Since 2010, a comprehensive program of studying musk deer has been conducted in the Sikhote-Alin region. This musk deer study program has employed the methods of radio telemetry, visual observation, life activity traces survey and photo-video traps. New data on the use of space by musk deer, as well as their daily activity, nutrition, labeling and distribution activities, have been obtained. The research herein demonstrates the necessity for the application of scientific knowledge on the ecology of musk deer for conservation and sustainability.
Musk deer ; Moschus moschiferus ; Home range ; Daily activity ; Radio telemetry ; Complex technique ; Sikhote-Alin ; Sikhote-Alin natural reserve
In Russia, musk deer (Moschus moschiferus ) dwell in Siberia and in the Far East and are mostly known to be a hunting species. Identification of the actual number of musk deer in Russia faces a number of technical difficulties ( Zaitsev, 2006 ). In the late 1990s, the musk deer population was estimated to be approximately 150,000. In 2000, population decline was observed over most of the area (Morgunova et al., 2011 ), including the territory of the Amur region, where hunting of these animals had been banned for a long period of time, and in the Sakhalin region, where the musk deer subspecies listed in the Red List of Russia dwell.
According to the authorities in the Far Eastern Federal District, the musk deer population has stabilized after its decline in the 1990s. For the period of 2008–2010, the musk deer population was estimated to be roughly 50,000–55,000 (Morgunova et al., 2011 ). In Primorsky Krai, the musk deer population is unstable. According to the Central Hunting Control, after a substantial population decline during the 1980s (Zaitsev, 2006 ), the musk deer population maximum was registered in 2003 as 17,430 individuals, and the minimum in 2008 was 11,810 individuals. Only during recent years has there been stabilization of the population (Morgunova et al., 2011 ).
The musk deer population decrease of the 1970s–80s is currently being addressed with controversial protection measures. However, in areas outside of Russia, unfavorable conditions to sustain musk deer abundance have developed. All of the musk deer subspecies that are distinguished according to the classification adopted in Russia (Tsalkin, 1947 and Prikhodko, 2003 ), including those living in Russia, are included in the IUCN Red List. Far Eastern musk deer (M. moschiferus turovi ) is a subspecies that is listed in the Red Book of the Russian Federation Annex III (2001), including taxons of the animals that require special attention to their state in the natural environment. Musk deer derivative trade is controlled by the Convention on International Trade in Endangered Species of Flora and Fauna (CITES).
Musk deer species survival is in jeopardy in some regions of Russia. The downsizing of the musk deer population is caused by unregulated hunting in most places, and this practice has been increasing since the late 1980s because of the illegal export of musk deer and the destruction of their habitats. The highest population density of musk deer in the Russian forests is confined to dark coniferous wooded areas. Extensive fires that have occurred within the past decades as well as intensive logging of coniferous forests both contribute to the destruction and transformation of habitats, significantly affecting the number of musk deer. The number of musk deer species is used as an indicator of the stability of ecosystem relations that are common for large arrays of pine forests. The naturally occurring reasons for the decline in the period of unstable dynamics of climatic factors are less significant (Zaitsev, 2006 ).
The Program for the Study and Conservation of the Far Eastern musk deer was implemented in 2010 on the territory of the Sikhote-Alin state reserve and its surrounding areas (Terneisky district of Primorsky Krai). Research has been extended from studies conducted in the 1940s (Salmin, 1972 ) and carried out in 1975–2009 (Zaitsev, 1975 , Zaitsev, 1991 and Zaitsev, 2006 , etc.). This research is a joint program of the reserve, the Pacific Institute of Geography FEB RAS and the A.N. Severtsov Institute of Ecology and Evolution RAS.
The aim of the research program is to study the versatile ecology of musk deer. The tasks include the survey of the distribution of animal habitats, population structure, behavior and adaptive abilities of the musk deer, ecosystem relationships in the current period, the characteristics and rates of reproduction and other aspects of the ecology of the species using complex methodology. The obtained knowledge will be used to promote the preservation of this unique species and its habitat as well as for the prevention of further population decline and habitat reduction and will likely serve as one of the foundations for change and development of forest legislation.
The objectives of this paper include the analysis of the comprehensive methodological approach for the research of animal ecology and characterization of the main research methods with a brief analysis of the obtained data, the understanding of which is important for musk deer population management in current and future periods.
Research was conducted in the northeastern part of Sikhote-Alin reserve (Tayozhnaya river basin). This study of musk deer ecology has applied various methods including radio tracking, winter tracking, the study of life activity traces, visual observations, video recording and data analysis of the automatic photo and video recorders.
Trapping with the purpose of radio labeling. In 2012–2014, with the purpose of radiolabeling live trapping of musk deer, two methods of capture were applied: (1) stationary trapping; and (2) remote tracking method with subsequent immobilization. For trapping, the previously used method of Prikhodko (2008) was applied with some modifications. In particular, two falling doors at both ends of the trap were installed and the side poles were replaced by a twine net (Fig. 1 ).
Trap for musk deer capturing.
A line was extended through the center of the trap and connected to the trigger mechanism on the side of the trap attached to the ropes from each door. Lichens were used as a scent. Feeding began a few months before the capturing stage.
The second method of trapping was carried out during the snow season with availability of traces. A group of two or three people followed one musk deer for 10–12 days. During this period of time, the animal adjusted to the human presence and was allowed to approach within a sufficient distance to proceed with immobilization using a pneumatic gun filled with syringe darts. The aforementioned method is based on the ability of musk deer to adjust to the human presence (Zaitsev, 1975 and Zaitsev and Zaitseva, 1980 ).
Immobilization of musk deer was performed with a remote rifle, model 4V.310 (manufactured Telinject, Germany) and an anesthetic mixture of Zoletil 100 at 2.5 mg/kg and Xylazine at 4 mg/kg. Antisedan, at a concentration of 0.4 mg/kg, was used as an antidote to Xylazine.
Captured animals were fitted with Telonics collars (USA) with radio transmitters that operated in the frequency range of 150–152 MHz. Radio tracking was performed using a directional, aerial and receiving set, tuned in to the individual transmitter frequencies. Radio tracking allowed for the collection of information about the location of the marked animals, nutrition data, daily activity, marking activities, selection of places of rest (bed), how to make visual observations of the animals and how to detect the distance of daily movements in the snow-free period of the year, etc. Locations of radiolabeled animals were detected by two methods: (1) following the direction of the signal until the visual detection of the animal; and (2) triangulation. The coordinate data obtained was registered in the musk deer finding forms that became the basis for the database creation in a Microsoft Excel spreadsheet. The information was then transferred into the databases of Geographic Information Systems software package ArcGIS 9.3 for further analysis.
In 2012–2013, the coordinates of 458 places of stay for four animals were obtained by telemetry. The musk deer habitat area was calculated by two methods: (1) Minimum Convex Polygon method (MCP), including 95% of locations; and (2) Kernel method with the inclusion of probable density distribution of locations at 95% (home range) and at 50% (core zone of the habitat area or center of activity) levels. The first method connects the most remote points of the external locations of animals, forming a convex polygon. This method is the most simple, most established and most commonly used method for studying musk deer habitat areas. It allows one to obtain results comparable with previous studies. The time period between each used location exceeded 3 h. Further, to calculate the size of home ranges, the program RStatisticalSoftware 3.0.2 was used.
Musk deer activity was determined according to one of the two modes of radio collars. In the passive state, the transmitter sent one signal per second. This mode corresponded to the moments when the animal was not moving (i.e., during resting periods). In the active state, the collar transmitted seven impulses per five seconds. The latter mode indicated that the musk deer were active (feeding, movement, etc.). The transition from active to passive mode was observed with a delay of 2 min and this allowed for brief stops not to be confused with the passive state.
While collecting data on daily activity, signal type (passive or active) was recorded every five minutes. In 2012, measurements of the activity of two radiolabeled musk deer (adult males and young females) were produced 23,541 times. If the observations were realized during the whole day without loss of signal, 288 activity measurements per night were received.
To characterize the degree of seasonal activity of the musk deer, the average activity percent was used. Because the number of measurements of activity in different time intervals was distributed unevenly, the average percentage of activity for each of the selected seasons was calculated as the ratio of activity percent sum of all time intervals to the number of these intervals, i.e., 24. The average percentage of activity reflects the portion of time during which the object was observed in the active state during a certain season. Furthermore, for each of the 24 hour intervals, the proportion of measurements with an active signal was calculated, and this allowed for the determination of the percentage of time during which the animal was active in each of the time intervals.
Large amounts of data on many aspects of musk deer ecology have been obtained because of the survey of life activity traces included during their tracking. The ecology issues of the musk deer studied by this method include feeding, manifestation of different behavioral activity in space, choice of places for rest, detours and patrolling of the habitat area, marking territory, etc. Tracking of the musk deer that are familiar to the observers provides data on the movement of individuals and population structure (Zaitsev and Zaitseva, 1980 and Zaitsev, 1991 , etc.). Both marked musk deer and animals without radio collars were tracked by the researchers. All traces of life activity (beds, objects of territory marking, piles of excrement, urinary point, etc.) were observed. During the tracking period, herbarium was collected for further identification of the species of plants grazed by musk deer.
Labeling objects encountered during tracking (marks of the excreta of caudal glands complex, excrements) were described and periodically thereafter were inspected to identify the frequency of the repeated visits and marking. The location of the marking object was described with geographical coordinates using a GPS receiver and photographed to indicate the object type (fallen branch, bush, etc.), diameter, height from the substrate of the mark and presence of “scrapes” on the ground or snow. Additionally, the features of the terrain and location (trail ridge, slope exposure, slope steepness, etc.) of the marked object were recorded. Regarding piles of excrement and toilets that also have a communicative significance, the location (slope, ridge, valley, etc.), diameter of the toilet and time of use were described.
Descriptions of places of rest of musk deer (beds) were performed during the tracking period in addition to during accidental discoveries of the animals. The descriptions of beds were documented using coordinates of the location, substrate (snow, soil, litter), characteristics of the location (topography, distance from trails, etc.) and parameters of the bed, including its protective properties. If possible, the length of stay of the animal and the nature of its use (disposable or reusable) were also identified.
The method of prolonged visual observations of musk deer developed by Zaitsev (1975) is based on the ability of musk deer to quickly adjust to the human presence. The behavior of such “domesticated” animals was observed from a distance of 5–25 m. This method was applied with the purpose of obtaining the information about the physical condition of the animals as well as their molting, nutrition, marking activities, time spent in a particular location, length of the diurnal course during the snowless period of the year and other aspects of the ecology of the musk deer. Recordings of the duration and characteristics of different types of animal activity were made by visual observation, including feeding, movements, resting and marking behaviors. Observation of feeding allowed for the preferred types of grazed plants to be determined. This method provided the most reliable data when used in conjunction with radio telemetry and study of life activity traces.
Devices were installed in places frequently visited by musk deer: the permanent beds, marking objects, musk deer latrines and on the trails. This method allows the frequency of visits of beds and duration of resting in one place to be determined in order to observe the physical condition and molting of the animals. Video recorders installed on trails and marking objects record the process of territory marking and interaction between the individuals in the group.
Quantifying musk deer is complicated (Zaitsev, 2006 ). In this regard, it is considered that there are inadequate accounting methods for quantifying the animals (Humes, 2004 ). The method of winter route accounting that is widely used in Russia lacks accuracy in determining the number of musk deer. Other methods, including those recommended by the regulatory bodies of the hunting sector, are difficult with respect to the broad introduction into the accounting system on the territory of Russia (Bersenev et al., 2011 ).
To determine the density of the population of musk deer on the stationary study sites in Sikhote-Alin, a method of accounting of the musk deer number using feces in the snowless season was employed. The latter method is based on the ratio of the standard accounting encountered piles and latrines in the spring on the routes mapped randomly in the habitat of the musk deer as well as the musk deer population density, calculated during the winter in key areas (Zaitsev et al., 2013 ). This method of accounting a certain retrospective population density, existing from autumn to spring, instead of a momentary musk deer density, is often a limitation of other methods.
In efforts to identify the habitat preferences of the musk deer and the factors affecting the distribution and dissemination of individuals, a joint study in association with the World Wilde Fund of Nature (WWF) was conducted in 2012–2013. This study was based on the detection of the frequency of occurrence of musk deer tracks in different types of habitats. The tracks registration was performed in conjunction with data collection on various environmental factors (abundance of feed, the character of the vegetation, snow conditions, forest exploitation regime, distance from roads, etc.). Analysis of the relationship of the habitat properties with the presence of the musk deer, held under the MacKenzieand Roylescheme (MacKenzie and Royle, 2005 ), has revealed a combination of factors that affects the distribution of musk deer in Sikhote-Alin (Slaght et al., 2012 ).
Data collection within the musk deer study project in Sikhote-Alin is still in progress, thus, the results presented are preliminary.
During the study herein, six musk deer were captured and marked with radio-collars (Table 1 ).
|# of individual||Date of capture||Sex||Age||Period of tracking|
Five deer were captured in traps and one deer was captured by stationary remote immobilization. Two musk deer (#1 and #4) were subsequently taken by predators.
The calculation of the wildlife habitat area required the use of data from 152 positions of musk deer #1, 14 positions of musk deer #2, 129 positions of musk deer #3 and 100 positions of musk deer #4 (Maksimova et al., 2014 ).
The area of the year-long habitat was as follows: (1) musk deer #1–0.75 km2 (MCP, 95%), 1.4 km2 (Kernel, 95%) with nucleus 0.2 km2 (Kernel, 50%); (Humes, 2004 ) musk deer #3–0.51 km2 (MCP, 95%), 0.82 km2 (Kernel, 95%), 0.1 km2 (Kernel, 50%); and (2) musk deer #4–1.2 km2 (MCP, 95%), 1.24 km2 (Kernel, 95%), and 0.18 km2 (Kernel, 50%). In regard to musk deer #2, the area is calculated only through April at 0.3 km2 (MCP, 95%) and 1.4 km2 (Kernel, 95%). The seasonal habitat area of musk deer #1 ranged from 0.94 km2 in spring to 1.32 km2 in winter; for musk deer #3, it ranged from 0.37 km2 in summer to 1.17 km2 in spring; and for musk deer #4, it ranged from 1.09 km2 in fall to 4.75 km2 in spring (Kernel, 95%).
Musk deer activity manifested in the alternation of phases of rest and various forms of behavior. Such rhythms are well synchronized, alternating the phases of the diurnal cycle, and are also coordinated with changing environmental factors (Zaitsev and Zaitseva, 1983 and Prikhodko, 2008 ). Application of radio telemetry has provided an opportunity to remotely assess the daily activity of animals in a natural setting without human disturbance.
In the spring, musk deer were active for 49.2% of the time. During the summer, they were active for 50.5% of the time. In the fall, they were active for 53.2% of the time. Thus, from spring to autumn, activity did not differ significantly.
Musk deer have a twilight-night lifestyle. According to the data, the curve of daily activity has two peaks in the morning and evening (Fig. 2 ). The evening peak is more intense.
The distribution of the daily activity of the musk deer in Sikhote-Alin reserve in spring, summer and autumn periods.
At this time, the activity reaches its maximum values: 81% in the spring (from 21 to 22 h); 85% in the summer (from 22 to 23 h) and 78% in the summer (from 21 to 22 h). Musk deer are less active between 9 and 19 h. During particular time intervals (from 9 to 11 in the spring and from 9 to 10 h in the fall), musk deer were active less than 30% of the time (Fig. 2 ). In general, the nature of the daily activity rhythm of musk deer in different seasons is very similar.
Musk deer are territorial animals (Zaitsev, 1991 and Prikhodko, 2003 ). Olfactory-optic communication has a particular significance in communication between the individuals of musk deer. Maintenance of the stability of individual plots is managed by a developed system of olfactory-mediated communication and marking of different objects by excreta of the specific skin glands.
Among the marking methods, excreta complex caudal glands have a particular importance in the regulation of relations between males of musk deer (Zaitsev, 1985 , Zaitsev, 2006 , Prikhodko, 2003 and Zaitsev et al., 2014 ) (Fig. 3 ).
Marking with excreta of tail gland.
On the stationary area in the Taiga river basin, when marking with the excreta of the caudal glands, males used fallen branches (in 37% of reported cases), the trunks of trees or shrubs (58%), as well as the stems of herbaceous plants (5%). The diameter of the marked objects ranged from 0.7 to 7.3 cm. The main sequence of activities performed by musk deer while leaving such marks was identified both by eye and by video recording as follows: (1) orientation when approaching the mark; (2) olfactory examination; and (3) turning the back of the body towards an object of marking. The male was observed to touch tightly the marking object with the tail and produce a motion with the rear part of the body from side to side (Fig. 3 ). During and after marking, the musk deer rakes snow or the ground towards the object, using a forelimb. Such “scrapes” facilitate search for old tags.
Photo and video traps filmed 16 defecation areas and two acts of marking with the tail gland. The process of marking with excrements starts with sniffing of the old litter (in areas of common toilets) and then the animal follows ahead and stands above the old toilet. At the same time, it spreads the hind legs, squats and defecates. The diameter of the described “toilets” commonly was 15–25 cm (in some cases up to 55 cm). They were usually located on the animal pathways (Fig. 4 ).
Toilets of musk deer exhibiting multiple use.
According to the results of the study on identification of the factors that affect the choice of habitat by musk deer and its distribution, it is demonstrated that the anthropogenic environmental factors have the greatest impact on musk deer. Thus, the low probability indicator of musk deer presence is mostly defined by a combination of two factors: intensive logging and proximity to roads (Slaght et al., 2012 ).
Market demand for musk deer derivatives has been maintained at a high level while the number of animals of this species has been decreasing (Humes, 2004 ). Existing threats for the population (Zaitsev, 2006 ) require specific measures, particularly referring to rational management and exploitation of musk deer habitats. To develop recommendations for the conservation of musk deer in the Sikhote-Alin region, scientific knowledge is needed.
Complex research on musk deer in Sikhote-Alin allows for a comprehensive insight into previously under-studied aspects of its environment as well as the degree of influence of natural and anthropogenic factors on population parameters such as abundance, distribution and survival. Using multiple research methods, valuable scientific information on such important issues as the use of habitat areas, distribution, feed, marking activities and daily activity was obtained.
This is the first time that use of territory by musk deer and their daily activities have been studied using radio telemetry in Russia. In general, the sizes of habitat areas calculated using this method are consistent with the results obtained using other methods of tracking and studying of musk deer ecology (Zaitsev, 1991 and Zaitsev, 2006 ).
Accumulation of new data and analysis of the results obtained within the current program have practical importance for musk deer population management in Sikhote-Alin and can be used in the development of recommendations for environmental protection in other regions.
For assistance in implementing the present program, the authors are grateful to the administration and staff of the Sikhote-Alin reserve (A. Astafiev, D. Gorshkov, E. Pimenova, S. Soutyrina, M. Gromyko, etc.). Field studies involved employees of the Representation office of Wildlife Conservation Society in Russia, Amur branch of WWF, and students of various higher educational institutions.
Are you one of the authors of this document?