Advertisement

Oecologia

, Volume 187, Issue 4, pp 1095–1105 | Cite as

Stable isotope ecology of black rhinos (Diceros bicornis) in Kenya

  • Thure E. Cerling
  • Samuel A. Andanje
  • Francis Gakuya
  • John M. Kariuki
  • Linus Kariuki
  • Jackson W. Kingoo
  • Cedric Khayale
  • Isaac Lekolool
  • Anthony N. Macharia
  • Christopher R. Anderson
  • Diego P. Fernandez
  • Lihai Hu
  • Shawn J. Thomas
Special Topic

Abstract

Stable isotope and elemental ratios in hair are influenced by the environment, including both climate and geology. Stable carbon isotopes can be used to give estimates of the C4/CAM fraction of diets of herbivorous mammals; stable nitrogen isotopes are related to the local water deficit; strontium isotopes are determined by the local geology. We studied hair from rhinos in Kenya to determine spatial patterns in δ13C, δ15N, and 87Sr/86Sr ratios. The samples of rhino hair were collected during Kenya Wildlife Service translocation or veterinary activities. δ13C values showed diets dominated by C3 foods, but in some regions the diet, at least seasonally, contained significant quantities (i.e., > ca. 20%) of C4/CAM foods. δ15N values were related to water deficit, with higher δ15N values in regions with high water deficit. 87Sr/86Sr isotope ratios were found to be related to the local geological substrate suggesting that 87Sr/86Sr isotope ratios are provisionally useful for determining the origins of illegal wildlife materials in Kenya and elsewhere in Africa.

Keywords

Isotope ecology East Africa Diceros Conservation National parks Diet Rhinoceros 

Introduction

Stable isotope analyses of animal tissues provide important information about the diets of herbivorous mammals (Cerling et al. 2003; Sponheimer et al. 2003), even in the absence of direct observations. In some cases, regional dietary differences can be determined (van der Merwe et al. 1988) and these can be further used in forensics investigations of wildlife products (e.g., ivory, van der Merwe et al. 1990; Cerling et al. 2007b).

Stable carbon isotopes are very useful in estimating the fraction of C3 plants versus C4 plants in the diet of herbivorous mammals. This is because the C3 pathway has, on average, significantly lower 13C/12C ratios than do C4 plants. C3 plants comprise trees, shrubs, forbs, and herbs. C4 plants in East Africa are primarily grasses or sedges, although some C4 dicots are present in East Africa (for example, Blepharis). In Africa, the isotope ratio of CAM plants is generally similar to that of C4 plants. Some CAM plants (e.g., Sansevieria, Euphorbia, and Salsola) are known to be eaten by black rhinos (Oloo et al. 1994; Ganqa et al. 2005; Lieverloo et al. 2009; Buk and Knight 2010). Because stable isotopes do not distinguish between C4 plants and CAM in Africa, we will refer to C4 plants and CAM plants collectively as C4/CAM for the purposes of rhino isotope ecology.

Likewise, 15N/14N ratios are inherited from diet but the interpretation of dietary source is not as straightforward as in carbon because of variation in both soil nitrogen and in N-fixation in plants. Acacia are legumes and some are known to fix nitrogen from the atmosphere. However, stable isotope analyses of Acacia from Kenya show that many Acacia do not fix nitrogen based on their nitrogen isotope ratios. Other important potential nitrogen-fixing plants in rhino diet include Indogofera.

87Sr/86Sr isotopes differ in geological regions due to the decay of 87Rb over time; 87Sr/86Sr from old geological terrains, especially granitic terrains have high amounts of radiogenic 87Sr, whereas young basaltic rocks, such as are found in the Rift Valley, have very little radiogenic 87Sr. 87Sr/86Sr ratios of very young basaltic rocks (i.e., Quaternary) are generally about 0.703–0.705; very old terrains, such as in South Africa may have 87Sr/86Sr ratios greater than 0.720. van der Merwe et al. (1990) and Vogel et al. (1990) both proposed that 87Sr/86Sr isotope ratios could provide a means to determine the origin of elephant ivory and thus could be a good indicator of wildlife origins. In East Africa, Koch et al. (1995) and Coutu et al. (2016) published 87Sr/86Sr isotope ratios for elephant bone and ivory from some known locations. Thus, strontium isotopes in diet are determined primarily by the local geological substrate; however, Graustein and Armstrong (1983), Quade et al. (1995), and Vitousek et al. (1999) showed that the local geological substrate is modified by atmospheric import of dust so that the ecologically available strontium, expressed as 87Sr/86Sr ratios, is a combination of the local geological substrate and imported dust. Sr isotope ratios have proven to be important ecological tracers in studying fish ecology (e.g., Kennedy et al. 1997, 2000; Brennan et al. 2015a, b); bird migration (Blum et al. 2001); and paleoecology (Hoppe et al. 1999).

In this study we report the results of keratin samples (hair) from black rhinos (Diceros bicornis) from Kenya that were obtained between 2008 and 2012. We compare isotope ratios of strontium, carbon, and nitrogen from different regions within Kenya that have differing geological substrates and different ecological conditions related to an aridity cline; we also study single hairs for carbon and nitrogen isotopes to understand seasonal diet changes. These regional results suggest that isotopes would be useful in distinguishing the origins of illegal wildlife products, such as rhino horn. We also report the results for carbon and oxygen from tooth enamel from rhinos that died principally during the late 1960s drought in East Africa.

Methods

Sample collection

Hair samples from black rhinos (Diceros bicornis) were collected by KWS staff between 2008 and 2012 during translocation, ear-notching, transmitter installation, or treatment procedures between 2008 and 2012. Hair was collected from the tail by plucking so that the proximal end was preserved. Figure 1 shows the localities sampled with the local bedrock geology. Appendix SI 1 gives information about individual rhinos (age, sample collection date, residential history, etc).
Fig. 1

Localities of samples collected for isotope analysis in relationship to bedrock geology. Geology is from the Kenya National Atlas (1962) and the Quennel (1959). This figure will appear in color in the online version of the journal

Plant samples were collected from the Laikipia region, the Nairobi region, and from Tsavo East NP between 1997 and 2007. Plants known to be within 50 m of known historic bomas (thorn-bush corrals made by local pastoralists) are not included in this report. C3 and C4 plants were identified in the field. Details of the rhino program in Kenya are provided by Western (1982), Oloo et al. (1994), Muya and Oguge (2000), Walpole et al. (2001), Birkett (2002), Walpole (2002), Amin et al. (2006), Patton and Jones (2007), Patton et al. (2007, 2008), Okita-Ouma et al. (2008, 2010), Dharani et al. (2009), Ngene et al. (2011), and Emslie (2013).

Teeth were sampled from the Tsavo East Research Center, using the collections from animals that died in the great drought of the late 1960s (Spinage 1994). Sampled species included the black rhino, elephant (Loxodonta africana), giraffe (Giraffa camelopardalis), and zebra (Equus burchelli), with the three latter species being analyzed for diet comparisons to black rhinos.

Treatment

Hair samples were washed with chloroform–methanol to remove organic contaminants.

Light stable isotope analysis

Single bulk hair samples and plant samples were ground and ca. 0.5–2 mg were analyzed using an elemental analyzer (EA) coupled to a flow-through isotope ratio mass spectrometer (IRMS) for measurements of the 13C/12C and 15N/14N ratios. Tooth enamel was treated with H2O2 and 0.1 m acetic acid and the dried powder was reacted with 100% H3PO4; purified CO2 was analyzed for 13C/12C and 18O/16O ratios on an IRMS operating in dual-inlet mode.

Results are reported using the conventional permil (‰) notation:
$$ \updelta^{ 1 3} {\text{C}} = (R_{\text{sample}} / \, R_{\text{standard}} - 1 )\times 1000, $$
where Rsample and Rstandard are the 13C/12C ratios in the sample and standard, respectively. Analogous results are reported for δ15N and δ18O isotope values. The standards are Pee Dee Belemnite (PDB) for carbon and oxygen and AIR (atmospheric nitrogen) for nitrogen. We do not correct for the δ13C change in the atmosphere because the change in δ13C over the observation period (1997–2012) is less than 0.5 ‰ for the average annual atmosphere δ13C value (NOAA 2017). We also do not correct the δ13C values for the rhino tooth enamel samples, which were predominantly from about 1970; the rhino tooth enamel samples are directly compared with other taxa from the sample collection period and so all taxa would have an identical offset.

Sequential hair samples were cut into 2 mm segments and analyzed using the EA-IRMS as described above. Diet histories were modeled using the reaction progress model (Ayliffe et al. 2004; Cerling et al. 2007a); we use the same parameters as was found for equids (3 ‰ isotope enrichment for diet–hair for both δ13C and δ15N; multi-pool model with 0.5, 4, and 138 days half-lives with fractional contributions of 0.41, 0.15, and 0.44, respectively). Zazzo et al. (2007, 2008) found that bovids (sheep and cow) also should be modeled with a multi-pool model with similar pool half-lives (i.e., one very short, one intermediate, and one ca. 100 days or longer). Turnover experiments of this nature have not defined the isotope pools for nitrogen with the same precision as carbon. Therefore, we use the same parameters for isotope turnover for nitrogen isotopes as for carbon isotopes; we base this rationale on the assumption that the turnover pools relate to essential versus non-essential amino acids and that the structural remodeling of the different amino acids is similar for carbon and nitrogen. We assume a constant growth rate for rhino tail hair: West et al. (2004) and Wittemyer et al. (2009) have shown that changes in growth rates in horse tail hair related to diet, and in elephant tail hair related to physiological stress, respectively, must be less than 5%, which was the limit of detection in those two studies. Average growth rates are not known for rhinoceros tail hair; we assume a growth rate of 0.8 mm per day, which is the same as that for equids and elephants (Ayliffe et al. 2004; West et al. 2004; Cerling et al. 2009; Wittemyer et al. 2009).

Sr isotope analysis

Sr isotope measurements on bulk rhino hair were made in the Sr isotope laboratory on a Neptune multi-collector ICP-MS in the Department of Geology and Geophysics at the University of Utah. Samples were digested and analyzed as described by Tipple et al. (2013), using the Sr-FAST method of Mackey and Fernandez (2011).

Results

Bulk samples, resident individuals

Hair of several individuals from Nairobi National Park, Tsavo West National Park (Ngulia), Tsavo East National Park, Masai Mara National Reserve, Nakuru National Park, and from Solio, Ol Pejeta, Ol Jogi, and Mugie in the greater Laikipia region were sampled and analyzed for δ13C and δ15N in their hair. Hair samples were generally 50–100 mm in length; reported values represent the average for an entire hair. Figures 2 and 3 show the average values of resident individuals from these regions which span a large ecological gradient from mesic (Nakuru, Nairobi, Masai Mara) to xeric (Tsavo West); those results show that δ13C and δ15N values for many of the collecting regions differ from each other. Table 2 gives the average δ13C, δ15N, and 87Sr/86Sr isotope values for each locality and Table SI 1 gives the individual values for each analysis. The differences between the regions are discussed further in “Discussion”.
Fig. 2

Average δ13C and δ15N for rhino hair collected from different localities in Kenya

Fig. 3

Water deficit (WD) and average δ15N for rhino hair from different localities. Below at WD of 275 mm/year there is no trend for WD vs. δ15N; however, above a WD of 275 mm/year there is a strong positive correlation (r2 = > 0.98)

Sequence samples of hair

Seven tail hairs were analyzed sequentially, one each from Tsavo East, Tsavo West, Nakuru, Mugi, Mara Mara, Ol Jogi, and Nairobi (Appendix SI 2). Figure 4 shows the δ13Ckeratin and δ15Nkeratin values for the hair along with the estimated diet for each segment using the reaction progress model for diet–hair enrichment (Ayliffe et al. 2004; Cerling et al. 2007a). Only one sample, from Ol Jogi in the Laikipia, shows evidence for a significant (> 20%) C4/CAM contribution to the diet.
Fig. 4

δ13C and δ15N values for hair and estimated diets for sequential hair analysis of black rhinos from Kenya. a Masai Mara; b Mugie; c Nakuru; d Nairobi; e Ol Jogi; f Tsavo East; g Tsavo West; h δ13C and δ15N ranges for diet. This figure will appear in color in the online version of the journal

Plant samples

Table SI 2 summarizes the results from plants collected in the Laikipia, Nairobi, and Tsavo regions from 1997 to 2007 and Appendix SI 3 lists the individual results of analyses. As is typical for C3 plants, the average δ13C for the different regions ranged between − 25.5 and − 28.6 ‰ with the most mesic region (Nairobi) having the most negative δ13C values. Likewise as is typical for C4 plants, the average δ13C values for C4 plants range from − 11.6 to − 12.9 ‰ for this period of time (1997–2007). Average δ15N values for C3 and C4 plants at each site are similar: Nairobi (2.4 and 1.8 ‰, respectively), Laikipia (6.3 and 4.8 ‰, respectively), and Tsavo (9.8 and 10.8 ‰, respectively); these δ15N values represent a strong gradient from mesic (Nairobi) to xeric (Tsavo) as shown by the climatological data in Table 1. All sites had some samples that can be interpreted as being N-fixing based on δ15N values near 0 ‰; however, most Acacia specimens (often considered to be an N-fixer) had δ15N values that do not provide evidence for N-fixation by the plants sampled.
Table 1

Location of site with rhino hair collections (2008–2012)

Location

Lat (N)

Long (E)

Altitude (m)

MAT (mm/year)

MAP (mm/year)

Water deficit (mm/year)

Masai Mara

− 1.41

35.08

1510

16.5

736

− 72

Mugie

0.74

36.63

1840

16.7

660

450

Nairobi NP

− 1.36

36.85

1643

18.9

785

124

Nakuru NP

− 0.46

36.09

1790

17.5

909

− 82

Ol Jogi

0.24

36.98

1720

17.4

770

372

Ol Pejeta

0.04

36.93

1790

16.2

819

260

Solio

− 0.24

36.88

1995

16.0

758

275

Tsavo East

− 3.36

38.61

530

25.1

549

799

Tsavo West–Ngulia

− 3.05

38.30

585

25.0

621

629

Mugie and Ol Jogi MAP from Georgiadis (personal communication), Solio from Lamuria (Kenya Met Office), Tsavo West from Makindu

WD calculated as in Blumenthal et al. (2017)

CAM plants were not analyzed as part of this study. In the discussion below we assume that CAM plants have a δ13C value similar to C4 plants growing in the same region. We discuss diet in terms of C3 and C4/CAM contributions as mentioned in “Methods”.

Tooth enamel samples

Table SI 3 shows the results for analysis of tooth enamel for 14 different teeth from 10 individual rhinos, 11 elephants, 5 zebras, and 5 giraffes, all from Tsavo East National Park. All specimens were from the same approximate year of death (ca. 1970). The average δ13Cenamel values are − 10.6 ± 1.3, − 9.9 ± 1.3, 0.4 ± 0.3, and − 11.3 ± 0.5 ‰, for rhinos, elephants, zebra, and giraffes, respectively.

Discussion

Hair samples used for comparison report the average of δ13Ckeratin and δ15Nkeratin over the length of the hair, which was generally 5 cm or greater. Assuming growth rates of 0.8 mm per day, the δ13C and δ15N values give the average value over several months of time, assuming 50% contribution to a longer metabolic pool increases the time represented in a sample.

End-member δ13C values for diet estimates

The selection of an end-member diet value is problematic in many studies. In this case, we have seven detailed diet histories from different regions that can take into account seasonality in diet. We use the forward diet model of Ayliffe et al. (2004) and Cerling et al. (2007a) on seven individuals to estimate the “pure C3” end member. Assuming growth rates of 0.8 mm per day, a sample interval of 2 mm gives an average length of 2.5 days. The 10th percentile diet values for all sites are between − 26.5 and − 28.6 ‰; due to the variability in individual plants (Appendix SI 3) we use − 27.5 ‰ for the C3 end member. The pure C4/CAM end member is assumed to be − 12.0 ‰ based on the data in Appendix SI 3, and the Cerling and Harris (1999) survey of Kenyan C3 and C4 plants. For hair samples using a 3 ‰ isotope enrichment for 13C (Cerling and Harris 1999), these correspond to δ13C hair values of − 24.6 and − 9.0 ‰ for pure C3 and pure C4 diets, respectively.

Bulk hair samples: stable isotopes—δ13Ckeratin and δ15Nkeratin

The δ13Ckeratin and δ15Nkeratin in hair from different regions in Kenya show clear differences in isotope values between Laikipia (Solio, Ol Pejeta, Ol Jogi, Mugie), Ngulia/Tsavo, Mara, Nakuru and Nairobi (Fig. 2 and Table 2). Mara, Nakuru, and Nairobi are depleted in both 13C and 15N; using the δ13C values above the long-term fraction of C4/CAM derived resources are ca. 5–10%. The Laikipia group (Solio, Ol Pejeta, Ol Jogi, Mugie) cluster together with δ13C values indicating a C4/CAM diet contribution of ca. 15–20% and with δ15N values lower than the greater Tsavo region. The two Tsavo populations (Ngulia, and Tsavo East) have a low contribution of C4/CAM (less than 10%), but have the highest δ15N values.
Table 2

δ13C, δ15N, and 87Sr/86Sr isotope average values for rhino hair from localities in this study

Locality

δ13C ± 1sd (N)

δ15N ± 1sd (N)

87Sr/86Sr ± 1sd (N)

Lake Nakuru NP

− 23.8 ± 0.7 (13)

4.9 ± 0.8 (13)

0.7074 ± 0.0001 (7)

Masai Mara NR

− 23.8 ± 0.5 (11)

5.4 ± 0.6 (11)

0.7111 ± 0.0023 (5)

Mugie GR

− 22.1 ± 0.8 (15)

7.0 ± 0.3 (15)

0.7073 ± 0.0001 (10)

Nairobi NP

− 23.7 ± 0.3 (16)

4.3 ± 1.1 (16)

0.7072 ± 0.0003 (5)

Ol Jogi GR

− 22.5 ± 0.3 (9)

6.0 ± 0.8 (9)

0.7076 ± 0.0002 (5)

Ol Pejeta

− 22.0 ± 0.9 (3)

5.0 ± 0.2 (3)

0.7062 ± 0.0002 (3)

Solio GR

− 22.9 ± 0.3 (5)

5.0 ± 0.6 (5)

0.7057 ± 0.0005 (5)

Tsavo–Ngulia

− 23.6 ± 0.2 (10)

9.3 ± 0.4 (10)

0.7073 ± 0.0009 (5)

Tsavo East NP

− 23.2 ± 0.6 (6)

10.5 ± 0.3 (6)

0.7105 ± 0.0009 (6)

The number of samples analyzed in each group is given in parentheses

Black rhinos are known to eat C4 grasses, but some CAM plants are also likely to be of significant importance—specifically Sansevieria, Aloe, Euphorbia, and Salsola, all of which are known to be favored by black rhinos in certain regions or seasons (Goddard 1970; Oloo et al. 1994; Ganqa et al. 2005; Lieverloo et al. 2009; Buk and Knight 2010).

δ15N is strongly related to the water deficit; Fig. 3 shows that there is little relationship between water deficit and δ15N in rhino hair below ca. 275 mm/year, but that there is a very strong relationship above 275 mm/year (r2 > 0.98). The δ15N of plants from the Nairobi, Laikipia, and Tsavo regions show similar increases in δ15N with increasing water deficit; the advantage of using δ15N of hair as a proxy for the δ15N of plants is that it averages the δ15N values over many thousands of plants.

Strontium isotopes

We analyzed 51 bulk rhino hair samples and the 87Sr/86Sr ratio in the hair ranges from 0.7050 to 0.7136 (Table SI 2).

Comparison of the different sites shows that each site has a narrow range of 87Sr/86Sr ratios (Fig. 5). Sites in predominantly volcanic terrains (Fig. 1: Nakuru NP, Nairobi NP, and the Laikipia region) have 87Sr/86Sr values averaging between 0.706 and 0.707 (Table 2). Tsavo East and the Masai Mara regions, which have bedrock dominated by old Mozambique Belt metamorphic rocks (Fig. 1), have distinctive radiogenic 87Sr/86Sr ratios averaging about 0.710–0.711 (Table 2). The rhinos from Tsavo West are in the enclosed Ngulia Rhino Sanctuary; rhinos from this site may be on a mixing line between the Mozambique Belt end-member values (ca. 0.711) and a Quaternary volcanic end member because the 87Sr/86Sr values for Mzima Springs is 0.70480 ± 0.00009 (n = 3). The Tsavo River originates in the Mzima Springs and is the principal drinking (and wallowing) water for rhinos in the Ngulia Rhino Sanctury, whereas the principal bedrock in the sanctuary is Mozambique Belt metamorphic rocks.
Fig. 5

Average δ15N and 87Sr/86Sr for rhino hair collected from different localities in Kenya

Possible forensics applications

These results suggest that stable isotopes could have forensic applications is helping to assign regions of origin to rhino horn. Because rhino horn is composed of keratin, the isotope distributions based on hair can give the δ13Ckeratin, δ15Nkeratin, and 87Sr/86Sr values characteristic of specific geographic regions (e.g., Figs. 2, 4). The current data set from East Africa for 87Sr/86Sr isotopes from wildlife from national parks and reserves in East Africa is restricted to only a few studies (van der Merwe et al. 1990; Koch et al. 1995; Coutu et al. 2016; and this study). Those studies suggest that the strontium isotope values in East Africa may range from ca. 0.704 (Quaternary volcanic rocks) to > 0.72 (Archean granitic rocks). This preliminary data set suggests that a more extensive sampling of wildlife from East Africa will be able to be used to show the ranges of 87Sr/86Sr isotope ratios related to geological bedrock maps. Thus, probability assignments to specific national parks/reserves or game ranches may be possible in the near future using the relationship of 87Sr/86Sr in animal tissues to the local bedrock.

Any application of this would have to take into account that rhino horn would record the location where the horn was growing, which may or may not be the same location as where a poaching incident occurred. Rhino horn grows at a rate of about 30–60 mm per year, with lower growth rates in older individuals; rhino horns are up to 500 mm in length, with the tips being worn off with age (Pienaar et al. 1991). Thus, a 500 mm horn could have as much as 15 years of growth recorded. Rhino horn grows at its base and thus the most recent keratin in the horn (at the base) would be from the location where the poaching occurred; a sample from the horn tip could be many years earlier and if the individual had been translocated, such a sample would indicate the location where the keratin actually formed. Thus, a preferred sample would be from the newest possible growth of keratin.

Seasonal diet changes in single individuals

We analyzed a detailed diet history of seven individuals (Table SI 2). Figure 4a–g shows the measured δ13Ckeratin and δ15Nkeratin in hair and the calculated isotope value for instantaneous diet for the previous 50+ days for these individuals in Masai Mara (4A), Mugie (4B), Nakuru (4C), Nairobi (4D), Ol Jogi (4E), Tsavo East (4F), and Tsavo East–Ngulia (4G). Hairs were subsampled at 2–3 mm resolution, which corresponds to about 2–4 days using the estimated growth rate of 0.8 mm per day. We note that the growth is likely to be between 0.3 (humans) and 0.8 (elephants, equids) mm/day (West et al. 2004; Wittemyer et al. 2009); thus, estimated times in Fig. 4a–g should be construed as estimates only. Likewise, each sample interval represents the integrated diet in δ13C and δ15N over the ca. 2–4 days period represented in the interval analyzed; therefore, the range of true dietary δ13C and δ15N values from individual plants that contribute to diet over each hair is greater than the integrated dietary estimates that are shown in Fig. 4a–g.

The Mara rhino had a constant diet that changed about 2 weeks before the sample was collected (Fig. 4a). For the ca. 50 days prior to the late diet change, this individual had a δ13C value of diet ranging between − 26.5 and − 28.4 ‰, consistent with a pure C3 diet, or one with a small fraction (< 5%) of C4/CAM. However, in the last 2 weeks before sampling, the diet was a high as − 24.6 ‰, which we estimate indicates up to 20% C4/CAM. Nitrogen isotope values also are different in the final 2 weeks before sampling: the first 2 months diet was between 0 and 1 ‰ for δ15N, whereas the last several weeks (the period where C4/CAM became significant) was as high as + 5 ‰. We do not have representative samples of plants from the Mara and so do not speculate on the δ15N of plants in the region, except to note that mesic environments in Kenya (e.g., Nairobi) have more positive δ15N values than the more xeric regions (Laikipia, Tsavo) as shown in Appendix SI 3.

The Mugi rhino hair was 116 mm long and represents a time interval of about 5 months (150 days) based on the estimated growth rates of 0.8 mm per day; Fig. 4b shows the δ13C and δ15N of hair and also the estimated δ13C and δ15N of their diet throughout this period. Estimated diet δ13C values of range from − 23.0 to − 27.5 ‰, which is in the range of the δ13C value for C3 plants from water-stressed regions. Using a δ13C value of − 27.5 ‰ for C3 plants, one could make the case for a diet that includes up to ca. 20% C4/CAM-derived vegetation. This sample was collected in January 2012, after the inception of the “short rains”, and the highest δ13C values were during the period when grasses would have been at their highest level of nutrition. Thus, this individual likely had a small, but significant, fraction of C4/CAM vegetation in its diet for a short period of time. δ15N values of the estimated diet range from 2.3 to 5.5 ‰; the δ15N values of Acacia in the Laikipia region have a very wide range of values, and their δ15N values indicate that most Acacia in Laikipia are not N-fixing (see Table SI 3)

The Nakuru rhino hair was 85 mm long, representing a time interval of ca. 110 days based on an assumed growth rate of 0.8 mm per day. This individual exhibited little variation in the δ13C and δ15N of the hair, and hence little variation in the δ13C and δ15N of the diet over this time interval (Fig. 4c). The average δ13C of the diet was calculated to be − 27.6 ± 1.1 ‰, indicating a nearly pure C3 diet. The average δ15N diet was estimated to be 1.6 ± 0.5‰.

The Nairobi rhino can be interpreted as having a pure C3 diet throughout the recorded interval (Fig. 4d); diet ranged between − 25.8 and − 27.8 ‰ which is in the observed range of C3 plants for Nairobi (Cerling and Harris 1999). The calculated δ15N value for the diet ranged from 2.4 to − 0.8 ‰; many plants in the Nairobi region have values in this range (Appendix SI 3) including N-fixing plants as well as plants that are not N-fixing.

The Ol Jogi rhino (Fig. 4e) shows significant diet change though the period recorded in its hair, which is ca. 135 days based on an estimated growth rate of 0.8 mm per day. Weekly diet ranges from − 28.1 ‰ which is a pure-C3 diet, to − 21.4 ‰ which is about 40% C4/CAM. Approximately, 30% of the period of record has a diet comprising > 20% C4/CAM (that is, more positive than − 24 ‰); about 40% of the time the diet was essentially all C3 (more negative than ca. − 26 ‰). The estimated δ15N of diet ranges from − 1.5 to 6.7 ‰, which is well within the range of δ15N for plants in the Laikipia region. Most acacias in the Laikipia region have δ15N values greater than 0 ‰, indicating that they do not directly fix nitrogen (Appendix SI 3).

The Tsavo East rhino hair was 115 mm in length, corresponding to a time interval of ca. 140 days based on the estimated growth rate of 0.8 mm per day and thus represents on the order of 5 months of time. The range in δ13C of estimated diet is from − 27.8 to − 25.0 ‰ (Fig. 4f), indicating a pure C3 diet or nearly so; the δ13C of plants in Tsavo East ranges between − 29 and – 24 ‰ (Appendix SI 3). The δ15N of the estimated diet range ranges from 5.4 to 9.6 ‰ which is well within the observed range of plants, including Acacia, within the Tsavo East region (Appendix SI 3). Overall, the diet of this rhino is suggestive of a pure C3 diet, or nearly so, over the time interval represented by this individual.

The Tsavo West rhino was from the Ngulia sanctuary and the hair analyzed was 63 mm in length. The estimated δ13C of the integrated diet ranged from − 27.6 to − 25.7 ‰ which represents a pure C3 diet, or nearly so, over the ca. 3 months period represented by this sample (Fig. 4g). The range in integrated δ15N values is slightly higher (ca. 3 ‰) over this interval.

Altogether, we analyzed the detailed seasonal diets, integrated over a ca. 2–4 days interval, over time periods of the order of 2–5 months per individual. Only one individual (from Ol Jogi) showed unequivocal evidence for some C4/CAM diet resource use, although several others had δ13C values high enough that they could indicate 10–20% C4 diet intake, or that the local δ13C was several per mil enriched due to water-stressed conditions. Overall, the detailed isotope profiles show relative constant diets over time, in significant contrast to elephants in East Africa (Cerling et al. 2004, 2006, 2009).

Rhino diet in Tsavo from the 1960s using tooth enamel

The diets of rhinos from Tsavo National Park in the 1960s were characterized using tooth enamel from 16 different individuals using the Tsavo East Research Centre collections (Table SI 3). Proportions of C3-browse and C4/CAM in the diet of rhinos and elephants were determined assuming that giraffe and zebra represent end-member values for the C3-browse and the C4/CAM diet, respectively. Using these end members gives estimates for black rhinos and elephants of 7 ± 11 and 13 ± 11% C4/CAM in the diets, respectively, using the standard deviation from the mean. The range for C4/CAM consumption was 0–23% for rhinos. The amplitude of diet change is attenuated during enamel formation and maturation (Passey and Cerling 2002) and, therefore, the seasonal input of C4/CAM could be higher than 25%. Goddard (1970), reporting observations from 1967 to 1969, noted that rhinos in Tsavo Park consumed > 100 different species of plants from 32 plant families, that grass was rarely eaten, and that rhinos had a clear preference for legumes. These data show that several of the rhinos from the Tsavo National Parks had a small portion of C4 grasses or other C4–CAM plants contributing to their diet in Tsavo National Park in the 1960s; for comparison, the hair samples collected from 2008 to 2012 from the Tsavo National Parks suggest a pure C3 diet for the rhinos sampled.

Conclusions

Stable isotope ratios from rhino hair from Kenya shows distinct differences between some geographic regions. All Nairobi rhinos have very similar δ13C and δ15N values; the δ13C values are compatible with a diet that is nearly 100% from C3 plants, and with δ15N diet values near 0 ‰. This does not indicate a reliance on Acacia because many other plants have values similar to this in this mesic part of Kenya. Rhinos from Mara also have a diet comprising almost entirely C3 plants, but most of them have a slightly higher δ15N value than do Nairobi rhinos. Rhinos from Laikipia (Ol Jogi, Ol Pejeta, Mugie) show evidence for a portion of C4/CAM in the diet; the percentage of C4/CAM could reach quantities of up to 40% for short periods of time based on a detailed isotope profile in hair. Rhinos from Tsavo East have the highest δ15N values; other mammals from the Tsavo region also have high δ15N values (Cerling et al. 2004; Yeakel et al. 2009) and this is likely due to the xeric character of the landscape with associated high δ15N values in many plants.

87Sr/86Sr values of rhinos from our study areas range from about 0.705 to 0.714. Lower values, from 0.705 to 0.708, are associated with the regions where the local geology comprises primarily of Neogene and Quaternary volcanic rocks. The highest values, from 0.709 to 0.714, are from Tsavo East and the Masai Mara, both of which are dominated by Mozambique Belt basement rocks. The 87Sr/86Sr ratios from Ngulia sanctuary rhino in Tsavo West appear to represent a mixing between the basement Mozambique Belt rocks and the waters derived from Mzima Springs that are sourced in young Quaternary volcanics. Rhinos from Nakuru and Nairobi have 87Sr/86Sr ratios between 0.7069 and 0.7075; Laikipia rhinos have 87Sr/86Sr values between 0.705 and 0.708, with increasing 87Sr/86Sr ratios from the southeast (Solio–Ol Pejeta) to the northwest (Ol Jogi–Mugie).

In summary, the light stable isotopes (δ13C and δ15N) are good indicators of local ecology, whereas 87Sr/86Sr ratios indicate local geological bedrock. Together, these suggest a possible application for forensic use as well as having ecological applications.

Notes

Acknowledgements

We thank the members of Kenya Wildlife Service for assistance in collecting rhino hair samples, Truman Young and Meave Leakey for collecting plant samples, and Nicholas Georgiadis for water and temperature data from the Laikipia region. We thank the government of Kenya for permission to do this work. We thank IsoForensics for making the Neptune MC-ICP-MS available for this study. This work was done under CITES permits US831854, US053837/9, US159997/9, and US08996A/9. Dr. Samuel Andanje died on 4 May 2015 while this manuscript was in the initial stages of preparation; the living authors are grateful to him for his work on this project.

Author contribution statement

TEC and SAA conceived and designed the experiments. TEC, SAA, FG, JMK, LK, JWK, CK, IL, and ANM carried out the field work and laboratory analyses were performed by ANM, CRA, DPF, LH, and SJT. TEC, SAA, and DPF analyzed the data. TEC and SAA wrote the manuscript.

Supplementary material

442_2018_4185_MOESM1_ESM.xlsx (45 kb)
Supplementary material 1 (XLSX 44 kb) Appendix SI 1. Data on individual histories of rhinos sampled in this study
442_2018_4185_MOESM2_ESM.xlsx (114 kb)
Supplementary material 2 (XLSX 114 kb) Appendix SI II. δ13C and δ15N values of sequential hair samples for 7 rhinos from Kenya; diet input δ13C and δ15N values, and estimated fraction of C4/CAM in diet calculated as described in text
442_2018_4185_MOESM3_ESM.xlsx (48 kb)
Supplementary material 3 (XLSX 48 kb) Appendix SI III. δ13C, δ15N, and C/N ratios of plants collected from the regions of study from 1997 to 2007
442_2018_4185_MOESM4_ESM.docx (92 kb)
Supplementary material 4 (DOCX 91 kb) Table SI 1. δ13C, δ15N, and 87Sr/86Sr of bulk hair samples from black rhinos in this study
442_2018_4185_MOESM5_ESM.docx (46 kb)
Supplementary material 5 (DOCX 45 kb) Table SI 2. Average δ13C and δ15N values for plants from Laikipia, Nairobi, and Tsavo regions collected between 1997 and 2007
442_2018_4185_MOESM6_ESM.docx (99 kb)
Supplementary material 6 (DOCX 99 kb) Table SI 3. δ13C and δ18O of teeth from rhinos (Diceros bicornis), elephant (Loxodonta africana), Burchell’s zebra (Equus burchelli), and giraffe (Giraffa camelopardalis) from Tsavo East National Park, all from approximately 1970. δ13C1750 is the δ13C value corrected to 1750 due the change in the δ13C of the atmosphere due to fossil fuel burning (NOAA, 2017); percent C4 is calculated assuming the average δ13C values for equids and giraffe represent pure C4 and pure C3 diets, respectively and assuming all taxa have the same isotope enrichment factor for diet to enamel (see Cerling et al., 1999)

References

  1. Amin R, Okita-Ouma B, Adcock K, Emslie RH, Mulama M, Pearce-Kelly P (2006) An integrated management strategy for the conservation of Eastern black rhinoceros Diceros bicornis michaeli in Kenya. Int Zoo Yearb 40(1):118–129CrossRefGoogle Scholar
  2. Ayliffe LK, Cerling TE, Robinson T, West AG, Sponheimer M, Passey BH, Hammer J, Roeder B, Dearing MD, Ehleringer JR (2004) Turnover of carbon isotopes in tail hair and breath CO2 of horses fed an isotopically varied diet. Oecologia 139:11–22CrossRefPubMedGoogle Scholar
  3. Birkett A (2002) The impact of giraffe, rhino and elephant on the habitat of a black rhino sanctuary in Kenya. Afr J Ecol 40:276–282CrossRefGoogle Scholar
  4. Blum JD, Taliaferro EH, Holmes RT (2001) Determining the sources of calcium for migratory songbirds using stable strontium isotopes. Oecologia 126:569–574CrossRefPubMedGoogle Scholar
  5. Blumenthal SA, Levin NE, Brown FH, Brugal J-P, Chritz KL, Harris JM, Jehle GE, Cerling TE (2017) Aridity and hominin environments. Proc Natl Acad Sci 114:7331–7336CrossRefPubMedGoogle Scholar
  6. Brennan SR, Fernandez DP, Zimmerman CE, Cerling TE, Brown RJ, Wooller MJ (2015a) Strontium isotopes in otoliths of a non-migratory fish (slimy sculpin): implications for provenance studies. Geochim Cosmochim Acta 149:32–45CrossRefGoogle Scholar
  7. Brennan SR, Zimmerman CE, Fernandez DP, Cerling TE, McPhee MV, Wooller MJ (2015b) Strontium isotopes delineate fine-scale natal origins and migration histories of Pacific salmon. Sci Adv 1(4):e1400124.  https://doi.org/10.1126/sciadv.1400124 CrossRefPubMedPubMedCentralGoogle Scholar
  8. Buk KG, Knight MH (2010) Seasonal diet preferences of black rhinoceros in three arid South African National Parks. Afr J Ecol 48(4):1064–1075CrossRefGoogle Scholar
  9. Cerling TE, Harris JM (1999) Carbon isotope fractionation between diet and bioapatite in ungulate mammals and implications for ecological and paleoecological studies. Oecologia 120:347–363CrossRefPubMedGoogle Scholar
  10. Cerling TE, Harris JM, Passey BH (2003) Dietary preferences of East African Bovidae based on stable isotope analysis. J Mamm 84:456–471CrossRefGoogle Scholar
  11. Cerling TE, Passey BH, Ayliffe LK, Cook CS, Ehleringer JR, Harris JM, Dhidha MB, Kasiki SM (2004) Orphans’ tales: seasonally dietary changes in elephants from Tsavo National Park, Kenya. Palaeogeogr Palaeoclimatol Palaeoecol 206:367–376CrossRefGoogle Scholar
  12. Cerling TE, Wittemyer G, Rasmussen HB, Vollrath F, Cerling CE, Robinson TJ, Douglas-Hamilton I (2006) Stable isotopes in elephant hair documents migration patterns and diet changes. Proc Natl Acad Sci 103:371–373CrossRefPubMedGoogle Scholar
  13. Cerling TE, Ayliffe LK, Dearing MD, Ehleringer JR, Passey BH, Podlesak DW, Torregrossa A-M, West AG (2007a) Determining biological tissue turnover using stable isotopes: the reaction progress variable. Oecologia 151:175–189CrossRefPubMedGoogle Scholar
  14. Cerling TE, Omondi P, Macharia AN (2007b) Diets of Kenyan elephants from stable isotopes and the origin of confiscated ivory in Kenya. J Afr Ecol 45:614–623CrossRefGoogle Scholar
  15. Cerling TE, Wittemyer G, Ehleringer JR, Remien CH, Douglas-Hamilton I (2009) History of Animals using Isotope Records (HAIR): a 6-year dietary history of one family of African elephants. Proc Natl Acad Sci 106:8093–8100CrossRefPubMedGoogle Scholar
  16. Coutu AN, Lee-Thorp J, Collins MJ, Lane PJ (2016) Mapping the elephants of the 19th century East African ivory trade with a multi-isotope approach. PLoS One 11(10):e0163606CrossRefPubMedPubMedCentralGoogle Scholar
  17. Dharani N, Kinyamario JI, Wagacha PW, Rodrigues AJ (2009) Browsing impact of large herbivores on Acacia xanthophloea Benth in Lake Nakuru National Park, Kenya. Afr J Ecol 47:184–191CrossRefGoogle Scholar
  18. Emslie RH (2013) African rhinoceroses—latest trends in rhino numbers and poaching. CITES—CoP16 Inf. 51. An update to Doc 54-2-Annexe 2 from the IUCN Species Survival Commission’s (IUCN/SSC) African Rhino Specialist Group to the CITES Secretariat pursuant to Resolution Conf. 9.14 (Rev. CoP15), pp 54-2Google Scholar
  19. Ganqa NM, Scogings PF, Raats JG (2005) Diet selection and forage factors affecting woody plant selection by black rhinoceros in the Great Fish River Reserve, South Africa. S Afr J Wildl Res 35:77–83Google Scholar
  20. Goddard J (1970) Food preferences of black rhinoceros in the Tsavo National Park. Afr J Ecol 8:145–161CrossRefGoogle Scholar
  21. Graustein WC, Armstrong RL (1983) The use of strontium-87/strontium-86 ratios to measure atmospheric transport into forested watersheds. Science 219:289–292CrossRefPubMedGoogle Scholar
  22. Hoppe KA, Koch PL, Carlson RW, Webb SD (1999) Tracking mammoths and mastodons: reconstruction of migratory behavior using strontium isotope ratios. Geology 27:439–442CrossRefGoogle Scholar
  23. Kennedy BP, Folt CL, Blum JD, Chamberlain CP (1997) Natural isotope markers in salmon. Nature 387:766–767CrossRefGoogle Scholar
  24. Kennedy BP, Blum JD, Folt CL, Nislow KH (2000) Using natural strontium isotopic signatures as fish markers: methodology and application. Can J Fish Aquat Sci 57:2280–2292CrossRefGoogle Scholar
  25. Kenya National Atlas (1962) Geological map of Kenya. Survey of Kenya. Nairobi, Kenya. 1 sheetGoogle Scholar
  26. Koch PL, Heisinger J, Moss C, Carlson RW, Fogel ML, Behrensmeyer AK (1995) Isotopic tracking of change in diet and habitat use in African elephants. Science 267:1340–1343CrossRefPubMedGoogle Scholar
  27. Lieverloo RJ, Schulling BF, de Boer WF, Lent PC, de Jong CB, Brown D, Prins HHT (2009) A comparison of faecal analysis with backtracking to determine the diet composition and species preference of the black rhinoceros (Diceros bicornis minor). Eur J Wildl Res 55:505–515CrossRefGoogle Scholar
  28. Mackey GN, Fernandez DP (2011) American Geophysical Union Fall Meeting, San Francisco, CA, USAGoogle Scholar
  29. Muya SM, Oguge NO (2000) Effects of browse availability and quality on black rhino (Diceros bicornis michaeli Groves 1967) diet in Nairobi National Park, Kenya. Afr J Ecol 38:62–71CrossRefGoogle Scholar
  30. Ngene S, Bitok E, Mukeka J, Okita-Ouma B, Gakuya F, Omondi P, Kimitei K, Watol Y, Kimani J (2011) Census and ear-notching of black rhinoceros (Diceros bicornis michaeli) in Tsavo East National Park, Kenya. Pachyderm 49:61–69Google Scholar
  31. NOAA (2017) Carbon-13/carbon-12 ratios in carbon dioxide. https://www.esrl.noaa.gov/gmd/webdata/iadv/ccgg/graphs/pdfs/ccgg.MLO.co2c13.1.none.discrete.all.pdf. Downloaded on 11 March 2017
  32. Okita-Ouma B, Mijele D, Amin R, Gakuya F, Ndeereh D, Lekolool I, Omondi P, Woodley D, Litoroh M, Bakari J, Kock R (2008) Minimizing competition by removing elephants from a degraded Ngulia rhino sanctuary, Kenya. Pachyderm 44:80–87Google Scholar
  33. Okita-Ouma B, Amin R, Van Langevelde F, Leader-Williams N (2010) Density dependence and population dynamics of black rhinos (Diceros bicornis michaeli) in Kenya’s rhino sanctuaries. Afr J Ecol 48:791–799Google Scholar
  34. Oloo TW, Brett R, Young TP (1994) Seasonal variation in the feeding ecology of black rhinoceros (Diceros bicornis L.) in Laikipia, Kenya. Afr J Ecol 32:142–157CrossRefGoogle Scholar
  35. Passey BH, Cerling TE (2002) Tooth enamel mineralization in ungulates: implications for recovering a primary isotopic time-series. Geochim Cosmochim Acta 18:3225–3234CrossRefGoogle Scholar
  36. Patton F, Jones M (2007) Determining minimum population size and demographics of black rhinos in the Salient of Aberdare National Park, Kenya. Pachyderm 43:63–72Google Scholar
  37. Patton F, Campbell P, Parfet E (2007) Establishing a monitoring system for black rhinos in the Solio Game Reserve, central Kenya. Pachyderm 43:87–95Google Scholar
  38. Patton F, Campbell P, Parfet E (2008) Biological management of the high density black rhino population in Solio Game Reserve, central Kenya. Pachyderm 44:72–79Google Scholar
  39. Pienaar DJ, Hall-Martin AJ, Hitchins PM (1991) Horn growth rates of free-ranging white and black rhinoceros. Koedoe 34:97–105Google Scholar
  40. Quade J, Chivas AR, McCulloch MT (1995) Strontium and carbon isotope tracers and the origins of soil carbonate in South Australia and Victoria. Palaeogeogr Palaeoclimatol Palaeoecol 113:103–117CrossRefGoogle Scholar
  41. Quennel AM (1959) Geological map of Tanganyika. Geological Survey Department. Dodoma, Tanganyika (1 sheet)Google Scholar
  42. Spinage C (1994) Elephants. T. & A.D. Prosser, LondonGoogle Scholar
  43. Sponheimer M, Lee-Thorp JA, DeRuitter DJ, Smith JM, van der Merwe NJ, Reed K, Grant CC, Ayliffe LK, Robinson TF, Heidelberger C, Marcus W (2003) Diets of the Southern African bovidae: stable isotope evidence. J Mamm 8:471–479CrossRefGoogle Scholar
  44. Tipple BJ, Chau T, Chesson LA, Fernandez DP, Ehleringer JR (2013) Isolation of strontium pools and isotope ratios in modern human hair. Anal Chim Acta 798:64–73CrossRefPubMedGoogle Scholar
  45. van der Merwe NJ, Lee-Thorp JA, Bell RHV (1988) Carbon isotopes as indicators of elephant diets and African environments. Afr J Ecol 26:163–172CrossRefGoogle Scholar
  46. van der Merwe NJ, Lee-Thorp JA, Thackeray JF, Hall-Martin A, Kruger FJ, Coetzee H, Bell RHV, Lindeque M (1990) Source area determination of elephant ivory by isotopic analysis. Nature 346:744–746CrossRefGoogle Scholar
  47. Vitousek PM, Kennedy MJ, Derry LA, Chadwick OA (1999) Weathering versus atmospheric sources of strontium in ecosystems on young volcanic soils. Oecologia 121:255–259CrossRefPubMedGoogle Scholar
  48. Vogel JC, Eglington B, Auret JM (1990) Isotope fingerprints in elephant bone and ivory. Nature 346:747–749CrossRefGoogle Scholar
  49. Walpole MJ (2002) Factors affecting black rhino monitoring in Masai Mara National Reserve, Kenya. Afr J Ecol 40:18–25CrossRefGoogle Scholar
  50. Walpole MJ, Morgan-Davies M, Milledge S, Bett P, Leader-Williams N (2001) Population dynamics and future conservation of a free-ranging black rhinoceros (Diceros bicornis) population in Kenya. Biol Conserv 99:237–243CrossRefGoogle Scholar
  51. West AG, Ayliffe LK, Cerling TE, Robinson TF, Karren B, Dearing MD, Ehleringer JR (2004) Short-term diet changes revealed using stable carbon isotopes in horse tail-hair. Funct Ecol 18:616–624CrossRefGoogle Scholar
  52. Western D (1982) Patterns of depletion in a Kenya rhino population and the conservation implications. Biol Conserv 24:147–156CrossRefGoogle Scholar
  53. Wittemyer G, Cerling TE, Douglas-Hamilton I (2009) Establishing chronologies from isotopic profiles in serially collected animal tissues: an example using tail hairs from African elephants. Chem Geol 267:3–11CrossRefGoogle Scholar
  54. Yeakel JD, Patterson BD, Fox-Dobbs K, Okumura MM, Cerling TE, Moore JW, Koch PL, Dominy NJ (2009) Cooperation and individuality among man-eating lions. Proc Natl Acad Sci 106:19040–19043CrossRefPubMedGoogle Scholar
  55. Zazzo A, Harrison SM, Bahar B, Moloney AP, Monahan FJ, Scrimgeour CM, Schmidt O (2007) Experimental determination of dietary carbon turnover in bovine hair and hoof. Can J Zool 85:1239–1248.  https://doi.org/10.1139/Z07-110 CrossRefGoogle Scholar
  56. Zazzo A, Mononey AP, Monahan FJ, Scrimgeour CM, Schmidt O (2008) Effect of age and food intake on dietary carbon turnover recorded in sheep wool. Rapid Commun Mass Spectrom 22:2937–2945CrossRefPubMedGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  • Thure E. Cerling
    • 1
    • 2
  • Samuel A. Andanje
    • 3
  • Francis Gakuya
    • 3
  • John M. Kariuki
    • 3
  • Linus Kariuki
    • 3
  • Jackson W. Kingoo
    • 3
  • Cedric Khayale
    • 3
  • Isaac Lekolool
    • 3
  • Anthony N. Macharia
    • 4
  • Christopher R. Anderson
    • 1
  • Diego P. Fernandez
    • 1
  • Lihai Hu
    • 1
  • Shawn J. Thomas
    • 1
  1. 1.Department of Geology and GeophysicsUniversity of UtahSalt Lake CityUSA
  2. 2.Department of BiologyUniversity of UtahSalt Lake CityUSA
  3. 3.Kenya Wildlife ServiceNairobiKenya
  4. 4.Department of GeographyKenyatta UniversityNairobiKenya

Personalised recommendations