See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/325359354

Diet of sympatric wild and domestic ungulates in southern Mongolia by DNA

barcoding analysis

Article  in  Journal of Mammalogy · March 2018

CITATIONS READS

0 18

7 authors, including:

Lkhagvasuren Badamjav
Institute of General and Experimental Biology, Mongolian Academy of Sciences
51 PUBLICATIONS   886 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Migration ecology and conservation of Mongolian wild ungulates View project

All content following this page was uploaded by Lkhagvasuren Badamjav on 25 May 2018.

The user has requested enhancement of the downloaded file.

 Journal of Mammalogy, 99(2):450–458, 2018
DOI:10.1093/jmammal/gyx182
Published online January 20, 2018

Diet of sympatric wild and domestic ungulates in southern

Mongolia by DNA barcoding analysis
Taro Sugimoto,* Takehiko Y. Ito, Takeshi Taniguchi, Badamjav Lkhagvasuren, Tsendsuren Oyunsuren,
Yumi Sakamoto, and Norikazu Yamanaka
Arid Land Research Center, Tottori University, 1390 Hamasaka, Tottori 680-0001, Japan (TS, TYI, TT, YS, NY)
Institute of General and Experimental Biology, Mongolian Academy of Sciences, Ulaanbaatar 13330, Mongolia (BL, TO)
* Correspondent: tasugi@ees.hokudai.ac.jp

The impact of increasing numbers of livestock on wild ungulates has been a concern in Mongolia. We used DNA
barcoding via next generation sequencing to reveal detailed food habits and assess the extent of potential resource
competition among ungulates in the desert steppe of southern Mongolia. From feces of 2 wild (Mongolian gazelle
[Procapra gutturosa], Asiatic wild ass [Equus hemionus]) and 4 domestic (sheep [Ovis aries], goat [Capra
hircus], horse [Equus caballus], camel [Camelus bactrianus]) ungulates collected in June 2014, we identified 25
plant families (39 to genus and 18 to species). Dietary diversity was the highest for sheep (Levins’ standardized
index = 0.351), followed by goats (0.342), and the lowest for gazelles (0.220). Gazelles had a high dietary overlap
with goats (Pianka’s index = 0.654) and sheep (0.630). Wild asses had the highest dietary overlap with horses
(0.847), followed by goats (0.727) and sheep (0.719). The high dietary overlaps may be explained by similar
digestive systems and body sizes among gazelles, sheep, and goats and between wild asses and horses, and by
the high dietary diversity of sheep and goats. The present study shows more diverse food compositions than
previous studies, suggesting that because of the high dietary diversity of sheep and goats and their overlap with
wild species, their increasing number is a potential risk for the survival of wild ungulates.

Key words:  DNA barcoding, dryland, food habits, Mongolian gazelle, resource competition, trnL approach, wild ass

Mongolia’s Gobi-Steppe Ecosystem is the largest area of intact
steppe in the world (Batsaikhan et al. 2014), where traditional
nomadic herders and wildlife coexist. Since the economic
transformation of 1990, the number of livestock, such as sheep
(Ovis aries) and goats (Capra hircus), has been increasing.
Overgrazing by livestock has resulted in changes in vegetation
structure (Sasaki et al. 2008) and land degradation (Hilker et al.
2014). Therefore, the impact of livestock grazing on native
wildlife has been a concern in Mongolia.
The desert steppe in southern Mongolia is an important habi-
tat for multiple ungulate species, such as the Mongolian gazelle
(Procapra gutturosa), Asiatic wild ass (or khulan, Equus
hemionus), and goitered gazelle (Gazella subgutturosa). These
species are known to migrate over long distances (Ito et al.
2013; Imai et al. 2017), which means that they require vast
areas of continuous habitat to survive. These wild ungulates
are now faced with the risk of habitat fragmentation by rail-
ways and border fences (Ito et al. 2005, 2008, 2013; Kaczensky
et al. 2011), but increasing livestock populations and vegetation
degradation by overgrazing also are likely to have a negative
impact on wild ungulates as a result of enhanced competition
for habitat and resources.
A few studies have investigated the patterns of resource use by
Mongolian gazelles and livestock in southern Mongolia to infer
their competitive relationships using microhistological analyses
(Campos-Arceiz et al. 2004; Yoshihara et al. 2008). However,
the food compositions observed were derived from limited cat-
egorical data, and fine-scale dietary differences among sym-
patric ungulates were not detected. Microhistological analysis
is labor intensive and requires strong taxonomic skills because
it is generally difficult to identify digested plant fragments in
feces at low taxonomic levels (Soininen et al. 2009). These
drawbacks hamper the accrual of accurate knowledge regarding
food habits and reliable estimation of dietary overlap among
sympatric ungulates.
We used a DNA barcoding approach by means of next gen-
eration sequencing. Next generation sequencing generates a
large amount of sequence data from samples composed of a

© 2018 American Society of Mammalogists, www.mammalogy.org

450

Downloaded from https://academic.oup.com/jmammal/article-abstract/99/2/450/4818386

by Tottori University Tottori Campus user
on 04 April 2018
 SUGIMOTO ET AL.—DIET OF MONGOLIAN WILD AND DOMESTIC UNGULATES
mixture of DNA from various species, allowing the simultane-
ous identification of multiple species (Valentini et al. 2009a).
With this approach, we aimed to reveal the detailed food habits
and assess the extent of potential resource competition among
sympatric wild and domestic ungulates in the desert steppe of
southern Mongolia.
Materials and Methods
Study area.—The study area was located in the des-
ert steppe of southern Mongolia (Fig. 1; 42°55′–43°32′N,
108°42′–110°13′E) with sparse vegetation, dominated by
central Asian desert plants such as Allium spp., Stipa spp.,
Salsola spp., Caragana spp., Artemisia spp., Nitraria sibirica,
Cleistogenes soongorica, and Reaumuria soongorica. The cli-
mate is strongly continental and arid, with cold winters (min-
imum temperatures below −35°C), dry and windy springs,
and relatively wet and hot summers (maximum temperatures
above 40°C); annual precipitation is about 100 mm (Dorjgotov
et al. 2004). Seminomadic pastoralists live at some of the low-
est population densities in the country (0.79 people/km2—
National Statistical Office of Mongolia 2010). Wild ungulates
found in the study area are Mongolian gazelle, Asiatic wild
ass, and goitered gazelle, which are listed as Least Concern,
Near Threatened, and Vulnerable, respectively, on the IUCN
Red List (IUCN 2017). The Mongolian gazelle is distrib-
uted in eastern Mongolia and adjacent areas of Russia and
northeastern China, and its population size is estimated to be
approximately 1,000,000 in Mongolia (IUCN SSC Antelope
Specialist Group 2016). The Asiatic wild ass and goitered
gazelle are distributed in multiple countries in Central Asia,
and a recent study estimated their population sizes in southern
Mongolia to be 35,899 and 28,462, respectively (Buuveibaatar
et al. 2017). The Mongolian desert steppe hosts the world’s
largest populations of these 3 ungulates (IUCN SSC Antelope
Specialist Group 2016; Buuveibaatar et al. 2017). Major live-
stock in the area are sheep, goats, horses (Equus caballus), and
camels (Camelus bactrianus). Cattle (Bos taurus) were rarely
observed in the study area (Yoshihara et al. 2008). Horses and
Fig. 1.—Location of collection sites of feces of Mongolian gazelles (Procapra gutturosa; white diamonds) and wild asses (Equus hemionus; black
circles) in southern Mongolia, collected 19–26 June 2014.
Downloaded from https://academic.oup.com/jmammal/article-abstract/99/2/450/4818386
by Tottori University Tottori Campus user
on 04 April 2018
451
camels graze freely, whereas sheep and goats are herded by
herders.
Sampling.—From 19 to 26 June 2014, we searched southern
Mongolia for herds of gazelles and wild asses with a 4-wheel
drive vehicle and collected feces left behind by the herds we
found (Fig. 1). Around those areas where we collected feces
of wild ungulates, we searched for livestock (i.e., sheep, goats,
horses, and camels) and collected their feces. We also collected
ungulate feces in areas around water points. We assumed 1 pile
of feces to be from 1 individual. We did not collect old feces
or piles that could not be clearly distinguished from surround-
ing piles. We collected at least 5 pellets from gazelles, sheep,
and goats, and 1–3 pellets from wild asses, horses, and camels.
Because we collected only a few feces from goitered gazelles,
we did not include this species in our analyses. At every site
where we collected ungulate feces, we also collected leaves
from plant species to develop a local sequence database, which
improves the precision of DNA barcoding (Valentini et al.
2009b). All samples were collected in plastic bags at ambient
temperature and kept in a freezer (−20°C) in a laboratory until
DNA extraction.
DNA extraction and species identification.—Fecal DNA
was extracted from about 20 mg of feces, obtained from 1–3
pellets, using DNeasy Plant Mini Kit (Qiagen, Tokyo, Japan)
following the manufacturer’s protocol. We first conducted spe-
cies identification by sequencing a part of the mitochondrial
cytochrome b gene with the primers developed in this study
(Table 1). Polymerase chain reaction (PCR) was performed in a
final volume of 20 µl, including 1× AmpliTaq Gold 360 Master
Mix (Applied Biosystems, Foster City, California), 0.35 µM of
each forward and reverse primer, and 1.0 µl of DNA extract.
The reaction conditions were 95°C for 10 min, followed by 40
cycles of 30 s at 95°C, 30 s at 56°C, and 30 s at 72°C, and a final
extension for 10 min at 72°C. PCR products were sequenced
in the reverse direction using a BigDye Terminator v3.1 Cycle
Sequencing Kit (Applied Biosystems) and an ABI 3130 auto-
matic sequencer (Applied Biosystems).
Sequencing of the P6 loop of the trnL (UAA) intron.—
For each of the samples, for which species identification was
 452 JOURNAL OF MAMMALOGY
Table 1.—Primer sequences for species identification amplifying a part of the cytochrome b gene. Species: Mongolian gazelle (Procapra
gutturosa), sheep (Ovis aries), goat (Capra hircus), wild ass (Equus hemionus), horse (Equus caballus), camel (Camelus bactrianus).
Name Sequence 5′-3′
YHG2F GGCCTATTCCTAGCAATACAC
YHG1R CTAGGTTTGTGCCAATATATGG
RUN1F TACCTACATGCCAACGGAGC
RUN1R GAAGGATATTTGTCCTCATGG
a
Amplified length includes primers.
successful, we amplified the P6 loop of the chloroplast trnL
(UAA) intron with primers g and h (Taberlet et al. 2007). In
accordance with the Ion Amplicon Library Preparation Fusion
Method (Life Technologies, Foster City, California; User
Guide #4468326), the 5′ end of the forward primer was fused
with the A-adaptor, barcode, and barcode adapter, and the 5′
end of the reverse primer was fused with the trP1 adaptor.
Barcodes allow the assignment of sequences to the respec-
tive samples. PCR was performed in a final volume of 25 µl,
including 22 µl of Platinum PCR SuperMix High Fidelity
(Invitrogen, Carlsbad, California), 0.4 µM of each modi-
fied forward and reverse primer, and 1.0 µl of DNA extract
diluted to 100 times. The reaction conditions were 94°C for
2 min followed by 40 cycles of 30 s at 94°C, 30 s at 55°C, and
1 min at 68°C. The PCR products were purified (Agencourt
AMPure XP; Beckman Coulter, Brea, California) and quanti-
fied (Qubit dsDNA HS Assay Kit; Invitrogen) and were then
mixed together to obtain approximately the same number of
molecules per PCR product corresponding to the different
fecal samples. The quality and size of the mixed PCR prod-
uct were assessed with the 2100 Bioanalyzer High Sensitivity
DNA kit (Agilent Technologies, Santa Clara, California), and
we adjusted the concentration to 6 pM and used it as the tem-
plate for emulsion PCR. Emulsion PCR was performed on the
Ion OneTouch 2 system using the Ion PGM Template OT2 400
kit (Life Technologies), and sequencing was performed on the
Ion Torrent PGM system using the Ion Torrent 318 chip kit
v2 and an Ion PGM Sequencing 400 kit (Life Technologies),
according to the respective manufacturer’s protocols.
Development of the reference database.—For DNA barcod-
ing, we built local and public reference databases from plant
DNA sequences obtained in this study and from those in the
EMBL nucleotide library, respectively. About 20 mg of plant
tissues, collected from the study area, was used to extract plant
DNA with the DNeasy Plant Mini Kit (Qiagen) following the
manufacturer’s protocol. The trnL (UAA) intron was ampli-
fied using primers c and h (Taberlet et al. 2007). PCR was per-
formed in a final volume of 20 µl, including 1× AmpliTaq Gold
360 Master Mix (Applied Biosystems), 0.3 µM of each forward
and reverse primer, and 5–10 ng of DNA extract, and the reac-
tion conditions were those described by Taberlet et al. (2007).
PCR products were sequenced using a BigDye Terminator v3.1
Cycle Sequencing Kit (Applied Biosystems) and an ABI 3130
automatic sequencer (Applied Biosystems). We extracted the P6
loop sequences from the trnL (UAA) intron sequences obtained
and from the EMBL nucleotide library (release 125) using the
Downloaded from https://academic.oup.com/jmammal/article-abstract/99/2/450/4818386
by Tottori University Tottori Campus user
on 04 April 2018
Amplified length (bp)a Target species
340 Gazelle, sheep, goat
180 Wild ass, horse, camel
program ecoPCR (available at http://www.grenoble.prabi.fr/
trac/ecoPCR) and built “local” and “public” sequence data-
bases, respectively.
Sequence analysis and taxon assignation.—We only consid-
ered sequence reads with a perfect match on tags (i.e., bar-
code and barcode adaptor sequences). The Ion Torrent Suite
Software (Life Technologies) trimmed the tags and sorted out
sequences based on sample-specific tags. We analyzed the
sequences using the programs TagCleaner (Schmieder et al.
2010) and PRINSEQ (Schmieder and Edwards 2011) in addi-
tion to using multiple programs from the OBITOOLS package
(Boyer et al. 2016). First, we trimmed the primer sequence
using TagCleaner, where a maximum of 2 errors on the prim-
ers was allowed. Then, we discarded sequences with lengths
out of the 10–150 bp range and those with a mean quality score
< 25, and we output the sequences in FASTA format using
PRINSEQ. We clustered identical sequences together using
obiuniq program and discarded sequences with occurrence
< 10 using obigrep program. We then used obiclean program
to detect erroneous sequences by giving the status of head,
singleton, or internal to each sequence (Boyer et al. 2016).
Internal corresponds to erroneous sequences derived from
amplification or sequencing errors; therefore, we did not con-
sider these sequences for subsequent taxon assignation. Taxon
assignation was performed using the ecoTag program with
the local and public sequence databases (Pegard et al. 2009).
We discarded sequences with a best match score < 0.95 and a
total number of counts < 1% of the most common sequence
in each sample to avoid erroneous sequences (e.g., chimeric
sequences and contamination) and to increase the accuracy of
taxon assignation. We investigated the distribution of all the
taxonomic identifiers (taxids) selected with the highest match
score in local and public databases by referring to 3 refer-
ences: Grubov (2001), Undarmaa et al. (2015), and Virtual
Guide to the Flora of Mongolia (http://greif.uni-greifswald.
de/floragreif/). We discarded any taxa that were not distributed
in southern Mongolia. If 2 or more taxa were assigned to a
given sequence with the same match score, we assigned the
sequence the lowest taxonomic level that included both taxa.
Species or genus names were accepted if the best match score
was ≥ 0.98, whereas a family name was accepted if the best
score was ≥ 0.95 and < 0.98.
Data analysis.—Food composition was expressed as the
frequency of occurrence (percentage of feces in which a plant
taxon was detected) at the level of plant family and the level
of molecular operational taxonomic units (MOTUs). We
 SUGIMOTO ET AL.—DIET OF MONGOLIAN WILD AND DOMESTIC UNGULATES
evaluated dietary niche breadth, representing dietary diversity,
using the Levins’ index (Levins 1968) standardized by Colwell
and Futuyma (1971). The standardized index value ranges from
0 (minimum niche breadth) to 1 (maximum niche breadth). We
estimated the extent of dietary overlap among 6 ungulates using
Pianka’s niche overlap index (Pianka 1973), which ranges from
0 (no dietary overlap) to 1 (complete dietary overlap). These 2
indexes were calculated on the basis of the frequency of occur-
rence at the family level.
We assessed the differences in food composition at the
family level between all pairs of 6 ungulates using Fisher’s
exact test (with the use of contingency tables), apply-
ing the Benjamini–Hochberg correction for multiple tests
(Benjamini and Hochberg 1995). Raw occurrence data were
used and rare plant families, defined as those for which the
total number of raw occurrences between the 2 comparative
objects was < 5, were excluded from the food compositions
in making the comparison. Rarefaction analyses were con-
ducted using the program EstimateS 9.1.0 (Colwell 2013) to
evaluate the effect of sample size on the detected number of
plant families and MOTUs.
Results
For the local sequence database, we collected 71 plants belong-
ing to 53 species, 45 genera, and 23 families. Of the 71, 5 plants
were identified only to the family level and 13 plants to the
genus level due to uncertainties in the plant species identifi-
cation. All the plant sequences were deposited in GenBank
(accession numbers LC309022–LC309093).
We used 192 fecal samples for next generation sequencing
(gazelle = 27, wild ass = 39, sheep = 31, goat = 39, horse = 24,
and camel = 32). The Ion Torrent PGM system generated a total
of 13.2 × 106 sequence reads. These were derived from 4 runs
of 318 tips. After the filtering processes, 6,525,211 sequence
reads (49% of the initial outcome) were retained for subsequent
taxonomic assignation. The average number of sequence reads
per sample was 33,985.
Taxonomic assignation identified 57 MOTUs at the species
(18), genus (23), and family (16) levels (Table 2). The number
of MOTUs per sample ranged from 1 to 12 with an average of
4.6 (average number of MOTUs per sample: gazelle = 3, wild
ass = 5.5, sheep = 5.4, goat = 5.3, horse = 4.6, and camel = 3.4).
Frequency of occurrence at the family level indicated that of
the 25 families we detected, Poaceae and Fabaceae were fre-
quently consumed by all 6 ungulates (Fig. 2). The most fre-
quently observed plant taxon was Ephedra for gazelles, Stipa
for wild asses and horses, and Reaumuria soongarica for cam-
els (Table 2). Sheep and goats showed similar food composi-
tions and frequently consumed Stipa and Fabaceae (Table 2).
Food compositions were significantly different between any
pair of the 6 ungulates (P < 0.001), except between sheep and
goats (P = 0.523).
High dietary diversity was observed for sheep (Levins’ stan-
dardized index = 0.351) and goats (0.342), followed by wild
asses (0.268), horses (0.246), camels (0.222), and gazelles
Downloaded from https://academic.oup.com/jmammal/article-abstract/99/2/450/4818386
by Tottori University Tottori Campus user
on 04 April 2018
453
(0.220). Gazelles and wild asses had relatively high dietary
overlap values with sheep and goats, whereas wild asses had
a higher overlap value with horses (Table 3). Sheep and goats
generally had high overlap values with all the ungulates. The
highest overlap value was observed between sheep and goats,
whereas the lowest value was observed between gazelles and
camels (Table 3).
Curves describing the cumulated numbers of plant families
approximately reached asymptotes (Fig. 3a), suggesting that
additional sampling would not significantly affect the results,
whereas those describing the cumulated numbers of MOTUs
did not appear to show the same tendency toward asymptotes
(Fig. 3b), leaving open the possibility that additional sampling
could reveal more MOTUs.
Discussion
We successfully revealed food habits of sympatric wild and
domestic ungulates that were more detailed than previous stud-
ies, in which only a few plant families were identified (Campos-
Arceiz et al. 2004; Yoshihara et al. 2008). These detailed food
compositions allowed us to reveal fine-scale dietary differ-
ences among sympatric ungulates in the desert steppe with a
DNA barcoding approach. This approach required a compre-
hensive reference sequence database for accurate plant species
identification. We used both local and public databases of the
P6 loop sequences; these 2 databases were complementary to
each other. Of the 71 plants in the local database, 24 were not
registered in the public database, whereas the public database
included some plant species consumed by ungulates that the
local database did not include (e.g., Dontostemon integrifolius).
Our local database still is far from complete, and further plant
sampling is required to improve the comprehensiveness of this
sequence database for the desert steppe of southern Mongolia.
With respect to the taxonomic assignation, not all plants
could be identified to the species level (Table 2). This was
not only due to the incompleteness of the reference databases
but also the relatively weak resolution power of the P6 loop
(Taberlet et al. 2007) within some plant families (e.g., Poaceae
and Asteraceae). For instance, 3 Stipa species (S. glareosa,
S. gobica, S. krylovii) in Poaceae had the same P6 loop
sequence, resulting in identification to the genus level; differ-
ent genera in Asteraceae (Ajania sp. and Artemisia sp.) also had
the same sequence, resulting in identification only to the family
level. Therefore, sequencing additional loci, such as the inter-
nal transcribed spacers of nuclear ribosomal DNA, is necessary
to improve taxonomic resolution (Baamrane et al. 2012).
Our study described dietary diversity at the level of plant
families, but did not fully describe dietary diversity at the level
of MOTUs. Not reaching an asymptote in the cumulated num-
ber of MOTUs may be related to the low number of MOTUs
per sample. The average number of MOTUs per sample in our
study (average = 4.6, range = 1–12) was much lower than those
in other dietary studies of ungulates using DNA barcoding.
Hibert et al. (2013) detected an average number of plant fami-
lies per sample to be 24 (range = 11–34), whereas Raye et al.
 454 JOURNAL OF MAMMALOGY

Table 2.—Food composition of 6 ungulates during summer 2014 in the desert steppe of southern Mongolia (% of frequency of occurrence).
Species: Mongolian gazelle (Procapra gutturosa), sheep (Ovis aries), goat (Capra hircus), wild ass (Equus hemionus), horse (Equus caballus),
camel (Camelus bactrianus).

Family Plant taxon Gazelle Wild ass Sheep Goat Horse Camel
Amaranthaceae Amaranthaceae 10.3 2.6 8.3 9.4
Anabasis brevifolia 5.1 3.2 4.2 6.3
Bassia 2.6 3.2 2.6
Salsola 2.6 12.5 31.3
Corispermum 2.6
Amaryllidaceae Amaryllidaceae 7.4 2.6 2.6 12.5
Allium 11.1 41.0 6.5 10.3 41.7 15.6
Apocynaceae Cynanchum 2.6
Asparagaceae Asparagus 9.7 2.6 6.3
Asteraceae Asteraceae 25.9 28.2 29.0 23.1 16.7 15.6
Ajania 3.7 4.2
Artemisia 33.3 7.7 19.4 7.7 12.5 3.1
Scorzonera 5.1 3.2 29.2
Brassicaceae Dontostemon crassifolius 6.5
Convolvulaceae Convolvulaceae 3.7 2.6 3.2
Convolvulus 44.4 12.8 35.5 43.6 8.3
Cyperaceae Carex 11.1 8.3
Ephedraceae Ephedraceae 5.1
Ephedra 77.8 41.9 48.7 3.1
Euphorbiaceae Euphorbia 4.2
Fabaceae Fabaceae 7.4 12.8 74.2 82.1 12.5 65.6
Caragana 3.2 7.7 3.1
Glycyrrhiza 12.9 7.7
Hedysarum 2.6 6.5 7.7 3.1
Hedysarum fruticosum 2.6
Lespedeza dahurica 3.7 2.6
Oxytropis 8.3
Iridaceae Iris bungei 3.2 7.7
Iris tenuifolia 2.6
Nitrariaceae Nitrariaceae 3.2 2.6
Nitraria sibirica 6.5 12.8 16.7 3.1
Peganum nigellastrum 8.3
Plumbaginaceae Plumbaginaceae 4.2
Limonium 3.7 9.7 5.1 8.3
Poaceae Poaceae 66.7 22.6 2.6 66.7 12.5
Chloris virgata 2.6
Echinochloa crusgalli 7.7
Stipa 14.8 87.2 90.3 76.9 91.7
Polygonaceae Polygonaceae 5.1 6.5 2.6
Atraphaxis 3.7 23.1 25.8 28.2 4.2 28.1
Calligonum mongolicum 35.9 6.5 15.4 28.1
Knorringia sibirica 5.1
Polygonum 3.7 5.1
Rheum nanum 18.5 17.9 3.2 3.1
Primulaceae Glaux maritima 22.6 7.7
Rosaceae Rosaceae 3.7 61.5 16.1 41.0 16.7 18.8
Prunus pedunculata 3.7 9.7 10.3
Rutaceae Haplophyllum dauricum 5.1 19.4 10.3
Salicaceae Salicaceae 3.7
Scrophulariaceae Scrophulariaceae 2.6
Solanaceae Solanaceae 3.7
Tamaricaceae Tamaricaceae 23.1 6.5 5.1
Reaumuria soongarica 11.1 61.5 25.8 20.5 45.8 75.0
Ulmaceae Ulmus 3.2 2.6
Zygophyllaceae Zygophyllaceae 2.6 3.1
Tribulus terrestris 2.6
Zygophyllum 10.3 3.1
Total number of molecular operational taxonomic units 21 30 32 36 23 20

Downloaded from https://academic.oup.com/jmammal/article-abstract/99/2/450/4818386

by Tottori University Tottori Campus user
on 04 April 2018
 SUGIMOTO ET AL.—DIET OF MONGOLIAN WILD AND DOMESTIC UNGULATES 455

Fig. 2.—Frequency of occurrence (%) of plant families consumed by 6 ungulates during summer 2014 in the desert steppe of southern Mongolia.
Plant families are ordered from left to right by decreasing frequency of occurrence in goats.

Table 3.—Pianka’s niche overlap indices based on food compo- extraction. We think that using multiple pellets and homogeniz-
sition between pairs of 6 ungulates in the desert steppe of southern ing them for DNA extraction could help to increase the number
Mongolia. Species: Mongolian gazelle (Procapra gutturosa), sheep of MOTUs per sample, hence, detect detailed dietary profiles
(Ovis aries), goat (Capra hircus), wild ass (Equus hemionus), horse with smaller sample sizes.
(Equus caballus), camel (Camelus bactrianus). Mongolian gazelles consumed Poaceae in low proportions;
however, they consumed Ephedra and Convolvulus in high pro-
Gazelle Wild ass Sheep Goat Horse Camel
portions. Although we should exercise caution when comparing
Gazelle 0.406 0.630 0.654 0.387 0.315 the results from 2 different techniques (i.e., DNA barcoding and
Wild ass 0.719 0.727 0.847 0.675
Sheep 0.959 0.754 0.622
microhistological analysis—Soininen et al. 2009), our result
Goat 0.678 0.654 was consistent with previous studies using microhistological
Horse 0.540 analysis that found low proportions of graminoids and high
Camel proportions of forbs and shrubs consumed in eastern Mongolia
during summer (Schaller 1998) and southern Mongolia dur-
ing early winter (Campos-Arceiz et al. 2004). Consumption of
(2011) detected an average number of plant species per sample low to modest amounts of graminoids has also been reported
to be 21.8 (range = 5–38). The average number of plant taxa in other gazelle species such as the Tibetan gazelle (Harris and
per sample is probably influenced by factors such as variations Miller 1995; Schaller 1998; Li et al. 2008) and goitered gazelle
in plant communities across study sites and seasons, ecologi- (Schaller 1998; Xu et al. 2012). However, 2 studies reported
cal characteristics of the ungulates studied, and filter settings that Mongolian gazelles consumed high amounts of graminoids
for cleaning sequence reads. Another potential factor could be during summer in Inner Mongolia, China (Jiang et al. 2002),
the DNA extraction protocol. Deagle et al. (2005) showed that and southern Mongolia (Yoshihara et al. 2008). The differences
DNA subsampled from different parts of the same fecal sample in graminoid consumption among these studies may be due to
resulted in the detection of different food compositions. In this different analytical methods, different vegetation compositions
study, we collected fecal material from 1–3 pellets for DNA in the study habitats, or heterogeneous distributions of plants

Downloaded from https://academic.oup.com/jmammal/article-abstract/99/2/450/4818386

by Tottori University Tottori Campus user
on 04 April 2018
 456 JOURNAL OF MAMMALOGY
Fig. 3.—Rarefaction curves of the number of plant families (a) and
molecular operational taxonomic units (MOTUs) (b) detected in fecal
samples of 6 ungulates in the desert steppe of southern Mongolia col-
lected 19–26 June 2014.
even in similar habitats. Our study revealed summer food hab-
its; during winter, the consumption of graminoids may increase
because highly nutritious plants become limited (Jiang et al.
2002; Xu et al. 2012).
Wild asses consumed mainly Poaceae, especially Stipa at
87%, a typical characteristic of grazer species (Hofmann and
Stewart 1972). Tibetan wild asses (or kiang, Equus kiang) in the
Tibetan plateau also have been found to consume high amounts
of graminoids during summer (Harris and Miller 1995; Schaller
1998; but see Shi et al. 2016). The difference in food habits
between the Mongolian and Tibetan wild asses is most evident
in their consumption of shrubs. Wild asses in southern Mongolia
consumed moderate amounts of shrubs (e.g., R. soongarica),
but kiangs in the Tibetan plateau consumed shrubs only in
small proportions. This difference is likely due to the greater
availability of shrubs in the desert steppe of southern Mongolia
than in the Tibetan plateau (Schaller 1998). Kiangs and also
horses, another grazer species in the Tibetan plateau, have been
shown to maintain graminoid-dominant diets, with a high die-
tary similarity (Schaller 1998). This relationship persisted in
the desert steppe, where wild asses and horses consumed large
amounts of Poaceae with moderate amounts of shrubs, leading
to a high overlap value.
Sheep and goats showed a high dietary diversity con-
sisting of various plants, mainly Poaceae and Fabaceae.
Downloaded from https://academic.oup.com/jmammal/article-abstract/99/2/450/4818386
by Tottori University Tottori Campus user
on 04 April 2018
Their food compositions are probably influenced by herding
practices (Harris and Miller 1995) because their movements
are controlled by herders. Wingard et al. (2011) suggested
that Mongolian nomadic herding practices have the potential
to limit diet variability of livestock based on the finding that
the number of key plant species in the diet of sheep and goats
was lower than that of mountain sheep (argali, Ovis ammon).
However, our study suggested that herding practices have the
potential not only to limit but also expand dietary diversity. The
dietary diversity is probably reflected by the vegetation diver-
sity and conditions where herders guide their livestock.
Many studies have merged data from sheep and goats because
both species are herded together, making it difficult to distin-
guish their feces (e.g., Campos-Arceiz et al. 2004; Yoshihara
et al. 2008; Wingard et al. 2011). Herein, we investigated their
summer food habits separately. Although there were slight dif-
ferences in their food compositions, such as the relative propor-
tions of Poaceae and Rosaceae, treating the dietary data from
sheep and goats in the same herd as a single category seems
to be reasonable during summer, given the fact that their food
compositions were not significantly different and the overlap
value was the highest we found for any pair.
A stark contrast in food habits between camels and the other
livestock was seen in the consumption of Stipa. Camels con-
sumed low proportions of graminoids with no Stipa. Instead,
they consumed high amounts of R. soongarica (shrubs) and
Fabaceae (which include both shrubs and forbs), showing a
browser characteristic. This result was comparable to a previ-
ous review of camel feeding behavior (Iqbal and Khan 2001),
which indicated that camels predominantly are browsers. Their
shrub-dominant diets probably lead to relatively low overlap
values between camels and the other ungulates.
In general, we found a wide range of overlap values (0.315–
0.959) in the pairwise comparisons. Similarities in the digestive
system or body size are likely to be related to similarities in diet
(Bell 1970; Hofmann 1973; Jarman 1974). Gazelles, sheep,
goats, and camels are ruminants, whereas wild asses and horses
are nonruminants, and this fact, along with similarities in body
size, may explain the high overlap values between gazelles and
sheep–goats and between wild asses and horses. In contrast,
the low overlap value between wild ungulates (i.e., gazelles and
wild asses) is likely due to the differences in their digestive
systems (i.e., ruminant and nonruminant) and body size, which
could be important in enabling these ungulates to live sympatri-
cally in the desert steppe. In addition to body size and digestive
system, high dietary diversity of sheep and goats may contrib-
ute to their generally high overlap values with all ungulates.
Given the high dietary diversity of sheep and goats and their
overlap with wild species, their increasing number is a poten-
tial threat to the survival of wild ungulates through increased
resource competition. Improving herding practices by creating
nongrazing areas or controlling livestock numbers could be
effective measures for protecting wild ungulates in the desert
steppe. It should be noted that estimating the dietary overlap
value alone does not quantify the actual extent of competition
(Harris and Miller 1995). However, revealing the detailed food
habits of animals during a season when resources are high is
 SUGIMOTO ET AL.—DIET OF MONGOLIAN WILD AND DOMESTIC UNGULATES
essential for understanding fundamental ecological character-
istics. To extend our findings, future work should address sea-
sonal, annual, and regional variations in dietary overlap, as well
as dietary analyses in relation to resource availability. A DNA
barcoding approach via next generation sequencing can help to
address these research questions.
Acknowledgments
We express gratitude to the members of Molecular biology
laboratory of Institute of General and Experimental Biology
in Mongolian Academy of Sciences for their useful advice
and assistance. This research was supported by Grants-
in-Aid for Scientific Research 24510326, 15K06931 from
MEXT of Japan.
Literature Cited
Baamrane, M. A. A., et al. 2012. Assessment of the food habits of the
Moroccan dorcas gazelle in M’Sabih Talaa, west central Morocco,
using the trnL approach. PLoS One 7:e35643.
Batsaikhan, N., et al. 2014. Conserving the world’s finest grassland
amidst ambitious national development. Conservation Biology
28:1736–1739.
Bell, R. H. V. 1970. The use of herb layer by grazing ungulates in
the Serengeti. Pp. 111–124 in Animal populations and relation to
their food resources (A. Watson, ed.). Blackwell, Oxford, United
Kingdom.
Benjamini, Y., and Y. Hochberg. 1995. Controlling the false discov-
ery rate - a practical and powerful approach to multiple testing.
Journal of the Royal Statistical Society Series B-Methodological
57:289–300.
Boyer, F., C. Mercier, A. Bonin, Y. Le Bras, P. Taberlet, and
E.  Coissac. 2016. OBITOOLS: a UNIX-inspired software pack-
age for DNA metabarcoding. Molecular Ecology Resources
16:176–182.
Buuveibaatar, B., et al. 2017. Mongolian Gobi supports the world’s
largest populations of khulan Equus hemionus and goitered gazelles
Gazella subgutturosa. Oryx 51:639–647.
Campos-Arceiz, A., S. Takatsuki, and B. Lhagvasuren. 2004. Food
overlap between Mongolian gazelles and livestock in Omnogobi,
southern Mongolia. Ecological Research 19:455–460.
Colwell, R. K. 2013. EstimateS: statistical estimation of species rich-
ness and shared species from samples. Version 9. User’s Guide and
application. http://purl.oclc.org/estimates. Accessed 31 July 2017.
Colwell, R. K., and D. J. Futuyma. 1971. On the measurement of
niche breadth and overlap. Ecology 52:567–576.
Deagle, B. E., D. J. Tollit, S. N. Jarman, M. A. Hindell, A. W.
Trites, and N. J. Gales. 2005. Molecular scatology as a tool to
study diet: analysis of prey DNA in scats from captive steller sea
lions. Molecular Ecology 14:1831–1842.
Dorjgotov, D., N. Orshih, and T. Oyunchimeg. 2004. Geographic
atlas of Mongolia. Administration of Land Affairs, Geodesy and
Cartography, Ulaanbaatar, Mongolia.
Grubov, V. I. 2001. Key to the vascular plants of Mongolia. Science
Publishers, Inc., Enfield, New Hampshire.
Harris, R. B., and D. J. Miller. 1995. Overlap in summer habitats
and diets of Tibetan plateau ungulates. Mammalia 59:197–212.
Hibert, F., P. Taberlet, J. Chave, C. Scotti-Saintagne, D. Sabatier,
and C. Richard-Hansen. 2013. Unveiling the diet of elusive
Downloaded from https://academic.oup.com/jmammal/article-abstract/99/2/450/4818386
by Tottori University Tottori Campus user
on 04 April 2018
457
rainforest herbivores in next generation sequencing era? The tapir
as a case study. PLoS One 8:e60799.
Hilker, T., E. Natsagdorj, R. H. Waring, A. Lyapustin, and Y. Wang.
2014. Satellite observed widespread decline in Mongolian grasslands
largely due to overgrazing. Global Change Biology 20:418–428.
Hofmann, R. R. 1973. The ruminant stomach: stomach structure
and feeding habitat of East African game ruminants. East African
Literature Bureau, Nairobi, Kenya.
Hofmann, R. R., and D. R. M. Stewart. 1972. Grazer or browser: a
classification based on the stomach structure and feeding habits of
east African ruminants. Mammalia 36:222–240.
Imai, S., T. Y. Ito, T. Kinugasa, M. Shinoda, A. Tsunekawa, and
B.  Lhagvasuren. 2017. Effects of spatiotemporal heterogeneity
of forage availability on annual range size of Mongolian gazelles.
Journal of Zoology 301:133–140.
Iqbal, A., and B. B. Khan. 2001. Feeding behaviour of camel - review.
Pakistan Journal of Agricultural Sciences 38:58–63.
Ito, T. Y., et  al. 2005. Preliminary evidence of a barrier effect of
a railroad on the migration of Mongolian gazelles. Conservation
Biology 19:945–948.
Ito, T. Y., A. Okada, B. Buuveibaatar, B. Lhagvasuren, S. Takatsuki,
and A. Tsunekawa. 2008. One-sided barrier impact of an inter-
national railroad on Mongolian gazelles. Journal of Wildlife
Management 72:940–943.
Ito, T. Y., et al. 2013. Fragmentation of the habitat of wild ungulates
by anthropogenic barriers in Mongolia. PLoS One 8:e56995.
IUCN. 2017. IUCN Red List of Threatened Species. Version 2017.
www.iucnredlist.org. Accessed 31 July 2017.
IUCN SSC Antelope Specialist Group. 2016. Procapra gutturosa.
In: IUCN 2016. The IUCN Red List of Threatened Species. Version
2017. www.iucnredlist.org. Accessed 31 July 2017.
Jarman, P. J. 1974. The social organization of antelope in relation to
their ecology. Behaviour 48:215–266.
Jiang, Z.-W., S. Takatsuki, J. S. Li, W. Wang, Z. X. Gao, and J. Z. Ma.
2002. Seasonal variations in foods and digestion of Mongolian
gazelles in China. Journal of Wildlife Management 66:40–47.
Kaczensky, P., R. Kuehn, B. Lhagvasuren, S. Pietsch, W. Yang, and
C. Walzer. 2011. Connectivity of the Asiatic wild ass population
in the Mongolian Gobi. Biological Conservation 144:920–929.
Levins, R. 1968. Evolution in changing environments. Princeton
University Press, Princeton, New Jersey.
Li, Z. Q., Z. G. Jiang, and C. W. Li. 2008. Dietary overlap of Przewalski’s
gazelle, Tibetan gazelle, and Tibetan sheep on the Qinghai-Tibet
plateau. Journal of Wildlife Management 72:944–948.
National Statistical Office of Mongolia. 2010. Mongolian
Statistical Yearbook 2009. National Statistical Office of Mongolia,
Ulaanbaatar, Mongolia.
Pegard, A., et al. 2009. Universal DNA-based methods for assessing
the diet of grazing livestock and wildlife from feces. Journal of
Agricultural and Food Chemistry 57:5700–5706.
Pianka, E. R. 1973. The structure of lizard communities. Annual
Review of Ecology and Systematics 4:53–74.
Raye, G., C. Miquel, E. Coissac, C. Redjadj, A. Loison, and
P. Taberlet. 2011. New insights on diet variability revealed by
DNA barcoding and high-throughput pyrosequencing: chamois
diet in autumn as a case study. Ecological Research 26:265–276.
Sasaki, T., T. Okayasu, U. Jamsran, and K. Takeuchi. 2008. Threshold
changes in vegetation along a grazing gradient in Mongolian range-
lands. Journal of Ecology 96:145–154.
Schaller, G. B. 1998. Wildlife of the Tibetan steppe. University of
Chicago Press, Chicago, Illinois.
 458 JOURNAL OF MAMMALOGY
Schmieder, R., and R. Edwards. 2011. Quality control and prepro-
cessing of metagenomic datasets. Bioinformatics 27:863–864.
Schmieder, R., Y. W. Lim, F. Rohwer, and R. Edwards. 2010.
TagCleaner: identification and removal of tag sequences from
genomic and metagenomic datasets. BMC Bioinformatics 11:341.
Shi, J. B., et  al. 2016. Dietary overlap and co-existence of sym-
patric wild yak, Tibetan wild ass and Tibetan antelope in Arjin
Shan national nature reserve, Xinjiang province, China. Wildlife
Research 43:323–331.
Soininen, E. M., et  al. 2009. Analysing diet of small herbivores: the
efficiency of DNA barcoding coupled with high-throughput pyro-
sequencing for deciphering the composition of complex plant mix-
tures. Frontiers in Zoology 6:16.
Taberlet, P., et  al. 2007. Power and limitations of the chloroplast
trnL (UAA) intron for plant DNA barcoding. Nucleic Acids
Research 35:e14.
Undarmaa, J., T. Okuro, N. Manibazar, and N. Yamanaka. 2015.
Rangeland plants of Mongolia, Vol. I & II. Munkhiin Useg,
Ulaanbaatar, Mongolia.
Downloaded from https://academic.oup.com/jmammal/article-abstract/99/2/450/4818386
by Tottori University Tottori Campus user
on 04 April 2018
View publication stats
Valentini, A., F. Pompanon, and P. Taberlet. 2009a. DNA barcoding
for ecologists. Trends in Ecology & Evolution 24:110–117.
Valentini, A., et  al. 2009b. New perspectives in diet analysis based
on DNA barcoding and parallel pyrosequencing: the trnL approach.
Molecular Ecology Resources 9:51–60.
Wingard, G. J., et  al. 2011. Argali food habits and dietary overlap
with domestic livestock in Ikh Nart Nature Reserve, Mongolia.
Journal of Arid Environments 75:138–145.
Xu, W. X., et  al. 2012. Diet of Gazella subgutturosa (Guldenstaedt,
1780) and food overlap with domestic sheep in Xinjiang, China.
Folia Zoologica 61:54–60.
Yoshihara, Y., T. Y. Ito, B. Lhagvasuren, and S. Takatsuki. 2008.
A comparison of food resources used by Mongolian gazelles and
sympatric livestock in three areas in Mongolia. Journal of Arid
Environments 72:48–55.
Submitted 1 August 2017. Accepted 11 December 2017.
Associate Editor was Leslie Carraway.