Morphological variation in Craugastor frogs Kathan 1 Morphological Variation in Craugastor frogs Across an Elevational Gradient Jessica C. Kathan Department of Wildlife, Fish, and Conservation Biology EAP Tropical Biology and Conservation Program, Spring 2016 University of California, Davis 3 June 2016 ABSTRACT I investigated the effect of altitude on the morphological variation of three species of Craugastor leaf litter frogs. These species have highly variable dorsal coloration and patterning. I collected 31 individuals of C. crassidigitus, C. podiciferus, and C. underwoodi on the Sendero Principal in the Cloud Forest in the Biological Station in Monteverde, Costa Rica. Individuals of C. podiciferus tended to be lighter at higher elevations, but no such trend was found for the other two species. I hypothesize that differences in leaf litter conditions may select for certain morphs at certain elevation, but that selection pressures are so ephemeral that a strong trend is not apparent for all species. Individuals of the species C. underwoodi tended to be larger at higher elevations, which may be due to better breeding conditions at lower elevations and a resulting higher abundance of juveniles. Variacin morfolgica de ranas Craugastor en un gradiente altitudinal RESUMEN Investigu el efecto de la altura sobre el nivel del mar en la variacin morfolgica de tres especies de ranitas de hojarasca Craugastor. Estas especies presentan amplia variacin en coloracin y patrones dorsales. Evalu 31 individuos de las especies C. crassidigitus, C. podiciferus y C. underwoodi en el Sender Prinipal de la Estacin Biolgica Monteverde, Costa Rica. Los individuos de C podiciferus tendieron a ser ms claros a mayor elevacin, pero no ocurri as en las otras dos especies. Mi hiptesis es que las condiciones de la hojara sca donde se encuentran estas ranas podra seleccionar ciertos morfos a cierta elevacin, pero sera una seleccin efmera que no es aparente en todas las especies. Los individuos de C. underwoodi tendieron a ser ms grandes a mayor elevacin, lo que podr a deberse a mejores condiciones reproductivas en elevaciones bajas, que resulta en mayor abundancia de juveniles que son de menor tamao. ____________________________________________________________________________ Crypsis is an evolutionary strategy employed by some prey animals in which their coloration and pattern matches that of their background. This morphology allows individuals to resist predation by avoiding detection from predators. Leaf litter frogs of the genus Craugastor depend heavily upon cryptic coloration to avoid predation (Scott 1983). This crypsis also has a behavioral component; individuals generally do not flee a predator until directly approached (Cooper et. al 2008). The coloration of many species of Craugastor is highly variable, a characteristic which often makes distinguishing different species of the genus difficult (Leenders 2001). This varied coloration likely reflects the heterogeneity of the forest floor (Scott 1983). I looked at three species of leaf litter frogs present in the cloud forest at the Estacion Biologica in Monteverde, Costa Rica: C. crassidigitus, C. podiciferus, and C. underwoodi. These species have fairly similar natural histories; all are diurnal insectivores that live and forage
Morphological variation in Craugastor frogs Kathan 2 within the leaf litter. All of them are also direct developers, meaning they are born as fully developed froglets and do not depend upon bodies of water to breed. The three species have slightly different elevational ranges; C. underwoodi is most abundant below 1500 meters a.s.l.; C. podiciferus is present between 10892650 meters a.s.l. and C. underwoodi has a narrower range of 920-1800 meters a.s.l. (Pounds and Fogden 2000; Savage 2002). Although dorsal coloration is highly variable, the three species are easily identifiable in the field, as C. crassidigitus has prominent tubercles on its feet, and C. podiciferus has a characteristic dark mask. Woolbright and collaboraters (2008) described a closely related species as having an equilibrium of polymorphism in which different morphs are selected for ephemerally across time and space, preserving morphological diversity. The color and pattern of the forest floor changed constantly enough that particular color morph was not selected for consistently enough to collapse the polymorphism. However, in other species, elevation is a major factor impacting coloration. In ecotherms specifically, melanism is sometimes be more prevalent at higher elevations, both to deal with cooler temperatures, and to lessen the harmful effects of UV -B radiation (Garcia et. al 2009). I wondered if Craugastor frogs would exhibit the even polymorphism that Woolbright described, or if their coloration would differ across an elevational gradient. My question, therefore, was whether the dorsal pattern and coloration of frogs in the family Craugastor changes across elevation. MATERIALS AND METHODS I spent approximately twenty hours over a period of five days from 3/15/2016 to 3/20/2016 collecting frogs on the Sendero Principal at the Biological Station in Monteverde. The site is a mix of primary and secondary cloud forest. I collected frogs between the 18:00 hs and 23:00 hs, as this was the period during which I was most easily able to find frogs. Each sampling day, I walked from 1530m to 1653 meters on the Sendero Principal, looking for frogs on either side of the trail. Specific microhabitat conditions (depth of leaf litter, area of open leaf litter, canopy cover) was heterogeneous throughout the trail. I captured each frog by hand, measured the snout-to-vent length (SVL) in millimeters using calipers, identified it to species, and took a picture of the dorsal coloration. I also measured the elevation of each capture using an altimeter and noted each frogs behavior upon discovery. I released each frog at the point of capture. Using the pictures, I sorted each frog into one of nine color categories, ranging from all dark and mostly dark with some medium to all light. I similarly sorted the dorsal patterns of each frog into four categories: uniform, mottled, horiz ontally barred, and vertically barred. I compared the size and pattern of all individuals to decrease the chance that a specific individual was caught more than once.
Morphological variation in Craugastor frogs Kathan 3 Figure 1. A sample of the shade and pattern variation displayed by Craugastor frogs. Morphs from left to right: uniform light, mottled medium, vertical barred medium, uniform dark. I used an ANOVA to test whether the shades differed across the elevational gradient, and a linear regression to determine whether SVL was correlated with el evation. RESULTS I collected a total of 31 individuals, of which 17 were C. underwoodi, 10 were C. podiciferus, and four were C. crassidigitus. Each species was found throughout the sampled elevational range. All four C. crassidigitus captured were dark and of uniform dorsal coloration, but the pattern and color of C. underwoodi and C. podiciferus individuals was highly variable, with colors ranging from light tan, to red, to dark brown, and patterns uniform coloration, vertical or horizontal barring, or mottling. C. podiciferus individuals displayed a tendency to be lighter at higher elevations and darker at lower elevations (N=8; d.f. 3, 4; p=.084). Although my results werent strongly significant, this could be in part due to the small sample size. C. underwoodi and C. crassidigitus individuals showed no such trend.
Morphological variation in Craugastor frogs Kathan 4 Figure 2. Average altitude of capture of each morph of C. podiciferus. Individuals of this species displayed a tendency to be lighter at higher elevations. Medium and light/medium morphs have no error bars because I found only one individual of each of these morphs. I found no significant relationship between altitude and SVL for C. podiciferus and C. crassidigitus individuals. However, C. underwoodi showed a trend toward larger individuals at higher altitudes that bordered on significance (p=.0505) (Figure 3). Figure 3. There is a slight positive correlation between SVL (snout to vent length) and elevation of capture of C. underwoodi individuals. This is due to the abundance of juveniles (SVL<16 mm) at lower elevations C. crassidigitus and C. podiciferus individuals displayed no such trend. DISCUSSION The absence of a prevailing morphological trend across elevation agrees with findings by Savage and collaborators (1970), who found no difference in a closely related species, E. bransfordii, among different ecoclines or across elevation. The habitat of these frogs exists on a scale that is extremely variable both temporally and spatially; the color of leaf litter and ambient sunlight can change by the centimeter or the second, respectively. This environmental heterogeneity may preserve morphological heterogeneity across the elevational gradient. The slight trend towards darker individuals at lower elevations and lighter individuals at higher elevations for C. podiciferus may indicate a subtle, but more prevailing, difference between leaf litter conditions. Although I did not notice an obvious difference in substrate across elevation, a more intense investigation of leaf litter color and depth across elevation may reveal differences. Another possible explanation is that individuals at a higher elevation are subject to a different degree or type of predation pressure. Since crypsis is an evolutionary reaction to visual predators, the absence of a certain type of predator, or the presence of different predators, may lead to the selection of a different color morph at different elevations A third possible explanation is that higher and lower populations are reproductively isolated, leadin g to an effect similar to that present in so-called ring species, where individuals at either end of a
Morphological variation in Craugastor frogs Kathan 5 geographical range are reproductively isolated but can both breed with the same intermediate populations, creating a gradient of morphologies (Irwin 20 05). In this way, lower-elevation populations, while not being entirely cut off by individuals at higher elevations, may experience genetic drift towards a darker coloration. Genetic analyses could help determine whether the morphological gradient shown is a result of genetic differences, and to what extent low and high altitude populations interbreed. If genetic differences are shown, relocation experiments could help determine whether this difference is adaptive, or merely a result of genetic drift. I also found a tendency for C. underwoodi individuals captured at higher elevations to be larger. This is likely due to the abundance of juveniles (SVL<16mm) at lower elevations. This may be due to better reproductive conditions at lower elevations. Although t his species is a direct developer, without an external tadpole stage, it still requires moist leaf -litter conditions; more extensive testing of microclimate conditions of the leaf litter across an elevational gradient is warranted. Additionally, Watling an d colleagues (2001) found that arthropod prey availability was the main driving force of reproductive activity in a related species of leaf litter frog; sampling small terrestrial invertebrates across elevation could yield information about resource availability for developing froglets, and help determine whether this could cause better breeding conditions at lower elevations. Two factors were widely unaccounted for in this study. Firstly, it is possible that these species, like many neotropical frogs, are able to change their coloration. Such an ability has not yet been documented. If this ability does exist, individuals may be able to alter their coloration to match the surrounding substrate, regardless of genotype. The second unaccounted factor is that the territory sized of individual frogs of this genus has not been documented. Its possible, therefore, that certain individuals do not stay at the same elevation. Further study on both of these points as they relate to Craugastor species is needed. ACKN OWLEDGEMENTS I owe the utmost gratitude to Federico Chinchilla and Andrs Camacho for their guidance and undying patience as my advisors. Thanks also to Frank Joyce as a voice of reason, to Flix Salazar as a master herp wrangler, and to all my peers, friends, and partners in crime in the EAP program. Additional thanks to the Instituto Monteverde and Estacin Biolgica Monteverde for expertise and equipment. Lastly, thanks to Cathy and Dan Kathan for instilling in me a love of all things wild. CITATIONS Cooper, William E., Janalee P. Caldwell, and Laurie J. Vitt 2008. Effective crypsis and its maintenance by immobility in Craugastor frogs. Copeia 2008.3: 527-532. Irwin, Darren E., et al. 2005. "Speciation by distance in a ring species." Science 307.5708: 414416. Endler, J. A., and J. J. D. Greenwood 1988. Frequency-dependent predation, crypsis and aposematic coloration [and discussion]. Philosophical Transactions of the Royal Society of London B: Biological Sciences 319.1196: 505-523. Leenders T. 2001. A Guide to Amphibians and Reptiles of Costa Rica. Distribuidores Zona Tropical, S.A.
Morphological variation in Craugastor frogs Kathan 6 Pounds, J. A., Fogden, M.P. 2002. Amphibians and Reptiles of Monteverde. Monteverde: Ecology and Conservation of a Tropical Cloud Forest. Nadkarni, N.M. and Wheelwright, N.T, eds., Oxford University Press, Oxford, UK. Savage, Jay M., and Emerson Sharon B. 1970. Central American Frogs Allied to Eleutherodactylus bransfordii (Cope): A Problem of Polymorphism. Copeia 1970.4: 62344. Web. Savage, J. M. 2002. The Amphibians and Reptiles of Costa Rica. University of Chicago Press, Chicago and London. Scott, N.J. 1983. Costa Rican Natural Histor. In, Janzen, D.H. (Ed.) University of Chicago Press, Chicago. Watling, J.I. and Donnelly, M.A., 2002. Seasonal patterns of reproduction and abundance of leaf litter frogs in a Central American rainforest. Journal of Zoology, 258(02), pp.269-276. Woolbright, L.L. and Stewart, M.M., 2008. Spatial and temporal variation in color pattern morphology in the tropical frog, Eleutherodactylus coqui. Copeia, pp.431-437