Geographic Variation in the Diet of the Cauca Poison Frog Andinobates Bombetes (Anura: Dendrobatidae) in the Andes of Colombia

Herpetology Notes, volume 7: 559-564 (2014) (published online on 3 October 2014)

Geographic variation in the diet of the Cauca Poison Frog Andinobates bombetes (Anura: Dendrobatidae) in the Andes of Colombia

Diego A. Gómez-Hoyos1, Margarita M. López-García2, Carlos A. Soto-Garzón3, Diana M. Méndez-Rojas4, Ted R. Kahn5 and Julián A. Velasco6,*

Abstract. We describe the diet of Andinobates bombetes (Myers and Daly, 1980) based on stomach contents taken from sixty- nine individuals from three localities in the central northwestern Colombian Andes. 1471 individual prey items were found in the stomach contents from 16 taxonomic groups: arthropods (14), crustaceans (1) and mollusks (1). Taxonomic groups of prey were identiﬁed to the lowest taxonomic rank possible. We used the Index of Relative Importance (IRI) and found that Formicidae (ants), and the subclass Acari (mites) dominated the prey taken in our samples. There was not a signiﬁcant correlation between snout-vent length (SVL) and prey volume sampled. The dietary ant ratio was different between localities, whereas that of mites was consistent across the three localities studied. These results indicate that A. bombetes is a dietary ant and mite specialist, similar to other dendrobatid frogs.

Keywords: Trophic ecology, Dendrobatidae, poison frog, ants, mites, diet

Introduction skins of poison frogs (Saporito et al., 2004, 2007a, 2007b, 2009; Smith and Jones, 2004). Particularly, dendrobatid Poison frogs (Dendrobatidae) have received poison frogs have developed an over-expressed system considerable attention in the scientific community of alkaloid accumulation that facilitates the sequestration because of their brilliantly colored aposematic patterns, of alkaloids directly from their diet, such as ants, mites the presence of alkaloids in the skin, and their complex and millipedes (Daly, 1998, 2000; Saporito et al., 2004, reproductive behaviors (Biavati et al., 2004). Several 2007a, 2007b, 2009; Takada et al., 2005; Savitzky et al., studies have shown that a number of arthropod dietary 2012). sources explain the origin of alkaloids sequestered in the Studies concerning diet and prey consumption are important in understanding the relationship between diet and alkaloid sequestration in poison frogs, as well as competition with congenerics. The variation in alkaloid profiles in poison frogs is related to differences 1 Wildlife Conservation Society, Colombia Program, Cali, Valle in availability of dietary arthropods consumed among del Cauca, Colombia. E-mail: [email protected] populations of these frogs (Saporito et al., 2007a). 2 Grupo de Investigación Insectos de Colombia, Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Studies have found geographic variation in alkaloid Bogotá, Colombia. profiles related to diet in all species studied (Saporito 3 Fundación Huella Verde, Armenia, Colombia. et al., 2006, 2007b; Mebs et al., 2008). Few studies 4 Red de Ecoetología, Instituto de Ecología A.C, Xalapa, have explored diet patterns across the geographic range Veracruz, México. of a single poison frog species, and how these patterns 5 Integrated Taxonomic Information System, Data Development would correspond to differences in dietary arthropod contractor, Smithsonian Institution, National Museum of communities and subsequent alkaloid sequestration in Natural History, Washington, D.C. 6 Laboratorio de Análisis Espaciales, Instituto de Biología, poison frogs. Universidad Nacional Autónoma de México. México. D.F. Herein we provide insights to the diet of A. bombetes *Corresponding author e-mail: [email protected] from three localities within its range. Andinobates 560 Diego A. Gómez-Hoyos et al.

Figure 1. Study area in the Central and Western Andes of Colombia. Solid circle: Cañón del Río Barbas; Square: La Samaritana Farm; Triangle: Reserva Forestal de Yotoco.

bombetes is a small-bodied frog and occurs in Central All frogs were captured by hand and standard snout- and Western Andes of Colombia (Suárez-Mayorga, vent (SVL) in millimeters (mm) was measured, except 2004; Marín-Gómez and Gómez-Hoyos, 2011). We for RFY locality. quantified the diet of A. bombetes at three localities in Stomach contents were sampled immediately after the Central and Western Andes of Colombia (Fig. 1), capture (from 9:00 to 16:00 h) by stomach flushing and then compared our findings of local dietary items using a small blunt needle (Popper® 3.048 mm gauge). among the three study localities. This procedure is effective to flush stomach contents without producing injury (Solé et al., 2005). Stomach Materials and Methods contents were preserved in 70% ethanol. We identified each prey item to the lowest taxonomic rank possible. We used visual encounter surveys (VES) (Heyer We measured length and width in mm of each intact et al., 1994) during July and October 2009 to collect prey item. These measurements were used to calculate A. bombetes individuals for gut content analyses. prey volume using the spheroid volumetric equation as Sampling occurred at three sites in the Central and follows: Western Andes of Colombia: 1) La Samaritana farm (SF), Salento municipality, Quindío; secondary forest; V = 4/3π * (length/2) * (width/2)2 1626 m (4°34’N, 75°38’W); 2) Cañón del Río Barbas We quantified the number of each prey item in the diet (CRB), Filandia municipality, Quindío; mature forest; using the Index of Relative Importance (IRI) described 1998 m (4°42’N, 75°38’W); and 3) Reserva Forestal de by Pinkas et al. (1971) with the following equation: Yotoco (RFY), Yotoco municipality, Valle del Cauca; secondary forest; 1600 m (3°53’N, 76°38’W) (Fig. 1). IRIt =[%Ot * (%Nt + %Vt)] Geographic variation in the diet of the Cauca Poison Frog in the Andes of Colombia 561

where %Ot is the occurrence percentage (i.e. the homogeneity of variances, we applied a Spearman rank number of stomachs containing each prey category correlation to examine the relationships between SVL deﬁned as t), %Nt is the percentage of the number of t and prey volume consumed in each locality. We tested items in all stomachs, and %Vt is the percentage of the for differences in proportions of prey items between volume of t items in all stomachs examined. localities using a Kruskal-Wallis test. All statistical We also calculated the trophic niche breadth using analyses were conducted with Deducer package on R Levins´ standardized formula (Hurlbert, 1978): language v. 3.0.2 (R Core Team, 2013).

2 BA=[(1/Σpi )-1]/n-1 Results where pi is the proportion in volume of each prey item with respect to the total volume of prey in each A total of 69 stomach contents were analyzed from location; n is the number of prey items in the diet of the live A. bombetes: 18 from CRB (6 males, mean frogs. The index produces values ranging from zero to SVL=19.24±0.89; 6 females, mean SVL=18.67±1.51; 6 immatures), 27 from FS (11 males, mean 1; BA=1 means that all prey categories were consumed in equal proportions and that selection of prey was not SVL=17.44±0.83; 10 females, mean SVL=19.24±0.89; significant, whereas a value near zero indicates that 6 immatures) and 24 from RFY. We identified 16 one or few prey were consumed in large quantities prey taxa to the subphylum, class, subclass, order and indicating intentional prey selection (Valderrama- family level (Table 1). We identified 1471 individual Vernaza et al., 2009). After testing for normality and prey items with a mean of 4.43 (±1.90) prey categories

Table 1. Diet of Andinobates bombetes from three localities in the Central and Western Andes of Colombia. N: number of prey; %F: percentage of frequency of each prey item; %V: percentage of volume of prey; IRI: Index of Relative Importance.

Cañón Río Barbas Samaritana Farm Reserva Forestal Yotoco Dietary Sample N %F %V IRI N %F %V IRI N %F %V IRI Classification Arthropoda Arachnida Araneae 2 11.11 4.8 60.26 4 14.81 0.41 16.08 8 25 0.87 57.53 Acari 164 88.89 46.58 8695.89 283 88.89 13.55 5460.59 286 100 29.5 8057.11 Pseudoscorpionida 1 5.56 0.47 4.36 1 3.7 0.13 1.09 5 20.83 1.44 48.51 Insecta Coleoptera 12 38.89 9.38 510.7 23 44.44 6.49 461.38 30 66.67 12.48 1189.39 (includes larvae)

Collembola 18 33.33 8.84 482.08 12 22.22 0.17 48.89 61 50 3.87 738.13

Diptera 12 50 3.84 379.44 17 40.74 0.69 145.37 42 62.5 4.51 750.87 (includes larvae) Hemiptera 21 55.56 0.84 411.04 5 14.81 3.31 61.53 9 29.17 0.93 74.01 Hymenoptera 82 77.78 19.71 3525.89 208 77.78 71.92 8331.29 84 58.33 41.5 3295.63 (Formicidae) Hymenoptera 1 5.56 0.06 2.07 22 55.56 2.11 324.08 6 20.83 1.79 59.55 (non Formicidae) Orthoptera 1 5.56 1.95 12.55 0 0 0 0 0 0 0 0 Thysanoptera 3 16.67 0.54 24.55 12 33.33 0.45 82.85 17 33.33 1.55 152.98 Psocoptera 0 0 0 0 3 7.41 0.63 8.46 1 4.17 0.1 1.15 Diplura 0 0 0 0 0 0 0 0 1 4.17 0.16 1.42 Malacostraca Isopoda 0 0 0 0 0 0 0 0 1 4.17 0.01 0.79 Myriapoda 0 0 0 0 0 0 0 0 6 20.83 0.96 42.42 Mollusca Gasteropoda 3 11.11 3 43.8 1 3.7 0.14 1.14 3 4.17 0.32 3.58 562 Diego A. Gómez-Hoyos et al.

Figure 2. Proportion of ants (A) and mites (B) consumed by Andinobates bombetes in three localities in the Central and Western Andes of Colombian Andes (CRB: Cañon Río Barbas; FS: Finca La Samaritana; RFY: Reserva Forestal Yotoco). Solid circle: mean; error bars: 95% conﬁdence interval.

and 21.45 (±18.08) prey items per individual frog. values. Our data suggest that A. bombetes is a dietary ant Formicidae and Acari had the highest IRI scores (Table and mite specialist as many other dendrobatid species 1). Only Acari and Formicidae were consumed in a from several genera (e.g., Ameerega, Andinobates; higher proportion than other prey categories. Pheidole Kahn et al., 2014 in press). Ants and mites are an were consumed most commonly (79.03% of total ants) abundant prey resource in Neotropical the leaf litter and other ant genera were rarely found (Solenopsis, habitat where dendrobatid frogs live (Daly et al., 1999: Carebara, Cyphomyrmex, Strumygenis, Crematogaster, Hölldober and Wilson, 1990; Franklin et al., 2006). Octostruma, Linepithema, Discothyrea, Pyramica, Andinobates bombetes usually is found in microhabitats Brachymyrmex). For example, the formicine genus (e.g., bromeliads, leaf litter, soil substrate) (Myers and Brachymyrmex was represented by only a single record Daly, 1980) where many of the prey items were found. (0.26% of total ants). Niche breadth was close to zero Variation in ant and mite consumed by A. bombetes (CRB=0.1774; FS=0.1631; RFY=0.1624). across the study sites is consistent with findings in other dendrobatid frog species (Valderrama-Vernaza et We found no significant correlation between poison al., 2009; Lieberman, 1986; Daly, 1998; Biavati et al., frog SVL and prey volume (R=0.10, p > 0.05) in 2004; Darst et al., 2005; Saporito et al., 2004, 2007a, the individuals we examined. We found significant 2009; Forti et al., 2011). differences in the dietary percentage of ants consumed between localities (Kruskal-Wallis Chi2 = 6.8, df = 2, Ants and mites are primary sources of dietary alkaloids p = 0.03), with a higher proportion of ants consumed in a number of poison frog species (Daly et al., 1999: at FS than at RFY (Fig. 2, A). The percentage of mites Hölldober and Wilson, 1990; Franklin et al., 2006). consumed for poison frogs did not vary among sites Variation in alkaloid profiles of A. bombetes has been (Kruskal-Wallis Chi2 = 1.72, df = 2, p = 0.42; Fig. 2, observed over broad geographic areas (Myers and Daly, B). 1980) and can be attributed to differences in the kinds of arthropods, and their respective alkaloids, consumed by the frogs. A previous study of A. bombetes conducted at Discussion two sites in Colombia, revealed a total of 22 piperidine The diet of Andinobates bombetes is primarily alkaloids sequestered in the skins of these poison frogs; composed of ants and mites, which occur at a higher little variation occurred in the 15 and 17 compounds percentage than other prey items at our study sites. This evaluated from the two localities (Myers and Daly, result is well-supported by our very small niche breadth 1980). Since the time of that study, the link between diet Geographic variation in the diet of the Cauca Poison Frog in the Andes of Colombia 563 and alkaloid sequestration has been demonstrated (Daly diversity: standard methods for amphibians. Smithsonian et al., 1999; Saporito et al., 2011). However, with the Institution Press. 1–384. aim to elucidate the relationship between toxicity and Hölldober, B., Wilson, E.O. (1990): The Ants. Cambridge, USA, Belknap Press of Harvard University Press. diet in dendrobatid poison frogs further studies will be Hurlbert, S.H. (1978): The measurement of niche overlap and some necessary to characterize the alkaloid profiles of the relatives. Ecology 59: 67-77. varied prey consumed by A. bombetes, principally ants Kahn, T.R., La Marca, E., Lotters, S., Brown, J.L., Twomey, E. and mites. and Amezquita, A. (2014 in press): Aposematic Poison Frogs (Anura; Dendrobatidae) of the Andean Countries Bolivia, Acknowledgments. This study was funded by an internship grant Colombia, Ecuador, Peru and Venezuela. Conservation from the Wildlife Conservation Society for Colombia Program International Tropical Field Guide Series. (ONE WCS grant). The authors are grateful to Jennifer Pramuk Lieberman, S.S. (1986): Ecology of the leaf litter herpetofauna of (Woodland Park Zoo), Catalina Silva (Cali Zoo), and Beatriz a Neotropical rainforest: La Selva, Costa Rica. Acta Zoológica Velásquez (Cali Zoo) for their support during our fieldwork. Mexicana 15: 1-71. A special thanks to Carlos Jaramillo (Director of the Reserva Marín-Gómez, O.H., Gómez-Hoyos, D.A. (2011): Estado Actual Forestal de Yocoto) for logistic support during our fieldwork. de Ranitomeya bombetes (Anura: Dendrobatidae): Plan de We are grateful for the insightful and encouraging review of our manejo y conservación para las poblaciones de Ranitomeya manuscript by Anna J. Phillips (Dept. of Invertebrate Zoology, bombetes Myers y Daly 1980 en Quindío, Colombia. Editorial NMNH), Mason Ryan (University of New Mexico), and an Académica Española. anonymous reviewer. This work was also made possible by a Mebs, D., Pogoda W., Baista A., Ponce M., Köhler G., Kauert graduate scholarship from the Consejo Nacional de Ciencia y G. (2008): Variability of alkaloid profiles in Oophaga pumilio Tecnología (Conacyt) to JAV and a scholarship from the U.S. (Amphibia: Anura: Dendrobatidae) from western Panama and Fish and Wildlife Service to DAG-H. southern Nicaragua. Salamandra 44: 241-247. Myers, C.W., Daly, J.W. (1980): Taxonomy and ecology of References Dendrobates bombetes, a new Andean frog with new skin toxins. American Museum Novitates 2692: 1–23. Biavati, G.M., Wiederhecker, H.C., Colli, G.R. (2004): Diet Pinkas, L., Oliphant, M.S., Iverson, I.L.K. (1971): Food habits of Epipedobates flavopictus (Anura: Dendrobatidae) in a of Albacore, Bluefin tuna, and Bonito in California waters. Neotropical Savanna. Journal of Herpetology 38: 510-518. California Fish and Game, Fish Bulletin 152: 1-105. Caldwell, J.P. (1996): The evolution of myrmecophagy and its R Core Team (2013): R: A language and environment for statistical correlates in poison frog (Family Dendrobatidae). Journal of computing. R Foundation for Statistical Computing, Vienna, Zoology (London) 240: 75–101. Austria. Available at: http://www.R-project.org/. Daly, J.W. (1998): Thirty years of discovering arthropod alkaloids Saporito, R.A., Garraffo, H.M., Donnelly, M.A., Edward, A.L., in Amphibian skin. Journal of Natural Products. 61: 162-172. Longino, J.T., Daly, J.W. (2004): Formicine ants: an arthropod Daly, J.W., Garraffo, H.M., Spande, T.F. (1999): Alkaloids from source for the pumiliotoxin alkaloids of dendrobatid poison amphibian skins. In: Alkaloids: Chemical and Biological frogs. Proceedings of the National Academy of Sciences 101: Perspectives, p. 1-161. Pelletier S.W., Ed., Pergamon, New 8045-8050. York. Saporito, R.A., Donnelly, M.A., Garraffo, H.M., Spande, T.F., Daly, J.W., Garraffo, H.M., Jain, P., Spande, T.F., Snelling, R.R., Daly, J.W. (2006): Geographic and seasonal variation in Jaramillo, C., Rand, A.S. (2000): Arthropod-frog connection: alkaloid-based chemical defenses of Dendrobates pumilio from Decahydroquinoline and pyrrolizidine alkaloids common to Bocas del Toro, Panamá. Journal of Chemical Ecology 32: 795- microsympatric myrmicine ants and dendrobatid frogs. Journal 814. of Chemical Ecology 26: 73-85. Saporito, R.A., Donnelly, M.A., Norton, R.A., Garraffo, H.M., Darst, C.R., Menéndez-Guerrero, P.A., Coloma, L.A., Cannatella, Spande, T.F., Daly, J.W. (2007a): Oribatid mites as a major D.C. (2005): Evolution of dietary specialization and chemical dietary source for alkaloids in poison frogs. Proceedings of the defense in poison frogs (Dendrobatidae): a comparative analysis. National Academy of Sciences 104: 8885-8890. American Naturalist 165: 56-69. Saporito, R.A., Donnelly, M.A., Jain, P., Garraffo, H.M., Spande, Forti, L.R., Tissiani, A.S.O., Mott, T., Strûssmann, C. (2011): T.F., Daly, J.W. (2007b): Spatial and temporal patterns of Diet of Ameerega braccata (Steindachner. 1864) (Anura: alkaloid variation in the Poison Frog Oophaga pumilio in Costa Dendrobatidae) from Chapada dos Guimaraes and Cuiabà, Mato Rica and Panama over 30 years. Toxicon 50:757-778. Grosso State, Brazil. Brazilian Journal of Biology 71: 189-196. Saporito, R.A., Spande, T.F., Garraffo, H.M., Donnelly, M.A. Franklin, E., Santos, E.M.R., Albuquerque, M.I.C. (2006): (2009): Arthropod alkaloids in poison frogs: a review of the Diversity and distribution of oribatid mites (Acari: Oribatida) ‘dietary hypothesis’. Heterocycles 79: 277-296. in a lowland rain forest in Peru and in several environments of Saporito, R.A., Norton R.A., Andriamaharavo, N.R., Garraffo., the Brazilians states of Amazonas, Rondonia, Roraima y Pará. H.M., Spande, T.F. (2011): Alkaloids in the mite Scheloribates Brazilian Journal of Biology 66: 999-1020. laevigatus: further alkaloids common to oribatid mites and Heyer, W. R., Donnelly, M. A., McDiarmid, R. W., Hayek, L. A. poison frogs. Journal of Chemical Ecology 37: 213–218. C., Foster, M.S. 1994: Measuring and monitoring biological Savitzky, A.H., Mori, A., Hutchinson, D.A., Saporito, R.A., 564 Diego A. Gómez-Hoyos et al.

Burghardt, G.M., Lillywhite, H.B., Meinwaild, J. (2012): Sequestered defensive toxins in tetrapod vertebrates: principles, patterns, and prospects for future studies. Chemoecology 22: 141-158. Smith, S.Q., Jones, T.H. (2004): Tracking the cryptic pumiliotoxins. Proceedings of the National Academy of Sciences 101: 7841- 7842. Solé, M., Beckmann, O., Pelz, B., Kwet, A., Engels, W. (2005): Stomach-ﬂushing for diet analysis in anurans: an improved protocol evaluated in a case study in Araucaria forests, southern Brazil. Studies of Neotropical Fauna and Environmental 40: 23- 28. Suárez-Mayorga, A.M. (2004): Rana venenosa del Cauca. Dendrobates bombetes. In: Libro rojo de anﬁbios de Colombia. p. 302-306. Rueda-Almonacid, J.V., Lynch, J.D., Amézquita, A., Eds., Serie Libros Rojos de especies Amenazadas de Colombia. Conservación Internacional Colombia, Instituto de Ciencias Naturales-Universidad Nacional de Colombia, Ministerio del medio Ambiente. Bogotá, Colombia. Takada, W., Sakata, T., Shimano, S., Enami, Y., Mori, N., Nishida, R., Kuwahara, Y. (2005): Scheloribatid mites as the source of pumiliotoxins in dendrobatid frogs. Journal of Chemical Ecology 31: 2403-2415. Valderrama-Vernaza, M., Ramírez-Pinilla, M.P., Serrano-Cardozo, V.H. (2009): Diet of the Andean Frog Ranitomeya virolinensis (Athesphatanura: Dendrobatidae). Journal of Herpetology 43: 114-123.

Accepted by Diogo Provete; Managing Editor: Diogo Provete