Earth mounds up to about 2 m in height and 50 m in diameter occur in densities up to about 50 per ha in many locations in western North America. Often termed Mima mounds because of their conspicuous development on Mima Prairie, Thurston Co., Washington, these mounds are known from the Puget Sound lowlands, the central valley of California (where the terrain is commonly known as hog-wallows), coastal marine terraces and cismontane valleys of Southern California, the Columbia and Snake River Plateaus of the interior Pacific Northwest, various valleys and plateaus of the Rocky Mountain region, and other isolated locations (Scheffer, 1947; Price, 1949). Mounds of similar form, usually termed prairie or pimple mounds, occur from Saskatchewan to Minnesota and south through Iowa and Missouri to Arkansas, eastern Oklahoma, and the coastal plain of Louisiana and Texas (Price, 1949; Ross et al., 1968; Mollard, 1982). Numerous hypotheses have been suggested for the origin of these mounds, most involving physical processes of soil deposition and erosion (Krinitzsky, 1949; Newcomb, 1952; Scheffer, 1947; Vitek, 1978). The most highly regarded biological hypothesis, formulated by Dalquest and Scheffer (1942), is that these mounds are the gradual product of the backward displacement of soil that results from outward tunneling of pocket gophers from sites representing territory centers and the locations of permanent nest and food storage chambers. According to this hypothesis the essential descriptive features of Mima mounds are that they (1) are composed only of textural materials that such animals can move, (2) are restricted to thin or poorly drained soils that limit tunneling to a shallow surface zone, (3) are confined to flat or moderately sloping land, (4) are nearly circular in outline regardless of slope, (5) tend to be evenly spaced because of the strong territoriality of pocket gophers, and (6) occur only within the historical range of these animals (or forms with similar characteristics). While the above features are not absolutely inconsistent with hypotheses of origin of mounds by depositional and erosional mechanisms, no hypothesis of this type predicts that Mima mounds will possess all of these characteristics and limitations. Mound areas that have been examined in detail in western North America conform to these descriptive criteria (Scheffer, 1947; Mielke, 1977; ,Cox, 1983). Mima mounds are so far known only from areas west of the Mississippi River. Pocket gophers occupy only limited ranges east of the Mississippi River, Geomys bursarius entering parts of Wisconsin, Illinois, and Indiana from the west, and G. pinetis occurring in parts of Alabama, Georgia, and Florida. Scheffer (1958) suggested that one test of the hypothesis of mound formation by fossorial rodents would be a search for Mima-type mounds in other parts of the world, especially those where animals with characteristics similar to those of pocket gophers are found. In Kenya, East Africa, earth mounds up to 6 m in diameter and 1.5 m in height have been described at elevations of 1,950-3,600 m on Mt. Kenya, and the occupation and possible formation of some of these mounds by fossorial rodents mentioned (Fries and Fries, 1948; Coe, 1969; Jarvis and Sale, 1971). We have recently observed such mounds in several grassland areas at elevations of 2,250-2,440 m on Mt. Kenya, in the Aberdare Mountains, and along the eastern edge of the Mau Escarpment; in July 1981 we studied in detail two such areas near Nyahururu, Kenya, in the Aberdare highlands (Cox and Gakahu, 1983). All of these localities lie at elevations above those at which species of fungus-gardening termites (genera Macrotermes and Odontotermes), which are definitely known to produce large mounds, have been reported (Hesse, 1955; Glover et al., 1964; Pomeroy, 1977). All lie within the geographical and altitudinal range of the rhizomyid mole rat, Tachyoryctes splendens (Riippell), a rodent convergent in morphology and behavior with pocket gophers (Rahm, 1971; Kingdon, 1974). The site examined in greatest detail, a fenced pasture 1.3 km E Nyahururu, elev. 2,375 m, exhibited nearly circular mounds up to 2.0 m high and 17.8 m in diameter; the mean density of mounds was 30.5 per ha (Cox and Gakahu, 1983). Analysis of distance-to-nearest-neighbor relationships (Clark and Evans, 1954) gave a dispersion index, R, of 1.22, indicating a significant tendency toward uniform spacing of mounds. The mounds were underlain by a horizontal laterite rock hardpan that varied in depth from a mean of 38.6 cm at mound edges to a mean of 24.6 cm at the centers of intermound basins. At the second February 1984 149
Read full abstract