EDP Sciences logo
Submit your paper
Search
Menu
All issues No 424 (2023) Knowl. Manag. Aquat. Ecosyst., 424 (2023) 20 Full HTML
Issue
Knowl. Manag. Aquat. Ecosyst.
Number 424, 2023
Freshwater ecosystems management strategies
Article Number 20
Number of page(s) 12
DOI https://doi.org/10.1051/kmae/2023016
Published online 25 July 2023

© G. Laufer et al., Published by EDP Sciences 2023

Licence Creative CommonsThis is an Open Access article distributed under the terms of the Creative Commons Attribution License CC-BY-ND (https://creativecommons.org/licenses/by-nd/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. If you remix, transform, or build upon the material, you may not distribute the modified material.

1 Introduction

Biological invasions are a major component of global change and are expected to maintain their growth rate in the near future (Seebens et al., 2017). Human activities move an increasing number of species outside their historical distribution ranges, and some of them manage to get established and invade new regions. These invasions can affect ecosystem services and decrease native species abundance and richness through mechanisms such as predation, competition, indirect interactions, disease transmission and hybridization (Ricciardi et al., 2017). Although important advances have been made in the last decade concerning how invasions influence the conservation of biodiversity, economy and human health, there is still a great lack of knowledge about the state and the impacts of many invasive species worldwide (Latombe et al., 2016). This lack of information is even greater in some regions, such as South America, where studies are comparatively scarce, and usually report only occurrence or anecdotal data of invasive alien species (Speziale et al., 2012; Ballari et al., 2016; Schwindt and Bortolus, 2017).

The American bullfrog, Lithobates catesbeianus (Shaw 1802) is an invasive species of concern, as it is expanding its global range. This species, native to eastern North America, has been introduced into several regions, mostly to be cultured for human consumption (Kraus, 2009). The bullfrog is a large-bodied aquatic anuran with great ecological plasticity. It has the ability to exploit diverse resources, and to tolerate a wide range of environmental conditions. This invasive species shows a preference to occupy permanent lentic systems (Adams and Pearl, 2007), commonly named ponds, where it inhabits throughout its ontogenic cycle (Minowa et al., 2008; Cook et al., 2013). Bullfrog juveniles and adults have the ability to disperse between these systems on land, and thus expand its distribution. The availability of these ponds determines bullfrog colonization and expansion in the landscape (Blaustein and Kiesecker, 2002; Ficetola et al., 2010). Thus, a bullfrog invasion is a spatially structured population that can be seen as a network of nodes (ponds, discrete patch habitats) interconnected by its dispersal capacity in a suboptimal environment (Descamps and De Vocht, 2016). For this reason, the graph theory approach can be an interesting management tool for quantifying connectivity between ponds and exploring the bullfrog invasion risk (Drake et al., 2017).

Adult bullfrogs are important predators and strong competitors for food resources (Kiesecker et al., 2001; Jancowski and Orchard, 2013; Laufer et al., 2021). Tadpoles are also predators and habitat disruptors by bioturbation (Kupferberg, 1997; Kiesecker et al., 2001; Ruibal and Laufer, 2012; Gobel et al., 2019). Although we know that bullfrog invasion can alter different components of invaded communities (e.g. Gobel et al., 2023), we know little about its effects on different taxonomic groups (mostly from diet records; Jancowski and Orchard, 2013). Most of the studies have focused on its impacts on native amphibians. In addition, L. catesbeianus has recently been identified as a vector and reservoir of global amphibian diseases, especially Ranavirus and Chytridiomycosis (Garner et al., 2006; Lesbarrères et al., 2012; Schloegel et al., 2012; Ribeiro et al., 2019).

The spread of this invasive frog was related to local changes in the structure of amphibian assemblages and their acoustic niches, and population declines (Fisher and Shaffer, 1996; Hecnar and M'Closkey, 1997; Li et al., 2011; Both and Grant, 2012; Liu et al., 2017). Evidence suggests that large sized bullfrog adults may behave as important predators of amphibians (Liu et al., 2018; Bissattini et al., 2018; Oda et al., 2019; Laufer et al., 2017, 2021). In addition, bullfrog juveniles can also be strong competitors of native amphibians (Silva et al., 2016). In fact, Bissattini and collaborators (2019) found a shift from competitor to predator behavior, during bullfrog ontogeny. Kats and Ferrer (2003) reviewed bullfrogs' negative effects in Western North America, and predicted that its global spread would impact amphibians in new regions.

The bullfrog is becoming a frequent invader in the Neotropical region, with several foci recently reported in Argentina, Brazil, Venezuela, Ecuador, Colombia, and Uruguay (Barbosa et al., 2017). Therefore, it is important to understand its effects on native amphibian assemblages, as well as which species would be most affected, to generate control strategies and mitigation practices (Adams and Pearl, 2007). In this sense, the literature suggests that the most affected species would be those that share microhabitat use with bullfrogs (Pearl et al., 2004). In this context, it is important to explore the effects of this invasion on native amphibian assemblages in permanent water bodies and to assess which aquatic species are being most affected.

The bullfrog was introduced in Uruguay in the 1980 s at 19 aquaculture farms for producing frog legs for human consumption, a project which did not become a profitable operation and collapsed in the early 2000 s. This industry was not strictly controlled by national authorities, and the environmental risks involved were underestimated. Thus, invasion foci emerged in some of the sites where bullfrog farms were established (Laufer et al., 2008, 2018a, 2018b). One of these farms, Laguna Dorada, was located in Southeastern Uruguay, in San Carlos (Maldonado Department), and was one of the largest bullfrogs farms from 1993 to 2001. In 2007, as part of a national survey of old bullfrog farm areas, a search for bullfrogs in San Carlos was conducted and no feral individuals were detected (Laufer et al., 2018a). Later, in 2015, the presence of feral bullfrogs was reported in a pond located over a kilometer away from the Laguna Dorada farm facilities (Lombardo et al., 2016). A few months later, six additional invaded water bodies were reported, thus confirming that the first pond represented a larger population. Emphasis was placed on further efforts to evaluate the bullfrogs' real distribution and ecological impacts in the area (Laufer et al., 2018a).

The goal of this study was to evaluate the current invasion degree, and its effects on the native amphibians in San Carlos. In addition, management tools were to be generated from the understanding of the local network of environments susceptible to be invaded by bullfrogs. As a working hypothesis, it was proposed that changes in San Carlos amphibian assemblages' structure are explained by bullfrogs preying on and competing with species of this group. Therefore, it is expected to find lower richness and abundance of native amphibians in ponds invaded by bullfrogs. This effect should be stronger for those native anurans which most interact with bullfrogs, due to their similar microhabitat use.

2 Materials and methods

2.1 Study site

San Carlos is located in Southeast Uruguay in the Merín Lagoon Graven ecoregion, 15 km from the Atlantic Ocean (Brazeiro, 2015). It is a plain area (average height 20 m a.s.l.), where San Carlos Stream flows into the Maldonado Stream. The region has a national conservation priority due to its unique assemblages and high biodiversity (Di Minin et al., 2017; Grattarola et al., 2020). It is a peri-urban area, with a mosaic landscape of agricultural activities, which means that there is a great density of permanent water bodies, frequently used as reservoirs for farms (Álvarez et al., 2015). Those water bodies, called ponds further on, were characterized as having a pH of 7.24 ± 1.06 (mean ± SD), a conductivity of 203.09 ± 212.70 μS · m−1, an area of 902 ± 1262 m2, 46.0 ± 44.0% of macrophyte coverage, and fish in 54% of them (unpublished data, based on 26 sampled ponds during the spring of 2017). These ponds (Fig. 1) are usually inhabited by native anurans, which use them for breeding and foraging (Arrieta et al., 2013; Grattarola et al., 2020).

thumbnail Fig. 1

Photographs of two ponds sampled in San Carlos, Maldonado Department (Uruguay), showing the differences in macrophyte cover.

2.2 Post-metamorphs: sampling, mapping and analysis

Four surveys were completed during the spring and the summer: 8–10 December, 2017; 19–21 October, 2018; 20 March, 2019; and 6–8 December, 2019. We sampled all the ponds peripheral to those where we knew the bullfrog was reported in previous studies (Lombardo et al., 2016; Laufer et al., 2018a). The limit of the sampling area was established by a buffer around the invaded zone (where bullfrogs were not detected). We then complemented this limit with visits to peripheral areas during periods of bullfrog vocalization, confirming that there were no other foci in the area. This sampling area allowed us to cover 75 ponds, invaded and not invaded by bullfrog, within the considered area. These ponds were standardly sampled once during these campaigns (in total by the end of the surveys, we visited once each pond), between 8:00 p.m. and 12:00 a.m. that is within the period of highest bullfrog (Laufer et al., 2017) and native anuran activity (Moreira et al., 2007). At each pond, a slow walk was taken by two researchers around the whole perimeter, for a minimum of five and a maximum of ten minutes, depending on the pond size (Dodd, 2010; Heyer et al., 2014). The total number of post-metamorphic (i.e. juveniles and adults) amphibians observed were recorded for each species (direct count of bullfrog and native anurans). Bullfrog calling activity was categorized as: 0–no records; 1–a single individual vocalizing; 2–two or three individuals vocalizing; 3–more than three individuals vocalizing, and still being able to individualize the calls; and 4–chorus, where calling individuals were not identifiable (Dodd, 2010). In each of the campaigns, both types of ponds were sampled, (invaded and uninvaded by bullfrogs) and environmental variables at that specific time were recorded. These variables, along with the sampling date, are useful for standardizing and understanding differences between campaigns. During sampling, air temperature, humidity, wind intensity and percent of cloud coverage were registered.

The 75 sampled ponds were classified into four categories, according to the bullfrog field data collected: 0–no records; 1–one individual observed and/or heard; 2–between 2 and 5 individuals observed and/or heard in calling category 2 or 3; and 3–more than 5 individuals observed and/or heard in calling category 4 (chorus) (Heyer et al., 2014). Bullfrog distribution was mapped using ArcGIS Pro software (ESRI, 2021), and ponds were allocated to the above-mentioned categories. The distribution areas were obtained from the polygons generated by the union of the external ponds, and adding a buffer area of 727 meters (considering the maximum terrestrial dispersion distance of bullfrog in a similar environment in Belgium; Descamps and De Vocht, 2016).

Richness (total number of species observed or heard) and abundance (total number of individuals observed) of the native anuran assemblages were compared between invaded and uninvaded ponds. Invaded ponds were considered those with any bullfrog record, regardless of density. For this analysis, 60 comparable ponds (for which data was complete) were considered, 27 of which were invaded by bullfrogs. Post-metamorphic anuran richness (number of species observed or heard) and abundance (direct counts from visual observations) were tested to determine they were related to the presence-absence of bullfrogs by fitting a generalized linear model (GLM), with a negative binomial distribution. This distribution is commonly used to count variables with overdispersion. Air temperature, humidity, wind intensity and percent cloud coverage were tested as possible explanatory variables in the model. Model selection was done using the Likelihood Ratio Test (LRT). Sampling date was always included as an explanatory variable in the model, in order to consider the variation introduced by the sampling design. Finally, a residual analysis was performed to check the homoscedasticity and normality of each model's residuals (Zuur et al., 2007).

The specific abundances (direct counts from visual observations) of the three most-frequent native species were explored (i.e. post-metamorphic individuals recorded in more than 15 ponds): Boana pulchella, Leptodactylus luctator and Pseudis minuta. These three species differ in many ecological traits. Boana pulchella is a mid-sized arboreal species, inhabiting peripheral trees and shrubs, using the ponds' macrophytes for calling and reproduction. Leptodactylus luctator is a big sized cursorial species, which reproduces in ponds and displays parental care. Pseudis minuta is a fully aquatic mid-sized species, inhabiting the pond during its entire ontogeny. This frog commonly swims and floats on the surface (Melchiors et al., 2004; Moreira et al., 2007). For these three species, a negative binomial GLM was used to test the association between post-metamorph abundance and observed bullfrog abundance following the above-mentioned procedure (Zuur et al., 2007). In this case, the sampling date was used as an explanatory variable, in order to consider the variation introduced by the sampling design.

2.3 Tadpole: sampling and analysis

From the 8th to 10th of December, 2017 a subset of 26 ponds (19 of them invaded by bullfrogs) were sampled diurnally with hand nets to study tadpole assemblages (Dodd, 2010; Heyer et al., 2014). These ponds were selected because they were located within the center of the invasion and because they presented similar and comparable conditions. Pond area was determined using a hand-GPS. Tadpole sampling was done at five equally distant stations and covering different environments: three in the vegetated shallow areas and two in the deep areas. At each station, three standardized hand net passes of 2 m each were performed. The identity and number of tadpoles collected were recorded and then released. The percent of coverage by emerging and floating macrophytes was visually estimated. The presence of fish was considered as a binomial variable, according to the presence-absence of any fish in the hand net samples. The most frequent fish were Characidae and Poeciliidae. Fishes reach significant abundances in these ponds and have a generalist diet, thus excluding many amphibian species (Teixeira de Mello et al., 2011). This variable was recorded due to the strong available evidence showing that fish presence determines the structure of the amphibian community (Hecnar and M'Closkey, 1997).

Species richness and abundance of the native anuran tadpole assemblages between invaded and uninvaded ponds were compared. Invaded ponds were those considered as having any bullfrog record, regardless of density. A negative binomial GLM was used to test the association between tadpole richness and abundance and the presence-absence of bullfrogs following the above-mentioned procedure. Pond area, fish presence, and percentage of macrophyte cover were tested as possible explanatory variables in the model. The best models were selected using the LRT (Zuur et al., 2007). All data analyses were performed in R open software, with α = 0.05 as the criterion for statistical significance (R Core Team, 2019).

2.4 Graph analysis

Since American bullfrog dispersion consists in the occupation of discrete areas (permanent ponds), understanding and predicting its invasion process requires understanding the spatial structure and connectivity of these systems at the landscape level (Adams and Pearl, 2007). San Carlos' pond network was explored using graph theory, an approach that allows quantifying the risk of invasion based on the landscape structure, and the species' dispersal ability (Minor and Urban, 2008; Drake et al., 2017). Pond locations were obtained from the hydrological shape available at the Infraestructura de Datos Espaciales del Uruguay (https://visualizador.ide.uy/ideuy/core/load_public_project/ideuy`), with a buffer of 4650 m (the greatest distance between two invaded ponds) around the bullfrog invasion area that was mapped in San Carlos. Then, the euclidean distances between the centroids of each pond was determined, and those pairs of ponds that were distanced at less than 727 m (maximum dispersal distance of L. catesbeianus according to Descamps and De Vocht, 2016) were considered connected. Based on this criterion, the topological pond network was constructed. This model tacitly assumed that bullfrog movement ability was homogeneous across landscape, not considering other factors that could promote its dispersion (Fletcher et al., 2011; Drake et al., 2017), such as human activity and surrounding environmental features (e.g. lotic water bodies). In any case, the bullfrog is known to preferentially inhabit ponds, where it reproduces and grows, and only occasionally uses other aquatic environments such as creeks, streams or temporary ponds (Blaustein and Kiesecker, 2002). In addition, during our repeated surveys in the area we have not recorded the presence of bullfrog in lotic systems (unpublished data). Considering that the topography of the area is predominantly low-lying, it is unlikely that the bullfrog will move mostly through the water currents to invade the ponds. In any case, we do not have standardized data for the area that allows us to evaluate this possible bias.

Based on the graph model, the ponds that would play an important role (because of their location and connectivity in the network) in the local dispersion of the bullfrog in San Carlos were selected. These would be the ponds in which management is a priority to slow down the invasion: the stepping stones, peripheral hubs and connectors. The so-called stepping stones, those ponds that allow the shortest paths through the network, were obtained by the betweenness centrality metric, using the igraph R package. For the prioritization, 5% of the ponds with the highest betweenness centrality values were selected (Newman and Girvan, 2004; Csardi and Nepusz, 2006). Then, the modular structure of the network was analysed, and the topological role of the different ponds was determined using the rnetcarto R package. To evaluate whether the structure of the network was significantly modular, null models with the oecosimu function of the vegan R package were performed. The algorithm used was quasiswap, with 2000 iterations. Finally, peripheral hubs and connectors were selected, those ponds that should be prioritized due to their greatest number of connections between modules (Guimera and Amaral, 2005).

3 Results

3.1 Invasion status

Bullfrog was recorded in 32 of the 75 sampled ponds in San Carlos, occupying a total area of 16.5 km2. Seven ponds contained the highest bullfrog densities, eight had intermediate densities, and the remaining seventeen showed the lowest densities (Fig. 2). Invasion areas were located surrounding the San Carlos stream and its tributaries, especially upstream of the La Paja creek. Also, the distribution of the bullfrogs was mainly observed near national and neighborhood roads, where there were more human settlements and more water reservoirs (Fig. 2).

thumbnail Fig. 2

Distribution area of Lithobates catesbeianus in San Carlos, Maldonado Department (Uruguay). Estimated bullfrog density levels are shown: 0 in white, 1 in yellow, 2 in orange, 3 in dark red. Location of the old farms facilities (sky-blue stars): 1, registered farm (Laufer et al., 2018a), 2, unregistered farm. Circles: sample ponds, shaded areas encompass polygons of invaded ponds and a 727 m buffer area.

3.2 Effects on native post-metamorphic amphibians

The native post-metamorphic amphibians recorded in the 60 water bodies sampled were Boana pulchella, Pseudis minuta, Scinax granulatus, Scinax squalirrostris, Dendropsophus sanborni, Phyllomedusa iheringii, Leptodactylus luctator, Leptodactylus gracilis, Leptodactylus latinasus, Leptodactylus mystacinus, Odontophrynus asper Physalaemus gracilis, Pseudopaludicola falcipes, Rhinella arenarum and Rhinella dorbignyi. The ponds' species richness was explained by a statistically significant GLM model (Tab. 1) that included bullfrog presence (residual deviance ResDev = 56.2; P = 0.028) and sampling date (ResDev = 61.0; P = 0.27) as explanatory variables. The native species richness was 1.9 ± 1.1 (mean ± standard deviation) in invaded ponds and 2.5 ± 1.9 in uninvaded ponds (Fig. 3a). A significant relationship between the abundance of post-metamorphic native anurans and the presence-absence of bullfrogs at the pond level was unable to be identified (ResDev = 75.9, P = 0.87) (Fig. 3b). The selected GLM model for post-metamorphic native anurans abundance (Tab. 1), included wind intensity (ResDev = 67.3, P = 0.041), air temperature (ResDev = 71.4, P = 0.033) and sampling date (ResDev = 76.0, P = 0.20).

The three most frequent anuran species showed greater abundance variation in uninvaded ponds than in invaded ones, and a tendency to decrease with increasing bullfrog abundances (Tab. 2). For instance, P. minuta‘s abundance showed a statistically significant and abrupt decrease with an increase in bullfrog abundance (ResDev = 28.34; P = 0.002). However, this relation was not statistically significant for B. pulchella (ResDev = 40.00; P = 0.13) and L. luctator (ResDev = 52.06; P = 0.38).

Table 1

Selected GLM models (with their explained deviance and adjustment quality) for richness and abundance of native post-metamorphic amphibians, in San Carlos. All the used exploratory variables are listed. For each selected exploratory variable, the estimated coefficient, the residual deviance, the P-value, and the effect (positive “+” or negative “–”) are included. When a variable was not selected by the model it is indicated with a “NI” (not included). When a variable was not statistically significant it is indicated with a “NS”. The sampling date was always included as an explanatory variable, because it denotes the experimental design.

thumbnail Fig. 3

Richness (A) and abundance (B) of native post-metamorphic anurans in relation to bullfrog presence(1)-absence(0) in San Carlos, Maldonado Department, Uruguay. White boxes indicate uninvaded ponds and grey boxes bullfrog invaded ponds. Inside the boxes the median is shown as a horizontal line. n.s: not statistically significant.
thumbnail Fig. 4

Abundance of the three most frequently observed post-metamorphic native anurans (Boana pulchella, Leptodactylus luctator, and Pseudis minuta) in relation to bullfrog abundance in San Carlos, Maldonado Department, Uruguay. Black points are observations for each sampled pond, n.s. is not statistically significant, and the red line in the rightmost graph is the negative binomial GLM model (ResDev: Residual Deviance, P: P-value). The grey shade is the adjustment of the model for its standard deviation.
Table 2

GLM models (with their Explained deviance and Adjustment quality) for the abundance of the three most frequent native post-metamorphic amphibians, in San Carlos. For both exploratory variables, the estimated coefficient, the residual deviance, the P-value, and the effect (positive “+” or negative “–”) are included. When a variable was not statistically significant it is indicated with a “NS”. The sampling date was always included as an explanatory variable, because it denotes the experimental design. The sign “*” indicates a statistically significative effect of this factorial variable.

3.3 Effects on native tadpoles

The native tadpoles observed in the 26 sampled ponds were B. pulchella, P. minuta, Scinax spp. (S. granulatus and/or S. squalirostris), P. gracilis, Rhinella spp. (R. arenarum and/or R. dorbignyi) and Leptodactylus spp. (L. gracilis, L. latinasus and/or L. mystacinus). The number of native tadpole species was significantly higher in uninvaded and more vegetated ponds. The best GLM model obtained, explained native tadpole richness in the studied ponds when including bullfrog presence and macrophyte cover as variables (Tab. 3). Both exploratory variables were statistically significant (bullfrog presence: ResDev = 27.7; P = 0.048 and macrophyte cover: ResDev = 21.2; P = 0.011). In all cases, larval richness increased with increased vegetation cover (coefficient = 1.01). Water bodies invaded by bullfrogs had lower tadpole richness than those that were not invaded (Fig. 5a).The average number of native tadpole species was 1.2 ± 1.1 (mean ± SD) in invaded ponds, and 2.3 ± 1.2 in uninvaded ones.

Native tadpole abundance was also affected by bullfrog invasion (Tab. 3). This abundance was explained by a model that included the following statistically significant explanatory variables: presence of bullfrog (ResDev = 44.6; P = 0.026), pond area (ResDev = 31.2; P < 0.001) and presence of fish (ResDev = 26.3; P = 0.026). Native tadpole abundance was significantly lower in systems with fish (4.7 ± 9.4) than in systems without fish (7.0 ± 7.4), and the same pattern was observed in bullfrog-invaded ponds (3.7 ± 6.0) than in uninvaded ones (10.9 ± 11.8) (Fig. 5b). Both in invaded and uninvaded ponds, pond area was also a determinant of abundance. For instance, the abundance of tadpoles was inversely related to the pond area (coefficient = 1.00), and it dropped steeply in ponds over 500 m2.

thumbnail Fig. 5

Specific richness and abundance of native anuran tadpole assemblage in relation to bullfrog (A) and fish (B) presence(1)-absence (0) in San Carlos, Maldonado Department, Uruguay. White boxes refer to the uninvaded ponds and grey boxes to the invaded ponds. Inside the boxes the median is shown as a horizontal line.
Table 3

Selected GLM models (with their Explained deviance and Adjustment quality) for richness and abundance of native amphibian larvae, in San Carlos. All the used exploratory variables are listed. For each selected exploratory variable, the estimated coefficient, the residual deviance, the P-value, and the effect (positive “+” or negative “–”) are included. When a variable was not selected by the model it is indicated with a "NI" (not included).

3.4 Pond prioritization

The San Carlos pond network (mean area = 1026 m2, range = 10 to 64262 m2) was dense, highly interconnected and presented a strongly modular structure (modularity coefficient = 0.81; z = 387.1; P-value < 0.001). The invaded ponds belonged to four modules. The module that contained the area where the bullfrog farm was located, also contained 22 invaded ponds. The remaining six invaded ponds were located in three adjacent modules, at the edges of the mapped distribution (Fig. 6). A list of 55 ponds that should be prioritized for local management of the bullfrog invasion was obtained, due to its role in the network (Electronic material, Tab. E1 ). Seven of those ponds that act as connectors and two as peripheral hubs were identified. Some connectors were located close to invaded ponds. Furthermore, all invaded ponds were identified as ultraperipheral or peripheral nodes. From the 5% of ponds with the highest betweenness values (i.e. stepping stones), five were already invaded by bullfrogs. A significant proportion of these stepping stones were located close to the invasion front (Fig. 6). The list of ponds ranked by high connectivity between modules or by their role as stepping stones, and further information about the pond network, were included as supplementary material (Electronic material, Tab. E1, Fig. E1).

thumbnail Fig. 6

Topological network of ponds of San Carlos, Maldonado Department (Uruguay), according to distances related to the dispersal capacity of L. catesbeianus (left). Uninvaded ponds appear in blue and those invaded by L. catesbeianus, in red, orange and yellow (according to their degree of invasion, described in methods). San Carlos' pond network map shows the stepping stones (black middle point), the connectors (square symbol) and the peripheral hubs (triangle symbols). Bullfrog invaded ponds are in red on the map (right).

4 Discussion

Our study presents the first report on the distribution, density, impacts on the native amphibian assemblage, and prioritization of management strategies for ponds invaded by bullfrogs in Uruguay. The observations and analyses determined that the bullfrog population is relatively spread out in San Carlos. This knowledge is relevant in the context of a strong need to clearly diagnose the status of the small invasive populations in southern South America, especially in order to address their management (Pluess et al., 2012; Barbosa et al., 2017; Schwindt and Bortolus, 2017). In this sense, our observations are an interesting contribution, which also confirms bullfroǵs negative effects on local biodiversity (Li et al., 2011). In addition, the pond network analysis revealed that modules and a series of nodes that would have an important role in the dispersion of this invasive species were identified.

The invasive bullfrog population studied herein is the largest population known in Uruguay. Its estimated distribution area was 1.9 km2 (six ponds) in 2015 (Laufer et al., 2018a), while the current area determined is approximately six fold larger, occupying nearly 16.5 km2 (32 ponds). Before this study, the largest feral population in Uruguay was reported in Aceguá (Cerro Largo Department), where expansion began in 2012 (Laufer et al., 2018a). Considering that bullfrog distribution in San Carlos is even greater than in Aceguá, we can assume that the San Carlos population is in an expansion phase.

Through unstructured surveys with local people, a second unregistered farm was noted to have been working in the area, maintaining and fattening animals from the main farm, Laguna Dorada (unpublished data). Therefore, there were potentially two bullfrog propagule sources approximately one kilometer apart. This additional farm could have influenced the current distribution. In fact, the high bullfrog density area is close to both old farms. Interestingly, dispersal could have occurred from both areas in the same direction, upstream over the highways and roads (Fig. 2). This association between bullfrog presence and human activities could have favored dispersal. Specifically, in the case of San Carlos, the existence of many artificial lentic systems and water reservoirs, may favor the establishment of this invasion and dispersion (Jeschke and Strayer, 2006). Considering that the bullfrog is not a very selective species, choosing all kinds of permanent ponds (Nie et al., 1999), it is clear that its expansion is strongly affected by socio-economic activities. Ficetola and collaborators (2010) found that land use can explain the distribution of this species’ invasion. Undoubtedly, the human dimension must be included in the research and management of invasive alien species to achieve significant progress in understanding this phenomenon (Peterson et al., 2013; Ballari et al., 2016).

Results show a negative impact of bullfrog invasion on post-metamorphic native amphibian richness. This effect coincides with the evidence reported from other regions (reviewed by Kraus, 2009; but see Both and Melo, 2015) and suggested in Uruguay (Laufer et al., 2008; Lombardo et al., 2016; Laufer and Gobel, 2017). The observed decline in species richness is expected at the site scale (pond) due to the invasion of a large and abundant top predator (Oda et al., 2019). The impacts of a large predator can be drastic throughout the food web, and invasive predators often exert greater pressure than native predators (Sih et al., 2010; Thomsen et al., 2014). This difference may be due to the inability of native species to detect the new predator (Polo-Cavia et al., 2010), or the use of other invasion weapons which are novel to the native community (Strauss et al., 2012).

Although post-metamorph richness was statistically different in invaded and uninvaded systems, the effect of invasion on post-metamorph abundance was not significant. This lack of significance could be due to the great variation in abundance recorded in uninvaded systems. Furthermore, not all native anuran species are equally affected by bullfrogs. There may be species that are more tolerant and are, therefore, more abundant in invaded systems (Bucciarelli et al., 2014). For this reason, the analysis of the most common species can contribute to the understanding of the system.

The most affected species was Pseudis minuta, which is characterized by its post-metamorphic aquatic habits. Although we observed a drop in the abundance of Boana pulchella and Leptodactylus luctator, it was not statistically significant. Pseudis minuta has very similar microhabitat use and morphological conformation to L. catesbeianus, but it is much smaller; L. catesbeianus mean adult male Snout-Vent Length (SVL) is 131 mm (Kaefer et al., 2007), P. minuta mean adult male SVL is 31 mm (Melchiors et al., 2004). This species would be much more exposed to predation than other species because of its aquatic habits, smaller body size, and probable inability to recognize a new exotic predator (Polo-Cavia et al., 2010). Surely, the Pseudis genus of Neotropic aquatic frogs could be one of the genera most affected by bullfrog invasion in the region (Both and Melo, 2015; Silveira and Guimarães, 2021). This type of asymmetric effect, along with the encounter rate, have already been reported for the Northern Hemisphere (Pearl et al., 2004) and Southeastern Brazil (Silva et al., 2011). These asymmetries should be considered in the creation of management plans that seek to mitigate the impacts of this invasion.

The impacts of bullfrogs on native tadpoles were more noticeable than on post-metamorphs. This contrast is reasonable if the habits of the bullfrog as an invasive anuran are considered, and its greater interaction with aquatic organisms (Descamps and De Vocht, 2016). In addition to post-metamorph predation pressure, the existing evidence suggests that bullfrog tadpoles can also prey on eggs and small animals (Schiesari et al., 2009; Ruibal and Laufer, 2012), and may even affect the development and survival of native tadpoles (Kiesecker et al., 2001; Blaustein et al., 2020; Tasker et al., 2022).

Bullfrogs can coexist and also be benefited from predator fish (i.e. indirect trophic effects: Smith et al., 1999; Semlitsch et al., 2015), but they can also use other water habitats where fish are infrequent. The impacts of this invasion could be greater since bullfrogs show land dispersal and colonize water bodies where fish are absent or at low densities (Hecnar and M'Closkey, 1997). Thus, the ponds with restricted areas and/or hydroperiods, as well as suboptimal pH and oxygen conditions for fish can be invaded by bullfrogs (Descamps and De Vocht, 2016). All of these factors would severely restrict the availability of viable breeding sites for various native anurans, at the landscape level. Macrophyte coverage showed an important role in maintaining native tadpole richness. The importance of maintaining spatial heterogeneity, as a source of refuge and microhabitat for native species, and as a way to mitigate the impacts of the bullfrog invasion, should be emphasized. Macrophyte coverage favored the coexistence of a greater number of native species, as observed in the results (Hartel et al., 2007).

Most empirical evidence about bullfrog effects comes from studies of post-metamorphic stages. Nevertheless, aquatic pond communities, including tadpoles and other organisms, seem to be strongly affected as well (Gobel et al., 2019). Bullfrog invasions could affect the recruitment of species that depend on permanent ponds. They could even affect the ecosystem services associated with the aquatic systems in which they take place, at least in the long term (Moore and Hunt, 2012).

We must consider possible biases generated by the methodology used in the sampling, both of larvae and post-metamorphs, that could affect our results. The samplings were limited in time and seasons and therefore may reflect an underrepresentation (even an overrepresentation) of some amphibian species (i.e. Both et al., 2009). To better understand the effects and explore the underlying mechanisms, it would be necessary to increase sampling effort.

The obtained pond web and prioritization are valuable to prevent bullfrog expansion. From the network analysis, a list of 55 ponds that were of great importance to prevent or delay the expansion process of L. catesbeianus in San Carlos was generated (Electronic material, Tab. E1). Most of the prioritized ponds are uninvaded yet, and then are sites where the arrival of bullfrogs should be avoided (e.g. physical, chemical, and other measures). The relatively small number of invaded ponds and their restricted allocations to four modules, indicate that eradication is still possible. Kraus (2009) reviewed previous bullfrog control actions, identifying five cases of successful eradication in Europe (Great Britain, Netherlands and Germany). The success of these actions was explained by the low number of invaded ponds (maximum 53) detected at early stages, and because different management techniques were applied (i.e. pond draining and/or fencing, removing adults and tadpoles by aquatic traps, pit-fall traps, hand-capture, shooting and electrofishing). In recent years, the eradication of a population in Yosemite National Park, United States of America, was also reported with the use of seine net for tadpoles, and hand catching and shooting for post-metamorphs (Kamoroff et al., 2020). In addition, in Belgium, fyke nets have been used successfully to control bullfrogs (Louette et al., 2013; Descamps and De Vocht, 2023). These efforts can be complemented with a series of promising new techniques that can amplify the results (e.g. Groffen et al., 2019; Sutherland et al., 2019; Everts et al., 2022). Invaded pond distributions and position within the web, as well as bullfrog abundance at each pond, are useful data for planning this possible eradication in San Carlos. In fact, eradication should begin in the five ponds that have already been invaded by bullfrogs and were prioritized due to their high risk of regional dispersal (i.e. stepping stones).

Since the bullfrog is one of the priority species for control and/or eradication in Uruguay (Aber et al., 2012), the available rapid response protocol should be applied to this population restricted to a relatively small area in San Carlos (Comité de Especies Exóticas Invasoras, 2018a). The information provided here is useful for the diagnosis and planification to the Specialized Technical Group for Rapid Response, of the National Committee for Invasive Alien Species of the Ministry of the Environment. This committee already has the experience of an action plan for another bullfrog focus in Aceguá, in Northeastern Uruguay (Comité de Especies Exóticas Invasoras, 2018b). Unfortunately, the actions implemented there were few, isolated, and lacked coordination and involvement of the local social actors (personal observations), underestimating not only the management methods but also the socio-environmental nature of the problem (Pluess et al., 2012). This previous experience suggests that a successful bullfrog control in San Carlos would be difficult to achieve. In any case, we strongly recommend that the authorities implement eradication actions in San Carlos. This eradication must integrate multidisciplinary approaches, considering different interacting factors, such as scientific knowledge, social aspects, the ethics of wildlife management, as well as a fine knowledge of the invasion in the landscape context.

As a general conclusion, there is a significant risk of bullfrog expansion, with negative consequences to native San Carlos amphibians, that could be prevented. The information reported herein is essential to understand the situation and to predict scenarios and, therefore, achieve better conservation decisions. Our findings confirm the importance of maintaining strong monitoring and strengthening the determination of the actual distribution of invasive alien species (Simberloff et al., 2005). Therefore, conservation efforts should aim to eradicate bullfrogs, and prevent its expansion based on field data analysis.

Supplementary Material

Table E1: Ponds prioritized for bullfrog control, due to their high connectivity at the landscape level, in San Carlos, Maldonado (Uruguay). These are the priority ponds to attend to in an eradication plan; eliminating or preventing the arrival of the bullfrog would slow down the invasion process. For each pond, the geographic coordinates in UTM 21S, the invasion category of L. catesbeianus and the reason for its prioritization, are included.

Figure E1: Topological ponds network at San Carlos, Maldonado (Uruguay), according to distances related to the dispersal capacity of L. catesbeianus. The size of each node is proportional to the degree centrality value of each pond (top left). The size of each node is proportional to the value of betweenness centrality of each pond (top right). The squares represent the ponds with the topological role of connector, and the triangles represent the peripheral hubs (bottom left). In these first three, uninvaded ponds appear in blue and those invaded by L. catesbeianus, in red, orange and yellow (according to their degree of invasion, described in Methods). In the latter, the colors identify the different modules of the network. Invaded ponds appear in larger sizes (bottom right).

Access here

Acknowledgements

GL, NG and NK thank the Agencia Nacional de Investigación e Innovación (ANII) and the Programa de Desarrollo de las Ciencias Básicas (PEDECIBA), Uruguay, for their postgraduate grants. GL is a member of the Sistema Nacional de Investigadores (SNI), Uruguay. This research was partially supported by a small grant from the Rufford Foundation and by the Agencia Nacional de Investigación e Innovación, Uruguay, ANII_FMV_3_2020_1_162548. We thank the support of the local residents of the studied areas for their contributions during fieldwork. We appreciate the expert advice on topological network methodology, provided by Mariana Illarze, Ana Borthagaray and Matias Arim. All the animal welfare protocols used were authorized by the National Commission for Animal Experimentation, Museo Nacional de Historia Natural (Code 013/11).

References

  • Aber A, Ferrari G, Porcile JF, Rodríguez E, Zerbino S. 2012. Identificación de prioridades para la gestión nacional de las especies exóticas invasoras, MVOTMA-DINAMA, Montevideo. [Google Scholar]
  • Adams M, Pearl C. 2007. Problems and opportunities managing invasive Bullfrogs: is there any hope? In: Gherardi F, ed. Biological invaders in inland waters: Profiles, distribution, and threats. Dordrecht: Springer pp. 679–693. [CrossRef] [Google Scholar]
  • Álvarez A, Blum A, Gallego F. 2015. Atlas de Cobertura de Suelos del Uruguay, DINOT, FAO, Montevideo. [Google Scholar]
  • Arrieta D, Borteiro C, Kolenc F, Langone JA. 2013. Anfibios. In Soutullo A, Clavijo C, Martínez-Lanfranco JA eds. Especies prioritarias para la conservación en Uruguay: Vertebrados, moluscos continentales y plantas vasculares, SNAP, MVOTMA, Montevideo pp. 113–127. [Google Scholar]
  • Ballari SA, Anderson CB, Valenzuela AEJ. 2016. Understanding trends in biological invasions by introduced mammals in southern South America: a review of research and management: Invasive mammals in southern South America. Mamm Rev 46: 229–240. [CrossRef] [Google Scholar]
  • Barbosa FG, Both C, Araújo MB. 2017. Invasive American bullfrogs and African clawed frogs in South America: high suitability of occurrence in biodiversity hotspots. Zool Stud 56: e28. [PubMed] [Google Scholar]
  • Bissattini AM, Buono V, Vignoli L. 2018. Field data and worldwide literature review reveal that alien crayfish mitigate the predation impact of the American bullfrog on native amphibians. Aquat Conserv 28: 1465–1475. [CrossRef] [Google Scholar]
  • Bissattini AM, Buono V, Vignoli L. 2019. Disentangling the trophic interactions between American bullfrogs and native anurans: Complications resulting from post‐metamorphic ontogenetic niche shifts. Aquat Conserv 29: 270–281. [CrossRef] [Google Scholar]
  • Blaustein AR, Jones DK, Urbina J, et al. 2020. Effects of invasive larval bullfrogs (Rana catesbeiana) on disease transmission, growth and survival in the larvae of native amphibians. Biol Invasions 22: 1771–1784. [CrossRef] [Google Scholar]
  • Blaustein AR, Kiesecker JM. 2002. Complexity in conservation: lessons from the global decline of amphibian populations. Ecol Lett 5: 597–608. [CrossRef] [Google Scholar]
  • Both C, Grant T. 2012. Biological invasions and the acoustic niche: the effect of bullfrog calls on the acoustic signals of white-banded tree frogs. Biol Lett 8: 714–716. [CrossRef] [PubMed] [Google Scholar]
  • Both C, Melo AS. 2015. Diversity of anuran communities facing bullfrog invasion in Atlantic Forest ponds. Biol Invasions 17: 1137–1147. [CrossRef] [Google Scholar]
  • Both C, Solé M, Dos Santos TG, Cechin SZ. 2009. The role of spatial and temporal descriptors for neotropical tadpole communities in southern Brazil. Hydrobiologia 624: 125–138. [CrossRef] [Google Scholar]
  • Brazeiro A. 2015. Eco-regiones de Uruguay: biodiversidad, presiones y conservación: aportes a la Estrategia Nacional de Biodiversidad, Universidad de la República, Montevideo. [Google Scholar]
  • Bucciarelli GM, Blaustein AR, Garcia TS, Kats LB. 2014. Invasion complexities: the diverse impacts of nonnative species on amphibians. Copeia 2014: 611–632. [CrossRef] [Google Scholar]
  • Comité de Especies Exóticas Invasoras. 2018a. Protocolo de Respuesta ante Invasiones Biológicas de Especies Exóticas Invasoras. [online] Available at: https://www.gub.uy/ministerio-ambiente/comunicacion/publicaciones [ Accessed 12 May 2022]. [Google Scholar]
  • Comité de Especies Exóticas Invasoras. 2018b. Plan Piloto de Erradicación de Rana toro, en Aceguá (Cerro Largo). [online] Available at: https://www.gub.uy/ministerio-ambiente/comunicacion/publicaciones [ Accessed 12 May 2022]. [Google Scholar]
  • Cook MT, Heppell SS, Garcia TS. 2013. Invasive bullfrog larvae lack developmental plasticity to changing hydroperiod. J Wildl Manag 77: 655–662. [CrossRef] [Google Scholar]
  • Csardi CN, Nepusz T (2006) The igraph software package for complex network research. Interjournal Complex Syst 1695:1–9. [Google Scholar]
  • Descamps S, De Vocht A. 2016. Movements and habitat use of the invasive species Lithobates catesbeianus in the valley of the Grote Nete (Belgium). Belg J Zool 146: 90–100. [Google Scholar]
  • Descamps S, De Vocht A. 2023. State-of-the-art approach on the management of invasive faunistic aquatic alien species: the American bullfrog in Belgium. Environ Challenges 11: 100690. [CrossRef] [Google Scholar]
  • Di Minin E, Soutullo A, Bartesaghi L, Rios M, Szephegy MN, Moilanen AT. 2017. Integrating biodiversity, ecosystem services and socio-economic data to identify priority areas and landowners for conservation actions at the national scale. Biol Conserv 206: 56–64. [CrossRef] [Google Scholar]
  • Dodd CK. 2010. Amphibian ecology and conservation: a handbook of techniques. Oxford: Oxford University Press. [Google Scholar]
  • Drake JC, Griffis-Kyle KL, McIntyre NE. 2017. Graph theory as an invasive species management tool: case study in the Sonoran Desert. Landsc Ecol 32: 1739–1752. [CrossRef] [Google Scholar]
  • Everts T, Van Driessche C, Neyrinck S, et al. 2022. Using quantitative eDNA analyses to accurately estimate American bullfrog abundance and to evaluate management efficacy. Environ DNA 00: 1–13. [Google Scholar]
  • Ficetola GF, Maiorano L, Falcucci A, et al. 2010. Knowing the past to predict the future: land-use change and the distribution of invasive bullfrogs. Glob Chang Biol 16: 528–537. [CrossRef] [Google Scholar]
  • Fisher RN, Shaffer HB. 1996. The decline of amphibians in California's Great Central Valley. Conserv Biol 10: 1387–1397. [CrossRef] [Google Scholar]
  • Fletcher RJ, Acevedo MA, Reichert BE, Pias KE, Kitchens WM. 2011. Social network models predict movement and connectivity in ecological landscapes. Proc Natl Acad Sci 108: 19282–19287. [CrossRef] [PubMed] [Google Scholar]
  • Garner TW, Perkins MW, Govindarajulu P, et al. 2006. The emerging amphibian pathogen Batrachochytrium dendrobatidis globally infects introduced populations of the North American bullfrog, Rana catesbeiana. Biol lett 2: 455–459. [CrossRef] [PubMed] [Google Scholar]
  • Guimera R, Amaral LAN. 2005. Cartography of complex networks: modules and universal roles. J Stat Mech 2005: P 02001. [Google Scholar]
  • Gobel N, Laufer G, Cortizas S. 2019. Changes in aquatic communities recently invaded by a top predator: evidence of American bullfrogs in Aceguá, Uruguay. Aquat Sci 81: 8. [CrossRef] [Google Scholar]
  • Gobel N, Laufer G, Gonzalez-Bergonzoni I, Soutullo A, Arim M. 2023. Invariant and vulnerable food web components after bullfrog invasion. Biol Invasions 25: 901–916. [CrossRef] [Google Scholar]
  • Grattarola F, Martínez-Lanfranco JA, Botto G, et al. 2020. Multiple forms of hotspots of tetrapod biodiversity and the challenges of open-access data scarcity. Sci Rep 10: 1–15. [CrossRef] [PubMed] [Google Scholar]
  • Groffen J, Kong S, Jang Y, Borzee A. 2019. The invasive American bullfrog (Lithobates catesbeianus) in the Republic of Korea: history and recommendations for population control. Manag Biol Invasions 10: 517. [CrossRef] [Google Scholar]
  • Hartel T, Nemes S, Cogălniceanu D, et al. 2007. The effect of fish and aquatic habitat complexity on amphibians. Hydrobiologia 583: 173–182. [CrossRef] [Google Scholar]
  • Hecnar SJ, M'Closkey RT. 1997. Changes in the composition of a ranid frog community following bullfrog extinction. Am Midl Nat 1997: 145–150. [CrossRef] [Google Scholar]
  • Heyer R, Donnelly MA, Foster M, Mcdiarmid R. 2014. Measuring and monitoring biological diversity: standard methods for amphibians. Washington DC: Smithsonian Institution Press. [Google Scholar]
  • Jancowski K, Orchard S. 2013. Stomach contents from invasive American bullfrogs Rana catesbeiana (= Lithobates catesbeianus) on southern Vancouver Island, British Columbia, Canada. NeoBiota 16: 17. [Google Scholar]
  • Jeschke JM, Strayer DL. 2006. Determinants of vertebrate invasion success in Europe and North America. Glob Chang Biol 12: 1608–1619. [CrossRef] [Google Scholar]
  • Kaefer ÍL, Boelter RA, Cechin SZ. 2007. Reproductive biology of the invasive bullfrog Lithobates catesbeianus in southern Brazil. Annales Zoologici Fennici 44: 435–444. [Google Scholar]
  • Kamoroff C, Daniele N, Grasso RL, Rising R, Espinoza T, Goldberg CS. 2020. Effective removal of the American bullfrog (Lithobates catesbeianus) on a landscape level: long term monitoring and removal efforts in Yosemite Valley, Yosemite National Park. Biol Invasions 22: 617–626. [Google Scholar]
  • Kats LB, Ferrer RP. 2003. Alien predators and amphibian declines: review of two decades of science and the transition to conservation. Divers Distrib 9: 99–110. [CrossRef] [Google Scholar]
  • Kiesecker JM, Blaustein AR, Miller CL. 2001. Potential mechanisms underlying the displacement of native red-legged frogs by introduced bullfrogs. Ecology 82: 1964–1970. [CrossRef] [Google Scholar]
  • Kraus F. 2009. Alien reptiles and amphibians: a scientific compendium and analysis. Dordrecht: Springer. [Google Scholar]
  • Kupferberg S. 1997. Facilitation of periphyton production by tadpole grazing: functional differences between species. Freshw Biol 37: 427–439. [CrossRef] [Google Scholar]
  • Latombe G, Pyšek P, Jeschke JM, Blackburn TM, Bacher S, Capinha C, et al. 2016. A vision for global monitoring of biological invasions. Biol Conserv 213: 295–308. [Google Scholar]
  • Laufer G, Canavero A, Núñez D, Maneyro R. 2008. Bullfrog (Lithobates catesbeianus) invasion in Uruguay. Biol Invasions 10: 1183–1189. [CrossRef] [Google Scholar]
  • Laufer G, Gobel N. 2017. Habitat degradation and biological invasions as a cause of amphibian richness loss: a case report in Aceguá, Cerro Largo, Uruguay. Phyllomedusa 16: 289–293. [CrossRef] [Google Scholar]
  • Laufer G, Gobel N, Berazategui M, et al. 2021. American bullfrog (Lithobates catesbeianus) diet in Uruguay compared with other invasive populations in Southern South America. North-West J Zool 17: 196–203. [Google Scholar]
  • Laufer G, Gobel N, Borteiro C, Soutullo A, Martínez-Debat C, de Sá RO. 2018a. Current status of American bullfrog, Lithobates catesbeianus, invasion in Uruguay and exploration of chytrid infection. Biol Invasions 20: 285–291. [CrossRef] [Google Scholar]
  • Laufer G, Gobel N, Kacevas N, Lado I. 2018b. Una nueva población feral de rana toro (Lithobates catesbeianus) en Uruguay, encontrada con participación ciudadana. Rev Latinoam Herpetol 1: 47–50. [CrossRef] [Google Scholar]
  • Laufer G, Gobel N, Soutullo A, Martínez-Debat C, de Sá RO. 2017. Assessment of the calling detection probability throughout the day of two invasive populations of bullfrog (Lithobates catesbeianus) in Uruguay. Cuadernos Herpetol 31: 29–32. [Google Scholar]
  • Lesbarrères D, Balseiro A, Brunner J, et al. 2012. Ranavirus: past, present and future. Biol Lett 8: 481–483. [CrossRef] [PubMed] [Google Scholar]
  • Li Y, Ke Z, Wang Y, Blackburn TM. 2011. Frog community responses to recent American bullfrog invasions. Curr Zool 57: 83–92. [CrossRef] [Google Scholar]
  • Liu X, Wang S, Ke Z, et al. 2018. More invaders do not result in heavier impacts: The effects of non‐native bullfrogs on native anurans are mitigated by high densities of non‐native crayfish. J Anim Ecol 87: 850–862. [CrossRef] [MathSciNet] [PubMed] [Google Scholar]
  • Lombardo I, Elgue E, Villamil J, Maneyro R. 2016. Registro de una población asilvestrada de rana toro (Lithobates catesbeianus) (Amphibia: Anura: Ranidae) en el departamento de Maldonado, Uruguay. Bol Soc Zool Uruguay 25: 61–65. [Google Scholar]
  • Louette G, Devisscher S, Adriaens T. 2013. Control of invasive American bullfrog Lithobates catesbeianus in small shallow water bodies. Eur J Wildl Res 59: 105–114. [CrossRef] [Google Scholar]
  • Melchiors J, Di-Bernardo M, Pontes GMF, de Oliveira RB, Solé M, Kwet A. 2004. Reproduction of Pseudis minuta (Anura, Hylidae) in southern Brazil. Phyllomedusa 3: 61–68. [CrossRef] [Google Scholar]
  • Minor ES, Urban DL. 2008. A graph-theory framework for evaluating landscape connectivity and conservation planning. Conserv Biol 22: 297–307. [CrossRef] [PubMed] [Google Scholar]
  • Minowa S, Senga Y, Miyashita T. 2008. Microhabitat Selection of the Introduced Bullfrogs (Rana catesbeiana) in Paddy Fields in Eastern Japan. Curr Herpetol 27: 55–59. [CrossRef] [Google Scholar]
  • Moore TL, Hunt WF. 2012. Ecosystem service provision by stormwater wetlands and ponds − a means for evaluation? Water Res 46: 6811–6823. [CrossRef] [PubMed] [Google Scholar]
  • Moreira LFB, Machado IF, Lace ARGM, Maltchik L. 2007. Calling period and reproductive modes in an anuran community of a temporary pond in southern Brazil. S Am J Herpetol 2: 129–135. [CrossRef] [Google Scholar]
  • Newman M, Girvan M (2004) Finding and evaluating community structure in networks. Phys Rev E 69:1–16. [Google Scholar]
  • Nie M, Crim JD, Ultsch GR. 1999. Dissolved oxygen, temperature, and habitat selection by bullfrog (Rana catesbeiana) tadpoles. Copeia 1999: 153–162. [CrossRef] [Google Scholar]
  • Oda FH, Guerra V, Grou E, et al. 2019. Native anuran species as prey of invasive American Bullfrog, Lithobates catesbeianus, in Brazil: a review with new predation records. Amphib Reptile Conserv 13: 217–226. [Google Scholar]
  • 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. [CrossRef] [Google Scholar]
  • Peterson AC, Richgels KL, Johnson PT, McKenzie VJ. 2013. Investigating the dispersal routes used by an invasive amphibian, Lithobates catesbeianus, in human-dominated landscapes. Biol Invasions 15: 2179–2191. [CrossRef] [Google Scholar]
  • Pluess T, Jarošík V, Pyšek P, et al. 2012. Which factors affect the success or failure of eradication campaigns against alien species? PloS ONE 7: e48157. [CrossRef] [PubMed] [Google Scholar]
  • Polo-Cavia N, Gonzalo A, López P, Martín J. 2010. Predator recognition of native but not invasive turtle predators by naïve anuran tadpoles. Anim Behav 80: 461–466. [CrossRef] [Google Scholar]
  • R Core Team. 2019. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/. [Google Scholar]
  • Ribeiro LP, Carvalho T, Becker CG, et al. 2019. Bullfrog farms release virulent zoospores of the frog-killing fungus into the natural environment. Sci Rep 9: 1–10. [PubMed] [Google Scholar]
  • Ricciardi A, Blackburn TM, Carlton JT, et al. 2017. Invasion science: A horizon scan of emerging challenges and opportunities. Trends Ecol Evol 32: 464–474. [CrossRef] [PubMed] [Google Scholar]
  • Ruibal M, Laufer G. 2012. Bullfrog Lithobates catesbeianus (Amphibia: Ranidae) tadpole diet: description and analysis for three invasive populations in Uruguay. Amphib-Reptil 33: 355–363. [CrossRef] [Google Scholar]
  • Schiesari L, Werner EE, Kling GW. 2009. Carnivory and resource-based niche differentiation in anuran larvae: implications for food web and experimental ecology. Freshw Biol 54: 572–586. [CrossRef] [Google Scholar]
  • Schloegel LM, Toledo LF, Longcore JE, et al. 2012. Novel, panzootic and hybrid genotypes of amphibian chytridiomycosis associated with the bullfrog trade. Mol Ecol 21: 5162–5177. [CrossRef] [PubMed] [Google Scholar]
  • Schwindt E, Bortolus A. 2017. Aquatic invasion biology research in South America: geographic patterns, advances and perspectives. Aquat Ecosyst Health Manag 20: 322–333. [CrossRef] [Google Scholar]
  • Seebens H, Blackburn TM, Dyer EE, et al. 2017. No saturation in the accumulation of alien species worldwide. Nat Commun 8: 1–9. [CrossRef] [PubMed] [Google Scholar]
  • Semlitsch RD, Peterman WE, Anderson TL, Drake DL, Ousterhout BH. 2015. Intermediate pond sizes contain the highest density, richness, and diversity of pond-breeding amphibians. PLoS ONE 10: e0123055. [CrossRef] [PubMed] [Google Scholar]
  • Sih A, Bolnick DI, Luttbeg B, et al. 2010. Predator-prey naïveté, antipredator behavior, and the ecology of predator invasions. Oikos 119: 610–621. [CrossRef] [Google Scholar]
  • Silva ETD, Both C, Ribeiro Filho OP. 2016. Food habits of invasive bullfrogs and native thin-toed frogs occurring in sympatry in southeastern Brazil. S Am J Herpetol 11: 25–33. [CrossRef] [Google Scholar]
  • Silva ETD, Ribeiro Filho OP, Feio RN. 2011. Predation of Native Anurans by invasive bullfrogs in southeastern Brazil: spatial variation and effect of microhabitat use by prey. S Am J Herpetol 6: 1–10. [CrossRef] [Google Scholar]
  • Silveira SDS, Guimarães M. 2021. The enemy within: consequences of the invasive bullfrog on native anuran populations. Biol Invasions 23: 373–378. [CrossRef] [Google Scholar]
  • Simberloff D, Parker IM, Windle PN. 2005. Introduced species policy, management, and future research needs. Front Ecol Environ 3: 12–20. [CrossRef] [Google Scholar]
  • Smith GR, Rettig JE, Mittelbach GG, Valiulis JM, Schaack SR. 1999. The effects of fish on assemblages of amphibians in ponds: a field experiment. Freshw Biol 41: 829–837. [CrossRef] [Google Scholar]
  • Speziale KL, Lambertucci SA, Carrete M, Tella JL. 2012. Dealing with non-native species: what makes the difference in South America? Biol Invasions 14: 1609–1621. [CrossRef] [Google Scholar]
  • Strauss A, White A, Boots M. 2012. Invading with biological weapons: the importance of disease-mediated invasions. Funct Ecol 26: 1249–1261. [Google Scholar]
  • Sutherland WJ, Dicks LV, Ockendon N, Smith RK. 2019. What Works in Conservation. Cambridge: Open Book Publishers. [Google Scholar]
  • Tasker BR, Honebein KN, Erickson AM, et al. 2022. Effects of elevated temperature, reduced hydroperiod, and invasive bullfrog larvae on pacific chorus frog larvae. PLoS ONE 17: e0265345. [CrossRef] [PubMed] [Google Scholar]
  • Teixeira de Mello F, González-Bergonzoni I,Loureiro M. 2011. Peces de agua dulce del Uruguay. Montevideo: PPR-MGAP. [Google Scholar]
  • Thomsen MS, Byers JE, Schiel DR, et al. 2014. Impacts of marine invaders on biodiversity depend on trophic position and functional similarity. Mar Ecol Prog Ser 495: 39–47. [CrossRef] [Google Scholar]
  • Zuur AF, Ieno EN, Smith GM. 2007. Analyzing ecological data. New York: Springer. [Google Scholar]

Cite this article as: Laufer G, Gobel N, Kacevas N, Lado I. 2023. American bullfrog (Lithobates catesbeianus) distribution, impact on native amphibians and management priorities in San Carlos, Uruguay. Knowl. Manag. Aquat. Ecosyst., 424, 20.

All Tables

Table 1

Selected GLM models (with their explained deviance and adjustment quality) for richness and abundance of native post-metamorphic amphibians, in San Carlos. All the used exploratory variables are listed. For each selected exploratory variable, the estimated coefficient, the residual deviance, the P-value, and the effect (positive “+” or negative “–”) are included. When a variable was not selected by the model it is indicated with a “NI” (not included). When a variable was not statistically significant it is indicated with a “NS”. The sampling date was always included as an explanatory variable, because it denotes the experimental design.

In the text
Table 2

GLM models (with their Explained deviance and Adjustment quality) for the abundance of the three most frequent native post-metamorphic amphibians, in San Carlos. For both exploratory variables, the estimated coefficient, the residual deviance, the P-value, and the effect (positive “+” or negative “–”) are included. When a variable was not statistically significant it is indicated with a “NS”. The sampling date was always included as an explanatory variable, because it denotes the experimental design. The sign “*” indicates a statistically significative effect of this factorial variable.

In the text
Table 3

Selected GLM models (with their Explained deviance and Adjustment quality) for richness and abundance of native amphibian larvae, in San Carlos. All the used exploratory variables are listed. For each selected exploratory variable, the estimated coefficient, the residual deviance, the P-value, and the effect (positive “+” or negative “–”) are included. When a variable was not selected by the model it is indicated with a "NI" (not included).

In the text

All Figures

thumbnail Fig. 1

Photographs of two ponds sampled in San Carlos, Maldonado Department (Uruguay), showing the differences in macrophyte cover.
In the text
thumbnail Fig. 2

Distribution area of Lithobates catesbeianus in San Carlos, Maldonado Department (Uruguay). Estimated bullfrog density levels are shown: 0 in white, 1 in yellow, 2 in orange, 3 in dark red. Location of the old farms facilities (sky-blue stars): 1, registered farm (Laufer et al., 2018a), 2, unregistered farm. Circles: sample ponds, shaded areas encompass polygons of invaded ponds and a 727 m buffer area.
In the text
thumbnail Fig. 3

Richness (A) and abundance (B) of native post-metamorphic anurans in relation to bullfrog presence(1)-absence(0) in San Carlos, Maldonado Department, Uruguay. White boxes indicate uninvaded ponds and grey boxes bullfrog invaded ponds. Inside the boxes the median is shown as a horizontal line. n.s: not statistically significant.
In the text
thumbnail Fig. 4

Abundance of the three most frequently observed post-metamorphic native anurans (Boana pulchella, Leptodactylus luctator, and Pseudis minuta) in relation to bullfrog abundance in San Carlos, Maldonado Department, Uruguay. Black points are observations for each sampled pond, n.s. is not statistically significant, and the red line in the rightmost graph is the negative binomial GLM model (ResDev: Residual Deviance, P: P-value). The grey shade is the adjustment of the model for its standard deviation.
In the text
thumbnail Fig. 5

Specific richness and abundance of native anuran tadpole assemblage in relation to bullfrog (A) and fish (B) presence(1)-absence (0) in San Carlos, Maldonado Department, Uruguay. White boxes refer to the uninvaded ponds and grey boxes to the invaded ponds. Inside the boxes the median is shown as a horizontal line.
In the text
thumbnail Fig. 6

Topological network of ponds of San Carlos, Maldonado Department (Uruguay), according to distances related to the dispersal capacity of L. catesbeianus (left). Uninvaded ponds appear in blue and those invaded by L. catesbeianus, in red, orange and yellow (according to their degree of invasion, described in methods). San Carlos' pond network map shows the stepping stones (black middle point), the connectors (square symbol) and the peripheral hubs (triangle symbols). Bullfrog invaded ponds are in red on the map (right).
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.