Download UNCORRECTED PROOF Frog community responses to recent

Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
UNCORRECTED PROOF
Frog community responses to recent American bullfrog invasions
Yiming LI 1*, Zhunwei KE 1,2, Yihua WANG 1,2, Tim M. BLACKBURN3
1
Key laboratory of Animal Ecology and Conservation Biology, Institute of Zoology, The Chinese Academy of
Sciences, Chaoyang District, Beijing 100101, China
2
Graduate University of the Chinese Academy of Sciences, 19 Yuquan Road, Beijing 100039, China
3
Institute of Zoology, ZSL, Regent’s Park, London, NW1 4RY, United Kingdom
Abstract
Native species may decline quickly when confronted with an exotic species to which they are not adapted. The extent
of decline may depend on the abundance of an invader and the length of time since it first arrived in the community (residence
time), and the interaction between these two variables. We tested these effects using data on the effects of American bullfrog
Lithobates catesbeianus invasion on native frog communities in 65 permanent lentic waters on islands in the Zhoushan
Archipelago, China. We examined variation in native frog abundance and species richness in relation to features of the American
bullfrog invasion, habitat disturbance, characteristics of the water body and fish communities and the presence of red swamp
crayfish. Bullfrog invaded sites had lower native frog density and species richness, higher submerged vegetation cover and greater
frequency of repairs to the water body than did non-invaded sites. The minimum adequate general linear mixed models showed
that both native frog density and species richness were negatively related to post-metamorphosis bullfrog density, and that native
frog species richness was also positively related to the vegetation cover. There was no effect on either native frog density or
species richness of residence time or its interaction with bullfrog density, or of the abundance of bullfrog tadpoles. The results
suggested that post-metamorphosis bullfrogs had impacts on native frog communities in the islands, and that the extents of these
impacts are proportional to post-metamorphosis bullfrog density [Current Zoology 57 (1): – , 2011].
Key words
Invader abundance, Residence time, Impact, American bullfrog, Native frog decline
Many invasive species have large impacts on native communities and human economies, posing a major
Received Feb. 25; 2010; accepted June 2, 2010.

Corresponding author. [email protected]
© 2011 Current Zoology
1
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
threat to global biodiversity (Williamson, 1996; Wilcove et al., 1998; Mack et al., 2000; Pimentel et al., 2002).
Understanding the factors governing the impacts of invaders is thus of considerable importance for the
management of invasive species. Such factors may include the abundance of the invader, the time since the start of
the invasion (here termed ‘residence time’), the invader’s functional distinctiveness, and disturbance or other
changes to the environment (Parker et al., 1999; Mack et al., 2000; Ricciardi and Atkinson, 2004; McKinney,
2005; Sax et al., 2005; Strayer et al., 2006). However, few studies have been conducted to evaluate these effects
on assemblages of native species (Parker et al., 1999; Mack et al., 2000; Strayer et al., 2006; Wilson et al., 2007).
Strayer et al. (2006) distinguished between invasions in an ‘acute’ phase, immediately after a new species
arrives, and a ‘chronic’ phase, after various ecological and evolutionary processes have come into play. Any
decline in native species may be especially severe in the acute phase, when confronted with an exotic species
(predator, competitor, or disease transmitter) to which they have not been previously exposed, and to which they
are not yet adapted. The extent of this decline in the invaded community is likely to increase with both the
abundance of the invader and the residence time. Invader abundance and residence time may also act
synergistically: lower invader abundance over a longer residence time, or higher invader abundance over a
shorter residence time, may have greater impacts on the invaded community than lower invader abundance over
shorter residence time.
We tested the impacts of invader abundance and residence time by evaluating the effects of exotic
American bullfrogs (Lithobates catesbeianus; hereafter referred to as bullfrogs) on the richness and abundance
of species in native frog communities on islands in the Zhoushan Archipelago, China. Naturalized populations
of bullfrogs on the islands derive from escapes from bullfrog farms or from deliberate releases for religious
purposes, and have recently invaded multiple water bodies on several of the islands (Li et al., 2006). This
provides a unique replicated natural experiment on the effects of the abundance and residence time of an
invasive species on native frog communities. Naturalized bullfrogs have caused population declines or local
extinctions of some native amphibians and snakes in the United States and other countries (ISSG 2002; Kats &
Ferrer 2003; Ficetola et al. 2007b) through the combined impacts of larvae and adults, although a few studies
have shown the impacts of bullfrog tadpoles on local frog tadpoles (Kupferberg, 1997; Kiesecker and Blaustein,
1997; Kiesecker et al., 2001). However, little is known about influence of post-metamorphosis bullfrogs alone
on the local frogs, or on the quantitative effects of bullfrog density and residence time.
We used the experiment in nature provided by naturalised bullfrogs on the Zhoushan Archipelago to test
three predictions about the impacts of bullfrogs on native frog communities. If bullfrog density and residence
time had harmful impacts on the native frog communities, it would be expected that (1) the native frog density
2
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
and species richness of local communities were lower in bullfrog-invaded sites than non-invaded sites; (2)
native frog density and species richness were negatively correlated with bullfrog density and residence time;
and (3) there were synergistic effects of bullfrog density and residence time, such that the interactions between
these variables explained variation in native frog abundance and species richness over bullfrog density and
residence time alone. We focused on effects on the abundance and richness of native frog species because the
effects of invasive species are most frequently quantified in these terms (Sax et al., 2005).
1
Materials and Methods
1.1
Study species
The bullfrog is native to the eastern and Great Plains regions of the United States and Nova Scotia in
Canada (Bury and Whelan, 1984). Since the end of the 19th century it has been introduced widely across the
world (ISSG, 2002; Li and Xie, 2004; Govindarajulu, 2004; Ficetola et al., 2007a,b; Giovannelli et al., 2008;
Liu and Li, 2009). Bullfrogs prefer larger permanent still waters such as ponds or reservoirs and breed from
early spring through late summer (Bruneau and Magnin, 1980; Bury and Whelan, 1984; Shirose and Brooks,
1995; Wang and Li, 2009). Bullfrog larvae typically require several months to three years before metamorphosis
(Willis et al., 1956; Degraaf and Yamasaki, 2001). Bullfrogs often co-occur with other aquatic invasive species
(Moyle, 1973; Adams et al 1999). The invasion of bullfrogs can be facilitated by the presence of co-evolved
non-native fish (Adams et al., 2003). In the Zhoushan Archipelago and nearby mainland, propagule pressure
and human hunting pressure (negative effect) are major determinants of bullfrog invasion success, while
characteristics of permanent water bodies, such as submerged vegetation cover, maximum depth and surface
area, do not affect the invasion success (Li et al., 2006).
1.2
Study islands
The Zhoushan Archipelago (29°31′–31°04′N and 121°30′–123°25′ E) is located in the East China Sea,
Zhejiang Province of China (Fig. 1). The detailed descriptions of the archipelago can be found elsewhere (Li et
al., 1998; 2006). Eight native frog and toad species occurred in our study sites, and frogs can be found in
permanent water bodies and surrounding areas. They exhibit some habitat and resource overlap with bullfrogs
(Wu et al., 2005). Breeding sites for amphibians are mainly provided by freshwater ponds, pools, reservoirs,
rivers, streams, rice fields, ditches and puddles. Rivers and streams are relatively rare on the islands, while
reservoirs and ponds are common. Here, pond refers to a permanent lentic water without a dam, and reservoir to
a permanent lentic water with a dam. Most ponds or reservoirs are artificially constructed, and were used for
breeding fishes and invasive red swamp crayfish Procambarus clarkii by the fishery department of local
3
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
government from the 1980s to the early 1990s. This ceased in most ponds or reservoirs in the mid 1990s
because of poor economic returns. Ponds or reservoirs now mainly provide fresh water for local people and
agricultural irrigation.
Native and exotic fish are commonly found in ponds or reservoirs, including species of Cyprinidae (e.g
Cyprinus carpio, Ctenopharyngodon idellus, Carassius auratus, Parabramis pekinensis, Aristichys nobilis and
Hypophthalmichthys molitrix) and Catostomidae (e.g. Misgurnus anguillicaudatus). The exotic red swamp
crayfish has spread in rice fields and permanent water bodies on the islands. Submerged vegetation is similar
across the water bodies (Li et al., 2002), and is dominated by annual plants such as Polygonum ssp., Cyperus
ssp., Beckmannia spp., Potentilla ssp. and perennial plants like Carex spp. and Alternanthera philoxeroides.
Most water bodies have a patrolman who is responsible for protecting them from damage due to human
activities or natural disasters. Human activities such as fishing and swimming in water bodies are prohibited by
the local government, but illegal fishing often occurs.
1.3
Methods
Data on site characteristics, densities of bullfrogs, and densities and species richness of native frogs in
permanent lentic waters on eight islands were collected from early June to late July in the bullfrog breeding
season of 2005 and non-breeding season between mid August and early October 2006. Additional data on time
since bullfrog invasion, habitat disturbance, characteristics of fish communities and the presence of the crayfish
were gathered in or after the non-breeding season in 2006. Data gathered in 2006 were used for analysis, while
the 2005 data were used to help determine which water bodies were not invaded, and which had more recent
bullfrog invasions (see below).
Bullfrogs have established naturalised reproducing populations on nine of the islands in the archipelago:
Zhoushan, Daishan, Liuheng, Taohua, Xiushan, Xiashi, Cezi, Fodu and Putuoshan. Zhoushan was excluded
from survey because the main sites invaded by bullfrogs (Li et al., 2006) had been destroyed due to urban
development and no bullfrogs could be found there. Frogs were surveyed on the remaining eight islands in a
randomly determined order. Investigators looked for all permanent lentic waters (>40 m2 in area), such as ponds
or reservoirs, which bullfrogs typically inhabit, within 500 m of the main roads on each island. Water bodies
that are being used for aquaculture or are often flooded with seawater, which are frequently artificially dried by
people, were not considered as permanent lentic waters. Frogs were not found in reservoirs with banks made
entirely of stone or cement, possibly because they are unsuitable habitats. These waters were excluded from the
survey.
Bullfrog invaded waters were defined as water bodies where bullfrogs were found or their calls were heard
4
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
for a long period (see methods on bullfrog residence time). 61 permanent bodies of water were found and
surveyed in 2005. In 2006, the survey extended to 79 permanent bodies of water, including 59 of the permanent
waters surveyed in the previous year (the other two sites had been destroyed). In total, 56 water bodies were
identified as invaded by bullfrogs and 23 as uninvaded in 2006. Both adult and sub-adult bullfrogs or tadpoles
were found in most invaded sites, suggesting that bullfrog breeding populations had established in them.
1.3.1
Density of post-metamorphosis bullfrog and bullfrog tadpoles and density and species richness of
native frogs Line transect methods were used to investigate the density of post-metamorphosis bullfrogs, and
the density and species richness of native frogs in water bodies (Jaeger, 1994). Transects (2 m×10 m) followed
the shoreline with half the width of the transect (1 m) in the water and half on the shore. The accessible
shoreline of each water body was divided into segments of equal length, excluding parts constructed from
stones or cement. For waters <200 m2 in area, the shoreline was divided into 2–4 segments, and for those >200
m2 into five segments. One line transect was located at random within each segment. All permanent waters were
surveyed for two or three consecutive nights each year. Transects were located in a different randomly chosen
position each night.
All transect sampling was carried out at night between the hours of 1930 and 2230. To avoid disturbing the
frogs, an observer walked slowly and as lightly as possible along each transect (1–2 km/h) with an electric torch
(12 volt DC lamp) counting all native frogs and post-metamorphosis bullfrogs encountered. Frogs were
carefully searched for in lighted areas at least 1.5 m ahead of the observer. Frogs heard diving into water or
jumping away from the bank in the un-searched part of a transect were followed to determine their identity. If
the frog escaped unidentified, the transect was discarded and another transect would be fixed in that segment.
Occasionally, a transect was re-searched after 10 minutes. This happened where accessible area in the segment
was limited. Experience on frogs at the study site suggested that 10 minutes was sufficient for frogs to recover
from being disturbed (Wu et al., 2005; Wang et al., 2007). Frog species were identified by sight with the help of
guidebooks (Huang, 1990), and specimens were collected to confirm identifications. Careful attention was also
paid to calls from water bodies sampled to determine if they were invaded by the bullfrog. The mean number of
native frog individuals, post-metamorphosis bullfrog individuals and native frog species per transect were used
as a measure of native frog density, bullfrog density and native frog species richness respectively, at each site.
We also counted bullfrog tadpoles in the water part of each transect following the amphibian survey each night,
and calculated the average number of the tadpoles per transect.
1.3.2 Fish species richness
Information on the occurrence and species richness of fish species was obtained
from the fishery departments of local governments on each island, which used to manage fishery of breeding
5
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
fishes and crayfish in water bodies. All water bodies in our analysis had fish present, and so this variable was
excluded from further consideration.
1.3.4 Presence of red swamp crayfish
We set up a line transect along the bank (accessible shoreline) of each
pond or reservoir. We carefully looked for red swamp crayfish in the water and on the banks along the line
transect between the hours of 1930 and midnight with an electric torch (12 volt DC lamp), and collected the
crayfish we found. The collected crayfish were identified using a guidebook (Li and Xie, 2002), and then
released back into the water. We recorded the presence of the crayfish as a dichotomous variable (1: present, or
0: absent).
1.3.5 Water body characteristics The location of each site was recorded using a Global Positioning System
(GPS). For reservoirs (most water bodies), maximum depth was estimated as the difference in height between
the water surface and the bottom of the dam. For ponds, which are small and shallow, and usually less than 3 m
in depth, maximum depth was directly measured in an accessible area. The surface area of a water body was
estimated according to its geometrical shape (most water bodies were approximately triangular or rectangular).
Following the frog line transect survey (see above), the cover of submerged vegetation in a 1 m wide strip from
the water’s edge in the water part of each transect was estimated and assigned to one of 10 categories: 1 (0%),
2(1%–10%), 3 (11%–20%),…9 (81%–90%) and 10 (>90%) (Li et al., 2006). We estimated shade in a water
body by measuring the angle from the water body center at eye height to the top of the tree line or horizon east,
south and west using a handheld clinometer (Adams et al., 2003). We used the average of the angles as an index
of shade.
1.3.6 Water repair and illegal fishing Water repair and illegal fishing are main human disturbances to water
bodies in the islands. Repairs that are conducted by local governments usually include the re-building of
damaged banks and the clearing of sludge from the water. The water development department of local
government on each island was visited and information on the repairs carried out on every water body between
August 2005 and before survey in 2006 was obtained. As all native frog species in the study take less than three
months from eggs to metamorphosis (Huang, 1990), the data on the past year were more likely to reflect
immediate effects of the repair on survival of tadpoles or post-metamorphosis frogs. The occurrence of repairs
was scored as a dichotomous variable (1, occurrence of repairs in the past three years; 0, no repairs).
We investigated illegal fishing activities around water bodies between 8:00 and 17:00 during daytime. We
visited each water body two times in two days, half hour each time each day. We looked around a water body if
there were fishing activities. We determined fishing activities around the water body if there were people with
fishing tools such as fish rod or fishnet. We defined the fishing activities as a dichotomous variable (1,
6
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
occurrence of fishing; 0, no fishing).
1.3.7 Bullfrog residence time We obtained information on residence time for bullfrogs in invaded sites using
questionnaire surveys (Yiming Li, unpublished data). One person was usually interviewed for each water body
not invaded by bullfrogs, and 2 or 3 people were for a bullfrog-invaded water body. However, for some invaded
water bodies where residents were relatively few, only one interview was possible. In total 120 people were
interviewed, averaging 1.5 persons per water body. The time since bullfrog invasion as estimated by each
interviewee was calculated as the average of the range of residence times given by the person. When an
interviewee gave the answer of > n years on the residence time for a water body, we treated the residence time
as n+1 years for the water body. When several interviewers gave different answers on the residence time for the
bullfrog invasion for a water body, the average value of these answers was used. There were 31 water bodies
where ≧2 people were interviewed. In these cases, the range of residence times obtained from different
interviewees fell between 0 and 4.5 years, with a mean (± standard error) of 1.63 years (± 1.47 years). For 32%
of these water bodies, interviewees gave the same residence time, while the range was ≤ 2 years for 65% of
water bodies, and ≤ 4 years for 94% of water bodies. The range of residence times increased with increasing
average residence time. Ranges above 3 years generally come from water bodies that had been invaded for more
than 10 years. Thus, estimates of residence times were quite consistent across different interviewees. The
questionnaire survey identified two reservoirs not invaded by bullfrogs but where sub-adults and adults of
bullfrogs were found. The residence time for the bullfrog invasion was assumed as one year in these cases.
Residence times for bullfrog invasion for water bodies on the islands range from 1 to 14.5 years.
1.3.8 Statistical analysis The logarithm of the mean number of native frogs, post-metamorphosis bullfrogs
and bullfrog tadpoles per transect (calculated as log [1+(total number of frogs counted at a site / number of
transects measured at a site)]) were used in all analyses. The synergistic effect of post-metamorphosis bullfrog
density and time since invasion was incorporated as the interaction term between post-metamorphosis bullfrog
density and residence time. This synergistic effect thus increases with increasing post-metamorphosis bullfrog
density and/or residence time. We did not pursue the synergistic effect of bullfrog tadpole density and residence
time because the occurrence of the tadpoles was very low.
In 14/79 water bodies sampled, no post-metamorphosis bullfrogs or bullfrog tadpoles were detected by the
line transect survey in waters known from other evidence to have been invaded. It is impossible to know from
the data whether or not these populations had become extinct, and so they were excluded from our analysis.
This left 65 water bodies, 42 bullfrog invaded sites and 23 uninvaded sites.
Differences in maximum depth, vegetation cover (the distribution of vegetation cover is not normal in
7
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
bullfrog uninvaded sites), shade and surface area of a water body, fish species richness, density of native frogs,
and native frog species richness were compared between sites that had and had not been invaded by bullfrogs
using non-parametric Kruskal-Wallis tests. The occurrence of water body repairs or illegal fishing in the past
year were similarly compared using Chi-squared tests.
As water bodies on each island are located closer to each other than those on other islands, species
distributional data of water bodies may display spatial autocorrelation (Dormann, 2007; Dormann et al., 2007).
We used linear mixed models in R version 2.7.1 (function glmer in the lme4 package, R Development Core
Team 2006) to derive the relationships between native frog density and species richness (dependent variables)
and independent variables relating to the environment (area, depth, shade, vegetation cover; all log-transformed),
human impacts (occurrence of repairs, occurrence of illegal fishing) and other species occurrence (fish species
richness, presence of red swamp crayfish, density of post-metamorphosis bullfrogs, bullfrog residence time, and
the interaction of the last two variables), with island as a random effect. We used backward stepwise selection to
reduce a full model containing all independent variables to the model with the lowest BIC value, by sequential
deletion of the variable with the lowest contribution to the model at each step. We checked that the addition of
no variable to this minimum adequate model further lowered its BIC. We used BIC in preference to AIC as it
reduces the likelihood of overfitting.
2 Results
All islands included sites that were and were not invaded by bullfrogs, except Xiashi, where our study sites
only included water bodies invaded by bullfrogs. Bullfrogs were detected in 42/65 water bodies surveyed (Table
1). Bullfrog tadpoles were only found in six water bodies, all of which also had post-metamorphosis bullfrogs
present.
Ponds and reservoirs that had been invaded by bullfrogs did not differ significantly from non-invaded sites
in surface area, shade, maximum depth, characteristics of fish communities and fishing (Table 1). However,
sites invaded by bullfrogs had lower native frog density, lower native frog species richness, higher submerged
vegetation cover (or more submerged vegetation cover) and a higher frequency of water repairs than
non-invaded sites. Nine percent of bullfrog invaded sites (4/42) had no native frogs, whereas native frogs were
present in all non-invaded sites (23/23).
A minimum adequate general linear mixed model for native frog density only contained
post-metamorphosis bullfrog density (Table 2). The native frog density was negatively related to
post-metamorphosis bullfrog density (Fig.2A). There was no significant effect of bullfrog residence time on
8
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
native frog density, or of the interaction between residence time and bullfrog density. There was also no effect
of bullfrog tadpole density on native frog density, over and above the effect of post-metamorphosis bullfrog
density. The minimum adequate model for native frog species richness included two variables: native frog
richness and vegetation cover. Native frog richness was negatively correlated with post-metamorphosis bullfrog
density and positively with vegetation cover (Fig 2B, C).
3 Discussion
Both native frog density and native frog species richness were, on average, significantly lower at sites
invaded by bullfrogs than at uninvaded sites on islands in the Zhoushan archipelago (Table 1). Minimum
adequate models for both native frog density and species richness each included a significant term for
post-metamorphosis bullfrog density, but no significant effects of residence time or its interaction with density.
These results support the first of our predictions, that native frog species richness and density are lower in water
bodies invaded by bullfrogs. They also partially support our second prediction, in that native frog species
richness and density were consistently negatively correlated with bullfrog density. However, there is no support
for the prediction of a negative effect of bullfrog residence time on native frog community structure, or for
synergistic effects of bullfrog density and residence time. These results suggested that the native frogs have not
adapted to bullfrog invasion after 15 years of the invasion, and strongly implicate post-metamorphosis bullfrogs
as a cause of declines in the abundance and species richness of native frog species in bullfrog-invaded sites.
Although vegetation cover and the occurrence of recent repairs to the water body differed between sites with
and without bullfrogs (Table 1), there is no evidence that these were responsible for lower native frog density or
species richness in sites invaded by bullfrogs. Neither native frog density nor species richness was associated
with the recent occurrence of water repairs, once bullfrog density and vegetation cover were included in the
analysis. The greater vegetation cover in sites invaded by bullfrogs (Table 1) is unlikely to lead to lower species
richness because the general linear mixed models indicated that higher vegetation cover actually promotes the
native frog richness.
Several previous studies have reported negative effects of red swamp crayfish invasion on native amphibians
through predation on eggs, larvae and adults (Gamradt et al., 1997; Cruz et al., 2006). There was no difference
in the presence of the crayfish between sites with and without bullfrogs and no relationship between the native
frogs and the presence of the crayfish in the islands. One possible explanation for these results is that the
Zhoushan islands are home to a native crayfish Macrobrachium nipponensis, which is functionally similar to
red swamp crayfish. Native frogs may previously have evolved specific anti-predator traits to co-exist with the
9
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
native crayfish, thus may not be damaged by the alien crayfish. In the absence of data on red swamp crayfish
density and residence time in the waters sampled, this possibility should be treated with caution. The effects of
red swamp crayfish invasion on native frogs in the islands certainly warrant further study.
We found no evidence that characteristics of fish communities were responsible for the lower native frog
density and species richness. The negative effects of introduced fish on native amphibians have been
extensively documented (Moyle, 1973; Fisher and Shaffer, 1996; Adams, 1999; Martinez-Solano et al., 2004;
Davison and Knapp, 2008), and there is evidence that introduced fish have larger impacts on native amphibians
than bullfrogs in areas where the fish and bullfrogs co-occurred (Adams et al., 1999). While we could not
distinguish between the presence of native and introduced fish species, we found no effect of fish species
richness on native frog richness or density, while fish were present in all water bodies in our analysis. We
cannot rule out the hypothesis that the presence of a specific fish species drives differences in the density and
richness of native frog species, but the density of that fish would have to co-vary strongly with bullfrog density
to be responsible for the strong negative effects of bullfrog density we observed. There was no difference in the
richness or presence of fish communities between sites with or without bullfrogs (Table 1).
Post-metamorphosis bullfrogs most likely threaten populations of native frogs through predation,
competition and by spreading disease (predation: Werner et al., 1995; Pearl et al., 2004; Hirai, 2004; Wu et al.,
2005; Wang et al., 2007; food competition: Wu et al, 2005; disease: Daszak et al., 2004; Garner, 2006). All three
types of threats may contribute to native frog decline in the islands. Post-metamorphosis bullfrogs prey upon at
least four native frog species (R. limnocharis, R. nigromaculata, R. japonica and B. bufo gargarizans) in the
islands (Wu et al., 2005; Wang et al., 2007). They also have some dietary overlap with these native frog species,
competing with them for food resources (Wu et al., 2005). Post-metamorphosis bullfrogs sampled in autumn
2006 on the islands have been found to be infected with the fungal pathogen Batrachochytrium dendrobatidis
(Yiming Li, unpublished data), which has affected at least 200 amphibian species worldwide (see Kriger and
Hero, 2006). It is possible that post-metamorphosis bullfrogs might have carried and spread the disease to the
Zhoushan islands, and that the disease may be responsible for the native frog decline. The threats to native frogs
of predation, competition and infection may all plausibly increase with post-metamorphosis bullfrog density.
Bullfrog tadpole competition is also recognized as an important mechanism in affecting native amphibian
tadpoles (Kupferberg, 1997; Kiesecker and Blaustein 1997; Lawler, et al., 1999; Kiesecker et al., 2001). Effects
of bullfrog tadpoles on native frogs were not detected in the present study, perhaps due to the lower occurrence
of tadpoles relative to post-metamorphosis frogs. Bullfrog populations in many of the water bodies studied did
not successfully produce eggs in a given year (unpublished data). Also, new tadpoles in some water bodies
10
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
finish their metamorphosis before winter, which is likely to lead to the lower presence of tadpoles in water
bodies in the islands.
In conclusion, the present study is consistent with the hypothesis that post-metamorphosis bullfrogs have
negative impacts on native frog communities in islands of the Zhoushan archipelago. The presence of exotic
bullfrogs in lentic water bodies is associated with reductions in the total density and species richness of native
frog species. The extents of these reductions are proportional to post-metamorphosis bullfrog density, but not to
residence time over the 15 years that the invasion has been occurring.
Acknowledgements We thank Feng XU and Yanping WANG for helping a part of field works and Richard Duncan for
comments on the manuscript. This work was supported by a grant from National Science foundation (No. 30870312) and by a
grant from the “973” program (No. 2007CB411600).
References
Adams MJ, 1999. Correlated factors in amphibian decline: Exotic species and habitat change in western
Washington. Journal of Wildlife Management 63: 1162–1171.
Adams, MJ, Pearl CA.,Bury RB, 2003. Indirect facilitation of an anuran invasion by non-native fishes. Ecology
Letters 6: 343–351.
Bruneau M, Magnin E, 1980. Croissance, nutrition et reproduction des ouaouarons Rana catesbeiana Shaw
(Amphibia Anura) des Laurentides au nord de Montreal. Canadian Journal of Zoology 58:175–183.
Bury RB, Whelan JA, 1984. Ecology and Management of the Bullfrog. Washington DC: Publications Unit,
U.S.Fish and Wildlife Service.
Cruz MJ, Rebelo R, Crespo EG, 2006. Effects of an introduced crayfish Procambarus clarkii on the distribution
of south-western Iberian amphibians in their breeding habitats. Ecography 29: 329–338.
Daszak P, Strieby A, Cunningham AA, Longcore JE, Brown CC, Porter D, 2004. Experimental evidence that the
bullfrog Rana catesbeiana is a potential carrier of chytridomycosis, an emerging fungal disease of amphibians.
Herpetological Journal 14: 201–207.
Davison C and Knapp RA. 2008. Multiple stressors and amphibian declines: Dual impacts of pesticides and fish
on yellow-legged frogs. Ecological Applications 17:587–597.
Degraaf RM, Yamasaki M, 2001. New England Wildlife: Habitat, Natural History, and Distribution. Hanover,
New Hampshire: University Press of New England.
11
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
Dormann CF, McPherson JM., Araujo MB, Bivand R, Bolliger J et al., 2007. Methods to account for spatial
autocorrelation in the analysis of species distributional data: A review. Ecography 30:609–628.
Dormann CG, 2007. Effects of incorporating spatial autocorrelation into the analysis of species distribution data.
Global Ecology and Biogeography 16:129–138.
Ficetola GF, Coïc C, Detaint M, Berroneau M, Lorvelec O, Miaud DC, 2007a. Pattern of distribution of the
American bullfrog Rana catesbeiana in Europe. Biological Invasions 9: 767–772.
Ficetola GF, Thuiller W, Miaud C, 2007b. Prediction and validation of the potential global distribution of a
problematic alien invasive species - the American bullfrog. Diversity and Distributions 13: 476–485.
Fisher RN, Shaffer HB, 1996. The decline of amphibians in California’s Great Central Valley.
Biological
Conservation 10: 1387–1397.
Gamradt SC, Kats LB, Anzalone CB, 1997. Aggression by non-native crayfish deters breeding in California
newts. Conservation Biology 11: 793–796.
Garner TWJ, Perkins M, Govindarajulu P, Seglie D, Walker SJ et al., 2006. The emerging amphibian pathogen
Batrachochytrium dendrobatidis globally infects introduced populations of the North American bullfrog Rana
catesbeiana. Biology Letters 2: 455–459.
Giovannelli JGR, Haddad CFB, Alexandrino J, 2008. Predicting the potential distribution of the alien invasive
American bullfrog Lithobates catesbeianus in Brazil. Biological Invasions 10:585–590.
Govindarajulu P, 2004. Introduced bullfrogs Rana catesbeiana in British Columbia: Impacts on native Pacific
treefrogs Hyla regilla and red-legged frogs Rana aurora. Dissertation. University of Victoria, Victoria,
British Columbia, Canada.
Hirai T, 2004. Diet composition of introduced bullfrog Rana catesbeiana in the Mizorogaike Pond of Kyoto,
Japan. Ecological Research 19: 375–380.
Huang MH, 1990. Fauna of Zhejiang: Amphibia, Reptilia. Hangzhou: Zhejiang Science and Technology
Publishing House (In Chinese).
ISSG Invasive Species Specialist Group, 2002. Global Invasive Species Database: Rana catesbeiana,
www.issg.org/database/species/ecology.asp? si=80&fr=1&sts=tss.
Jaeger R, 1994. Transect sampling. In: Heyer WR, Donnelly MA, McDiarmid RW, Hayek LA, Foster MS ed.
Measuring and Monitoring Biological Diversity: Standard Methods for Amphibibas. Smithsonian
Institution Press, Washington and Lndon: 103–107.
Kats LB, Ferrer RP, 2003. Alien predators and amphibian decline: Review of two decades of science and
transition to conservation. Diversity and Distribution 9: 99–110.
12
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
Kiesecker JM, Blaustein AR, 1997. Population differences in responses of red-legged frogs Rana aurora to
introduced Bullfrogs. Ecology 78: 1752–1760.
Kiesecker JM, Blaustein AR, Miller CI, 2001. Potential mechanisms underlying the displacement of native
red-legged frogs by introduced bullfrogs. Ecology 82: 1964–1970.
Kriger KM, Hero JM, 2007. Large-scale seasonal variation in the prevalence and severity of chytridiomycosis.
Journal of Zoology 271: 352–359.
Kupferberg SJ, 1997. Bullfrogs Rana catesbeiana invasion of California river: The role of larval competition.
Ecology 78:1736–1751.
Lawler SP, Dritz D, Strange T, Holyoak M, 1999. Effects of introduced mosquitofish and bullfrogs on the
threatened California red-legged frog. Conservation Biology 13: 613–622.
Li C, Xie F, 2004. Invasion of bullfrogs Rana catesbeiana Shaw in China and its management strategies.
Chinese Journal of Applied and Environmental Biology 35: 95–98 (In Chinese with English abstract).
Li GY, Chen ZH, Liu AX, Sunh MJ, Ying SD, 2002. Classification system on wet vegetation in Zhejiang
Province and main vegetation types and their distribution characteristics. Journal of Zhejiang Forestry College
19: 356–362.
Li YM, Niemila J.& Li DM. 1998. Nested distribution of amphibians in the Zhoushan archipelago, China: can
selective extinction cause nested subsets of species. Oecologia 113:557-564.
Li YM, Wu ZJ, Duncan RP, 2006. Why islands are easier to invade: Human influences on bullfrog invasion in
the Zhoushan Archipelago and neighboring mainland China. Oecologia 148:129–136.
Li ZY, Xie Y. 2002.Invasive Alien Species in China. Beijing: China Forestry Publishing House.
Liu X, LiY, 2009. Aquaculture enclosures relate to the establishment of feral populations of introduced species.
PLoS One 4: e5199.
Mack RN, Simberloff D, Lonsdale WM, Evans H, Clout M et al., 2000. Biotic invasions: Causes, epidemiology,
global consequences, and control. Ecological Applications 10: 689–710.
Martinez-Solano I, Barbadillo LI, Lapena M, 2004. Effect of introduced fish on amphibian species richness and
densities at amontane assemblage in the Sierra de Neila, Spain. The Herpetological Journal 13:167–173.
McKinney ML, 2005. Species introduced from nearby sources have a more homogenizing effect than species
from distant sources: Evidence from plants and fishes in the USA. Diversity and Distributions 11: 367–374.
Moyle PB, 1973. Effects of introduced bullfrogs Rana catesbieana on the native frogs of the San Joaquin Valley,
California. Copeia 1973:18–22.
13
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
Parker IM, Simberloff D, Lonsdale WM, Goodell K, Wonham M et al., 1999. Impact: Toward a framework for
understanding the ecological effects of invaders. Biological Invasion 1: 3–19.
Pearl CA, Adams MJ, Bury RB, McCreary B, 2004. Asymmetrical effects of introduced bullfrogs Rana
catesbeiana on native ranid frogs in Oregon. Copeia 2004:11–20.
Pimentel D, 2002. Biological Invasions: Economic and Environmental Costs of Alien Plant, Animal, and
Microbe Species. Boca Raton FL: CRC Press.
R Development Core Team 2006. R: A language and Environment for Statistical Computing. Vienna, Austria: R
Foundation for Statistical Computing.
Ricciardi A, Atkinson SK, 2004. Distinctiveness magnifies the impact of biological invaders in aquatic
ecosystems. Ecology Letters 7: 781–784.
Sax DF, Stachowicz JJ, Gaines SD, 2005. Species Invasions: Insights into Ecology, Evolution, and Biogeography.
Sunderland MA: Sinauer Associates, Inc..
Shirose LJ, Brooks RJ, 1995. Age structure, mortality,and longevity in syntopic populations of three species of
ranid frogs in central Ontario. Canadian Journal of Zoology 73: 1878–1886.
Strayer DL, Eviner VT, Jeschke JM, Pace ML, 2006. Understanding the long-term effects of species invasions.
Trends in Ecology and Evolution 21: 645–651.
Wang YH, Li YM, 2009. Habitat selection by the introduced American bullfrog Lithobates catesbeianus on
Daishan Island, China. Journal of Herpetology 43:205–211.
WangYP, Guo ZW, Pearl CA, Li YM, 2007. Body size affects the predatory interactions between introduced
American bullfrogs Rana catesbeiana and native anurans in China: An experimental study. Journal of
Herpetology 41: 514–520.
Werner EE, Wellborn GA, McPeek MA, 1995. Diet composition in postmetamorphic bullfrogs and green frogs:
Implications for interspecific predation and competition. Journal of Herpetology 29: 600–607.
Wilcove DS, Rothstein D, Dubow J, Phillips A, Losos E, 1998. Quantifying threats to imperiled species in the
United States. BioScience 48: 607–615.
Williamson M, 1996. Biological Invasion. London, UK: Chapman & Hall.
Willis YL, Moyle DL, Baskett TS, 1956. Emergence, breeding, hibernation, movement, and transformation of the
bullfrog in Missouri. Copeia 1956: 30–41.
Wilson JRU, Richardson DM, Rouget M, Procheş Ş, Amis MA et al., 2007. Residence time and potential range:
Crucial considerations in plant invasion ecology. Diversity and Distributions 13: 11–22.
Wu ZJ, Li YM, Wang YP, Adams MJ, 2005. Diet of introduced bullfrogs Rana catesbeiana: Predation on and
14
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
diet overlap with native frogs on Daishan Island, China. Journal of Herpetology 39: 668–674
Zhou JL, 1989. Integrated agriculture plans in Zhoushan city. Hangzhou: Zhejiang People Press (In Chinese).
15
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
Table 1 Comparison of site characteristics, densities of native frogs and bullfrogs, and native frog species richness for 65
permanent waters with and without bullfrog invasion in the Zhoushan archipelago, China
Site characteristic
Bullfrogs absent
Bullfrogs present
(n = 23)
(n = 42)
χ2
P
Log permanent water area (m2)
3.253±0.627
3.441±0.712
0.612
0.434
Maximum water depth (m)
5.174±3.221
5.607±4.408
0.000
0.995
Mean vegetation cover (category)
1.973±1.654
2.776±1.992
5.455
0.020
Shade (angle)
23.188±9.808
24.841±10.629
0.512
0.474
Occurrence of water repairs (%)
1/23
7/42
13.125†
<0.001
Occurrence of fishing (%)
19/23
37/42
0.080†
0.777
Occurrence of red swamp crayfish
10/23
17/42
0.384†
0.535
Fish species richness
3.261±1.214
3.024±0.897
1.090
0.297
Mean native frog species richness /
1.042±0.603
0.738±0.631
4.476
0.034
2.701±2.860
1.337±1.306
4.355
0.037
0±0
2.073±2.521
-
-
0±0
0.07±0.182
-
-
0±0
8.79±3.8
-
-
(%)
transect
Mean native frog density richness /
transect
Mean number of postmetamorphic
bullfrogs / transect
Mean number of bullfrog
tadpoles/transect
Time since bullfrog invasion (years)
Values are means ± standard deviations. χ2 is the chi-squared value (d.f. = 1) calculated from a
Kruskal-Wallis test or a Chi-square test (†) comparing sites invaded by bullfrogs to those without
bullfrogs. P is the associated P-value.
16
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
Table 2 The minimum adequate general linear mixed models with native frog density or richness as the dependent variable
and other factors as independent variables
Models
Parameters ±SD
t value
Probability
Native frog density (minimum BIC= -17.34)
Intercept
0.4161±0.0767
5.428
<0.001
-0.3551±0.0733
4.845
<0.001
0.7341±0.161
4.561
<0.001
Bullfrog density
-0.8677±0.1717
5.055
<0.001
Vegetation cover
0.6514±0.1922
3.39
0.001
Bullfrog density
Native frog richness (minimum BIC=92.53)
Intercept
17
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
Fig. 1
The islands of the Zhoushan archipelago, China
The eight islands included in this study are Daishan, Liuheng, Taohua, Xiushan, Xiashi, Cezi, Fodu and Putuoshan.
Fig. 2 The relationships among native frog density, native frog richness, postmetamorphic bullfrog density and vegetation
cover in permanent lentic water bodies on Zhoushan Archipelago, China
A. Native frog density and postmetamorphic bullfrog density.
B. Native frog richness and postmetamorphic bullfrog density.
Native frog richness and vegetation cover.
18
C.
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
Figure 1
19
Y.M. LI et al.,: Impacts of American bullfrogs on native frogs in Zhoushan Archipelago
Fig 2.
A
Log (1+native frog density)
1.2
1
0.8
0.6
0.4
0.2
0
0
0.2
0.4
0.6
0.8
1
1.2
1.4
Log (1+bullfrog density)
B
Log native frog richness
2.500
2.000
1.500
1.000
0.500
0.000
0
0.2
0.4
0.6
0.8
1
1.2
1.4
Log (1+bullfrog density)
C
2.500
Native frog richness
2.000
1.500
1.000
0.500
0.000
-0.2
0
0.2
0.4
0.6
0.8
1
Log vegetation cover
20