South American Journal of Herpetology, 4(3), 2009, 286-294 © 2009 Brazilian Society of Herpetology

Diet of the invasive Lithobates catesbeianus (Shaw, 1802) (anura: ranidae) in Viçosa, Minas Gerais State, Brazil

Emanuel Teixeira da Silva1,3, Evelyze Pinheiro dos Reis2, Renato Neves Feio1 and Oswaldo Pinto Ribeiro Filho1

1 Programa de Pós-graduação em Biologia , Departamento de Biologia Animal. 2 Programa de Pós-graduação em Genética e Melhoramento, Departamento de Biologia Geral, Universidade Federal de Viçosa, 36571‑000, Viçosa, MG, Brasil. 3 Corresponding author. E‑mail: [email protected]

Abstract. Based on the stomach content analysis of 113 individuals, the diet of the invasive Lithobates catesbeianus (American Bullfrog) was examined in four sites located within the municipality of Viçosa (20°45’S and 42°51’W), state of Minas Gerais, Brazil, from August 2005 to March 2007. The effects of frog size and sexual maturity on stomach contents were determined. Prey items were grouped according to their primary , being classified as aquatic, terrestrial and amphibious. In general, the most frequent prey categories were post-metamorphic Anura, Diplopoda, Hemiptera, Hymenoptera Formicidae and Araneae. The diet of adults of both sexes was similar, but differed from the diet of young . Terrestrial prey were most abundant both in number and occurrence. For adult Bullfrogs, amphibious prey were most significant in volume. There was a significant correlation between prey and predator sizes, as well as a greater consumption of native anurans by larger Bullfrogs. The results confirmed that Bullfrogs have a generalist feeding habit that can have important negative effects on native amphibian communities in environments occupied by this invasive species.

Keywords. Feeding habits; biological invasion; American Bullfrog; Lithobates catesbeianus.

Introduction It is well known that intentional introductions and farming escapes due to harvest deficiencies lead to Native to North America, Lithobates catesbeia- population establishment in aquatic sites near frog nus (Shaw, 1802), commonly known as the Bullfrog, farms (Hammerson, 1982; Bury and Whelan, 1984). was introduced in several countries for commer- The occurrence of invasive populations of the Bull- cial harvest (Moyle, 1973; Bury and Whelan, 1984; frog have been reported in the south, southeast and Giovanelli et al., 2008). The Bullfrog is the largest central regions of Brazil (Guix, 1990; Batista, 2002; anuran species in North America. It is mostly aquatic Martins et al., 2002; Boelter and Cechin, 2007; and is a generalist predator, reflecting prey availabil- Giovanelli et al., 2008; Kaefer et al., 2007; Reis ity, including cannibalism (Bury and Whelan, 1984). et al., 2007; Silva et al., 2007). Given the general- The feeding habits of L. catesbeianus have been ist predatory habits of the Bullfrog, its introduction studied several times and many of these studies de- into natural environments is a cause for concern and scribed the occurrence of uncommon prey types for prompts further investigation. Its ability to prey on an anuran, such as moles, mice, bats, birds, snakes, other anurans, as well as competitive effects among lizards, turtles, small alligators, salamanders and tadpoles (Kupferberg, 1997; Lawler et al., 1999; Kie- fish (Korschgen and Moyle, 1955; Cohen and How- secker et al., 2001) and possible pathogen transmis- ard, 1958; Korschgen and Baskett, 1963; Brooks Jr., sion (Hanselmann et al., 2004; Garner et al., 2006; 1964; Corse and Metter, 1980; Bury and Whelan, Barrasso et al., 2009), makes the Bullfrog a possible 1984; Silva et al., 2007; Camargo Filho et al., 2008). agent of amphibian population decline at its introduc- The Bullfrog was chosen for commercial exploi- tion sites (Hayes and Jennings, 1986; Kats and Ferrer, tation because of its high fecundity, which results in 2003; Pearl et al., 2004). greater performance in captivity than other frog spe- In the municipality of Viçosa, state of Minas cies (Vizotto, 1984; Fontanelo, 1994). In Brazil, the Gerais, the first Bullfrogs arrived in the early 1980’s first individuals of L. catesbeianus arrived in 1935 after the construction of the “Ranário Experimental” (Fontanello, 1994), and since then tadpoles and ma- (RE; experimental frog farming) at the Universidade ture frogs were freely given to producers in order to Federal de Viçosa (UFV) campus (Lima, 1994). Since stimulate the cultivation of this species. This spe- then, the RE has become a source of tadpoles and im- cies easily colonized different ecosystems in Brazil, mature frogs that have established invasive popula- which lead to the recommendation of its production tions in aquatic sites located in surrounding areas, due in many Brazilian states (Fontanello, 1994). to escapes from the RE. The aim of the present study Silva, E. T. et al. 287 was to describe the diet of L. catesbeianus in aquatic (Boelter and Cechin, 2007). The samples were then environments on the UFV campus in an attempt to deposited in the herpetological collection at “Museu analyze the effects of frog size, sex and sexual matu- de Zoologia João Moojen” (MZUFV). In the labora- rity on the prey types consumed. tory, frogs were weighed to the nearest 1 g, and the snout-vent length (SVL) and jaw width (JW) were measured to the nearest 0.05 mm using calipers. The Material and methods individuals were grouped in seven size classes based on the SVL, adapting the classification of Lima et al. Study sites (1998): class 1 (< 50 mm); class 2 (50.05‑70 mm); class 3 (70.05‑90 mm); class 4 (90.05‑110 mm); class Bullfrog specimens were collected at four sites lo- 5 (110.05‑130 mm); class 6 (130.05‑150 mm) and cated on the campus of the Universidade Federal de class 7 (> 150.05 mm). Viçosa, in the municipality of Viçosa (20°45’S and Frogs were dissected to remove the stomachs 42°07’W), Minas Gerais state, southeastern Brazil. which were preserved in 70% ethanol. The individu- The municipality of Viçosa lies within the Atlantic als were sexed by gonad analysis and development Rainforest domain; the region was originally covered of the secondary sexual traits (diameter of tympanic by semi-deciduous forest. Currently, secondary forest membrane, throat coloration and swollen thumbs; fragments surrounded by agriculture, pastures, and Bury and Whelan, 1984). Females (young and ma- eucalyptus plantations make up the landscape (Ribon ture) were separated based on gonad development et al., 2003; Coelho et al., 2005). (Costa et al., 1998). Males weighing more than 45 g The first sampling site, known as the “Represa do were classified as adults (Lima et al., 1998). Three Belvedere”, consists of a group of small connected groups were considered in diet analyses: adult males, dams surrounded by a secondary forest fragment and adult females and juveniles. a grassplot. The dams have abundant aquatic vege- The stomach content was analyzed with the help tation composed of Salvinia sp. and Nympheaceae. of a stereomicroscope. Food items were determined The second site, known as “Estação Experimental to the lowest possible taxonomic level. The plant de Fruticultura” is an area designated for the experi- remains found were considered to be accidently in- mental cultivation of fruit trees, with four small dams gested. Prey items were grouped according to their originally built to serve as water reservoirs for irriga- primary habitat, being classified as aquatic, terrestrial tion. The shoreline herbaceous vegetation is managed and amphibious. The length and maximum width of regularly. There are a few aquatic plants, including each intact prey item was measured with calipers (to Nympheaceae. The third sampling site consists of the nearest 0.05 mm) (Wu et al., 2005). Individual two artificial reservoirs located in an area of experi- prey volume (mm3) was calculated using the formula mental grain production. The shoreline is dominated for an ellipsoid (Magnusson et al., 2003): by Poaceae and sparse tree coverage, and there is no evident aquatic vegetation. The last site is an experi- Prey volume = 4/3 π (lenght/2) (width/2)2 mental fish station located near the Ranário Experi- mental site. It consists of 95 dams of different sizes, The total volume (cm3) of prey categories was surrounded by regularly cut herbaceous vegetation. estimated by water displacement (Magnusson et al., Some of the dams have aquatic vegetation, with the 2003). This method was used because of the presence presence of Eichhornia crassipes and Nympheaceae. of fragmented prey items that could be identified but not measured, and could be grouped to estimate the total volume of each prey sample. Data collection

Frogs were captured at night (18:30‑22:00 h) from Data analysis August 2005 to March 2007. Field activities in sites 1 and 2 were carried out twice per month, and sites 3 Frogs with empty stomachs or with only plant re- and 4 were sampled sporadically, starting in August mains were not considered in the analysis. The fre- 2006. Frogs were captured by hand with the help of quency of prey occurrence per Bullfrog (FO = 100 nets and an air rifle. The specimens collected were x number of stomachs containing the prey category killed in situ and stored on ice to retard digestion t divided by the total number of stomachs), relative 288 Diet of Lithobates catesbeianus

Table 1. Snout-vent length (SVL), prey length (all these in mm), and prey volume (mm3) of Lithobates catesbeianus collected in Viçosa,

Minas Gerais State, Brazil (N = 113), presented as mean ± one standard deviation. Nf = number of frogs; Np = number of prey measured.

SVL Individual prey length Individual prey volume Adult males 71.45‑175.75 2.35‑97.60 1.69‑17193.80

(Nf = 46; Np = 85) (136.65 ± 19.79) (30.11 ± 25.65) (1810.94 ± 3532.26) Adult females 88.65‑187.25 1.25‑83.40 0.32‑26625.37

(Nf = 33; Np = 65) (144.27 ± 20.05) (24.42 ± 19.95) (1696.81 ± 3979.04) Juveniles 35.90‑99.80 0.80‑30.85 0.49‑1917.49

(Nf = 34; Np = 68) (59.37 ± 13.71) (9.58 ± 7.19) (103.55 ± 278.29) Size class 1 35.90‑49.60 2.05‑30.85 0.49‑1125.66

(Nf = 11; Np = 24) (44.70 ± 4.71) (8.42 ± 7.51) (94.01 ± 247.69) Size class 2 51.50‑69.30 0.80‑28.95 0.68‑1917.49

(Nf = 16; Np = 36) (61.36 ± 6.09) (10.13 ± 7.05) (118.54 ± 325.67) Size class 3 70.70‑88.65 2.80‑72.95 1.62‑1627.78

(Nf = 9; Np = 34) (76.98 ± 6.71) (14.48 ± 12.73) (198.72 ± 390.87) Size class 4 99.65‑108.60 26.05‑97.60 397.53‑2538.67

(Nf = 4; Np = 2) (103.10 ± 4.27) (61.83 ± 50.59) (1468.10 ± 1514.01) Size class 5 111.05‑129.90 8.10‑34‑35 14.08‑2116.24

(Nf = 11; Np = 10) (118.85 ± 6.12) (17.40 ± 9.10) (555.05 ± 714.34) Size class 6 130.25‑149.90 1.55‑96.90 0.52‑17193.80

(Nf = 44; Np = 68) (141.22 ± 5.07) (29.48 ± 21.68) (2461.97 ± 4160.17) Size class 7 152.20‑187.25 1.25‑95.60 0.32‑26625.37

(Nf = 18; Np = 44) (162.34 ± 10.28) (32.68 ± 28.06) (1865.52 ± 4263.98)

abundance of prey in Bullfrog stomachs (NF = 100 The average prey length and volume, as well as the x total number of individuals of the prey category t length and volume of the largest prey items obtained divided by the total number of all individuals con- from each stomach were compared to the SVL with sumed), and relative volumetric abundance of prey the Pearson Correlation test. Only the specimens that in Bullfrog stomachs (VF = 100 x total volume of had at least one prey that could be measured were individuals of the prey category t divided by the total considered in this analysis. volume of all prey categories in all stomachs) were calculated for each prey category. The diet overlap among the three Bullfrog groups Results (adult males, adult females and juveniles) was calcu- lated using the index of Pianka (1973). Diet overlap Bullfrog body size (SVL) varied from 35.9 to between predator group j and group k was calculated 187.2 mm (113.9 ± 42.6 mm), and was highly corre- as lated to JW (rs = 0.982; p < 0.05). Adult males and fe- males had no significant differences in size, but were 2 2 0.5 Ojk = Okj = [S (pij x pik)] / (S pij x pik ) significantly larger than juveniles (F = 232.4; df = 2; P < 0.05; Q[0.05, 110] = 3.364; Table 1). Where pij and pik are the relative abundance (NF) A total of 129 specimens of L. catesbeianus were of the category i used as prey by the groups j and k collected and 11 of them (8.5%) had empty stom- paired in each treatment, and Ojk = Okj means that the achs. Plant remains were found in the stomachs of effect of group j on group k is equal to the effect of 81 individuals (62.8%), of which five (3.9%) proved group k on group j. to have nothing but plant remains present. After the An analysis of variance (ANOVA) was used to ex- exclusion of those individuals with empty stomachs amine differences between Bullfrog groups in terms or with stomachs containing only plant remains, in- of the number of ingested prey items and SVL. Prey formation on diet composition was obtained from 113 length and volume were also compared among adult specimens (46 adult males, 33 adult females and 34 males, females and juveniles by using ANOVA. If juveniles). Of the 129 collected frogs, 107 (82.9%) one of these differences was significant, a Tukey’s had at least one identifiable prey, and the other six test was used to make a side-by-side comparison of specimens (4.6%) had only arthropod fragments or the corresponding variable among Bullfrog groups. non-identifiable remains in their stomachs. Silva, E. T. et al. 289

Table 2. Diet composition of Lithobates catesbeianus in Viçosa, Minas Gerais State, Brazil (N = 113); NF = numeric frequency; FO = fre- quency of occurrence; VF = volumetric frequency.

Adult males (N = 46) Adult females (N = 33) Juveniles (N = 34) Prey categories NF FO VF NF FO VF NF FO VF Annelida hirudinea 1.57 2.17 <0.01 1.11 3.03 0.01 — — — oligochaeta 0.79 2.17 0.20 — — — 2.17 5.88 3.03 Mollusca gastropoda 3.94 8.70 0.85 6.67 15.15 3.51 2.17 5.88 13.44 Pulmonata 2.36 4.35 0.48 3.33 9.09 2.51 1.09 2.94 12.71 Prosobranchia 0.79 2.17 0.34 1.11 3.03 0.91 1.09 2.94 0.73 Arthropoda Diplopoda 16.54 23.91 26.72 12.22 21.21 11.38 3.26 8.82 2.42 Juliformes 13.39 21.74 26.58 11.11 21.21 10.98 1.09 2.94 <0.01 Polydesmida 3.15 2.17 0.15 1.11 3.03 0.40 2.17 5.88 2.42 arachnida 11.02 23.91 0.90 13.33 24.24 2.99 10.87 26.47 2.66 araneae 10.24 21.74 0.80 13.33 24.24 2.99 9.78 23.53 1.94 acarina — — — — — — 1.09 2.94 0.73 opiliones 0.79 2.17 0.10 — — — — — — Crustacea 0.79 2.17 4.18 1.11 3.03 <0.01 — — — Decapoda 0.79 2.17 4.18 — — — — — — ostracoda — — — 1.11 3.03 <0.01 — — — insecta 43.31 67.39 11.49 46.67 51.52 21.51 77.17 79.41 60.53 ephemeroptera (adults) — — — 3.33 3.03 0.28 — — — ephemeroptera (naiads) — — — — — — 1.09 2.94 <0.01 isoptera 0.79 2.17 <0.01 1.11 3.03 <0.01 — — — odonata (adults) 4.72 10.87 1.55 4.44 9.09 0.21 7.61 20.59 18.77 odonata (naiads) 0.79 2.17 0.22 1.11 3.03 0.06 1.09 2.94 4.12 orthoptera 0.79 2.17 0.29 1.11 3.03 0.17 10.87 17.65 19.85 hemiptera 8.66 13.04 2.16 13.33 24.24 16.32 10.87 20.59 4.84 Coleoptera 7.87 21.74 3.62 6.67 12.12 1.53 6.52 17.65 2.54 hymenoptera 7.09 17.39 0.39 10.00 18.18 0.96 30.43 47.06 8.35 formicidae 6.30 15.22 0.33 7.78 12.12 0.33 26.09 44.12 4.00 others 0.79 2.17 0.06 2.22 6.06 0.63 4.30 11.76 5.14 lepidoptera (adults) 2.36 4.35 1.24 2.22 6.06 1.41 1.09 2.94 <0.01 lepidoptera (larvae) 8.66 17.39 3.73 3.33 9.09 0.85 — — — Diptera 0.79 2.17 0.05 — — — 7.61 14.71 2.06 Chordata amphibia 22.05 34.78 50.97 18.89 36.36 58.09 4.35 11.76 10.65 anura 22.05 34.78 50.97 18.89 36.36 58.09 4.35 11.76 10.65 tadpoles 3.15 6.52 4.51 4.44 12.12 13.76 1.09 2.94 0.73 Post‑metamorphic 18.90 28.26 46.46 14.44 27.27 44.34 3.26 8.82 9.93 Terrestrial prey 59.06 86.96 39.14 58.89 63.64 20.98 82.61 79.41 69.37 Aquatic prey 18.90 28.26 11.38 25.56 39.39 32.43 13.04 26.47 13.44 Amphibious prey 21.26 30.43 46.55 15.56 33.33 44.34 4.35 3.26 9.93 Arthropod remains — 21.74 1.44 — 21.21 1.67 — 20.59 3.03 Undetermined remains — 15.22 1.55 — 9.09 0.59 — 11.76 0.97

A total of 309 prey items were identified (127 in The most frequent prey categories of adult males adult males, 90 in adult females and 92 in juveniles; were post-metamorphic Anura and Diplopoda (Ta- Table 2). The number of prey items per individual was ble 2). Post-metamorphic Anura were found to be similar among Bullfrog groups (range, average ± SD; most common in adult females, along with Araneae, adult males: 1‑9; 2.98 ± 2.20; adult females: 1‑13; Hemiptera, and Diplopoda. For juveniles, Hymenop- 3.13 ± 2.42; juveniles: 1‑10; 2.94 ± 2.06; F = 0.066; tera Formicidae, Orthoptera, Hemiptera, Odonata, df = 2; P > 0.05). and Araneae were the most frequent prey categories. 290 Diet of Lithobates catesbeianus

Table 3. Correlation (Pearson) among SVL of Lithobates cates- F = 6.715; df = 2; P < 0,05; Q (0.05, 215) = 3.310]. The beianus (N = 87) and prey length and volume, from Viçosa, length and volume of the largest prey items, as well as Minas Gerais State, Brazil. the average prey length and volume, were positively correlated to the SVL of frogs (Table 3). Because of Variables rs T p Mean prey length 0.4805 5.0519 0.0001 the small number of individuals collected (N = 4), Largest prey length 0.4916 5.2049 0.0001 size class 4 was not considered in the analyses. In- Mean prey volume 0.3348 3.2752 0.0005 sects were the most abundant prey group in the diet of Largest prey volume 0.3331 3.2566 0.0006 all size classes, although a progressive reduction was observed in consumption of insects from class 3 to class 7 (Figure 1). On the other hand, the presence of A total of 49 anurans were consumed by Bullfrogs, Diplopoda and Amphibia tended to increase ontoge- being 40 post-metamorphic and nine tadpoles, one netically, and Amphibia was not consumed by class 1. of them a L. catesbeianus tadpole. The post-meta- morphic anurans identified were the native hylids Dendropsophus minutus (N = 2), D. elegans (N = 3), Discussion crospedospilus (N = 1), S. eurydice (N = 9), Hypsiboas faber (N = 2), and the bufonid Rhinella are generally considered opportunis- pombali (N = 1), besides 22 undetermined individu- tic predators. Their diets often reflect the availability als. The post-metamorphic anurans (N = 15) ranged of prey in their , and they ingest any prey of from 12.7 to 56.1 mm in length (average ± sD: appropriate size (Korschgen and Baskett, 1963; Du- 37.3 ± 11.8 mm) and 216.8 to 17,193.8 mm3 in vol- ellman and Trueb, 1994; Stebbins and Cohen, 1995). ume (6848.4 ± 5793.1 mm3). Compared to other frogs, L. catesbeianus seems to be Terrestrial prey items were the most frequent an extremely opportunistic predator (Korschgen and in number, especially in the stomachs of juveniles Moyle, 1955; Cohen and Howard, 1958; Korschgen (82.6%; Table 2). Regarding total volume, terrestrial and Baskett, 1963; Brooks Jr., 1964; Corse and Met- prey were most frequent only for juveniles (69.37%), ter, 1980; Bury and Whelan, 1984; Silva et al., 2007; while amphibious prey were most frequent in adult Camargo Filho et al., 2008) and the results herein Bullfrogs (males: 46.5%; females: 44.3% of the diet). on the diet of invasive populations corroborate this Diet overlap was higher between adult males and fe- suggestion. males (0.944; 83.3%) than between adults and juve- Insects were the most diverse and abundant prey niles (adult females 0.671, 61.6%; adult males 0.554, group in the diet of L. catesbeianus, as observed by 55.9%). Adults of both sexes showed no significant Korschgen and Moyle (1955), Korschgen and Bas- differences in relation to prey size, but consumed kett (1963) and Corse and Metter (1980), probably larger prey than juveniles [prey length: F = 21.01; because of their high availability, although this was df = 2; P < 0,05; Q (0.05, 215) = 3.310; prey volume: not evaluated in this study. Amphibia and Diplopoda also represented a considerable portion of total prey volume (53.9% and 19.9%, respectively). Our results were similar to those obtained in a similar study by Boelter and Cechin (2007) in areas where the spe- cies was introduced in the state of Rio Grande do Sul, southern Brazil. Plant ingestion by Bullfrogs was considered ac- cidental by several authors (Korschgen and Moyle, 1955; Korschgen and Baskett, 1963; Corse and Met- ter, 1980; Wu et al., 2005). In area 1, 30.2% of the frogs with plant in their stomachs (N = 43) showed remains of Salvinia sp., a common aquatic pterido- phyte in local dams. Potential prey of L. catesbeianus, such as spiders, moths, and treefrogs were observed Figure 1. Diet composition of the size classes of Lithobates cates- beianus collected in Viçosa, Minas Gerais State, Brazil (N = 109) on this floating vegetation in several occasions during according to the numeric frequency (NF). The group “Others” fieldwork, and were also found in Bullfrog stomachs refers to Annelida, Gastropoda and Crustacea. together with Salvinia sp. remains. Thus, Salvinia sp. Silva, E. T. et al. 291 may have been ingested accidentally when the Bull- palatability as a factor that affects prey selection by frogs were ingesting arthropods and anurans. Korsch- amphibians, our results and those of other authors gen and Baskett (1963) suggest that the presence of (Korschgen and Baskett, 1963; Brooks Jr., 1964), plant remains in Bullfrog stomachs is a result of its suggest that L. catesbeianus can tolerate these repug- proximity to the prey at the moment of capture or due nant compounds. to its movement on the water’s surface. The occurrence of a bufonid (Rhinella pombali) Ontogenetic shifts in frog diet are often related among the stomach contents, as well as the observa- to size differences among predators (Duellman and tion of a Bullfrog preying upon a toad of the same Trueb, 1994). Larger individuals can continue feeding species (Reis et al., 2007), also indicates tolerance on the same types of prey used by the smaller frogs, to bufonid chemical defenses. Korschgen and Moyle while also including larger prey in their diets (Steb- (1955) also found a bufonid in a stomach of L. cates- bins and Cohen, 1995) which can explain the differ- beianus, and questioned if the occurrence of only one ences in diet composition observed between adults individual would not indicate a low acceptance of this and juveniles of L. catesbeianus. This trend became amphibian group as prey. Bury and Whelan (1984), evident also by the positive correlation between prey in their review on L. catesbeianus ecology, cite in- size (length and volume) and frog SVL, as observed stances of predation on bufonids and the immobiliza- for Leptodactylus ocellatus by França et al. (2004) tion effect on Bullfrogs by the secretion of bufonid and Maneyro et al. (2004). Consumption of Diptera, poison glands. Bufonids were also reported as prey Odonata, Orthoptera, and Hymenoptera was more for other frog species, including Rana pretiosa (Pearl frequent in juveniles, and the consumption of Hyme- and Hayes, 2002), Leptodactylus ocellatus (Gallardo, noptera Formicidae stands out as the greatest differ- 1958; França et al., 2004), L. labyrinthicus (Cardoso ence between adult and juvenile frogs. The opposite and Sazima, 1977) and Ceratophrys ornata (Braun was true for the consumption of post-metamorphic et al., 1980). anurans and diplopods, which were more frequent Terrestrial prey items were numerically dominant in the adult diet, a factor most likely due to the fact and most frequent in the stomachs, although other that most of these prey items are too large for juvenile studies have indicated a high importance of aquatic frogs. This may also contibute to the low diet overlap prey items in the diet of L. catesbeianus (Werner et al., between adults and juveniles. Our comparison of the 1995; Hirai, 2004; Wu et al., 2005). The low frequen- diet among size classes (Figure 1) may reflect differ- cy of prey with amphibious habits in the stomachs of ences between adults and juveniles, since classes 1 juveniles is a consequence of the low consumption of and 2 were composed only of juveniles, class 3 was post-metamorphic anurans, probably because of size composed mostly of juveniles, and classes 5‑7 were constraints (see results). On the other hand, the con- composed only of adults. Similar results were ob- sumption of anurans also explains the greater volu- tained by Govindarajulu et al. (2006) for Bullfrogs metric contribution of amphibious prey in adult Bull- introduced in Canada. frogs. The results of Hirai (2004), in Kyoto, Japan, A high similarity and overlap between the diet could have been influenced by the high abundance of male and female Bullfrogs were also reported by of the crayfish (Procambarus clarkii) detected in the Brooks Jr. (1964), Werner et al. (1995), and Boelter habitat studied, which resulted in a high frequency of and Cechin (2007). Spatial distribution of Bullfrogs this crustacean in the diet of L. catesbeianus. in sampling sites may explain this similarity. Most Our study confirms that the American Bullfrog adult males and females were collected at the same has a generalist feeding habit, which together with place, on the edge of water bodies, which is related its capacity to occupy modified environments (Bar- to the reproductive behavior of this species (see Bury rasso et al., 2009), appears to favor the establishment and Whelan, 1984). Similarity in body size and jaw of invasive populations in Brazil. The use of native width may also contribute to the higher diet overlap anurans as a food source was high among adult Bull- between adult males and females than between adults frogs, which suggests the possible occurrence of a and juveniles. negative effect on native anuran populations in sites A peculiarity in this study was the high consump- occupied by this invasive species. Observations in tion of diplopods. These have chemical de- the western United States suggest a direct correlation fenses, such as phenolic and quinone compounds, between Bullfrog dispersal and decrease of native produced by repugnant glands (Ruppert and Barnes, ranids, and predation is cited as a possible cause of 1996). Although Stebbins and Cohen (1995) report this decline (Moyle, 1973; Hammerson, 1982; Hayes 292 Diet of Lithobates catesbeianus and Jennings, 1986; Kats and Ferrer, 2003; Pearl maturidade sexual das rãs foram determinados. As et al., 2004). Nevertheless, stating that predation is presas foram classificadas em três grupos de acordo the cause of population decline based only on stom- com seu principal habitat: aquáticas, terrestres e ach content analysis, without data on the size of prey anfíbias. As categorias de presas mais freqüentes na populations, may not be safe (Hayes and Jennings, dieta foram Anura (pós-metamórficos), Diplopoda, 1986). In fact, the high presence of native anurans in Hemiptera, Hymenoptera Formicidae e Araneae. the diet of the Bullfrog, as reported here, can also in- As dietas de adultos de ambos os sexos foram dicate some level of co-existence, since the invasion semelhantes, diferindo em maior grau da dieta de in Viçosa has lasted about 30 years. Unfortunately, indivíduos jovens. Presas terrestres foram mais data on the size of native anuran populations are lack- abundantes em número e ocorrência e, para os adultos, ing for the sites studied, as well as on their fluctuation presas de hábitos anfíbios foram mais representativas since the Bullfrog introduction, which does not allow em volume. A dieta variou de acordo com o tamanho us to draw further conclusions. dos indivíduos, havendo relação positiva entre o On the other hand, competition for food resources tamanho das rãs e o tamanho das presas ingeridas, e among adults of invasive Bullfrogs and ecologically maior consumo de anfíbios anuros pelas rãs maiores. similar species is also suggested as a possible negative Estes resultados confirmam que a rã-touro possui interaction (Hayes and Jennings, 1986; Werner et al. hábito alimentar generalista, podendo ter algum 1995; Barrasso et al., 2009). Some species of South efeito negativo importante sobre as comunidades de American leptodactylids, such as Leptodactylus ocel- anfíbios nativos nos locais ocupados por esta espécie. latus, which is common in lentic water bodies and has a generalist diet, could be affected by this com- petition, as proposed by Barrasso et al. (2009). Local Acknowledgements extinction of leptodactylids was reported by Batista (2002) in a locality of the Brazilian State of Goiás, We are grateful to the colleagues of the Museu de Zoologia following Bullfrog introduction. In Viçosa, Lima and João Moojen for their assistance in field work; to Patrícia S. Santos for her comments and suggestions during the development Verani (1988) estimated a population around 200 in- of this work; to Jorge A. Dergam and Rubem A. Boelter for their dividuals of L. ocellatus at the second site sampled help in obtaining references; to Paulo R. Cecon for his help with in the present study, based on data collected between the statistical analyses; to João Vitor G. Lacerda and Ansley January 1978 and January 1979, before the introduc- Ditmore for the English revision; to Departamento de Biologia Animal for logistic support; to IBAMA for the collection permits tion of Bullfrogs. During the fieldwork of the present (036/05 and 233/06). Marcio Martins and two anonymous study few individuals of this species were observed reviewers also made useful comments on the final version of in that site, and their encounter was not frequent, sug- this paper. ETS would also like to thank the Conselho Nacional gesting a possible population decrease since those de Desenvolvimento Científico e Tecnológico (CNPq) for the times. The possibility of negative interactions with fellowship during part of this study (PIBIC/UFV). Bullfrogs cannot be discarded, although other factors can be involved in this apparent decline. Guix (1990) Literature Cited reported that L. ocellatus was frequent in a locality of São Paulo State, in spite of the presence of invasive Batista, C. G. 2002. Rana catesbeiana (Bullfrog). Effects on Bullfrogs. Thus, the possible relationship between the native anuran community. Herpetological Review, 33:131. spread of the Bullfrog and native amphibian popu- Barrasso, D. A., R. Cajade, S. J. Nenda, G. Baloriani, and R. Herrera. 2009. Introduction of the American Bullfrog lation declines in Brazil is a question that remains Lithobates catesbeianus (Anura: Ranidae) in natural and unclear. modified environments: an increasing conservation problem in Argentina. South American Journal of Herpetology, 4:69‑75. Boelter, R. A. and S. Z. Cechin. 2007. Impacto da dieta de rã- touro (Lithobates catesbeianus – Anura, Ranidae) sobre a Resumo fauna nativa: estudo de caso na região de Agudo, RS, Brasil. Natureza & Conservação, 5:45‑53. A dieta do anfíbio invasor Lithobates catesbeianus Braun, P. C., C. A. S. Braun, and M. D. S. Pineda. 1980. (rã-touro americana) foi examinada em quatro Observações sobre o comportamento alimentar de locais em Viçosa (20°45’S e 42°51’W), Estado de Ceratophrys ornata (Bell, 1843) em cativeiro. Revista Brasileira de Biologia, 40:401‑403. Minas Gerais, Brasil, através da análise do conteúdo Brooks Jr., G. R. 1964. An analysis of the food habits of the estomacal de 113 exemplares coletados entre agosto Bullfrog, Rana catesbeiana, by body size, sex, month, and de 2005 e março de 2007. Os efeitos do tamanho e habitat. The Virginia Journal of Science, 15:173‑186. Silva, E. T. et al. 293

Bury, R. B. and J. A. Whelan. 1984. Ecology and management Hayes, M. P. and M. R. Jennings. 1986. Decline of ranid of the Bullfrog. U. S. Fish and Wildlife Service Resource frog species in Western North America: are Bullfrogs Publication, 155:1‑24. (Rana catesbeiana) responsible? Journal of Herpetology, Camargo Filho, C. B., H. C. Costa, E. T. Silva, O. P. R. Filho, 20:490‑409. and R. N. Feio. 2008. Lithobates catesbeianus (American Hirai, T. 2004. Diet composition of introduced Bullfrog, Rana Bullfrog). Prey. Herpetological Review, 39:338. catesbeiana, in the Mizorogaike Pond of Kyoto, Japan. Cardoso, A. J. and I. Sazima. 1977. Batracofagia na fase Ecological Research, 19:375‑380. adulta e larvária da rã pimenta, Leptodactylus labyrinthicus Kaefer, I. L., R. A. Boelter, and S. Z. Cechin. 2007. Reproductive (Spix, 1824) – Anura, Leptodactylidae. Ciência e Cultura, biology of the invasive Bullfrog Lithobates catesbeianus in 29:1130‑1132. southern Brazil. Annales Zoologici Fennici, 44:435‑444. Coelho, D. J. S., A. L. Souza, and C. M. L. Oliveira. 2005. Kats, L. B. and R. P. Ferrer. 2003. Alien predators and amphibian Levantamento da cobertura florestal natural da microrregião declines: review of two decades of science and the transition de Viçosa, MG, utilizando-se imagens de Landsat 5. Revista to conservation. Diversity and Distributions, 9:99‑110. Árvore, 29:17‑24. Kiesecker, J. M., A. R. Blaustein, and C. L. Miller. 2001. Cohen, N. W. and W. E. Howard. 1958. Bullfrog food and growth Potential mechanisms underlying the displacement of at the San Joaquin experimental range. Copeia, 1958:223‑225. native red-legged frogs by introduced Bullfrogs. Ecology, Corse, W. A. and D. E. Metter. 1980. Economics, Adult Feeding 82:1964‑1970. and Larval Growth of Rana catesbeiana on a Fish Hatchery. Korschgen, L. J. and D. L. Moyle. 1955. Food habits of the Journal of Herpetology, 14:231‑238. Bullfrog in Central Missouri farm ponds. The American Costa, C. S. L., S. L. Lima, D. R. Andrade, and C. A. Agostinho. Midland Naturalist, 54:332‑341. 1998. Caracterização Morfológica dos Estágios de Korschgen, L. J. and T. S. Baskett. 1963. Foods of impoundment Desenvolvimento do Aparelho Reprodutor Feminino da Rã- and stream dwelling Bullfrogs in Missouri. Herpetologica, touro, Rana catesbeiana, no Sistema Anfigranja de Criação 19:89‑99. Intensiva. Revista Brasileira de Zootecnia, 27:642‑650. Kupferberg, S. J. 1997. Bullfrog (Rana catesbeiana) Invasion Duellman, W. E. and L. Trueb. 1994. Biology of Amphibians. of a California river: the role of competition. Ecology, The Johns Hopkins University Press, Baltimore, 670 pp. 78:1736‑1751. França, L. F., K. G. Facure, and A. A. Giaretta. 2004. Trophic Lawler, S. P., D. Dritz, T. Strange, and M. Holyoak. 1999. and spatial niches of two large-sized species of Leptodactylus Effects of introduced mosquitofish and Bullfrog on the (Anura) in Southeastern Brazil. Studies on Neotropical Fauna threatened California red-legged frog. Conservation Biology, and Environment, 39:243‑248. 13:613‑622. Fontanello, D. 1994. Histórico da Ranicultura no Brasil. pp. 3‑6. Lima, S. L. 1994. Universidade Federal de Viçosa; pp. 95‑104. In: Lima, S. L, M. R. C. Figueiredo and O. M. Moura In: Lima, S. L, M. R. C. Figueiredo and O. M. Moura (Eds.), Diagnóstico da Ranicultura: problemas, propostas de (Eds.), Diagnóstico da Ranicultura: problemas, propostas de soluções e pesquisas prioritárias. Associação Brasileira de soluções e pesquisas prioritárias. Associação Brasileira de Estudos Técnicos em Ranicultura, Viçosa. Estudos Técnicos em Ranicultura, Viçosa. Gallardo, J. M. 1958. Observaciones sobre el comportamiento Lima, S. L. and J. R. Verani. 1988. Dinâmica populacional da de algunos anfibios argentinos. Ciencia e Investigación, rã-manteiga, Leptodactylus ocellatus (Linnaeus, 1758) em 14:291‑302. Viçosa, MG (Anura, Leptodactylidae). Revista Brasileira de Garner, T. W. J., M. W. Perkins, P. Govindarajulu, D. Seglie, Biologia, 48:113‑118. S. Walker, A. A. Cunningham, and M. C. Fisher. 2006. Lima, S. L., C. L. S. Costa, C. A. Agostinho, D. R. Andrade, The emerging amphibian pathogen Batrachochytrium and H. P. P. Filho. 1998. Estimativa do Tamanho da Primeira dendrobatidis globally infects introduced populations of Maturação sexual da Rã-touro, Rana catesbeiana, no Sistema the North American Bullfrog, Rana catesbeiana. Biology Anfigranja de Criação Intensiva. Revista Brasileira de Letters, 2006:455‑459. Zootecnia, 27:416‑420. Giovanelli, J. G. R., C. F. B. Haddad, and J. Alexandrino. 2008. Magnusson, W. E., A. P. Lima, W. A. Silva, and M. C. Araújo. Predicting the potential distribution of the alien invasive 2003. Use of geometric forms to estimate volume of American bullfrog (Lithobates catesbeianus) in Brazil. invertebrates in ecological studies of dietary overlap. Copeia, Biological Invasions, 10:595-590. 2003:13‑19. Guix, J. C. 1990. Introdução e colonização de Rana catesbeiana Maneyro, R., D. E. Naya, A. Canavero, and A. Camargo. 2004. Shaw, 1802 em um pequeno vale no município de Suzano Diet of the South American frog Leptodactylus ocellatus (SP), sudeste do Brasil. Grupo de Estudos Ecológicos Série (Anura, Leptodactylidae) in Uruguay. Iheringia Série Documentos, 2:32‑34. Zoologia, 94:57‑61. Govindarajulu, P., W. S. Price, and B. R. Anholt. 2006. Martins, M. B., M. Di-Bernardo, G. Vinciprova, and J. Introduced Bullfrogs (Rana catesbeiana) in Western Canada: Measey. 2002. Geographic distribution: Rana catesbeiana. Has Their Ecology Diverged? Journal of Herpetology, Herpetological Review, 33:319. 40:249‑260. Moyle, P. B. 1973. Effects of introduced Bullfrogs, Rana Hammerson, G. A. 1982. Bullfrog eliminating leopard frogs in catesbeiana, on native frogs of the San Joaquin valley, Colorado? Herpetological Review, 13:115‑116. California. Copeia, 1973:18‑22. Hanselmann, R., A. Rodriguez, M. Lampo, L. F. Ramos, A. A. Pearl, C. A. and M. C. Hayes. 2002. Predation by Oregon Spotted Aguirre, A. M. Kilpatrick, J. P. Rodriguez, and P. Daszak. frogs (Rana pretiosa) on Western Toads (Bufo boreas) in 2004. Presence of an emerging pathogen of amphibians Oregon. The American Midland Naturalist, 147:145‑152. in introduced Bullfrogs Rana catesbeiana in Venezuela. Pearl, C. A., M. J. Adams, R. B. Bury, and B. McCreary. Biological Conservation, 120:115‑119. 2004. Asymmetrical Effects of Introduced Bullfrogs (Rana 294 Diet of Lithobates catesbeianus

catesbeiana) on Native Ranid Frogs in Oregon. Copeia, Silva, E. T., H. C. Costa, and R. N. Feio. 2007. Rana catesbeiana 2004:11‑20. (Bullfrog). Prey. Herpetological Review, 38:443. Pianka, E. R. 1973. The Structure of Lizard Communities. Annual Vizotto, L. D. 1984. Ranicultura. Ciência e Cultura, 36:42‑45. Review of Ecology and Sistematics, 1973:53‑74. Werner, E. E., G. A. Wellborn, and M. A. McPeek. 1995. Diet Reis, E. P., E. T. Silva, R. N. Feio, and O. P. R. Filho. 2007. Composition in Postmetamorfic Bullfrogs and Green Frogs: Chaunus pombali (Pombali’s toad) Predation. Herpetological Implications for Interspecific Predation and Competition. Review, 38:321. Journal of Herpetology, 29:600‑607. Ribon, R., J. E. Simon, and G. T. Mattos. 2003. Bird extinctions in Wu, Z., Y. Li, Y. Wang, and M. J. Adams. 2005. Diet of introduced Atlantic Forest fragments of the Viçosa region, Southeastern Bullfrogs (Rana catesbeiana): predation on and diet overlap Brazil. Conservation Biology, 17:1827‑1839. with native frogs on Daishan Island, China. Journal of Ruppert, E. E. and R. D. Barnes. 1996. Zoologia dos Herpetology, 39:668‑674. Invertebrados. Roca, São Paulo, 1029 pp. Stebbins, R. C. and N. W. Cohen. 1995. A natural history of Submitted 21 July 2009 amphibians. Princeton University Press, New Jersey, 316 pp. Accepted 12 November 2009