ArticlePDF Available

The Danger of Having All Your Eggs in One Basket—Winter Crash of the Re-Introduced Przewalski's Horses in the Mongolian Gobi

  • International Takhi Group

Abstract and Figures

Large mammals re-introduced into harsh and unpredictable environments are vulnerable to stochastic effects, particularly in times of global climate change. The Mongolian Gobi is home to several rare large ungulates such as re-introduced Przewalski's horses (Equus ferus przewalskii) and Asiatic wild asses (Equus hemionus), but also to a millennium-old semi-nomadic livestock herding culture. The Gobi is prone to large inter-annual environmental fluctuations, but the winter 2009/2010 was particularly severe. Millions of livestock died and the Przewalski's horse population in the Gobi crashed. We used spatially explicit livestock loss statistics, ranger survey data and GPS telemetry to provide insight into the effect of a catastrophic climate event on the two sympatric wild equid species and the livestock population in light of their different space use strategies. Herders in and around the Great Gobi B Strictly Protected Area lost on average 67% of their livestock. Snow depth varied locally, resulting in livestock losses following an east-west gradient. Herders had few possibilities for evasion, as competition for available winter camps was high. Przewalski's horses used three different winter ranges, two in the east and one in the west. Losses averaged 60%, but differed hugely between east and west. Space use of Przewalski's horses was extremely conservative, as groups did not attempt to venture beyond their known home ranges. Asiatic wild asses seemed to have suffered few losses by shifting their range westwards. The catastrophic winter 2009/2010 provided a textbook example for how vulnerable small and spatially confined populations are in an environment prone to environmental fluctuations and catastrophes. This highlights the need for disaster planning by local herders, multiple re-introduction sites with spatially dispersed populations for re-introduced Przewalski's horses, and a landscape-level approach beyond protected area boundaries to allow for migratory or nomadic movements in Asiatic wild asses.
Content may be subject to copyright.
The Danger of Having All Your Eggs in One Basket—
Winter Crash of the Re-Introduced Przewalski’s Horses in
the Mongolian Gobi
Petra Kaczensky
*, Oyunsaikhan Ganbataar
, Nanjid Altansukh
, Namtar Enkhsaikhan
, Christian
, Chris Walzer
1Research Institute of Wildlife Ecology, University of Veterinary Medicine, Vienna, Austria, 2Great Gobi B Strictly Protected Area Administration, Takhin Tal, Gobi-Altai
Province, Mongolia, 3Department of Zoology, National University of Mongolia, Faculty of Biology, Ulaanbaatar, Mongolia, 4International Takhi Group - Mongolia, Baigal
Ordon, Ulaanbaatar, Mongolia, 5International Takhi Group - Switzerland, Sihlwald, Switzerland
Large mammals re-introduced into harsh and unpredictable environments are vulnerable to stochastic effects, particularly in
times of global climate change. The Mongolian Gobi is home to several rare large ungulates such as re-introduced
Przewalski’s horses (Equus ferus przewalskii) and Asiatic wild asses (Equus hemionus), but also to a millennium-old semi-
nomadic livestock herding culture. The Gobi is prone to large inter-annual environmental fluctuations, but the winter 2009/
2010 was particularly severe. Millions of livestock died and the Przewalski’s horse population in the Gobi crashed. We used
spatially explicit livestock loss statistics, ranger survey data and GPS telemetry to provide insight into the effect of a
catastrophic climate event on the two sympatric wild equid species and the livestock population in light of their different
space use strategies. Herders in and around the Great Gobi B Strictly Protected Area lost on average 67% of their livestock.
Snow depth varied locally, resulting in livestock losses following an east-west gradient. Herders had few possibilities for
evasion, as competition for available winter camps was high. Przewalski’s horses used three different winter ranges, two in
the east and one in the west. Losses averaged 60%, but differed hugely between east and west. Space use of Przewalski’s
horses was extremely conservative, as groups did not attempt to venture beyond their known home ranges. Asiatic wild
asses seemed to have suffered few losses by shifting their range westwards. The catastrophic winter 2009/2010 provided a
textbook example for how vulnerable small and spatially confined populations are in an environment prone to
environmental fluctuations and catastrophes. This highlights the need for disaster planning by local herders, multiple re-
introduction sites with spatially dispersed populations for re-introduced Przewalski’s horses, and a landscape-level approach
beyond protected area boundaries to allow for migratory or nomadic movements in Asiatic wild asses.
Citation: Kaczensky P, Ganbataar O, Altansukh N, Enkhsaikhan N, Stauffer C, et al. (2011) The Danger of Having All Your Eggs in One Basket—Winter Crash of the
Re-Introduced Przewalski’s Horses in the Mongolian Gobi. PLoS ONE 6(12): e28057. doi:10.1371/journal.pone.0028057
Editor: Georges Chapouthier, Universite
´Pierre et Marie Curie, France
Received October 6, 2011; Accepted October 31, 2011; Published December 28, 2011
Copyright: ß2011 Kaczensky et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: No specific funding was obtained for the explicit analyses contained in this manuscript. Rather it is a by-product built on data being collected by a
range of ongoing projects funded by several sources. Main funding was provided by the Austrian Science Foundation FWF project P14992 and P18624 (http:// Additional funding and resources came from the Great Gobi B SPA administration, the International Takhi Group, the University of
Veterinary Sciences in Vienna, and the Mohammed bin Zayed Species Conservation Fund, project no.11251783. The funders had no role in study design, data
collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail:
Small populations have a high extinction risk due to
demographic stochasticity, the loss of genetic variability, and the
potential detrimental effect of recessive genes [1]. A small
population with a restricted range resembles the proverbial ‘‘all
eggs in a basket’’, as it is particularly susceptible to environmental
stochasticity [2]. However, many re-introduced populations of
large mammals start small [3] due to logistical and financial
constraints or the controversial nature of the species concerned.
Large mammals re-introduced into harsh and unpredictable
environments are vulnerable to stochastic effects [4,5], particularly
in times of global climate change and the associated increase in
extreme weather events [6,7].
Arid rangelands with a high level of interannual variation in
precipitation are believed to follow non-equilibrium dynamics with
precipitation being the main factor controlling both ungulate and
vegetation dynamics [8]. The Mongolian Gobi in Central Asia
constitutes a vast, largely intact and continuous stretch of non-
equilibrium dry land [9] which is home to several endangered or
critically endangered large migratory ungulates [10–12] as well as
a millennium-old semi-nomadic livestock herding culture [13,14].
Extreme weather conditions in the form of droughts followed by
cold and snow rich winters (called ‘‘dzud’’ in Mongolia) occur at
irregular intervals and have resulted in mass die-offs of livestock
[15,16]. Although periodic mass winter mortality in wild ungulates
on open range has been documented elsewhere [17–19,11], are
frequently mentioned by local people in Mongolia and acknowl-
edged in the scientific literature (for Asiatic wild asses Equus
hemionus in [18], Mongolian gazelles Procapra gutturosa in [20],
goitered gazelle Gazella subgutturosa in [21]), the actual impact of
these climatic extremes on wild ungulate population dynamics is
PLoS ONE | 1 December 2011 | Volume 6 | Issue 12 | e28057
poorly documented, not least because of the difficulties of
obtaining reliable population estimates of wild ungulates over
the vast expanses of the Mongolian rangeland [22].
Przewalski’s horses have been re-introduced to Mongolia since
1992 and free-ranging populations now exist in Hustai National
Park (NP) in central Mongolia [23] and in the Great Gobi B
Strictly Protected Area (SPA) in the Dzungarian Gobi in south-
western Mongolia [24]. The initial phase of the re-introduction
programme in the Dzungarian Gobi was plagued with various
problems [25] and population growth could only be achieved by
introducing additional captive animals. Management changes
were implemented in 1999/2000, but in 2000/2001 the area was
hit by a dzud winter. The population suffered a net loss of 21%
and almost no foals were produced in the spring of 2001 [26].
Since 2002/03 the Przewalski’s horse population finally started to
show positive population growth, independent of released animals.
The positive population development in the two Mongolian re-
introduction sites has resulted in the down-listing of the
Przewalski’s horse from ‘‘extinct in the wild’’ to ‘‘endangered’’
on the IUCN Red List of Threatened Species [27]. In December
2009 the Hustai NP population had reached 259 animals (D.
Usukhjargal unpubl. data) and the Great Gobi B SPA population
138 animals or almost the ‘‘.140 horses necessary to achieve a
95% probability of survival over 100 years under the low severity
level of catastrophes scenario’’ [26].
However, in the winter of 2009/2010 one of the worst dzud
conditions ever hit Mongolia. Millions of livestock died, driving
their owners into economic disaster and causing a humanitarian
crisis [28,29]. Concurrently, the Przewalski’s horse population in
the Great Gobi B SPA crashed, providing a textbook example of
the risks faced by small and spatially confined species in
unpredictable environments. By coincidence we also followed
the whereabouts of 10 GPS-collared sympatric Asiatic wild asses
from July 2009 until July 2010. In the following we provide insight
into the effect of a catastrophic climate event on two sympatric
wild equid species and the livestock population of the local semi-
nomadic pastoralists in light of their different space use patterns.
Materials and Methods
Ethics statement
All data sets were collected within the frames of the legal
requirements of Austria and Mongolia. Capture and collaring of
Asiatic wild asses was conduced within a cooperation agreement
between the International Takhi Group and the Mongolian
Ministry of Nature, Environment and Tourism signed on
15.02.2001 and renewed on 27.01.2011.
Study area
The Dzungarian Gobi in SW Mongolia forms a rather distinct
entity of the Gobi ecosystem due to its geographic location in a
basin surrounded by high mountains and by being located at the
edge of the influence of the Atlantic/Mediterranean and the Asian
Monsoon weather systems [30]. Almost the entire eastern and
central part of the Dzungarian Gobi falls within the 9,000 km
Great Gobi B SPA.
Plains dominate the landscape of the Great Gobi B SPA in the
east and rolling hills in the west. The Altai Mountains flank the
park to the north, and the Takhin Shar Naruu Mountains form
the international border with China (Figure 1A). Elevations range
from 1,000 to 2,840 m. The climate of the Great Gobi B SPA is
continental with long cold winters and short, hot summers.
Average annual rainfall is 96 mm with a peak during summer.
Average snow cover lasts 97 days. Both rain and snowfall can be
highly variable from year to year in space and time.
Desert areas are widely dominated by Chenopodiaceae, such as
Saxaul Haloxylon ammodendron and Anabasis brevifolia. The steppe
areas are dominated by Asteraceae, such as Artemisia and Ajania,
and Poaceae like Stipa and Ptilagrostis [31]. In locations where
several springs occur, these are surrounded by intermittent
swamps and form permanent oases.
The wild ungulate community of the steppe areas consists of
goitered gazelle, Asiatic wild ass, and re-introduced Przewalski’s
horse. Starting in 1992, a total of 89 Przewalski’s horses on 10
transports had been airlifted from abroad to the Takhin Tal
adaptation facilities at the NE edge of the Great Gobi B SPA.
Przewalski’s horses live in stable harems groups, have non-
exclusive home ranges of 152–826 km
, select for the most
productive plant communities and are slow to expand their range
[24]. To speed up range expansion, the last group of re-introduced
Przewalski’s horses, was released at the Takhi us oasis complex,
about 120 km west of the established Przewalski’s horse popula-
tion in 2005 (Figure 1B). In 2007 three wild born stallions were
flown in from Hustai NP to test the feasibility of intra-Mongolian
population exchanges. Wild asses seem to live in fission–fusion
groups, have non-exclusive home ranges of 4,449–6,835 km
show little preference for any particular plant community [24].
The wild ass population of the Dzungarian Gobi seems to
constitute a distinct subpopulation [12], numbering several
thousand individuals (P. Kaczensky and R. Ransom unpubl. data).
The Great Gobi B SPA is also used by ,100 semi-nomadic
herder families with ,60,000 livestock. Herders show north-south
seasonal movements between winter pastures along the fringes of
the Dzungarian basin and alpine summer pastures in the Altai
Mountains [32]. Local economy is heavily based on livestock, with
cashmere generating the main income. Since the collapse of the
socialist system local herders have limited access to veterinary care
and largely operate without winter fodder reserves [14].
Winter severity
We have been recording temperature on an hourly basis using a
data logger (HOBO temperature logger, Hoskin Scientific
Limited, Vancouver, Canada) at Takhin Tal research station
since April 2003. Furthermore, O. Ganbaatar records unusual
weather events in his personal research journal. No further
weather stations are present anywhere in the vicinity of the Great
Gobi B SPA.
Livestock losses
To indirectly assess spatial variation in the severity of the 2009/
2010 dzud conditions we obtained spatially explicit livestock loss
data from the majority of local herders in and adjacent to the
Great Gobi B SPA. We obtained information on livestock
numbers and losses by personally interviewing local families
(N = 115) and from livestock statistics collected by the local
(administrative units called ‘‘bag’’) governors (N = 387). We
obtained winter camp coordinates either from 1:100.000 topo-
graphic maps during our interviews or via GPS mapping in the
field. Livestock numbers and losses are largely based on self-
reported numbers by the local herding families.
Przewalski’s horse monitoring
We recorded births and mortalities of individual Przewalski’s
horses based on a horse year lasting from 1 May until 30 April the
following year. Foals born before 1 May were manually assigned to
the correct horse year. Population numbers for each horse year were
calculated as the number of animals alive on 30 April, showing the
Winter Crash of Re-Introduced Przewalski’s Horses
PLoS ONE | 2 December 2011 | Volume 6 | Issue 12 | e28057
Figure 1. Impact of the 2009/2010 dzud winter on local nomads and two wild equid species in the Great Gobi B Strictly Protected
Area in south-western Mongolia. A) Livestock loss prediction map as a proxy for winter severity based on the average % total livestock loss at 219
herder camp locations using ordinary kriging, B) Winter losses among the re-introduced Przewalski’s horse population as a function of their
respective winter ranges. The total distribution range in 2009 is based on group locations of 12 harem and 1–3 bachelor groups of Przewalski’s horses
on 129 observation days from January through December 2009. C) Movement patterns of Asiatic wild asses based on GPS positions of 10 wild asses
followed from July 2009 to July 2010 (N = 355,618). Blue dots mark locations during the dzud period (N = 99,220) and red dots locations during the
rest of the monitoring period.
Winter Crash of Re-Introduced Przewalski’s Horses
PLoS ONE | 3 December 2011 | Volume 6 | Issue 12 | e28057
net gain since the previous horse year. In addition, all births and
mortalities, and in the past transports, were recorded by horse year.
Przewalski’s horse groups were checked by park rangers 1–2
times a week. Rangers were able to individually identify each
Przewalski’s horse based on overall appearance (size, shape, coat
colour, special marks) until the population numbered around 100
animals in 2007. Thereafter, they were still able to identify adult
stallions and all mares, but became increasingly unsure about
young stallions.
Rangers determined the location of individual Przewalski’s
horses and groups based on a raster map, noted group size and
composition, and any peculiarities of individual horses (e.g.
injuries, poor body condition, etc.). An additional 15 Przewalski’s
horses had been followed by satellite telemetry between 2001 and
2008 (for details see [24,33]) and comparison between telemetry
data and ranger monitoring showed that the latter is sufficient to
document the broad patterns of spatial organization of the
different groups and distribution range development (P. Kac-
zensky unpubl. data).
When rangers failed to localize individual Przewalski’s horses
they attempted to locate the horse’s carcass by searching the area
the animal was last seen in. However, during the dzud winter
search efforts were hindered by deep snow and a subsequent snow
melt that transformed large parts of the Gobi into mud flats.
Furthermore, with the ground thawing, wildlife and domestic
animal carcasses at Chonin us quickly started to sink into the many
mud holes. When access finally became possible in April 2010,
rangers walked the area for two full days in search of Przewalski’s
horse carcasses. Rangers collected various tissue samples from all
carcasses encountered for histo-pathological examinations [34].
Although management of the re-introduced Przewalski’s horse
population has adopted a low intervention policy, emergency hay
was purchased in early March 2010, and the remaining
Przewalski’s horses were supplemented with hay from 7 March–
10 April at Chonin us and from 7–31 March at Takhin Tal and
Takhi us [35]. In addition, several Przewalski’s horses gained
access to old hay reserves at the Takhin Tal camp from November
2009 on and received supplementary fresh hay as early as the
begin of January. However, feeding was frequently disrupted by
heavy snow storms that confined rangers to their homes [35].
Wild ass monitoring
Between 2002 and 2003 we collared 7 Asiatic wild asses in the
Great Gobi B SPA with Argos and GPS-Argos collars [24]. To
gain more detailed insight into small scale habitat use we captured
an additional 24 asses in July 2007 and July 2009 and equipped
them with GPS store-on-board (SOB) collars that attempted a
GPS fix every 15 minutes over a 12 month period. We retrieved
21 of the 24 collars but due to technical problems only 1 had
monitored ass movements in 2007/2008 [33] while 10 had
monitored wild ass movements during the dzud year 2009/2010
(collecting a total of 355,618 GPS locations, Table S1). All collars
were deployed with drop-off devices (CR-2a, Telonics, Mesa,
USA) programmed to release 12 months after deployment.
We attempted collar retrieval by systematically climbing high
points throughout the Great Gobi B SPA and subsequently
homing in on the signal from the VHF beacon of the shed units.
During our search for shed GPS collars in July 2010 we also
recorded all carcasses of winter-killed wild asses. Furthermore,
rangers made general notes about wild ass carcasses, while
searching for Przewalski’s horses. Carcasses were identified as
potential winter-kills, when they were fresh enough to show
considerable amounts of skin with winter fur and formerly freeze-
dried tissue remains.
Data analysis
For visualization and analysis of spatial data, we used ArcMap
9.3 (ESRI, Environmental Systems Research Institute, Inc.,
Redlands, California, USA). We digitized rivers, springs, oases,
villages and elevations from Russian 1:100 000 topographic maps.
We created a livestock loss prediction map by using ordinary
kriging in the Geostatistical Analyst function. We averaged the
total livestock loss for any given winter camp if more than one
family used the same location. We used the % average livestock
loss of 2010 as the attribute variable to obtain the livestock loss
prediction map, which we subsequently used as a spatially explicit
proxy of winter severity. The proxy map was qualitatively
validated by comparing it to MODIS/Terra satellite snow cover
images [36] from the onset of the lasting snow cover in November
2009 and from the melt-off phase in March and April 2010 (Figure
For selection of the key variables explaining death or survival of
an individual Przewalski’s horse between December 2009 and
April 2010 we used a generalized additive model (GAM) and a
generalized linear model (GLM) in R [37] with subsequent least
square model averaging based on Akaike weights of all candidate
models (R library glmulti [38] and MuMIn). The relative
importance of each variable is expressed as the sum of the AIC
weights from all models that included this variable.
Winter severity
Dzud conditions in the Great Gobi B SPA started on 22
December 2009 and lasted until the end of March. In December
2009 and January 2010 three major snowstorms (22 December, 29
December–7 January and 17–20 January) deposited large amounts
of snow, each packing down the previous layer and in places
reaching accumulated snow depths of 1 m and more. In February
another 5 periods of heavy snow storms occurred, each lasting for
2–3 days. From 6–8 March the last severe snow storm hit, but
temperatures stayed low until mid March (Table S2, Figure S2).
The period from December 2009 until March 2010 was 2.7–5.7uC
colder than during the previous 7 years (Table S2).
Herders in and around the Great Gobi B SPA lost on average
67% of their entire livestock, with only camels less affected
(Table 1). Most affected was the north-eastern part of the Great
Gobi B SPA, where herders lost 80–100% of their livestock. Least
affected were the hills in the central part of the SPA, where
livestock losses were in the magnitude of 20–40%, and the areas in
the west, where losses were in the magnitude of 40–60%
(Figure 1A). The spatial pattern suggests that the weather largely
came from the west and that the snow clouds were stopped by the
high mountains forming the south-eastern tip of the Altai
Mountains thereby depositing the bulk of their snow load at the
north-eastern edge of the Dzungarian basin. This pattern is also
supported by the timing and spatial distribution of the lasting snow
cover in early November 2009 and the thawing pattern at the end
of March/beginning of April 2010 (Figure S1). The majority of
livestock losses occurred during the snow storms in February.
Przewalski’s horses
Annual population growth for the horse years 2002/03 until
2008/09 was positive and averaged 12% (range 1–20%; Table 2).
During the horse year 2009/10 the population suffered a net loss of
60%. The main crash happened during the dzud period, with the
population dropping from 138 Przewalski’s horses in December
2009 down to 49 by the end of April 2010. Furthermore only one
foal was born in 2010 (Table 2).
Winter Crash of Re-Introduced Przewalski’s Horses
PLoS ONE | 4 December 2011 | Volume 6 | Issue 12 | e28057
The main die-off started in January (N = 16), peaked in
February (N = 64) and tailed off in March (N = 6) and April
(N = 3). Winter losses averaged 64% of the early December 2009
population and were most severe in the eastern winter ranges
around Takhin Tal (82%) and Chonin us (73%). However, only
one out of 19 animals (5%) died in the western winter range
around Takhin us (Figure 1B). Most animals seem to have been
lost during snow storms.
Due to the limited vehicle access of large parts of the Gobi from
mid December until the middle of April, rangers could only locate
the carcasses of 33 individuals. However, no living Przewalski’s
horses were reported from anywhere in or around the Great Gobi
B SPA throughout 2010 and 2011, and thus it is safe to assume
that the remaining animals also died.
Area was obviously the key factor for mortality or survival of
Przewalski’s horses during the dzud 2009/2010. In addition, for
the 119 Przewalski’s horses in the eastern winter range, age was
the most likely predictor of mortality, with the youngest age classes
being most affected (Table 3, Figure S3). The influence of sex and
origin of the horses was less certain, but if it played a role, the
effect of a zoo origin (horses born in the Gobi had a higher chance
of survival than those born in a zoo) was twice as strong as the
effect of sex (stallions had a lower survival probability than mares).
Whether or not a mare had a foal in 2009 did not seem to have
any predictive value (Table S3).
Asiatic wild asses
We were able to retrieve 10 out of the 14 collars deployed in
2009. These collars had dropped off the live animals in July 2010,
as they were not associated with a wild ass carcass and the data
showed movements until the drop-off day. We found one collar
with a non-functioning VHF unit by pure chance near a spring,
suggesting that some of the 4 missing units may have had similar
technical troubles. Even with the fate of 4 animals remaining
unknown, a minimum of 71% of our collared asses survived the
dzud winter. Furthermore, rangers did not find any ass carcasses
when searching for deceased Przewalski’s horses at Chonin us and
registered only 1 or 2 wild ass carcasses in the Takhi us winter
range. During 10 days of intensive ground search for dropped
Table 1. Livestock losses in and around Great Gobi B Strictly
Protected Area during the dzud winter 2009/2010 based on
self-reported losses of 502 families.
Livestock population
end Dec. 2009 N lost % lost
Goats 80,797 54,435 0.67
Sheep 59,033 40,068 0.68
Horses 5,211 3,081 0.59
Cows/Yaks 3,377 2,066 0.61
Camels 1,049 258 0.25
Total 149,467 99,908 0.67
Table 2. Population development of the re-introduced Przewalski’s horse population in the Great Gobi B Strictly Protected Area
Horse year Number of Przewalski’s horses Annual l
alive by end of April Born dead Winter
loss trans- ported
1992/93 6 1 0 0 5
1993/94 10 1 5 0 8 21.67
1994/95 9 2 3 0 0 21.10
1995/96 19 2 5 1 13 21.33
1996/97 26 4 5 0 8 21.05
1997/98 26 6 12 4 6 21.23
1998/99 39 5 7 3 15 21.08
1999/00 43 6 6 0 4 1.00
38 15 24 22 4
2001/02 35 140021.08
2002/03 54 13 8 2 14 1.14
2003/04 59 13 8 0 0 1.09
2004/05 86 24 9 5 12 1.25
2005/06 95 22 14 0 0 1.10
2006/07 96 33 32 3 0 1.01
2007/08 113 28 14 2 3 1.15
2008/09 124 36 25 8 0 1.10
49 28 103 89 0
2010/11 48 120021.02
excluding transported horses.
1 December until 15 April, all birth related deaths excluded.
Bold letters indicate horse years with a dzud winter.
Winter Crash of Re-Introduced Przewalski’s Horses
PLoS ONE | 5 December 2011 | Volume 6 | Issue 12 | e28057
collars in July 2010 we also only came across the carcasses of 2
wild asses from the preceding winter.
The GPS data from the 10 collars revealed that wild asses had
moved west during the dzud period (Figure 1C). This is a pattern
we had not observed in previous years (Figure S4). Three animals
even went well beyond all previous wild ass locations and one
crossed the border fences between Mongolia to China and back
(Figure 1C). The rangers also reported that they had not seen any
wild asses in the eastern part of the park during the winter and that
the very first wild ass did not arrive back in the Chonin us area
until the beginning of April 2010.
Herders and livestock
The dzud winter 2009/2010 was certainly the most extreme
winter in Mongolia during the past 50 years. Fifteen out of
Mongolia’s 21 provinces were declared disaster zones and over 7.8
million livestock, 17% of the national stock, are believed to have
perished [29]. The dzud disaster was caused by the combination of
a very dry summer followed by a long, very cold winter with deep
snow. In many places the situation was aggravated by excessive
livestock stocking rates, reduced mobility and the lack of winter
fodder reserves [16,28].
Due to its geographic location, the eastern part of the
Dzungarian Gobi was particularly heavily hit. In and around the
Great Gobi B SPA stocking rates are moderate, although numbers
were on the rise (from ,60,000 in 2001, see [32], to ,75,000 by
the end of 2009). While there is little indication for pasture
degeneration so far, there is evidence for competition for good
winter camps (e.g. suboptimal campsites being used, and camps
being burnt down as acts of sabotage), which constitute a key
resource for local herders [13].
Competition for winter camps allows little spatial flexibility and
when combined with a lack of infrastructure and supporting
services families basically had to stay put, even when the
conditions were bad. Furthermore, during the 2009/2010 dzud
winter deep snow arrived so quickly that people were unable to
move, even if they had an alternative place to go to. Several
families in the eastern part of the Great Gobi B SPA got trapped at
fall camps, which are normally only used for a few weeks after
leaving the summer pastures and before settling into the winter
camps. These families had extended their presence at the
temporary fall camps unusually long because of the rather poor
condition of their winter pastures caused by the preceding
summer’s drought.
Przewalski’s horses
The weather conditions during the dzud 2009/2010 resulted in
the north-eastern part of the Dzungarian basin receiving large
amounts of snow. This area also happened to be the winter range
of the majority of the re-introduced Przewalski’s horses, resulting
in a major population crash. A modelling exercise had previously
identified natural catastrophes as having the greatest influence on
the extinction risk of the small Przewalski’s horse population in
Takhin Tal [26]. However, the frequency, spatial extent and
severity of such unusual weather events are difficult to predict,
making long term model predictions of population growth of small
populations in harsh environments even more unreliable than
under ‘‘normal environmental stochasticity’’ [4].
Provision of hay after the main die-off may have helped to
reduce late winter mortality at Chonin us and Takhi us, explaining
the slightly lower losses among Przewalski’s horses as compared to
livestock. However, around Takhin Tal access to hay from early
winter on did not prevent massive losses, possibly because feeding
was impossible during snow storms [35]. Consequently, interven-
tion possibilities for free-ranging animals during natural disasters
of the magnitude of the 2009/2010 dzud seem very limited.
On the regional scale the damage to Przewalski’s horse recovery
in Mongolia was somewhat dampened by the fact that the winter
range of the third Przewalski’s group in the Great Gobi B SPA was
located in an area less affected by the dzud, and on a national scale
by the fact that winter losses at Hustai NP were much lower and
only amounted to 10% of the early December population (D.
Usukhjargal unpubl. data). Close cooperation between the two re-
introduction sites already resulted in a transport of horses from
Hustai NP to the Great Gobi B SPA and negotiations for further
transports to speed up population recovery are ongoing.
Furthermore, there are plans to transport Przewalski’s horses
from a breeding facility in adjacent China to Mongolia.
As during the dzud 2000/2001, the youngest age classes
suffered the highest mortality [26]. What came as a surprise
though was that zoo born Przewalski’s horses, despite having lived
in the Gobi for multiple years, may still have a lower survival
probability than those born in the Gobi. It also appears that mares
may be less susceptible to succumbing to dzud conditions than
stallions. Histo-pathological examination of samples collected from
33 Przewalski’s horse carcasses did not suggest that disease played
a major role in the dzud 2009/2010 die-off (A. Ku¨bber-Heiss,
unpubl. data), as had been the case during the dzud 2000/2001
[34]. Our findings suggest that with the increasing proportion of
Gobi born Przewalski’s horses the re-introduced population may
become more robust in facing future dzud conditions, although we
have yet to understand the underlying adaptive mechanisms.
Although, contrary to livestock, Przewalski’s horses were not
constrained to any particular place by the choice of a herder, they
nevertheless stayed put in the area most affected by the dzud.
Access to hay from early winter on may have been a reason for the
17 Przewalski’s horses around Takhin Tal to remain. However,
the 102 horses at Chonin us did not receive hay until after the
main die-off had already happened. The area most heavily
impacted included the horses’ winter and summer range. The re-
introduced Przewalski’s horses seem very conservative in their
range use, having comparatively small home ranges, clear habitat
preferences, and showing only a slow tendency for range
expansion [24]. Przewalski’s horse groups in the north-eastern
part of the SPA overlap and interact, but there seems to be no
contact to the group at Takhi us (O. Ganbaatar unpubl. data).
Consequently, the re-introduced population still has a rather
limited spatial knowledge, and venturing beyond the known range
during extreme conditions would be somewhat of a risky strategy.
Table 3. Averaged model parameters of a general additive
model (GAM) for survival or mortality of 119 Przewalski’s
horses that wintered in the eastern part of the Great Gobi B
Strictly Protected Area during the dzud winter 2009/2010.
Coefficient z value
Relative variable
Intercept 20.669 0.413 0.652
na na 0.062
Sex_Stallion 20.855 0.472 0.070 0.68
Origin_Zoo 21.763 0.994 0.076 0.61
See Figure S3 for relationship.
Pvalue based on the full model including all 3 variables.
Winter Crash of Re-Introduced Przewalski’s Horses
PLoS ONE | 6 December 2011 | Volume 6 | Issue 12 | e28057
Whether autochthonous Przewalski’s horses were more mobile
than the present-day re-introduced animals is unknown. From
other species we know that sedentary and migratory animals or
subpopulations can coexist within the same species and habitat
[39]. Thus it is possible that during captive breeding either the
behavioural tradition or the genetic component for exploratory
movements was lost. However, the severe effect of this localized
catastrophic event was largely due to the small size and limited
range of the present day Przewalski’s population. A large and
continuous population would be able to counteract local
population lows or extinctions via re-colonization.
Asiatic wild asses
Although we have no means of quantifying the impact of the
dzud 2009/2010 on the Asiatic wild ass population, evidence
suggests that mortality was low. GPS tracking data and ranger
observations show that wild asses moved away from the most
severely affected areas in the north-eastern part of the Dzungarian
Gobi, a pattern we had not observed in previous years [12,24],
(also see Figure S4). Contrary to sympatric Przewalski’s horses,
wild asses have large home ranges, show little dependence on a
particular habitat type and seem to live in fission-fusion groups
[24]. Due to the different scale of habitat use, winter severity
within the asses’ home range was patchy, rather than uniform as
was the case within the much smaller Przewalski’s horse home
ranges, or punctual like for the fixed winter camps of local herders
and their livestock. The lack of a distinct spatial substructure
within the wild ass population [12] likely facilitates information
transfer between individuals and subgroups [40] and may allow
individuals access to information well beyond their individual
home range [41], making exploratory long distance movements
during extreme conditions less risky.
Management Recommendations
The dzud winter 2009/2010 is a text book example for how
vulnerable small and spatially confined populations are in an
environment prone to fluctuations and catastrophes. Losses of this
magnitude are difficult to predict or model in any reasonable
framework and will in any case remain probabilities. As long as
populations remain small and spatially confined, success is not
guaranteed, necessitating a long term conservation commitment.
The difference in how severely the Przewalski’s horse popula-
tion was affected, on a local as well as a national scale, highlights
the advantage of distributing your ‘‘eggs in more than one basket’’.
Consequently, the national strategy for Przewalski’s horse
conservation in Mongolia should continue to aim for multiple
re-introduction sites with spatially dispersed populations and close
cooperation among projects on a national as well as an
international scale.
Wild asses were obviously able to avoid the worst of the dzud
winter by moving west, up to 50 km beyond the Great Gobi B SPA
boundary. These long distance movements and range shifts highlight
again how vulnerable migratory or nomadic ungulates are to
fragmentation and how important it is to manage them on a
landscape-level, including multi-use areas outside of protected areas.
The spatial flexibility of local herders is restricted by the limited
availability of suitable winter camps and further complicated by
administrative boundaries and a lack of cooperation beyond the
extended family. Consequently, outrunning a dzud disaster is
hardly an option and people will need to prepare for dzud events
by means of banking during good years, improved husbandry and
control of stocking rates and diversification of their means of
income. Certainly, the herding sector will not be able to provide a
livelihood for a growing population in the future.
Supporting Information
Figure S1 Snow cover dynamics from the first snowfalls
in November 2009 to snow melt in April 2010, over-
imposed with high loss areas from the % livestock loss
prediction map. Generally, high loss areas correspond with
areas that received snow early and where snow stayed long. Snow
depth can be indirectly inferred from snow melt patterns.
Quantitative analyses were hindered by 1) the inability to remotely
measure snow depth, and 2) the high percentage of satellite images
with total or partial cloud cover (e.g. images top left & images in
the middle), resulting in large no-data zones.
Figure S2 Snow conditions in and around Takhin Tal in
February and March 2010.
Figure S3 Probability of mortality for the 119 Przewals-
ki’s horses that wintered in the east part of the Great
Gobi B SPA during the dzud winter 2009/10 based on
age. The solid line shows the value predicted by the general
additive model (GAM) based on a spline with 8 knots. The dashed
lines are the 95% credibility intervals.
Figure S4 GPS positions of 8 wild asses between July
2002 and July 2009, years with no dzud winters. No
avoidance of the eastern part of the park, as in 2009/10, is seen.
For detailed description of data collection and monitoring period
see [24] and [33].
Table S1 GPS data from 10 Asiatic wild asses moni-
tored from July 2009 until July 2010.
Table S2 Temperatures based on hourly measurements
at Takhin Tal research station at the NE edge of the
Great Gobi B SPA in SW Mongolia. A) Average monthly
temperatures from April 2003 through August 2010. B) Daily
mean temperature from November 2009 through April 2010.
Table S3 Averaged model parameters of a general
linear model (GLM) for survival or mortality of 39 adult
Przewalski’s horse mares (age
4 years) that wintered in
the eastern part of the Great Gobi B SPA during the dzud
winter 2009/10.
This work would not have been possible without the dedicated work of the
local rangers under the worst possible conditions. We are grateful to Felix
Knauer who helped with the statistical analysis, Henrik von Wehrden who
helped with acquiring the MODIS/Terra satellite snow cover images, John
Linnell who provided comments and corrections on an earlier version of
this manuscript and Karin Svadlenak-Gomez who helped with the final
editing. This work is dedicated to the brave people in and around the Great
Gobi B SPA who suffered so much hardship during the dzud winter 2009/
Author Contributions
Analyzed the data: PK. Wrote the paper: PK. Collection of field data: PK
OG NA NE. Capture and collaring of animals: CW PK OG NE. Winter
monitoring of study site 2009/2010: OG NA NE. Manuscript correction
and comments: CW CS.
Winter Crash of Re-Introduced Przewalski’s Horses
PLoS ONE | 7 December 2011 | Volume 6 | Issue 12 | e28057
1. Lacy RC (2000) Considering threats to the viability of small populations using
individual-based models. Ecol Bull 48: 39–41.
2. Goodmann D (1987) The demography of chance extinction. Viable populations
for conservation. In: Soule´ ME, ed. Viable populations for conservation,
Cambridge University Press, Cambridge, UK. pp 11–34.
3. Fischer J, Lindenmayer DB (2000) An assessment of the published results of
animal relocations. Biol Conserv 96: 1–11.
4. Asbjørnsen EJ, Sæther BE, Linnell JDC, Engen S, Andersen R, et al. (2005)
Predicting the growth of a small introduced muskox population using population
prediction intervals. J Appl Ecol 74: 612–618.
5. Sæther BE, Engen S, Filli F, Aanes R, Schro¨ der W, et al. (2002) Stochastic
population dynamics of an introduced Swiss population of the ibex. Ecology 83:
6. Saltz D, Rubenstein DI, White GC (2006) The impact of increased
environmental stochasticity due to climate change on the dynamics of Asiatic
wild ass. Conserv Biol 20: 1402–1409.
7. IPCC (2007) Climate Change 2007: Impacts, Adaptation and Vulnerability.
Contribution of Working Group II to the Fourth Assessment Report of the Inter-
governmental Panel on Climate Change’. Parry M, Canziani O, Palutikof J, van
der Linden P, Hanson C, eds. Cambridge University Press, Cambridge.
8. Vetter S (2005) Rangelands at equi librium and non-equ ilibrium: recent
developments in the debate. J Arid Environ 62: 321–341.
9. Fernandez-Gimenez ME, Allen-Diaz B (1999) Testing a non-equilibrium model
of rangeland vegetation dynamics in Mongolia. J Appl Ecol 36: 871–885.
10. Berger J, Young JK, Berger KM (2008) Protecting migration corridors:
challenges of optimism for Mongolian saiga. Plos Biol 6: 1365–1367.
11. Mallon DP, Jiang Z (2009) Grazers on the plains: challenges and prospects for
large herbivores in central Asia. J Appl Ecol 46: 516–519.
12. Kaczensky P, Kuehn R, Lhagvasuren B, Pietsch S, Yang W, et al. (2011)
Connectivity of the Asiatic wild ass population in the Mongolian Gobi. Biol
Conserv 144: 920–929.
13. Fernandez-Gimenez ME (1999) Sustaining the Steppes: a geographical history of
pastoral land use in Mongolia. Geogr Rev 89: 315–342.
14. Scharf K, Ferna´ ndez-Gime´nez M, Batbuyan B, Enkhbold S (2010) Herders and
hunters in a transitional economy: The challenge of wildlife and rangeland
management in post-socialist Mongolia. In: Du Toit JT, Kock R, Deutsch JC,
eds. Wild Rangelands – Conserving Wildlife While Maintaining Livestock in
Semi-Arid Ecosystems, Wiley-Blackwell, Oxford, UK. pp 312–339.
15. Begzsuren S, Ellis JE, Ojima DS, Coughenour MB, Chuluun T (2004) Livestock
responses to droughts and severe winter weather in the Gobi Three Beauty
National Park, Mongolia. J Arid Environ 59: 785–796.
16. Tachiiri K, Shinoda M, Klinkenberg B, Morinaga Y (2008) Assessing Mongolian
snow disaster risk using livestock and satellite data. J Arid Environ 72:
17. Barrett MW (1982) Distribution, Behavior, and Mortality of Pronghorns during
a Severe Winter in Alberta. J Wildlife Manage 46: 991–1002.
18. Zhirnov LV, Ilyinsky VO (1986) The Great Gobi National Park - a refuge for
rare animals in the central Asian deserts. Report for the United Nations
Environmental Programme (UNEP) by the Centre for International Projects,
GKNT, Moscow, Russia. Vneshtorgizdat, Moscow, USSR.
19. Reim ers E (1982) Winter mortality and population trends of reindeer in
Svalbard, Norway. Arctic and Alpine Res 14: 295–300.
20. Milner-Gulland EJ, Lhagvasuren B (1998) Population dynamics of the
Mongolian gazelle Procapra gutturosa: an historical analysis. J Appl Ecol 35:
21. Mallon DP (2008) Gazella subgutturosa. In: IUCN 2010. IUCN Red List of
Threatened Species. Version 2010.4. Available:
Accessed 2011 Aug 14.
22. Singh NJ, Milner-Gulland EJ (2011) Monitoring ungulates in Central Asia:
current constraints and future potential. Oryx 45: 38–49.
23. Wit P, Bouman I (2006) The tale of the Przewalski’s horse. KNNV Publishing,
Utrecht, The Netherlands.
24. Kaczensky P, Ganbaatar O, von Wehrden H, Walzer C (2008) Resource
selection by sympatric wild equids in the Mongolian Gobi. J Appl Ecol 45:
25. van Dierendonck MC, Wallis de Vries MF (1996) Ungulate reintroductions:
Experiences with the takhi or Przewalski horse (Equus ferus przewalskii)in
Mongolia. Conserv Biol 10: 728–740.
26. Slotta-Bachmayr L, Boegel R, Kaczensky P, Stauffer C, Walzer C (2004) Use of
population viability analysis to identify management priorities and success in
reintroducing Przewalski’s horses to southwestern Mongolia. J Wildlife Manage
68: 790–798.
27. Boyd L, King SRB (2011) Equus ferus. In: IUCN 2011. IUCN Red List of
Threatened Species. Version 2011.1. Available:
Accessed 2011 Nov 22.
28. Sternberg T (2010) Unravelling Mongolia’s Extreme Winter Disaster of 2010.
Nomadic Peoples 14: 72–86.
29. UN Mongolia Country Team (2010) Mongolia – Dzud Appeal. Office for the
Coordination of Humanitarian Affairs (OCHA), New York, USA and Geneva,
Switzerland. Available:
asp?Page = 1856. Accessed 2011 Jun 10.
30. von Wehrden H, Wesche K (2007) Relationships between climate, productivity
and vegetation in southern Mongolian drylands. Basic and Applied Dryland
Research 1,2: 100–120.
31. von Wehrden H, Wesche K, Tungalag R (2006) Plant communities of the Great
Gobi B Strictly Protected Area, Mongolia. Mongolian Journal of Biological
Sciences 4: 3–17.
32. Kaczensky P, Enkhsaikhan N, Ganbaatar O, Walzer C (2007) Identification of
herder-wild equid conflicts in the Great Gobi B Strictly Protected Area in SW
Mongolia. Exploration into the Biological Resources of Mongolia 10: 99–116.
33. Kaczensky P, Ito TY, Walzer C (2010a) Satellite telemetry of large mammals in
Mongolia: what expectations should we have for collar function? Wildl Biol
Pract 6: 108–126.
34. Robert N, Walzer C, Ru¨ egg SR, Kaczensky P, Ganbaatar O, et al. (2005)
Pathologic findings in reintroduced Przewalski’s horses (Equus caballus przewalskii)
in Southwestern Mongolia. J Zoo Wildlife Med 36: 273–285.
35. Kaczensky P, Ganbaatar O, Altansukh N, Enkhsaikhan N (2010b) Winter disaster
in the Dzungarian Gobi – crash of the Przewalski’s horse population in Takhin Tal
2009/2010. Field report. Available:
files/Dow nloads/PDF/Studien/2010_08_20_Przewalski%20horse%20report.
pdf. Accessed 2011 Jun 16.
36. Hall DK, Riggs GA, Salomonson VV (2006) updated weekly. MODIS/Terra
Snow Cover 8-day L3 Global 500 m Grid V005, November 2009 – April 2010.
BoulderColorado USA: National Snow and Ice Data Center. Digital media.
37. R Development Core Team (2007) R: A Language and Environment for
Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.
Available: Accessed 2011 Jun 16.
38. Calcagno V, De Mazancourt C (2010) glmulti: An R Package for Easy
Automated Model Selection with (Generalized) Linear Models. Journal of
Statistical Software 34(12): 29. Available:
paper. Accessed 2011 Mar 2011.
39. Bunnefeld N, Bo¨ rger L, van Moorter B, Rolandsen CM, Dettki H, et al. (2011) A
model-driven approach to quantify migration patterns: individual, regional and
yearly differences. J Ecol 80: 466–476.
40. Sundaresan SR, Fischhoff IR, Dushoff J, Rubenstein DI (2007) Network metrics
reveal differences in social organization between fission-fusion species, Grevy’s
zebra and onager. Oecologia 151: 140–149.
41. Bryson J, Kaczensky P (2009) Exploring knowledgedissemination as a selective force
for aggregation:Preliminaryresults from modelling wildAsiatic asses.10th European
Conference on Artificial Life (ECAL); SEP 23–26, 2009; Budapest, Hungary.
abstracts/Bryson.pdf. Accessed 2011 Aug 14.
Winter Crash of Re-Introduced Przewalski’s Horses
PLoS ONE | 8 December 2011 | Volume 6 | Issue 12 | e28057
... Dzud (extremely severe winter) induced wild-horse mortality rates of 21% in 2000/2001 and of 60% in 2009/2010 (livestock mortality was even higher, 67% in 2009/2010). This is a huge challenge for small and isolated animal populations such as the reintroduced population of Przewalski's horse [27]. Therefore, horses living in these extreme conditions serve as an ideal model species for monitoring the relationship between reintroduced animals and their environment. ...
... In equids, forage abundance is the most important determinant for habitat use because of their digestive anatomy [30,31] and they may spend more than half of their daily time budgets grazing to get enough nutrients [32]. However, inconsistent and contradictory information has been described by previous observations of the feeding behaviour and daily budget of horses (feral horses [24][25][26][27][28][29][30][31][32][33][34]; wild horses [25,32,35]:). In Hustai National Park, Mongolia, harems of Przewalski's horses usually feed in the morning and evening, before and after the walk to the water source [25,35]. ...
... The disastrous winter of 2009 and 2010, resulting in the loss of the 60% of Przewalski's horse population, demonstrated how sensitive are the small and spatially restricted populations to severe climatic and environmental changes [27]. As stated by Slotta-Bachmayr [93] the severity level of natural climatic conditions has the highest influence on extinction risk and population size of Przewalski's horses in the Great Gobi B, according to population simulation model VORTEX. ...
Full-text available
Background Reintroduction is a common technique for re-establishing threatened species. However, the adaptation to novel habitats with distinct conditions poses a risk of failure. Weather conditions affect the behaviour of animals, and thus, their adaptation to new conditions and survival. Reintroduced Przewalski’s horses living in Mongolia’s continental arid climate with extreme temperature and precipitation variability, serve as an ideal model species for studying the behavioural response of selected groups to these harsh conditions. Methods The research was conducted in The Great Gobi B Strictly Protected Area, Mongolia. In summer 2018, three groups were recorded (Azaa, Tsetsen and Mares18) involving 29 individuals. In Spring 2019, 4 groups were recorded (Azaa, Tsetsen, Hustai1 and Mares19) involving 34 individuals. In Autumn 2019, 4 groups were recorded (Azaa, Tsetsen, Hustai2 and Tanan) involving 35 individuals. Thirteen weather variables were recorded in 10-min intervals, together with the percentage representation of selected behavioural categories (feeding, locomotion, resting, and social). The effect of weather on behaviour was analysed through GLMM. Influence of the group-history factors (recently reintroduced, long-term reintroduced and wild-born) was also analysed. Results Feeding significantly increased with cloudy and windy conditions and was more frequent in autumn than spring and summer. Locomotion was positively explained by temperature and cloudiness and was higher in summer than spring and autumn. Resting behaviour decreased with altitude and cloudiness, and the dispersion of the group was lower when resting. Increased social interactions were observed with higher temperatures and were more frequent in summer compared to spring and autumn. Differences were found in the display of the behaviours among the selected harems, showing interesting patterns when grouping them according to their origin and experience. Conclusions Weather patterns seem to influence the behaviour of Przewalski’s horse. These results might assist in further management plans for the species, especially in the view of intensifying climate change and alteration of weather patterns. As previously suggested, after approximately 1 year, horses adapt to novel conditions and display the typical behavioural pattern of wild-born Przewalski’s horses.
... Historically, the population in Mongolia was continuous with khulan in the three northern provinces of China, namely Xinjiang Uygur Autonomous Region (hereafter referred to as "Xinjiang"), Gansu Province ("Gansu"), and Inner Mongolia Autonomous Region ("Inner Mongolia"). However, this situation has changed with the construction of the border fence along the international border in the 1990 s and its subsequent reinforcement (Yang, 2007;Kaczensky et al., 2011aKaczensky et al., , 2011bLinnell et al., 2016). ...
... By law, all expressways and railways must be fenced in China, with the exception of specific wildlife overpasses or bridges. The new expressway and railway that run parallel to the old highway G216 constitute a major barrier to the seasonal movement of khulan, making them more vulnerable to local catastrophic events (i.e., droughts or severe winters, Kaczensky et al., 2011a) thereby reducing the perspective for long term survival of the species in KNR (Chen et al., 2021;Appendix D). ...
... The khulan population in northern China was once continuous with khulan in Mongolia and in the highly fragmented areas along the Mongolian border, long-term survival will depend on immigration of khulan from Mongolia. In Xinjiang, connecting KNR with the Great Gobi protected areas on the Mongolian side, would benefit the species in the Junggar basin by increasing the range and allowing for higher movement flexibility and genetic exchange over a large transboundary protected area network (Kaczensky et al., 2011a(Kaczensky et al., , 2011bCMS, 2019). Hence there is an urgent need to develop a national khulan conservation strategy and action plan which initiates cross-border cooperation with Mongolia to safeguard the long-term survival of the species in the Gobi region. ...
Full-text available
Understanding the changes in population size, distribution and threats, is essential for assessing the status of threatened species. Northern China is believed to be an important stronghold for the Near Threatened Asiatic wild ass or khulan (Equus hemionus), but a recent assessment of the species has been lacking. To document change and updated the current status of khulan in China, we conducted a literature review targeting peer-reviewed and grey literature, newspaper articles, and summarized the results own field surveys and interviews from part of the species range. For a better understanding of the threats to khulan in China, we summarized the results of studies on environmental habitat factors and human disturbances for khulan, most of which are only available in Chinese language. Our results suggest that khulan in China have experienced a dramatic decline and fragmentation of their distribution range caused by excessive anthropogenic interferences. The remaining khulan range in China covers probably less than 40,000 km² and is scattered over several nature reserves and the border areas in northern Xinjiang, northwestern Gansu, and western Inner Mongolia. We estimate the remaining population at about 4000 individuals, with ~80% found in Kalamaili Mountain Ungulate National Nature Reserve in Xinjiang. The occurrences along the border with Mongolia are small and dependent on cross-border movements, which are currently severely hindered by border fences. Over the past 15 years, Kalamaili Mountain Ungulate National Nature Reserve was exposed to various human pressures and experienced dramatic population fluctuation in the khulan population size. Key factors which negatively influenced khulan were habitat loss, fragmentation, and disturbance due to mining exploration and infrastructure development. No systematic monitoring of khulan is done in the rest of the khulan range, but whereas illegal hunting seems no longer a serious threat, infrastructure development and land use changes (including increasing livestock numbers) are happening throughout the remaining range of khulan in China. Hence there is an urgent need to develop a national khulan conservation strategy and initiate cross-border cooperation with Mongolia to safeguard the long-term survival of the species in the Gobi region.
... Energy expenditure may be particularly affected by ground icing from rain-on-snow events (Hansen et al. 2014) or periods of heavy snowfalls and extremely cold temperatures called dzuds (Bekenov et al. 1998). Dzuds are believed to have led to the extinction of several khulan populations throughout Central Asia (Bannikov 1981) and recently nearly wiped out a small reintroduced population of Przewalski's horse (Equus ferus przewalskii) in the Mongolian Gobi (Kaczensky et al. 2011). Warmer temperatures and increased precipitation in the Arctic have favored an earlier onset and increased number of biting insects. ...
... In the US, for example, migration maps are used by local municipalities to develop planning guidelines to minimize impacts on migrating animals. (Kaczensky et al. 2011) was one of the key arguments for doubling the size of the Great Gobi B Strictly Protected Area. The success of conservation efforts relies on decision makers having easy access to the migration data themselves. ...
... GPS tracking data have helped visualize the enormous range and unpredictability of nomadic movements ( Joly et al. 2020, Nandintsetseg et al. 2019. Recent studies suggest wide-ranging nomadic movements are related to habitat predictability and heterogeneity (Mueller et al. 2011b) and can be triggered by extreme weather events (Kaczensky et al. 2011). Together, these data highlight the need for dynamic conservation strategies, such as mobile ranger units during critical life stages (Bull et al. 2013) or temporary road closures during mass movements (Whittington et al. 2019). ...
Full-text available
Our understanding of ungulate migration is advancing rapidly due to innovations in modern animal tracking. Herein, we review and synthesize nearly seven decades of work on migration and other long-distance movements of wild ungulates. Although it has long been appreciated that ungulates migrate to enhance access to forage, recent contributions demonstrate that their movements are fine tuned to dynamic landscapes where forage, snow, and drought change seasonally. Researchers are beginning to understand how ungulates navigate migrations, with the emerging view that animals blend gradient tracking with spatial memory, some of which is socially learned. Although migration often promotes abundant populations—with broad effects on ecosystems—many migrations around the world have been lost or are currently threatened by habitat fragmentation, climate change, and barriers to movement. Fortunately, new efforts that use empirical tracking data to map migrations in detail are facilitating effective conservation measures to maintain ungulate migration. Expected final online publication date for the Annual Review of Ecology, Evolution, and Systematics, Volume 52 is November 2021. Please see for revised estimates.
... The effects of catastrophic environmental factors on mortality are described under 'population dynamics' in Domain 4, but climatic events such as drought, storms, heavy snow, extreme cold, thunderstorms, lightning strikes and fire have all been reported to cause mass mortalities in wild horse populations [14,[113][114][115][116]. ...
... When mass mortality events do occur, populations can be relatively slow to recover because of low recruitment rates, however, sudden reductions in populations also result in a compensatory increase in reproductive rate [218,220]. Catastrophic climate events can have the greatest effect on wild horse populations with population declines of 12-61 % reported after extreme cold winters in some parts of the US [113,114,116], mass mortalities during drought, with a 51% reduction reported during drought in an Australian population [14], 28% of a population dying during a thunderstorm in Argentina [115] and 4% dying with a lightning strike in Pryor Mountain, US [218]. ...
Full-text available
A detailed understanding of what is usual for a species under optimal conditions is critical for identifying and interpreting different features of body function that have known impacts on animal welfare and its assessment. When applying the Five Domains Model to assess animal welfare, the key starting point is therefore to acquire extensive species-specific knowledge relevant to each of the four physical/functional Domains of the Model. These Domains, 1 to 4, address areas where objective information is evaluated and collated. They are: (1) Nutrition; (2) Physical environment; (3) Health; and (4) Behavioural interactions. It is on the basis of this detailed knowledge that cautious inferences can then be made about welfare-relevant mental experiences animals may have, aligned with Domain 5, Mental State. However, this review is focused entirely on the first four Domains in order to provide a novel holistic framework to collate the multidisciplinary knowledge of horses required for undertaking comprehensive welfare assessments. Thus, inferring the potential mental experiences aligned with Domain 5, the final step in model-based welfare assessments, is not considered here. Finally, providing extensive information on free-roaming horses enables a better understanding of the impacts of human interventions on the welfare of horses in both free-roaming and domestic situations.
... To meet their breeding, growth, and resource requirements, wildlife often undertakes a variety of movements among habitat patches (Jachowski and Singh, 2015;Ofstad et al., 2016). The specific linear habitats that are used during travel between habitat patches are typically described as ecological corridors (Rosenberg et al., 1997;Beier and Noss, 1998); allowing animals to traverse to and from seasonal ranges that provide access to higher quality forage or specific locations for completing their life cycle (Sawyer and Kauffman, 2011), or to survive catastrophic climate events (Kaczensky et al., 2011a). As a result, such ecological corridors are critical for improving habitat connectivity, maintaining population growth and gene flow, and ensuring individual proliferation and settlement in new areas (Jordan, 2000;Mech and Hallett, 2001). ...
... Water availability and switching among the sparsely located water bodies may be a key driver of the high mobility of khulan (Nandintsetseg et al., 2016). Such highly mobile behavior enables them to cope with localized catastrophic weather events, but this requires habitat connectivity at a landscape scale (Kaczensky et al., 2011a;Wingard et al., 2014). Khulan face the threats of overhunting, habitat loss, fragmentation, and linear roads throughout their distribution range (Ito et al., 2013;Xu et al., 2022). ...
Full-text available
Habitat use by wild animals may be subject to spatial and temporal changes. The Kalamaili Mountain Ungulate Nature Reserve (KNR), inhabited by >80 % of Chinese khulan (Equus hemionus), is one of the most important desert ungulate reserves in Northwestern China. However, frequent human disturbance caused by mining development and road construction in KNR has interfered with or completely blocked their movement and access to parts of the reserve. In our study, we assessed the distribution of suitable habitat and ecological corridors of khulan before the mining development (2005), at the peak of mining development (2011), and after ecological restoration and expressway construction (2019). The results showed that most of the core habitat and ecological corridors of khulan were concentrated in the middle of the reserve and on the east side of the road. The habitat of khulan in KNR went from a good natural habitat in 2005 to deterioration due to mining development in 2011. Then it partially recovered due to ecological restoration from 2015 onwards. However, in 2019, road construction likely hindered its recovery to pre-mining levels. The location of corridors accordingly varied with the change of core habitat in different years. Nevertheless, the corridors intersecting roads generally had higher centrality values, indicating their higher importance. Our study revealed the significant impacts that mining development and road construction have had on the distribution of core habitats, ecological corridors, and movement of khulan. These results provide a scientific basis and valuable arguments for strengthening habitat connectivity for khulan.
... Our frog example can be viewed as a form of a collective memory. We suggest that collective memories that are important for a population's persistence are a category of distributed adaptation (e.g., Kaczensky et al., 2011); however, the more interesting among them, and those in which the distributed adaptations' perspective may be particularly useful, are cases in which different individuals carry different information and their sum is greater than its parts: those in which no single individual has the relevant information (Seeley et al., 2006). In particular, no individual has the "correct" or best information. ...
... involves memories of older individuals, produced by longer experience or experience of different environmental conditions (Morales et al., 2010). It may be possible for the population to utilize such information, by imitating the older individuals, when an adverse condition reoccurs (e.g., Kaczensky et al., 2011). There are two questions that can help distinguish between cases of collective memory and distributed adaptations. ...
Full-text available
Species’ adaptation to their environments occurs via a range of mechanisms of adaptation. These include genetic adaptations as well as non-traditional inheritance mechanisms such as learned behaviors, niche construction, epigenetics, horizontal gene transfer, and alteration of the composition of a host’s associated microbiome. We propose to supplement these with another modality of eco-evolutionary dynamics: cases in which adaptation to the environment occurs via what may be called a “distributed adaptation,” in which the adaptation is not conferred via something carried by an individual of the adapted species (as with genes, behavior, or associated microbes), but by some structural or compositional aspect of the population. Put differently, the adaptively relevant information cannot be reduced to information possessed by a single individual, whether genetic or otherwise. Rather, the adaptively relevant information is distributed, and is found strictly at the population level. While human culture is presumably such a case, as may be cases found in social insects, we want to suggest that there are other cases that belong to this category and to explore its evolutionary implications. In particular, we discuss the factors that affect whether adaptive information is stored in a distributed way, to what degree, and what kinds of adaptive information are most likely to be found in this modality of adaptation.
... Although initial exploratory movements were to be expected, the two reintroduced kulan which we monitored over two winters showed little indication of restricting their movements, but rather kept exploring some new areas to the south-east (mare 26860) and south-west (mare 26859). The range sizes of the reintroduced kulan in the Torgai region are in the same order of magnitude of those of kulan from Mongolia's South Gobi Region (Kaczensky et al., 2011b;Payne et al., 2020), but contrary to kulan in the Gobi, which seem to be primarily nomadic (Noonan et al., 2020), the ctmm analysis indicated a clear homerange for the reintroduced kulan. ...
... We see the large ranges and high mobility of the reintroduced kulan as a sign of success as kulan along the steppe desert gradient should not be expected to stay only in the steppe or only in the desert habitat, but rather migrate between those two on a seasonal basis. Large ranges and high mobility are the best adaptation to highly variable climatic conditions or extreme events and can lower the risk of mass mortality due to droughts or extreme winters ("dzud") as has been shown for kulan in Mongolia (Kaczensky et al., 2011b). ...
Full-text available
Asiatic wild ass, or kulan ( Equus hemionus kulan ) were once a key species of the Eurasian steppes and deserts. In Kazakhstan they went extinct by the 1930s. Early reintroductions have reestablished the species in two protected areas, but the species has reclaimed <1% of their former range and remained absent from central Kazakhstan. To initiate restoration in this vast region, we captured and transported a first group of nine wild kulan to a large pre-release enclosure in the Torgai region in 2017, and two more in 2019. We used direct observations and post-release movement data of four kulan equipped with GPS-Iridium collars to document their adaptation process in a vast novel habitat without conspecifics. For comparison with movements in the source populations, we additionally equipped two kulan in Altyn Emel National Park and six in Barsa Kelmes State Nature Reserve. The nine transported kulan formed a cohesive group with very high movement correlation in the enclosure. After release, the group initially stayed tightly together but started to break up by mid-May and all kulan travelled independently by mid-August. With 48,680–136,953 km ² , the 95% Autocorrelated Kernel Density Estimation ranges of the reintroduced kulan were huge and about 10–100 times larger than those in the source populations. The reintroduced mares never reconnected, there was no evidence of successful reproduction, and two of the four collared mares were killed by poachers and one died of natural causes. At least one stallion survived in the wild, but the fate of the other uncollared animals remains unclear. We speculate that the fission-fusion dynamics and low movement correlation of kulan societies and the need for migratory movements harbours the risk that animals released into a novel environment loose contact with each other. This risk is likely enhanced in steppe habitats where movement constraining factors are absent. Further kulan reintroductions to the steppes and deserts of central Kazakhstan should aim to release larger groups and build up the free-ranging population quickly to reach a critical mass, increasing the chance of kulan encountering conspecifics to successfully breed and increase their chances of survival.
... In unpredictable environments, the population-level benefits of migration and nomadism outweigh more restricted movement tactics so long as the landscape remains connected (Teitelbaum and Mueller 2019). For example, during a high-snowpack winter in Mongolia, a spatially confined population of Przewalski's horses (Equus ferus przewalskii) experienced greater mortalities than a nearby population of Asiatic wild ass (Equus hemionus), which were more exploratory and able to relocate to alternative habitat (Kaczensky et al. 2011). Because the benefits of nomadism depend on landscape connectivity, highly mobile nomadic species should be more threatened by movement barriers than residents or migrants (Teitelbaum and Mueller 2019). ...
Full-text available
Animal movement can mediate the ecological consequences of fragmentation; however, barriers such as fences, roads, and railways are becoming a pervasive threat to wildlife. Pronghorn (Antilocapra americana) habitat in western North America has been fragmented by roads, railways, and fences. Although pronghorn are sensitive to barriers, neither the relative permeability of different barriers to crossing nor their influence on space use have been quantified. We used a large global positioning system (GPS)‐collar dataset of pronghorn (n = 1,010 animal‐years) in Wyoming, USA, to first quantify the likelihood that pronghorn cross each of 5 different anthropogenic barriers, including fences, county roads, railroads, state highways, and interstate highways (i.e., interstates). Next, we assessed how each barrier influenced pronghorn space use during the winter as indexed by the area occupied, and daily displacement relative to the density of barriers on an individual's winter range. The semi‐permeability of the 5 barriers varied substantially, with the interstate being the most severe barrier to pronghorn movement. Pronghorn were >300 times less likely to cross interstates compared to state highways. Although pronghorn space use was rarely influenced by barriers within individual core winter ranges, pronghorn space use was constrained by barriers on the buffered periphery of individual winter ranges. Despite their different permeability to movement, the density of fences and combined interstates and railroads had similarly negative effects on pronghorn space use. Our results illustrate that the degree to which pronghorn avoid crossing barriers may scale up to affect access to habitat. Additionally, our results indicate that the effects of barriers on habitat access are not proportional to their permeability. Our results add to a growing consensus that effective management of mobile species depends on understanding how different kinds of semi‐permeable barriers influence access and use of habitats. We observed that the degree to which pronghorn crossed different barriers varied substantially, and this semi‐permeability may scale up to affect access to habitat. We suggest that barriers on the periphery of core winter ranges can be productive candidates to improve habitat access for pronghorn, given the strong effect peripheral barriers had on pronghorn space use.
... Most of the khulan tracks west of Great Gobi B SPA are from the dzud winter 2009/10, when large amounts of snow covered the Dzungarian Gobi well into March (KACZENSKY et al. 2011 ). Observational evidence and data from regular wildlife surveys has already shown that khulan are free to roam far from water during winter, when snow covers the ground (KACZENSKY et al. 2015a). ...
Full-text available
Water is the lifeline for the world's drylands and the key for the distribution of water-dependent equids like khulan. We developed a simple algorithm using khulan tracks from GPS telemetry to identify waterpoints. This approach allowed us to obtain the first landscape-scale information on the use of waterpoints by khulan in Great Gobi B SPA. We discuss the merits and limitations of the algorithm and the implication for landscape level conservation.
Full-text available
Feral horses (Equus ferus caballus) in the western United States are managed by the Bureau of Land Management (BLM) and United States Forest Service in designated areas on public lands with a goal of maintaining populations in balance with multiple uses of the landscape. Small, isolated populations can be at risk of extirpation from stochastic events and deleterious genetic effects resulting from inbreeding and reduced heterozygosity. The genetic diversity of feral horse herds is periodically monitored using blood or hair samples collected during management gathers (i.e., occasions when the herd is rounded up). We conducted a study to examine genetic characteristics of the feral horse population at the BLM Little Book Cliffs Herd Management Area (HMA) in Colorado, USA, using non‐invasively collected fecal samples. Additionally, we explored whether genotypes could be used to document space use and potential sub‐population development. We used a random sampling scheme, walking transects in sampling areas covering most of the HMA to find and collect fecal samples of all ages, except those that were deteriorating. We collected >1,800 fecal samples from across the study area in May, August, and October 2014. We then identified unique individuals using a suite of microsatellite loci. Our estimates of genetic diversity from fecal samples were higher than those reported from blood and hair samples taken during recent horse gathers, likely because our sample size and spatial distribution was larger. Genotypes revealed that some individuals were found only in certain parts of the study area and at a higher proportion than random; thus, they could be considered residents in those sampling areas. Using discriminant function analyses, we detected 5 genetic groups in the sample population, but these did not correspond to individuals in specific parts of the study area. Our results support the use of fecal DNA to augment direct observations of horse presence and could be used to detect habitat use and areas of high density. Non‐invasive techniques such as fecal DNA sampling can help managers decide whether new individuals need to be translocated to a closed population to maintain genetic diversity without the human safety and animal welfare concerns associated with gathers and invasive techniques. © 2021 The Wildlife Society. This article is a U.S. Government work and is in the public domain in the USA. We found higher estimates of genetic diversity from feral horse fecal samples collected non‐invasively than those reported from blood or hair samples collected during a gather; genotypes derived from fecal samples can be used to show feral horse space use. Fecal samples can therefore be used to examine genetic characteristics of a horse herd in the absence of a gather and can elucidate areas of high horse use.
Full-text available
important management concerns. The Great Gobi B Strictly Protected Area (SPA) in SW Mon-golia is a re-introduction site for the Przewalski’s horse (Equus ferus przewalskii), a stronghold of the Asiatic wild ass (Equus hemionus), and remains an important grazing area for semi-nomadic herders. We show the power of simple inventory and monitoring methods to assess herder-wildlife conflicts, by combining data of: (1) human and livestock demographic data, (2) migration patterns, of semi-nomadic herders, (3) monthly surveys of wild- and domestic ungu-lates, and (4) observations of re-introduced, free-ranging Przewalski’s horses. A total of 111 semi-nomadic families with 57,657 head of livestock use the park, mainly in winter. Grazing impact of small stock affects 33 % of the park area and is virtually absent in the core area. How-ever, due to the unequal distribution of open water, livestock is present at almost all water points. Seasonal wild horse and wild ass distribution seems to be positively linked to water availability and negatively to herder presence. We documented several cases of wild ungulate poaching, but the magnitude of the problem remains unknown. There are still many knowledge gaps and local people need to be more actively involvement in park management. As this is a rather new approach in Mongolia, we suggest park management to move towards adaptive co-management, accompanied by simple, but sound monitoring and evaluation schemes.
Full-text available
Aggregation is a costly strategy for any species, because individuals must compete for resources such as food, shelter and mating opportunities. Hamilton (1971) provides the best-established explanation of aggregation, which is security. For reasons ranging from increased vigilance to cooperative mobbing of aggressors, living in a group can provide benefits as well as costs. But security may not be the only explanation for aggregation. Aggregation can also occur by chance rather than choice, as individuals may come to the same location simply be seeking the same resources, and it may not be worth eort of driving others away from these. One strong indication that there may be more than one selective pressure for aggregation is when a group can be described at multiple levels of hierarchy. A fission- fusion (FF) society is characterised this way. FF societies consist of both parties — relatively small groups travel together, and a community or troop from which parties are drawn. Party size may well be determined by security, but in some species it is not clear that the entire community ever aggregates at the same time. Lehmann et al. (2007) suggest that knowledge exchange may be one selective pres- sure for community-level aggregation in fission-fusion species where party composition is flexible. This is an attractive theory in species like great apes where some commu- nities are known to posses behaviour not seen in other communities. Such 'culture' is particularly obvious when the communities are completely isolated from each other by geographic features such as rivers, but otherwise live in identical ecosystems (Whiten et al., 1999). Here we present new preliminary results concerning aggregation in Asiatic wild asses Equus hemionus. In examining available data simulating multiple hypotheses on their social behaviour, we have identified potential new costs as well as benefits. Our collaboration began because Kaczensky and colleagues were looking for a mecha- nism that could explain the exceptional ability of Asiatic asses to exploit unexpected resource bonanzas in the form of transient oases caused by rainfall in the Gobi desert. Mongolian or Asiatic wild asses are primarily adapted to arid desert steppes and semi-deserts of the Gobi. They can venture well away from water sources and are able to exploit resources that vary in space and time. Asiatic wild asses also live in
Experiences with the réintroduction of the takhi, or Przewalski horse (Equus fcrus przewalskit), in Mongolia can serve as valuable lessons for réintroduction of ungulates in general. We discuss the present taxonoinic, historical, and biological evidence and conclude that the takhi should be viewed as a typical steppe herbivore. Its last refuge, the Dzungarian Gobi, should therefore be seen as a marginal habitat because it consists mainly of desert and semidcsert. Since 1992 two réintroduction projects have been in the acclimatization phase in Mongolia. Despite promising developments, problems with cooperation, management, habitat choice, insufficient knowledge of the ethology of the species, and current land use within the different project areas could jeopardize the successful réintroduction of takhi. We review the conditions required for a potentially successful ungulate réintroduction. Ttie planning of a reintroduction within the framework of safeguarding an entire ecosystem with an integrated management plan appears essential. Each potential réintroduction site should be assessed thoroughly for its suitability, including size, habitat types, current land use, socioeconotnics, legislation, and potential problems. Each site should be provided with one or more acclimatization facilities to harbor genetically and physically healthy, socially adapted animals in biologically sound groups. An organization structure should be established for each reintroduction site. Its objective should be to develop an effective management plan and to carefully monitor the population and its surrounding ecosystem. Special attention should be given to local socioeconomic situations, community participation, and train- . ing of staff for management, research, and ranger and warden activities.
As wildlife populations become smaller, the number of interacting stochastic processes which can destabilize the populations increases: genetic effects (inbreeding and loss of adaptability) and instability of the breeding structure (sex ratio imbalances, unstable age distribution, and disrupted social systems) can decrease population growth and stability. Recent analyses have shown that some populations can be very sensitive to these stochastic processes, at larger population sizes than had been suggested previously, and often in unexpected ways. Interactions among processes can reduce population viability much more so than would be assumed from consideration of isolated factors. For example, in monogamous species, random fluctuations in sex ratio will depress the mean number of breeding pairs in populations with as many as 500 adults. At low population densities, individuals may not be able to find mates, or may not encounter individuals sufficiently unrelated to be accepted as suitable mates. Inbreeding depression of demographic rates can become a significant contributor to population decline in populations with several hundred individuals, even if genetic problems are not the primary threat. Most models of genetic decay in small and fragmented populations assume demographic stability. However, when the increases in demographic fluctuations of small populations are considered, rates of loss of genetic variation and accumulation of inbreeding can be much faster than has been suggested before. These processes can be examined in detailed, individual-based PVA models. Accurate data to parameterize these models, however, are often not available. Thus, we need to interpret cautiously PVA conclusions for populations that are small, highly fragmented, or projected for many generations.
In Mongolia, several record-breaking disastrous dzuds (mass livestock loss directly induced by harsh winter conditions but often influenced by drought in the previous summer) occurred from 1999 to 2003. To understand the mechanism of this climatic disaster, we conducted a tree regression analysis. The predictor variables included two indices developed from remote sensing data—the Normalized Difference Vegetation Index (NDVI) and the Snow Water Equivalent (SWE)—as well as the previous year's livestock numbers and mortality rates. According to the model, serious livestock mortality was associated with low NDVI values (i.e., poor vegetation) in August of the previous year, high SWE values (i.e., significant snow accumulation) in December of the previous year, a high previous year's mortality, and high previous year's livestock population. This result suggests that for dzud risk assessment, we need to monitor snowfall in winter, the vegetation condition in the previous summer, and the density and health condition of the livestock. The tree-based model developed in this study is effective only for a white dzud (deep snow), the most common type of dzud. The large cross-validation error indicates that more data are needed before using the model in order to make predictions.
We investigated how both droughts and dzuds (severe winter weather) control livestock mortality in a non-equilibrium steppe ecosystem of Mongolia, Gobi Three Beauty National Park. These steppe ecosystems have developed under high interannual variability of rainfall and nomadic grazing systems. Interannual precipitation variation was 39%, with 128mm mean annual precipitation. The effect of climate variability and extreme events on livestock mortality is a critical aspect for the Mongolian economy. Analysis of drought and precipitation variability on livestock mortality rate was not significantly influenced by the index of mean annual precipitation and annual winter temperature. Overall, unlike hot dry regions, pastoral livestock mortality in the cold dry regions was affected more by dzuds and annual growing seasonal rain than by droughts. Dzuds can be frequent events, occurring as often as once every 2 and 3 years within a decade. The average annual livestock mortality for the combined drought and dzuds years (18%) was 4.8% greater than the years with dzuds alone, and 7% greater than in years with only drought. Thus livestock mortality appears to be more sensitive to dzuds than to droughts, and that dzuds contributes more to livestock mortality even years where combined drought and winter storms occur.
During the winter of 1975-76 on Edgeoya, Svalbard, an estimated 23% of the Rangifer tarandus platyrhynchus population starved to death. This implies a population net decrease of 16% and a 65% loss of the calves entering the winter in 1975. Animal body weights and fat accumulation in midsummer indicate that the surviving reindeer seemed to recover rapidly. Weather conditions may limit arctic populations of reindeer without a preceding overgrazing of ranges.-Author