The impacts of Cenozoic climate and habitat changes on small mammal diversity of North America

Highlights

• As climate and habitats changed through time, so did rodent and lagomorph community structure.

• Crown heights and open-habitat locomotor adaptations increased through time.

• Changes parallel those of ungulates, but started millions of years earlier.

Abstract

Through the Cenozoic, paleoclimate records show general trends of global cooling and increased aridity, and environments in North America shifted from predominantly forests to more open habitats. Paleobotanical records indicate grasses were present on the continent in the Eocene; however, paleosol and phytolith studies indicate that open habitats did not arise until the late Eocene or even later in the Oligocene. Studies of large mammalian herbivores have documented changes in ecomorphology and community structure through time, revealing that shifts in mammalian morphology occurred millions of years after the environmental changes thought to have triggered them. Smaller mammals, like rodents and lagomorphs, should more closely track climate and habitat changes due to their shorter generation times and smaller ranges, but these animals have received much less study. To examine changes in smaller mammals through time, we have assembled and analyzed an ecomorphological database of all North American rodent and lagomorph species.

Analyses of these data found that rodent and lagomorph community structure changed dramatically through the Cenozoic, and shifts in diversity and ecology correspond closely with the timing of habitat changes. Cenozoic rodent and lagomorph species diversity is strongly biased by sampling of localities, but sampling-corrected diversity reveals diversity dynamics that, after an initial density-dependent diversification in the Eocene, track habitat changes and the appearance of new ecological adaptations. As habitats became more open and arid through time, rodent and lagomorph crown heights increased while burrowing, jumping, and cursorial adaptations became more prevalent. Through time, open-habitat specialists were added during periods of diversification, while closed-habitat taxa were disproportionately lost in subsequent diversity declines. While shifts among rodents and lagomorphs parallel changes in ungulate communities, they started millions of years earlier than in larger mammals. This is likely a consequence of the smaller mammals' greater sensitivity to environmental changes and more rapid evolution. These results highlight the importance of examining understudied members of vertebrate faunas for understanding the evolution of terrestrial communities through time.

Introduction

The transition from the closed forests of the early Eocene to today's domination of open habitats is a story of ecological transformation that has occupied generations of paleontologists studying the Cenozoic terrestrial fossil record. Originally inspired by the dramatic evolutionary transformation of horses in North America (Marsh, 1876, Osborn, 1918, Matthew, 1926, Stirton, 1940, Simpson, 1951, MacFadden, 1992, MacFadden, 1998), studies of the spread of grasslands have given us a picture of the many coevolving pieces of the terrestrial ecosystem. North America, in particular, has been a laboratory for ecosystem evolution during global change, and records of paleobotanical remains, both macroscopic (Wing, 1998, Graham, 1999, Jacobs et al., 1999) and microscopic (Strömberg, 2002, Strömberg, 2004, Strömberg, 2011, Edwards et al., 2010), and of large mammals (Webb, 1977, Janis, 1984, Janis, 1993, Janis et al., 2002a, Janis et al., 2004, MacFadden, 2000, Jernvall and Fortelius, 2002, Strömberg, 2006) offer a picture of macroevolutionary dynamics. Missing from the existing picture, though, is the concomitant evolutionary history of the small mammals in these ecosystems.

Global and regional paleoclimate records from a variety of proxies show a general trend of cooling and increased aridity through the last 50 million years, punctuated by several warmer and wetter intervals (Fig. 1, Retallack, 2001, Retallack, 2007, Retallack, 2013, Zachos et al., 2001, Zachos et al., 2008). Changes in climate correspond with large scale environmental changes, shifting from predominantly forests to more open habitats over much of the continent (Leopold et al., 1992, Jacobs et al., 1999, Retallack, 2007, Dillhoff et al., 2009, Strömberg, 2011). A well-documented history of environmental change in the Great Plains shows a transition from predominantly forest habitats in the Eocene, to mosaic woodlands and savannas in the late Eocene and Oligocene, grasslands in the early Miocene, savanna woodlands in the middle Miocene, and open grasslands in the late Miocene (Wing, 1998, Jacobs et al., 1999, Retallack, 2007, Strömberg, 2011).

Many studies have examined large mammal communities over the last 50 million years. The response of large herbivorous mammal faunas (ungulates) to the environmental transition toward more open, grass-dominated habitats involved shifts toward increased tooth crown height (hypsodonty) combined with adaptations for running (cursoriality) in many species (Janis et al., 2002a, Janis, 2007, Damuth and Janis, 2011, Strömberg, 2011, Jardine et al., 2012, Levering et al., 2017). High-crowned (hypsodont) teeth are a common adaptation to accommodate increased rates of tooth wear, either due to feeding on abrasive or tough plant matter (like grasses or wood) or exogenous grit (dust), which is commonly found in open or arid environments (Stirton, 1947, Van Valen, 1960, Janis, 1988, Janis and Fortelius, 1988, Williams and Kay, 2001, Mendoza and Palmqvist, 2008, Damuth and Janis, 2011, Hummel et al., 2011, Jardine et al., 2012, Kaiser et al., 2013, Damuth and Janis, 2014, Lucas et al., 2014, Hoffman et al., 2015). A recent study also found higher-crowned teeth restrict stress within teeth and increase shearing pressure during mastication, offering a biomechanical advantage when feeding on mechanically resistant foods such as grass (DeMiguel et al., 2016). As environments become more open and arid, grasses commonly become the most abundant food resource, thus the tooth crown height of taxa in a community reflects the relative openness and/or abundance of grass in these habitats (Janis et al., 2002a).

Analyses of large herbivore communities, primarily from the Cenozoic Great Plains of North America, exemplify many important ecological and evolutionary patterns and processes. These studies have revealed how species and communities react to climate and habitat change, both through convergent evolution of similar adaptations in unrelated groups and changes in the diversity of browsing and grazing groups that correspond with changes in vegetation and habitat. However, studies of large herbivorous mammals also show a significant temporal mismatch between environmental changes and the response of ungulate faunas in North America (Wang et al., 1994, Jernvall et al., 1996, MacFadden, 2000, Janis et al., 2002a, Janis, 2007, Strömberg, 2006, Strömberg, 2011, Mihlbachler et al., 2011). Evidence from both paleosols and phytoliths indicate a transition toward more open, grass-dominated habitats millions of years before ecomorphological changes in ungulates (Retallack, 2001, Retallack, 2007, Strömberg, 2002, Strömberg, 2004, Strömberg, 2005, Strömberg, 2011, Edwards et al., 2010). These studies suggest that open grass-dominated habitats predominated by 23 Ma (late Arikareaan) in the Great Plains region; this is well before the radiation of higher-crowned ungulates around 18 Ma (late Hemingfordian) and the appearance of the majority of specialized grazing ungulates about 14 Ma (late Barstovian) (Janis et al., 2002a, Damuth and Janis, 2011). A recent study of mesowear in equids found an increase in average mesowear score during the early Miocene (23 to 19.4 Ma), without a corresponding increase in crown height (Mihlbachler et al., 2011). That suggests that those larger herbivores were exploiting open, grass-dominated habitats as soon as they became widespread, becoming subject to a selective regime prior to the evolution of obvious dental adaptations (Mihlbachler et al., 2011).

At the same time as ungulates evolved high-crowned teeth, their limb morphology had evolved to meet the demands and opportunities offered by the new habitat structure. Studies of limb bones in Cenozoic large mammals indicate an increase in cursorial ungulates from the late Oligocene through the Miocene (Janis and Wilhelm, 1993, Carrano, 1999, Janis et al., 2002b, Levering et al., 2017), concurrent with the spread of open habitats. Large predators evolved during this time as well; mammalian carnivores increased in size and ecological diversity through the Oligo-Miocene (Van Valkenburgh, 1991, Van Valkenburgh, 1999, Wesley-Hunt, 2005). While some late Oligocene amphicyonids display cursorial adaptations (Hunt, 2009, Hunt, 2011), most large predators remained ambush specialists until the late Miocene and through to the Pleistocene, when pursuit predators diversified (Janis and Wilhelm, 1993, Andersson, 2005, Figueirido et al., 2015). This delay offers more evidence for a degree of evolutionary disjunction between predators and prey, rather than the tight coevolutionary story told in the past (Bakker, 1983, Janis and Wilhelm, 1993).

These and other studies have shown a history of ecomorphological evolution in large mammals, but complexities in the relationship between ecological evolution and environmental change (Janis et al., 2002a, Strömberg, 2006) suggest that there is a need to consider small mammals in order to understand the evolution of terrestrial ecosystems through time. The temporal discontinuity between changes in environments and the response of larger mammals, with adaptations appearing millions of years after the appearance of open habitats and grasses (Edwards et al., 2010, Strömberg, 2011), raises the question of whether larger mammals closely track environmental and climatic changes, or if their response is delayed by their relatively high resource requirements and dispersal abilities. Ecosystems include many smaller mammalian herbivores, specifically rodents and lagomorphs, which occupy the same habitat and use the same food resources as ungulates. Janis et al. (2008a) previously noted that dental and locomotor changes in rodents and lagomorphs were more synchronous with habitat changes in the early Miocene.

Rodents first appeared in North America by the Late Paleocene (Clarkforkian) and then radiated extensively (Jepsen, 1937, Korth, 1994, Janis et al., 2008a). Lagomorphs first appear in the Late Eocene (Uintan) and while they are widespread and relatively common in North American faunas, they have had a comparably low diversity throughout the Cenozoic, never > 18 species. Since the Late Eocene, rodent and lagomorph diversity has been considered steady at about 40 to 50 genera (Korth, 1994, Janis et al., 2008a). Despite this relatively stable number of genera, there has been substantial taxonomic turnover and ecological diversification over time. Estimated divergence dates based on molecular data point to a diversification of rodents in the Paleogene, with all extant families established by the end of the Oligocene (Fabre et al., 2012). While evidence points to diversification of sciurids and Castorimorpha (castorids and geomyoids) in the Oligocene, the majority of radiations (shifts in diversification rates) which led to extant diversity took place in the late Miocene around 10 Ma.

Rodents and lagomorphs have a diverse array of diets, and rodents are known for their opportunistic omnivory (Landry, 1970), a reputation likely based on the diversity of omnivorous rats and mice (Muroidea). However, among rodents there are specialized insectivores, carnivores, and many herbivores (Schmidt-Kittler, 2002, Evans et al., 2007, Samuels, 2009, Price et al., 2012). Extant herbivorous rodents include taxa that feed on a wide range of tough or fibrous plant matter, including grasses (ex. cricetids like Microtus), roots (ex. geomyids like Geomys), and wood (ex. Castor). These highly herbivorous taxa display high-crowned teeth (Stirton, 1947, Williams and Kay, 2001), as do many rodent and lagomorph species specialized for life in open habitats, where small mammals are more likely to consume exogenous grit (Bair, 2007, Jardine et al., 2012, Tapaltsyan et al., 2015). Unlike ungulates (with the exception of some meridungulates), multiple clades of rodents and lagomorphs have independently derived hypselodont (ever-growing) cheek teeth. Hypselodonty allows continuous replenishment of tooth structures throughout an animal's life, and as a result would be expected to facilitate herbivorous rodent and lagomorph consumption of abrasive materials (Tummers and Thesleff, 2003, Vianey-Liaud and Michaux, 2003, Damuth and Janis, 2011, Tapaltsyan et al., 2015). Herbivorous rodents tend to show higher rates of tooth wear than browsing ungulates, more like mixed feeders (Damuth and Janis, 2014). Damuth and Janis (2014) found hypselodont Glires show rates of tooth wear an order of magnitude greater than taxa with rooted molars, and suggested that this was not explicable based on differences in dietary abrasiveness or soil consumption, but might be explained by differences in tooth function, enamel hardness, or feeding behavior.

In addition to dietary diversity, Glires display a wide range of locomotor ecologies. Among extant rodents and lagomorphs burrowing (fossorial), jumping (saltatory), and running (cursorial) locomotor habits are observed in multiple clades (Stein, 2000, Samuels and Van Valkenburgh, 2008). Many studies have documented how the locomotor habits of mammals are reflected in their postcranial morphology, and used that to morphology infer the habits of extinct taxa (ex. Van Valkenburgh, 1987, Samuels and Van Valkenburgh, 2008, Hopkins and Davis, 2009, Samuels et al., 2013). Cursorial and burrowing adaptations facilitate movement and survival in open habitats, and their evolution has been shown in some groups to reflect the transition from forests to more open environments (Janis et al., 2002b, Samuels and Van Valkenburgh, 2009, Calede et al., 2011).

Given the differences in the life histories of many small and large species (i.e. shorter generation times, more localized populations, shorter dispersal distances, smaller resource requirements), smaller mammals are likely to respond more rapidly to environmental changes than larger species (Smith et al., 1998, Bowman et al., 2002, Barnosky et al., 2004, Yom-Tov and Yom-Tov, 2004, Renaud et al., 2005, Bofarull et al., 2008, Janis et al., 2008a, Wolf et al., 2009, Finarelli and Badgley, 2010, Badgley and Finarelli, 2013, Badgley et al., 2014), making them very informative as to the timing and magnitude of environmental changes over the Cenozoic (e.g. Millien and Damuth, 2004). Smaller mammals should also respond to smaller-magnitude climate and habitat perturbations than larger mammals and evolve in situ, while larger taxa would be buffered from those changes by their ability to range over a larger area and to make use of sparsely distributed microhabitats (Raia et al., 2012a, Fortelius et al., 2014).

While our expectation is that small mammals will respond more directly, rapidly, and locally to environmental change on the basis of their smaller resource bases and home ranges, it is worth noting that some researchers (Liow et al., 2008, Liow et al., 2009) have suggested that origination and extinction rates are lower in Glires and Eulipotyphlans than in larger mammals, a phenomenon they explain with the argument that Sleep-or-Hide (SLOH) behaviors like burrowing act to buffer species somewhat against extinction. While this argument may apply to patterns of taxonomic diversity, it would still be reasonable to expect more rapid morphological responses to changing environments in small mammals given the finer scale at which they experience the environment, regardless of taxonomic rates.

The Cenozoic trend toward cooling, drying, and increased openness of habitats in North America (Fig. 1, Retallack, 2001, Retallack, 2007, Zachos et al., 2001, Zachos et al., 2008, Edwards et al., 2010, Strömberg, 2011) predicts evolutionary adaptations in the small mammals inhabiting those habitats. Dietary and locomotor adaptations of mammals should reflect the decreasing plant cover, changing vegetation types and food availability, and corresponding changes in soil types. In general, mammal communities should show increased tooth crown heights in response to the spread of more open and arid environments and increased dominance of grasses in ecosystems. Open environments should also favor adaptations for cursoriality, saltation, and fossoriality as small mammals respond to the increased predation pressure brought on by the loss of trees and bushes that provide refuge in closed environments.

We predict that, as environments became more open through the Cenozoic, small herbivorous mammals would opportunistically exploit small patches of open habitats and show rapid adaptations for life within new habitat types. A shift toward communities with more high-crowned taxa and locomotor types more typical of open habitats is expected during the initial opening of habitats in the Oligocene, followed by a possible reversal during the wetter, more closed habitats of the Middle Miocene, and continuation of open-habitat adaptation through the long term cooling and drying of the late Miocene. This is supported by a number of studies that have documented burrowing adaptations and increased crown height within multiple lineages of rodents in the middle Oligocene, and the appearance of highly fossorial (subterranean) taxa in the late Oligocene (Martin, 1987, Hopkins, 2005, Hopkins, 2008, Gobetz and Martin, 2006, Samuels and Van Valkenburgh, 2009), millions of years before open habitat adaptations are observed among ungulates.

Until recently, the response of small mammal communities to Cenozoic climate and habitat changes in North America had not been well studied. Several recent studies have examined the crown height evolution of Glires from the Cenozoic of North America (Jardine et al., 2012, Tapaltsyan et al., 2015), but these studies do not have a comprehensive geographic, taxonomic, or temporal scope. Jardine et al. (2012) directly compared patterns of crown height evolution and species richness for herbivorous mammals (ungulates, proboscideans, xenarthrans, rodents, and lagomorphs) that lived in the Great Plains region of the United States from the middle Eocene to early Pleistocene (39.6 to 1.9 Ma). Tapaltsyan et al. (2015) studied genera of rodents that lived across North America from the early Eocene to early Pleistocene (50 to 2 Ma). They examined generic richness of rodents and percent composition of crown height classes through time in the context of environmental change and produced a mathematical model of crown height evolution. Both of those studies binned occurrence data into a serious of divisions based on the North American Land Mammal Ages, with 23 bins used in Jardine et al. (2012) and 31 bins used in Tapaltsyan et al. (2015). Those bins represent variable amounts of time, ranging from 0.5 to 5 million years.

Jardine et al. (2012) found the initial appearance of high-crowned Glires and ungulates in the Great Plains preceded the spread of grasslands by several million years, and the appearance of highly hypsodont ungulates and hypselodont rodents post-dated the spread of grasslands. They suggested that this indicated high-crowned teeth in both clades were not a simple adaptation to grass consumption, but rather an adaptation to accommodate the ingestion of exogenous grit (Jardine et al., 2012). Tapaltsyan et al. (2015) found a steady and progressive increase in higher-crowned taxa through time. Hypselodonty showed a pronounced increase in the late Miocene, at the same time hypsodont ungulates started to dominate communities (Jardine et al., 2012, Tapaltsyan et al., 2015). They suggested that the steady change in crown height through time indicated higher-crowned teeth were constantly advantageous in the face of consuming abrasive material, an expected consequence of cooling and increased aridity through the Cenozoic (Tapaltsyan et al., 2015).

This study covers a broader scope than any previous work, examining the complete fossil record of Glires (rodents and lagomorphs) throughout their history in North America, from the late Paleocene to recent. Species diversity, tooth crown height, and locomotor habits are examined through time, and compared to global and regional records of climate and habitat change. The time bins used in this study represent equal amounts of time, each 0.5 Ma in length, in contrast to the coarser, variable-length time bins based on the North American Land Mammal Ages used in other recent studies (Jardine et al., 2012, Tapaltsyan et al., 2015). Examination of the ecology of Glires through time allows determination of when and how small mammal communities responded to past environmental changes. Comparison to the record of large mammals will reveal the timing and degree of response of both large and small mammals to past climate and habitat changes.