Length of the Larval Cycle of Ascaphus truei in Coastal Streams of the Redwood Region, Northern California

Abstract

-We studied length of the larval period of tailed frogs (Ascaphus truei) by examining length frequency distributions of larvae collected from late 1993-1996 in nine streams within the redwood zone of north coastal California. Three streams had populations with a 1 yr larval cycle, four streams with a 2 yr cycle, and two streams with populations that switched from a 2 to a 1 yr larval cycle during our study. Of 14 additional streams sampled twice in 1996, 11 had populations with a 1 yr cycle and three had a 2 yr cycle. In populations with a 1 yr larval cycle, the youngest cohort was first observed in late fall, grew during winter and spring, and completed metamorphosis by late August-early September. No larvae were found during the fall. Streams with a 2 yr larval cycle contained larvae throughout the year, were significantly higher in elevation, and had significantly greater discharge than those with a 1 yr cycle. Summer water temperatures, stream aspect, and cover type were not related to length of larval cycle. We suggest that stream surveys for presence and abundance of Ascaphus truei be completed by early August in streams of the coastal zone of northern California. l f erpetology, Vol. 32, No. 3, p. 40 -409, 1998 i t 1998 Society for the Study of Amph bians and Reptiles the Larval Cycle of Ascaphus truei in Coastal Streams of ood Region, Northern California The tailed frog, Ascaphus truei, is unique among North American anurans in being specialized for life in cold, clear, mountain streams (Nussbaum et al., 1983). It occurs in the Pacific Northwest from southern British Columbia south to northwestern California from the Cascade Mountains west to the coast (Metter, 1968). Disjunct populations also occur in the Blue Mountains of southeastern Washington/northeastern Oregon and in the Rocky Mountains of northern Idaho and western Montana (Nussbaum et al., 1983). Length of the larval period shows variation with respect to geographic location and altitude. Tailed frog larvae from northern and inland populations spend two to three years in streams before metamorphosing into juveniles (Metter, 1964, 1967; Dougherty and Sheldon, 1982; Bull and Carter, 1996). In contrast, larvae from a ile frog, Ascaphus truei, is unique orth A erican anurans in being speor life in cold, clear, mountain streams et al., 1983). It oc urs in the Pacif c 3 To whom correspondence should be addressed. montane population in northwestern Washington required four years to complete developent and finished metamorphosis in their fifth year (Brown, 1990). Brown (pers. comm.) subsequently compared lowland (elevation = 152 m), montane (366-853 m), and subalpine (1067 m) populations in northern Washington and found that most lowland tadpoles required two to three years to complete embryonic and larval life while montane and subalpine populations required three to four years before beginning metamorphosis. Bury (pers. comm.), sampling coastal streams in the Olympic Mountains of Washington and the Oregon Coast Range, found that northern sites had tadpoles with a 2 yr larval cycle, but the three most southerly sites apparently had Ascaphus truei with a 1 yr cycle. Although Metter (1967) included one population near the California border, no other reports of studies of larval cycles of this species exist from coastal California streams. Based on . 404 This content downloaded from 157.55.39.4 on Thu, 08 Sep 2016 05:48:32 UTC All use subject to http://about.jstor.org/terms LARVAL ASCAPHUS TRUEI the nature of the variation reported in previous studies, we predicted that the coastal regions of northern California should have relatively short larval periods. During the fall of 1992 and 1993, we noted that larval Ascaphus truei could not be found in some streams in this area, suggesting that larval cycles might be as short as one year. In December 1993, we began this study on selected streams to determine the length of the larval period of Ascaphus truei. In addition, we provide information on the timing of metamorphosis of larval A. truei and give a preliminary assessment of factors that may control the length of the larval period in streams of the coastal region of northern California. MATERIALS AND METHODS Study Area.-This study was conducted on lands owned by Simpson Timber Company in western Humboldt and Del Norte Counties, California, in the north coast redwood (Sequoia sempervirens) zone. Mean summer and winter temperatures are about 18 C and 5 C, respectively, and precipitation varies from about 130254 cm annually, 90% falling from October through April (Elford, 1974). Coast redwood and Douglas-fir (Pseudotsuga menziesii) are codominant conifers in the study area, but Douglas-fir is more prevalent at higher elevations. Common riparian species include red alder (Alnus rubra), big leaf maple (Acer macrophyllum), and willows (Salix spp.). We began sampling in December 1993 in three low-elevation (122-171 m) tributaries of the Mad River (Jiggs, Blackdog, and Pollock Creeks). Six additional streams, tributaries to Redwood Creek, Mad, and Little Rivers, at varied elevations (266-732 m), were selected for survey from 1994-1996 to document the range in variation of life history traits and assess environmental factors that may influence any differences. Streams were sampled five to eight times a year depending on accessibility and suitability for sampling (adequate water clarity and moderate to low discharge). In 1996, as part of a habitat study for tailed frogs, we sampled 14 additional streams during July and early August and again during October and November. These streams were tributaries to the Mad River, Redwood Creek, Klamath River, Smith River, and one flowed directly into the Pacific Ocean. Methodology.-Larval Ascaphus truei were collected with the aid of a viewing box and an aquarium net. Rocks were overturned and the substrate beneath gently disturbed by hand to find the larvae. For each larva, total length (TL, tip of snout to end of tail) was measured to the nearest mm and developmental characteristics were recorded as follows: tadpole with or without hind limb buds (limbs 3 mm, TH); and metamorphosing tadpoles with both front and hind limbs and obviously reduced tail (TM). All larvae were returned to near their place of capture. We usually searched 30-60 m of stream or until we collected 15-25 larvae. Occasionally larger samples were collected if there was large variation in larval sizes. We recorded water temperature to the nearest 0.1 C at the time of sampli g with a Schultheis thermometer. In February 1996, we placed electronic thermograph units (HOBO data loggers, Onset Computer Corp.) in the three low elevation streams to record annual water temperature profiles (nearest 0.2 C). Additional units were placed in the other six streams by June, and in the 14 additional streams in August to record water temperatures during the warmest time of the year. Stream discharge (cm3/sec) was estimated in all but one stream (from mid-August to mid-September) by measuring the stream width (cm), depth at this point at /4, Y, and 3 intervals across the stream (Platts et al., 1983), and timing the surface speed of a small floating object for three trials. Elevation (m), stream aspect (azimuth in degrees), and cover types (redwood, Douglas-fir, redwood/Douglas-fir mix, and Douglas-fir/hardwood mix) were taken from a Geographic Information System (G.I.S.) data base maintained by Simpson Timber Company and aerial photographs (1:12,000 scale). We prepared length frequency distributions of larval Ascaphus truei from capture data and followed each size class of larvae through time. We especially noted the presence or absence of larvae during fall (September through November) sampling. To develop a general picture of the size of larvae in their first year of life from streams with 1 and 2 yr cycles, we combined samples over three years from three streams with a 1 yr cycle and over two-three years from four streams with a 2 yr cycle. We selected the period of the highest average water temperatures (17 August through 2 September) and calculated an overall mean daily water temperature (380 records) and a mean daily maximum water temperature (19 records) for each stream. ANOVA (NCSS, Version 6.0, Hintze, 1995) was used to determine if mean elevation, summer and maximum water temperatures, and stream discharge differed between streams with 1 and 2 yr larval cycles. A Chi-square test (NCSS, Version 6.0) was used to see if aspect (divided into four equal quartiles) and cover types differed between the streams with different larval cycles.