The Effects of a Dam on Breeding Habitat and Egg S

The Effects of a Dam on Breeding Habitat and
Egg Survival of the Foothill Yellow-legged Frog
(Rana boylii) in Northwestern California
Pacific Southwest Research Station, USDA Forest Service
1700 Bayview Drive, Arcata, California 95521, USA
*Present address: Rocky Mountain Research Station, USDA Forest Service
Southwest Forest Science Complex, 2500 South Pine Knoll
Flagstaff, Arizona 86001-6381, USA
Although the debate continues on the relative roles of natural
fluctuations and anthropogenic influences in recently observed
amphibian population declines (Blaustein 1994; McCoy 1994;
Pechman et al. 1991; Pechman and Wilbur 1994; Travis 1994),
many species are clearly demonstrating local extinctions and largescale declines (Blaustein and Wake 1990; Blaustein et al. 1994b;
Fellers and Drost 1993; Hayes and Jennings 1986; Jennings 1988).
Causes of these declines can be placed in two broad categories:
direct habitat degradation (Blaustein et al. 1994b), and more nebulous large-scale environmental changes and contamination (e.g.,
airborne pollutants, acid precipitation, ultraviolet radiation)
(Blaustein et al. 1994a; Dunson and Wyman 1992; Fellers and
Drost 1993; Hagstrom 1977; Harte and Hoffman 1989).
Dams and water diversion projects have been proposed as causes
of declines of many native frogs (Hayes and Jennings 1986;
Jennings 1988;Moyle 1973), but the effects of these projects have
not been quantitatively evaluated. Water diversions also have been
implicated in region-wide declines of anadromous fish in the Pacific Northwest (Moyle and Williams 1990; Nehlsen et al. 1991)
and are documented to adversely affect riverine
macroinvertebrates (Petts 1984). Impacts to fish include loss and
degradation of habitat because of reservoir filling and downstream
changes in river morphology along with detrimental off-season
andfluctuatingwaterreleases(BurtandMundie 1986; Petts 1984).
A recent petition to list the arroyo southwestern toad (Bufo
microscaphus califomicus) as endangered indicated that impoundments associated with dam construction have caused a 40% loss
in the original habitat of this toad. In addition, it was suggested
that habitat quality downstream of dams has been severely compromised by changes hi water temperatures, riparian vegetation,
and habitat stability (Federal Register 1993). Our study provides
information on both direct and indirect effects of a dam on the
foothill yellow-legged frog (Rana boylii) in a northwestern California river system.
Rana boylii is generally found in association with stream riffles
that have rocky substrates and partly shaded banks (Hayes and
Jennings 1988; Moyle 1973; Zweifel 1955). This frog ranges from
northern Baja California, Mexico (Loomis 1965) to central Oregon, west of the Sierra Nevada and Cascade crests. Rana boylii
is currently listed as a Species of Special Concern in California
(Laudenslayer et al. 1991) and is also a candidate for Federal listing (Federal Register 1994). The most notable declines are in
southern California and the west slope drainages of the Sierra
Nevada and southern Cascade Mountains (Jennings and Hayes
1994; Leonard et al. 1993).
Our objectives were: (1) to describe the changes in Rana boylii
breeding habitat in a dammed river prior to and some years following dam construction, and (2) to document the effects of the
timing of water releases from the dam on the reproductive ecology of this population.
Our study took place on the main stem of the Trinity River, in
Trinity County, California, USA, and covered the 63 km of river
from Lewiston Dam (near Lewiston, California) downstream to
the confluence with the North Fork Trinity, near Helena, California (elevation 420-550 m). Construction of two dams, an upper
(Trinity) and a lower (Lewiston), was completed in 1963. At Helena, this river drains an area of 2968 km2. Before dam construction the river had an average yearly (water year, Oct-Sept) discharge of 1.2 million acre-feet at Lewiston. During the four years
of our study, the yearly discharge ranged from 270,800 to 367,600
acre-feet at Lewiston. The basin surrounding our study reach is
52% public land (USDA Forest Service and USDI Bureau of Land
Management) and 48% privately owned. Land use activities within
this watershed include logging and associated road building, recreation, and widely scattered home construction. Portions of river
channel and adjacent areas were mined in the early to mid-1900's.
Since 1984, the USDI Fish and Wildlife Service, in cooperation with the USDI Bureau of Reclamation (the dam administrator), has been conducting a flow evaluation study on the main
stem of the Trinity River. Water has been released at varying levels above base flow (8.5 m3/sec [cms]), each spring in order to
facilitate migration of wild and hatchery-raised anadromous fish
and to permit evaluation of fish habitat availability and changes
as a function of discharge. The timing of these high flow releases
has been influenced by many factors, though efforts have been
made to mimic historical flow release patterns (Trinity River Restoration Program 1994).
FIG. 1. Riparian habitat composition before (1960) and 26 years after
(1989) construction of Trinity and Lewiston Dams on the main stem
Trinity River. Estimates were derived from GIS analysis of aerial photo
A geographical information system (GIS) was used to analyze
amount and distribution of riparian habitats prior to and following construction of the dams. Aerial photographs taken before
(1960) and 26 years after (1989) the construction of the dams
were used to derive estimates of the areal extent of four riparian
habitats: (1) unvegetated rocky areas, (2) willow stands, (3) wil-
Herpetological Review 27(2), 1996
low/alder mixed stands, and (4) alder/cottonwood stands. We used
U.S. Geological Survey (USGS) records of flow releases at
Lewiston and in an undammed tributary stream within our study
area, Grass Valley Creek, California, to determine the riming and
magnitude of discharges during our study. Values for discharge at
Helena (the lower end of our study area) were derived by extrapolating tributary flows from Grass Valley Creek to the drainage area between Lewiston and Helena.
Held surveys of all potential breeding habitat were conducted
from 1991-1994 along the study reach described above. Designation of potential breeding habitat was based on literature accounts of suitable habitat (Fuller and Lind 1992; Hayes and
Jennings 1988; Kupferberg, in press; Zweifel 1955) and our experience with this species in other drainages in northwestern California. Potential breeding habitat, for riverine environments, was
defined as low-velocity (0.0-0.21 m/sec), shallow water (8-40
cm), with rocky substrates, adjacent to sparsely vegetated, gravel/
cobble bars (hereafter referred to as bars). Bars are formed through
the natural meandering process of rivers and are areas of repeated
sediment deposition and erosion; usually they occur on the inside
of meander bends (Leopold et al. 1964). We sampled only those
bars (= sites) that were at least 10 m long and composed of coarse
gravel (>32 mm diam: Plaits et al. 1983) or larger substrates. Many
of the bars are no longer geomorphologically active because of
decreased flows. In these cases, only the sparsely vegetated portions of bars were surveyed and the heavily vegetated and steeply
sloped portions were omitted because of poor habitat quality. In
1993 and 1994, we also surveyed several "bank feathering" restoration projects that were within our study reach. These feathering projects are designed to restore river bars to their historic form.
Through removal of riparian berms and encroaching vegetation,
river banks are mechanically recontoured to form a gentle slope
and provide shallow, low water-velocity habitats (Trinity River
Restoration Program 1994).
Several surveys were conducted throughout the breeding and
rearing season of Rana baylii on the main stem Trinity River in
1991-1994 (Table 1). Two surveys were conducted each spring
in 1991-1993 and one in 1994. Twenty-five to 33 natural bars
were surveyed each year. In addition, one bank-feathering project
was surveyed in 1993 and five more were surveyed in 1994. Surveys were timed to coincide with the peak of breeding activity
each year, if possible, and were carried out from late April through
TABLE 1. Rana boylii reproductive ecology on the main stem Trinity River, between Lewiston and Helena, in northwestern California, 1991-1994.
Survey Type
Survey Dates
# Sites
Egg Masses
May 2-14
Potnl Breed Sites
May 15-24
Post-high flow
Jun 4-6
Post-high flow
Jun 7-14
Post-high flow
Jun 17-21
Larvae (egg sites)
Jul 31-Aug 30
Larvae (other sites)
Jul 31-Aug 30
Potnl Breed Sites
Apr 28-May 7
Potnl Breed Sites
May 26-Jun 4
Post-high flow
Jun 25-26
Larvae (egg sites)
Jul 28-29
Jun 9-Jun 28
Jul 7-Jul l5
Aug 30-Sept l
Potnl Breed Sites
Jun l-Jun 9
Monitoring Site
Apr 15-Nov 14
Subset Potnl Breed
Potnl Breed Sites
# Masses
# Masses
1. ( ) = larvae only found at this site.
2. Includes one recently completed bank feathering restoration project she-and two natural sites not surveyed in previous years.
3. Includes five recently completed bank feathering restoration project sites not surveyed in previous yean; 10 of 24 egg masses were laid at these sites.
Herpetalogical Review 27(2), l996
June. At least one mid-summer survey (late July/early August)
was conducted to search for larvae. We also conducted at least
one survey at known breeding sites, immediately following highflow releases from the dam to track egg mass survival. In 1993,
high-flow releases from Lewiston Dam in April were followed
by a long period of rain and natural high flows from tributary
input in May and early June; as a result we surveyed substantially
later than in other years. At each bar site, egg masses, juveniles,
and adult frogs were counted. The length of each bar was measured by pacing so that estimates of frogs and egg masses per
linear meter of habitat could be derived. We also examined historical locality records for Rana boylii to qualitatively assess the
distribution of this species in the Trinity Basin prior to dam construction on the main stem.
Water releases since construction of the dams have been 1030% of pre-dam flows, based on both total yearly volume and
magnitude of periodic high flows (Trinity River Restoration Program 1994; USGS records 1991-1994). Within our study area,
vegetation covered less than 30% of the pre-dam riparian area
and more than 95% of the post-dam area. Open bars made up
70% of the pre-dam riparian area and only 4% of the post-dam
area (Fig. 1). This constitutes a loss of 94% of bar habitat, and
potential breeding habitat for Rana boylii. Encroachment by riparian vegetation has resulted hi the development of stable sandy
berms, forcing the river into a narrower and deeper channel that
lacks habitat complexity. Other changes seen along the main stem
Trinity include loss of seasonally flooded marshes and high-flow
side channels, filling of pools with fine sediments, lowered water
temperatures, and riffle sedimentation (Petts 1984).
Historical records (Bury 1967; Museum of Vertebrate Zoology,
University of California at Berkeley) demonstrate that Rana boylii
occurred on the main stem Trinity and in many tributary streams,
though relative abundance could not be gauged from these records.
Initial surveys indicated that there were few Rana boylii in the
upper portion of our study area and that the lower section had
more potential breeding habitat due to the more natural flow conditions created by tributary streams; subsequently we surveyed
only the lower 22.5 km. Of the 25-39 bar and bank feathering
sites searched, 7-10 had egg masses annually. Total egg mass
counts over four years ranged from 21-31 annually in 22.5 km
(Table 1).
During the four years we tracked breeding relative to the tim-
FIG. 2. Chronology of foothill yellow-legged frog (Rana boylii) breeding and egg mass abundance as documented by river shoreline surveys, and
discharge (in cubic-meters per second [cms]) of the main stem Trinity River at Helena for (a) 1991, (b) 1992, (c) 1993, and (d) 1994.
Herpetological Review 27(2), 1996
ing and magnitude of flow releases on the main stem, we found
significant negative impacts in 1991 and 1992 and undetermined
impacts in 1993 and 1994. In 1991 and 1992, all egg masses laid
before high flow releases were lost (Table 1, Fig. 2a & b). In
1991, three egg masses were laid at one site after the largest flow
release, and a second flow increase from 14 to 18 cms eliminated
two of the three (Fig. 2a). Mid-summer surveys of breeding sites
indicated that larvae survived after the high flows only at sites
where egg masses had hatched prior to high-flow releases. In 1991,
three to four larvae were found at two of the seven known breeding sites. Of the 10 known 1992 breeding sites, three sites had
one larvae each and three sites had two, eight, and fifteen larvae.
In 1993 we found no egg masses or larvae during spring surveys,
though surveys were conducted later than in previous years. However, we did find larvae at several sites during mid-summer surveys. We do not know whether these larvae were missed during
spring surveys or whether they moved into the main stem from
unsurveyed tributary streams. The only year in which a substantial proportion of egg masses and larvae appear to have survived
to metamorphosis was 1994. High-flow releases were earlier in
1994, and most breeding activity occurred after these releases
(Fig. 2d). At one site visited biweekly through late October 1994,
all life stages were seen, though no more than two metamorphic
individuals were seen on any given day.
Though downstream effects of dams can be quite variable, several changes appear consistent across most dammed systems. For
example, peak flood flows are decreased and year-round flows
are typically lower (Ligon et al. 1995; Williams and Wolman
1984). Sediment concentrations and suspended sediments decrease
substantially for many miles downstream and bed material typically becomes more coarse (Ligon et al. 1995; Williams and
Wolman 1984). In general, the areal extent of riparian vegetation
increases below dams as a result of reduced flows and lack of
scouring floods; this was the case on the Trinity (Fig. 1, Petts
1984). In addition, riverine species experience direct habitat loss
as a result of reservoir filling. Erode and Bury (1984) documented
a potential loss of more than 50 km of stream habitat for Rana
boylii inhabiting the area of a proposed dam and reservoir.
On the Trinity River, two aspects of the dams are having significant impacts on the Rana boylii population. First, river morphology has changed significantly since flows became controlled
in 1963, and as a result a significant amount of breeding habitat
has been lost (Fig. 1). Data from a comparably-sized undammed
river fork in the same system (the South Fork Trinity) demonstrated that both the number of potential sites and the total number of egg masses were substantially higher on this fork than in
our main stem study area. We found 548-730 egg masses annually, at 81-87 sites in 15 km during three years of surveys from
1992-1994 on the undammed fork (unpublished data). The second factor is the timing of high-flow releases from the dam. In at
least two of the last four years, entire cohorts of Rana boylii were
lost along our main stem study reach because of unseasonal high
flow releases (Fig. 2). Kupferberg (in press) demonstrated that
survival of Rana boylii egg masses in a natural river was dependent on channel morphology and timing of egg deposition, during both increases and reductions in flows.
Other dam-related factors that may influence the breeding ecology of Rana boylii in this river are water temperature, exotic species, and changes in prey base. To accommodate fisheries requirements, water temperatures are kept artificially low (relative to
pre-dam conditions and present conditions on other local rivers)
using higher than normal releases in mid-summer; this may retard development of eggs and larvae (Duellman and Trueb 1986).
FIG. 3. Discharge (in cubic meters per second [cms]) hydrographs for the
main stem Trinity River at Lewiston, and Grass Valley Creek, a tributary
to the main stem below Trinity and Lewiston dams for (a) 1991 and (b)
Bullfrogs (Rana catesbeiana), a suspected predator on many native western ranids (Hayes and Jennings 1986; Moyle 1973), are
present hi the main stem Trinity (pers. obs.). Controlled flows
and lack of winter flooding may actually increase suitable habitat
for this exotic frog by providing stable pool areas with established aquatic vegetation. Changes in macro-invertebrate prey
composition and abundance have been documented in many
dammed river systems (Petts 1984) and these changes may affect
native amphibians.
We believe the negative impacts we have identified on the main
stem Trinity River can be reduced with thoughtful management.
Habitat loss is a critical problem and Rana boylii has responded
to "bank feathering" restoration projects within one year of construction. Ten of the 24 egg masses found hi 1994 surveys were at
restoration sites. However, we do not know whether frogs using
these newly created habitats are additions to the breeding population or whether the frogs have moved into these areas simply because they are more suitable than what is currently available on
the river.
Tuning of high flows is more difficult to address because we
do not know what environmental or other cues frogs use to ini-
Herpetological Review 27(2), 1996
tiate breeding. The timing of amphibian breeding is thought to be
influenced by factors such as temperature and day length, which
directly affect hormonal cycles (see Duellman and Trueb 1986
for a review). Metter (1961) speculated dial for western toads
(Bufo boreas boreas) breeding in riverine systems, timing and
volume of run-off may also influence timing and location of breeding. An Asian stream-breeding frog, Buergeria buergeri, is known
to decrease its breeding activity during heavy rainfall and high
stream flows (Fukuyama and Kusano 1992). On the main stem
Trinity, releases usually have been timed to mimic average predam patterns of spring run-off events. Using average conditions
to determine flow schedules may or may not be consistent with
environmental conditions in a given year. Fof example, if a threshold temperature or water level is required before frogs can initiate breeding and these conditions occur in April because of local climatic shifts, frogs may initiate breeding. If high-flow releases then occur in May of that year, egg masses or larvae are
likely to be lost Compare the discharge from an undammed tributary within the watershed to controlled discharges from Lewiston
dam for the same time period in 1991 and 1994 in Figure 3. The
tributary stream reflects the natural environmental conditions. In
1991 (and 1992, not shown), high flow releases from the dam
occurred much later than the natural flow regime demonstrated
by the tributary and in 1994 timing was similar between the main
stem and the tributary. Note that 1994 was the only year hi which
we believe that a substantial portion of larvae survived to metamorphosis. A more appropriate strategy for Rana boylii and, we
suspect, for most species that have evolved hi stochastically fluctuating riverine environments, would be to base real-time changes
in flow releases on current environmental conditions (e.g., if it is
raining, release more water). Understanding the interactions between environmental factors (both natural and artificial) and frog
breeding ecology is critical for managing water releases to reduce impacts on frogs and other species existing in dammed river
BURY, R. B. 1967. An analysis of the herpetofauna of Trinity County,
California. Unpubl. report. Zoology Department, Univ. of California
Berkeley. 182 pp.
DROST, C. A., AND G. M. FELLERS. 1994. Decline of frog species in the
Yosemite section of the Sierra Nevada. Cooperative National Park
Resource Studies Unit, Univ. of California, Davis. Tech. Rep. No. NFS/
WRUC/NRTR 94-02 (UC CPSU TR # 56).
DUELLMAN, W. E, AND L TRUEB. 1986. Biology of Amphibians. McGrawHill Book Company, New York. 670 pp.
DUNSON, W. A., AND R. L. WYMAN. 1992. A symposium on amphibian
declines and habitat acidification. J. Herpetol. 26:349-352.
FEDERAL REGISTER. 1993. 50 CFR Part 17. Endangered and threatened
wildlife and plants; proposed endangered status for the arroyo southwestern toad. USDI Fish and Wildlife Service. Proposed rule. Federal
Register 58:41231-41237.
____. 1994.50 CFR Part 17. Endangered and threatened wildlife and
plants; animal candidate review for listing as endangered or threatened species. USDI Fish and Wildlife Service. Proposed rule. Federal
Register 59:58982-59028.
FELLERS, G. M., AND C. A. DROST. 1993. Disappearance of the Cascades
frog (Rana cascadae) at the southern end of its range, California, USA.
Biol. Conserv. 65:177-181.
FUKUYAMA, K., AND T. KUSANO. 1992. Factors affecting breeding activity
in a stream-breeding frog, Buergeria buergeri. J. Herpetol. 26:88-91.
FULLER, D. D., AND A. J. LIND. 1992. Implications of fish habitat improvement structures for other stream vertebrates. In R. Harris and D.
Erman (eds.), Proceedings of the Symposium on Biodiversity of Northwestern California, pp 96-104. Santa Rosa, California.
HAGSTROM, T. 1977. Grodomas fors vinnande i en forsurad sjo [The extinction of frogs in a lake acidified by atmospheric pollution]. Sverages
Natur. 11:367-369.
HAUTE, J., AND E. HOFFMAN. 1989. Possible effects of acidic deposition
on a Rocky Mountain population of the tiger salamander, Ambystoma
trigrinum. Conserv. Biol. 3:149-158.
HAYES, M. P., AND M. R. JENNINGS. 1986. Decline of ranid frog species in
western North America: are bullfrogs (Rana catesbeiana) responsible?
J. Herpetol. 20:490-509.
____, AND ______. 1988. Habitat correlates of the distribution of California red-legged frog (Rana aurora draytonit) and foothill yellowAcknowledgments.—We thank G. Hodgson, M. Kamprath, and S.
legged frog (Rana boylii): implications for management In R. C. Szaro,
Mook for help with field work. D. Reese, L. Reid, and C. Zabel and two
K. E. Severson, D. R. Patton (tech. coords.), Management of Amphibanonymous reviewers provided helpful comments on this manuscript
ians, Reptiles, and Small Mammals in North America, pp 144-158.
The Museum of Vertebrate Zoology at the University of California at
USDA Forest Service, Rocky Mountain Forest and Range Experiment
Berkeley provided locality records for relevant counties (database printStation, Gen. Tech. Rep. RM-166, FortCollins, Colorado.
out dated 13 March 1995). The USDI Geological Survey provided discharge records for October 1990 through September 1994 for the Trinity JENNINGS, M. R. 1988. Natural history and decline of native ranids in
River at Lewiston (Trinity County), California, station number 11525500,
California. In R. F. De Lisle, P. R. Brown, B. Kaufman, and B. M.
and Grass Valley Creek (Trinity County), California, station number
McGurty (eds.), Proceedings of the Conference on California Herpe11525600.
tology, pp. 61-72. Southwestern Herpetol. Soc., Van Nuys, California.
____, AND M. P. HAYES. 1994. Amphibian and Reptile Species of SpeLITERATURE CITED
cial Concern in California. A Report to the California Department of
BLAUSTEIN, A. R. 1994. Chicken little or Nero's fiddle? A perspective on
Fish and Game, Rancho Cordova, California. 255 pp.
declining amphibian populations. Herpetologica 50:85-97.
KUPFERBERG, S. J. in press. Hydrologic and geomorphic factors affecting
conservation of a river breeding frog (Rana boylii). Ecol. Appl.
J. B. HAYS. 1994a, UV repair and resistance to solar UV-B in amphib- LAUDENSLAYER, W. R., W. E. GRENFELL, JR., AND D. C. ZEINER. 1991. A
ian eggs: a link to population declines? Proc. Natl. Acad. Sci. 91:1791check-list of the amphibians, reptiles, birds, and mammals of Califor1795.
nia. Calif. Dept Fish and Game 77:109-141.
_____, AND D. B. WAKE. 1990. Declining amphibian populations: a glo- LEONARD, W. P., H. A. BROWN, L. L. C. JONES, K. R. MCALLJSTER, AND R.
bal phenomenon? Trends Ecol. Evol. 5:203-204.
M. STORM. 1993. Amphibians of Washington and Oregon. Seattle Au_____, ____, AND W. P. SOUSA. 1994b. Amphibian declines: judging
dubon Society, Seattle, Washington. 168 pp.
stability, persistence, and susceptibility of populations to local and LEOPOLD, L. B., M. G. WOLMAN, AND J. P. MILLER. 1964. Fluvial proglobal extinctions. Conserv. Biol. 8:60-71.
cesses in geomorphology. W. H. Freeman and Co. San Francisco, CaliBRODE, J. M., AND R. B. BURY. 1984. The importance of riparian systems
to amphibians and reptiles. In R. E. Warner and K. M. Hendrix (eds.), LIGON, F. K., W. E. DIETRICH, AND W. J. TRUSH. 1995. Downstream ecoProceedings of the California Riparian Systems Conference, pp. 30logical effects of dams, a geomorphic perspective. Bioscience 45:18336, Univ. of California, Davis.
BURT, D. W, AND J. H. MUNDE. 1986. Case histories of regulated stream LOOMIS, R. B. 1965. The yellow-legged frog, Rana boylei, from the Siflow and its effects on salmonid populations. Can. Tech. Rep. Fish.
erra San Pedro Martir, Baja California None, Mexico. Herpetologica
Aquatic Sci. 1477:1-98.
Herpetological Review 27(2), 1996
MCCOY, E. D. 1994. Amphibian decline : a scientific dilemma in more
ways than one. Herpetologica 50:98-103.
METIER, D. E. 1961. Water levels as an environmental factor in breeding
season of Bufo boreas boreas. Copeia 1961:488.
MOYLE, P. B. 1973. Effects of introduced bullfrogs, Rana catesbeiana,
on the native frogs of the San Joaquin Valley, California. Copeia
____, AND J. E. WILLIAMS. 1990. Biodiversity loss in the temperate
zone: decline of the native fish fauna of California. Conserv. Biol.
NEHLSEN, W., J. E. WILLIAMS, AND J. A. LICHATOWICH. 1991. Pacific salmon
at the crossroads: stocks at risk from California, Oregon, Idaho, and
Washington. Fisheries 16:4—21.
VITT, AND J. W. GIBBONS. 1991. Declining amphibian populations: the
problem of separating human impacts from natural fluctuations. Science 253:892-895.
____, AND H. M. WILBUR. 1994. Putting declining amphibian populations in perspective: natural fluctuations and human impacts. Herpetologica 50:65-84.
PETTS, O. E. 1984. Impounded Rivers: Perspectives for Ecological Management John Wiley and Sons. Chichester, England. 285 pp.
PLATTS, W. S., W. F. MEGAHAN, AND G. W. MINSHALL. 1983. Methods for
Evaluating Stream, Riparian, and Biotic Conditions. USDA Forest
Service, Intermountain Forest and Range Experiment Station, Ogden,
Utah. Gen. Tech. Rep. INT-138.70 pp.
TRAVIS J. 1994. Calibrating our expectations in studying amphibian populations. Herpetologica 50:104-108.
TRINITY RIVER RESTORATION PROGRAM. 1994. Restoration of the Main Stem
Trinity River, Background Report USDI Fish and Wildlife Service,
Trinity River Fishery Field Office, Weaverville, California, 14 pp.
WILLIAMS, G. P., AND M. G. WOLMAN. 1984. Downstream effects of dams
on alluvial rivers. USDI Geol. Surv. Prof. Pap. 1286:1-64.
ZWEIFEL, R. G. 1955. Ecology, distribution, and systematics of frogs of
the Rana boylii group. Univ. California Publ. Zool. 54:207-292.
Prey of the Introduced Bufo marinus
on American Samoa
Department of Marine and Wildlife Resources
P.O. Box 3730, Pago Pago, American Samoa 96799, USA
Present address: Department of Biological Sciences,
University of North Carolina, Wilmington, North Carolina 28403, USA
e-mail: [email protected]
On Tutuila Island, 26 toads were collected in suburban and park
areas in the village of Tafuna and 28 were collected in secondary
forest and primary rainforest (forest) of Amalau and Olovalu Crater
near Futiga. Toads were collected from 24 October 1992 to 13
July 1994 and always 2-4 h after sunset to allow some time for
foraging by the toads. Collected toads were immediately placed
in plastic bags and frozen until processing.
TABLE 1. Mean (standard deviation) snout-vent lengths (SVL), jaw
width (JW), and body mass minus stomach contents (BM) for 54 Bufo
marinus in two habitats on Tutuila, American Samoa. Probabilities based
on unpaired t- test
N = 54
N = 26
N = 28
SVL (mm)
JW (mm)
36.9 (5.5)
33.8 (4.9)
39.8 (4.3)
TABLE 2. Mean (standard deviation) weight of entire contents, animal
contents, and plant/grit contents from stomachs of Bufo marinus from
two habitats on Tutuila, American Samoa. Probabilities based on t- tests.
N = 26
N = 28
Entire contents (g)
Animal contents (g)
Plant/grit contents (g)
Intact toads, entire stomach contents, animal content of stomachs,
and plant/grit content of stomachs were weighed to the nearThe giant cane or marine toad, Bufo marinus, is native to Central and northern South America (Zug and Zug 1979) but has been est 0.1 g on a triple-beam balance. Snout-vent length (SVL) and
widely introduced to islands in the West Indies, the Pacific, Aus- maximum width of the jaws was measured with dial calipers to
tralia, and New Guinea (reviewed by Zug et al. 1975). In many the nearest 0.1 mm. Animal prey were identified to the lowest
areas it was introduced primarily to control sugar cane insect pests possible taxonomic level and quantified.
Forest toads were longer, had wider gapes, and weighed more
and has apparently been fairly successful hi some areas in reducthan
suburban/park toads (Table 1). In addition, the stomachs of
ing damage to this crop (Wolcott 1937).
toads contained twice as much food by mass as did those of
Bufo marinus was introduced from Hawaii to Tutuila Island
(14°20'S, ITO^S'W), American Samoa, in 1953 and the popula- suburban/park toads. Animal contents, but not plant/grit contents,
tion was estimated to be over 2 million by 1976 (Amerson et al. of the stomachs were significantly greater in forest toads (Table 2).
Toads in both habitat types took a wide variety of prey (Table
1982). Toads occur island-wide on Tutuila, have colonized neighboring Aunu'u Island, but do not presently occur on the nearby 3). Nearly one half of the prey taken by forest toads was milliislands of Tau, Ofu, and Olosega, nor on Savaii and Upolu of pedes, whereas suburban/park toads took more moths, caterpillars, and beetles. The largest centipede taken was 51 mm long.
Western Samoa.
Marine toads have catholic tastes (Alexander 1965; Hinckley Termites, geckos, and dog food were also eaten by toads in sub1963; McCoid 1994; Zug et al. 1975). The purpose of this study urban/park habitats (P. Craig, B. Grant, P. Trail, pers. comm.), but
was to examine the food habits of marine toads in two different the latter two items were not found hi the stomachs of this study.
Nearly equal numbers of animal prey were found in suburban/
habitats on Tutuila, American Samoa.
park toads (227) as in forest toads (217). Because the animal conHerpetological Review 27(2), 1996