J. Mamm. Soc. Japan 18(2) : 87-98 December 1993 Size and Distribution of Home Ranges of the Japanese Shrew-mole Urotrichus talpoides Nobuo ISHII Japan Wildlife Research Center, Bunkyo - ku, Tokyo 113, Japan (Accepted 23 September 1993) Abstract. Size and spatial distribution of home ranges of the Japanese shrew- mole Urotrichus talpoides were studied by live-trapping in a mature mixed (broadleaved and coniferous) forest in Chiba Prefecture, Honshu. During the non-breeding season, no significant differences in home range size were found between sexes. The home ranges of breeding males were significantly larger than those of non-breeding males, while no significant differences were found in range size between non-breeding and breeding females. During the non- breeding season, home ranges were mutually exclusive among individuals of the same sex, but overlapped extensively with those of the opposite sex. In the breeding season, male ranges largely overlapped each other. Female ranges also overlapped each other, but to a lesser extent. The tolerance between the sexes during the non-breeding season is unusual in comparison with other mole and shrew species, whose individual ranges are exclusive to each other irre- spective of sex. Key words : Japanese shrew-mole ; Urotrichus talpoides ; home range size ; spatial organization. The shrew-moles Urotrichus, Neurotrichus and Uropsilus, though talpids (Fam- ily Talpidae), are less well adapted to the fossorial way of life than are the typical moles. The life-style of the shrew-moles is, therefore, of interest from the view point of comparative ecology, however, detailed information on the population ecology of shrew-moles has not previously been available. In this paper seasonal changes in the size and distribution pattern of the home ranges of the Japanese shrew-mole Urotrichus talpoides are described. Study Area and Methods Study area The study was carried out in Compartment 24 of the Tokyo University Forest located on the Boso Peninsula, Chiba Prefecture, Honshu. A live-trap- ping plot, consisting of 88 trapping stations set ten meters apart in an 8 x 11 array encompassing 0.7 ha, was established in a mature mixed forest of 51.6 ha surrounded by young broadleaved forests and conifer plantations. The domi- nant trees were Abies firma, Quercus acuta, Q. glauca, Q. salicina, and Castanop- sis cuspidata var. sieboldii. The understory was composed mainly of Eurya 88 Ishii japonica, Camellia japonica, Cinnamomum japonicum, Dendropanax trifdus, and Illicium religiosum. The undergrowth consisted mainly of Trachelospermum asiaticum, Maesa japonica, Arachniodes sporadosora and Gleichenia japonica. Trapping procedure One Sherman live-trap (SFA type) baited with a mixture of chopped cheese and oatmeal was placed at one opening of a burrow system near each station. Live-trapping was conducted once a month from June 1977 until December 1978, and then biweekly between January and July 1979. Trapping was operat- ed for five consecutive nights monthly until December 1978, and then for three to five nights in each trapping session thereafter. The traps were set in the evening and checked at midnight and again in the morning, until October 1978. To avoid trap mortality, from November 1978 onwards, I checked the traps three times each night at four-hourly intervals. The traps had to be closed during the day otherwise jungle crows Corvus macrorhynchos disturbed them. Shrew-moles were individually marked by toe-clipping ; their sex, repro- ductive condition, weight, age, and the point of capture were recorded before their release. Sexing of live shrew-moles was difficult (except in the breeding season), but could be done by physically checking for the presence of a penis, using the finger tips. Further details of methods for estimating reproductive condition are given by Ishii (1982). The current year's young were identifiable because they had milk teeth or only slightly worn teeth and they could be easily distinguished from adults that had overwintered and had more worn teeth. Methods of analysis Home range size was estimated using two methods : (1) the minimum range area (Stickel, 1954), and (2) the probability ellipse (Jennrich & Turner, 1969 ; Mazurkiewicz, 1969). The minimum range area was obtained by connecting the outermost capture sites of an individual and measuring the area inside. Using the probability ellipse, the area was calculated to contain 95 percent of expected captures. To examine the distribution pattern of home ranges, I mapped the capture sites of each individual as a minimum range area. Only those individuals captured at more than three different sites were included. In addition, the degree of range overlap (DRO) was calculated among the individuals of the same sex and between the sexes using the following equation : DRO=100 x CAoICAh where Ah is the range area of an individual (holder) concerned, and Ao the area occupied by other individuals (invaders) within the holder's range. Also, social interactions were considered when two individuals were captured simultaneous- ly with one trap (double captures). Home Range of Shrew-mole Results Trapping success Shrew-moles were easily captured using Sherman live-traps. During the course of the study, 85 shrew-moles (49 males and 36 females) were captured and marked individually. Between June 1977 and October 1978, however, 15 individuals died of starvation in the traps. After November 1978, when traps were checked every four-hours, trap mortality was reduced to just one case. Home range size To determine the number of captures needed to calculate home range size, I plotted range size, calculated by the minimum area method, as a function of the number of consecutive captures for several individuals having high fre- quencies of recapture within a constant reproductive condition. Some of the results are shown in Fig. 1. On the basis of those results, about 20 captures was established as the necessary minimum for calculating actual home range size for observing comparison with other species, and set 10 captures was regarded as the minimum for observing the size differences between the sexes and between the seasons of this study. This compromise was unavoidable due to data limitation. The calculation was only made for those individuals whose ranges were entirely within the trapping area. CONSECUTIVE CAPTURES Fig. 1. Change in home range size plotted against consecutive captures for five individuals. The range size was calculated as the minimum range area. Breeding condition is given in parentheses after each individual's number ; N = non-breeding, B = breeding. Home range sizes were calculated separately for the breeding and non- breeding periods. Males in the study area became sexually mature in early February, and active males decreased rapidly during May. Mature females were observed in February and March, pregnant females in March and April, and lactating females in April and May (Ishii, 1982). During the non-breeding season, no significant difference was found between the range sizes of males and females (Mann-Whitney U-test: the minimum range area, U = 11 ; the probability ellipse, U = 13). The home range of breeding males was, however, significantly larger than that of non-breeding males (the minimum range area, U = 9, p < 0.05 ; the probability ellipse, U = 1, p <0.01). Six males were captured 10 or more times both as non-breeders and as breeders, and in five cases the breeding home range size was the larger of the two. In contrast, no significant difference was found in range size between non-breeding and breeding females (the minimum range area, U = 5 ; the probability ellipse, U =2), though female #69 enlarged her home range during the breeding season. There was no significant difference between male and female range sizes during the breeding season (the minimum range area, U= 4 ; the probability ellipse, U = 3). Home range sizes of individuals that were captured 10 or more times within a constant reproductive condition (non-breeding or breeding) are listed in Table 1, and home range size statistics are given in Table 2. Table 1. Home range sizes (m2) calculated using two different methods for individuals captured 10 or more times. The number of captures= n. Non-breeding Breeding Individual No. Min. range 95%.prob. Min. range 95% .prob. n area ellipse area ellipse Male # 59 74 77 83 96 99 101 105 105* 116 Female # 43 55 69 72 80 81 - * different period Home Range of Shrew-mole Table 2. Mean home range sizes (m2)with standard deviations (SD) as estimated for individuals captured 10 or more times and those captured 19 or more times. The sample size = n. Min. range area Prob. ellipse n Mean SD Mean SD Individuals captured 10 or more times Non-breeding Males 10 1160 560 3537 1128 Females 4 1563 539 4700 2056 Total 14 1275 566 3869 1468 Breeding Males 6 2825 1396 8729 2670 Females 3 1450 540 6304 2197 Individuals captured 19 or more times Non-breeding Males 5 1510 607 3899 1513 Females 4 1563 539 4700 2056 Total 9 1533 542 4255 1705 Breeding Males 3 3933 520 10120 2403 Distribution pattern of home ranges During the non-breeding season, male ranges more or less overlapped each other, but were exclusive at least between resident individuals which remained until the beginning of the breeding season. During the breeding season, male ranges overlapped extensively as they expanded their ranges. Female ranges showed little overlap during the non-breeding season, especially among resi- dents, with this tendency being more distinct than that shown among males. During the breeding season female ranges tended to overlap to a lesser extent than male ranges. Male and female ranges overlapped throughout the year, and during the non-breeding season there were some cases in which one male and one female occupied almost identical areas. For example, in the non- breeding season of 1977-78 male #77 and female #55, and in 1978-79 male #96 and female #72, and male #I05 and female #69 were such pairs.