誌名 Animal behaviour and management ISSN

(1)

japonicus)の出没と里地里山環境との関係

誌名

誌名 Animal behaviour and management ISSN

ISSN 18802133

著者著者

梅田, 健太郎羽山, 伸一山本, 俊昭梶ヶ谷, 博巻/号

巻/号 47巻1号

掲載ページ

掲載ページ p. 1-11 発行年月

発行年月 2011年3月

農林水産省農林水産技術会議事務局筑波産学連携支援センター

Tsukuba Business-Academia Cooperation Support Center, Agriculture, Forestry and Fisheries Research Council Secretariat

(2)

Relationships between intrusion of Japanese blaclくbear

( U r s u s t h i b e t a n u s j a p o n i c u s )

and Satochi‑Satoyama environment in Numata City

，

Gunma Prefecture

，

Japan

Kentaro UMEDA¹

*

， ShinイchiHAYAMA¹， Toshiaki YAMAMOT0²， Hiroshi KAJIGAYA² lLaboratory ofWildlife Medicine， Nippon Veterinary and Life Science University， Tokyo， Japan

2 Department of Applied Science， Nippon Veterin釘yand Life Science University， Tokyo， Japan

*

Corresponding author. E‑mail address:[email protected]

Abstract

Several factors are implicated in intrusion of Japanese black bears (Ul幻ISthibetanus japonicus) into human living areas， and one of them is the deterioration of the Satochi‑Satoyama environment. We analyzed the relationships between bear intrusion加dSatochi.^幽Satoyamaenvironment in Numata City， Gunma Prefecture， to elucidate factors involved in bear ir市usion.We classified出eSatochi‑Satoyama environment by land usage into six groups， namely， crop fields， paddy fields， orchards， fallow fields， wastelands and forest edge; their locations， planting condition， and management situation were investigated. Data of the Satochi‑Satoyama environment quantified using the Geographic Infonnation System and data associated with the bear intrusion sites within the area investigat巴dwere analyzed. The analyses were carried out macroscopically and microscopically on divided lots and bear intrusion sites. Macroscopic analysis revealed a positive significant relationship of bear intrusion with geographical complexity (P < 0.01) and the area ratio of orchards (P < 0.01). Microscopic analysis showed that the bear intrusion

企

equencyincreased significantly with orchards (P < 0.01) and on the forest edge with heavy tussock growth (P < 0.05). The results suggest that the influential factors in bear intrusions in the s加dyarea include geographical complexity and the presence of orchards and tussocks.

Key words: geographic information system， intrusion， Japanese black beaにSatochi‑Satoyama

Animal Behaviou1' and Management， 47 (1): 1・11，2011 (Received 2 Jun 2010; Accepted fo1' publication 11 Jan 2011)

Introduction

The Japanese black bear (Ursus thibetanus japonicus) is one of the largest terrestrial mammals in Japan. The body length and body weight of組 adult bear are in the range of 120・・・145 cm and 70・120 kg， respectively (Yoneda 2008). Japanese black bears inhabit Honshu and Shikoku in Japan

，

but are thought to be at risk of becoming extinct in some areas (Oi 2009). They used to be found in Kyushu

，

but are most likely extinct in this region. In recent years， many cases of bearm加lsioninto human living ar'eas have been reported in various regions in Japarl. The bear intrudes into human living創'easand damages agriculhlral crops and fruits， and occasionally harms humans. Therefore， be創・ intrusionis a serious social problem in highィisk reglOns

The annual number of bear' intrusion incidents and resu1ting damage flucωate. A high incidence of bear' intrusion in 2006 r巴sultedin danlage that was more severe than in other ye創.'swhen the damage arnount

was usual. The amount of agricultural damage caused by bears in 2006^巴xceeded700 million yen to the average range of 300‑400 million yen (Ministry of Agriculture， Foresむy and Fisheries 2009). Also， the number of personal injuries from bear intrusion in 2006 was 140 to the average range of 40・50(Minisむyofthe Environment 2010). lf

也

elugh incidence observed in 2006 becomes more common， the scale of damage caused by the bear will likely increase. ln addition， capturing a large number of the bear inおudinginto human living areas because of a pest con位01willlead to extem

U l l

ation of the bear population in the region To reduce the contact betw巴enhumans and J apanese black bears， the control of bear intrusion is essential To successfully achieve this， it is necessary to identi命 factors affecting bear in往usionin order to facilitate the development of appropriate control measures.

One possible cause of bear intrusion into human living areas is nut production failure， of which its relationship to bear in加lsionhas been investigated previously (Nagai 1998; Tani伊chi& Oz誌i2003; Oka

(3)

e l a

l. 2004). All of these studies used the number of bears captured because of a pest control or the l1umber of witnessed bear i^l¹^釘usioncases as indexes of bear intrusion

企

equency，al1d revealed that these indexes correlated with the condition of nut production. Moreover， another factor suggested was deterioration ofthe Satochi‑Satoyama environrnent (Oi 2005; Miura 2008; Kawai 2009). Satochi‑Satoyama is a boarder zone between urban areas and wildemess. The Satochiδatoyama environrnent has been established through various human activities， and is composed of areas such as secondaty forests surrounding villages， agricultural fields scattered in these secondaty forests， reservoirs， and meadows (Mil1istry of the Environrnent 2001)ーTraditionally，Satochi‑Satoyama has been used as firewood sources， and for timber harvesting and charcoal production，創ld therefore was frequently visited atld managed by people. However， in re巴c^印e凶n

1 t t

yea創rsラ丸Sa幻toch

悩

1ηi‑Sa従toyamahas not been well m口1a^飢II泌1It

幻

ained c

∞

omp抑ar，吋 p

戸

re^肝Vlぬo山usl

か

yand has b^巴enneglected d

如

ueto progressive depopulation and a昭glllgIII areas a^釘round Satochi‑Satoyama. Consequently， broadleaf forests that produce food for wildlife have spread to forest edge areas and concealed the inner forest f‑iom view. Furtherrnore

，

the area of abandoned agricultural fields has been increasing (Ministry of Agriculture， Fores

t I

y and Fisheries 2008)， and thick gr0wt11S of herbaceous plants in these abandoned fields and their surrounding at.eas have served as hiddel1 areas and patlls for wildlife intruding into human living areas. These changes in the Satochi‑Satoyama environrnent have become favorable for wildlife to live， and this is believed to be one of the causes for the frequent bear in

t I

usion

Despite this situation， there are a few studies investigating the relationship between bear intrusion and the Satochi^輔Satoyamaenvironrnent. Honda

e l a

l. (2009) developed a predictive risk model and a map of human‑Asiatic black bear contact covering the whole area of Yamanaslu Prefecture. This model indicated that the high‑risk sites of hUlllan‑bear COl1tact were fixed. Moreover， the model showed that distance from the forest edge， ratio of conifers to the total forested area， bear distributiol1， and distance from settlement are important f:

control of this unbeneficial beat. behavior. We believe that identification of the factors u^l¹derぢing bear intrusion wi1l address tlle problems betwe巴nhumans and Japanese black bears.

Materials and Methods Study area and investigatiol1 period

The study area was the Hocclu area (2.7 km~) of Numata City (36⁰38' N， 139⁰02' E; 443.3 km²) located in the north patt of Gunma Prefecれrre(Fig. 1).

ト

Nu限m潟a幻則ta Cit

勿

Y1おslocated i設n1a Japan se伺a^曹t

守

ypeclirnate zone known foωr S^l¹0wy weather dmロr澗 g w

叩

II泌It匂加e^釘r加 df自lt出1おewe^伺at

白

heぽrlll SUlll立meiぽr.The city' s altitude is in the r釦 geof250・2，100 m， and the average temper従ぽe al1d armual precipitation are 11.1 oC atld 1098.9 mm， respectively. The Hocchi area is located in the northwest area of Numata City a^l¹d is surrounded by ridges of Mt. Hotaka (2，144 m). The Hocchi area with an altitude of approximately 500・800m actively produces fruits such as apples. The investigation of the Satochi‑Satoyama environrnent in the Hocchi area was pelfomled for・10 days from October・toDecember 2008.

Categorization of the Hocchi area by land usage The Satochi^嗣Satoyamaenvironrnent in the Hocchi area was first grouped into six areas by land usage， namely， crop fields (excluding paddy fields)， paddy fields， orchards， fallow fields， wastelands and forest edge. These six areas were fi.uther categorized according to pl^加 ting condition and management situatiol1， atld their locations were recorded (Table 1).

Table 1 Details of categories of land usage ca組伊ri

問。f

l自主d田 age

plan話ng字

condilion

主主anagement^{キ季} situalion crop fields

p自.ddyfields 十十

ロ

rchards

fallow fields

wSBtel811ds

一

^‑or‑or^十

forestedge

* : " 1

+

J with planting

，

1" ‑J without planting

** " 1

+

J managed by humans: tussocks were mowed， 1" ‑J not managed by humans (moderate tussock growth): the height of tussocks was less than 50 cm，ー …Jnot managed by humans (heavy tussock growth): the height of tussocks was 50 cm and above

(4)

N _A

も ¥ 〆マ

. ¥ ミ ¥ v

ず

^E

j¥¥ 官 ^j 〆

デゾ，

む/¥斗

1 〓

^J

mata C i t y ，イ ; ¥ nma P r e f e c t u r e !グ

j k 刀

ノー~会トぐ

: ゴゲヲモ

^v

^¥ ^; ^ぐ

… ーが

デ ; / 宅 f ¥

ぜ

:;J ¥

トー一一一‑1今

、

。

250 500km

Figure 1 Study area. The location of Numata City

( 0 )

is shown on a map of Japan (Ie代)， and the Hocchi area is encircled with a dotted line on an aerial photograph (right). The area investigated in the current study includes the south of Kamihocchi‑machi， Nakahocchi‑machi， and the north of Hocchishinden‑machi.

For the assessment of management siれ1ation，three levels were used: managed; not managed with moderate tussock growth; not managed with heavy tussock growth. The levels of tussock growth were assessed using a tussock height of 50 cm as threshold based on the fact that the height of adult bears is approximately 50‑60 cm (Hazumi 1996). We considered that tussock height affects the intmsive behaviors of bears because growth of more than 50 cm in height (heavy tussock growth) but not under 50 cm (moderate tussock growth) likely provide screens for bears to hide upon intrusion into human living areas. In addition， management situation of the forest edge was assessed within a 5‑m inner rim (towards the forest) of the forest edges; however， tussock grm九rthin otheI・

parts ofthe forest was not recorded

While investigating the Satochi^楠Satoyama environment， local residents were also interviewed whether they had witnessed bear intrusion and/or had been a貸出tedby bear intmsion. If available， detai!ed information on these incidents was recorded Furthermore， when comparatively new signs of the bear (claw marks， feeding marks， tracks and faeces) were found dming the investigation， their locations were recorded on a map

Data related to bear IntrusIon

Data

企

oma total of 70 sites of bear in加ls10n (hereafter intrusion site data) kept by the Natmal Environment Division of Gunma Prefecture. the Department of Forestry and Environmental A

伯郡

(5， 4 and 31 cases in 2004， 2005 and 2006， respectively)， the Agricultural Administration Division of Num昌ta City， the Department of Economic Mfairs (6 and 20 cases in 2007 and 2008， respectively)， and om group (4 cases in 2008， obtained in this study)， were analyzed The inなusion site data include detail巴d location (location on the map， address and land usage)， date， dmation， and situation (crop damage or witnessed). The types of crops affected and the approximate scale of damage， if applicable， were also included. These data were based on interview with local residents: dlerefore， data ofbear captivity were not included. Data analysis

We initially used the Geographic Infonllation System (GIS) (Ar・cGIS9.3， Environmental Systems Research Institute Inc.， New York， USA) to set the area to be analyzed (analysis area). The distance企omthe forest edge to each bear intrusion site恒例s10n distance) was measUI

吋

basedon the intrusion site data in order to obtain mean intrusion distance and the standard deviation. The zone between the forest edge and the line positioned at a combined distance of the

(5)

mean intrusion distance and standard deviation was defined as the analysis area.

We next used GIS (忘れvironmental Systems Research Institute Inc.) to incorporate data of b巴ru

ll1truslOn sltes剖 dtypes of land usage (crop fields， paddy fields， orchards， fallow fields， wastelands and forest edge) within the analysis area to the map of the Hocchi area (1:25，000 scale， Geographical Survey Institute， Tsukuba， Japan) in order to analyze the relationship between bear in加 sion and the Satochi‑Satoyama env

I r

onment.

Data of the Satochi‑Satoyama env

I r

onment were grouped into the macroscopic and microscopic env

I r

onments， and separately rulalyzed (see below). (i) Relationship be何reen bear intrusion macroscopic Satochiδatoyama environment

and Since the analysis are昌 line

おか

extends in the north‑south direction， lines perpendiculru' to the longitude lines were drawn at 100‑m interva1s to divide the analysis area into lots. The area was also divided to the east block and west block， and 40 lots were allocated to each block. The total number of lots analyzed in this study was 80. Macroscopic analysis was performed on the lot‑based data.

We focused on geographical complexity and the rate of land usage as macroscopic factors of the Satochi‑Satoyama enviromnent involved in beru

ll1truslOn.

Geographica1 complexity in each lot was analyzed based on the measurement of the length of the forest edge (LFE， Fig. 2) using GIS. The LFE becomes large when the forest line frequently pe即位atesinto the il1l1er

日十

^E

length of forest edge

(LFE)

edge of analysis area

‑‑‑‑‑‑‑f^{時初}

100m

v

analysis urca

Figure 2 The length of the forest edge し(FE)，The concept of LFE is illustrated， The forest edge is indicated by bold lines. The analysis area is shaded with hatched lines and outlined with dotted lines defining the edge of analysis area. Each lot was created by dividing the analysis area into 100‑m intervals along the longitude lines. The length of the forest edge was measured in by individual lots.

forest， and a large LFE indicates the penetration of human living area into the for・est.Ther巴fore，we considered that LFE represents the geographical complexity ofthe individuallots.

The area of e昌chtype of land usage (crop fields， paddy fields， orchards， managed fallow fields， non‑managed fallow fields， managed wastelands， and non‑managed wasteland) in each lot was measured， and the rate to the total area of the lot was calculated. Two types of non‑managed" land usage (fallow fields and wasteland) here were not categorized further by the extent of tussock growth (moderate or heavy) used for the assessment of management si加ation described earlier.

We considered that these data of the Satochi‑ Satoyama environment would largely be Ul1changed， unless a 加ge^帽scaledeforestation and re‑plal1l1ing were carried out. Therefore， we used the intrusion site data in the period

企

om2004 to 2008 and data of the Satochi‑Satoyama environment quantified using the GIS in this study. Fmihermoreラwhena beru・ll1truslOn incident occurred outside the analysis area， it was treated as an incident occmTIng in th巴nearestlot based on the measurement on the horizontal line to the longitudinallines.

Stepwise multiple regression analysis was perforwed to identi今出e factors affecting bear intrusion in the macroscopic Satochi^刷Satoyama env

I r

onment. The number of bear intrusion sites in each lot was used as a dependent variable， while LFE and rate of land usage in each lot were used as independent variabl巴s

(ii) Relationship be抑 een bear intrusion microscopic Satochi‑Satoyama environment

and The relationship between bear in出 sionand the microscopic Satochi‑Satoyruna environment was analyzed on a site‑basis rather than on a lot^剛basis.

We focused on the land usage of the bear inなuSlOn site and the management si印刷onof the near巴stforest from the intrusion site as macroscopic enviromnental factors affec

t I

ng bear in釘usion in the Satochiδatoyama environment.

The frequency of bear intrusion incidents in each type of land usage was calculated to examine frequency distribution. The neru'est forests 合omthe intrusion sites were classified into subgroups by management levels (presence of management and tussock growth level)， and the frequency of beru intrusion in each subgroup was calculated to eXaIlUne frequency distribution， Probabilities of nonUl1iforwity in frequency distribution were statistically analyzed using the Binomial test.

Because th巴seanalyses were perforwed to identi命 the relationship between bear intrusion and the microscopic Satochi‑Satoyama environment at tlle time of this study， data related to the bear intrusion incidents that occurred in the analysis area in 2008 were only used. Hence， data of 23 intlUsion cases that occmTed in

(6)

2008

，

after巴xcludingone case outside the analysis area frOlll 24 cases， were used for the lllicroscopic 昌nalysis. Extraction of high聞riskzones

羽市en factors affecting bear intrusion in the Satochi個Satoyama environment are， identified， they should be utilized for taking the appropriate llleasures to prevent bear輔氏lateddamage in the study area. F or this purpose， we aillled to extract high‑risk zones (HRZs)， that is， areas at particularly high risk of bear m釘usion. Certain conditions were illlposed on the m加lsionfactors of the Satochi^欄Satoyamaenvironment， which were identified by data analysis

，

in order to exむactHRZs

Results

20.9%， 2.2%， 7.8%， 0.6%， 10.7% 釦 d 21.3%， respectively. Statistical analysis revealed that the bear intrusion :fi‑equency was significantly increased with orchards (one叩 dedtest， P < 0.01).

Assesslllent of the lllanagelllent situation of the forest edge nearest to each bear intrusion site revealed that tussock growth was heavy in 95.7% of the cases (22 of 23 cases)ラwhilellloderate in 4.3% of the cases (1 of 23 cases). The length ratios of the lllanaged forest edge

，

non‑lllanaged forest edge with llloderate tussock growth

，

and non‑lllanaged forest edge with heavy

如ssockgrowth to the total LFE were 1.6%， 17.6% and 80.8%， respectively. Statistical analysis revealed that the bear intrusion frequency was significantly increased on the nonせnanagedforest edge with heavy tussock growth (one‑sided test， P < 0.05).

Extraction of HRZs

Analysis area HRZs were extracted based on the analysis results Mean and standard deviation of the intrusion

distance frOlll the forest edge were 35.1 III and 33.5 lll， respectively. Based on these results， the analysis ar巴a was defined as a strip with a width of 69 III

illllllediately adjacent to the human living area side of the forest edge. Thus， the area of 1.0 km²was analyzed in this s加dy.

Relationship between bear intrusion and lllacroscopic Satochi‑Satoyallla environment

Using lllultiple linear regression analysis， we examined the correlation betwe巴nthe nUlllber of bear intrusion sites in each lot and both LFE and area ratio of crop fields， paddy fields， orchards， lllanaged and non‑lllanaged fallow fields， and lllanaged and non‑lllanaged wastelands. Our results revealed that significant positive correlations were found only between the nUlllber of bear intrusion sites and both LFE (P < 0.01) and the area ratio (P < 0.01) of orchards The partial regression coe伍cientsof LFE and the area ratio of orchards were 0.007 and 0.013， and their standard errors were 0.003 and 0.002， respectively. The coefficient of deterrnination was 0.38. With regard to other factors， the nUlllber ofbear in住usionsites showed no significant correlations with the land usage rate of crop fields (P 0.75)， paddy fields (P 0.77)， lllanaged fal10w fields (P = 0.67)， non‑lllanaged fal10w fields (P = 0.90)， lllanaged wastelands (P出 0.68)，or non^刷lllanagedwastelands

( P

0.64)ー

Relationship between bear intrusion and microscopic Satochi‑Satoyama environment

A total of 87% of the bear intrusion sites were within orchards (n = 20)， 4.3% within crop fields (n =

1)，4.3% within non‑lllanaged fal10w fields (n口 1)，and 4.3% within non‑managed wastelands (n 1). The rates of the areas used for crop fields， paddy fields， orchards， managed fallow lands， non‑lllanaged fal10w lands

，

lllanaged wastelands and noルlllanage wastelands and others (e.g.， roads) were 20.1%，16.1%，

口

^mal

刊 …

口…

^t

口刊山山

²

ロ悶町

^Z

輔醐醐 ' ‑ ‑ ‑ ‑ ' 蹴醐醐」ーJ

町、ス A持Y、出必ゲ円~/ ~ r‑‑勺「宅，...J.。弔..JY，;脅執ぷ日告ミ祝日

Figu治 3 H RZs extracted based on the results. Sites of bear intrusion reported in 2004 (.)， 2005 (4.)， 2006 (8)， 2007 (+) and 2008 (X) are shown with the analysis area (blank polygon)， forest (cross^側hatched polygon)， human living areas (striped polygon) and HRZs (hatched polygon) on the map of the study area. Areas in white are those not belonging to the above categories.

(7)

the lots having an area ratio of orchards of 16% or higher・werede:fined as HRZs.

Thirdly， since high bear intmsion frequency in the forest edge with heavy tussock growth suggested that risk of bear intmsion is high around the forest edge heavily covered with tussocks

，

a condition dependent on the levels of tussock grm九Ithwas established. The bear intrusion sites^{前倒} thefor・estedge with heavy tussock growth reported in 2008 (n 22) were analyzed. The intrusion distance was measur・edin each case and a mean intmsion distance of 19.4 m and a standard deviation of 13.4 m were obtained. The combined value of the mean intmsion distance and the standard deviation (32.8 m) was used as a condition to extract the area located within 33 m from the forest edge with heavy tussock growth.

Finally， since a high bear intmsion frequency in orchards suggested a high risk of bear intmsion in orchards， the utilization of land as an orchard is used as a condition dependent on land usage. When only one part of a single orchard satisfies the above‑mentioned conditions (detennined by LFE， the area ratio of orchards and the tussock growth level in the forest edge)

，

the entire orchard was considered to satis命the HRZ conditions.

so far obtained in this study and data related to the Satochi‑Satoyarna enviromnent in 2008 using the following conditions (Fig. 3).

Firstly

，

since a positive couelation between the numberof bear in加 sionsites and LFE suggested that risk of bear intmsion becomes higher as LFE becomes larger， an LFE‑dependent condition was established. To establish the condition， the LFEs of individual lots were first expediently grouped using a 10‑m interval， arld then the median LFE was calculated. The median LFE plus 1 m or longer was defined as the LFE‑dependent condition for HRZs. The calculated median LFE was 130 m (Fig. 4). Therefor・e，lots having an LFE of 131 m or longer were defined在sHRZs.

Secondly， since the positive couelation between th巴

number of bear intmsion sites and the area ratio of orchards to the entire lot suggested that risk of bear intmsion becomes higher as the area ratio of orchar'ds becomes larger， a condition dependent on the area ratio of orchards was set. To establish the condition， the area ratios of orchards in individual lots were first expediently grouped using a 10% interval， and then the median area ratio was calculated. The median area ratio of orchards plus 1 % or higher was de:fined as the conditions dependent on the area ratio of orchards. The calculated median ratio was 15% (Fig. 5). Therefore

，

3.5 25

go

一

E3 h判長吋ね録︒

SM ob MM φ3 MM φh h 3.0

0.5 2.5

2.0

1.5

1.0

露盤Numberoflots

~ Frequencyofbear mtruSlon

20

15

10

6

B o z o E A g

同

M

Z

0.0

。

. . . . f b

t也 V n.C:j'/

' 0

LFEs(xl0

皿}

Figure 4 Distribution of the number of lots and frequency of bear intrusion in relation to L

ド

E.Number of lots and frequency of bear intrusion are indicated by vertical bars and dotled lines， respectively. The frequency of bear intrusion are calculated as follows: frequency of bear intrusionロ(numberof bear intrusion sites) / (number of lots).しF正sare expressed 1 0ィnunits on the X axis; for example， lots with LFEs of 10 units included those with LFEs of over 100 m but under 110 m (100 m ~ LFEく110m).

% 、、 . < o ヘ

1 )

，

九九 % 、

ヘ

も

ヘヘ

<o

A

、

.~

、

弘為やや

、

'¥

、。

(8)

4.5

4β

d G

3.5

‑ !

;;....

.

，

3sg

;;....

< < 1

2.5 _~

屯ト4

E

0 _ω

ロ

0

・

Q) 品a 弘 0.5 1.5 1.0 2.0

際瑠Numberoflots

一命 ‑Frequencyofbear intmsion

45

35

15

5 30

20

10 25 40

3 3 3 題点 g p z

0.0

。

70

Ar ea : r a

鼠

oofo : r c l

国:rds(%)

Figure 5 Distribution the number of lots and frequency of bear intrusion in relation to the ratio of orchards. Number of lots and frequency of bear intrusion are indicated by vertical bars and dotted lines， respectively. The frequency of bear intrusion are calculated as follows: frequency of bear intrusion

=

(number of bear intrusion sites) / (number of lots). The ratios of orchards are shown on the X axis and were grouped using an interval of 10%; for example， lots with an area ratio of orchards of 0% included those with the ratios of 0% or over but under 10% (0%::;; area ratio of orchards < 10%).

50 60 40

30 20

G 10

area and the human living area. Although there was no previous study using LFE for analysis， we used LFE as an index reflecting two‑dimensional and geographical characteristics of the study area

Since the correlation between orchards and bear intrusion sites was demonstrated by both macroscopic and microscopic analyses， the presence of orchards appeared to be one of the important factors that affect bear intrusion.τhe Japanese black bear is omnivorous with a preference for vegetable foods and is known to consume various types of food， such as herbaceous plants

，

insects

，企

uitsand nuts depending on the season (Hashimoto & Takatsuki 1997; Hashimoto 2002; Sakamoto & Aoi 2006). Furthermore

，託

wasreported that the Japanese black bear seasonally migrates to find sufficient food (Tamatani et a .l2001; Saito & Aoi 2003; lzumiyama & Shiraishi 2004). Considering the ecology of the bears demonstrated in these study， orchards provide a stable food supply to the bears inhabiting areas around Satochi‑Satoyama in the poor harvest seasons (e.g.， au加mnis a bad year for nut production)， and therefore， are the focal site of bear intrusion. lzumiyama et

a

.l(2008) assessed the time of bear in加lsion into Satochi‑Satoyama and age of captured bears， and suggested that food shortage was more severe in the inner part of the mountains in Nagano Prefecture in 2006 when the incidence of bear Discussion

Bear intmsion from the perspective of Satochi‑ Satoyama Environment

The macroscopic analysis revealed that the

会equencyof in位usionof Japanese black bears was high in geographically complex areas and areas highly used as orchards. The microscopic昌nalysisshowed that the sites of bear in往usionwere unevenly dis^仕ibuted， and the dis位ibutionwas high in orchards and the areas adjacent to the forest edge with heavy tussock growth.

We considered

白紙

thegeographical charact巴nstlcs of the Hocchi area con紅ibuteto the high frequency of bear in包usionin the geographically complex areas^町The Hocchi area is surrounded by mountainous regions， and agricul加'alfields are densely distributed in the forest sides. These characteristics are similarly observed in the geographically complex areas. This implies that the areas wherein agricultural fields penetrate the forest have a high frequency of bear intrusion， suggesting that Japanese black bear intrude into the geographically complex areas in order to find crops in the agricultural field. Honda et a .l(2009) used geographical factors

，

such as dist如何企omthe river and elevation， for the analysis of human‑bear contact， whereas we used indexes， such as LFE， that reflect the two^刷dimensional and geographical relationship between the bear living

(9)

intrusion was markedly high. This also indicates that Jap^昌neseblack bear inhabiting the study area is likely to intrude into orch

紅白

whenit is unable to find sufficient food in its living zon巴

In addition， Satochi‑Satoyama has possibly become a favorable place for the Japanese black bear to live， and the bear se役ledin areas around Satochiδatoyama may intrude into human living areas in search of food To confirm this possibili思 itis necessary in theおture to monitor nut production in auturnn， as well as to investigate and analyze the home range ofthe Japan^巴se black beru・whichhas a confirmed in凱lsivehabit.

Thus far， electric fences have been used to prevent bear‑related damage.百leeffectiveness of these devices to block the intrusion of Japanese black bear was studied in Japan (Huygens & Hayashi 1999)， and their e

妊

ectivenessto prevent intrusion of American black bear (Ursus americanus) (Storer et a .l1938)， brown bear (Ursusαrctos) (Gard 1971) and polar bear (Urslis maritimus) (Wool

合

idg巴 1983; Davies & Rockwell 1986) was exarnined in other countries. In recent years

，

measures transfonning the devastated Satochi‑ Satoyama enviroument to an environmentally un合iendlyarea to wildlife have been taken. The buffer zones were fixed， for exrunple， abandoned orchru'd trees were removed and tussocks in abandoned agricu1tural fields and the forest edge were cut. Kyoto Prefecture (Minis

句

ofthe Environment of Japan 2006) and Nagano Prefecture (Nagano Prefecture， ForesむyDivision 2009) are among thos^巴thatexecuted such Satochi.^帽Satoyamamanagement programs against bear inむusion.In addition， Saito et a. l(2006) reported that mowing a 13^・hahinterland region of a community that suffered from serious damage caused by Japanese macaques (Macαca戸scαta)altered their home range， andp似1

How 巴V巴釘r^，^ラt

白

h巴6

貸他

'ecはtsoft

白

h巴s 巴m巴as^凱u江reson r 巴duc

叩

ttionof da創magecaused by wildlife we倒r巴rru'

芯叫

el却yinvestigated. In particular， only a few studies have scientifically investigated the effects of buffer zone management and therefore building a scientific basis is註nticipated.

We confirmed that heavy tussock growth in the forest edge， a sign of a devastated Satochi‑Satoyama environment， correlates with bear intrusion. However，

会Irther analysis of this correlation requires on‑site experiments

，

such as mowing tussocks grown in the forest edge a

HRZs for bear intrusion

Our resu1ts revealed the zon巴swithin the study area have a high risk of bear intrusion. Cases of bear intrusion that occun

吋

inthe analysis ru'ea of

出

isstudy (n ⁼42) were extracted from the bear i姉 usiondata in the period from 2004 to 2007 and superirnposed on the HRZ map. The bear intrusion sites in the HRZs or within a 15 m distance

企

om the HRZs were

expedient1y combined to obtain the ratios of bear intrusion incidents that occurred in or approximate to the HRZs. The obtained ratios in 2004， 2005， 2006 and 2007 were 50.0% (2/4)， 33.3% (113)， 82.8% (24/29) and 33.3% (2/6)， respectively. The overall ratio in four years was 69.0% (29/42). These resu1ts indicate that， although the ratios varied depending on the year， approximately 70% of bear intrusion sites were within the HRZs and adjacent (within 15 m) areas

To prevent bear in凱lsionand reduce bear‑related damage in the study area， it is necessary to focus on the HRZs and work on establishing countermeasures. However， monitoring each HRZ separately may cause the spread ofbears into other areas current1y with no or low risk of bear intrusion. For this reason， when regulating wildlife including bears， the entire area， rather than individual HRZs， should be controlled.

FurthermOle， this study was performed in a region with a relatively large area of orchards， and it became appar巴ntthat the presence of orchards was one of the major factors involved in bear intrusion. However， we did not analyze other factors such as distanc

巴企

oma river or surrounding forest vegetation. For this reason

，

it is necessary to crury out HRZ measures while considering the above‑mentioned factors in beru・

intlUsion cases in^組 ru'eawith no orchards or a sma11 area of orchards

Acknowledgements

We sincerely thank Mr. T. Kubodera (D巴partment of Forestry and Environmental A

伯

irs

，

Technical Suppo託 Division of Gunma Prefecture)， Mr. H. Sakaniwa (D巴partmentof Foresむyand Environmental A

伯

irs， Na加工al Environment Division of Gunma Prefecture)， Mr. S. Kuroiwa and Mr. H. Sakata (Depruちnentof Forestry and Environmenta1 Affairs)， Mr. R. Moro and Mr. N. Yamada (the Deprutment of Economic Affairs

，

Agricultural Administration Division of Numata City) for supporting this investigation ofthe Satochi‑Satoyama environment and data analysis， and for supplying data. We also express our herutfe1t thanks to Mr. N. Tamura， Mr. S. Ono (Ono Apple Fann) for providing valuable infOlmation on various aspects of Japanese black bear inhabiting the study area， and members of the Laboratory of Wildlife Medicine， Nippon Veterinary and Life Science University， Japan for suppOlting us.

References

Davies JC， Rockwell RF. 1986. An electric fence to deter polar bears. Wildlife Society Bulletin 14， 406^幽409

Gard R. 1971. Brown bear predation on sockeye salmon at Karluk Lake， Alaska. Journα10/附ldl件 Management 35， 193‑204

(10)

Hashimoto Y. 2002. Seasonal food habits ofthe Asiatic black bear (Ursus thibetanus) in the Chichibu Mountains， Japan. Mammal Study 27，65^嗣72

Hashimoto Y， Takatsuki S. 1997. Food habits of Japanese black bears: A review. Mammaliαn Science 37，

ト

19(in Japαllese witlz Englislz ahstract).

Hazumi T. 1996. Japanese black bear. In: Hidaka T(巴ds)，The Encyclopaedia of Animals in Japan Volume 1: Mammals 1， pp.l44

ぺ

47. Heibonsha Limited Publishers， To防o(in Japanese)

Honda T， Yoshida ，YNagaike T. 2009. Predictive risk model and map of human‑Asiatic black bear contact in Yamanashi Prefecture， centraI Japan Mammal Stu砂34，77・84.

Huygens OC， Hayashi H. 1999. Using elecむicfences to reduce Asiatic black bear depredation in Nagano prefecture， central Japan. Wildl件 Society Bulletin 27， 959^・964.

Izumiyama Sラ ShiraishiT. 2004. Seasonal changes in elevation and habitat use of the Asiatic black bear (Ursus thibet抑制 inthe Northern Japan Alps. Mammα1 Study 29， 1‑8.

lzumiyan1a S， MochiZl永iT， Kishimoto R， Gotoh M， Hayashi H. 2008. Elucidation of massive haunt factor to Asiatic black bear depend to mraI area capture time and age assessment in the Nagano Prefecture. Bulletin Shinshu University Alpine Field Center 6， 19・24(in Japallese witlz Ellglislz ahstract).

Kawai M. 2009.

羽生

atis Satoyama? In: Kawai M， Hayashi Y(eds)， Animal revolt， a rapidly growing deer population and bears intruding into residential areas， pp. 32・44. PHP Institute， Tokyo (in Japanese)

Ministry of Agriculture， Forestry and Fisheries of Japan. 2008. Statistical tables (trends in areas of agricultural fields and abandoned agricu1tm^司alfields， and rates of abandoned agricultural fields) [homepage on the Internet]. The Minis^むy of Agriculture， Fores

句

Tand F出 eries，Tokyo， Japan; [cited 23 May 2010]. Available from 主主L:

http://www.maff.goj.p/chus

h i /

ar油le/07siryo/panC yomigaem/pdf/p51̲56siryoul.pdf

Minisむy of Agriculture， Forestry and Fisheries of Japan.却2O

∞

Oω9.T

百

r巴叩nd

出

sin da叙mag伊eof agri

民

CUltUl江ralcαro

叩

ps c

印邑u出se

吋

db匂ywild birds and b己ast岱s伊[homep伊a伊ge

∞

0n the Int^，お巴rnτne吋t].The Mi

出

m

な

lS甜むyof Agricl

F

日lS

品

he

叩

I

ロ

i巴邸s

，

Tokyo ラ Japan; [cited 23 May 2010]. Available

企

om URL: h仕p://www.maff.goj.p/j/ seisarνtyo^可u

/ h i

gai/h̲zyokyo/h20/pdf/091218e.pdf Ministry of the Environment of Japan. 2001

Investigation and analysis of Satochi‑Satoyama in Japan (interim report)

using buffer zones [homepage on the Internet]. Biodiversity Center of Japan， Yamanashi， Japan; [cited 23 May 2010]. Available合om 型RL:

http://www.biodic.goj.p/cbd/2006/pdf/ll02̲2̲6 Ministry of the Environment of Japan. 2010. Personal

illJur1es caused by bear^司s in fiscal year 2009 [Immediate release] [homepage on the Internet]. Ministry of the Environment， Tokyo， Japan; [cited 23 May 2010]. Available from URL:

http://www.env.goj.p/nature/ choju/ docs/ docs4linjur y‑qe.pdf

Miura S. 2008. Guide to wildlife management

，

how 10

deal with wild αnimals. Iwanami Shoten， Publishers， Tokyo (in Japanese).

Nagai S. 1998. Relationship between the number of capωred Asian Black bear and fluctuations of fruit bearing of Fagus crenala and Querus mongolica var. grosseserrata

，

in Toyama Prefecture. Living Organisms in Toyama p

， ψ

cture 37， 17・22 (in Japanese)

Nagano Prefecture， Fores町 Division. 2009. A collection of cases of community building to become resistant to wild birds and beasts (2008 edition) [homepage on the Internet].子4砲ano Prefec加ralGovernments， Nagano， Japan; [cited 23 May 2010]. Available from !l匙 http://www.pref.naganoj.p/rinmu/shinrin/04choj

刊 /

13 zireνzireiTOP.h加1

Oi T. 2005. The Animαl's Forests in Transition Ecology

，

Evolution and Conservation

0 1

Forest λ1ammals in Japαn^… Tokai University Press，

Kanagawa (in Japanese)

Oi T. 2009. The Japanese Black Bear: Biology

0 1

Bears in Forest Ecosystems. Tokai University Press， Kanagawa (in Japanese).

Oka T， Miura S， Masaki T， SUZl氷iW， Osumi K， Saitoh S. 2004. Relationship between changes in beechnut production and Asiatic black bears in northern Japan. Journal

0 1

Wildlife Management 68， 979^胸986.

Saito C， Morimitsu Y， Kiyono H. 2006. Crop damage management: experimentaI habitat management and its effect on ranging behavior of Japanese macaques (Macacα戸'scαta).Mammalian Science 46，63^胸64(in Japanese).

Saito M， Aoi T. 2003. Seasonal home range use by black bears living in a rural area forest in Tono， Japan. Bulletin

0 1

the Iwαle University Forests 34，

11・22(i1l Japallese with Ellglislz ahstract)

Sakamoto Y， Aoi T. 2006. Food habits of the Asiatic black bear Ursus Ihibetanus japonicus in the northern Ohu Mountains， Japan. Bulletin

0 1

Ihe Iwate UnIversity Forests 37， 17^刷27(ill Jt司pallese witlz Ellglislz ahstract)

Storer TL， Vansell GH， Moses BD. 1938. Protection of mountain ap1ar1es from bears by use of electric fence.

， j

(11)

characteristic and habitat use of Japanese black bears (Uηus thibetanus j中onicus)in the northem Kinki district. Forest Research 73， 1‑11 (in Japαnese witlt Englislt abstt・，(lct).

Taniguchi S， Oz北iS. 2003. Relationship between the number of black bear (Ursus thibetanus jαrponicus) witnessed in the surrounding villages and fiuit bearing of Fagus crenata and Quercus mongolica var. grosseserrata in Mt. Hyonosen Hyogo prefec^何回ー Japanese Journal of Forest

Environment 45， 1・6(in Japallese witlt Ellglislt abstract).

Wooldridge DR. 1983. Polar bear elecなonicdet巴:rrent and detection systems. International Conference on Bear Research and Management 5， 264‑269. Yoneda M. 2008. Ursidae. ln: Abe H(巴ds)，A Guide to

the Mammals of Japan 2nd edn

，

pp.77・78.Tokai Universiザ Press，Kanagawa (in Japanese with Englislt abstract).

(12)

群馬県沼居市におけるニホンツキノワグマ ( U r s u st h i b e t a n u s j a p o n i c u s )

の出没と盟地盤山環境との関係

梅田健太郎羽山伸一山本俊昭乙梶ケ谷博2 l告本獣産生命科学大学野生動物学教室，武蔵野市

2日本獣医生命科学大学獣医学部獣医保健看護学科，武蔵野市

*

Corresponding author. E‑mail address:[email protected] 要約

ニホンツキノワグマ Ursusthibetanus japonicus (以下ラクマ)の人の生活関への出没には、いくつかの要因が考えられているが、その一つに里地里山(以下車地)環境の荒廃が挙げられている。本研究は、クマの出没と里地環境との関係を解析し、クマの出没要因の解明を閥的とした。本研究は、群馬県沼田市にて、 2008年 10月から 12丹の期間に行った。クマの出没に関連すると考えられる里地環境を畑地、水田、果樹園、不作付け地、荒地、林縁と定義し、これらの位置、作付け状流、管理状況を調査した。解析には、地理構報システムによって定量化した里地環境のデータと調査地におけるクマの出没地点データを用いた。解析は、旺視的解析と微視的解析に分け、それぞれ、作成した区画、

クマの出没地点を慕にして行った。巨視的な解析では、クマの出没と地形の捜雑性

( p

< 0.00および土地利用における果樹園の割合 (p< 0.00 との間に正の相関が確認された。微視的な解析では、

クマの出没が、果樹関 (pく0.00および薮化レベルが高い林縁屑辺 (p< 0.05)に偏っていることが確認された。解析の結果、クマの出没には、地形の擾雑な場所であることと、果樹関と薮の存在が影

していると考えられた。

キーワード:里地里山，出没，地理情報システム，ニホンツキノワグマ

Animal Behaviour and Management， 47 (1): 1・11，2011 (2010. 6.2受付;2011. 1.11受理)

誌名 Animal behaviour and management ISSN

japonicus)の出没と里地里山環境との関係

誌名