193

Diet of Jungle Crows in an Urban Landscape

Reiko KUROSAWA1*, Ryo KONO2, Takahiko KONDO3 and Yutaka KANAI4

1 Biodiversity Lab, Institute of Low Temperature Science, Hokkaido University,

Nishi 8-chome, Kita 19-jo, Kitaku, Sapporo, Hoddaido 060‐0819, Japan 2 Ecosys 3-11-18-201, Higashi, Kunitachi, Tokyo 186-0002, Japan

3 Environmental Ecosystem Lab, Rakunogakuen University, 582-1 Midoricho, Bunkyodai, Ebetsu, Hokkaido 069-8501, Japan

4 Wild Bird Society of Japan, 2-35-2, Minamidaira, Hino, Tokyo 191-0041, Japan * e-mail: rkurosawa@nifty.com

Abstract

An animal species’ diet constitutes the basic ecology of that species and is indispensable for understanding its population dynamics. As jungle crows have increased in the urban areas of Tokyo, conflicts between humans and crows have increased. Since highly urbanized areas are poor in natural food resources, it is possible that jungle crows depend heavily on human-related food in Tokyo. We have, therefore, determined their diet and compared the results with those of other regions, taking steps firstly, to assess the reliability of information from pellets, feces and direct observations of feeding crows.

Jungle crows exploit a wide variety of food in response to temporal and seasonal availability in urban as well as rural areas. Their winter diet consists primarily of plant matter, suggesting that they retain their original food habit of eating fruits and berries of trees and shrubs, even at a heavily urbanized site. The proportion of food scraps in their diet was similar to that of animal matter in our study, suggesting an unusually ample availability of anthropogenic food sources in this city. If food scraps in Tokyo could be reduced, a great proportion of the food currently available to the crows would be lost, driving down the carrying capacity. We suggest ecological studies be conducted before making some action plans to solve wildlife issues.

Key words: diet, food scrap, fruit eater, jungle crow (Corvus macrorhynchos), urban landscape

1. Introduction

Jungle crows (Corvus macrorhynchos) are large crows (body length ca. 55 cm) distributed in the forested regions of East Asia through the Far East Asia （Madge & Burn, 1994; Iwasa et al., 2002）. They inhabit a variety of forests, such as broad-leaved deciduous, broad-leaved evergreen and mixed coni-ferous forests. Forest dwellers, as the English name implies, they occur in farmlands with hedges and in wooded areas of city parks as well, making them one of the most familiar bird species to Japanese people （Higuchi, 1979; Karasawa, 2003）.

Having a close relationship with behavior as well as morphology, feeding habits are a basic ecological parameter of animal species. Jungle crows are year- round residents and are reported to have a home range of less than 8 km in Tokyo (Morishita et al., 2003). Therefore, the diet of this species probably reflects the ecosystems of its local habitats, which implies, on the other hand, that the diet varies from region to region.

Jungle crows are known to have a wide range of food preferences. They feed on insects, bird eggs and chicks, while they scavenge dead animals. They eat various fruits and seeds of trees as well, and thus their role as a seed disperser has been pointed out (Ueda & Fukui, 1992).

Although Jungle crows have a relationship, direct or indirect, with a wide variety of wildlife in their habitats, few studies on their feeding habits have been carried out in predominantly urban areas except for Osaka (Ueda & Fukui, 1992).

In Tokyo, a megalopolis with 12 million people, approximately 36,000 crows were counted in the winter of 2001-2002 (Ueta et al., 2003). Food scraps in this city are disposed of in transparent plastic bags and lie outdoors until garbage trucks come to collect them. Since the plastic bags are flimsy, they are vulnerable to scavenging by crows, cats and rats (Kurosawa et al., 1999a). The availability of food scraps in plastic bags on the streets is suggested to be responsible for the marked increase in crows in Tokyo (Kawachi & Matsuda, 1999; Kurosawa et al., 2001,

194 R. KUROSAWA et al. Ueta et al., 2003). The conflict between humans and crows has increased as the crow population increased in urban areas. The major complaint by the citizens was scattering of trash on streets by scavenging crows (Ministry of the Environment, 2001; Kurosawa et al., 2003). The citizens in areas with a higher frequency of garbage scattering had strong negative feelings towards crows. Since highly urbanized areas are poor in natural resources, it is possible that jungle crows depend heavily on human-related food in Tokyo.

In this paper, we have attempted to determine the ecological status of this crow in a highly urbanized area by analyzing its diet, and comparing it with those in other regions of Japan. In so doing, we would like to find a new light to reflect on the state of co-existence of man and crows in Tokyo.

Jungle crows eat food, frequently tearing it into small pieces, and they feed their young in their nests directly from the gular pouch, making it hard to identify the food items by video recording. They gather at communal roosts in the evening, and regularly regurgitate indigestible portions of food as pellets (Yamagishi, 1982). Since the pellets are fre-quently found on the ground under the roost, they are readily available for research workers. Food remains in the pellets can shed some light on the feeding environment of the crows that have discarded them. Since the food remains in the pellets are indigestible portions of the food the crows have taken in, however, they do not necessarily represent their food completely. Therefore, we have compared the results of the analyses of pellets and feces with direct observations of feeding crows (Kurosawa, 1999), in order to determine the reliability of pellets as an indicator of feeding habits.

2. Methods

The study sites were communal roosts of crows in

parks and a corporate institute with a large wooded area in Tokyo; namely, Mizumoto Park, Rikugien Park,

Tama Zoo and Hitachi Central Research Center (Fig. 1). The roosts were located in canopy-closed parts of broad-leaved deciduous or evergreen stands mixed with clusters of conifers (Cryptomeria, and Chamaecyparis). Since almost all the crows (97%) in Tokyo in 1999 were jungle crows (Kurosawa et al., 1999b), so we assumed that the pellets and feces belonged to this species. Collecting pellets and feces beneath the roosts between December 1999 and March 2000, we preserved them in a 75% alcohol solution for later analysis. Cleaning by water and screening all the specimens (44 pellets and 15 feces), we identified them using 8-50 power microscopes. In 1998 we observed jungle crows feeding in Inokashira Park (Kurosawa, 1999). We used observation data obtained during the winter months (Dec. to Mar.) for comparison.

We classified the substances identified in the pellets and feces into the following four categories: 1) Plant matter (fruits, berries, nuts, seeds, buds and other plant parts) 2) Animal matter (bones, insects, bird feathers, mammal hair, fish scales and others) 3) Artificial matter (cloth, fiber, paper, plastic, metal, glass and others) 4) Unidentified substances including inorganic matter (clay, pebbles and others).

We generally identified plants and animals at a family level, and when possible, at a species level (Asano, 1996; Yamatokeikokusha, 2000; Nakane et al., 1992). We assorted all fishes into animal matter without discriminating between marine and fresh water fishes. We placed all mammal hairs in animal matter. We excluded crow feathers from analysis as an alien object taken in by chance.

We worked out the frequency of each category. When a pellet contained parts of fruits, leaves, buds, bones, insects, plastic and paper, for instance, the frequency for plant matter was 3, animal matter 2, artificial matter 2, and inorganic matter 0. After we had compared the characteristics of the analyses of pellets, feces and direct observations, we made a list

Fig. 1 Map of study sites in Tokyo. shows crow roost sites.

Diet of Jungle Crows in an Urban Landscape 195 of food items according to the most suitable method and compared the results with former studies in regions other than Tokyo.

3. Results

3.1 Reliabilities of pellets, feces and direct

observations as indicators of diet The pellets of crows are reported to be shaped like

bullets (Fig. 2; Yamagishi, 1982). Most of the pellets collected at the study sites, however, came in indefinable shapes (Fig. 3), with a length up to 30 mm and the width less than 30 mm.

Each pellet contained a variety of substances (Table 1), which suggests that a crow ingests a wide spectrum of food in a short period of time (Kondo, 2003). Most of the pellets consisted of fragile substances which tend to crumble easily. The largest hard object was a 22 mm by 3 mm piece of twig and the bulkiest object was a spherical seed with a radius

of 5.0 mm (family Myrsinaceae). Pieces of plastic and mandarin orange endocarp were intact and folded (Fig. 4). Substances such as seeds, pieces of bones and the outer chitin of insects were easy to identify. Most of the substances contained in the feces were, however, very small fragments of leaves, tiny seeds less than 1 mm in diameter and other objects which were difficult to identify.

Plant matter was found in most of the pellets analyzed (43/44, Table 1). Animal matter and arti-ficial matter were contained in two thirds of the pellet specimens (34/44). We compared the categories based on the results of the analyses of the pellets and feces and the result of the direct observation of feeding crows (Fig. 5). The frequency of each category statistically differed among the three sources (χ2 = 52.7057, df = 6, Ｐ < 0.001), but the differences were not significant between any of the pairs （Fisher’s PLSD test: F = 0, df = 2, Ｐ > 0.05; F = 0.4569, df = 3, Ｐ > 0.05）.

Table 1 Food categories observed in the diet of jungle crows. Values show the cases with % occurrence observed in parentheses.

Food categories Pellets Feces Major items (Genus/species & family)

fruit 19 2 Citrus (Rutaceae), Rubus (Rosaceae), Rhus (Anacardiaceae), Diospyros kaki (Ebenaceae), Ardisia (Myrsinaceae), Quercus myrsinaefolia (Fagaceae), Castanea crenata (Fagaceae)

seeds 18 4 Poaceae buds 17 1 leaves, stalks, roots etc. 42 15 Total plant matter 43 (97.7) 15 (100) bones 12 1

insect parts 17 5 Hierodula patellifera (egg case), Rhomborrhina japonica & other Scarabaeidae

feathers, hair 18 3 birds, mammals scales 1 0 fish, reptiles Total animal matter 34 (77.3) 8 (53.3) cloth, paper 18 3 plastic, styrofoam, aluminum foil, paper plastic, metal etc. 23 3 glass, fiber (mainly cotton) Artificial objects 34 (77.3) 6 (40.0) pebble, clay, unknown 25 8 stones, clay, shells Inorganic and/or unidentified matters 25 (56.8) 8 (40.0)

Total 44 15

Fig. 2 Bullet-shaped pellet of jungle crows reported as Fig. 3 Pellet with an indefinable shape most frequently typical in former studies (photo by R. Kurosawa). observed in this study (photo by M. Matsuda). Bar indicates 3 cm.

196 R. KUROSAWA et al.

A balanced proportion of all categories was represented in pellets, whereas a large proportion of unidentified matter was seen in direct observations. On the other hand, artificial matter tended to be identified best in the direct observations, which frequently revealed the origin of the food items.

Parts of insects, fruits and seeds were best represented in pellet analysis, because the pellets contained larger pieces of indigestible substances than did feces. Direct observation had the advantage of identifying large food items as crows handled them. The analysis of pellets was, on the other hand, effective at identifying small food items containing indigestible substances that defied direct observation. In addition, pellets were easy to collect. In contrast, feces had the lowest rate of identification, indicating that they are not suitable materials for determining crows’ diets.

Therefore, we used pellet analysis to determine the crows’ diets, complementing its short-comings with the direct observation of feeding crows. The analysis of pellets could allow for comparison between different regions, if conducted under a standardized method and categorization.

3.2 Details of jungle crow diets in Tokyo Plant matter was found in almost all of the pellets

analyzed (97.7%; Table 2). Most nutritional among this were fruits and berries of trees and shrubs. The plant species found in the crows’ diets seemed to reflect the regional flora, because they were common in the Tokyo region both in natural and cultivated landscapes. Mandarin oranges (Citrus deliciosa, Rutaceae), chestnuts (Castanea crenata, Fagaceae) and persimmons (Diospyros kaki, Ebenaceae) are probably common food sources for crows. Seeds of Poaceae were also contained in the pellets, indicating that jungle crows fed on grass seeds in winter in Tokyo as they do in more agricultural regions also (Suzuki, 2002).

Second in frequency to plant matter in the pellets were artificial and animal matters (77.3%). The most common substances (17/44 cases) among the artificial matter were small pieces of white or transparent plastic bags usually used as garbage bags in Tokyo. The pellets also contained aluminum foil, Styrofoam and paper, all of which are used for food wrappings, suggesting their association with food scraps. Other inorganic matter, such as pebbles and clay were also found in the pellets.

The most frequent animal matter was insects, such as the ootheca (egg cases) of mantises (Hierodula patellifera, Mantidae), and beetles (Sacarabedae). The same groups of insects were reported in previous studies in rural regions (Ikeda, 1957; Suzuki, 2002). Most of the mammal hairs were human hair, which has a probability of being taken in by chance as the crows scavenged food scraps at garbage collection sites.

Table 2 shows the results of direct observation of feeding jungle crows (Kurosawa, 1999). Jungle crows were observed eating a variety of fruits and berries of trees, supporting the result of the pellet analysis. They were also observed to carry fruits of Rhus succedanea (Anacardiaceae) ， Zanthoxylum ailanthoides (Rutaceae)，Diospyros kaki，and genus Citrus species to their roosts (Reiko Kurosawa pers. observ.). We found saplings of Rhus succedanea, Zanthoxylum ailanthoides and Ardisia japonica (Myrsinaceae) under the roost sites.

Jungle crows were often observed scavenging food scraps in dustbins, in picnic areas, and at sites where people feed birds and fish. On a few occasions, they were observed snatching bags of snack food directly away from stalls in a park (Kurosawa, 1999). Thus, they readily use human food resources in Tokyo.

Animal matter identified in direct observations included rats, cats, feral rock doves, frogs, snails and earthworms. In addition, songbirds, such as fledgl-ing Gray Starlings and Tree Sparrows were taken in May and June (Table 2).

Fig. 4 Orange endocarp and egg case of mantis observed in a

single pellet (photo by R. Kono). Fig. 5 Proportion of food categories of jungle crows in each

sample source.

Diet of Jungle Crows in an Urban Landscape 197

4. Discussion The habit of procuring a wide variety of food in

response to temporal and seasonal availability was a common feature among jungle crows in urbanized as well as rural areas (Ikeda, 1957; Liu et al., 1986). Among these, the major food substance was plant matter such as fruits and berries of trees and shrubs. The fruits of genus Rhus in particular were reported as a food item of jungle crows in previous studies (Ikeda, 1957; Yamagishi, 1982; Ueda & Fukui, 1992; Suzuki 2002), suggesting that these fruits are a common and important food resource for this species. Therefore, jungle crows also retained their original food habit of eating fruits and berries of trees in a heavily urbanized site as Tokyo.

The second most important food substances for the jungle crows in Tokyo were both artificial and animal matters. The latter represents a natural food source commonly observed in the previous studies through-out this crows’ range (Ikeda, 1957; Yamagishi, 1982; Liu et al., 1986; Suzuki, 2002; Kondo, 2003). However, the high frequency of artificial substances would be a characteristic of crows’ diets in Tokyo.

There was a significant difference in the proportion of artificial substances found in jungle crows’ diet among regions: 77% in Tokyo, and 7.7% and 7.9% in the agricultural areas of Tsuruoka City, Yamagata Prefecture (Suzuki, 2002), and Ebetsu City, Hokkaido (Kondo, 2003) respectively (χ2 = 103.42, df = 2, P < 0.001; Fig. 6). If we consider marine fish, mandarin oranges and human hair as taken from food scraps, the proportion of crows’ dependence on human-related food source is even greater. The high frequency of artificial substances in jungle crows’ diets suggests a greater availability of anthropogenic food sources in Tokyo. Therefore, it is highly probable that the carrying capacity of jungle crows was unnaturally raised, contributing to an increase of this species in urban landscapes. If food scraps in Tokyo could be reduced, a significant proportion of food currently

available to the crows would be lost. Since no life can sustain itself far above the carrying capacity, the population of crows would follow suit.

Jungle crows feed on insects such as Scarabedae beetles and mantis in winter when insects are not active. These insects are common in a wide range of habitats and are probably food sources for jungle crows in a natural state. The beetles are primary consumers of plants, and mantis is a predator in the food web. The predation of these insects has a probability of increasing biodiversity at the levels of producers and primary consumers of forest eco-systems （Kullberg & Ekman, 2000）. Control of one organism will inevitably affect the other organisms in the food web, and history warns us against excessive hunting/controlling of crows because their removal has resulted in a population explosion among insects, such as beetles and grasshoppers, greatly damaging the regional economies (Forbush, 1927; Inugai & Haga, 1953).

During the breeding season, jungle crows are known to prey on the nests and young of other bird species, including larger birds of prey in Tokyo (Ueta,

Table 2 List of food items of jungle crows found by direct observation (Kurosawa, 1999; with permission from WBSJ).

Plant matter Fruits (including berries, nuts and beans)

Eriobotrya japonica (Rosaceae) Diospyros kaki (Ebenaceae) Viburnum odoratissimum (Caprifoliaceae) Rhus succedanea (Anacardiaceae) Aphananthe aspera (Ulmaceae) Sweda controversa (Cornaceae) Magnolia praecocissima (Magnoliaceae) Gleditsia japonica (Leguminosae)

Animal matter Mammal Rats & mice (Muridae) Cats (Felis catus) (young) Bird Rock Doves (Columba livia)

Grey Starlings (Sturnus cineraceus) (young & eggs) Tree Sparrows (Passer montanus) (young & eggs)

Others Land snails (Gastropoda) Earthworms (Oligocchaeta)

Fig. 6 Different frequencies of artificial substances found in

pellets of jungle crows in urban (Tokyo) and rural (Tsuruoka and Ebetsu; Suzuki, 2002; Kondo, 2003) areas.

198 R. KUROSAWA et al. 1998). In winter, however, bird remains were not found in the pellets in our study, suggesting that the crows capture eggs and chicks of other birds primarily during the breeding season, probably to meet the demand of growing young. Thus, jungle crows were not a major predator on small bird species in winter in our study sites.

The natural inorganic matter such as pebbles and clay found in crows’ pellets in our study supports previous reports of their frequency in the pellets of crows (Liu et al., 1986; Suzuki, 2002; Kondo, 2003). Among the hypotheses is that crows deliberately take in ashes for nutrition (Takenaka, 2003). An analysis of the nutrient values and experiments on the importance of inorganic matter will be required to determine this issue.

Acknowledgements Shigeharu Koshikawa helped us identify plant

seeds. Nanae Kato lent us a microscope. Michio Matsuda as well as the assistant researchers of the Wild Bird Society of Japan cooperated in collecting pellet and fecal samples at crow roosts. Kazuji Fukui translated an article into Japanese. The Hitachi Central Research Center, park offices of Rikugien, Mizumoto and Tama Zoological parks gave us permission to conduct research at these sites. Takashi Kurosawa read and gave advice on the early manuscript. We would like to express our sincere gratitude to all these people.

References

Asano, S. (1996) Genshoku Zukan; Mebae to Tane (Colored guide to the seeds and seedlings. Zenkoku Noson Kyoiku Kyokai, Tokyo. (in Japanese)

Forbush, E.H. (1927) Birds of Mass. and other New England States, Vol. 2.

Higuchi, H. (1979) Habitat segregation between the Jungle Crow and the Carrion Crow (Corvus macrorhynchos, C. corone), in Japan. Japanese Journal of Ecology, 29:353-358.

Ikeda, S. (1957) Karasuka ni zokusuru chourui no shokusei ni tsuite (On the food of Corvus species). Forestry Agency Report on Wildlife, Tokyo. (in Japanese)

Inugai, T. and R. Haga (1953) Hokkaido niokeru karasu no higai to bojono kenkyu (III) Tokuni karasu no shokusei to nogyo tono kankei (A study on crows concerning agricultural impact and guarding of the crop (III) -Relation between crows diet and agriculture). Report of Hokkaido University Agriculture Department, 1(4):459-484. (in Japanese)

Iwasa, M.A., A.P. Kryukov, R. Kakizawa and H. Suzuki (2002) Differentiation of Mitochondrial Gene of Jungle Crow Corvus macrorhynchos (Corvidae) in East and South Asia. Journal of Yamashina Institute for Ornithology, 34:66-72.

Karasawa, K. (2003) Karasu wa dorehodo kashikoika (How wise are crows?). Chukobunko, Tokyo. (in Japanese)

Kawachi, H. and M. Matsuda, eds. (1999) A report on the symposium “How should we deal with the crow problems in Tokyo? (I).” Tokyo Chapter, Wild Bird Society of Japan, Tokyo. (in Japanese with English summary)

Kondo, T. (2003) The diet of Carrion and Jungle Crows Corvus corone and C. macrorhynchos living in Nopporo Forest Park and its surroundings during the breeding season. Bachelor’s thesis, Rakuno Gakuen University, Ebetsu, Hokkaido.

Kullberg, C. and J. Ekman. (2000) Does predation maintain tit community diversity? OIKOS, 89:41-45.

Kurosawa, R. (1999) The life of Jungle Crows (Corvus macrorhynchos) in Inokashira Park. In: H. Kawachi and M. Matsuda, eds., How should we deal with the crow problems in Tokyo? (1), Tokyo Chapter, Wild Bird Society of Japan, Tokyo, pp.16-31. (in Japanese with English summary)

Kurosawa, R., M. Narusue, H. Kawachi and K. Suzuki (1999a) The relationship between Jungle Crows Corvus macro-rhynchos and garbage in Tokyo. Strix, 18:71-78. (in Japanese with English summary)

Kurosawa, R., M. Narusue, H. Kawachi and K. Suzuki (1999b) Crows and garbage in Tokyo (a preliminary report, June 1999. In: Kawachi, H. and M. Matsuda, eds., How should we deal with the crow problems in Tokyo? (1), Tokyo Chapter, Wild Bird Society of Japan, Tokyo, pp.13-22. (in Japanese with English summary)

Kurosawa, R., M. Narusue, H. Kawachi and K. Suzuki (2001) Relation of Jungle Crows and garbage in Tokyo II – Comparison between summer and winter seasons. Strix, 19:71-79. (in Japanese with English summary)

Kurosawa, R., Y. Kanai, M. Matsuda and M. Okuyama (2003) Conflict between Humans and Crows in Greater Tokyo: – Garbage Management as a Possible Solution –. Global Environmental Research, 7(2): 139-147.

Liu, H.J., J.Q. Ren and Y.E. He. (1986) A basic research on the diet of Jungle Crows. Journal of Zoology, (2):7-10. (in Chinese)

Madge, S. and H. Burn (1994) Crows and Jays. Helm A & C Black, London.

Ministry of the Environment. (2001) Jichitai tantosha notameno karasu taisaku manual (Crow management manual for the administrators). Ministry of the Environment, Tokyo. (in Japanese)

Morishita, E., Kiyoshi Itao and H. Higuchi. (2003) Movements and habitat use of crows in urban areas, based on PHS tracking. Global Environmental Research, 7(2): 181-191.

Nakane, T., J. Aoki and R. Ishikawa. (1992) Common Insects of Japan in Color. Hoikusha Publishing Co. Ltd, Osaka. (in Japanese)

Suzuki, Y. (2002) Hashibosogarasu Corvus corone and Hashibutogarasu C. macrorhynchos no shokusei oyobi shoka koso (Diets and digestional enzymes of Corvus corone and C. macrorhynchos). Master’s thesis, Yamagata University, Tsuruoka City, Yamagata. (in Japanese)

Takenaka, M. (2003) Crows Problems in Sapporo Area. Global Environmental Research, 7(2): 149-160.

Ueda, K. and N. Fukui. (1992) Fruit food of crows Corvus corone and C. macrorhynchos: Analysis of seeds collected at the pre-roosting sites. Japanese Journal of Ornithology, 40(2):67-74. (in Japanese with English summary)

Ueta, M. (1998) Crow-related low nesting-success of small open-nesting birds in Tokyo area. Strix, 16:67-72. (in Japanese with English summary)

Ueta, M., R. Kurosawa, S. Hamao, H. Kawachi and H. Higuchi. (2003) Population Change of Jungle Crows in Tokyo. Global Environmental Research, 7(2): 131-137.

Yamagishi, S. (1982) Pellet no shirabe kata (How to study pellets). In: Haneda, K., ed., Yacho no seikatsu (The life of birds), Tsukijishokan, Tokyo, pp.169-172. (in Japanese)

Yamatokeikokusha. (2000) Yamakei handy zukan, series 3-5 (Yamakei handy guide to tree flowers). Genshokuzukan. Yamatokeikokusha, Tokyo. (in Japanese)

(Received 31 October 2003, Accepted 2 December 2003)