The water deer (Hydropotes inermis) is one of the most primitive extant deer of the family Cervidae. Males of these animals have large canine teeth, and neither males nor females have antlers (Cooke and Farrell 1998). Two subspecies of the water deer currently exist: the Chinese subspecies (H. inermis inermis Swinhoe, 1870), and the Korean subspecies (H. inermis argyropus Heude, 1884) (see Geist 1998). Some of the Chinese subspecies were introduced into England and France in the late nine-teenth century (Cooke and Farrell 1998, Geist 1998). The Chinese subspecies has been seriously restricted in dis-tribution and population size in China (Wang 1998). In contrast, the Korean subspecies has mostly remained in Korea (Won and Smith 1999). However, a clear survey has not yet been conducted. The Chinese subspecies have recently been continuously studied (Xu et al. 1996, Wang 1998, Zhu et al. 2004, Hu et al. 2006, 2007, Koh et al. 2009), but the ecology and genetics of Korean water deer have rarely been studied (Lee 2003, Kim 2007, Koh et al. 2009). Water deer have preferred Asteraceas family followed by Fagaceae and Polygonaceae (Kim et al. 2011). Diets in the lowland and mountainous areas are different. We need to estimate the home range of wildlife in general to better understand habitat use patterns. In Korea, other wildlife species such raccoon dogs (Nyctereutes procyonoides) (see Choi and Park 2006, Kim et al. 2008), wild boars (Sus scrofa) (see Choi et al. 2006), and Asiatic black bear (Ursus thibetanus) (see Yang et al. 2008) have been studied, but only limited amounts of data were gathered.

An individual’s home range, the area used by an ani-mal in its normal activities of food collection, mating, and raising its young (Burt 1943), is a fundamental piece of its ecology. To estimate home range size, both mini-mum convex polygons (MCP) and kernel (K) methods have been commonly used because they are easily com-pared among studies (Harris et al. 1990, White and Garrott 1990). The K method usually allows for the determination of the center of activity (Worton 1989, 1995, Seaman and Rowell 1996).

Home range sizes of the water deer have been reported in a few studies (Cooke and Farrell 1998). In China, mean seasonal home ranges varied from 18 (1.8 km²) to 46 ha (4.6 km²) in a population. Females established feeding sites and set up small territories of about half a hectare. In England, the mean annual home range size was 21 ha (2.1 km²) at Whipsnade. However, home range studies have not yet been conducted in Korea. Therefore, such studies are essential to understanding the ecology of the Korean water deer. The aim of the present study was to estimate home range sizes. Sex, time (6 AM-6 PM, day and night), and season were considered in the home range compari-son.

Two different study sites were targeted in the present study (Fig.1 ): one is Daebu Island, Ansan of Gyeonggi Province (126°34？30？-126°39？0？ E, 37°15？0？-37°16？30？ N), and the other is a rural area of Chuncheon, Gangwon Province (127°46？04？-127°45？23？ E, 37°48？28？-37°47？50？ N) in the central parts of the Korean peninsula. In the

former site, two different types of habitats were found, including a reclaimed area (lowland) and a mountainous area (highland). Dominant species in the reclaimed area included halophytes (e.g., Salicornia europaea and Suae-da japonica) and reeds (e.g., Miscanthus sinensis, Phrag-mites communis, and Arundinella hirta). In contrast, shrubs and trees were dominant in mountainous areas. A variety of herbs were commonly found in both areas. In general, the mountainous area consisted of mixed for-ests dominated by oaks (Quercus spp.) and pines (Pinus spp.). Various types of herbs were commonly found in the mountainous area.

In Ansan, a total of five individuals were captured using an anaesthetizing gun (N = 2, succicholine injection 200 mg) or rescued from injury (N = 3). Among them, the three individuals rescued from injury were banded on the neck and released into the Shihwa reclaimed land. Despite the injury, no movement problems were observed. Only one female individual could be radio-tracked, and the other two individuals died within one month after release. Two individuals captured using the anaesthetizing gun were killed before release. In a rural area of Chuncheon, two males and one female were rescued from injury (N = 3) by the wildlife medicine center of Gangwon National Univer-sity. The three individuals were banded on the neck and released into their original habitats. Two males and one female could be tracked after release. All of the released individuals were handled after anaesthetizing. The ani-mals from Ansan were sexed, weighed, aged, and banded before release. One female was continuously tracked for two sequential days with about 1-h gaps using a radio col-lar (July and November, 2009), and two males and one fe-male were intensively tracked for 2-15 days with about 1-h gaps using a GPS collar (August and September, 2009). We could not detect the radio-tracked female after December 2009 due to signal loss. No signal could be detected on the GPS-tracked male after 15 days due to battery consump-tion. One male and one female were killed by passing traf-fic after 2-3 days. The ground homing method was used to radio-track the female (Mech and Barber 2002, Kim et al. 2008). The radio tracking was generally done from a truck using a VHF antenna (Wildsystem, Gurye, Korea) and a R20 radio receiver (ICOM Inc., Tokyo, Japan). We followed the strongest signal to close in on the individual and tried to reduce bias by tracking as carefully as possible using radio tracking (White and Garrott 1990). For GPS track-ing, which provides scientists with the ability to track an animal’s position and status in real time, we collected GPS coordinates from our homes and offices without any field work. The GPS coordinate data were obtained using a cell phone via Short Message Service (SMS) during the survey periods.

ArcGIS9.3 computer software (ESRI Inc., San Diego, CA, USA) was used for the home range analyses. All data col-lected were analyzed to estimate the home range of the Korean water deer using the Home Range Tools (HRT) for ArcGIS9.x. For topographic maps, 1:5,000 maps from the National Geographic Information Institute (NGI, Seoul, Korea) were used. Both the 95% fixed MCP and the 50% fixed K methods were used to calculate home range sizes of the Korean water deer (Worton 1989, Harris et al. 1990, White and Garrott 1990, Worton 1995, Seaman and Rowell 1996). Tracking data of the four individuals were used to estimate home range sizes based on sex, time, or season.

In this study, we analyzed home range sizes of four individuals (the number of fixes for subject 2, 6, 7, and 8 were 96, 217, 42, and 15 GPS points, respectively) of the Korean water deer. The mean home range size of the four individuals was 2.77 km² by 95% MCP and 0.34 km² by 50% K (Table 1). There seemed to be a clear difference of home range size between the male (3.30 km²) and female (2.25 km²) using the MCP method but little difference of home range size between the male (0.23 km²) and fe-male (0.45 km²) using the K method (Table 2). For the 95% MCP method, we found a meaningful difference in home range size between day (6 AM-6 PM, 1.99 km²) and night (2.43km²) (Table 2). Also, a great difference in home range size of the Korean water deer was observed between summer (June-August, 4.65 km²) and spring (March-May, 0.48 km²) or fall (September-November, 0.85 km²) (Table 2). For the 50% K method, little difference in home range size was found between day (0.32 km²) and night (0.26 km²) or between summer (0.57 km²) and spring (0.14 km²) or fall (0.21 km²) (Table 2).

The home range size of the Korean water deer has been unavailable until now. All of the current studies used a radio collar to track each animal. GPS tracking using a cell phone collar has recently been used to track wild mammals worldwide. However, such a method has not yet been used in Korea. Therefore, it is necessary to ap-ply such a method using a GPS collar and a code division multiple access (CDMA) system for Korean wildlife track-ing. Here we mentioned our results using both radio and GPS tracking.

Based on our results, the Korean water deer showed a mean home range size similar to that of the Chinese wa-ter deer in China and England. In the present study, mean home range size (N = 4) was approximately 2.8 km² for 95% MCP during the study periods. There have been few studies on home range size of the Chinese water deer. At Whipsnade in England, the mean annual home range size was 21 ha (2.1 km²; 1.2-2.5 km²) (Cooke and Farrell 1998). In China, it was also reported that mean home range sizes of a population varied between 18 (1.8 km²) and 46 ha (4.6 km²) (Cooke and Farrell 1998). Based on the information about home range sizes provided above, those of the Korean water deer (0.2-6.4 km²) seem to be similar to those of the water deer in England (1.2-2.5 km²) and China (1.8-4.6 km²).

In the wild, feeding activity of nocturnal mammals tends to be higher at night than during the day. For exam-ple, raccoon dog feeding areas differed largely between daytime (0.01 km²) and nighttime (0.35 km² for 95% MCP) hours (Kim et al. 2008). The Korean water deer in the pres-ent study showed the same pattern. However, the Korean water deer moved somewhat more to feed during the daytime (1.90 km²), even though it moved much move ac-tively at night (2.43 km² for 95% MCP). In fact, it seemed to spend its night-time hours mainly feeding by ruminating, whereas it spent its day-time hours both ruminating and resting. Based on our observations, the Korean water deer tended to move actively after dawn (500-700) and before

sunset (1,700-1,900). The water deer usually moved to its feeding site(s) or its resting site(s) during those hours. Cooke and Farrell (1998) reported that Chinese water deer spend about half of their daytime hours feeding. The water deer also fed at night, but details of quantity have not yet been determined (Cooke and Farrell 1998). The feeding peaks occurred in the early morning and in the evening as seen in the present study.

To investigate the seasonal differences in home range sizes, we analyzed home range sizes using both radio tracking (N = 1) and GPS tracking (N = 1) in summer, using GPS tracking (N = 2) in spring, and using radio tracking (N = 1) in fall (Table 2). We found a large difference in home range sizes between summer and spring or fall. Home range sizes were much more extensive in summer (4.65 km²) than in spring (0.48 km²) or fall (0.85 km² for 95% MCP) in our tracking. Such seasonal variations of home range size have not yet been analyzed in detail. Cooke and Farrell (1998), however, reported that mean seasonal home ranges were found to vary between 10 ha (1 km²) and 46 ha (4.6 km²) depending on season. Presumably, water deer use larger areas and actively move to feed on plants under good conditions during the summer and use smaller areas and move in a more restricted manner feed on plants under relatively poor habitat conditions during the spring, fall, and winter.

Finally, we analyzed the differences in home range size between females and males (Table 2). Based on our study, the female (2.25 km² for 95% MCP) showed a smaller home range than the male (3.30 km² for 95% MCP). Cooke and Farrell (1998) mentioned that home range size was largest for older females (about 1 km²) and second-year deer (about 1.5 km²) from late spring (May) to mid-sum-mer (July), whereas home range size was < 0.5 km² for older males (> 2 years). Our results showed that the mean home range size of adult males (≥ 2 years) was larger than that of adult females (≥ 2 years), even though it is difficult to compare with previous results.

In conclusion, we first estimated home range size of the water deer using radio and GPS tracking in Korea. The Korean water deer appeared to have a home range of 0.2-6.4 km², which is similar to that of the Chinese water deer in China and England. There were variations between day, night, seasons, and sexes seen in the present study. However, further extensive studies should be conducted to evaluate detailed home range sizes due to our limited data (e.g., the small number of individuals tracked and no tracking in the winter). We suspect that GPS tracking based on the CDMA system could be a very useful tool for understanding the ecology of the water deer using the ra-dio tracking method.