items in the PR. Differences for these species (and also for pocket mice Perognathus sp., occurring more often in the MER) were statistically significant. The number of seats containing squirrels (Spermophilus sp.) and white-footed mice (Peromyscus sp.) were low in both areas and did not differ statistically (Table 1).
Reptiles and arthropods were eonsumtd rather more often in the P1k, but scarcely or not at all in the MBR. By contrast, fruits were consumed in the MBR and not in the PR. In all cases, the differences were highly significant (Table 1). In the PR, snakes (including nine Croralus) occurred in 19 seats, lizards in 27, and unidentified reptiles in eight. In the MBR we only detected reptiles in two scats (Table 1). Arthropod remains were also a usual component of coyote scats in the PR. Most of the samples included imagos of Coleoptera
(21 scars), while others had larvae (12; mainly of Coleoptera and Lepidoptera) and scorpions (9). On some occasions Orthoptera reach very high densities in the MBR and coyotes include them in their diet (authors, unpubi.), but we did not detect any arthropods in the MBR autumn scats analysed. Vegetative items (excluding some occasional grasses) occurring in the seats of MBR included remains of prickly-pear and mesquite fruits (10 and eight stats respectively). Birds, ungulates (cattle and deer, probably carrion) and garbage occurred in low proportion in both areas and did not differ statistically between them (Table I).
The remarkable role of Iagomorphs in the diet of coyotes in both deserts, despite their different abundance, suggests that they are selected by coyotes as a preferred prey. The importance of leporids to many coyote populations has been documented by Clark (1972) and Johnson & Hansen (1979), among others. Additionally, Short (1979) presented evidence that leporids were the main food item for coyotes in an area oI Arizona which is rather similar to the PR.
Rodents were probably consumed in proportion to their availability; whereas bipedal
kangaroo rats are more abundant in the Sonoran Desert, where rainfall and plant cover are very low, while quadrupedal rodents (Neoroma sp., Peromyscus sp.) are more common
in the less arid region of the Chihuahuan Desert (see Kotler, 1984; A. Gonzalez, pers. comm.). A high occurrence of leporids and rodents and a low occurrence of ungulates are
cited in some other studies of coyote food habits in semi-arid regions of the U.S.A. (Johnson & Hansen, 1979; Short, 1979; Andelt, 1985).
Som& of our results could be indicative of certain latitudinal trends in the feeding of
coyotes. So, Knowlton (in Voigt & Berg, 1987) reported a general north to south increase in the availability of small prey in North America. However, we think that some of the trends detected reveal mainly the general features of predation by vertebrates in hot arid environments. This may explain the
high proportion of reptiles and arthropods in coyote diet in the PR.
To our knowledge, the results in the PR represent the highest occurrence of reptiles ever previously detected in coyote scats. Young (1951) referred to the occurrence of ‘fish, frogs and reptiles’ in only 116 out of more than 50,000 coyote stomach contents analysed (approx. 02%) while Bekoff (1982) and Voigt & Berg (1987) did not refer to reptiles as prey of coyotes in their revisions of the species ecology. In the
same way, reptiles occurred in about 10% of the seats of the black-backed jackal (Canis mesomelas) and in 12% of those of the Cape fox (Vulpes chama) in the Namib Desert (Bothma a al., 1984). They are cited as a common prey of almost all the canids in the Sahara Desert (Le Berre, 1990), while it is difficult to find reptiles as the prey of canids in more mesic environments. When comparing the food habits of the eagle owl Bubo bubo and the great homed owl Bubo virginianus in different Palearctic and Nearctic biomes, however, Donázar a al. (1989) related their diets in deserts to the presence of reptiles and arthropods as prey. Moreover, in a pioneer study on the ecology of the Saharan birds5 Valverde (1957) referred to the
important role of reptiles as the food of the diurnal and nocturnal birdsof prey, staring that
‘in hot climates, vegetarian reptiles play the same role as prey of carnivores than rodents in
The relevant ecological function of reptiles in hot desert foodwebs is possible because of the high species density and abundance of reptiles in these areas, due to the increased insolation (Schall & Pianka, 1978) and the high production efficiency of small ectotherms
under these environmental conditions (Turner at al., 1976).
Arthropods are also an important component of desert faunas [for instance, in the Negev
Desert, density biomass of isopods is 400 times that of hares (Lapin europaeus) and 14 times that of jerboas (Jaculusfaculus) (Steinberger, 1991)] and a relevant prey group for desert predators. So, two canids in the Namib Desert (the aardwolf, Proteles cristatus, and the bat-eared fox, Otocyon megalotis) are specialized for the consumption of insects, while Scorpionidae are usually consumed by other canid species (Bothma at at., 1984). As indicated previously, both species of great owls (Bubo bubo and B. virginianus) often capture arthropods in the Palaearctic and Nearctic deserts (Donäzar at at., 1989). Valverde (1957) also found invertebrates (mainly Tenebrionidae and Acrididae) to be common prey of the Saharan carnivores, from foxes (Vulpes spp.) to hyaenas (Hyaena brunnea). By contrast, in the analysis of more than 50,000 coyote stomachs from California, Young (1951) referred to the presence of invertebrates only in about P1% of the samples. In the same way, scarcity of fruit in the diet of the Pinacate coyotes could be typical of extreme deserts, although Valverde (1957) reported a fruit (Rhus typartirum) as the May—June staple food of jackals (Canis aureus) in some areas of the Sahara.
In conclusion, we think that some of the differences detected in the diet of coyotes in MapimI and Pinacate, especially the major role of reptiles and invertebrates as prey, and probably the low importance of fruit and the high trophic diversity, may be characteristic features of the trophic webs in the most arid hot ecosystems, which therefore have longer food chains than usually thought.
The study was supported by the Instituto de Ecologla A.C. (projects ‘Mapimf’ and ‘P
inacate’) and through an agreement between the Spanish CSIC and the Mexican CONACYT for research on Biological Reserves. A. L. Herrera, A. Herrera, F. Herrera, R. RuIz de Esparza, F. Pizarro, N. Bustamante and B. Ezcurra helped us in many ways.
Andelt, W. F. (1985). Behavioral ecology of coyotes in south Texas. WildlifeMonographs, 94: 1—45. Bekoff, M. (1982. Coyote (Canis latrans). In: Chapman, J. B. & Feldhaxnmer, G. A. (Eds), Wild
Mammals of North America, pp. 447—459. Baltimore: John Hopkins University Press. 1147 pp. Bothma, J. du P., Nd, J. A. J. & Macdonald, A. (1984). Food niche separation between four
sympatric Namib Desert carnivores. Journal of Zoology, 302: 327-340.
Clark, B. W. (1972). Influence of jackrabbit density on coyote population change.JournalofWildlife
Management, 36: 343—356.
Cloudsley-Thompson, J. (1976). Deserts and Grasslands. I. Desert life. New York: Doubleday, Garden City. 144 pp.
Cornet, A. (1988). Principales caractéristiques climatiques. In: Montana, C. (Ed.), Esiudiolntegrado
de los recursos vegetacion, sueloy agua en la Reserva de la Biosfera de Mapimi. I. Ambiente naturaty
humano, pp. 45—76. Mexico: Instituto de Ecologia. 290 pp.
Delibes, M., Hernández, L. & Hiraldo, F. (1989). Comparative food habits of three carnivores in western Sierra Madre, Mexico. Zeitschrift fib- Saugederkunde, 54: 107—110.
Donizar, J. A., Hiraldo, F,, Delibes, M. & Estrella, R. R. (1989). Comparative food habits of the
Eagle Owl Bubo bubo and the Great Honed Owl Bubo virginianus in six Palearctic and Nearctic
biomes. OrnisScandinavica, 20: 298—306.
Erliage, S. (1968). Food studies on captive otters (Lutra lutra L.). Oikos,19: 81—98.
Bzcurra, E. & Rodriguez, V. (1985). Rainfaff patterns in the Gran Desierto, Sonora, Mexico.
Journal of Arid Environments,10: 13—28.
Johnson, M. K. & Hansen, R. M. (1979). Foods of•coyotes in the Lower.Grand Canyon, Arizona.
Journal of Arizona Academy of Sciences, 12: 81—83.
Kotler, B. P. (1984). Risk of predation and the structure of desert rodent communities. Ecology,
Le Berre, M. (1990). Faune du Sahara. II. Mammiferes. Paris: Lechevalier—Chabaud. 360 pp. Macmalion, J. A. & Wagner, F. H. (1985). The Mohave, Sonoran and Chihuahuan deserts of North
America. In: Evenari, M., Noy-Meir, I. & Goodall, D. W. (Eds), Ecosystems of she World. 12A.
Hot deserts and arid shrublands, pp. 105—202. Amsterdam: Elsevier. 365 pp.
Margalef, R. (1957). La teoria de la información en Bcologia. Memories de la Real Academia de
CienciasyArtes de Barcelona, 32: 373—449.
Niebauer, T. J. & Rongstad, 0. J. (1977). Coyote food habits in northwestern Wisconsin. In:
Phillips, R. L. & Jonkel, C. (Eds), Proceedings of the 197S Predator Symposium. pp. 237—251.
Missoula: Montana Forest and Conservation Experiment Station, University of Montana. 268 pp. Perez-Gutierrez, C., Fierro, L. C. & Trevino, J. C. (1982). DeterminaciOn de Ia composición de la dieta del coyote (Canis latrans Say) a través del afloen la region central de Chihuahua por medio del
análisis de contenido estomacal. Pastirates, 13: 2—15.
Schall, I. J. & Pianka, E. K. (1978). Geographical trends in numbers of species. Science, 201: 679—
Serrano, V. & Sosa, V. (1985). Preliminary study of desert black-tailed jackrabbit (Lepus californicus
Gray) and desert cottontail (Sylvilagus auduboni Baird) in the BolsOn de MapimI, Durango,
Mexico. Fourth International Theriological Congress, Edmonton, Alberta, Canada (unpubl. poster).
Short, H. L. (1979). Food habits of coyotes in a semidesert grass-shrub habitat. Forest Service U.S. Department of Agriculture. Research note RM-364: 1—4.
Sokal, R. R. & Rohlf, F. J. (1981). Biomet,y (2nd Edn). San Francisco: W.H. Freeman. 859 pp. Steinberger, Y. (1991). Biology of arid region soils. Faunal components. In: Skujins, J. (Ed.),
Semiarid Lands and Deserts. Soil resource and reclamation, pp. 173—192. New York: Marcel
Dekker. 668 pp.
Turner, F. B., Medica, P. A. & Kowalewsky, B. W. (1976). Energy Utilization by a Desert Lizard
(Uta stansburiana). U.S.II.B.P. Desert Biome Monographs, No. 1. Logan: Utah University
Valverde, J. A. (1957). Ayes del Sahara Espanol (Estudio ecologico del desierto). Madrid: Instituto de
Estudios Africanos. Consejo Superior de Investigaciones Cientificas. 487 pp.
Vela, C. B. L. (1985). DeterminaciOn de Ia composiciOn de Ia theta del coyote Canis latrans Say, por
medio del análisis de heces en tres localidades del estado de Chihuahua. B.Sc. thesis, University
Auton, Nuevo LeOn, Monterrey, Mexico.
Voigt, D. R. & Berg, W. B. (1987). Coyote. In: Novak, M., Baker, J. A., Obbard, M. B. & Malloch, B. (Eds), Wild Furbearer Management and Conservation in North America, pp. 345—357. Ontario: Ministry of Natural Resources. 1150 pp.
Young, S. P. (1951). The clever coyote. I. Its history, life habits, economic status and control. In: Young, S. P. & Jackson, H. H. T. (Eds), The Clever Coyote, pp. 1—226. Washington D.C.:
Wildlife Management Institute. 411 pp.