Strontium isotopes support low mobility for Lemuriformes

Brooke E. Crowley Laurie R. Godfrey Madagascar 4th largest island on the planet. Highly variable climate and topography lead to quite diverse biomes. Flora and fauna found nowhere else on the planet

Euplerid carnivores Tenrecs

Lemurs

Chameleons Arthropods Who are the lemurs?

>100 species. Diversity of ecologies but all are relatively small-bodied and weigh <10 kg. Who were the lemurs?Who were the lemurs? Palaeopropithecus ~41.5-46 kg Megaladapis Palaeopropithecus ~41-46 kg ~46-85 kg Mesopropithecus Megaladapis 10-14 kg ~46.5-85 kg

MegaladapisHadropithecus ~35 kg Palaeopropithecus ~41-46 kg ~46-85 kg

PachylemurIndrisLemur ~7 kgillustrations used with permission from Stephen Nash/ CI International. 11.5-13.5 kg Hadropithecus ~35 kg Microcebus <100 g Archaeoindris IndrisInrdriLemur ~7~7 kgkgillustrations used with permission from Stephen~162 kg Nash/ CI International.

Hadropithecus ~35 kg Microcebus <100 Archaeoindrisg ~ 162 kg ArchaeolemurArchaeolemur ~18-26~18-26.5 kgkg Jungers et al. 2008; Images by Stephen Nash/CI International Sussman 2003; Jungers et al. 2008; Images by Stephen Nash/ CI International Jungers et al. 2008; Images by Stephen Nash/ CI International

Archaeoindris ~ 162 kg Archaeolemur ~18-26 kg

Jungers et al. 2008; Images by Stephen Nash/ CI International Who were the lemurs? Palaeopropithecus spp.

• Understanding the ecological roles that now-extinct species played is critical for making conservation and management decisions regarding Madagascar’s remaining biodiversity.

• We have learned a lot about the now-extinct taxa over the years, including estimated body masses, locomotion patterns, and reconstructed diets.

• However, we do not have a good handle on how mobile they may have been.

Megaladapis spp. Expectations and Predictions

Mobility

Body Size Smaller-bodied Larger-bodied Habit/locomotion Arboreal Terrestrial Diet Folivores; “generalist” herbivores Frugivores; omnivores Habitat Dense/closed forest Open/arid habitat

Among mammals (and anthropoid primates specifically), there are a number of factors that affect home ranges. Body size is the primary control but factors are likely inter-related.

Reviewed in Milton and May (1976) Nature Available Taxa Eight lemur genera are particularly well-represented in subfossil deposits

Extant Extinct Propithecus Megaladapis 3.0-6.5 kg 46.5-85 kg

Eulemur Palaeopropithecus 1.1-2.5 kg 41.5-46 kg

Lemur Archaeolemur ~2.2 kg 18-26.5 kg

Lepilemur Pachylemur 0.5-1.0 kg 11.5-13.5 kg

Smith and Jungers (1976); Jungers et al. (2008); Godfrey et al. (2006); Muchlinksi et al. (2011); Godfrey et al. (2012) AJPA Expectations and Predictions Mobility

Propithecus Lemur Lepilemur 3.0-6.5 kg ~2.2 kg 0.5-1.0 kg 2.4-9.0 ha 5.7-7.4 ha 0.24 ha Eulemur 1.1-2.5 kg 0.88 ha

Based on body size, diet, and habit, we would expect relative mobility to vary slightly among our four extant taxa. Observed home ranges align reasonably well with these expectations

Reviewed in Milton and May (1976) Nature Expectations and Predictions Mobility

Propithecus Lemur Lepilemur 3.0-6.5 kg ~2.2 kg 0.5-1.0 kg 2.4-9.0 ha 5.7-7.4 ha 0.24 ha Eulemur 1.1-2.5 kg Megaladapis 46.5-85 kg 0.88 ha Palaeopropithecus Based on trends observed for primates as a 41.5-46 kg whole, we might expect that now-extinct Archaeolemur large-bodied lemurs had larger home ranges 18-26.5 kg Pachylemur than extant smaller-bodied taxa. 11.5-13.5 kg

Reviewed in Milton and May (1976) Nature Expectations and Predictions Mobility

Propithecus Lemur Lepilemur 3.0-6.5 kg ~2.2 kg 0.5-1.0 kg 2.4-9.0 ha 5.7-7.4 ha 0.24 ha Eulemur Megaladapis 1.1-2.5 kg 46.5-85 kg 0.88 ha Palaeopropithecus 41.5-46 kg Expectations change if we consider diet and postcranial Archaeolemur 18-26.5 kg anatomy/locomotion. Pachylemur 11.5-13.5 kg Godfrey et al. (2006) in Lemurs: Ecology and Adaptation; Walker et al. (2008) Evol Anthropol; Hogg et al. (2015) PLoS ONE Expectations and Predictions Mobility

Propithecus Lemur Lepilemur 3.0-6.5 kg ~2.2 kg 0.5-1.0 kg 2.4-9.0 ha 5.7-7.4 ha 0.24 ha Eulemur Megaladapis 1.1-2.5 kg 46.5-85 kg 0.88 ha Palaeopropithecus vidence that extinct taxa were relatively inactive: 41.5-46 kg - Brain size Archaeolemur - Semicircular canal radii 18-26.5 kg - Retzius periodicity Pachylemur 11.5-13.5 kg Godfrey et al. (2006) in Lemurs: Ecology and Adaptation; Walker et al. (2008) Evol Anthropol; Hogg et al. (2015) PLoS ONE Expectations and Predictions

But… multiple lines of evidence that all extinct taxa were relatively inactive - Brain size - Semicircular canal radii - Retzius periodicity

We can use 87Sr/86Sr to compare mobility among co-occurring taxa. More mobile taxa should have more variable strontium isotope ratios (87Sr/86Sr). Strontium Isotopes (87Sr/86Sr)

• Strontium (Sr) isotopes (87Sr/86Sr) in animal tissues can track foraging and movement of individuals across climatically homogenous landscapes (unlike C, N, O, or H isotopes). • Primarily affected by geology. 86Sr is non-radiogenic; 87Sr is produced by decay of 87Rb. Abundance of 87Sr is thus determined by the age and original 87Rb content of parent rock. Low 87Sr/86Sr High 87Sr/86Sr 0.70 0.80 Younger Older Mafic igneous (e.g., basalt) Felsic igneous (e.g., granite) Carbonates Clastics (e.g., shale) Sr2+ is incorporated into plants and animal skeletal tissues with minimal fractionation (substitutes for Ca2+).

As rocks weather, they release Sr2+ into soils and water. Madagascar is ideally suited for a Sr-based approach

Geologically complex! > 3 billion years of time preserved in the island’s rocks Proximal geologies differ in chemistry and age.

Currently only two sites have reasonably robust sample size for both extant and extinct lemur genera.

N

Adapted from Roig et al. (2012) Ampasambazimba

10 km Madagascar is ideally suited for a Sr-based approach

Ankilitelo Geologically complex! > 3 billion years of time preserved in the island’s rocks

Proximal geologies differ in chemistry and age.Sedimentary 10 km Igneous & Metamorphic Quaternary alluvium

Quaternary cuirasse/hardground

Eocene limestone, shaley Limestone Qalv Ankilitelo/Mikoboka map c5-6 Taolambiby Eocene limestone, oolitic limestone, and sandstone e4 Upper Cretaceous basalt, basanite, and sakalavite Upper Cretaceous sandstone, shale, marl, coquina, and chalk Qcs River

Fault N 10 km10 km10 km Mikoboka/Ankilitelo (a series of sinkholes in a limestone plateau) Adapted from Roig et al. (2012) Madagascar is ideally suited for a Sr-based approach Ampasambazimba Geologically complex! > 3 billion years of time preserved in the

island’s rocks 10 km Proximal geologies differ in chemistry and age. Quaternary alluvium

Akm3 Akm2 Miocene to Pleistocene basalt, basanite, and ankaratrite

At18 Miocene to Pleistocene rhyolite, trachyte, phonolite, ignimbrite

Akm2 Ediacaran to Cambrian granite, monzonite, and undifferentiated syenite Ankilitelo Qalv Akm3 Ediacaran to Cambrian stratified granite and syenite

At1 Cryogenien paragneiss and schist At1 At18 Neoarchaean granitic to granodioritic migmatitic orthogneiss Vmpm 10 km River Fault Shear zone N Ampasambazimba 10 km Ampasambazimba (wetland underlain by a variety igneous and

metamorphic rocks) Adapted from Roig et al. (2012)

Taolambiby 10 km

Ankilitelo 10 km10 km

10 km

Taolambiby

10 km Results

Mikoboka/Ankilitelo

c2 =7.02, df = 6, p = 0.32 Palaeopropithecus There are no differences in variance or median Bartlett = 0.33 87Sr/86Sr between extinct and extant taxa. Sr 86

Sr/ A single Palaeopropithecus has an elevated 87 87Sr/86Sr compared to other individuals. It is an outlier for the genus, but not the Extinct lemur

Extant Extinct group.

c2 =0.09, df = 1, p = 0.76 Lemur Eulemur

Bartlett = 0.14 Lepilemur Propithecus Megaladapis Archaeolemur Palaeopropithecus

Asterisks (*) = non-Ankilitelo sites on the Mikoboka plateau. Results

Mikoboka/Ankilitelo

c2 =7.02, df = 6, p = 0.32 Palaeopropithecus Bartlett = 0.33 Sr 86 Sr/ 87

Extant Extinct

c2 =0.09, df = 1, p = 0.76 Lemur Eulemur

Bartlett = 0.14 Lepilemur Propithecus Megaladapis Archaeolemur Palaeopropithecus

Asterisks (*) = non-Ankilitelo sites on the Mikoboka plateau. Results Excluding the Palaeoprpithecus Mikoboka/Ankilitelo does not affect results.

c2 = 9.75, df = 6, p = 0.14 Palaeopropithecus Bartlett p = 0.41 Sr 86 Sr/ 87

Extant Extinct

c2 = 0.51, df = 1, p = 0.47 Lemur Eulemur

Bartlett p = 0.12 Lepilemur Propithecus Megaladapis Archaeolemur Palaeopropithecus

Asterisks (*) = non-Ankilitelo sites on the Mikoboka plateau. Ampasambazimba

10 km

Ankilitelo

10 km

Results Qalv c5-6 Taolambiby Ankilitelo/Mikoboka Mikoboka/Ankilitelo Qcs map e4

c2 =7.02, df = 6, p = 0.32 Palaeopropithecus Bartlett = 0.33 10 km 10 km10 km Sr 86 Sr/ 87

Extant Extinct Lemur Eulemur Lepilemur

87 86 Propithecus

Sr/ Sr for all individuals Megaladapis Archaeolemur

consistent with local geology. Palaeopropithecus Outlier Palaeopropithecus could have foraged on Quaternary seds. It may have fallen into the sinkhole on its own, or it may have been dropped by a predator(e.g., now-extinct eagle). Ampasambazimba

10 km

Ankilitelo

10 km

J.P. Wallace et al. Results Qalv Taolambiby c5-6 Precipitation & dustMikoboka/Ankilitelo Qcs e4

c2 =7.02, df = 6, p = 0.32 Palaeopropithecus Bartlett = 0.33 10 km 10 km10 km Sr 86 Sr/ 87

Sea spray Atmospheric dust and aerosols Surface geology Bedrock Extant Extinct

Fig. 3. Schematic showing anticipated sources of bioavailable strontium in Miocene Florida. The influence of marine-derived strontium should be highest closest to Lemur 87 86 87 86 the coast. Vegetation rooting depth (arrows) mayApparent differences in also influence Sr/ Sr; whereas weathering bedrock should beSr/ the dominantSr among source for deep-rooted trees, grassesEulemur with shallow root systems may be more susceptible to atmospherically deposited strontium. Lepilemur Propithecus Megaladapis

taxa may reflect spatial partitioning or Archaeolemur direction and the Asiatic wild ass (E. hemionus) in the southwest Gobi Widga et al., 2010). Studies on older taxa are more rare and have fo- 2 Desert has a home range between 18,186 and 69,988consumption of different foods. km (reviewed in cused on Miocene hominids (e.g. Copeland et al., 2011). Palaeopropithecus Ransom and Kaczensky, 2016). We are only aware of three studies that have used strontium iso- Few studies have explored the mobility of Miocene equids. It has topes to investigate the mobility of extinct horses. As mentioned above, been postulated that a small early Miocene equid, Parahippus sp., was 87Sr/86Sr suggest Anchitherium aurelianense had a small home range in relatively sedentary based on modern forest-dwelling artiodactyls with southern Germany (Tütken and Vennemann, 2009). Pellegrini et al. similar body masses and diets (Hulbert, 1984). Conversely, the mid- (2008) measured 87Sr/86Sr for four Pleistocene Equus hydruntinus in Miocene equid, Neohipparion cf. leptode, is thought to have migrated western and central Italy. Low variability in 87Sr/86Sr among in- seasonally, like modern zebras, based on distinct age classes and the dividuals suggests they were also relatively sedentary. Lastly, Hoppe absence of foals within an attritional fossil assemblage (Hulbert, 1982). and Koch (2007) analyzed 87Sr/86Sr for a diversity of Pleistocene her- Lastly, Tütken and Vennemann (2009) analyzed 87Sr/86Sr for a single bivores, including three Equus sp. from Northern Florida. The authors Anchitherium aurelianense from the Middle Miocene in Southern Ger- concluded that one individual likely stayed in Florida while the other many and determined this individual was relatively sedentary, although two individuals traveled ca. 150 km north into Georgia. it may have traveled up to 50 km to the northeast. 2. Methods 1.3.2. Using strontium isotopes to track mobility Strontium isotope ratios in an animal's bones or teeth can be used to 2.1. Regional setting distinguish relatively sedentary from highly mobile individuals, or trace an individual's movement across regions with variable geology (Britton The geographic focus of this study is Northern Florida, USA, which et al., 2009; Hoppe et al., 1999; Hoppe and Koch, 2007; Price et al., has relatively homogeneous bedrock and is known for its abundant 1985). Dispersal from a natal herd, seasonal migration, and daily Miocene fossil deposits. The region is underlain by the Florida Platform, movement within a home territory or range can be tracked using which is composed of a thick sequence of Paleocene to Oligocene car- strontium isotopes. Foraging on different types of vegetation may also bonates (Fig. 1; Scott, 1992). The strontium isotope ratios leached from influence 87Sr/86Sr in herbivores (Crowley et al., 2018; Hedman et al., these rocks should be relatively low and comparable to con- 2009; Maurer et al., 2012; Radloff et al., 2010). Whereas deeper rooting temporaneous seawater: ca. 0.7077–0.7078 for the Eocene and plants gain the majority of their strontium from bedrock, shallow 0.7078–0.7083 for the Oligocene (Fig. 1; McArthur et al., 2001). Sev- rooting plants are more likely to be influenced by atmospheric de- eral hundred kilometres to the north, bedrock is composed of older position (Fig. 3; Graustein and Armstrong, 1983; Reynolds et al., 2012). sediments as well as metamorphic and igneous rocks that date back to This geochemical tool has been used to examine foraging territories the formation and uplift of the Appalachian Mountain range during the and migration patterns for modern herbivores, including African ante- early Paleozoic (Fig. 2). These rocks should have considerably higher lope, zebra, and elephants (Koch et al., 1995; Radloff et al., 2010; Vogel (≥0.7110) and more variable 87Sr/86Sr than limestone (Capo et al., et al., 1990), as well as North American and European caribou (Britton 1998; Hoppe et al., 1999; McArthur et al., 2001). Published 87Sr/86Sr et al., 2011, 2009; Gigleux et al., 2017). Strontium isotopes have also for modern surface water, plants, and rodents from the region reflect been used to reconstruct the mobility of various Quaternary herbivores underlying geology. Samples from northern and central Florida have in Europe and North America, including bison, caribou, horse, deer, strontium isotope ratios that are broadly consistent with those expected mammoth, and mastodon (Baumann and Crowley, 2015; Britton et al., for carbonates (0.7075–0.7092; Hoppe et al., 1999; McArthur et al., 2011; Hoppe, 2004; Hoppe and Koch, 2007; Pellegrini et al., 2008; 2001). In contrast, samples from Georgia have higher and more variable

Ampasambazimba 87Sr/86Sr Bartlett p = 0.24 c 2 = 0.18, Extant df = 1, p = 0.67 Eulemur Extinct No Results Eulemur extinct and extant taxa differences in variance or median outlier for living lemurs, but not for the genus.

Lepilemur

Archaeolemur Bartlett p = 0.57 c

Megaladapis 2 = 2.17, . Single

Palaeopropithecus df = 5, p = 0.83

Pachylemur Eulemur 87 Sr/ is a statistical 86 Sr between Ampasambazimba 87Sr/86Sr Bartlett p = 0.17 c 2 = 1.14, Extant df = 1, p = 0.29 Eulemur Extinct E xcluding Eulemur Results

Lepilemur it does not affect the results.

Archaeolemur

Megaladapis Bartlett p = 0.83 c 2 =2.95,

Palaeopropithecus df = 5, p = 0.71

Pachylemur Ampasambazimba variability. Not surprising given 87Sr/86Sr Considerable intraspecific geologic complexity. Extant Eulemur Extinct

Eulemur Results

Lepilemur

Archaeolemur Ampasambazimba Taolambiby Ankilitelo Bartlett p = 0.57 c

Megaladapis 2 = 2.17,

Palaeopropithecus df = 5, p = 0.83

Pachylemur At18 10 km Akm2 Vmpm Qalv 10 km Akm3 10 km 10 km 10 km At1 Discussion Palaeopropithecus spp. • 87Sr/86Sr data consistent with other lines of evidence suggesting now-extinct taxa were relatively slow and stationary. • Overall, Lemuriformes appear to differ from anthropoid primates in their relatively low metabolisms, small brains, low RP, and reduced activity levels. • May be an adaptation to Madagascar’s harsh environments. • Unfortunately, small home range size also makes lemurs more vulnerable to fragmentation. • Large-bodied frugivores among the first species to disappear. Loss of these key seed dispersers may result in an increased number of “orphaned” plants, which in turn may affect forest regeneration, ecosystem health, and climate stability. • Ensuring remaining forest is protected and connected is critical.

Megaladapis spp. Acknowledgements

A more comprehensive version of this study is published in Frontiers in Ecology and Evolution: www.frontiersin.org/articles/10.3389/fevo.2019.00490/full Specimen Collection and Access: University of Antananarivo Kathleen Muldoon Gregg Gunnell and Catherine Riddle, Division of Fossil Primates, Duke Primate Center Sample Preparation and Analysis: Undergraduate assistants Tom Johnson and Gideon Bartov, Multicollector ICPMS lab, University of Illinois Urbana-Champaign

Lemur illustrations generously provided by Stephen Nash/ CI International