Université | EDITED BY AO Hy bas CiaupE MIAUD AND y Société Herpétologique de France ; “\ Rosert GUYETANTsau C. & Guverant R. feds} # Le Bourse om Lac / PRance ZH 1998 § Packs 285 - 294 COMPARISON OF ENVIRONMENTAL CONDITIONS AND THEIR EFFECTS ON CALL CHARACTERISTICS OF FIVE FROG SPECIES JN A COLOMBIAN HIGHLAND COMMUNITY Horst LUDDECKE, ADOLFO AMEZQUITA, FRANCISCO GUZMAN, XIMENA BERNAL, LIZA AGUDELO Universidad de los Andes, Departamento de Ciencias Biolégicas, A.A, 4976, Santafé de Bogold, D.C., CoLomBis, e-mail: holuddec@uniandes.edu.co Abstract: Among five syntopic frog species in the Colombian highlands at 3500 m altitude we found one aquatic-calling species and four terrestrial-calling species. The ter- restrial frogs called from six different plant species, whose relative abundance differed between calling areas and the general study area. Plants provide elevated perches and shelter to the frogs. We tape-recorded advertisement calls of focal males of each species under field conditions, and measured substrate or water temperature at the calling site, as well as air temperature. During each recording session we also estimated the number of conspecific neighbours calling and likely to be heard by the focal male. We measured three call characteristics (call duration, call repetition rate, pulse repetition rate). We found that each of the five species had a distinct call pattern. The call characteristics of species with water-dependent reproductive modes were more affected by substrate temperature and by social environment than those of the species with terrestrial Feproductive modes, We conclude that these differences depend in part on temperature fluctuations experienced by callers of each species, and in part on relative abundance of essential resources according to reproductive mode. Key worps: Neotropical frogs, mountains, habitat use, temperature effect, social environment & INTRODUCTION Males in frog communities often use Rand, 1983). Microclimates may differ consi- Particular calling sites according to each Species’ reproductive mode (Hédl, 1990). In Montane environments there is a trend towards waterindependent reproductive ‘modes (Tihen, 1965), accompanied by a shift towards terrestrial calling sites (Drewry and derably from the macroclimate (Bartholo- ‘mew, 1987). High-altitude environments in the tropics experience large daily tempera- ture fluctuations (Sarmiento, 1986), but high- land frogs may largely avoid the effects of extreme temperatures on their vocal perfor- 285mance by restricting their calling activity to. certain hours (Navas, 1996a), or by choosing appropriate calling sites. Vocal behaviour can also be shaped by the social environment (Wells. 1988) Contrary to aquatic-calling species, the males of terrestrial callers may not be aggre- gated at specilic sites, and the distances between conspecific calling neighbours could be larger than in pond callers. This difference in social environment coutg determine haw males of aquatic and terres. trial-calling species adiust their calls to thei, neighbours’ calling activity (Stewart ang Bishop, 1994) Our question was whether a relationship would exist between habitat use for calling the environmental conditions at the calling site, and certain call attributes of the frog species in a highland community. Material AND METHODS ‘The study area was a smat! highland valley at 3500 m altitude (4° 42° 43" N, 73° 48° 25° Wy in the Chingaza Natural Park in the eastern Andean chain near the Colombian capital Bogota. Several ponds are scattered along the bottom of the valley. Large swampy areas persist. year-round, where hundreds of small puddles formed during the rainy season. The vegetation is of the paramo type, generally Jow and herbaceous (Monasterio and Vuilleu- mier, 1986), and dominated by grasses, dwarf bamboo, and woolly plants ifig. 1). Monthly mean air temperatures are about 8°C in all months of the year, but there may be strong daily Nuctuations. Midday air temperatures may rise to over 30°C on sunny days, but mostly stay below 12°C, due to the usually heavy cloud cover, strong winds and frequent rains. Nightly frosts occur occasionally, parti- cularly in the dry season (December to March). Seasonal climate changes are gover. ned by annual rainfall pattem, which is uni- modal peaking in June with about 260 mm and a yearly total of 1900 mm rainfall (Sar- mienta, 1986), From January to june of 1998 we visited the study area about once per week, at a time when the breeding seasons of all anuran species were underway. We located a representative number of calling males of Fig. nen Partial view at the study area with abundant éwarl bamboo (dl, woolly plants tv), and grass tutseach species and measured the substrate {plant or water) temperature at the calling gite, a8 well as air temperature about | m ‘above the ground, We identified the plant gpecies on which terrestrial males were caling. Further, during 5-minute samples repeated every hour over three daily calling periods of each species, separated by at Jeast one week, we tape-recorded the vocal activity of focal males, using different models of SONY tape recorders (WM D6C, TCM-7, TCM-S68V), and estimated the number of conspecific calling neighbours likely to be heard by each Five calls per focal male were analised using Canary 1.2 software (Charif et al., 1995) on a Power Macintosh computer, Temporal call characteristics were measured on oscillograms and related to environmental {actors using simple regression analysis We sampled the relative abundance of the plant species idemified as calling sites in the areas where focal males were calling Within four 17 x 17 m squares we randomly selected a representative number of exclu- sive circles of 3.8 m? surface area, and inside each circle we counted the plant specimens belonging to each of the plant species used as calling sites. In the same way we sampled the general study area at 15 ramdomly selected points along each of four randomly established transect lines of 100 m length. ~ Resutrs The local frog community consisted of five species belonging to three different families. We found one species of hylid frog (Hyla labiais, three leptodactylid species Eleutherodactylus elegans, €, bogotensis, and Myla labialis Colestethus subpunctatus |) Eleurherodectyius nervicus: E. nervicus}, and one dendrobatid species (Colostethus subpunctatus). The temporal structure of the advertise- ment call was species specitic (fig, 2). Hyla lobiais andi Coltethus subpunctatus made one-note E. bogotensis: Fig. 2: Osc ihuygeams of a single call af each of five syntopic frog species, recorded under field conditions. Tine axes equal 1.5 5. See test for details 927calls, grouped in series with a low and high number of calls, respectively, whereas the Eleutherodactylus-species made multinote calls, separated by long silent intervals. The calls of E. nervicus, E. elegans and Hyla labialis were pulsed, whereas those of Eleutherodacty- lus bogotensis and Colostethus subpunctatus were not pulsed. Only Hyla lufidlis called from the water surface of ponds, whereas all other species used terrestrial sites. Colostethus subpunctatus males called in sheltered places, and almost exclusively at ground level next to areas where small puddles formed in the rainy season. The males of the Eleutherodactylus- species usually called from elevated sites ‘on a variety of plants (fig. 3), We did not find more than one individual calling from the same plant specimen. The six plant species identified as terrestrial calling sites had dif- ferent relative abundances (fig. 4; Good- ness-of-ft test; 72 = 054.8; dl =5; P= 0.000) Further, relative plant abundance ditereq significantly between the areas where calling frogs were located, and the geneys study area (Chi-Square test; X° = 15.1; df=5 P= 001). The main differences consisted ip dwart bamboo (Chusquea 5p} and woolly plants (Espeletia grandiflora) being relatively more common, but grass tufts (Calamagrss 5p.), moss cushions (Sphagnum sp.) andl spiny rosette plants (Puya sp) being relatively less common in areas where frogs called, Shrubs (Aragoa sp.) and aquatic sites (ponds as calling sites of Hyla labialis and puddles as tadpole deposition sites of Coostethus sub punctatus, not included in the statistical tes were rare everywhere. Substrate temperatures were signi cantly bigher than air temperatures for aquatic callers {lig 5; two-tailed paired Colostethus subpunctatus puddles labialis ttest, di = 23; 1 = -6.47; P< 0.001), and also E. nervicus and E. elegans Eleutherodactylus bogotensis igram on local distribution of five syntopie Irog species, Each black circle symbolises an indivi ddual frog in an often encountered position relative 10 large and small bodies of water and relative to round level 28840 30 / Ine. dwarf grass moss woolly spiny shrub temporal permanent bamboo tuft cushion plant rosette pond pond Relative abundance (%) Fig. 4: Relative abundance of the plont species used as terrestrial calling sites in the arcas where focal frogs were sampled (grey Columns) and in the general study area (white columns), For comparison, ponds aquatic calling sites of Hyla Jabials) and small puddles (tadpole deposition sites of Colostethus subpuncta tus) are also included. Temperature (°C) 0 2 4 6 B 1012 14 16 18 20 22 24 Hour of day Fig. 5: Comparison of the daily course of air (thin fine! and water hick fine) temperatore in the study area temperature readings were LOWESS ceyyessed against daytime), with corespooding substrate temperatures At calling sites of focal frogs. Sample sizes ate noted hetwenn parentheses: #: Eleutherodactylus nervicus (nt) M: Es bogotenss in = 151 a: eens in» 15), @ ,Colstthunsubpunctats (n= 15), Hyla labialis (a = 7. 289100} 80) 60| 40) 20] 0 Call duration (9) 100} 80) 60) 40] 20) oO. Cail rate (%) 395 7 9°11 13.15 10 80) rc) 40] 20 = 0.89 P=0.001 Pulse rate (%) }* E. nervicus * E, bogotensis +E. elegans 0 3.5 7 9111315 3 57 9 1113.15, Temperature (°C) Fig. 6: Relationship hetween substiate temperature, relative call daratinn and call or puke repetition rate in five syntopic frog, species. For interspecific Comparison all crude data were converted separately for ‘each species to percentage values, the maximum value of each variable in cach species being equal to 100%. Each symbol represents the average value for one individual. Sample sizes are n= 7 for Hyle labia- lis, = 15 for Colostethus subpunciatus, Fleutherndactvlus nervicus, and E, elegans, and a= 16 for F. bogotensis. for terrestrial callers [two-tailed paired t-tests; Colostethus subpunctalus: df = 14; t= -2.18; P = 0.047; all Eleutheroductylus- species pooled: df = 36; t = -4.80; P< 0.001). Only the aquatic-calling Hyla labialis expe- rienced a significant decrease in call dura- tion with higher temperatures, and only in H. labialis as well as in Eleutherodactylus elegans did higher calling-site temperatures signiti- cantly increased pulse and note repetition rate, respectively (fig. 6). The normal social environment, expres sed by the estimated number of cailing 200 neighbours, was positively and significantly correlated with all three parameters of the vocalisation of Hyla labialis, and with call duration of Colestethus subpunctatus (fig, 7). NO Eleutherodactylas-species showed a significant neighbour effect on calling activity. The large number of neighbours in E. elegans is fore~ most due to its relatively loud call, and does not necessarily indicate shorter nearest: neighbour distances,g 100f * g § 80) « 3 4 20.32 & 20) P< 0.001 0) . 100) = 80 A le = lit, 3° 20 = 0.58 0 P< 0.001] 0.2 46 8 1012 = 100} . . gad 2 = 60) . @ Bal 3 eons Pe pervcus 20 = 0. * E, bogotensis 0 P= 0.012 t +E elegans 0246 81012 02 4 6 8 1012 Number of calling neighbours ig. 7: Relalionships between the number of calling neighbours, relative call duration and call or pulse Lepetition rate in five syntopic frog species. For interspecific comparison all crude data wece converted separately for each species to percentage values, the maximum value of each variable in each species being, equal to 100%, Each symbol represents the average value for one in lop to bottom n = 36, 32, 36 for Hyla labialis, n ~ 26 for Colostethus subpunctatus, n iduat. Sample sizes are from 18 for Fleuthero- dictylus nervicus, from top to bottom m= +3, 44, 18 for E, bogotensis, anu = 16 for €. elegans. Discussion Hyla labialis was the species most affected by temperature (fig. 8). primarily in call duration and pulse repetition rate, in addi- tion to. significant temperature effect on alling rate reported by Navas (1996a). This author also reports a significant temperature effect on the calling rate of Colostrthus sub- Punctalus (Navas, 1996a) and a significant cor- ‘elation between temperature and number ‘1 vocatising males of this species (Navas, '996b1, but his result is based on pooling data from two altitudinally separated popt- lations, Although in all terrestrial species we found trends in the same direction as repor- ted for other anuran species whose calls were recorded within a wider temperature range (Bellis, 1957), the differences between the four highland species cannot be explai- ned exclusively by the difference in tempe- rature ranges within which frogs called Although tropical highland frogs may be able to call at very low temperatures (Navas, 1996a), in our study only the Eleutheradacty- tus-species called exposed to very harsh aorcall duration ‘ vita) case call duration Eteutnorodoctyvs cacao | se rate { 26700 coy catvate | ovotensis call duanon rate | Eleuthorodacts alee pulse ale Fig. 8: Synuptic diagram summing up the results of smperature and social efiects on call chan acteristies of five syntopic rog species, Black arrows indicate significant positive elfects, the grey arrow indicates a signi cant inverse effect, Arrow widths correspond to the magnitude of the effect based on the values of the coal cient of det mination in Figures @ andl 7, assigned to three Categories; thin: lower than 0.33, intermediate: (0.44-0.66, thick: higher than 0.67. Absence of arvow indicates that no significant effect was encounter conditions, including strong winds, They did not seem to use temporal or spatial thermal niches to reduce exposure to cold during calling activity, as proposed by Navas (1996b}, because towards dawn they called at substrate temperatures equal to air tem- peratures (two-tailed paired test; di = 8; ¢ 1.18; P< 0.272). Further, even though Hyla labialis called at water temperatures signifi- cantly higher than air temperatures, it never- theless was heard calling at pond temperatures as low as 4°C (pers. observ.). 45 comparison to plant diversity in the high mountains (Rangel et al., 1995), the ter- restrial frogs called from a very limited number of plant species. No plant seemed to be used as a calling site because of its taxonomic identity. Their main function pro- bably was to offer either shelter (e. g. spiny rosettes, protective layers of old leaves around the stem of woolly plants), or oppor- tunities to climb up and call from elevated positions (e. g. dwarf bamboo, woolly plants, shrubs). Areas where certain plant species or combinations of plant species are domi- nant, also have particular soil and drainage conditions (Rangel et al, 1995) and are either relatively dry (e.g. where grass is 292 dominant), or wet and swampy (e.g. where dwarf bamboo is dominant), Both extremes could endanger egg survival (Martin and Cooper, 1972). Soils supporting mixed plant-species associations containing most of the six species may therefore be neither too wet nor too dry. The presence ‘ol terres- trial callers in such areas may relate to ovi- position site selection, particularly in the Eleutheredactylus-species, whose direct deve- lopment takes many months (pers. observ} The social effect on call features of Hylt lubialis and Colostethus subpunctatus (lig. 8) could be related to the fact that they called in ponds and in areas next to puddles, res pectively, which were limited resources in the study area. In contrast, the Eleutherodacly- lus-species mostly called from plants which were abundant. The absence of a correlation between calling activity and number of calling neighbours in these species may be further due to their exposed positions 0” elevated sites, which probably offered more favourable conditions for long-distance sound propagation (Michelsen, 1983), and could have led to sufficient spacing and reduced vocal competition between calling neighbours. However, reactions of males ©their calling neighbours may differ between species (Sullivan and Wagner, 1988; Perill and Shepherd, 1989; Sullivan and Hinshaw, 1990), and are still insufficiently understood. The considerable intraspecific variation in call parameters in most of the five high- land species could be related to factors other than temperature and social environ- ment, for instance to their prolonged bree- ding seasons Aithough their calling activity increased in the course of many weeks, it never seemed to become stable, possibly due to fluctuating environmentat conditions (rainfall, moon phase), and to recruitment of additional males and drop-out of mated or exhausted males. There could be large dif- ferences in reproductive vigor between conspecific males calling on consecutive days as well as on any given day, and irregu- lar changes in call rate may even occur in individual males in the course of one night (Lopez and Narins, 1991) AckNowLEDceMENTS: We thank Universidad de fos Andes, Colciencias, and the Fundacién para la Promocién de la Investigacién y la Tecnologia for financial support of this research Permission to study and collect frogs in the Chingaza National Park was granted by the Colombian Ministry of the Environment REFERENCES * BaRTHOLOMEN, G.A. (1987); Interspecific comparison as a tool for ecological physiologists. In: New Directions in Ecological Physiology, p. 11-35. Feder, MLE,, Bennett, A.F, Burgeren, WW, Huey, R.B., Eds,, Cambridge, Cambridge University Press * BeLUIs, E.D, (1957): The effects of temperature on salientian breeding calls. Copeia 1957: 85-89. * CHaRir, R.A., MITCHELL, S., CLARK, C.W (1995): Canary 1.2 User’s Manuel. Comell Laboratory of Ornithology. Ithaca, New York. + Drewny, G.E., RAND, A.S. (1983): Characteristics of an acoustic community of Puerto Rican frogs of the genus Eleutherodactylus, Copeia 1983: 941-953 ‘+ Hoot, W. (1990): Reproductive diversity in Amazonian lowland frogs. Fortschr. Zool, 38: 41-60. + Lonez, PT., Nagins, PM. (1991): Mate choice in the neotropical frog, Eleutherodactylus coqui Anim. Behav. 41: 797-772 * Mantin, A.A., Coorer, A.K. (1972): The ecology of terrestrial anuran eggs, genus Crinia (Leptodactylidae). Copeia 1972; 163-168 * Micnieusen, A. (1983): Biophysical basis of sound communication. In: Bioacoustics. A Comparative Approach, p. 3-38 Lewis, B., Ed., New York, Academic Press. + MonasreRto, M., Vuitteumies, F (1986): Introduction: High tropical mountain biota of the world. In; High Akitude Tropical Biogeography, p. 3-7. Vuilleumier, f., Monasterio, M., Eds., New York, Oxford Univ. Press. * Navas, CA. (1996a): Thermal dependency of field locomotor and vocal performance of high- elevation anurans in the tropical Andes. |, Herpetol, 30: 478-487 * Navas, C.A. (1990b}: The effect of temperature on the vocal activity of tropical anurans: a comparison of high and low-elevation species. |. Herpetol. 30: 488-497. * Pexuiut, S.A., SuceHeRD, WJ. (1989): Spatial distribution and male-male communication in the northern cricket frog, Acris crepitans blenchurdi. J. Herpetol. 23: 237-243, * Rance, J.0., Lowy, P, Cirek A.M, (1995): Comunidades vegetales en regiones paramunas del altiplano Cundiboyacense. En: E] Péramo, Un Ecosistema de Alta Montafa, p. 95-131 Reyes, P, Molano, |., Gonzalez, F, Conés, A., Rangel, O., Fldrez, A., Iriante, P., Kraus, E, Eds., Bogota, 0.C., Fundacién Ecosistemas Andinos - ECOAN, Colombia * Sarsuento, G, (1986); Ecological features of climate in high tropical mountains. In: High Altitude Tropical Biogeography. p. 11-45. Vuilleumier, F, Monasterio, M., Eds., New York, Oxford Univ. Press, 293"* Srewarr, MM., BisHor. PJ. (1994): Effects of increased sound level of advertisement calls on calling male frogs, Eleutherodactylus coqui. |. Herpetol. 28: 46-53. * SULLIVAN, B.K., HinstiAw, SH. (1990): Variation in advertisement calls and male calling beha- ior in the spring peeper (Pseudacrs crucifer). Copeia 1990: 1146-1150. * Suuuivan, BK, WacNer, WE. (1988): Variation in advertisement and release calls, and social influences on calling behavior in the Gulf Coast toad (Bufo valliceps). Copeia 1988: 1014-1020, # THEN, |.A. (1965): Evolutionary trends in frogs. Amer. Zool. 5: 309-318. + Weuts, K.D. (1988): The effect of social interactions on anuran vocal behavior. tn: The Evolution of the Amphibian Auditory System, p. 433-454. Fritzsch, B.. Ryan, MJ, W, Hetherington, TE., Walkowiak, W., Eds., New York, John Wiley and Sons. aaa