024), and the Eocene-to-Miocene John Day Beds of Oregon, USA (Fremd, 2010), are all represented by multiple assemblages in this database.
The resulting selection are eighteen dinosaur and eighteen mammalian assemblages for comparison (Table 1). Within each, the species occurrences (derived from the literature, the Paleobiology Database (https://paleobiodb.org/#/), and the New and Old Worlds Database of Fossil Mammals (https://nowdatabase.org/) were compiled (Supplementary Appendix 1). As in Holtz (2021) stratigraphically successive non-overlapping species within a clade were counted as a single taxon rather than as two species within the fauna.
| Assemblages listed in stratigraphic order. |
|---|
| Mammalian Faunas |
| Rancho La Brea, Pleistocene, USA |
| Lujan Formation, Guerrero Member, Pleistocene, Argentina |
| Swartkrans Formation, Member 1, Pleistocene, South Africa |
| Sterkfontein Formation, Member 4, Pleistocene, South Africa |
| Tetoiu Formation, Grăunceaunu site, Pleistocene, Romania |
| Lothagam Formation, Lower Nwata Member, Miocene, Kenya |
| Gray Fossil Beds, Miocene, USA |
| Pikermi Formation, Miocene, Greece |
| Ash Hollow Formation, Cap Rock Member, Miocene, USA |
| Ash Hollow Formation, Merritt Dam Member, Miocene, USA |
| Rattlesnake Formation, Miocene, USA |
| Mascall Formation, Miocene, USA |
| John Day Formation, Haystack Valley Member, Miocene, USA |
| John Day Formation, Kimberly Member, Miocene, USA |
| Chitarwata Formation, Bugti Member, Oligocene, Pakistan |
| Salla Beds, Oligocene, Bolivia |
| Clarno Formation, Hancock Mammal Quarry, Eocene, USA |
| Upper Chadron Formation, Eocene, USA |
| Dinosaurian Faunas |
| Hell Creek Formation, upper beds, Maastrichtian, USA |
| Nemegt Formation, Maastrichtian, Mongolia |
| Horseshoe Canyon Formation, middle beds, Maastrichtian, Canada |
| Allen Formation, Campanian-Maastrichtian, Argentina |
| Dinosaur Park Formation, Megaherbivore Assemblage Zone 1b, Campanian, Canada |
| Iren Dabasu Formation, Santonian, China |
| Huincul Formation, Cenomanian, Argentina |
| Cedar Mountain Formation, Poison Strip Member, Hauterivian-Barremian, USA |
| Wessex Formation, Berriasian-Barremian, UK |
| Cedar Mountain Formation, upper Yellow Cat Member, Valanginian, USA |
| Cedar Mountain Formation, lower Yellow Cat Member, Berriasian, USA |
| Villar del Arzobispo Formation, Kimmeridgian-Berriasian, Spain |
| Lourinhã Formation, Praia Azul Member, Kimmeridgian-Tithonian, Portugal |
| Lourinhã Formation, Praia da Amoreira and Porto Novo Members, Kimmeridgian, Portugal |
| Alcobaça Formation, Kimmerdigian, Portugal |
| Morrison Formation, Zone 2, Kimmerdgian, USA |
| Shishugou Formation, upper beds, Callovian-Oxfordian, China |
| Xiashaximiao Formation, Bajocian-Callovian, China |
Ideally, the occurrence data would be weighted by the relative percentages of each taxon in each assemblage. However, such census is missing in the majority of these fossil communities, and in some cases small sample sizes might render fossil collection data inadequate in reflecting the actual relative percentages within the living communities. Thus, simply the presence of each taxon in an assemblage is recorded, as in previous comparisons of dinosaurian and mammalian communities by Van Valkenburgh & Molnar (2002), Mallon (2019), and Schroeder et al. (2021).
Body sizes, adult and juvenile
The primary ecological parameter examined in the present study is body size. While only one of many functional aspects of a taxon (e.g., locomotion; habitat range; feeding adaptations; among many others), body size is a factor which can be directly compared among disparate groups.
Following Holtz (2021), size classes rather than specific mass estimates are used as the point of comparison between taxa and communities. Individual taxon mass estimates were derived from the literature, particularly following the work of O’Gorman & Hone (2012), Benson et al. (2018), and Campione & Evans (2020). Recent work by Dempsey et al. (2025) suggests that the dinosaurian masses estimated in the previous work may be underestimates; however, their new methodology requires relatively complete skeletal data in order to approximate body size, which are sadly lacking for many clades. Where no specific estimate has been published for a particular taxon, comparative measurements were used to approximate the size from a close relative with better established data.
Expanding on the size class framework of Holtz (2021) to accommodate the larger body size of sauropod dinosaurs, the classes used here are 1, 10 kg or less; 2, 11-50 kg; 3, 51-100 kg; 4, 101-500 kg; 5, 501-1000 kg; 6, 1001-5000 kg; 7, 5001-10,000 kg; 8, 10,001-50,000 kg; and 9, greater than 50,000 kg. These general size classes lack the precision of applying individual mass estimates; however, such precision may mask the variability within species and seasonal variability within individuals and the uncertainty of mass estimates using different techniques. The adult mass estimates were taken from the largest known individuals. In some cases, the largest known individuals might surpass “typical” known specimens. For example, the largest known specimen of Edmontosaurus annectens (MOR 1142) appears to have massed approximately 14 tonnes (and thus was size class 8), whereas more typical collected specimens are approximately 3-4 tonnes (and thus size class 6) (SpinoInWonderland, 2021). Hone & Mallon (2017) and Mallon & Hone (2024) discuss how the small sample size of most dinosaur taxa and their particular growth strategy (determinate growth, but with full somatic maturity representing a small fraction of their natural lifespan rather than the majority, as in mammals and in crown birds) results in an undersampling of fully grown Mesozoic dinosaurs compared to Cenozoic mammals. For fossil mammals, no attempt was made to reflect separate maximum body sizes for taxa with known or suspected sexual dimorphism where the sexes might be in different size classes.
All known Mesozoic dinosaurs had hatchlings within size class 1 (Horner, 2000), and thus there were young and subadult dinosaurs occupying every size class between 1 and the maximum adult size (see also Cau, 2024). In contrast, young mammals are typically much closer to adult body masses than in Mesozoic dinosaurs (Codron et al., 2012). For example, a newborn Loxodonta africana is typically 100-120 kg (size class 4) while the adults achieve a body mass of 5-7 tonnes (size class 7). As a simplifying assumption, baby through subadult mammals are coded as occupying the three size classes smaller than the fully adult mass.
While survivorship curves have been calculated for a few dinosaur species (Erickson et al., 2010; Woodward et al., 2015; Griebeler, 2021), there is insufficient data at present for most of the taxa in this study. For this reason, no attempt is made to weight the biomass of individuals in each size class.
As in Holtz (2021), taxa of size class 1 are excluded from this analysis. This results in removing certain clades (e.g., avialians, alvarezsaurids, many smaller paravians among dinosaurs; most rodents and lagomorphs, smaller carnivorans, marsupials, and ungulates, and others among mammals). Small taxa were undoubtedly critical parts of the ecosystem. However, at that size class many additional clades would have to be considered outside the comparison of mammals and dinosaurs: for instance, lepidosaurs, testudines, small terrestrial crocodylomorphs, larger arachnids, and more. As a simplifying factor for this preliminary study, only the taxa greater than 10 kg adult mass are considered. Thus, these results are not comprehensive and will likely result in a slight skew of the statistics compared to the real-world distribution of sizes. It is worth observing, however, that similar exclusions are present in studies of extant ecosystems (e.g., Baskerville et al., 2011), and such comprehensive data may not exist in even the neontological ecological literature.
In the mammalian faunas there are a few sauropsids large enough to be included in data set. The Grăunceaunu locality of the Tetoiu Formation (Lower Pleistocene, Romania) contains the large ratite Pachystruthio sp. (Terhune et al., 2020), while the Pikermi Formation (Upper Miocene, Greece) contains the large ratite Struthio karatheodoris (Koufos et al., 2016) and the gigantic testudinidid Titanochelon sp. (Vlachos et al., 2020). As crown birds, like mammals, have extended parental care until the offspring are nearly fully grown, the young ratites are assumed to have achieved the size class of the adults before they are free living; however, as testudines have no parental care, the hatchling (size class 1) and all juvenile stages are assumed to be free living for Titanochelon.
Analyses
The statistical analyses for this study were conducted using the PAST (Paleontological Statistics) 5.2 software package (Hammer et al., 2001; version 5.2 2025). The data for size classes and species counts for each of the scenarios are provided as Table 2.
| Assemblage | Category | Species Counts | Average Size Class | ||||
|---|---|---|---|---|---|---|---|
| Adults Only | Juveniles Included | Independent Juveniles Included | Adults Only | Juveniles Included | Independent Juveniles Included | ||
| Rancho La Brea | Mammal | 31 | 84 | 31 | 3.93548387 | 3.4047619 | 3.93548387 |
| Lujan Formation, Guerrero Member | Mammal | 31 | 110 | 31 | 4.35294118 | 3.64864865 | 4.35294118 |
| Swartkrans Formation, Member 1 | Mammal | 46 | 92 | 46 | 3.06521739 | 2.93548387 | 3.06521739 |
| Sterkfontein Formation, Member 4 | Mammal | 46 | 95 | 46 | 3.13043478 | 2.89583333 | 3.13043478 |
| Tetoiu Formation, Grăunceaunu site | Mammal | 29 | 74 | 29 | 3.67857143 | 3.18918919 | 3.67857143 |
| Lothagam Formation, Lower Nwata Member | Mammal | 42 | 98 | 42 | 3.62790698 | 3.45918367 | 3.62790698 |
| Gray Fossil Beds | Mammal | 15 | 38 | 15 | 3.8 | 3.47368421 | 3.8 |
| Pikermi Formation | Mammal | 38 | 115 | 40 | 3.91111111 | 3.3826087 | 3.80434783 |
| Ash Hollow Formation, Cap Rock Member | Mammal | 11 | 25 | 11 | 3.36363636 | 3.08 | 3.36363636 |
| Ash Hollow Formation, Merritt Dam Member | Mammal | 21 | 42 | 21 | 3 | 2.84090909 | 3 |
| Rattlesnake Formation | Mammal | 30 | 73 | 30 | 3.66666667 | 3.39189189 | 3.66666667 |
| Mascall Formation | Mammal | 28 | 62 | 28 | 3.42857143 | 3.28571429 | 3.42857143 |
| John Day Formation, Haystack Valley Member | Mammal | 34 | 77 | 34 | 3.35294118 | 3.06493506 | 3.35294118 |
| John Day Formation, Kimberly Member | Mammal | 17 | 39 | 17 | 3.35294118 | 2.92307692 | 3.35294118 |
| Chitarwata Formation, Bugti Member | Mammal | 16 | 49 | 16 | 4.5 | 3.88 | 4.5 |
| Salla Beds | Mammal | 27 | 56 | 27 | 3.22222222 | 3.19642857 | 3.22222222 |
| Clarno Formation, Hancock Mammal Quarry | Mammal | 13 | 35 | 13 | 3.92307692 | 3.4 | 3.92307692 |
| Upper Chadron Formation | Mammal | 38 | 77 | 38 | 3.10526316 | 2.96103896 | 3.10526316 |
| Hell Creek Formation, upper beds | Dinosaur | 18 | 54 | 45 | 4.11111111 | 3.81481481 | 4.02222222 |
| Nemegt Formation | Dinosaur | 26 | 80 | 66 | 4.23076923 | 3.95 | 4.1969697 |
| Horseshoe Canyon Formation, middle beds | Dinosaur | 15 | 43 | 41 | 4.13333333 | 3.5 | 3.70731707 |
| Allen Formation | Dinosaur | 14 | 57 | 47 | 5.21428571 | 4.07017544 | 4.34042553 |
| Dinosaur Park Formation, Megaherbivore Assemblage Zone 1b | Dinosaur | 23 | 71 | 60 | 4.08695652 | 3.6056338 | 3.9 |
| Iren Dabasu Formation | Dinosaur | 12 | 39 | 36 | 4.25 | 3.41463415 | 3.66666667 |
| Huincul Formation | Dinosaur | 18 | 91 | 64 | 6.05555556 | 4.38709677 | 5.04918033 |
| Cedar Mountain Formation, Poison Strip Member | Dinosaur | 6 | 27 | 23 | 5.5 | 3.85185185 | 4.13043478 |
| Wessex Formation | Dinosaur | 24 | 91 | 76 | 4.33333333 | 3.96774194 | 4.1971831 |
| Cedar Mountain Formation, upper Yellow Cat Member | Dinosaur | 11 | 42 | 35 | 4.90909091 | 3.9047619 | 4.22857143 |
| Cedar Mountain Formation, lower Yellow Cat Member | Dinosaur | 8 | 27 | 24 | 4.375 | 3.55555556 | 3.75 |
| Villar del Arzobispo Formation | Dinosaur | 11 | 59 | 41 | 6.36363636 | 4.45762712 | 5.2195122 |
| Lourinhã Formation, Praia Azul Member | Dinosaur | 11 | 38 | 32 | 4.63636364 | 3.94736842 | 4.1875 |
| Lourinhã Formation, Praia da Amoreira and Porto Novo Members | Dinosaur | 14 | 46 | 40 | 4.42857143 | 3.82608696 | 3.975 |
| Alcobaça Formation | Dinosaur | 9 | 35 | 28 | 5 | 4.08571429 | 4.46428571 |
| Morrison Formation, Zone 2 | Dinosaur | 20 | 87 | 61 | 5.35 | 4.34090909 | 4.98360656 |
| Shishugou Formation, upper beds | Dinosaur | 12 | 40 | 32 | 4.33333333 | 3.925 | 4.21875 |
| Xiashaximiao Formation | Dinosaur | 18 | 64 | 54 | 4.61111111 | 3.77272727 | 4.05555556 |
Institutional abbreviations
MOR, Museum of the Rockies, Montana State University, Bozeman, Montana, USA.
RESULTS
Three scenarios were examined in this study: using only adult size to represent taxa; including all growth stages to represent taxa; and using independent-living juveniles and adults (but not dependent offspring) to represent taxa. For each of these scenarios, the average size class in dinosaurian versus mammalian communities were compared via box plots (Fig. 1), as well as the average number of effective taxa in each scenario (Fig. 2). Histograms depicting the size distribution of taxa for select communities are provided in Figs. 3 and 4.
- Box-and-whisker plot of average size classes in dinosaurian (blue) and mammalian (orange) faunas, under three different scenarios. Left, adult body size only; middle, juvenile size included; right, independent young included as separate functional “species”. Data from Table 2 and Supplementary Appendix 1. Silhouettes of Tyrannosaurus rex and Loxodonta africana from PhyloPic.org. Tyrannosaurus rex by Scott Hartman and reproduced under Attribution-NonCommercial-ShareAlike 3.0 Umported License. (link: https://www.phylopic.org/images/ff4f8bd4-7788-41f0-87bc-4fc42bb1da72/tyrannosaurus-rex). Loxodonta africana by Chuanxin Yu and reproduced under CC0 1.0 Universal Public Domain Dedication License (link: https://www.phvlopic.org/images/910d853a-1a15-4953-a1d3-b81208994d35/loxodonta-africana)
A box-and-whisker plot compares average size classes ranging between 2 and 7 in dinosaurian and mammalian faunas under three scenarios.
- Box-and-whisker plot of average functional “species” counts in dinosaurian (blue) and mammalian (orange) faunas, under three different scenarios. Left, adult body size only; middle, juvenile size included; right, independent young included as separate functional “species”. Data from Table 2 and Supplementary Appendix 1. See caption for Fig. 1 for sources and licenses of silhouettes.
Three box and whisker plots of the average functional species counts of faunas ranging between 0 and 120, in three different scenarios.
- Histogram of distribution of taxa at each size class for selected representative mammalian (upper) and dinosaurian (lower) faunal assemblages. Solid blue lines, adult taxa only; orange stipples, richness if all juvenile sizes are included. Data from Table 2 and Supplementary Appendix 1. See caption for Fig. 1 for sources and licenses of silhouettes.
A histogram compares the species counts at each size class till size greater than 50 tonnes for selected mammalian and dinosaurian faunal assemblages.
- Histogram of distribution of taxa at each size class for selected representative mammalian (upper) and dinosaurian (lower) faunal assemblages. Solid blue lines, adult taxa only; red diagonals, richness if independent juvenile sizes are included. Presence of independent young in Pikermi Formation reflects free-living juveniles of the testudinid Titanochelon (see text for discussion). Data from Table 2 and Supplementary Appendix 1. See caption for Fig. 1 for sources and licenses of silhouettes.
A panel of twelve bar charts displays species counts across different weight classes, split into two main ecological sections.
To determine whether dinosaurian and mammalian communities represent statistically significant distributions, Student t tests for equal means and Mann-Whitney U tests for distributions around equal medians were conducted. Primary statistics for these comparisons are provided in Tables 3 and 4, and additional statistics are available as Supplementary Appendix S2.
| Scenario | Faunal Type | Mean | t-test Value | Median | U-test Value | Skewness | Kurtosis | Result |
|---|---|---|---|---|---|---|---|---|
| Adult Sizes Only | Dinosaurs | 4.77±0.16 | 0.000000379 | 4.52 | 0.00000414 | 1.116905 | 0.423525 | Dinosaurs > Mammals |
| Mammals | 3.58±0.10 | 3.53 | 0.6290091 | -0.2405497 | ||||
| Young as Separate Taxa | Dinosaurs | 3.91±0.07 | 0.0000000584 | 3.91 | 0.00000306 | 0.2554598 | -0.240112 | Dinosaurs > Mammals |
| Mammals | 3.25±0.07 | 3.24 | 0.4301177 | -0.2643486 | ||||
| Independent Young as Separate Taxa | Dinosaurs | 4.24±0.10 | 0.0000621 | 4.19 | 0.00021374 | 1.032405 | 0.5214471 | Dinosaurs > Mammals |
| Mammals | 3.57±0.10 | 3.52 | 0.6727456 | -0.08803714 |
| Scenario | Faunal Type | Mean | t-test Value | Median | U-test Value | Skewness | Kurtosis | Result |
|---|---|---|---|---|---|---|---|---|
| Adult Sizes Only | Dinosaurs | 15±1.34 | 0.000058256 | 14 | 0.00032363 | 0.4442117 | -0.6425715 | Mammals > Dinosaurs |
| Mammals | 28.5±2.62 | 29.5 | -0.0168157 | -1.033759 | ||||
| Young as Separate Taxa | Dinosaurs | 55.1±5.03 | 0.097022 | 50 | 0.14106 | 0.5211393 | -0.9785076 | Mammals = Dinosaurs |
| Mammals | 68.9±6.40 | 73.5 | 0.05143094 | -1.092422 | ||||
| Independent Young as Separate Taxa | Dinosaurs | 44.7±3.67 | 0.0011096 | 41 | 0.0030829 | 0.4626308 | -0.7755874 | Dinosaurs > Mammals |
| Mammals | 28.6±2.64 | 29.5 | -0.0186373 | -1.083933 |
Fig. 1 and Table 3 show that in each of the three scenarios that the average size class of dinosaurs in their communities were statistically higher than mammals within theirs. This is hardly a ground-breaking revelation but does confirm Owen’s observation of dinosaurs as being “fearfully great” holds true after two centuries of continued discoveries. While inclusion of all juvenile individuals causes a greater decrease in the average size class of the dinosaur communities than the mammalian ones (Figs. 1, 3), the average size class of the average dinosaur community is still statistically significantly larger than that of the average mammal community. Including the independent young as separate functional taxa (Figs. 1, 4) still shows a marked difference between dinosaurian and mammalian communities.
Fig. 2 and Table 4 present the average number of functional “species” under the three scenarios. Using only adult taxa, mammalian communities consistently show a higher species richness (count of species) than those of dinosaur communities, as shown previously by Benson et al. (2016) and Benson (2018). However, when all juveniles are included as separate taxa this distinction is eliminated. Under this scenario, the average “species” counts in these communities are statistically indistinguishable.
While the inclusion of all juvenile stages might be a more accurate representation of the species richness from the point of view of carnivores, it does not likely reflect the autecology of the representative components of these ecosystems. Extended maternal care in crown-mammals indicates that in terms of feeding, locomotion, and protection, the offspring of mammals are effectively extensions of the mother and thus represent the same functional species. In contrast, once dinosaurs are living independently of their parents they would have been functionally different species at each subsequent growth stage. Under the independent young as separate functional taxa scenario, “species” richness (both mean and median) is statistically higher for Mesozoic dinosaurian communities than in Cenozoic mammalian ones (Figs. 2, 4; Table 2).
DISCUSSION
Assuming the analytical results have some merit in reflecting the actual community of fossil dinosaurs and mammals, what might these mean for ecosystems of the Jurassic and Cretaceous compared to that of the Cenozoic? The preliminary nature of this study represents a simplified approach to the palaeoecology of Mesozoic and Cenozoic terrestrial communities. More advanced approaches, such as including size 1 taxa, incorporating census data and survivorship curves to better approximate the distribution of the biomass among these communities, utilizing morphological indicators of ecological roles, and other additional data would improve the rigor of these analyses. Nevertheless, this initial set of analyses does suggest that the distinction between extended maternal care of mammals and the ontogenetic niche shift of dinosaurs might result in the possibility that dinosaur communities had functionally higher “species” richness than mammalian ones, at least at the body sizes examined here.
One possible explanation lies in changes in terrestrial net primary productivity. Modeling by Beerling (2000) found that net primary productivity (measured in Gt C a-1) was approximately twice as high in the Jurassic and Cretaceous than in the modern world. Previous work (Janis et al., 2000) supports a link between species richness and net primary productivity in terrestrial vertebrate ecosystems. A greater net primary productivity could support a higher diversity of species (whether true taxonomic species or their ontogenetic functional equivalent) for a given area compared to a lower one.
Alternatively, potentially a lower metabolic rate in dinosaurs than in comparable-sized mammals might also allow a greater functional diversity for the same given area. Previous modeling (Matsukawa et al., 2006; Farlow et al., 2010, 2023) found that population densities of both herbivorous and carnivorous dinosaurs were higher in dinosaurian communities than in Cenozoic mammalian ones, which was attributed in part to potentially lower metabolic rates of the dinosaurs.
Should these results hold with the inclusion of additional communities, there are additional hypotheses to be examined. For instance, are there differences in the size distributions of dinosaurian communities in which diverse non-carnivorous maniraptoriform coelurosaurs (Cau, 2024) or abelisauroids (Cau & Paterna, 2025) are present compared to those in which the non-carnivorous niches are occupied only by ornithischians and/or sauropodomorphs? How might the results be significantly different if juvenile dinosaurs of gregarious species remained with adults during their entire lifespan, rather than having a period of juvenile-only aggregations as proposed by Varricchio (2011)? To what degree if any did the spread of open grasslands in the midlate Cenozoic and the rise of large ungulate herds affect the size distributions compared to forest assemblages?
CONCLUSIONS
Despite having some broad morphological and ecological similarities to mammals, dinosaurs were not strictly scaled (and feathered) analogs to their distant amniote relatives. Comparison of species richness and body sizes of medium-to-large sized dinosaur and mammalian communities show that overall species counts of adult Mesozoic faunal assemblages had lower diversity than Cenozoic ones, using adult body size. Extended maternal care in mammals suggests that adult body size may be an accurate reflection of the distribution of the niches exploited within these communities. In contrast, many non-avian dinosaurs may have undergone substantial ontogenetic niche shifts during their life history, allowing each taxonomic species to occupy multiple functional roles within the same community. Under this scenario, functional “species” richness in Mesozoic communities seem to exceed those of Cenozoic ones. Whether this was enabled by higher net primary productivity in the Mesozoic, by lower metabolic rates in dinosaurs than mammals, by a combination of these factors, or other phenomena is beyond the scope of this initial study.
ELECTRONIC SUPPLEMENTARY MATERIAL
This article contains electronic supplementary material which is available at: https://doi.org/10.3301/IJG.2026.09.