Taricha rivularis
Red-bellied newt, Redbelly newt
Subgenus: Twittya
family: Salamandridae
subfamily: Pleurodelinae

© 2009 Mendocino Redwood Company (1 of 38)

View distribution map using BerkeleyMapper.

Conservation Status (definitions)
IUCN (Red List) Status Least Concern (LC)
See IUCN account.
NatureServe Status Use NatureServe Explorer to see status.
Other International Status None
National Status None
Regional Status None


bookcover The following account is modified from Amphibian Declines: The Conservation Status of United States Species, edited by Michael Lannoo (©2005 by the Regents of the University of California), used with permission of University of California Press. The book is available from UC Press.

Taricha rivularis Twitty, 1935
Red-Bellied Newt

Sharyn B. Marks1
Darrin Doyle2

1. Historical versus Current Distribution. Red-bellied newts (Taricha rivularis) have the most limited geographical distribution among the three species of Taricha. They occur in coastal California north of San Francisco Bay, in Sonoma, Lake, Mendocino, and Humboldt counties, at elevations between 150–450 m. Their distribution is roughly similar to that of the coast redwood (Sequoia sempervirens), but red-bellied newts are not restricted to redwood forests (Myers, 1942b; Stebbins, 1951; Riemer, 1958; Twitty, 1964a). There have been no studies comparing the present distribution with the historical distribution.

2. Historical versus Current Abundance. Twitty (1935) described this species as being “very abundantly represented” in mountain streams in Mendocino County. Abundance of red-bellied newts has been quantified only for one locality, Pepperwood Creek (a tributary of the Wheatfield fork of the Gualala River) in Sonoma County. Twitty (1959) remarked that “Pepperwood Creek is a small stream, but it is literally crawling with newts during the breeding season.” Using census data from Twitty (1961a, 1966), Hedgecock (1978) estimated that 58,000–60,000 breeding adults occurred in the roughly 2.5-km-long segment of Pepperwood Creek intensively studied by Twitty and collaborators between 1953 and 1961. This creek was chosen in part because of the high abundance of red-bellied newts, so it may not be representative of newt abundance at other localities. No data are available on the current abundance of this species at any locality.

3. Life History Features.

A. Breeding. Reproduction is aquatic.

i. Breeding migrations. Breeding migrations are initiated after a period of terrestrial foraging (see "Seasonal Migrations" below). Beginning in late January, movements become less random, and some individuals begin moving toward the stream; thereafter, the magnitude of the migration increases until it peaks in early March (Packer, 1960). Red-bellied newts do most of their migrating during cloudy and rainy nights (Grant et al., 1968). In general, rainfall triggers the movement of animals toward the stream, but in heavy amounts it inhibits migration to streams. In the absence of rainfall, the magnitude of the migration is correlated with daily changes in mean evening temperature and minimum relative humidity; when both mean evening temperature and minimum relative humidity increase (relative to values of the previous day), captures increase (Packer, 1960). Adults enter streams only after the winter floodwaters have receded (Twitty, 1942). Males enter the water a considerable time before females, whose aquatic phase is restricted to a short period during mating and oviposition (Twitty, 1942, 1955; Packer, 1963).

Twitty’s (1959, 1961a, 1966) long-term mark-recapture studies at Pepperwood Creek demonstrated that red-bellied newts typically return to the same portion of the home stream for successive breedings, apparently throughout their life. Once in the stream, animals usually do not range > 15 m (50 ft) from the original entry point (Packer, 1963). Studies of artificially produced hybrids (T. rivularis x T. torosa and T. rivularis x T. t. sierrae) suggest that “imprinting” on the home area takes place sometime early in the life cycle, either during larval life or just after metamorphosis (Twitty, 1961a). This highly developed homing mechanism promotes isolation and genetic divergence among demes in streams (Hedgecock, 1978). Gene flow is likely low between demes within the same stream and extremely low between demes in different streams.

Twitty and colleagues conducted a series of experiments in which they displaced animals from their native stream segments to other portions of the same stream, or to other streams within the watershed (Twitty, 1959, 1961a; Twitty et al., 1964, 1966). These displacements demonstrated the extraordinary homing abilities of red-bellied newts—some are capable of homing to their native stream segment even after displacement distances as great as 8.0 km (5 mi). Displaced animals make the homing journey by moving overland, either parallel to the stream (after intra-stream displacement) or directly overland across intervening mountain ranges (after inter-stream displacement; Twitty, 1959; Twitty et al., 1966). Homing is mostly postponed until the animals leave the water at the end of the breeding season and may not really commence until the onset of the fall rains during the next breeding season (Twitty et al., 1964, 1966). In the case of inter-stream displacement, homing may be delayed following residence at the release site for ≥ 1 yr, but most displaced animals initiate homing the first year following displacement (Twitty et al., 1966). In a few instances, speed of homing has been documented; animals have homed for a distance of about 400 m (0.25 mi) in 1 d, and 4.0 km (2.5 mi) in < 1 mo (Twitty et al., 1966).

Of the various senses that might be involved in homing, olfaction is more crucial than sight. Red-bellied newts do most of their migrating during cloudy and rainy nights when celestial cues are not available. Permanently blinded animals, displaced about 0.8 km (0.5 mi) to another part of the stream, are able to find their way back to their home segments with about the same degree of success as sighted animals (Twitty, 1959, 1961a). Experiments in which the olfactory system was damaged surgically or chemically demonstrated that olfaction plays a major role in the ability of red-bellied newts to home to their native stream segments after being displaced (Grant et al., 1968). Olfaction is critical for both initial orientation as well as eventual return to the home site. Grant et al. (1968) proposed that terrestrial odors from vegetation surrounding the home stream segment might provide long-term olfactory cues for homing. Other experiments, which artificially altered the topography of the land, demonstrated that kinesthetic sense is not the mechanism by which migration is oriented (Twitty, 1959, 1961a). Displaced individuals apparently begin their search using oriented (rather than random) movements. Most animals displaced for distances up to 12.9 km (8 mi) oriented their departure from the release site in the homeward direction. However, orientation accuracy decreased with increasing displacement distance (Twitty et al., 1967a).

ii. Breeding habitat. Breeding occurs in mountain streams, where the substratum is clean and rocky (Twitty, 1935; Stebbins, 1951). The initiation of breeding and length of the breeding season depend on weather and stream conditions. Adults enter the water soon after the streams begin to recede from winter floods. The aquatic phase may last from late February to May, though the bulk of breeding occurs in March to early April (Twitty, 1955, 1966; Stebbins, 1985). Heavy rains and consequent flooding may temporarily drive animals from the water, which may extend the breeding season beyond April.

B. Eggs.

i. Egg deposition sites. Oviposition occurs in mountain brooks primarily between mid March and mid April, though the oviposition period may span from early March to early May. Eggs are attached in flattened clusters (generally one egg-layer thick) to the undersides of stones, usually in the middle of the stream in swiftly flowing water; occasionally flattened egg masses are attached to submerged roots along the side of the stream (Twitty, 1935, 1942). In suitable habitats, clutches may occur in great abundance; approximately 70 egg masses were found attached to the underside of a single stone, all within an area of 22 x 30.5 cm (8.5 x 12 in; Twitty, 1935).

ii. Clutch size. Mean clutch size is about 10 eggs (range 6–16 eggs), and mean egg diameter (measured at the blastula stage) is roughly 2.6 mm (range 2.4–2.7 mm; Twitty, 1935, 1964a; Riemer, 1958). Eggs typically are darkly pigmented (dark gray or grayish brown; though completely white eggs are not uncommon) and develop into normally pigmented larvae (Twitty, 1964a). As with other amphibian species, time from oviposition to hatching varies with environmental temperature. The incubation period of eggs raised in the laboratory at 15 ˚C ranged from 30–34 d, while eggs raised at 23 ˚C took only 16–20 d; developmental rates in the field were intermediate between these values (Licht and Brown, 1967). The total length of hatchlings ranges from 10.3–11.0 mm (Riemer, 1958) or 10.5–12.7 mm (mean = 11.8 mm; Twitty, 1964a).

C. Larvae/Metamorphosis.

i. Length of larval stage. The larval stage (from hatching to metamorphosis) lasts about 4–6 mo (Twitty, 1964a,b; Licht and Brown, 1967). Unlike congeneric larvae, pigmentation is distributed uniformly over the dorsum and sides (Twitty, 1935, 1936, 1942, 1964a; Riemer, 1958). Larvae possess a dorsal tail fin that is reduced and does not extend as far anteriorly on the trunk as that of congeners; balancers are either absent or rudimentary in hatchlings (Twitty, 1935, 1936, 1942, 1964a; Riemer, 1958). The more streamlined dorsal tailfin and absence or reduction of balancers have been interpreted as adaptations to a mountain brook habitat (Noble, 1931; Stebbins, 1951; Twitty, 1966). Hindlimb development is considerably advanced at hatching relative to the other species of Taricha; at the time of independent feeding, the digits on the forelimb are distinctly forming (Twitty, 1936). This also is probably an adaptation to life in moving water. Petranka (1998) categorized red-bellied newt larvae as “pond-type,” but his rationale for this designation is unclear.

ii. Larval requirements.

a. Food. No data are available on larval diet. However, like other larval salamanders (such as rough-skinned newts; T. granulosa), red-bellied newt larvae are undoubtedly generalist carnivores, consuming nearly any appropriate prey item that will fit in the mouth.

b. Cover. Stones and vegetation in the water provide retreat sites for larvae during the day. In the field, larvae were most abundant in microhabitats with temperatures between 22–26 ˚C (Licht and Brown, 1967).

iii. Larval polymorphisms. None reported.

iv. Features of metamorphosis. Metamorphosis occurs during late summer or early fall when larvae reach 45–55 mm in total length (Twitty, 1935, 1936, 1942, 1961a, 1964a; Licht and Brown, 1967). The duration of metamorphosis has not been quantified; however, Twitty (1964a) reported finding abundant larvae at the beginning of September, and only few larvae (all in the process of metamorphosing) < 6 wk later in Pepperwood Creek (Sonoma County, California). So, the metamorphic period for that population extends over about 6 wk, but the duration of metamorphosis for individual animals has not been reported. There is no evidence that larvae overwinter before transforming (Riemer, 1958; Twitty, 1964a).

v. Post-metamorphic migrations. Little is known about post-metamorphic migrations except that juveniles leave the stream and go into hiding underground just after metamorphosis (Twitty, 1955, 1961a, 1966; Twitty et al., 1967b; Petranka, 1998).

vi. Neoteny. None reported.

D. Juvenile Habitat. Juveniles are thought to remain in underground shelters almost continuously from the time of metamorphosis to sexual maturity (estimated to be approximately 5 yr or longer). During a 3-yr study at Pepperwood Creek, few juveniles were observed or captured in land traps during the rainy season when large numbers of adults were evident. Of 8,919 animals captured in land traps, only 1.7% of these were juveniles. This contrasts with the population of rough-skinned newts at Pepperwood Creek, in which adults and juveniles were captured in about equal numbers (Twitty, 1966; Twitty et al., 1967b).

E. Adult Habitat. Adults migrate from terrestrial to aquatic habitats seasonally for breeding. There are no detailed descriptions of terrestrial habitats, and what information is available is somewhat inconsistent between sources. Several sources (Myers, 1942b; Twitty, 1964; Riemer, 1958) state that this species’ range is confined to the coast redwood belt, but Riemer (1958) notes that red-bellied newts are not restricted to redwood forests, nor are they particularly abundant in that habitat. However, none of these authors specifically describe the terrestrial habitat for this species. Twitty (1966) comments that California laurel (Umbellularia californica) trees are common near his study site at Pepperwood Creek, but no other tree species are mentioned. Petranka (1998) states that red-bellied newts are found predominantly in redwood forests. I (S.B.M.) have observed terrestrial adults in forest dominated by Douglas-fir (Pseudotsuga menziesii), tan oak (Lithocarpus densiflorus), and madrone (Arbutus menziesii) in southern Humboldt County, and colleagues have seen them within redwood forest in Mendocino County (S. Sillett and J. Spickler, personal communication). Clearly, multiple forest types are used by this species. Adults use terrestrial sites for underground retreats during the dry season (May–October) and for foraging and migration prior to winter breeding. Both Twitty (1966) and Licht and Brown (1967) mentioned that red-bellied newts at their study sites (Pepperwood Creek and Skaggs Springs, respectively, both in Sonoma County) were found on steep, heavily wooded slopes that rise from the south bank of the breeding stream (i.e., north-facing slopes). Packer (1960) noted that at Pepperwood Creek, the banks and north-facing slopes are littered with many fallen trees and branches that provide cover for red-bellied newts and other amphibians. Aquatic habitats include streams and rivers; red-bellied newts apparently do not use ponds or other standing water habitats for breeding (Riemer, 1958; Stebbins, 1985; Petranka, 1998). Males tend to enter the streams before females and therefore spend more time in the aquatic habitat (Twitty, 1942, 1955; Packer, 1963). Males also tend to breed more frequently than females; males breed usually every 1–2 yr, whereas females usually breed only ≥ 2 yr. Consequently, females may spend several years on land before entering the water again for breeding (Twitty et al., 1964; also see "Age/Size at Reproductive Maturity" below).

F. Home Range Size. Estimates of home range size should take into account the regular movements of individuals between terrestrial and aquatic habitats. Home range size has not been calculated in this way, but there are quantitative data available on aquatic home range size, as well as qualitative data on terrestrial home range size. Mark-recapture studies (e.g., Twitty, 1959, 1961a, 1966) demonstrated that individuals return each breeding year to the same general region of a stream, within roughly 15 m (50 ft) of their original capture point. Once in the water, animals usually do not range > 15 m from the original point of entry, although longer jaunts are sometimes taken (Packer, 1963). Based on terrestrial trapping of marked adults, Twitty et al. (1967b) concluded that the terrestrial home range encompasses only a small area of the hillside adjacent to the breeding site. These estimates of home range size are based only on data from a single locality (Pepperwood Creek, Sonoma County) and may not be representative of the species as a whole.

G. Territories. Based on mark-recapture experiments, Twitty (1961a) concluded that adults exhibit territorial behavior in the sense that each member of the population tends to return year after year to the same stream segment for reproduction. However, a demonstration of territoriality must include evidence of the defense of space against intruders. There is no evidence that adults defend their native stream segment from others, so red-bellied newts are probably not truly territorial.

H. Aestivation/Avoiding Dessication. Adults and juveniles spend the dry season (summer and early autumn) in underground shelters, where temperatures are lower and humidity is higher than at the surface. Packer (1960) remarks that the physiological state of the animals during this period is unknown, so it is uncertain if the adults are aestivating. Gamete formation and even some yolk deposition apparently take place during this time.

I. Seasonal Migrations. Adults emerge from underground retreats in the fall, triggered by rainfall and hormonal activity associated with the reproductive drive (Twitty, 1942, 1959; Packer, 1960). Typically, it takes several bouts of rainfall before the animals emerge, and as the season progresses, less rainfall is needed to trigger emergence. Not all animals appear at once; emergence is spread out over several months, until the middle of February (Packer, 1960). Salamanders forage on the forest floor before descending to the streams for breeding in the late winter or early spring (Twitty et al., 1964). Beginning in late January, some animals begin to move toward the stream (Packer, 1960). Adults do not enter the water until the streams begin to recede from the winter floods (Twitty, 1942; also see "Breeding migrations" above). If conditions are suitable, animals may begin to enter the water in late February to early March; they may not leave the water until early May. Heavy rainfall, increased stream volume, or increased sediment load, stimulate the animals to temporarily leave the stream (Packer, 1960). Breeding mostly occurs in March–April, provided that heavy rains and flooding do not interrupt the animals’ activities too frequently (Twitty, 1966). At the end of the mating season, both sexes usually leave the water abruptly, over a period of only a few days (Riemer, 1958; Twitty, 1959; Packer, 1960). As animals leave the breeding stream, they move uphill, as well as at an angle carrying them in the upstream direction; the reason for the latter behavior is unclear, but may be a carry-over of orientation behavior in the stream, in which animals are oriented upstream (Twitty et al., 1967c). Animals may disperse relatively great distances uphill above the stream, perhaps as far as several hundred yards (Twitty, 1961a), but animals native to one stream confine their migrations to the watershed of that stream (Twitty et al., 1967b). Animals spend the dry summer and early autumn underground and are not seen again until the following fall (Twitty, 1966).

J. Torpor (Hibernation). There are no reports of hibernation in this species. Although freezing temperatures do occur occasionally in coastal northern California, long periods of freezing are unusual. Animals are known to be active above ground from October–February, and in the water from October–May. However, red-bellied newts are not immune to the effects of cold temperatures. Observations of aquatic adults in thermal gradients in the lab indicate that they are immobilized at body temperatures below about 2 ˚C (Licht and Brown, 1967).

K. Interspecific Associations/Exclusions. Pacific giant salamander (Dicamptodon tenebrosus) larvae and small trout are found in mountain brooks occupied by red-bellied newt adults (Twitty, 1935; Bishop, 1947). Where their ranges overlap, red-bellied newts and rough-skinned newts co-occur in mountain brooks. Both species breed in these streams, but the former tends to breed in faster-moving water whereas the latter prefers slower portions of the stream. There is an overlap in the timing of breeding, although red-bellied newts have a shorter breeding season, entering and leaving the streams earlier than rough-skinned newts (Twitty, 1942). Interspecific amplexus is uncommon between the two species, and existence of hybrids in nature is extremely rare (Davis and Twitty, 1964). Laboratory experiments suggest that species-specific sex attractants released by the females may contribute to reproductive isolation between these species (Davis and Twitty, 1964). Although the geographic ranges of red-bellied newts and California newts (T. torosa) overlap in parts of Sonoma and Mendocino counties, these species never occur in the same streams (Twitty, 1942, 1955). In the laboratory, red-bellied newts may be hybridized with any of the other Taricha species using artificial cross-fertilization, producing viable and fertile hybrids (Twitty, 1936, 1955, 1959, 1961a); however, natural hybridization is rare between these species (Twitty, 1942, 1955; Hedgecock and Ayala, 1974).

L. Age/Size at Reproductive Maturity. Estimates of age at reproductive maturity range from 4–6 yr (Twitty, 1966) to 6–10 yr (Hedgecock, 1978). Adults range from 5.9–8.1 cm SVL (Stebbins, 1985) or 14–19.5 cm total length (Petranka, 1998). Sexual dimorphism during the breeding season is not as pronounced in red-bellied newts as it is in other species of Taricha. The skin of aquatic males is almost entirely smooth, the dorsal fin is not greatly enlarged, and the cloacal lips are not as swollen (Twitty, 1935). There is a difference in breeding frequency between the sexes. Not all individuals breed annually, though annual breeding is more common among males than females. Based on mark-recapture studies, Twitty et al. (1964) concluded that males will commonly, but not always, breed in immediately successive years. About 50–60% of sexually mature males marked one breeding season returned the next; most but not all of the remainder returned in 2 yr time. For females, the interval between breeding seasons varied from one to several years, but no more than about 2–3% of females bred in immediately successive years.

M. Longevity. Nearly 40% of the adult males marked in 1953 were still being caught at Pepperwood Creek in 1964 (Twitty, 1966), meaning that they must have been at least 17 or 18 yr old (assuming it takes about 6 yr to reach sexual maturity). Based on this, Hedgecock (1978) speculated that longevity probably ranges from 20–30 yr.

N. Feeding Behavior. With the onset of winter rains, adults emerge from their burrows to forage in the forest for a short period (Twitty, 1961a). Foraging is restricted to periods of high humidity, late in the day or at night, or during a rainfall (Twitty, 1966). Stomach content analysis of terrestrial and aquatic adults revealed that the diet consists of terrestrial organisms only, predominantly insects. Adults apparently do not feed during their aquatic phase, though they will resume feeding if forced back onto land by winter floods (Packer, 1961). Aquatic adults maintained in the lab also do not feed (Licht and Brown, 1967).

O. Predators. No predators have been specifically reported for red-bellied newts. However, Twitty (1966) reported that at his field station in Sonoma County, garter snakes “brazenly stole” newts from storage containers as he and his assistants looked on. Both red-bellied and rough-skinned newts occurred at the field station, but most of Twitty’s work there was on red-bellied newts, so presumably this is the species to which he was referring. He did not identify the garter snakes to species, but most likely they were common garter snakes (Thamnophis sirtalis), which are known to feed on rough-skinned newts (see rough-skinned newt species account for references).

P. Anti-Predator Mechanisms. The anti-predator mechanisms of red-bellied newts are similar to those described for rough-skinned newts: toxicity, aposematic coloration, and defensive posturing. Ovarian eggs and embryos of red-bellied newts contain high levels of tetrodotoxin (Mosher et al., 1964); all 3 species of Taricha show similar levels of toxicity in this regard (Twitty, 1937; Brodie et al., 1974b). The skin of adult red-bellied newts also contains tetrodotoxin. The toxicity of the back skin of adult female red-bellied newts is similar to that of California newts (T. torosa), but adult rough-skinned newts are considerably more toxic. Brodie et al. (1974b) estimated that 1,200–2,500 mice could be killed by the skin of red-bellied newts (as compared to approximately 25,000 mice for rough-skinned newts). Ovarian eggs and adult skin of red-bellied newts have similar toxicity levels. These high levels of tetrodotoxin render newts inedible to nearly all predators (Brodie, 1977). Animals are dark-colored dorsally and bright tomato-red ventrally. The dark dorsum is cryptic against the forest floor. The defensive posture exposes the aposematic coloration of the ventral surface. During the unken reflex, the tail and head are elevated vertically revealing the bright red underside of the chin and tail; at the same time, toxic skin secretions are released. The defensive display varies somewhat depending on the state of the salamander or with the intensity of stimulation: the low intensity response is a U-shaped posture involving a vertical elevation of head and tail such that both point skyward. With a high intensity response, the head tips further back and the pelvis and hindlimbs are lifted off the substrate such that the tip of the snout and the base of the tail almost come into contact. The unken reflex display is generally similar in all species of Taricha, but red-bellied newts show one striking variation in posture. Instead of elevating the tail, sometimes the pelvis and hindlimbs remain in contact with the ground, while the head, forelimbs, and body are together lifted from the substrate; this variation in posture may be a result of the longer tail of red-bellied newts counterbalancing the body (Brodie, 1977).

Q. Diseases. None reported.

R. Parasites. None reported.

4. Conservation. The conservation status of red-bellied newts is uncertain, because no ecological studies have been conducted on this species for the last 35 yr or so. However, this species has a limited and somewhat spotty geographic distribution, and human population pressure has intensified considerably over much of its range. Specifically, conversion of native forests and grasslands to vineyards and subdivisions likely poses a serious threat to red-bellied newts. For example, this change in land use has led to the large-scale removal of trees, resulting in the alteration of temperature, sediment load, and physical structure of rivers and streams, such that they are less hospitable to native anadromous salmonid fishes (Giusti and Merenlender, 2002). It is likely that this degradation of aquatic habitat has also negatively impacted aquatic-breeding salamanders, such as red-bellied newts. Removal of trees also affects the microclimate of terrestrial habitats, perhaps rendering them less suitable for terrestrial newts. Finally, increased vehicular traffic associated with housing subdivisions undoubtedly has resulted in increased mortality of terrestrial newts.

1Sharyn B. Marks
Department of Biological Sciences
Humboldt State University
Arcata, California 95521

2Darrin Doyle
Department of Biological Sciences
Humboldt State University
Arcata, California 95521

Literature references for Amphibian Declines: The Conservation Status of United States Species, edited by Michael Lannoo, are here.

Feedback or comments about this page.


Citation: AmphibiaWeb: Information on amphibian biology and conservation. [web application]. 2016. Berkeley, California: AmphibiaWeb. Available: (Accessed: May 30, 2016).

AmphibiaWeb's policy on data use.