Dietary niche partitioning among fossil bovids in late Miocene C3 habitats: Consilience of functional morphology and stable isotope analysis

Palaeogeography, Palaeoclimatology, Palaeoecology 253 (2007) 529 – 538 www.elsevier.com/locate/palaeo

Dietary niche partitioning among fossil bovids in late Miocene C3 habitats: Consilience of functional morphology and stable isotope analysis Faysal Bibi ⁎ Department of Geology & Geophysics, Yale University, PO Box 208109, New Haven, CT 06520-8109, USA Received 11 May 2007; received in revised form 22 June 2007; accepted 26 June 2007

Abstract Teeth of late Miocene bovids referred to Bovini and “Boselaphini” were subjected to enamel stable carbon and oxygen isotope analysis to test paleoecological reconstructions based on dental morphology. Teeth of Bovini possess derived characters–including larger size, higher crowns, and increased enamel surface area–that are indicative of feeding on a more fibrous and gritty diet, probably grass. In contrast, teeth of “Boselaphini” reflect the plesiomorphic condition among bovids–being smaller, lower-crowned, and with simple occlusal morphology– and are indicative of a diet with a greater reliance on softer food items such as browse. Late Miocene bovines are also expected to have inhabited drier, more open habitats than did boselaphines. Stable carbon and oxygen isotopic compositions from 30 fossil teeth (18 bovine, 12 boselaphine) from well-dated localities of between 8.3 and 7.9 Ma in age from the Siwalik deposits, Pakistan, were analyzed to test these paleoecological hypotheses. All δ13C values (VPDB) are more negative than −8‰, indicating that both bovines and boselaphines at this time had pure C3 diets. The mean δ13C for bovine teeth (−10.4‰) is more positive than that for boselaphines (−10.9‰), and the differences between these two series are significant (Wilcoxon, pb 0.01; t test, p b 0.05) while the variances are not. Early bovines thus appear to have exploited more open habitats than did their boselaphine counterparts. Lack of a significant difference between variances suggests that the dietary niche breadth of early bovines was not different from that of boselaphines. Mean δ18O (SMOW) for bovine teeth (26.3‰) is slightly more negative–as might be expected for grazers–but not statistically significant from the boselaphine δ18O mean (28.1‰, t test, p= 0.066). Overlap in δ18O values between bovines and boselaphines is high, implying that these two bovid types did not differ greatly in their water intake behaviors. Rather, both fossil bovines and boselaphines probably shared similar obligate drinking habits and dependency on water bodies much as living boselaphines, bovines, and tragelaphines (clade Bovinae) do today. Stable isotope analysis results, particularly δ13C values, suggest that in the late Miocene neither bovines nor boselaphines inhabited dense forest habitats. And while both bovid taxa may have been mixed feeders to different extents, the δ13C values support the hypothesis developed on the basis of dental functional morphology that early bovines evolved inhabiting more open habitats than did contemporaneous boselaphines. The scenario whereby the bovine clade owes its origins to a boselaphine lineage that adapted to drier, more open habitats is supported by the general context of climatic and faunal change in Eurasia in the late Miocene, particularly between 11–8 Ma, when faunal assemblages from many sites exhibit significant turnover events through which open-habitat taxa become present in increasing proportions at the expense of closed-habitat taxa. © 2007 Elsevier B.V. All rights reserved. Keywords: Bovidae; Bovinae; Bovini; Boselaphini; Late Miocene; Siwaliks; Grasslands; Diet; Paleoecology

⁎ Fax: +1 203 432 3134. E-mail address: [email protected]. 0031-0182/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.palaeo.2007.06.014

530

F. Bibi / Palaeogeography, Palaeoclimatology, Palaeoecology 253 (2007) 529–538

1. Introduction The earliest representatives of the bovid clade Bovini were recently identified from late Miocene fossil collections from the Indian Subcontinent (Bibi, 2007). The extant crown clade Bovini comprises grazing species such as buffaloes, bison, and cattle. The earliest bovines can be identified by dental remains from the Siwalik deposits, Pakistan, dated at least as far back as 8.9 Ma. These dental remains are distinguished from those of ancestral and contemporaneous boselaphine bovids primarily by their significantly larger size, higher crowns, larger basal pillars, and more complicated enamel outlines (Fig. 1). Taken together, these adaptations reflect an increased intake of rougher, more fibrous or gritty foods in early bovines, such as grasses, leading to the supposition that early bovines evolved exploiting more open habitats than did ancestral and contemporaneous late Miocene boselaphine bovids that lacked these dentognathic adaptations (Bibi, 2007). These inferences based on functional morphology are tested here using stable isotope compositions of tooth enamel to examine the feeding characteristics of late Miocene bovines and boselaphines. At tropical latitudes today most grasses utilize the C4 photosynthetic pathway that promotes survival in seasonally arid or heat-stressed regions (Tipple and Pagani, 2007). However, stable carbon isotopic evidence suggests C4 grasses did not substantially expand in the Siwalik region until after 8.1 Ma, with C4-dominated habitats first appearing by 7.4 Ma (Barry et al., 2002; Quade and Cerling, 1995; Quade et al., 1989). This suggests that if the earliest bovines evolved by exploiting more open habitats, the terrestrial environment must have been dominated by C3 floras, potentially C3 grasslands. Stable

Fig. 1. Examples of late Miocene bovine and boselaphine teeth analyzed in this study. Bovine teeth (left, YPM 19681) are larger, with developed basal pillars, more convoluted enamel ridges, and higher crowns than contemporaneous boselaphine teeth (right, YPM 19613). Both teeth shown are right upper second molars in mid wear stages. Scale bar equals 1 cm.

carbon isotope analysis distinguishes readily between C3 and C4 plants, with C3 plants displaying bulk organic carbon δ13C values between −35‰ and −22‰ and C4 plants between −19‰ and −9‰ (Bender, 1971; Cerling and Harris, 1999). Plant δ13C is an inverse function of the ratio of the partial pressure of CO2 in leaf intercellular spaces (ci) relative to that in the atmosphere (ca) (Farquhar et al., 1982a; Farquhar et al., 1982b) and as a result even within either of the C3 or C4 ranges there is an observed correlation with increasingly enriched δ13Cplant values and increasingly heat- or water-stressed environments (Ehleringer and Cooper, 1988; Farquhar et al., 1989). For example, Medina and Minchin (1980) and van der Merwe and Medina (1991) found that leaves growing in subcanopy conditions had more depleted δ13C values than those growing at canopy level and in clearings. This phenomenon is ascribed to two factors: (1) the ‘canopy effect’ whereby δ13C-depleted CO2, released as a result of decomposition of leaf litter at ground level, is photosynthetically recycled by plants growing nearer to the forest floor (Vogel, 1978); and (2) the observed relationship between plant water-use efficiency and δ13C, whereby unstressed plants growing in shaded/humid/low-salinity conditions will maximize the kinetic (at the stomatal level) and physiological (at the level of the enzyme rubisco) fractionation that discriminates against fixation of the heavier isotope of carbon, 13C (Farquhar et al., 1982a; O'Leary, 1993; van der Merwe and Medina, 1991). Conversely, increases in stressors such as light intensity, salinity, and aridity require plants to combat excessive evaporation by closing leaf stomata (i.e. decreasing stomatal conductance) for longer periods of time, trapping a limited reserve of CO2 in the leaf and resulting in fixation of a greater proportions of the heavier isotope of carbon than would occur under more optimal conditions (Farquhar et al., 1982a; van der Merwe and Medina, 1991). Studies have shown that decreased shade or water availability can effect δ13C changes of about +1– 2‰ in any single plant species (Ehleringer and Cooper, 1988; Michelsen et al., 1996). Changes in the partial pressure as well as the 13C/12C ratio of atmospheric CO2 can also affect the resulting δ13C value of plant material (and hence herbivore tooth enamel). Given that both the partial pressure and 13C/12C ratio of pre-industrial atmospheric CO2 do not appear to have changed significantly since the late Miocene (Pagani et al., 1999; Pagani et al., 2005; Passey et al., 2002), δ13C values of fossil teeth examined in this study are interpretable in reference to δ13C endmember values in modern systems. The δ13C values of plants translate directly to the δ13C values in tissues of the mammalian herbivores that feed on them. Large ruminant enamel bioapatite is enriched in 13C

F. Bibi / Palaeogeography, Palaeoclimatology, Palaeoecology 253 (2007) 529–538

by about +14.1 ± 0.5‰ with respect to the plant source (Cerling and Harris, 1999), with living bovines averaging slightly greater values of approximately +14.6 ± 0.3‰ (Cerling and Harris, 1999; Passey et al., 2005). Bovids feeding on pure C3 diets should exhibit ranges between about −21‰ and −8‰, with open-habitat species having more positive values than forest/canopy species. This is confirmed by studies of living ungulates, such as a study of African bovids (Cerling et al., 2003) which determined that duiker species (Cephalophus spp.), which inhabit rainforests and forests, exhibit much more negative δ13C values than the eland (Tragelaphus oryx), a woodlandsavanna species, though both of these bovids are exclusive C3 feeders. Similarly, contrasts between closed and open-habitat bovids in the ancient record can be tested by evaluating tooth enamel δ13C values of specific species. Within a defined age and region, fossil bovids that inhabited more open habitats should exhibit more positive δ13C values than those that inhabited more closed habitats. Analogous studies comparing δ13C values within the C3 range among various fossil ungulates have been performed with relative success on Miocene herbivore communities from Panama (MacFadden and Higgins, 2004), Turkey (Quade et al., 1995), and Florida and California (Feranec and MacFadden, 2006). The oxygen isotope composition of bovid tooth enamel was analyzed in order to test for differing patterns of water intake between bovine and boselaphine fossil bovids. Due to evaporative enrichment, leaf water δ18O values are typically10–30‰ more positive than the meteoric source (Yakir, 1997). As a result, mammal species that obtain the majority of their water from the plants they eat should exhibit more positive δ18O values than those that are obligate drinkers (e.g. Sponheimer and Lee-Thorp, 1999; Sponheimer and Lee-Thorp, 2001). Sponheimer and Lee-Thorp (1999) found that browsers tend to have slightly more enriched δ18O values than do grazers. On this basis, Miocene fossil boselaphines should be slightly more enriched in 18O than fossil bovines. This expectation is not strong, however, given the water-use habits of the living relatives of these fossils. Living bovines are obligate drinkers with strong affinities to water, exemplified by the water buffalo (Bubalus bubalis), which spends much time fully immersed in water. Among bovine outgroups, the chousinga (Tetracerus quadricornis, a boselaphine) and most tragelaphines (Tragelaphus spp.) are water dependent as well (Estes, 1991; Prater, 1965). Given the behavior of these extant species, water intake behaviors of ancient boselaphines and bovines likely depended more on varying local conditions such as water sources (e.g. seasonal waterholes vs. rivers vs. perennial lakes)

531

and climatic variables (e.g. changes in annual precipitation patterns) rather than clade-specific ecological attributes. This study utilizes stable isotope analysis to test hypotheses and predictions made on the basis of functional morphological differences between teeth of late Miocene bovines and boselaphines from the Siwalik deposits. Primarily, stable isotopes are used to test whether dietary intake differed in any significant way between these two bovid groups. Bovines are expected to display more positive δ13C values, indicating that they fed in more open C3 habitats than did boselaphines, and more depleted δ18O values as noted for grazing herbivores that have a greater reliance on meteoric sources for their water needs. Results of the analysis are discussed with implications for the early evolution of Bovini within the context of late Miocene climatic and paleoenvironmental changes. 2. Methods Thirty fossil specimens were selected for analysis based on their morphology and stratigraphic provenance (Table 1). Eighteen specimens are of early bovines (teeth traditionally referred to Selenoportax or Pachyportax), while twelve are of large non-bovine boselaphine fossils (likely Tragoportax) that in this paper will be referred to simply as boselaphines. Sites chosen and their ages are as follows: L008 (8.331 ± 0.04 Ma), L011 (7.926 ± 0.12 Ma), L012 (7.926 ± 0.12 Ma), L056 (7.926 ± 0.12 Ma), L073 (8.018 ± 0.31 Ma), and L074 (8.018 ± 0.31 Ma) (age determinations from Barry et al., 2002). Enamel was sampled using a Foredom rotary drill at low speeds, fitted with a 33.5-gauge (0.5 mm diameter) inverted cone carbide burr. Between 2000 and 6000 μg of enamel was removed along a single transect from the base to the tip of the crown, typically along either of the buccal ribs (metacone or paracone) on upper teeth or lingual ribs (metaconid or entoconid) on lower teeth, making sure to include no cementum or dentine in the sample. Enamel powder was treated with hydrogen peroxide and acetic acid to remove effects of contaminant organics and inorganic carbonate (Koch et al., 1997). Enamel was treated with 30% H2O2 for between 24 to 48 h, decanted and washed with distilled de-ionized water, then reacted with 0.1 N acetic acid for 4 h, after which samples were decanted and washed. Ethanol was added and samples were dried overnight. The treated enamel was dissolved in 100% phosphoric acid and the resulting CO2 was analyzed with a Gas Bench II coupled to a Thermo Finnigan DeltaPlus XP mass spectrometer at the Earth Systems Center for Stable Isotopic Studies, Yale

532

F. Bibi / Palaeogeography, Palaeoclimatology, Palaeoecology 253 (2007) 529–538

Table 1 Sample identification information and δ13C and δ18O values Sample number FB-10 FB-02 FB-04 FB-08 FB-09 FB-21 FB-23 FB-24 FB-26 FB-12 FB-17 FB-14 FB-25 FB-11a FB-01 FB-05a FB-22 FB-13 FB-18 FB-19 FB-28 FB-32 FB-34a FB-36 FB-15 FB-16 FB-27 FB-29 FB-30 FB-35a

YPM number

Taxon

δ13C (PDB)

19591 20018/19994 19613 20001 20026 19615 19610 49799 20024 19989 19502 49758 19281 19601 20022 20016 20002 19980 19499 19507 19500 19506 19504 19505 19678 19681 19680 19677 19679 19674

“Boselaphini” “Boselaphini” “Boselaphini” “Boselaphini” “Boselaphini” “Boselaphini” “Boselaphini” “Boselaphini” “Boselaphini” “Boselaphini” “Boselaphini” “Boselaphini” Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini Bovini

−10.98 −10.77 −11.62 −11.15 −10.74 −10.11 −11.18 −11.41 −9.81 −11.16 −10.71 −11.23 −11.79 −10.60 −12.21 −10.30 −10.71 −10.50 −9.72 −9.63 −10.25 −10.09 −10.02 −10.53 −9.83 −10.20 −9.70 −9.85 −10.20 −10.68

δ18O (SMOW) 32.64 28.79 25.43 28.26 27.75 27.09 27.72 26.39 28.04 32.16 22.70 29.91 26.44 29.10 22.36 24.00 29.77 31.13 25.31 24.94 25.76 25.99 27.58 24.73 27.26 23.86 24.89 28.64 26.53 25.69

Mass (μg) 4538 4346 4316 2197 2474 1897 1055 1140 2173 3209 2627 2361 4241 3359 3002 3348 1415 4735 4272 4312 3493 3301 2231 3696 3754 2474 4287 5542 4339 4595

Tooth sampled X

M M3 M3 M3 M3 M3 M3 M3 M3 M3 M3 MX M3 M3 M1or2 M3 M1or2 M1or2 M3 M3? M1 M1or2 M3 M3 M3 M2 M1or2 M1 M3 M3

Locality L011 L012 L012 L012 L012 L012 L012 L012 L012 L056 L073 L074 L008 L011 L012 L012 L012 L056 L073 L073 L073 L073 L073 L073 L074 L074 L074 L074 L074 L074

YPM, Yale Peabody Museum. For tooth abbreviations, M denotes molar and superscripts and subscripts denote upper and lower teeth positions, respectively, with x indicating indeterminate tooth position.

University. Standards run with the bovid enamel sample included NBS-18, NBS-19, Lincoln Limestone, and the MEme fossil proboscidean standard. For MEme, average run values of −10.37 for δ13C and 25.75 for δ18O were used (L. R. G. DeSantis pers. comm.). Carbon and oxygen isotope values are reported relative to isotopic standards such that

sample test (equivalent to Mann–Whitney U test) was used in addition. Statistical tests were performed using JMP 6.0 software. Significance was set to α = 0.05.

d13 C or d18 O ¼ ðRsample =Rstandard  1Þ  1000

The total range of variation for all values is just under 3‰, ranging from −12.2‰ to −9.6‰ (Fig. 2, Table 1). The mean δ13C value for all 18 bovine samples is −10.4‰ (range −12.2‰ to −9.6‰) while for the 12 boselaphine samples the mean is −10.9‰ (range −11.6‰ to −9.8‰). Shapiro–Wilk tests reveal that while boselaphine δ 13 C values are normally distributed (p N 0.05), bovine δ13C values are not (p = 0.0058), and so a Wilcoxon test was performed along with the t test. An F test determined that variances between bovine and boselaphine δ13C values were not significantly different. Both the Wilcoxon and t test indicate that bovine and

where Rsample and Rstandard are the 13C/12C or 18O/16O ratios of the sample and standard, respectively. δ13C values are reported relative to the Vienna PeeDee Belemnite (VPDB) standard, while δ18O values are against the Standard Mean Ocean Water (SMOW) standard. Shapiro–Wilk tests were performed on all sample series values to determine whether distributions were normal or not. For all samples, F tests and Student's t test assuming equal variances were used. For nonnormal distributions, the non-parametric Wilcoxon two-

3. Results 3.1. Carbon isotope ratios

F. Bibi / Palaeogeography, Palaeoclimatology, Palaeoecology 253 (2007) 529–538

Fig. 2. Tooth enamel δ13C and δ18O values for all specimens analyzed (A) and δ13C values for all specimens separated by locality (B).

boselaphine δ13C values are significantly different (twosample Wilcoxon, p b 0.01; t test, p = 0.0316). The δ13C values of bovine and large boselaphine teeth separate at −10.7‰ although two bovine samples display δ13C values more negative than −10.7‰, these in fact being the two most negative values recorded among all thirty samples analyzed. Likewise, among the boselaphine samples, two exhibit δ13C values more positive than −10.7‰ though both these samples remain more negative than the most enriched bovine samples (Fig. 2). 3.2. Oxygen isotope ratios The δ18O values for all samples range between 32.6‰ and 22.4‰. Means for the 18 bovine and 12 boselaphine samples are 26.3‰ and 28.1‰, respectively. Shapiro– Wilk tests reveal that boselaphine and bovine δ18O values are each normally distributed (p N 0.05), a two-sided F test between bovine and boselaphine δ18O values indicate that sample variances are not significantly different, and a twotailed t test finds that the difference between the means is not significant (p = 0.066). 4. Discussion 4.1. Carbon isotope ratios—paleodiets In general, ranges of carbon isotope values follow the relationships predicted on the basis of functional mor-

533

phological differences between teeth recognized as early Bovini and teeth attributable to other fossil boselaphines (Bibi, 2007). However, in each series, outliers exist that fall far from their respective series' means. Most of the outlier values derive from locality 12 (Fig. 2B) where there is no clear separation between bovines and boselaphines along the δ13C axis. Locality 12 is the single locality from which the greatest number of samples came. This locality is quite rich in boselaphine material, only a portion of which was sampled for this study, but poor in bovine material (3 specimens only), all of which was included in this study. The relative abundance of boselaphines with respect to bovines may indicate that locality 12 sampled paleohabitats that were more favorable to the former at the expense of the latter. However, it is not clear why this locality should produce outliers in both series, that is, one bovine with very 13Cdepleted values and two boselaphines with relatively 13 C-enriched values (Fig. 2B). The one other outlier is a very 13C-depleted bovine specimen from locality 8, a locality from which no boselaphines were available for comparative analysis. It is important to recognize that though this work evaluates individuals that probably derive from only two higher taxonomic groups (Bovini and “Boselaphini”) it could potentially include three, four, or more species. This is almost certainly the case simply judging by the range in size (particularly among the bovine specimens) and morphology (particularly among the boselaphines) exhibited within the two series. Species-level determinations are currently not possible with any confidence on the isolated dental material used here so this must remain an uncontrolled variable in the current study. As a result, it is also possible that samples with outlying carbon isotope values represent particular species unrecognizable on the basis of dental morphology alone. Further isotopic sampling using more complete fossil material from other collections may help resolve this issue. Despite the coarse taxonomic resolution of this study, bovine and boselaphine dental δ13C values are on average distinctive, with boselaphines exhibiting more 13C-depleted values and bovines more 13C-positive values, with statistically significant differences between these two series. These findings attest to the strength of functional morphology to draw paleoecological conclusions regardless of lower-level taxonomy. Teeth of early bovines are characterized by the acquisition of several advanced characters that indicate an increased ratio of tougher food items in the diet. These characters include increased crown height, the enlargement of pillars (entostyles and ectostylids), the complication of enamel ridges, and greatly increased size (Bibi,

534

F. Bibi / Palaeogeography, Palaeoclimatology, Palaeoecology 253 (2007) 529–538

2007). Increases in crown height are an adaptation to increased rates of wear, and it is known that the consumption of grasses by herbivores produces greater tooth wear than would occur with strict adherence to a diet of leaves primarily because grasses are closer to ground level and so naturally result in greater intake of hard soil particles by the grazer (Mainland, 2003; Sanson et al., 2007). In contrast, teeth of fossil boselaphines, both ancestral to and contemporaneous with Bovini, lack these advanced characters and remain quite primitive in overall morphology. Among living bovids, tragelaphines, species of which are primarily browsers (Tragelaphus spp.), retain the primitive dental morphological condition least changed from that of the common ancestor of all the Bovinae (Bovini +Boselaphini + Tragelaphini). Though early bovines appear to have evolved to handle a diet composed of more abrasive foods, they did so in the early late Miocene (by at least 8.9 Ma, Bibi, 2007), well before the expansion and dominance of grasslands by heat-tolerant C4 grasses which took place at 7.4 Ma in the Siwaliks sequence (Barry et al., 2002). As such, it is most likely that early bovines evolved feeding increasingly on C3 grasses and inhabiting more open C3 grassland environments. The fact that all of the samples analyzed here presented δ13C values more negative than −8.0‰ establishes that bovine diets in the Siwaliks at around 8 Ma were purely C3-dominated, including no significant component of C4 plants. Even within the C3 range, significant differences do exist between δ13C values of bovine and boselaphine fossil enamel, with bovine teeth having slightly higher δ13C. Such enrichment is only to be expected if the bovines were feeding on plants that themselves were slightly enriched in 13C. It has been shown that C3 plant δ13C values become progressively more positive with a change from a sub-canopy mesic to open xeric environments, due to the canopy effect (Medina and Minchin, 1980; van der Merwe and Medina, 1991; Vogel, 1978) and plant water-use efficiency responses to increased heat stresses in open environments (Farquhar et al., 1982a; O'Leary, 1993; van der Merwe and Medina, 1991). As a consequence, the differences in δ13C values between bovine and boselaphine fossil bovids presented in this study indicate that early bovines evolved exploiting more open environments than did boselaphines. The fact that the means of these two series differ by less than 1‰ suggests that the environmental/dietary differences between late Miocene bovines and boselaphines, though distinctive, were not extreme. For example, in a study of mammals from the Ituri Forest (Cerling et al., 2004), sub-canopy dwellers including the dwarf ante-

lope (Neotragus batesi) exhibited an average δ13C value of −22.8‰, more than 5‰ lighter than that of gapclearing inhabitants including sitatunga (Tragelaphus spekei). In contrast, the less than 1‰ separation between δ13C means of late Miocene bovine and boselaphine teeth is comparable to the differences in mean values recorded for species inhabiting similar environments such as lesser kudu (Tragelaphus imberbis) and greater kudu (Tragelaphus strepsiceros) from modern xeric bushlands in Kenya (Cerling et al., 2003), both species being C3 feeders in thicket or riverine forest habitats in arid environments. The actual diets of the fossil bovines and boselaphines and those of these living tragelaphines are not comparable, however, as the fossil taxa may have included significant quantities of C3 grasses in their diets while living tragelaphines feeding on a substantial amount of grass would display δ13C values more positive than −8‰ given that tropical grasses are predominantly C4 plants. It is likely that both the early bovines as well as the fossil boselaphines sampled inhabited some form of open C3 habitat, perhaps open forest or mosaic woodland-grassland environments where the boselaphines would more selectively feed on browse and fresh grass shoots while the early bovines would have an expanded diet that overlapped with the boselaphines but included also more of the tougher grasses. The reconstruction of mixed diets for both these taxa but with relatively greater dietary roughage in Bovini is in concordance also with the results of tooth cusp mesowear analysis (Bibi, 2007; Fortelius and Solounias, 2000). Dietary separation between these two bovid groups may have only been seasonal. Many studies (Bell, 1969; Dekker et al., 1996; Schuette et al., 1998; Traill, 2004) have shown that dietary behaviors among different sympatric African bovids can be remarkably similar during the wet season–when resources are not limiting–but can come to differ greatly with the first onset of the dry season when the food items that were every species' favored food items become scarce or altogether depleted. Bell's (1969) study of feeding behaviors among gazelle (Gazella thomsoni), topi (Damaliscus korrigum), wildebeest (Connochaetes taurinus), zebra (Equus burchelli), and savanna buffalo (Syncerus caffer) found that the larger species were able to include greater quantities of the medium to tall grasses that are less nutritious but more plentiful during the dry season than the more protein-rich short grasses that all these herbivores favored during the wet season. This is largely explained by the fact that larger animals have relatively lower metabolic requirements than smaller ones (Hungate et al., 1959; Bell, 1969). The observation that herbivore

F. Bibi / Palaeogeography, Palaeoclimatology, Palaeoecology 253 (2007) 529–538

body size is inversely correlated with the nutrient content of their food (protein to fiber ratio) has been dubbed the Jarman–Bell principle (Bell, 1971; Codron et al., 2007; Geist, 1974; Jarman, 1974) and has been used to relate body size, population structure and biomass, and habitat preferences in a variety of ungulates. Using these same principles, one can reconstruct the ecological context of late Miocene boselaphines and bovines. With the evolution of added chewing surfaces, increased crown height, and larger size (Fig. 1), early bovines were adapting to pressures selecting for the ability to incorporate more fibrous and less nutritious food items in their diet. 4.2. Paleoclimate and evolution The selective pressures that drove the evolution of grazing morphologies were likely a consequence of physical environmental changes that resulted in the creation of more open habitats at the expense of more closed ones. Numerous studies point to exactly this type of mesic–xeric environmental change during the early late Miocene, between about 11–8 Ma, in the Siwaliks region but also in Eurasia as a whole (Damuth et al., 2002; Fortelius et al., 2002; Fortelius et al., 2006; Franzen and Storch, 1999; Quade and Cerling, 1995). The late Miocene evolution of Bovini within the context of ecological change is consistent within the general picture of faunal turnover in the Siwaliks at the same time. Isotopic evidence suggests shifts towards drier, more seasonal environments at around 9.2 Ma, with C4 grasses beginning to establish by 8.1 Ma, and the first C4-dominated habitats (paleosol carbonate δ13C values greater than −4‰) appearing at 7.4 Ma (Barry et al., 2002; Quade et al., 1989). Barry et al. (2002) note that two of three major episodes of faunal turnover take place around this time, in the period between 8 and 7 Ma. The nature of the faunal changes during this time appear also to be consistent with the isotopic vegetational reconstructions, with increasing proportions of open-habitat taxa at the expense of closed-habitat taxa (Barry et al., 2002). The primary physical climatic change that produced these environmental changes appears to have been increased seasonality (Quade et al., 1989), with stronger precipitational contrasts between wet and dry seasons occurring during this time. In particular, the significance of seasonal intensification in terms of a drier or prolonged dry season that could no longer support the previously greater extent of forested habitats should be emphasized. In concert with increasing aridity, new annual fire regimes may have played an important role in the development of late

535

Miocene open habitats (Bond and Keeley, 2005; Keeley and Rundel, 2005). The evolution of Bovini in the Siwaliks region in the late Miocene then appears to have been part of a greater wave of faunal change and adaptation in response to the development of drier, more seasonal, open habitats. Unlike early bovines, late Miocene boselaphines appear to have responded to these climatic and environmental changes with less ‘drastic’ measures, maintaining what is essentially primitive dentognathic morphology. These fossil boselaphines may have been ecologically analogous to living tragelaphines (such as the greater and lesser kudus mentioned above) that restrict themselves to forested micro-habitats within generally more arid and open environments, thus ensuring a perennial, though spatially restricted, supply of browse-based foods such as leaves, herbs, and fruits. 4.3. Morphological innovation and niche breadth It has been suggested (Bibi, 2007) that the morphological novelties characterizing the earliest Bovini probably resulted in the ability of these bovids to expand their dietary spectrum rather than to simply shift it over, a phenomenon that has been observed in other grazeadapted fossil mammals (Feranec, 2003; Feranec, 2004; Feranec, 2007). To use Hutchinson's (1957) terminology, early bovines, in evolving new morphologies, might have expanded their fundamental niche, though not necessarily their realized niche. The results of the current study in fact provide no good evidence for either a broadening or a narrowing of the fundamental niche: while the fossil bovines do show a greater absolute range of δ13C values than do the boselaphines (Fig. 2), the variances themselves are not statistically significant, implying that these groups maintained essentially similar dietary niche breadths (Feranec, 2007). Therefore, the fundamental niche of early bovines appears simply to have shifted over (indicated by more positive δ13C values) relative to that of boselaphines, without evidence for any broadening. However, the fact that the entirety of the measured values lie ‘crowded’ at the most positive end of the possible C3 range may confound the ability to perceive broadened niches using δ 13 C analysis. There is also the possibility that early bovines retained the ability to feed in more closed habitats but only did so rarely, erratically, and when forced to. With regard to living bovines, this may be illustrated in the example of the African savanna buffalo (Syncerus caffer), which throughout the year prefers grazing in grasslands but will move to forest browsing if their preferred riverine grassland habitats are occupied by

536

F. Bibi / Palaeogeography, Palaeoclimatology, Palaeoecology 253 (2007) 529–538

wildebeest (Connochaetes taurinus) during the dry season when resources are limiting (Sinclair, 1977). Further work, such as serial sampling for determination of within-specimen isotopic variability (amplitude of seasonal signals), may better address this hypothesis (e.g. Cerling et al., 2006; Nelson, 2005). Such work would include preferably specimens identified to the level of species and from a wider chronological range extending into times when C4 became a prominent dietary component. 4.4. Oxygen isotope ratios—water intake behaviors Oxygen isotope ratios in the fossil bovine and boselaphine teeth display means that are different but not significantly so. Sponheimer and Lee-Thorp (1999) found that browsers tend to exhibit slightly more enriched δ18O values than grazers, the interpretation being that browsers derive a greater ratio of their water directly from their food than do grazers. The fact that the fossil boselaphine teeth are on the average more enriched in δ18O than the fossil bovine teeth would support the hypothesis that these Miocene boselaphines browsed more and drank less while the fossil bovines grazed more and as a consequence also drank more. However, this argument is tenuous based on the current data. The most enriched δ18O values are those of boselaphines while the single most depleted value is from a bovine specimen (Fig. 2A). Tooth enamel δ18O values show no consistent relationship either to morphology or to locality. Rather, the variation in δ18O values probably reflects behavioral differences in water intake at the level of the different individuals sampled. Living Bovini, Boselaphini, and Tragelaphini are obligate drinkers more or less tied to water sources, though much of their water intake must derive also from the plant matter they ingest, and this may have been the situation for their late Miocene fossil relatives as well. Variations in enamel δ18O values can be the result of any number and combination of water intake behaviors such as feeding from canopy tops (enriched) vs. understory (depleted), or drinking from evaporating waterholes (enriched) vs. rivers (depleted). These behaviors are likely to have been highly variable even within individual bovids from place to place and season to season. Thus, the late Miocene bovines and boselaphines sampled appear to have been opportunistic rather than restricted in their water intake behaviors. 5. Conclusions Thirty fossil dental specimens attributable to late Miocene (∼8 Ma) bovines and boselaphines were

sampled for bioapatite carbon and oxygen isotope ratios. The results provide support for fine-scale nichepartitioning between these two bovid phylogenetic/ functional groups along a dietary axis reconstructed using δ13C carbon isotope ratios. All values are clustered between −9‰ and −13‰, indicating that both bovines and boselaphines lived in pure C3-dominated environments. In accordance with predictions made on the basis of the functional morphological differences between fossil bovine and boselaphine teeth (Bibi, 2007), fossil bovine dental enamel on the whole exhibits significantly more positive δ13C values while fossil boselaphine teeth have more negative δ13C values. Enrichment of the heavier isotope of carbon (more positive δ13C values) in bovine teeth is indicative of feeding on C3 plants growing in more open environments–such as canopy gap clearings or grasses in open C3 grasslands– than inhabited by boselaphines which exhibited more negative values signifying feeding in more closed habitats. Outlier specimens within both the bovine and boselaphine series indicate that while diets were on the whole different, important dietary variations at the individual level were often significant, resulting in some more depleted and more enriched δ13C values than expected in bovines and boselaphines, respectively. Niche breadth, estimated as the variance of bovine and boselaphine δ13C values about their respective means, does not differ significantly between these two bovid groups. There is thus no evidence here that the evolution of novel morphological features associated with increased grazing permitted a broadening of the fundamental niche in early bovines. However, the fact that all specimen δ13C values in this study are clumped at the most positive end of the C3 feeding range gives reason to consider whether the absence of a C4 component may preclude detection of expanded niche breadth in this case. Further studies including younger specimens (b7.4 Ma) and the C4 dietary spectrum may better address the question of fundamental niche broadening. Oxygen isotope δ18O values are not significantly different between late Miocene bovines and boselaphines. Living representative of Bovinae (Bovini + Boselaphini + Tragelaphini) are all obligate drinkers more or less dependant on proximity to water bodies, and it is likely that this was the case for both the fossil bovines and boselaphines sampled in the study. This study has utilized stable isotopes to test ecological hypotheses developed on the basis of functional analysis of dental occlusal morphology (Bibi, 2007). The result is a high-resolution analysis that discriminates between more closed- and more open-environment resource use within a narrow range of isotopic values

F. Bibi / Palaeogeography, Palaeoclimatology, Palaeoecology 253 (2007) 529–538

(∼3‰). In tracing the morphological and ecological origins and evolution of a single herbivorous clade (Bovini), this study contributes to ongoing and wideranging efforts to reconstruct and explain the emergence of proto-modern climate systems, biotas, community structures, and taxa in late Miocene times, between about 10 and 5 Ma. Acknowledgements I am most grateful to G. Olack for his help in running the specimens at the Earth Systems Center for Stable Isotopic Studies, Yale University. R. Feranec, L. Grawe, and W. Straight provided much advice on sample collection and preparation. L. Grawe generously provided me with samples of MEme standard. B. Tipple, R. Ferance, M. Pagani, E. Vrba, and two anonymous reviewers provided comments that much improved an earlier draft of this manuscript. This work was supported by a Yale Institute for Biospheric Studies Center for Field Ecology grant, a Geological Society of America Ross Research Award, and a National Science Foundation Graduate Research Fellowship. References Barry, J.C., Morgan, M.E., Flynn, L.J., Pilbeam, D., Behrensmeyer, A.K., Raza, S.M., Khan, I.A., Badgley, C., Hicks, J., Kelley, J., 2002. Faunal and environmental change in the late Miocene Siwaliks of northern Pakistan. Paleobiology 28 (S2), 1–71. Bell, R.H.V., 1969. The use of the herb layer by grazing ungulates in the Serengeti. In: Watson, A. (Ed.), Animal Populations in Relation to their Food Resources. Blackwell Scientific, Oxford, pp. 111–124. Bell, R.H.V., 1971. A grazing ecosystem in the Serengeti. Scientific American 225, 86–93. Bender, M.M., 1971. Variations in the 13C/12C ratios of plants in relation to the pathway of photosynthetic carbon dioxide fixation. Phytochemistry 10, 1239–1244. Bibi, F., 2007. Origin, paleoecology, and paleobiogeography of early Bovini. Palaeogeography Palaeoclimatology Palaeoecology 248, 60–72. Bond, W.J., Keeley, J.E., 2005. Fire as a global ‘herbivore’: the ecology and evolution of flammable ecosystems. Trends in Ecology & Evolution 20, 387–394. Cerling, T.E., Harris, J.M., 1999. Carbon isotope fractionation between diet and bioapatite in ungulate mammals and implications for ecological and paleoecological studies. Oecologia 120, 347–363. Cerling, T.E., Harris, J.M., Passey, B.H., 2003. Diets of East African Bovidae based on stable isotope analysis. Journal of Mammalogy 84, 456–470. Cerling, T.E., Hart, J.A., Hart, T.B., 2004. Stable isotope ecology in the Ituri Forest. Oecologia 138, 5–12. Cerling, T.E., Wittemyer, G., Rasmussen, H.B., Vollrath, F., Cerling, C.E., Robinson, T.J., Douglas-Hamilton, I., 2006. Stable isotopes in elephant hair document migration patterns and diet changes. Proceedings of the National Academy of Sciences of the United States of America 103, 371–373.

537

Codron, D., Lee-Thorp, J.A., Sponheimer, M., Codron, J., de Ruiter, D., Brink, J.S., 2007. Significance of diet type and diet quality for ecological diversity of African ungulates. Journal of Animal Ecology 76, 526–537. Damuth, J., Fortelius, M., Andrews, P., Badgley, C., Hadly, E.A., Hixon, S., Janis, C., Madden, R.H., Reed, K., Smith, F.A., Theodor, J., Van Dam, J.A., Van Valkenburgh, B., Werdelin, L., 2002. Reconstructing mean annual precipitation based on mammalian dental morphology and local species richness. Journal of Vertebrate Paleontology 22 (suppl), 48A. Dekker, B., vanRooyen, N., Bothma, J.D., 1996. Habitat partitioning by ungulates on a game ranch in the Mopani veld. South African Journal of Wildlife Research 26, 117–122. Ehleringer, J.R., Cooper, T.A., 1988. Correlations between carbon isotope ratio and microhabitat in desert plants. Oecologia 76, 562–566. Estes, R.D., 1991. The Behavior Guide to African Mammals. University of California Press, Berkeley and Los Angeles. Farquhar, G.D., Ball, M.C., Voncaemmerer, S., Roksandic, Z., 1982a. Effect of salinity and humidity on delta-C-13 value of halophytesevidence for diffusional isotope fractionation determined by the ratio of inter-cellular atmospheric partial-pressure of CO2 under different environmental conditions. Oecologia 52, 121–124. Farquhar, G.D., Ehleringer, J.R., Hubick, K.T., 1989. Carbon isotope discrimination and photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 40, 503–537. Farquhar, G.D., O'Leary, M.H., Berry, J.A., 1982b. On the relationship between carbon isotope discrimination and the intercellular carbon dioxide concentration in leaves. Australian Journal of Plant Physiology 9, 121–137. Feranec, R.S., 2003. Stable isotopes, hypsodonty, and the paleodiet of Hemiauchenia (Mammalia: Camelidae): a morphological specialization creating ecological generalization. Paleobiology 29, 230–242. Feranec, R.S., 2004. Geographic variation in the diet of hypsodont herbivores from the Rancholabrean of Florida. Palaeogeography Palaeoclimatology Palaeoecology 207, 359–369. Feranec, R.S., 2007. Ecological generalization during adaptive radiation: evidence from Neogene mammals. Evolutionary Ecology Research 9, 555–577. Feranec, R.S., MacFadden, B.J., 2006. Isotopic discrimination of resource partitioning among ungulates in C-3-dominated communities from the Miocene of Florida and California. Paleobiology 32, 191–205. Fortelius, M., Eronen, J., Jernvall, J., Liu, L.P., Pushkina, D., Rinne, J., Tesakov, A., Vislobokova, I., Zhang, Z.Q., Zhou, L.P., 2002. Fossil mammals resolve regional patterns of Eurasian climate change over 20 million years. Evolutionary Ecology Research 4, 1005–1016. Fortelius, M., Eronen, J., Liu, L.P., Pushkina, D., Tesakov, A., Vislobokova, I., Zhang, Z.Q., 2006. Late Miocene and Pliocene large land mammals and climatic changes in Eurasia. Palaeogeography Palaeoclimatology Palaeoecology 238, 219–227. Fortelius, M., Solounias, N., 2000. Functional characterization of ungulate molars using the abrasion-attrition wear gradient; a new method for reconstructing paleodiets. American Museum Novitates 3301, 1–136. Franzen, J.L., Storch, G., 1998. Late Miocene Mammals from Central Europe. In: Agustí, J. (Ed.), Evolution of Neogene Terrestrial Ecosystems in Europe, vol. 1. Cambridge University Press, Cambridge and New York, pp. 165–190. Geist, V., 1974. Relationship of social evolution and ecology in ungulates. American Zoologist 14, 205–220.

538

F. Bibi / Palaeogeography, Palaeoclimatology, Palaeoecology 253 (2007) 529–538

Hungate, R.E., Phillips, G.D., McGregor, A., Hungate, D.P., Buechner, H.K., 1959. Microbial fermentation in certain mammals. Science 130, 1192–1194. Hutchinson, G.E., 1957. Concluding remarks. Cold Spring Symposia on Quantitative Biology 22, 415–427. Jarman, P.J., 1974. The social organization of antelope in relation to their ecology. Behaviour 48, 215–266. Keeley, J.E., Rundel, P.W., 2005. Fire and the Miocene expansion of C-4 grasslands. Ecology Letters 8, 683–690. Koch, P.L., Tuross, N., Fogel, M.L., 1997. The effects of sample treatment and diagenesis on the isotopic integrity of carbonate in biogenic hydroxylapatite. Journal of Archaeological Science 24, 417–429. MacFadden, B.J., Higgins, P., 2004. Ancient ecology of 15-millionyear-old browsing mammals within C3 plant communities from Panama. Oecologia 140, 169–182. Mainland, I.L., 2003. Dental microwear in grazing and browsing Gotland sheep (Ovis aries) and its implications for dietary reconstruction. Journal of Archaeological Science 30, 1513–1527. Medina, E., Minchin, P., 1980. Stratification of delta-C-13 values of leaves in Amazonian rain forests. Oecologia 45, 377–378. Michelsen, A., Jonasson, S., Sleep, D., Havstrom, M., Callaghan, T.V., 1996. Shoot biomass, delta C-13, nitrogen and chlorophyll responses of two arctic dwarf shrubs to in situ shading, nutrient application and warming simulating climatic change. Oecologia 105, 1–12. Nelson, S.V., 2005. Paleoseasonality inferred from equid teeth and intra-tooth isotopic variability. Palaeogeography Palaeoclimatology Palaeoecology 222, 122–144. O'Leary, M.H., 1993. Biochemical basis of carbon isotope fractionation. In: Ehleringer, J.R., Hall, A.E., Farquhar, G.D. (Eds.), Stable Isotopes and Plant Carbon-Water Relations. Academic Press, San Diego, California, USA, pp. 19–28. Pagani, M., Freeman, K.H., Arthur, M.A., 1999. Late Miocene atmospheric CO2 concentrations and the expansion of C-4 grasses. Science 285, 876–879. Pagani, M., Zachos, J.C., Freeman, K.H., Tipple, B., Bohaty, S., 2005. Marked decline in atmospheric carbon dioxide concentrations during the Paleogene. Science 309, 600–603. Passey, B.H., Cerling, T.E., Perkins, M.E., Voorhies, M.R., Harris, J.M., Tucker, S.T., 2002. Environmental change in the Great Plains: an isotopic record from fossil horses. Journal of Geology 110, 123–140. Passey, B.H., Robinson, T.F., Ayliffe, L.K., Cerling, T.E., Sponheimer, M., Dearing, M.D., Roeder, B.L., Ehleringer, J.R., 2005. Carbon isotope fractionation between diet, breath CO2, and bioapatite in

different mammals. Journal of Archaeological Science 32, 1459–1470. Prater, S.H., 1965. The Book of Indian Animals. Bombay Natural History Society, Bombay, India. Quade, J., Cerling, T.E., 1995. Expansion of C-4 grasses in the late Miocene of northern Pakistan—evidence from stable isotopes in paleosols. Palaeogeography Palaeoclimatology Palaeoecology 115, 91–116. Quade, J., Cerling, T.E., Andrews, P., Alpagut, B., 1995. Paleodietary reconstruction of Miocene faunas from Paşalar, Turkey using stable carbon and oxygen isotopes of fossil tooth enamel. Journal of Human Evolution 28, 373. Quade, J., Cerling, T.E., Bowman, J.R., 1989. Development of Asian monsoon revealed by marked ecological shift during the latest Miocene in northern Pakistan. Nature 342, 163–166. Sanson, G.D., Kerr, S.A., Gross, K.A., 2007. Do silica phytoliths really wear mammalian teeth? Journal of Archaeological Science 34, 526–531. Schuette, J.R., Leslie, D.M., Lochmiller, R.L., Jenks, J.A., 1998. Diets of hartebeest and roan antelope in Burkina Faso: support of the long-faced hypothesis. Journal of Mammalogy 79, 426–436. Sinclair, A.R.E., 1977. The African Buffalo. University of Chicago Press, Chicago. Sponheimer, M., Lee-Thorp, J.A., 1999. Oxygen isotopes in enamel carbonate and their ecological significance. Journal of Archaeological Science 26, 723–728. Sponheimer, M., Lee-Thorp, J.A., 2001. The oxygen isotope composition of mammalian enamel carbonate from Morea Estate, South Africa. Oecologia 126, 153–157. Tipple, B., Pagani, M., 2007. The early origins of terrestrial C4 photosynthesis. Annual Review of Earth and Planetary Sciences 35, 435–461. Traill, L.W., 2004. Seasonal utilization of habitat by large grazing herbivores in semi-arid Zimbabwe. South African Journal of Wildlife Research 34, 13–24. van der Merwe, N.J., Medina, E., 1991. The canopy effect, carbon isotope ratios and foodwebs in Amazonia. Journal of Archaeological Science 18, 249–259. Vogel, J.C., 1978. Recycling of CO2 in a forest environment. Oecologia Plantarum 13, 89–94. Yakir, D., 1997. Oxygen-18 of leaf water: a crossroad for plant associated isotopic signals. In: Griffiths, H. (Ed.), Stable Isotopes: Integration of Biological, Ecological, and Geochemical Processes. BIOS, Oxford, pp. 147–168.