The flight distance in free-ranging big game. The Journal ... [In: Proceedings of the VIIth International Congress of Game Biologists, Beograd-Ljubljana: 343â349.
Acta Theriologica 46 (2): 195–206, 2001. PL ISSN 0001–7051
Flight behaviour and observability in human-disturbed sika deer Jakub BORKOWSKI*
Borkowski J. 2001. Flight behaviour and observability in human-disturbed sika deer. Acta Theriologica 46: 195–206. Sika deer Cervus nippon Temminck, 1838 observability and flight behaviour were studied in an area with a high level of human disturbance (Tanzawa Mts, Japan). Deer observation rate was positively affected by habitat-related food conditions, while it was negatively correlated with the number of tourists in the study area. Flight frequency in April–September was lower than in October–February. It was also influenced by period of day, behaviour of investigator and deer group size. Group composition, deer activity and habitat condition had no effect on flight frequency. Thus, only 317 (48%) of the deer groups encountered were caused to flee and among them as many as 86% fled for a distance shorter than 40 m. It was concluded that deer in Tanzawa Mts learned to tolerate people, what is known for the populations which are unhunted or under low hunting pressure. Division of Agriculture and Agricultural Life Sciencies, University of Tokyo, Yayoi 1-1-1, Bunkyo-ku 113 Tokyo, Japan Key words: human disturbance, flight behaviour, sika deer, Japan
Introduction In wildlife management and conservation, the importance of the wildlife itself and its habitats is augmented by the fundamental role played by people. There is a variety of ways in which people may influence wildlife (Reed 1981, Meffe and Carroll 1994) and one of them is tourism. In general, behaviour of ungulates towards people reflects the behaviour of people towards ungulates (Geist 1971). People’s visits to a given area usually cause deer to flee (Schultz and Bailey 1978), while the harassment caused by large numbers of visitors can make deer seek a dense cover or even leave their home ranges (Jeppesen 1987). In turn, hunting is a factor modifying deer flight behaviour, with individuals in hunted populations usually tending to be more wary than those in unhunted ones (Behrend and Lubeck 1968). The flight behaviour of deer may also differ between the sexes (Behrend and Lubeck 1968).
* Present address: Wildlife Management Section, Forest Research Institute, Bitwy Warszawskiej 3, 00-973 Warsaw, Poland 
Understanding of the factors influencing flight behaviour is important for at least two reasons. Firstly, there are notable costs of locomotion, especially when deer are moving through snow and/or at high speed (Moen 1976). Thus, in winter, when food availability is limited, a high level of human disturbance may to some extent contribute to poor condition. Such effect may be enhanced by the fact that disturbed individuals are forced to leave preferred areas (with more abundant forage) and move to dense cover, which usually has little food. Secondly, such data are needed when deer are managed for viewing, as is often the case in national parks or other recreational areas. Knowledge of the impact of people on deer is especially important in highly-urbanized areas. Japan’s Tanzawa Mountains may provide a good example of such a place, because they are near Tokyo, which has an extremely high population density of humans. Mountain walking is very popular in Japan in some seasons of the year, but despite the fact that sika deer Cervus nippon Temminck, 1838 is the country’s most numerous large mammal, there are no data on the influence of tourists on its behaviour and observability. The general aim of the present study was therefore to answer the questions as to whether observability and flight behaviour of sika deer is affected by: (1) season and time of day, (2) habitat structure, (3) deer activity, (4) composition and size of groups, and (5) tourist numbers and behaviour.
Study area The study began in April 1994 and continued to February 1995, when a combination of a dense deer population and deep snow caused high mortality, reducing subsequent encounters with deer. The study area of ca 100 ha was in the Tanzawa Mountains (35°N; 139°E) of central Honshu, between Mt Tanzawa (1567 m a.s.l.) and Mt Ryugabamba (1500 m). The study area was small, but it was selected for two reasons. Firstly, this area was known to have one of the highest sika deer densities in the whole range of Tanzawa Mountains (K. Furubayashi, unpubl.), with 30–40 individuals present during the growing season, and more in late autumn and winter (maximum 120 in January 1995), when the high mountains offered rich food resources (Borkowski and Furubayashi 1998a). Secondly, it included the largest continuous patch of co-occurring Sasa hayatae and Pseudosasa purpurascens, two species of dwarf bamboo found to be closely linked to the ecology of sika deer in Tanzawa (Furubayashi 1996, Borkowski and Furubayashi 1998a). The size of my study area was considered sufficient, because radio-collared individuals had very small home ranges (between 11.2 and 20.2 ha) and did not roam out of the area (Borkowski and Furubayashi 1998b). A ridge in the study area has a tourist trail dividing it into western and eastern parts. The number of tourists depends on the season, being quite high on weekends and holidays from late September until June, but lower in the rest of the year. Hunting (of antlered deer only) is allowed at altitudes below 800 m between November 15 and February 15. No predators of sika deer are present, as wolves were exterminated by the end of the 19th century in Japan. Habitat classification was based on both aerial photos and ground inspections. Clearings occurred in small (0.2–1.2 ha) patches, and were dominated by Sasa hayatae. Closed woodland mostly covering the eastern part of the study area, had a dense overstory composed of numerous species of trees, the dominants being Weigela dekora, Rhododendron sp., Enkianthus campanulatus, Elaeagnus umbellata, Acer japonicum and Tilia japonica. Most of the trees were < 10 m in height, while the ground cover
Flight behaviour of sika deer
was dominated by Sasa hayatae, and to a much more limited extent by Pseudosasa purpurascens (which occupied about 25% of the total area of this habitat). Open woodland had an overstorey composed mainly of Fagus crenata trees about 20 m tall, with numerous gaps in the canopy reflecting natural mortality, while the understory was either very scarce or absent. The ground layer was dominated by Sasa hayatae. The northern part featured a habitat similar in its over- and understorey to the open woodland, but the ground vegetation level mostly comprised plant species unpalatable to deer. Because clearings were very small, there were no differences in visibility conditions between them and the closed woodland, while visibility in the open woodland was better than in the latter two habitats. Differences in visibility among habitats were not especially assessed, but these were rather obvious and the assessment was not considered necessary. Winters in the study area are mild, with average temperatures: –2.3, –4.6, –1.7, and 3.9°C in January, February, March and April, respectively (Yamane 1999). Daytime temperatures in winter are usually above 0°C. Average snow depths in 1994 were 30, 19 and 9 cm on easterly exposures and 14, 7 and 8 cm on westerly exposures, in February, March and April, respectively (Borkowski et al. 1996).
Methods A tourist trail of about 1.5 km between Mts Tanzawa and Ryugabamba was walked several days a month, four times in the day: soon after sunrise, before noon, in the afternoon and a short time before sunset. Exact times depended on day length, but consecutive walks were separated by at least two hours, usually three or more. The following data were recorded for walks: number of deer encountered – if possible with sex and age (juv., adult), deer activity, time and habitat type. Visual estimates (to nearest 5 m) were made of the flight distance, ie the distance to which a person can approach deer without causing it to flee (Altmann 1958) and, if possible, of the escape distance – defined as the distance that covered by fleeing deer before they stopped (as measured between the nearest individual before flight and the farthest after flight). This was done after a period of training involving the verification of estimated distances to a given point by pacing. Accuracy was found to be satisfactory after a few hundred trials. When possible, note was made of the behaviour of deer after they had stopped (standing alert and observing the investigator, walking away from the investigator, feeding). Deer not fleeing by the time an investigator was passing by them were not made to do so, but the minimum distance (“passing-by distance”) between the investigator and the nearest deer in a group was estimated. Individuals were considered members of the same group if distances between them were less than 50 m, but in most cases decisions were uncontroversial, as distances were noticeably shorter. To simulate different types of disturbance, the investigator (1) walked continuously along the trail (without stopping); (2) stopped on the trail when encountering deer and stood for 1–2 min without movement, or (3) walked off the trail. During every encounter only one type of disturbance was simulated, ie it never happened that, for instance, the observer first stood without movement and then walked off the trail, recording behaviour of the same group. Data were collected for each period of the day by walking in one direction only. The number of tourists encountered during every walk between October 1994 and February 1995 was recorded. Relation between tourist number and deer observability (number of observed individuals) was analysed by Spearman’s rank correlation. In statistical analysis of flight behaviour, data from both kinds of open woodland were combined. When analysing deer observability, in estimation of relative abundance of habitat types, visibility within habitats was considered. G-test for goodness of fit was applied in analyses of influence of habitat and period of day on deer observability, while flight frequency data were analysed using the G-test of independence. When number of comparisons is high (what was the case of this study) some differences may be significant just by chance. To avoid it, when analysing flight frequency, sequential Bonferroni procedure (Rice 1989) was applied. Thus, the result of a given comparison was considered significant if the probability obtained was smaller or equal to the probability indicated by Bonferroni sequential test (PB). Influence of group kind and group size on flight, escape and passing-by distances was analysed using two-way ANOVA.
Results Observations concerned 662 groups of deer (involving 2888 individuals, apparently observed repeatedly). The observability of deer in the various habitats differed from the expected (G = 1379.4, df = 3, p < 0.0001), and although the situation was modified by seasonal changes in habitat use (Borkowski and Furubayashi 1998a), sika deer were seen most often in clearings, while the open woodland with unpalatable plant species was a habitat in which very few deer were observed (Fig. 1). More deer were observed in the morning and evening than during the daytime in every season (spring: G = 9.2, df = 3, p = 0.03; summer: G = 51.5, df = 3, p < 0.0001; autumn: G = 53.2, df = 3, p < 0.0001; winter: G = 38.2, df = 3, p < 0.0001) (Fig. 2). In addition, more deer were active (feeding and walking) and fewer inactive (resting) in the morning and evening than during the daytime (G = 18.2, df = 1, p < 0.0001). Percent of active deer dropped from 82% in the morning and evening to 33% in the daytime. Deer observability was also negatively correlated with the number of tourists visiting the study area in all the periods of day: morning (rS = –0.72, p = 0.0001, n = 22), before noon (rS = –0.72, p = 0.0003, n = 15), afternoon (rS = –0.66, p = 0.003, n = 18) and evening (rS = –0.87, p < 0.0001, n = 13). Only 317 (48%) of 662 deer groups encountered were caused to flee, with 345 (52%) not doing so. Flight frequency was significantly lower (G = 7.8, df = 1, p = 0.005 < PB) between April and September than between October and February (Fig. 3). It was also dependent on time of day, with deer observed during the daytime seemingly more tolerant of people than those seen in the morning and evening (G = 10.2, df = 1, p = 0.001 < PB) (Fig. 3). Flight frequency was similar among habitats in both periods April–September (G = 2.4, df = 2, p > 0.05) and 1200
1000 800 600 400 200 0
H abit ats Fig. 1. Observed (filled bars) and expected on the basis of proportion of every habitat within study area (open bars) numbers of sika deer seen in different habitats. Cw – closed woodland, Ow – open woodland with dwarf bamboo, Ow1 – open woodland with unpalatable plant species, C – clearings.
Obs ervation frequency (%)
Flight behaviour of sika deer
100 80 Morning
Afternoon B efore noon
E vening 20 0
S pring (18)
S ummer (102)
W inter (281)
60 40 20 0
F light frequency (%)
F light frequency (%)
Fig. 2. Sika deer visibility in different periods of day (based only on those days when data were collected in all day periods). Number of observations are given in parenthesis.
60 40 20 0 Morning +evening
Fig. 3. Sika deer flight frequency in different seasons and periods of day. Filled bars – fleeing groups, open bars – non-fleeing groups.
October–February (G = 7.8, df = 2, 0.05 > p > PB). Activity did not influence flight frequency, with resting individuals fleeing as often as active ones (G = 1.7, df = 1, p > 0.05). The influence of group size on flight frequency was not clear. As rising deer densities in autumn and winter made group size larger than in spring and summer (Borkowski 1996), the two periods were analysed separately. In spring and summer, deer fled with similar frequency irrespective of group size (G = 0.3, df = 1, p > 0.05), while in autumn and winter, animals in large groups (of 6 and more individuals) fled more frequently than those in small ones (of 1–3) (G = 8.5, df = 1, p = 0.004 < PB), though frequencies did not differ between medium-sized (of 4–5 ind.) and large groups (G = 1.6, df = 1, p > 0.05) or between medium-sized and small ones (G = 0.6, df = 1, p > 0.05) (Fig. 4). Flight frequency was similar between deer associated in mixed groups, and those in both female (G = 6.1, df = 1, p = 0.01 > PB) and male groups (G = 4.3, df = 1, p = 0.03 > PB). Individuals in
April- S eptember 60 40 20 0 1-5
F light frequency (%)
F light frequency (%)
October-F ebruary 60 40 20 0 1-3
Group s ize
Group s ize
Fig. 4. Influence of sika deer group size on flight frequency. Filled bars – fleeing groups, open bars – non-fleeing groups. In April–September, small and middle – sized groups have been combined due to inadequate sample size.
F light frequency (%)
female groups fled with similar frequency to those in male groups (G = 0.04, df = 1, p > 0.05). The behaviour of the observer was another factor influencing flight frequency. Although when the observer was walking on the trail deer were caused to flee as frequently, as when he was walking off it (G = 5.1, df = 1, p = 0.02 > PB), deer fled much more often when the observer stopped, as opposed to walking continuously (G = 20.1, df = 1, p = 0.0001 < PB) (Fig. 5). All the 69 groups located less than 10 m from the trail were caused to flee when encountered, but any further increment in distance between deer and observer did not influence flight frequency (April– –September: G = 1.4, df = 2, p > 0.05; October–February: G = 2.1, df = 2, p > 0.05). Flight frequency depending on the distance between deer and observer was not modified by the habitat type (in all cases p > PB or p > 0.05). Only 8% of deer groups fled beyond the observer’s sight range, and as few as 6% fled more than 40 m; 42% fled less than 20 m and 44% groups fled between 20 and 40 m. As much as 92% of deer stood alert and observed the investigator after flight, while 4% of them walked away from the investigator and 4% began to feed immediately after
80 60 40 20 0
, Obs erver s behaviour
Fig. 5. Influence of observer’s behaviour on sika deer flight frequency. Filled bars – fleeing groups, open bars – non-fleeing groups.
Flight behaviour of sika deer
stopping. Among those groups which did not flee, 83% were observing the investigator and so knew of his presence, while the other 17% seemed unaware and continued feeding or resting. None of the studied distances (flight, escape, passing-by) was influenced neither by size nor by kind of deer groups (two-way ANOVA: F < 1.7, p > 0.05 in all cases). Although interesting, most of theoretically possible interactions could not be analyzed because of inadequate sample sizes (for instance of deer observed inactive, or in daytime). Similarly, there were relatively few records made when investigator walked off the trail or stopped. Therefore, the analyses have been often done for pooled datasets. Discussion The rates at which sika deer in the Tanzawa Mountains were observed varied among habitats, with food availability seemingly responsible for this, rather than habitat structure (influencing visibility conditions). While vegetation conditions limit visibility, food biomass is positively correlated with habitat openness (Arnold 1950, Pace 1958, McConnell and Smith 1971) so it is usually difficult to assess the extent to which deer observability depends on food, opposed to visibility conditions (Sage et al. 1983). In this study (in which visibility conditions in closed woodland and clearings did not differ, but were in both cases poorer than in open woodland), the overriding importance of vegetative conditions would be indicated if observation rate in the closed woodland had been similar to that in the clearings, and lower in both of them than in the open woodland. In fact, more deer than expected were seen in the clearings, while observability rate in the closed woodland was according to expectation, as was that in the open woodland with Sasa hayatae in the understorey. In the Tanzawa Mts, food biomass is much higher in clearings than in any other habitat (Borkowski and Furubayashi 1998a). Furthermore, visibility conditions in the open woodland with S. hayatae in the understorey were similar to those in the open woodland with an undestorey dominated by unpalatable plant species, but very few deer were observed in the latter as compared to the former. Deer were mostly active in the morning and evening, and reduced their activity during the daytime. Consequently, fewer deer were observed during the day, when they rested in the closed woodland (Borkowski and Furubayashi 1998b) than at dawn and dusk. Two distinct peaks of activity are very common among cervids (eg Geist 1963, Craighead et al. 1973, Jackson 1977, Cederlund 1981, Georgii 1981, Eberhardt et al. 1984). Such behaviour is often characteristic of populations disturbed by human activity, and migratory populations have even been found to exhibit diurnal activity in areas relatively unaffected by people, and crepuscular activity in areas with greater disturbance (Georgii 1981, Green and Bear 1990). Similarly, diurnal activity has been reported for undisturbed red deer Cervus elaphus (Bubenik and Bubenikowa 1967), while chamois Rupicapra rupicapra and red deer studied by Douglas (1971) recommenced diurnal feeding two years after
the cessation of hunting. On Nozaki Island (southwestern Japan), sika deer which were not disturbed by people fed in openings in the early afternoon (Endo 1992). In the study area, deer activity seemed to be influenced by large numbers of people present in autumn and winter, but remained largely confined to sunrise and sunset in summer too, despite low levels of human disturbance. It may be the heat of the summer, which forces deer to reduce their activity during the daytime, as radio-collared individuals in Tanzawa were then found to stay under the canopy of closed woodland (Borkowski 1996), thereby presumably minimised their costs of thermoregulation. In areas, where diurnal summer temperatures are high, deer select bed sites with shade from trees and thus minimise heat gain from direct solar radiation (as in the case of white-tailed deer Odocoileus virginianus, Ockenfels and Brooks 1994). Thus, depending on season sika deer reduce mid-day activity because of human disturbance or high temperatures. Sika deer at high altitude in the Tanzawa Mts seemed very accustomed to the presence of people. Although no comparative study is available, deer in the lower mountains, where hunting of antlered males is allowed, were much more wary than in the high mountains (J. Borkowski pers. obs., K. Furubayashi, pers. comm.). In the high mountains, only about half of observed groups took flight, and most of these fled less than 40 m before standing alert and looking towards the observer. In most studies cited throughout this paper, all the encountered individuals took flight and thus only flight distance (not flight frequency) was used as a measure of deer wariness. The acceptance of human presence found in the present study accords with what is known of behaviour in unhunted deer (Schultz and Bailey 1978). Behrend and Lubeck (1968) found the that flight distance was significantly longer for hunted deer than for unhunted ones. Altman (1958: 208) stated that “The beginning of hunting season cuts into the rutting situation like a catastrophic storm. The flight distance is suddenly stepped up”. In the present study, the flight frequency for October–February was significantly higher than that for April– –September. Although this may in part result from the higher number of tourists in autumn-winter than in other seasons, another factor seems important. In October, deer in Tanzawa begin to concentrate in the high mountains to feed on the abundant resources of dwarf bamboo (Borkowski and Furubayashi 1998a). They come from lower areas where, as mentioned above, hunting of males is allowed, and their greater wariness probably tends to increase average flight frequency. Although generally tolerant of humans, the deer studied tried to avoid frequent encounters with them. In every period of day deer observability was negatively correlated with the number of tourists. It is probable that animals were only able to tolerate a given level of disturbance, and that above some threshold they fled and sought cover further from the trail. In contrast, Schultz and Bailey (1978) found no significant evidence of an influence of the intensity of human disturbance (eg volume of traffic) on elk behaviour. Deer observed during the daytime were more tolerant towards people, and less ready to flee, than those seen in the morning and evening. This may reflect individual differences in the acceptance of people among
Flight behaviour of sika deer
the studied deer. An alternative explanation is that resting deer (like the majority of the animals seen during the daytime in the present study) are less ready to flee than active ones, because they try to hide from an approaching person (in that case a hunter) by maintaining their position until the last possible moment (Grau and Grau 1980). However, in Tanzawa resting and active sika deer did not differ in flight frequency, what seems to confirm the first explanation. In theory, deer might feel safer in habitats with better cover. However, habitat condition in this study did not influence sika deer flight behaviour. This fact may come from low wariness of studied deer which are tolerant towards people and live without predators. However, in the case of white-tailed deer living under predator pressure, LaGory (1987) similarly found no direct difference in the flight behaviour between open and forested habitat. There was some, though not very clear, trend for deer in large groups to flee more frequently than those in small ones. This contrasted with finding of Altman (1958), that solitary individuals were more wary than animals in groups. In turn, LaGory (1987) who found that larger groups are more likely to flee, proposed that the deer in them were more likely to spot a predator. However, in the present study, a hypothesis that better perception ability explains the greater flight frequency of larger groups may be rejected, because 83% of deer which did not flee knew of the presence of the observer. Deer in Tanzawa formed larger groups in the morning and evening than during the daytime, and morning and evening flight frequencies were higher than those during daytime hours. Thus, the higher flight frequency among larger groups may to some extent be an artifact. Flight frequency of sika deer was not influenced by group composition. Behrend and Lubeck (1968) showed that male white-tailed deer had longer flight distances than antlerless individuals in an area where males were hunted. However, this seemed to be a result of hunting, rather than sex-related differences in behaviour, since antlered and antlerless deer in an area not hunted did not differ in flight distances, with this for males in unhunted area being shorter than in hunted one (Behrend and Lubeck 1968). In contrast, red deer females are believed to be more wary than males (Mitchell et al. 1977). The results of this study confirm what is known of the reaction of wildlife to modes of disturbance. Animals usually accept certain modes of disturbance more willingly than others, for instance they are more tolerant of vehicles than of people on foot (Kucera 1976, Schultz and Bailey 1978). Human activity becomes tolerated if it is regular and predictable. Thus, sika deer in Tanzawa were most tolerant of continuous walking on the trail, ie probably the most common tourist activity. Flight frequency was significantly higher when the observer stopped and stood without movement, presumably because deer were less familiar with such behaviour from tourists. Similarly, MacArthur et al. (1982) found that mountain sheep Ovis canadensis canadensis reacted more strongly to a person walking with a dog than without it, as well as more strongly to a person approaching from over a ridge than to one coming directly from the road. Thus, although animal response to human disturbance may depend on its intensity (eg number of tourists), it will be
largely modified by the mode of disturbance (eg tourist behaviour). This should be remembered wherever disturbance is detrimental to wildlife and therefore needs to be managed. Since deer in this study were very tolerant of human activity and only half of them fled, flight frequency was a better measure of sensitivity to disturbance than flight distance used most often in other studies. In fact, Altman (1958) recommended treating flight distance with caution. Indeed, it can lead to wrong conclusions. For instance, a problem may appear when group-size differences in flight distance are analysed. As group size often differs among habitats (Hirth 1977, LaGory 1986, Borkowski and Furubayashi 1998c), differences in flight distance for groups of various sizes may origin from dissimilarities in the sight distance in different habitats. Management implications Sika deer in the high Tanzawa Mts have learned to accept people. The fact that only half of deer groups fled (and most for a very short distance) suggests that there is at present no risk that high costs of locomotion due to human disturbance will impoverish the condition of the animals. In theory, in the places where deer pressure on vegetation is severe, canalising movement of tourists into the trails crossing patches of vegetation especially vulnerable to animal pressure could discourage deer’s use of them and hence reduce their impact. However, if deer are tolerant of people this approach seems impractical. Acknowledgements: I am grateful to Prof K. Furuta of Tokyo University, Dr K. Furubayashi of Tokyo University of Agriculture and Technology and Dr M. Yamane of Kanagawa Forest Research Institute for their help and suggestions as well as to the members of Tanzawa Nature Conservation Association for their logistic support.
References Altman M. 1958. The flight distance in free-ranging big game. The Journal of Wildlife Management 22: 207–209. Arnold J. E. 1950. Changes in ponderosa pine bunchgrass ranges in northern Oregon resulting from pine regeneration and grazing. Journal of Forestry 48: 118–126. Behrend D. F. and Lubeck R. A. 1968. Summer flight behavior of white-tailed deer in two Adirondack forests. The Journal of Wildlife Management 32: 615–618. Borkowski J. 1996. The ecology of sika deer in relation to their habitat at the high altitude of Tanzawa Mts. Ph D thesis, University of Tokyo, Tokyo: 1–105. Borkowski J. and Furubayashi K. 1998a. Seasonal changes in number and habitat use of foraging sika deer at the high altitude of Tanzawa Mountains, Japan. Acta Theriologica 43: 95–106. Borkowski J. and Furubayashi K. 1998b. Home range size and habitat use in radio-collared female sika deer at high altitudes in the Tanzawa Mountains, Japan. Annales Zoologici Fennici 35: 181–186. Borkowski J. and Furubayashi K. 1998c. Seasonal and diel variation in group size among Japanese sika deer in different habitats. Journal of Zoology, London 245: 29–34.
Flight behaviour of sika deer
Borkowski J., Furubayashi K. and Habuto H. 1996. Effect of snow cover on the distribution of non-migratory sika deer on the top of Mt. Tanzawa, Japan. Journal of Forest Research 1: 9–11. Bubenik A. B. and Bubenikowa L. M. 1967. 24-h periodicity in red deer. [In: Proceedings of the VIIth International Congress of Game Biologists, Beograd-Ljubljana: 343–349. Cederlund G. 1981. Daily and seasonal activity pattern of roe deer in a boreal habitat. Viltrevy 11: 315–353. Craighead J. J., Craighead F. C., Ruff R. L. and O’Gara B. W. 1973. Home ranges and activity patterns of nonmigratory elk of the Madison Drainage herd as determined by biotelemetry. Wildlife Monographs 33: 1–50. Douglas M. J. W. 1971. Behavior responses of red deer and chamois to cessation of hunting. New Zealand Journal of Science 14: 507–518. Eberhardt L. E., Hanson E. E. and Cadwell L. L. 1984. Movement and activity patterns of mule deer in the sagebrush-steppe region. Journal of Mammalogy 65: 404–409. Endo A. 1992. Spatial distribution of rutting male relevant to female home range in the sika deer (Cervus nippon) in the Nozaki Island. M Sc thesis, Kyusyu University, Fukuoka: 1–49. Furubayashi K. 1996. [The studies on protection of sika deer in the Tanzawa Mountains]. Ph D thesis, University of Kyoto, Kyoto: 1–186. [In Japanese] Geist V. 1963. On the behaviour of the north American moose (Alces alces andersoni Pererson 1950) in British Columbia. Behaviour 20: 377–416. Geist V. 1971. A behavioural approach to the management of wild ungulates. [In: The scientific management of animal and plant communities for conservation. 11th Symposium of British Ecological Society. E. Duffey and A. S. Watt, eds]. Blackwell Scientific Publications, Oxford: 413–424. Georgii B. 1981. Activity patterns of female red deer (Cervus elaphus) in the Alps. Oecologia 49: 127–136. Grau G. A. and Grau B. L. 1980. Effects of hunting on hunter effort and white-tailed deer behaviour. Ohio Journal of Science 80: 150–156. Green R. A. and Bear G. D. 1990. Seasonal cycles and daily activity patterns of Rocky Mountain elk. The Journal of Wildlife Management 54: 272–279. Hirth D. H. 1977. Social behaviour of white-tailed deer in relation to habitat. Wildlife Monographs 53: 1–55. Jackson J. 1977. When do fallow deer feed. Deer 4: 215–218. Jeppesen J. L. 1987. Impact of human disturbance on home range, movements and activity of red deer (Cervus elaphus) in a Danish environment. Danish Review of Game Biology 13(2): 1–38. Kucera E. 1976. Deer flushing distance as related to observer’s mode of travel. Wildlife Society Bulletin 4: 128–129. LaGory K. E. 1986. Habitat, group size, and the behaviour of white-tailed deer. Behaviour 98: 168–177. LaGory K. E. 1987. The influence of habitat and group characteristics on the alarm and flight response of white-tailed deer. Animal Behaviour 35: 20–25. MacArthur R. A., Geist V. and Johnson R. H. 1982. Cardiac and behavioural responses of mountain sheep to human disturbance. The Journal of Wildlife Management 46: 351–358. McConnell B. R. and Smith J. G. 1971. Response of understory vegetation to ponderosa pine thinning in eastern Washington. Journal of Range Management 23: 203–212. Meffe G. K. and Carroll C. R. 1994. Principles of conservation biology. Sinauer Associates, Sunderland: 1–600. Mitchell B. B., Staines W. and Welch D. 1977. Ecology of red deer. Institute of Terrestrial Ecology, Banchory: 1–74. Moen A. N. 1976. Heart rates of white-tailed deer in winter. Ecology 57: 192–198. Ockenfels R. A. and Brooks D. E. 1994. Summer diurnal bed sites of Coues white-tailed deer. The Journal of Wildlife Management 58: 70–75. Pace C. P. 1958. Herbage production and composition under immature ponderosa pine stands in the Black Hills. Journal of Range Management 11: 238–243.
Reed D. F. 1981. Conflicts with civilization. [In: Mule and black-tailed deer of North America. O. C. Wallmo, ed]. University of Nebraska Press, Lincoln: 509–535. Rice W. R. 1989. Analyzing tables of statistical tests. Evolution 43: 223–225. Sage R. W., Tierson W. C., Mattfeld G. F. and Behrend D. F. 1983. White-tailed deer visibility and behaviour along forest roads. The Journal of Widlife Management 47: 940–951. Schultz R. D. and Bailey J. A. 1978. Responses of national park elk to human activity. The Journal of Wildlife Management 42: 91–100. Yamane M. 1999. [A study on nutritional ecology of sika deer in the eastern Tanazawa Mountains, Japan]. Ph D thesis, Tokyo University of Agriculture and Technology: 1–120. [In Japanese] Received 23 March 2000, accepted 8 December 2000.