Rana catesbeiana was introduced into California between 1914 and 1920 and has since spread throughout the state. In the San Joaquin Valley it has become the dominant frog on the valley floor and has spread into the Sierra Nevada foothills. It is most abundant in the warm low elevation pools of the foothill streams, in areas heavily altered by man, although at least two populations are established above 1600 m elevation. Of the two frog species native to this region, R. aurora is either absent or very rare at the present time, while R. boylii is found mostly in small permanent foothill streams higher than 200 m elevation, in areas not occupied by R. catesbeiana. The disappearance of R. aurora from the region, and the continuing reduction in range of R. boylii, is attributed to habitat alteration coupled with predation and competition from R. catesbeiana. HE bullfrog (Rana catesbeiana Shaw), native to the United States east of the Great Plains, was introduced into California several times between 1914 and 1920 (Storer, 1922). After the initial introduction, it was spread rapidly throughout the state by wellmeaning naturalists and farmers (Storer, 1925). Unfortunately, in the San Joaquin Valley and the Sierra Nevada foothills, one of the main by-products of the introduction of the bullfrog seems to have been the elimination of the red-legged frog (Rana aurora Baird and Girard) from the valley floor and foothill ponds and the reduction of populations of the foothill yellow-legged frog (R. boylii Baird) in the foothill streams. This paper shows how the distributions of R. catesbeiana and R. boylii have become nearly mutually exclusive in the Sierra Nevada foothills and attempts to document the disappearance of R. aurora from the San Joaquin Valley, concomitant with the spread of the R. catesbeiana. METHODS Between 27 July and 4 September 1970, the distribution and ecology of the fishes occurring in the streams of the Sierra Nevada foothills below elevations of 1100 m were studied (Moyle and Nichols, in preparation). At each locality where fish were collected, a rough estimate was also made of abundance of each frog species present, on a 0-3 scale. On this scale, '0' indicated no frogs present, 1, only one or two individuals observed in sample area, 2, frogs present in low numbers, (usually, less than 12 observed), and 3, frogs abundant. The frog abundance was thus rated for most of the streams accessible by road in Tulare, Fresno, Madera, Mariposa, and Tuolumne counties, California. Frogs were found at 95 of the 130 stream localities. During the study period the streams were at their minimum flow. Over half of them were not flowing at all, and water was present only in isolated pools. At each sampling site estimates and measurements were made of environmental factors that were likely to affect the distribution of the fishes, as discussed in detail in Moyle and Nichols (in preparation). The following were measured: 1) Elevation, in meters. 2) Air and water temperature. Since
Read full abstract