Egg Size, Incubation Temperature, and Posthatching ...

308

SHORTER COMMUNICATIONS

Journal of Herpetology, Vol. 36, No. 2, pp. 308–311, 2002 Copyright 2002 Society for the Study of Amphibians and Reptiles

Egg Size, Incubation Temperature, and Posthatching Growth in Painted Turtles (Chrysemys picta) FREDRIC J. JANZEN1 AND CARRIE L. MORJAN, Department of Zoology and Genetics, 339 Science II, Iowa State University, Ames, Iowa 50011-3223, USA Incubation temperature has profound and pervasive effects on phenotypes of developing embryos and offspring of oviparous reptiles (reviewed in Deeming and Ferguson, 1991). Perhaps the most compelling impact involves sex determination in many of these species. Indeed, this temperature-dependent sex determination (TSD) in reptiles has received considerable attention (reviewed in Bull, 1983; Ewert and Nelson, 1991; Janzen and Paukstis, 1991; Shine, 1999). Most adaptive explanations for the existence and persistence of TSD hypothesize that incubation temperatures benefit the sexes differentially (sensu Charnov and Bull, 1977). Incubation temperature does, in fact, influence many traits that may affect fitness and examining such traits may disclose sex-specific effects in accord with the Charnov-Bull model. A number of studies have adopted this exploratory approach, yet no general explanation for the adaptive significance of TSD in reptiles has been validated empirically (reviewed by Shine, 1999). One promising hypothesis suggests that the evolutionary persistence of TSD in reptiles could be explained by a covariance between egg size, nest thermal environment, and sexual size dimorphism (Roosenburg, 1996; Roosenburg and Niewiarowski, 1998). That is, a key trait (i.e., egg size) might have sex-specific (i.e., incubation temperature-specific) effects on fitness by influencing posthatching growth rates and ages of maturity. Specifically, (1) egg size must be correlated with eventual size of an individual; (2) incubation temperature must influence posthatching growth rates; (3) any covariance between egg size and temperature affecting growth must be sex-specific; (4) nesting females should be able to discriminate among among environmental cues related to nest temperature; and (5) these females should use such environmental cues to manipulate offspring sex ratio according to their egg sizes. The only attempt to evaluate this hypothesis produced considerable support. In diamondback terrapins (Malaclemys terrapin), a turtle with TSD and largefemale sexual size dimorphism in adulthood, egg size influenced posthatching growth rates and ages of maturity for females but not for males (Roosenburg and Kelley, 1996). Fieldwork showed subsequently that nesting terrapins generally laid larger eggs in warmer, female-producing microhabitats and smaller eggs in cooler, male-producing microhabitats (Roosenburg, 1996). Our study is a first attempt at testing the generality

1 Corresponding Author. E-mail: fjanzen@iastate. edu

of Roosenburg’s hypothesis. We evaluated the first three key conditions of the hypothesis (see above) using painted turtles (Chrysemys picta), an emydid turtle with TSD and large-female sexual size dimorphism [the remaining two conditions involving nesting biology are evaluated elsewhere (unpubl. data)]. Specifically, we explored the effects of egg size and incubation temperature on posthatching body size and growth. We expected to obtain results concordant with those of Roosenburg and Kelley (1996) given the similarity of painted turtles to terrapins phylogenetically and in patterns of TSD (Paukstis and Janzen, 1990) and sexual size dimorphism (Ernst et al., 1994). Eggs from 13 fresh nests were collected during June 1989 near the Thomson Causeway, an island in the Mississippi River near Thomson, Carroll County, Illinois (41⬚57⬘N, 90⬚07⬘W; Janzen, 1994). These eggs were weighed to the nearest 0.01 g with an Ohaus portable electronic balance and then transported immediately to the University of Chicago. In the laboratory, a subset of these eggs was chosen for this experiment. Eggs were assigned randomly to a position in a 3 ⫻ 8 matrix in 4 shoeboxes containing moist vermiculite (⫺150 kPa ⫽ 300 g dry vermiculite : 337 g deionized water). Two shoeboxes each were then placed in incubators set at either 26 (male-producing) or 30⬚C (female-producing) (Schwarzkopf and Brooks, 1985; Paukstis and Janzen, 1990). These temperatures are well within the range experienced by natural nests in this population during embryonic sexual differentiation (Weisrock and Janzen, 1999:fig. 1; for a detailed explanation on comparing constant and fluctuating incubation temperatures, see Georges, 1989). Containers were rehydrated weekly to replace any lost water and were rotated daily within incubators to minimize the effects of thermal gradients. Dissections of carcasses of turtles that died during the study confirmed that 26 of 27 individuals incubated at 26⬚C were male (the single female suffered severe deformities) and 17 of 17 individuals incubated at 30⬚C were female. After hatching in August 1989, 77 turtles were weighed to the nearest 0.01 g and then placed in an enclosed 3 ⫻ 3 m outdoor concrete pond with a sloping bottom ranging from 0.3 m to ⬎1 m. This pond, which received partial sun during the day, was used to rear turtles from April to November. Turtles were hibernated indoors at 5 C from December through March. Juveniles were fed waxworms, trout feed, and Reptomin ad libitum three times per week during the growing period. In addition, insect larvae and algae were naturally abundant in the pond for feeding. Individual mass was measured in late summer for each subsequent year of survival through 1992. However, we only report data for the first year of growth because many turtles were stolen in 1991, which greatly reduced sample sizes. Growth for hatchlings for the first year was calculated as the mass gained by an individual from hatching to one year of age, and differences between males (26⬚C) and females (30⬚C) were compared using a t-test. We then evaluated the relative effects of egg mass and incubation temperature on hatchling and juvenile masses by performing two analyses of covariance. With incubation temperature, egg mass, and their interaction as effects, we used hatchling mass as a response in the first analysis and mass at one year

SHORTER COMMUNICATIONS TABLE 1. of age.

309

Mass (x¯ ⫾ SD) of female and male painted turtles (Chrysemys picta) at hatching and at one year

Year

1989 (hatchling) 1990

Sex

female male female male

Mass (g)

N

t

P

⫾ ⫾ ⫾ ⫾

37 40 13 20

0.016

0.87

4.61

0.0001

4.47 4.45 14.73 11.07

0.13 0.13 0.62 0.50

in the second. Because females hatched on average two weeks earlier than males, we also used analysis of covariance to evaluate the relative effects of age (in days), hatchling mass, and incubation temperature on juvenile mass at approximately one year of age. We could not include clutch as a factor in our analyses because of the small sample sizes caused by low hatching success and low first-year survival. We also could not statistically separate the effects of incubation temperature and sex on growth in this study because these factors were confounded. Consequently, we use the terms ‘‘incubation temperature’’ and ‘‘sex’’ interchangeably throughout the manuscript. All statistical analyses were performed using JMP version 3.2.1 (SAS Institute, Inc., Cary, NC, 1997). Of the 96 eggs incubated for this experiment, 40 (83%) hatched at 26⬚C and 37 (77%) hatched at 30⬚C. Of these 77 hatchlings, only 20 (50%) from 26⬚C and 13 (35%) from 30⬚C survived to the end of the following summer (i.e., August 1990). Most of this mortality, which had no obvious causes, occurred in November 1989 just before hibernation and in spring 1990 shortly after removal from hibernation. The 33 turtles that survived to at least one year posthatching form the basis of our growth analyses. Although no differences existed in mass at hatching between males (eggs incubated at 26⬚C) and females (eggs incubated at 30⬚C), females were significantly heavier than males after the first year of growth (Table 1). When considering strictly the subset of individuals that survived to their first year, no significant differences in hatchling mass existed between males and females (N ⫽ 33, t ⫽ 1.12, P ⫽ 0.27). During the first year, however, these females grew significantly more than males (x¯females ⫽ 10.20 g ⫾ 0.53, x¯males ⫽ 6.27 g ⫾ 0.43, t ⫽ 5.76, P ⬍ 0.0001, df ⫽ 31). No egg mass by incubation temperature interaction was present among the analyses (hatchling mass: F1,73 ⫽ 0.39, P ⫽ 0.54; mass at one year: F1,29 ⫽ 0.030, P ⫽ 0.86; Table 2); therefore egg mass affected hatchling and juvenile size similarly between the sexes (Fig. 1), and this interaction effect was excluded from further analyses. Egg mass significantly and positively affected hatchling mass, whereas incubation temperature

marginally affected hatchling mass in a negative direction. At one year of age, both incubation temperature and egg mass significantly and positively affected juvenile mass. Even accounting for the fact that females (eggs incubated at 30⬚C) hatched 14 days earlier than males on average (eggs incubated at 26⬚C), age (in days) did not significantly affect juvenile mass at one year after accounting for the effects of incubation temperature and hatchling mass (age: F1,29 ⫽ 0.52, P ⫽ 0.47; incubation temperature: F1,30 ⫽ 38.14, P ⬍ 0.0001; hatchling mass: F1,30 ⫽ 14.63, P ⬍ 0.001). In summary, although female painted turtles grew faster than males during their first year, larger eggs produced larger hatchlings for both sexes (Fig. 1). Our intent was to investigate three key conditions of a recent hypothesis to explain the adaptive maintenance of temperature-dependent sex determination (TSD) in reptiles (Roosenburg, 1996). Specifically, we evaluated the effects of (1) initial egg mass and (2) incubation temperature (sex) on juvenile growth for one year posthatching in painted turtles and (3) whether any egg and temperature effects on growth were sex-specific. We detected persistent effects of egg mass on juvenile mass through one year of age, as well as differential growth rates between males (eggs incubated at 26⬚C) and females (eggs incubated at 30⬚C). We did not observe a sex-specific covariance between effects of egg mass and temperature on juvenile mass. Our results for these three conditions of Roosenburg’s hypothesis are largely consistent with those of Roosenburg and Kelley (1996) for diamondback terrapins. However, we detected a positive relationship between egg mass and juvenile mass at one year of age for both sexes (incubation temperatures) in painted turtles (Fig. 1), whereas egg size significantly affected juvenile size for only females in diamondback terrapins (Roosenburg and Kelley, 1996). Furthermore, although egg mass was an important factor for juvenile body size for both sexes throughout early growth in painted turtles, the persistence of such effects until age of maturity is currently unknown. Preliminary information from mark-recapture work in our study population indicates rapid maturation (⬃3 years in males and ⬃5 years in females; FJJ, unpubl.), so rapid

TABLE 2. Results from two analyses of covariance evaluating the effects of egg mass and incubation temperature on mass of painted turtles (Chrysemys picta) at hatching and at one year of age. Age (N)

Hatchling (77) One year (33)

Effect

Egg Mass Incubation Temperature Egg Mass Incubation Temperature

Estimate ⫾ SE

0.72 ⫺0.040 1.27 0.98

⫾ ⫾ ⫾ ⫾

0.039 0.020 0.37 0.17

F (df ⫽ 1)

P

331.57 3.97 11.49 32.01

⬍ 0.0001 0.050 0.002 ⬍ 0.0001

310

SHORTER COMMUNICATIONS elucidate whether females take advantage of the documented effects of egg size on juvenile growth to benefit the sexes differentially. If Roosenburg’s (1996) hypothesis holds, we predict for painted turtles that (1) body size will correlate positively with reproductive success for females but not for males and (2) individuals with relatively large eggs will oviposit in warmer, less vegetated, female-producing sites than painted turtles with relatively small eggs.

FIG. 1. Mass at 1 yr for male (open diamonds) and female (filled diamonds) painted turtles as a function of initial egg mass. The regression lines for sexes (⫽ incubation temperatures) were significantly different from 0 but not from each other.

growth even to one year of age is likely to be important ecologically in this case. The sex-specific benefits of larger body size and faster growth in painted turtles have yet to be documented definitively. Mature females are certainly larger (reviewed by Ernst et al., 1994) and typically grow faster as juveniles (reviewed by Ernst et al., 1994; St. Clair et al., 1994; but see Wilbur, 1975) than males under natural and seminatural conditions (Table 1). Most crucially, fecundity scales positively with body size in females (reviewed in Iverson and Smith, 1993), but apparently scales negatively with body size in males (S. McTaggart, unpubl.). In other words, larger females produce more eggs and smaller males sire more offspring. More information on mating patterns would be helpful to confirm the body size/fitness results for males, but the current implications are that females are likely to benefit more than males from larger body size in painted turtles. If indeed larger body size benefits females more than males, then our results support several key conditions of Roosenburg’s hypothesis for the adaptive maintenance of TSD in reptiles (Roosenburg, 1996). Larger painted turtle eggs lead to larger juveniles of both sexes (incubation temperatures), but this effect of accelerating growth may only (or primarily) be beneficial for females. The stage is thus set for further empirical evaluation of Roosenburg’s hypothesis by examining the mating system, nesting biology, and general evolutionary ecology of painted turtles in the field. Painted turtle nests with more overstory vegetation cover tend to be cooler than nests with less overstory vegetation cover at our study site (Weisrock and Janzen, 1999). Because overstory vegetation cover around a given nest is constant throughout embryonic development at this site (Janzen, 1994), this variable provides a realistic environmental cue that a turtle could use during oviposition to predict relative nest temperatures during subsequent sexual differentiation of embryos. Evaluating patterns of nest-site choice would

Acknowledgments.—We thank the Army Corps of Engineers and the Thomson Park Rangers for permission to work at the field site and K. Dodd, A. Georges, and an anonymous reviewer for helpful comments. Research at the Thomson Causeway was conducted under authority of scientific permit W-9225 from the Illinois Department of Conservation. Turtles were handled in accordance with University of Chicago Animal Welfare Assurance 52711. This project was supported in part by National Science Foundation grants BSR-8914686 and DEB-9629529 to FJJ and an National Science Foundation predoctoral fellowship to CM. Journal Paper No. J-19487 of the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa, Project No. 3369, and supported by the Hatch Act and State of Iowa Funds.

LITERATURE CITED BULL, J. J. 1983. Evolution of Sex Determining Mechanisms. Benjamin/Cummings, Menlo Park, CA. CHARNOV, E. L., AND J. BULL. 1977. When is sex environmentally determined? Nature 266:828–830. DEEMING, D. C., AND M. W. J. FERGUSON. 1991. Physiological effects of incubation temperature on embryonic development in reptiles and birds. In D. C. Deeming and M. W. J. Ferguson (eds.), Egg Incubation: Its Effects on Embryonic Development in Birds and Reptiles, pp. 147–171. Cambridge University Press, Cambridge. ERNST, C. H., J. E. LOVICH, AND R. W. BARBOUR. 1994. Turtles of the United States and Canada. Smithsonian Institution Press, Washington, DC. EWERT, M. A., AND C. E. NELSON. 1991. Sex determination in turtles: diverse patterns and some possible adaptive values. Copeia 1991:50–69. GEORGES, A. 1989. Female turtles from hot nests: is it duration of incubation or proportion of development at high temperatures that matters? Oecologia 81:323–328. IVERSON, J. B., AND G. R. SMITH. 1993. Reproductive ecology of the painted turtle (Chrysemys picta) in the Nebraska sandhills and across its range. Copeia 1993:1–21. JANZEN, F. J. 1994. Vegetational cover predicts the sex ratio of hatchling turtles in natural nests. Ecology 75:1593–1599. JANZEN, F. J., AND G. L. PAUKSTIS. 1991. Environmental sex determination in reptiles: ecology, evolution, and experimental design. Quarterly Review of Biology 66:149–179. PAUKSTIS, G. L., AND F. J. JANZEN. 1990. Sex determination in reptiles: summary of effects of constant temperatures of incubation on sex ratios of offspring. Smithsonian Herpetological Information Service 83:1–28. ROOSENBURG, W. M. 1996. Maternal condition and nest site choice: an alternative for the maintenance

SHORTER COMMUNICATIONS of environmental sex determination? American Zoologist 36:157–168. ROOSENBURG, W. M., AND K. C. KELLEY. 1996. The effect of egg size and incubation temperature on growth in the turtle, Malaclemys terrapin. Journal of Herpetology 30:198–204. ROOSENBURG, W. M., AND P. NIEWIAROWSKI. 1998. Maternal effects and the maintenance of environmental sex determination. In T. A. Mousseau and C. W. Fox (eds.), Maternal Effects as Adaptations, pp. 307–322. Oxford University Press, Oxford. SCHWARZOPF, L., AND R. J. BROOKS. 1985. Sex determination in northern painted turtles: effect of incubation at constant and fluctuating temperatures. Canadian Journal of Zoology 63:2543–2547. SHINE, R. 1999. Why is sex determined by nest temperature in many reptiles? Trends in Ecology and Evolution 14:186–189. ST. CLAIR, R., P. T. GREGORY, AND J. M. MACARTNEY. 1994. How do sexual differences in growth and maturation interact to determine size in northern and southern painted turtles? Canadian Journal of Zoology 72:1436–1443. WEISROCK, D. W., AND F. J. JANZEN. 1999. Thermal and fitness-related consequences of nest location in painted turtles (Chrysemys picta). Functional Ecology 13:94–101. WILBUR, H. M. 1975. A growth model for the turtle Chrysemys picta. Copeia 1975:337–343. Accepted 24 August 2001.


Rain-Harvesting Behavior in Agamid Lizards (Trapelus) MILAN VESELY´ 1, Department of Zoology, Faculty of Natural Sciences, Palacky University, Svobody 26, 771 46 Olomouc, Czech Republic; E-mail: [email protected] DAVID MODRY´ 2, Department of Parasitology, University of Veterinary and Pharmaceutical Sciences, Palacke´ho 1-3, 612 42 Brno, Czech Republic; E-mail: [email protected]. muni.cz The ability of arid region reptiles to use the body surface as a collector of water has been known for many years, although it has been described only in a small number of species. Bentley and Blumer (1962) demonstrated that Moloch horridus (Agamidae) drink the water film passively transported by capillary action of the skin to the mouth, thus contradicting a theory that water is absorbed transcutaneously in this species (Buxton, 1923). Gans et al. (1982) reexamined Moloch skin with the aid of scanning electron microscope (SEM) photographs and attributed the water Corresponding Author. Institute of Parasitology, Academy of Sciences of the Czech Republic, Branisovska´ 31, 370 05 Ceske´ Budejovice, Czech Republic; E-mail: [email protected]. muni.cz 1 2

311

flow to capillary forces generated by grooves between scales. Schwenk and Greene (1987) described a similar system in another agamid, Phrynocephalus helioscopus, and reported that capillary forces pull water through interscalar channels. In addition, they described a stereotyped posture that was exhibited when this species was sprayed with water. This posture involved lowering the head, raising the splayed hindquarters, and protruding the tongue. Comparable behavior has not been observed in Moloch (Sherbrooke, 1993). In the iguanid lizard Phrynosoma cornutum a similar behavior was described by Sherbrooke (1990). His term ‘‘rainharvesting’’ is used for a complex of behavioral and morphological characters associated with this form of water collection (Sherbrooke, 1990, 1993; Withers, 1993). More recently, such behavior has been described in Phrynosoma platyrhinos (Peterson, 1998). Herein, we report rain-harvesting in three agamids (Trapelus pallidus, Trapelus flavimaculatus , and Trapelus mutabilis). Examination of two additional species of the same genus (Trapelus ruderatus and Trapelus sanguinolentus) did not reveal the presence of this behavior. All the animals examined have relatively similar biology, inhabiting more or less open arid to semiarid areas of Northern Africa, the Near East, and the Middle East. During 1998, 41 specimens of Trapelus were examined for the presence or absence of rain-harvesting behavior. Numbers of specimens and their origin are as follows: T. pallidus (Reuss, 1834), 16 specimens from the Eastern Desert, Jordan; T. flavimaculatus (Ru¨ppell, 1835), three specimens obtained from a commercial dealer; T. mutabilis (Merrem, 1820), 10 specimens from northern Egypt; Trapelus ruderatus ruderatus (Olivier, 1804), nine specimens, from the Amman region, western Jordan; T. sanguinolentus (Pallas, 1827), three specimens obtained from a private reptile keeper. Lizards were housed in glass terraria with sand or gravel substrate. Terraria were heated by incandescent lamps and illuminated by fluorescent tubes. Lizards were fed crickets, wormsmealworms, and Zophobas morio larvae, with vitamin-mineral supplement every two to three days. Water dishes were placed in each terrarium, and the enclosures were sprayed twice per week. The drinking behavior of each animal was observed and recorded. Water was not available for three days prior to each trial. During trial observations, the subject was sprayed with water using a hand pump sprayer. Spraying lasted 3 min, and time to the beginning of drinking (movement of the jaw and tongue) was recorded. Additionally, the type and duration of rain-harvesting posture, and the presence or absence of wet stone licking was recorded. Minimum and maximum duration of these behavioral characters were measured in seconds. Animals exhibiting typical rain-harvesting posture were photographed using Nikon F90 camera with Sigma 90 mm macro-lens and Nikon SB 27 speed light. Each subject was observed in three separate trials with an interval of two weeks between trials. The ability of skin to carry water by capillary action in interscalar channels was tested in two individuals of each species. The experiments were similar to those of Schwenk and Greene (1987). The animals were wetted by light spraying, then blotted dry with a paper tissue and placed back into the enclosure. Next, water

312


FIG. 1. Rain-harvesting posture of female of Trapelus pallidus, 2 min after beginning of spraying.

dyed with blue food coloring (E 576, Ovo, Prague) was dropped on the dorsum to visualize the movement of water on the integument. Speed and direction of water movement were noted. For morphological comparison of skin surface microstructure, scanning electron microscopy (SEM) photographs were used. A small piece of skin was taken from the scapular region of a specimen of T. pallidus and T. ruderatus, freshly killed by overdosing with barbiturates (Thiopental, Spofa). Samples for SEM were fixed in 4% buffered paraformaldehyde, dehydrated in ethanol and acetone series, and desiccated by critical point drying in carbon dioxide. Samples were then gold-coated and examined and photographed with a JEOL JMS 6300 scanning microscope using various magnifications (50⫻, 100⫻, and 350⫻). Behavioral rain-harvesting was recorded in T. pallidus, T. mutabilis, and T. flavimaculatus. In contrast, T. ruderatus and T. sanguinolentus did not exhibit any indication of this behavior. Typical rain-harvesting posture in T. pallidus was assumed 5–50 sec after the animal was sprayed. The head was depressed to within 2–9 mm above the substrate, and the hindquarters were elevated so that the sacral region, or base of the tail, was the highest point on the body. The limbs were splayed, particularly the hind limbs (Fig. 1). The posture varied little among individuals or among trials with the same individual. The rain-harvesting posture of T. mutabilis and T. flavimaculatus did not differ substantively from that observed in T. pallidus. However, in some trials the animals exhibited a modified rain-harvesting posture, in which head was slightly depressed, and the lizard protruded the tongue, but the angle of body axis remained unchanged. In T. ruderatus and T. sanguinolentus, the reaction to spraying involved only lowering the head and licking drops from the surface of wet stones. Stone licking was observed in 23% of trials with T. pallidus, 10% of trials with T. mutabilis, and 11% of trials with T. flavimaculatus. During observations with dyed water, no obvious differences in speed and direction of water flow in interscalar channels were observed among individuals and species. Movement of dye was dependent on angle of body and the amount of water dropped on the body surface. SEM photographs of integument from the dorsolat-

eral body surface revealed similar-shaped scales and interscalar channels (Fig. 2A–B) in both T. pallidus (rain-harvesting) and T. ruderatus (non–rain-harvesting). In both species, typical honeycomb microstructures were present on the upper surfaces of scales in the scapular region, as well as in deep interscalar grooves (Fig. 2C–D). Three species of Trapelus displayed rain-harvesting behavior. This behavior has not been reported in this genus before. Abel (1952) observed a similar posture in T. ruderatus, and suggested that this behavior occurred because of ‘‘unpleasant contact of venter with wet substrate.’’ We question both this interpretation and whether Abel’s specimens were T. ruderatus, because he gave no description, and none of the nine T. ruderatus we examined exhibited this behavior. Trapelus pallidus and T. ruderatus are strictly parapatric in Jordan, occupying similar niches in habitats differing primarily by altitude. Trapelus ruderatus occurs in a dry part of the Mediterranean ecozone and in a mesic part of the Irano-Turanian ecozone with annual rainfall varying between 150 and 300 mm. In contrast, T. pallidus inhabits semidesert and desert habitats of the Syrian Desert, where mean annual rainfall does not exceed 100 mm (Disi, 1996; Disi and Amr, 1998; Disi et al., 2001). Hence, the presence of rain-harvesting behavior in T. pallidus may represent a behavioral adaptation for inhabiting habitats more arid than those inhabited by T. ruderatus. Honeycomb microstructures observed in SEM microphotographs of the integument of T. pallidus are similar to those of other rain-harvesting lizards (Sherbrooke, 1990; Schwenk and Greene, 1987). However, honeycomb structures are widely distributed among Iguania (Steward and Daniel, 1975; Ananjeva et al., 1991) and thus seem not directly correlated with rainharvesting behavior. The nearly identical integumental micromorphology of T. pallidus (rain-harvesting) and T. ruderatus (non–rain-harvesting) also indicate that these structures are not necessarily adaptations for rain-harvesting. The steep angle of the longitudinal body axis associated with the rain-harvesting posture suggests that gravity may be sufficient for water transport, and capillary action is not necessary for transport of water collected on the dorsal surface in Trapelus. Based on comparison with other reptile species in which rain-harvesting behavior has been reported, we propose three forms of this behavior, defined by body posture and role of capillary action. First, Moloch horridus is an Australian agamid lizard with the ability to drink water delivered to the mouth by capillary action in interscalar channels. These channels may be filled by water from rain, dew, droplets of water collected by body surfaces on vegetation or by absorption of water from damp sand. Although the ability to extract water from a wet substrate is evident in Moloch, the only stereotyped behaviors accompanying this process are rubbing of the ventral surface on the substrate and kicking sand onto the back (Gans et al., 1982; Sherbrooke, 1993; Withers, 1993). Second, in Phrynosoma cornutum and Phrynosoma platyrhinos, stereotyped behavior accompanying collection of rainwater has been reported. In these species, the stereotyped posture includes flattening the body and spreading the dorsal surface to maximize


313

FIG. 2. Scanning electron photomicrograph of epidermal surface of Trapelus pallidus (A) and Trapelus ruderatus (B). Detail of interscalar space showing honeycomb microornamentation of scale surface in T. pallidus (C) and T. ruderatus (D).

interception of raindrops. The head is depressed to within several millimeters of the substrate, and the limbs are splayed (particularly the hind limbs). The posture is elicited primarily when the body surface is wetted; however, a similar posture termed ‘‘dorsal shield’’ has been reported by Sherbrooke (1991) as a defensive posture. Drinking of collected water is not accompanied by licking the substrate in P. cornutum, but this behavior has been documented in P. platyrhinos (Sherbrooke, 1981). Capillary transport of water is reported to be more significant during light precipitation, when the water amount collected by the dorsal surface is not large enough to flow by gravity alone. Presence of integumentary microstructures, as are reported in Moloch, was confirmed, but capillary flow seems to be less effective (Sherbrooke, 1990). Whether these animals are able to extract water from a wet substrate remains unclear (Sherbrooke, 1993) but seems rather improbable. Third, in some species of the genus Phrynocephalus (Schwenk and Greene, 1987) and Trapelus (this study), stereotyped postures similar to those mentioned above have been observed. The postures involved head depression and elevation of the hindquarters, so that the caudal part of the dorsum was the highest part of the body. The limbs are splayed, particularly the hind limbs. The head is depressed, sometimes touching the substrate or stones, and the tongue is protruded repeatedly. This posture is usually assumed usually several seconds after the animal has

been sprayed by water. In contrast to previous behavioral form, the body axis is steep, and water collected by the dorsum accumulates on the rostrum forming a drop, which is ingested by regular opening of mouth and protruding of the tongue. In this case, capillary action seems to play no role in water transport. Agamids of the genus Trapelus occasionally lick or touch the rostrum to water in depressions or on stone surfaces. Within the Agamidae, Phrynocephalus and Trapelus represent rather distant evolutionary groups (Joger, 1991); presumably rain-harvesting behavior exhibited in these genera evolved independently. In both Phrynocephalus and Trapelus, the rain-harvesting posture resembles a threat posture and may have evolved by modification of this preexisting trait. It is probable that similar rain-harvesting behavior occurs in other agamid species inhabiting arid regions. Acknowledgments.—We thank D. Hegner from providing experimental animals and B. Koudela (University of Veterinary and Pharmaceutical Sciences, Brno) for kind assistance with SEM microscopy. We are highly indebted to W. C. Sherbrooke (American Museum of Natural History) for language assistance and comments on manuscript. This study was supported by internal research grant (3210-3006) of Faculty of Natural Sciences of Palacky University, Olomouc.

LITERATURE CITED ABEL, J. 1952. Zur biologie von Agama agilis Ol. und Agama ruderata Ol. Zoologische Anzeiger 144:125– 133.

314


ANANJEVA, N. B., M. E. DILMUCHAMEDOV, AND T. N. MATVEYEVA. 1991. The skin sense organs of some Iguanian lizards. Journal of Herpetology 25:186– 199. BENTLEY, P. J., AND W. F. C. BLUMER. 1962. Uptake of water by the lizard, Moloch horridus. Nature 194: 699–700. BUXTON, P. A. 1923. Animal Life in Deserts. Edward Arnold, London. DISI, A. M. 1996. A contribution to the knowledge of the herpetofauna of Jordan. IV. The Jordanian herpetofauna as zoogeographic indicator. Herpetozoa 9:71–81. DISI, A. M., AND Z. S. AMR. 1998. Distribution and ecology of lizards in Jordan (Reptilia: Sauria). Faunistische Abhandlungen des Staatlichen Museum fu¨r Tierkunde Dresden 21 (Suppl. 1998):43–66. DISI, A. M., D. MODRY´, P. NECAS, AND L. RIFAI. 2001. Amphibians and Reptiles of Hashemite Kingdom of Jordan. An Atlas and Field Guide. Edition Chimaira, Frankfurt am Main, Germany. GANS, C., R. MERLIN, AND W. F. C. BLUMER. 1982. The water-collecting mechanism of Moloch horridus reexamined. Amphibia-Reptilia 3:57–64. JOGER, U. 1991. A molecular phylogeny of agamid lizards. Copeia 1991:616–622. PETERSON, C. C. 1998. Rain-harvesting behavior by a free-ranging desert horned lizard (Phrynosoma platyrhinos). Southwestern Naturalist 43:391–394. SCHWENK, K., AND H. W. GREENE. 1987. Water collection and drinking in Phrynocephalus helioscopus: a possible condensation mechanism. Journal of Herpetology 21:134–139 SHERBROOKE, W. C. 1981. Horned Lizards: Unique Reptiles of Western North America. Southwest Parks and Monuments Association, Globe, AZ . 1990. Rain-harvesting in the lizard, Phrynosoma cornutum: behavior and integumental morphology. Journal of Herpetology 24:302–308. . 1991. Behavioral (predator-prey) interactions of captive grasshopper mice (Onychomys torridus) and horned lizards (Phrynosoma cornutum and P. modestum). American Midland Naturalist 126:187– 195. . 1993. Rain-drinking behaviors of the Australian thorny devil (Sauria: Agamidae). Journal of Herpetology 27:270–273. STEWARD, G., AND P. DANIEL. 1975. Microornamentation of lizard scales: variation and taxonomic correlations. Herpetologica 31:425–428 WITHERS, P. 1993. Cutaneous water acquisition by the thorny devil (Moloch horridus: Agamidae). Journal of Herpetology 27:265–270. Accepted: 8 September 2001.


Environmental Factors Affecting Calling Activity of a Tropical Diurnal Frog (Hylodes phyllodes: Leptodactylidae) FABIO H. HATANO, CARLOS F. D. ROCHA,1 AND MONIQUE VAN SLUYS, Setor de Ecologia, Departamento de Biologia Animal e Vegetal, Instituto de Biologia Roberto Alcaˆntara Gomes, Universidade do Estado do Rio de Janeiro, Rua Saõ Francisco Xavier 524, 20550-013, Rio de Janeiro, Rio de Janeiro, Brazil Calling activity of frogs is influenced by local environmental factors such as relative humidity (e.g., Cree, 1989), temperature (e.g., Licht, 1969; Pough et al., 1983; Navas, 1996), photoperiod (e.g., Jaeger et al., 1976; Hutchison and Maness, 1979; Whittier and Crews, 1987), and rainfall (e.g., Whittier and Crews, 1987; Duellman, 1995; Donnelly and Guyer, 1994). Most studies regarding calling activity have focused on nocturnally active species, whereas information on diurnally active species is scarce. This limits our knowledge of the extent to which diurnal activity is affected by a different range of environmental factors, particularly light intensity. Hylodes phyllodes is a recently described leptodactylid inhabiting the Atlantic rain forest region of southeastern Brazil (Heyer and Cocroft, 1986; Heyer et al., 1990; Rocha et al., 1997). At the Atlantic rain forest of Ilha Grande, an island located in the south of Rio de Janeiro State, Brazil, H. phyllodes is commonly found calling during the day in small streams in the forest (Rocha et al., 1997). We investigated the extent to which calling activity of the frog Hylodes phyllodes correlates with light intensity, temperature, relative humidity and photoperiod. The study was carried out from April 1997 to April 1999 in the Atlantic rain forest at Ilha Grande (23⬚11⬘S, 44⬚12⬘W), an island in southern Rio de Janeiro State, located approximately 150 km south of Rio de Janeiro City, southeastern Brazil. The forest exhibits different levels of regeneration because of disturbances caused by human activities, which ceased with the transformation of the area in to a state park (Arau´jo and Oliveira, 1988). Some remnants of primary forest (where apparently only selective wood cutting has occurred) can still be found in the most inaccessible areas of the island. Annual rainfall is about 2200 mm and mean annual temperature is about 23⬚ C (Central Nuclear de Angra, NUCLEN). The study area is characterized by a 50-year-old patch of regenerating forest located about 180 m above sea level. Calling activity of frogs was estimated monthly by counting the number of calling frogs and the number of calls within each sampling period. We established three observation points to record frog activity at the side of three small streams that were ⬎ 100 m apart. At each point we recorded the activity of H. phyllodes for 5 min at hourly intervals from 0500–1900 h. During each 5-min observation we recorded number of

1

Corresponding Author. E-mail: [email protected]

SHORTER COMMUNICATIONS active frogs (number of individuals calling at that site) and number of calls emitted by frogs using a hand counter and a chronometer. Prior to each observation we measured air temperature using a thermometer (to the nearest 0.1⬚C), air humidity (%) using a hand psychrometer and the amount of light (in Lux) reaching the ground using a luxmeter. We also obtained the rainfall and photoperiod of each observation day. Daily and monthly rainfall (in millimeter) for the region was obtained from the Central Nuclear de Angra Station (NUCLEN). Frog calling activity was expressed as mean number of calls at the sites and as mean number of active individuals for hourly and monthly intervals (Heyer et al., 1994). During November, December, and February, a period of high calling activity of H. phyllodes in the area, observations were made for 24 h to detect potential nocturnal activity by frogs. Frog calling was regressed against mean air temperature, air humidity, rainfall and light intensity recorded at the same period (Zar, 1999). The additive effect of these variables on number of active frogs and on mean number of calls per male was analyzed by multiple regression (Zar, 1999). For effect of rainfall on frog activity, we used amount of rain during the same day of sampling and the amount of rain for the month to analyze the effect of rain through the year. The monthly effect of the photoperiod on H. phyllodes activity was evaluated by analyzing the relationship between number of males calling per hour and number of calls per hour against daylength (in minutes). We also related duration (in minutes) of calling activity each day with photoperiod (in minutes) of the respective day using regression analysis. The effect of light on onset and end of H. phyllodes activity was evaluated by recording the precise time of the start and the end of frog activity and relating them to the time of sunrise and sunset of the same day, respectively. The onset and end of frog activity were considered as the time when the first and last calls were recorded, respectively. The daily calling activity of H. phyllodes usually started between 0500 and 0600 h and ended between 1800 and 1900 h; activity was most intense from 0700– 0800 h, in terms of number of calling males (Fig. 1a) and from 1000–1300 h, in terms of calls per male (Fig. 1b). A slight decrease in activity was observed between 1200 and 1300 h (Fig. 1a–b). We found no active individuals from May to July 1997 or from May to August 1998, and, in both years, the peak of frog activity occurred in January, February, and March (Fig. 2a–b). Mean intensity of light (in Lux) at the three sites in the forest peaked around 1100 h and decreased afterward (Fig. 1c). Mean light intensity in the habitat was positively correlated with mean number of calls/male (R2 ⫽ 0.78, F1,12 ⫽ 41.674, P ⬍ 0.001) but not number of frogs calling each hour (R2 ⫽ 0.02, F1,12 ⫽ 0.247, P ⫽ 0.628) (Fig. 1a–c). The relationship between number of calling H. phyllodes per hour and mean temperature was negative and significant (R2 ⫽ 0.71, F1,10 ⫽ 24.260, P ⫽ 0.001; Fig. 1a–b,d). However, number of calls per male was not significantly related to mean temperature (R2 ⫽ 0.18, F1,10 ⫽ 2.148, P ⫽ 0.173). Number of calls per male was negatively and significantly related to mean humidity in the habitat (R2 ⫽ 0.36, F1,12 ⫽ 6.690, P ⫽ 0.02), but relative humidity was not associated with number of calling individuals

315

FIG. 1. Mean number of individuals calling (⫾ 1 SE) (a) and mean number of calls per male (⫾ 1 SE) (b) of male Hylodes phyllodes recorded at the three points in the forest, light intensity (in Lux) (c) and temperature (in C) (d) in relation to time of day, from April 1997 to April 1999, at the Atlantic rain forest of Ilha Grande, Rio de Janeiro, Brazil. (R2 ⫽ 0.01, F1,12 ⫽ 0.129, P ⫽ 0.73). Rainfall did not affect activity of H. phyllodes either in terms of number of calling individuals (R2 ⫽ 0.05, F1,16 ⫽ 0.85; P ⫽ 0.37) or number of calls per male (R2 ⫽ 0.03, F1,16 ⫽ 0.57, P ⫽ 0.46). Multiple regression analysis showed an overall significant effect of environmental parameters on number of calling individuals during the day (R2 ⫽ 0.87, F3,8 ⫽ 17.330, P ⫽ 0.001). However, only temperature (P ⬍ 0.001) and humidity (P ⫽ 0.05) were significant after removing the effect of other variables. Similarly, there was an overall significant effect of environmental variables on mean number of calls per male (R2 ⫽ 0.79, F3,8 ⫽ 9.928, P ⫽ 0.005), but only light intensity was significant in influencing the number of calls/ male (P ⫽ 0.001). Across months, mean intensity of light had no significant effect on either mean number of calls per male

316


FIG. 3. Monthly time of sunrise (▫) (a) and time of start of activity of Hylodes phyllodes (●), and monthly time of sunset (䡲) (b) and time of end of activity of Hylodes phyllodes (䡩), from April 1997 to April 1999.

FIG. 2. Mean number of active individuals per hour (⫾ 1 SE) (a) and mean number of calls/male/h (⫾ 1 SE) (b) of male Hylodes phyllodes recorded at the three sites in the forest, light intensity (in Lux) (c) and temperature (in C) (d) in relation to months of year, from April 1997 to April 1999, at the Atlantic rain forest of Ilha Grande, Rio de Janeiro, Brazil. (R2 ⫽ 0.14, F1,23 ⫽ 3.606, P ⫽ 0.07) or number of calling frogs (R2 ⫽ 0.13, F1,23 ⫽ 3.602, P ⫽ 0.07; Fig. 2a–c). Onset and end of H. phyllodes activity in each month usually was coincident with sunrise and sunset, respectively (Fig. 3). Onset of calling activity was positive and significantly related to the time of sunrise (R2 ⫽ 0.89, F1,17 ⫽ 132.389, P ⬍ 0.001). Similarly, the end of frog activity was significantly related to the time of sunset (R2 ⫽ 0.88, F1,17 ⫽ 121.136, P ⬍ 0.001). Rainfall did affect H. phyllodes activity both in terms of number of calling individuals (R2 ⫽ 0.33, F1,23 ⫽ 11.194, P ⬍0.001) and number of calls per male (R2 ⫽ 0.30, F1,23 ⫽ 9.686; P ⬍0.001). Field observations indicate that H. phyllodes usually ceases or diminishes calling with the onset of rain. Usually, individuals stopped calling during strong rainfall. At Ilha Grande the photoperiod differs about 2.5 h between the longest and shortest days (Fig. 4). The relationship between day length and number of males calling per hour was positive and significant (R2 ⫽ 0.47, F1,23 ⫽ 20.202, P ⬍0.001) and so was the relation-

ship between day length and number of calls per hour (R2 ⫽ 0.46, F1,23 ⫽ 19.541, P ⬍0.001). Similarly, the extent of frog activity (in minutes) varied along the year, coincident with the corresponding photoperiod at each month (Fig. 4). The relationship between the extent of frog activity and that of photoperiod was positive and significant (R2 ⫽ 0.91, F1,17 ⫽ 164.530, P ⬍ 0.001). Number of calls per male (R2 ⫽ 0.39, F1,23 ⫽ 14.486, P ⫽ 0.001) and number of calling individuals on each month (R2 ⫽ 0.50, F1,23 ⫽ 23.185, P ⬍0.001) were positively and significantly related to mean temperature in the habitat during observations (Fig. 2a–b,d). Number of calling males (R2 ⫽ 0.01, F1,21 ⫽ 0.312, P ⫽ 0.582) and number of calls along the year (R2 ⫽ 0.01, F1,21 ⫽ 0.177, P ⫽ 0.678) were not significantly related to relative humidity. Multiple regression showed that environmental variables significantly explained frog activity in terms of calling males (R2 ⫽ 0.58, F5,17 ⫽ 4.757, P ⫽ 0.007), but no variable explained additional portion of frog activity. Similarly, the environmental variables significantly explained the number of calls per male (R2 ⫽ 0.52, F5,17 ⫽ 3.619, P ⫽ 0.021) but no variable had an additive effect on the number of calls. Absence of nocturnal activity indicates that H. phyl-

FIG. 4. Extent (in minutes) of photoperiod (-▫-) and of activity of Hylodes phyllodes (-●-), from April 1997 to April 1999, at the Atlantic rain forest of Ilha Grande, Rio de Janeiro, Brazil.

SHORTER COMMUNICATIONS lodes is a diurnal frog. Nocturnal inactivity of H. phyllodes was reported by Heyer et al. (1990) for the Atlantic rain forest of Boraceía where they found inactive individuals resting on leaves at night (we also found one inactive individual on a leaf at Ilha Grande). Exclusive diurnal activity by H. phyllodes is reinforced by the strong relationship between the times of sunrise and sunset with those of the starting and ending of frog activity, respectively. Photoperiod proved to have an important influence on activity of H. phyllodes, explaining a considerable part (91%) of its activity. Ambient light (Jaeger et al., 1976; Jaeger and Hailman, 1981) and photoperiod (Horseman et al., 1978; Floyd, 1985; Pancharatna and Patil, 1997) are important factors affecting activity of some amphibians. In some amphibian species, photoperiod has been demonstrated to significantly affect spermiogenesis (Duellman and Trueb, 1986). In H. phyllodes the circadian activity pattern seems finely regulated by photoperiod, whereas the intensity with which males call to attract females during the day is strongly affected by light intensity reaching the frog’s microhabitat. Athough there is some information regarding activity of nocturnal frogs (Duellman and Trueb, 1986), our study indicates how light may affect the activity of diurnal frogs. The environmental variables we studied affected the activity of H. phyllodes in different ways. Temperature explained number of active individuals and number of calls both during the day and througrout the year. Other studies indicate that environmental temperature usually affects frogs vocal activity (e.g., Licht, 1969; Pough et al., 1983; Navas, 1996). Light intensity seems an important factor regulating both onset and cessation of H. phyllodes calling activity and also the intensity of calls. However, light intensity was not important in explaining number of active individuals during the day or the year, because number of calling individuals in the three sampling points usually remained constant during the day. Because number of individual frogs was not affected by light intensity, but number of calls was, our data indicate that the amount of light reaching the microhabitat during the day strongly affects number of calls emitted by each frog. Around midday there was a decrease in light intensity reaching the frogs microhabitat causing (or resulting in) a decrease in frog activity, which reinforces the idea that frog activity is strongly affected by this environmental factor on a daily basis. As a result, the higher the light intensity in the microhabitat, the more intense is the calling activity of each individual. We believe that the intensity of calling activity during the year was not significantly related to light intensity because of the high variance in light conditions of the sampling days, which included sunny, cloudy and rainy days. Air humidity did not affect significantly number of active individuals and number of calls during the year nor the active individuals along the day. However, our data showed that, as air humidity increased in the microhabitat, number of calls tended to decrease. It is possible that this negative effect is related to the proximity of rain (100% air humidity indicates rain), which seemed to restrict the activity of frogs. Rainfall of the day of observation did not affect activity of H. phyllodes, although amount of rain falling in each par-

317

ticular month affected frog activity. It is known that the increase in rainfall during the year affects the annual activity of some amphibian species (e.g., Whittier and Crews, 1987; Duellman, 1995; Donnelly and Guyer, 1994). This may be a result of the increase in water in the environment, which provides more suitable opportunities for reproduction, especially for a species with aquatic tadpoles. Moisture and rainfall are important environmental factors affecting nocturnal frog activity (Cree, 1989), and in the case of the diurnal H. phyllodes they also had a significant effect. Although this frog usually lives in small streams in the forest where humidity is usually high not only because of the water of the rivulets but also because of the spray of waterfalls of its microhabitat, this environmental factor remains important for its activity. Acknowledgments.—This study is part of the results of the ‘‘Programa de Ecologia, Conservaçaõ e Manejo de Ecossistemas do Sudeste Brasileiro’’ and of the Southeastern Brazilian Vertebrate Ecology Project (Vertebrate Ecology Laboratory), both of the Setor de Ecologia, Instituto de Biologia, Universidade do Estado do Rio de Janeiro. We thank the Coordination of the CEADS/UERJ, the Direction of Campi Regionais, and the Administrative Coordination for local support and many facilities available. We also thank the SubReitoria de Po´s-Graduaçaõ e Pesquisa (SR-2/UERJ) for institutional support and for many facilities along the study. We are also grateful to D. Vrcibradic and J. P. Pombal Jr. for kindly revising the manuscript offering helpful suggestions. During the development of this study C.F.D.Rocha (Process 300 819/94-3) and M. Van Sluys (301 117/95-0) received research grants from the Conselho Nacional do Desenvolvimento Cientı´fico e Tecnolo´gico–CNPq and F. H. Hatano received a graduate fellowship from the Coordenaçaõ de Aperfeiçoamento de Pessoal de Nı´vel Superior (CAPES).

LITERATURE CITED ARAU´JO, D., AND R. OLIVEIRA. 1988. Reserva Biolo´gica Estadual da Praia do Sul (Ilha Grande, Rio de Janeiro): lista preliminar da flora. Acta botaˆnica Brasilica 1:83–94. CREE, A. 1989. Relationship between environmental conditions and nocturnal activity of the terrestrial frog, Leiopelma archeyi. Journal of Herpetology 23: 61–68. DONNELLY, M. A., AND C. GUYER. 1994. Patterns of reproduction and habitat use in an assemblage of Neotropical hylid frogs. Oecologia 98:291–302. DUELLMAN, W. E. 1995. Temporal fluctuations in abundances of anuran amphibians in a seasonal Amazonian rainforest. Journal of Herpetology 29: 13–21. DUELLMAN, W. E., AND L. TRUEB. 1986. Biology of Amphibians. McGraw-Hill, New York. FLOYD, R. B. 1985. Effects of photoperiod and starvation on the temperature tolerance of larvae of the giant toad, Bufo marinus. Copeia 1985:625–631. HEYER, W. R., AND R. B. COCROFT. 1986. Descriptions of two new species of Hylodes from the Atlantic forest of Brazil (Amphibia: Leptodactylidae). Proceedings of the Biological Society of Washington 99:100–109. HEYER, W. R., A . S. RAND, C. A . G. CRUZ, O . PEIX-

318 OTO, AND


C. E. NELSON. 1990. Frogs of Boraceía. Arquivos de Zoologia (Saõ Paulo). HEYER, R., M. DONNELLY, R. W. MCDIARMID, L. C. HAYEK, AND M. S. FOSTER. 1994. Measuring and Monitoring Biological Diversity. Standard Methods for Amphibians. Smithsonian Institution Press, Washington, DC. HORSEMAN, N. D., C. A . SMITH, AND D. D. CULLEY JR. 1978. Effects of age and photoperiod on ovary size and condition in bullfrogs (Rana catesbeiana Shaw) (Amphibia, Anura, Ranidae). Journal of Herpetology 12:287–290. HUTCHISON, V. H., AND J. D. MANESS. 1979. The role of behavior in temperature acclimation and tolerance in ectotherms. American Zoologist 19:367– 384. JAEGER, R. G., AND J. P. HAILMAN. 1981. Activity of Neotropical frog in relation to light. Biotropica 13: 59–65. JAEGER, R. G., J. P. HAILMAN, AND L. S. JAEGER. 1976. Bimodal diel activity of a Panamanian dendrobatid frog, Colostethus nubicola, in relation to light. Herpetologica 32:77–81. LICHT, L. E. 1969. Comparative breeding biology of the red-legged frog (Rana aurora aurora) and the western spotted frog (Rana pretiosa pretiosa) in southwestern British Columbia. Canadian Journal of Zoology 47:505–509. NAVAS, C. A. 1996. The effect of temperature on the vocal activity of tropical anurans: a comparison of high and low-elevation species. Journal of Herpetology 30:488–497. PANCHARATNA, K., AND M. M. PATIL. 1997. Role of temperature and photoperiod in the onset of sexual maturity in female frogs, Rana cyanophlyctis. Journal of Herpetology 31:111–114. POUGH, H. F., T. L. TAIGEN, M. M. STEWART, AND P. F. BRUSSARD. 1983. Behavioral modification of evaporative water loss by a Puerto Rican frog. Ecology 64:244–252. ROCHA, C. F. D., M. VAN SLUYS, AND F. H. HATANO. 1997. Geographic Distribution Hylodes phyllodes (leaf frog). Herpetological Review. 28:208. WHITTIER, J. M., AND D. CREWS. 1987. Seasonal reproduction: patterns and control. In D. O. Norris and R. E. Jones (eds.), Hormones and Reproduction in Fishes, Amphibians and Reptiles, pp. 385–409. Plenum Press, New York. ZAR, J. H. 1999. Biostatistical Analysis. Prentice Hall, Inc., Englewood Cliffs, NJ. Accepted: 8 September 2001.


Ecology of the Horned Leaf-Frog, Proceratophrys appendiculata (Leptodactylidae), in an insular Atlantic Rain-Forest Area of Southeastern Brazil LEONARDO BOQUIMPANI-FREITAS,1 CARLOS FREDERICO D. ROCHA,2 AND MONIQUE VAN SLUYS, Setor de Ecologia, Departamento de Biologia Animal e Vegetal, Instituto de Biologia Roberto Alcaˆntara Gomes, Universidade do Estado do Rio de Janeiro, Rua Saõ Francisco Xavier, 524, Maracana˜, 20550-019, Rio de Janeiro, Rio de Janeiro, Brazil. One of the most important factors limiting our understanding about the ecology of leptodactylid frogs of the genus Proceratophrys is the low rate at which individuals are encountered, probably a result of their low local abundances. These frogs are terrestrial and usually inhabit the litter of the floor of the Atlantic and Amazonian forests and the Cerrado‘‘ (Savannalike vegetation) of central Brazil (Heyer et al., 1990; Giaretta et al., 2000). Proceratophrys are usually cryptic (Sazima, 1978) and dispersed on the forest floor, making it difficult to detect individuals (Izecksohn and Peixoto, 1996). Some species, such as P. appendiculata of the Brazilian Atlantic rain forest, have a markedly low density of one individual per 500 m2 of forest floor (Rocha et al., 2000, 2001). As a result, knowledge of the biology for most of the species in this genus is limited. Although the genus includes at least nine species occurring in the Atlantic rain forest in Brazil (Izecksohn and Peixoto, 1981; Izecksohn et al., 1998), there is only limited information on food habits for P. boiei (Giaretta et al., 1998), calling activity for P. boiei (Bertoluci, 1998) and P. cururu (Eterovick and Sazima, 2000) and defensive behavior for P. appendiculata (Sazima, 1978). Proceratophrys appendiculata (Gu¨nther, 1873) lives on the forest floor in southeastern and southern Brazil (Izecksohn et al., 1998) where it usually is the largest species in the leaf litter frog community (Rocha et al., 2000). In this paper, we summarize data gathered over five years about aspects of the ecology (including activity, food habits, reproduction, and habitat use) of an insular population of P. appendiculata in an area of Atlantic rain forest in southeastern Brazil. The study was carried out from July 1995 to September 2000 in an area of primary Atlantic rain forest close to the village of Vila Dois Rios (23⬚11⬘S, 44⬚12⬘W), on Ilha Grande, an island in southern Rio de Janeiro State, located approximately 150 km west of Rio de Janeiro City, Brazil. Ilha Grande is within the Atlantic rain forest domain with different levels of regeneration resulting from disturbances caused by human activities over the last centuries (coffee, corn, and sugar plantations). These activities ceased in 1971 with the transformation of the area into a state park, the Parque Estadual da Ilha Grande (Arau´jo and Oliv1 Present address: Programa de Po´s-Graduaçaõ em Ecologia, Universidade Federal do Rio de Janeiro. 2 Corresponding Author. E-mail: [email protected]

SHORTER COMMUNICATIONS eira, 1988). Some remnants of primary forest still occur in the most inaccessible central areas of the island. Annual rainfall in the area is approximately 2200 mm (NUCLEN, 1996) and the mean annual temperature ranges about 23⬚C. The study area is located at approximately 240 m above sea level. Food habits of P. appendiculata were assessed based on 18 individuals collected between September 1996 and June 2000. Individuals were sampled either from random walks in the forest, or from litter floor plots. Frogs were sexed, measured (snout–urostyle length, SUL; and mouth width; to the nearest 0.1 mm) using a Vernier caliper, and weighed (to the nearest 1.0 mg) on an electronic balance. Specimens were dissected, stomachs removed and their contents analyzed. Stomach contents items were counted, identified to order, and measured for its length, width, and height (to the nearest 0.1 mm), using a Vernier caliper. The volume (in mm3) of each prey was estimated by multiplying these three dimensions (Schoener, 1967). The diet composition was qualitatively (prey types) and quantitatively (by number and volume) analyzed. We used regression analysis between the length of the largest prey (in millimeters) and mouth width and frog SUL to evaluate the relationship between frog size and the size of prey consumed. Sexual differences in mouth width were looked for using Analysis of Covariance with SUL as covariate (Zar, 1999). Differences in mean prey size ingested by males and females were assessed using the t-test (Zar, 1999). Microhabitat use was studied by recording the microhabitat of each individual sampled (N ⫽ 18) at its original position into seven microhabitat categories: (1) on forest floor leaf litter; (2) on uncovered forest floor; (3) under rocks at the edge of shallow rivulets; (4) partially submerged in rivulet water; (5) under/on rocks on the ground; (6) under tree roots; (7) under fallen trunks. In addition, we recorded the presence/ absence of the same microhabitat categories (microhabitats available) at 20 locations taken at random, in a 1 m radius from each location. Activity of P. appendiculata was studied during the calling (reproductive) period of the species for the area (see below). For activity estimates, we established two sites in the forest close to streams, separated from one another by approximately 400 m. At these sites, we recorded the monthly occurrence of calling males in the morning (0800–1000 h) and in the afternoon (1300–1400 h) from September 1998 to September 2000 to evaluate whether there were any active individual P. appendiculata. Activity was measured by recording the number of calling frogs and number of calls during 24-h periods, one each month when there were active individuals of P. appendiculata. At these sites, the number of calls were recorded during 15-min periods at hourly intervals using a handcounter. Activity was expressed by the mean number of active individuals and the mean number of calls at each hourly interval during the 24 h-period. In addition, prior to each observation, we recorded air temperature (to the nearest 0.1⬚C) using a thermometer, relative humidity (in %) using a hygrometer, and light intensity (to the nearest lux) using a luxmeter. The relationship between frog calling activity and environmental variables recorded throughout each day was tested by regression analysis (Zar, 1999).

319

TABLE 1. Number, volume (mm3), and frequency of types of prey found in the stomachs of Proceratophrys appendiculata (N ⫽ 18) in the Atlantic rain forest of the Ilha Grande, Rio de Janeiro State, Brazil. Numbers in parenthesis are percentages.

Item

Insecta Orthoptera Blattaria Coleoptera

Number (%)

Volume (%)

Frequency (%)

11 (61.1) 2 (11.2) 1 (5.6)

17432.9 (87.4) 1459.1 (7.3) 244.0 (1.2)

44.4 11.2 5.6

Arachnida Araneae

1 (5.6)

40.0 (0.2)

5.6

Gastropoda Stylommatophora

2 (11.2)

92.0 (0.4)

11.2

1 (5.6)

677.2 (3.4)

5.6

18 (100)

19945.2 (99.9)

Amphibia Anura Total

Reproductive characteristics were studied in six adult females. For each gravid female, we recorded SUL, number of eggs in each ovary, diameter (to the nearest 0.1 mm), mass (to the nearest 1.0 mg) of eggs (both based on 10 eggs from each ovary), and total mass of eggs from both ovaries. Reproductive effort was estimated as the mean value (arcsine transformed) of the total wet mass of eggs of each female divided by its body plus wet mass of eggs (relative clutch mass, RCM). The mean SUL of adult females (x¯ ⫽ 66.7 ⫾ 4.3 mm; range 60.2–72.2 mm; N ⫽ 6) was significantly larger than that of adult males (x¯ ⫽ 54.6 ⫾ 7.1 mm; range 49.4–62.1 mm; N ⫽ 10; t-test, t ⫽ ⫺3.760; P ⫽ 0.002). Similarly, mouth width of females was significantly larger than that of males (ANCOVA, F ⫽ 11.297, P ⫽ 0.006). The diet of Proceratophrys appendiculata was composed of six prey types (Table 1). Orthopterans were the most important item numerically (61.1%), volumetrically (87.4%), and also the most frequent (44.4% of the frogs analyzed, Table 1). The other items were consumed in relatively low proportions (Table 1). In one of the stomachs, we found one adult of an unidentified species of syntopic leaf litter leptodactylid of the genus Eleutherodactylus. There was a significant relationship between the size of the largest prey ingested with frog SUL (F ⫽ 6.250; R2 ⫽ 0.385; P ⫽ 0.031) and with frog mouth width (F ⫽ 6.115; R2 ⫽ 0.379; P ⫽ 0.033). Mean prey size (⫾ 1 SD) consumed by females (24.3 ⫾ 2.1 mm) was significantly larger than that of males (13.8 ⫾ 7.2 mm; t ⫽ 2.836; P ⫽ 0.018; N ⫽ 12). Proceratophrys appendiculata used three types of microhabitats among those available, with the forest floor leaf litter being the most frequent (55.6 %; Fig. 1). The other two microhabitat categories used were under rocks at the edge of shallow rivulets (33.3%) and in shallow rivulets partially submerged in water (11.1

320


FIG. 1. Proportion (%) of microhabitats used (open bars) and microhabitats available (black bars) for Proceratophrys appendiculata in the Atlantic rain forest of Ilha Grande, Rio de Janeiro State, Brazil, according to the following categories: LF, on forest floor leaf litter; UF, on uncovered forest floor; RR, under rocks at the edge of shallow rivulets; PS, partially submerged in rivulet water; RG, under/on rocks on the ground; TR, under tree roots; FT, under fallen trunks. %). The frequency of microhabitats used by the frog differed from that of available microhabitats (Fig. 1). At Ilha Grande, calling activity of P. appendiculata only occurred in September 1998 and during June, July and August 1999. During these turns, frogs remained close to shallow rivulets in the forest calling intensively both during the day and at night. During this period, individual frogs called continuously. Most of the calling activity was concentrated between 1000 and 1300 h and between 1900 and 2000 h (Fig. 2). The number of calls during the day was not significantly related to air temperature (F ⫽ 0.878; R2 ⫽ 0.038; P ⫽ 0.359) nor to relative humidity (F ⫽ 0.016; R2 ⫽ 0.001; P ⫽ 0.901) nor to light intensity (F ⫽ 2.006; R2 ⫽ 0.084; P ⫽ 0.171). However, the number of active individuals was significantly related to both air temperature (F ⫽ 14.163; R2 ⫽ 0.392; P ⫽ 0.001) and to light intensity (F ⫽ 8.510; R2 ⫽ 0.279; P ⫽ 0.008) but not to relative humidity (F ⫽ 0.012; R2 ⫽ 0.001; P ⫽ 0.912). Three of the six adult females sampled had mature eggs; the mean number of eggs per female was 841.7 ⫾ 108.7 (N ⫽ 3; range 729–946), the mean diameter of the eggs was 1.68 ⫾ 0.28 mm (N ⫽ 60 eggs) and the mean mass of eggs was 0.0041 ⫾ 0.0021g (N ⫽ 60 eggs). The estimated reproductive effort was 8.1 ⫾ 4.2% (N ⫽ 3; range 5.0–12.9). The diet of P. appendiculata at Ilha Grande is composed mainly of orthopterans. For the only other species in this genus for which there is some information about the diet (Proceratophrys boiei), Giaretta et al. (1998) found that its diet at Atibaia, in Saõ Paulo State, was composed predominantly of orthopterans and coleopterans. In general, frogs of the genus Proceratophrys are among the largest species in the leaf litter frog community in the Atlantic rain forest areas (Giaretta et al., 1997; Bernarde and Anjos, 1999; Rocha et al., 2001). Also, these frogs have a sedentary foraging strategy (Giaretta et al., 1998). For such a large and sedentary species, having a comparatively large body mass, it might be advantageous to concentrate its diet on a large, mobile prey, if they are relatively abundant. Orthopterans may fit this contention because they are large arthropods of the leaf litter fauna and relatively abundant (Santos et al., 1998) and highly mobile. The finding of an individual of the genus Eleuther-

FIG. 2. Number of calls (bars) and number of active individuals (line) of Proceratophrys appendiculata in the Atlantic rain forest of Ilha Grande, Rio de Janeiro State, Brazil.

odactylus in the stomach of one P. appendiculata indicates that not only is the latter a predator of the former but also that P. appendiculata may constitute a source of mortality for some smaller syntopic litter frog species on Ilha Grande. Giaretta et al. (1998) also found two adult (15–20 mm SUL) Eleutherodactylus parvus in the diet of P. boiei from a forested area in Atibaia, in the Mantiqueira range, Saõ Paulo State. This observation together with our data from Ilha Grande suggests that Proceratophrys frogs may be relatively common predators of small Eleutherodactylus in leaf litter frog communities. Our data show that frog size (in terms of SUL and mouth width) significantly affects the size of the largest prey ingested and that females tend to consume larger prey than males. This probably results from size differences between the sexes in this species, with females being larger. The microhabitat most frequently used by P. appendiculata was the leaf litter of the forest floor (55.6%). This microhabitat use is consistent with the leaflike appearance of P. appendiculata, which also has a cryptic defensive behavior (Sazima, 1978) whose efficiency depends on a life on the leaf litter. In addition, our observations suggest that during the short reproductive period, individuals tend to remain near shallow rivulets where males call actively throughout day and evening, and where females lay a large number of eggs. The only species of Proceratophrys of which there is any information about calling activity are P. avelinoi (Bernarde and Anjos, 1999) and P. cururu at Serra do Cipo´ in Minas Gerais State (Eterovick and Sazima, 2000). Bernarde and Anjos (1999) found that at Mata dos Godoy, in Londrina, Parana´ State, individuals of P. avelinoi called from July to April. However, calling activity of P. cururu was restricted to only two months of the year (November and December), during the wet season (Eterovick and Sazima, 2000). Our data for 1998 and 1999 show that P. appendiculata only called from 1–3 months during the year which indicates a similar extent of calling activity for P. appendiculata and Proceratophrys cururu. In addition, until September 2000 no calling activity of P. appendiculata was recorded for that year. The observation that during the few months when P. appendiculata calls, individual frogs call throughout the 24-h period, is consistent with an idea of large investment in calling and may be a consequence of the short calling period of the species throughout the year.

SHORTER COMMUNICATIONS The number of calls during the day was not related to the environmental variables studied (air temperature, relative humidity, or light intensity), but the number of active calling individuals was significantly related to air temperature and to light intensity. Air temperature is an environmental variable which affects some frog species (Duellman and Trueb, 1986), and at Ilha Grande it seems to determine to which extent a male P. appendiculata will be active. Our data also showed that the increase in light intensity results in an increase of individuals calling even though frogs call throughout night and day. Although there is no previous information regarding the effect of environmental factors on activity of Proceratophrys, our data suggest that P. appendiculata calling activity is significantly affected by light, resulting in males calling more intensively during the day. The fact that air humidity did not affect the species calling activity is not surprising because males call at the interior or along the sides of rivulets and small waterfalls, microhabitats under water spray and, thus, hypersaturated by humidity. The only other study providing some information on the activity of P. appendiculata was that of Rocha et al. (2000), in which active individuals were detected diurnally. Our data on male calling activity indicate that, at least during the reproductive period, individuals are active throughout day and night, which is similar to the study of Bernarde and Anjos (1999). However, males of P. cururu called only at night at open areas at Serra do Cipo´, Minas Gerais State, which is a drier habitat where frogs may be more susceptible to water loss (P. C. Eterovick, pers. comm.). This reinforces our hypothesis that the diurnal activity of P. appendiculata is made possible by the high humidity of its microhabitat in the Atlantic rain forest. Acknowledgments.—This study is part of the results of the Programa de Ecologia, Conservaçaõ e Manejo de Ecossistemas do Sudeste Brasileiro and of the Southeastern Brazilian Vertebrate Ecology Project (Vertebrate Ecology Laboratory), both of the Setor de Ecologia, Instituto de Biologia, Universidade do Estado do Rio de Janeiro. This study is part of the master’s thesis of the first author at de Programa de Po´sGraduaçaõ em Ecologia of the Universidade Federal do Rio de Janeiro. We thank M. B. Vecchi, F. H. Hatano, R. V. Marra, D. Vrcibradic, M. A. S. Alves, and H. G. Bergallo for field and laboratory assistance. We also thank E. R. Wild, P. C. Eterovick, C. Gascon, and an anonymous referee for their revision and valuable comments. We thank the Coordination of the CEADS/ UERJ, the Director of Campi Regionais and to the Administrative Coordination for local support and for making facilities available. We also thank the Sub-Reitoria de Po´s-Graduaçaõ e Pesquisa (SR-2/UERJ) for institutional support and for many facilities along the study. During the development of this study LBF received a grant from the Coordenaçaõ de Aperfeiçoamento de Pessoal de Nı´vel Superior–Capes. CFDR and MVS received Research Grant of the Conselho Nacional do Desenvolvimento Cientı´fico e Tecnolo´gico-CNPq (Grants 300 819/94-3 and 301117/95-0, respectively).

LITERATURE CITED ARAU´JO, D. S. D., AND R. R. OLIVEIRA. 1988. Reserva Biolo´gica da Praia do Sul (Ilha Grande, Estado do

321

Rio de Janeiro): lista preliminar da Flora. Acta Botaˆnica Brasilica 1:83–94. BERNARDE, P. S., AND L. ANJOS. 1999. Distribuiçaõ espacial e temporal da anurofauna no Parque Estadual Mata dos Godoy, Londrina, Parana´, Brasil (Amphibia: Anura). Comunicaçes do Museu de Cieˆncia e Tecnologia PUCRS. Se´rie Zoolo´gica Porto Alegre 12:127–140. BERTOLUCI, J. 1998. Annual patterns of breeding activity in Atlantic rainforest anurans. Journal of Herpetology 4:607–611. DUELLMAN, W. E., AND L. TRUEB. 1986. Biology of Amphibians. McGraw-Hill, New York. ETEROVICK, P. C., AND I. SAZIMA. 2000. Structure of an anuran community in a montane meadow in southeastern Brazil: effects of seasonality, habitat, and predation. Amphibia-Reptilia 21:439–461. GIARETTA, A. A., R. J. SAWAYA, G. MACHADO, M. S. ARAU´JO, K. G. FACURE, H. F. MEDEIROS, AND R. NUNES. 1997. Diversity and abundance of litter frogs at altitudinal sites at Serra do Japi, Southeastern Brazil. Revista Brasileira de Zoologia 14: 341–346. GIARETTA, A. A., M. S. ARAU´JO, H. F. MEDEIROS, AND K. G. FACURE. 1998. Food habits and ontogenetic diet shifts of the litter dwelling frog Proceratophrys boiei (Wied). Revista Brasileira de Zoologia 15:385– 388. GIARETTA, A. A., P. S. BERNARDE, AND M. N. C. KOKUBUM. 2000. A new species of Proceratophrys (Anura: Leptodactylidae) from the Amazon rain forest. Journal of Herpetology 34:173–178. HEYER, R., A. S. RAND, C. A. G. CRUZ, O. PEIXOTO, AND C. E. NELSON. 1990. Frogs of Borace ´ ia. Arquivos de Zoologia do Museu de Zoologia da Universidade de Saõ Paulo, Saõ Paulo, Brazil. IZECKSOHN, E., AND O. L. PEIXOTO. 1981. Nova espećie de Proceratophrys da Hileía bahiana (Amphibia; Anura; Leptodactylidae). Revista Brasileira de Biologia 41:19–24. . 1996. Uma grande concentraçaõ de indivı´duos de Proceratophrys laticeps (Amphibia; Anura; Leptodactylidae). Revista da Universidade Federal Rural do Rio de Janeiro, Se´rie Cieˆncia e Vida 18: 105–107. IZECKSOHN, E., C. A. G. CRUZ, AND O. L. PEIXOTO. 1998. Sobre Proceratophrys appendiculata e algumas espećies afins (Amphibia; Anura; Leptodactylidae). Revista da Universidade Federal Rural do Rio de Janeiro, Se´rie Cieˆncia e Vida 20:37–54. NUCLEN. 1996. Central Nuclear de Angra, Estaçaõ Metereolo´gica, Angra dos Reis, Rio de Janeiro. ROCHA, C. F. D., M. VAN SLUYS, M. A. S. ALVES, H. G. BERGALLO, AND D. VRCIBRADIC. 2000. Activity of leaf-litter frogs: when should frogs be sampled? Journal of Herpetology 34:285–287. . 2001. Estimates of forest floor litter frog communities: a comparison of two methods. Austral Ecology 26:14–21. SANTOS, H. C., C. F. D. ROCHA, AND H. G. BERGALLO. 1998. A produtividade, diversidade e abundaˆncia da mesofauna de litter em dois segmentos da Mata Atlaˆntica ( Mata de Planıćie e Mata de Encosta) na Ilha do Cardoso, Cananeía, Saõ Paulo. Anais do VIII Semina´rio Regional de Ecologia. V. VIII, 823– 836.

322


SAZIMA, I. 1978. Convergent defensive behavior of two leaf-litter frogs of Southeastern Brazil. Biotropica 10:158. SCHOENER, T. W. 1967. The ecological significance of sexual dimorphism in size in the lizard Anolis conspersus. Science 155:474–477. ZAR, J. H. 1999. Biostatistical Analysis. 4th ed., Prentice Hall, Inc., Upper Saddle River, NJ. Accepted: 8 September 2001.


Size-Specific Differences in Tail Loss and Escape Behavior in Liolaemus nigromaculatus DOUGLAS A. KELT,1,2 L. KARINA NABORS,2 AND MATTHEW L. FORISTER2, Department of Wildlife, Fish, & Conservation Biology, University of California, Davis, California 95616, USA Facultative tail loss (autotomy) is a common and widespread antipredation adaptation in lizards (Arnold, 1988; Zani, 1996). Although offering an escape from predators, tail loss has been demonstrated to influence many life-history traits and to have fitnessrelated ecological and behavioral correlates (e.g., Martıń and Avery, 1998; Wilson and Booth, 1998; Martıń and Lopez, 2000, and references therein). Animals with autotomized tails also may have reduced abilities to escape predators (Dial and Fitzpatrick, 1984). Costs and benefits of tail autotomy have been discussed for many taxa, but these studies encompass only a small percentage of species employing this tactic, and further studies are critical to evaluate the generality of these observations. Theoretical models of escape behavior predict that animals will adjust the distance at which they flee from an approaching predator (‘‘fleeing distance’’) according to the relative costs of staying put versus the costs of fleeing (Ydenberg and Dill, 1986). When the former cost exceeds the latter, over any distance, the animal should seek refuge. A number of factors are known to influence the decision to stay put, including both intrinsic (e.g., age) and extrinsic factors such as predator characteristics, ambient temperature, distance and quality of available refuges, etc. (e.g., Rocha and Bergallo, 1990; Martıń and Lo´pez, 1995, 2000; Cooper, 1997). Costs associated with fleeing are predominantly missed opportunity costs associated with abandoning a food resource, but for ectotherms may also include metabolic costs such as reduced time spent basking in the sun. Although several authors have evaluated the role of tail autotomy on escape behavior (e.g., Vitt and Cooper, 1986; Formanowicz et al., 1 Corresponding Author. E-mail: dakelt@ucdavis .edu 2 Graduate Group in Ecology, University of California, Davis, California 95616, USA

1990; Smith, 1996), none have simultaneously evaluated animal age and tail condition as these relate to escape behavior, and most have employed the controlled conditions of a laboratory setting. As part of a larger study on the ecology of arid ecosystems in northern Chile, we made preliminary observations on the frequency of tail autotomy and responses to potential predator threat in Liolaemus nigromaculatus (Tropiduridae, Liolaeminae). We studied only males of this species, which are characterized by a solid black head, and a dorsum and sides speckled yellow-green to vibrant lime-green over a grayishblack base color (Donoso-Barros, 1966). The genus Liolaemus is diverse, with over 160 species broadly distributed throughout temperate South America (Schulte et al., 2000). Work elsewhere in Chile (summarized by Jaksic, 1997) has demonstrated differing patterns of tail autotomy that correlate with energetic needs rather than predation pressure (Jaksic and Fuentes, 1980), primarily because the tail of some species is an important storage organ (see Medel, 1992). Additionally, low frequencies of tail autotomy may be associated either with low or high predation pressure (Jaksic and Fuentes, 1980; Nun˜ez and Yan˜ez, 1984). Autotomy is an important predator defense, however (see Arnold, 1984, 1988), and it seems reasonable to assume that some proportion of a population of autotomized lizards has been subjected to failed predation attempts. If lizards ‘‘learn’’ from previous encounters, then we might expect autotomized individuals to be more sensitive than tailed individuals to an approaching predator. On the other hand, work elsewhere (Dial and Fitzpatrick, 1984; Formanowicz et al., 1990; Martıń and Avery, 1998) has suggested that autotomized individuals are slower to respond to predation threat, either because they have fewer energy reserves to draw upon or because tailless animals cannot move as swiftly (or deftly) as tailed animals. We designed our study to compare the response of normal (‘‘tailed’’) and autotomized lizards to the approach of a potential predator (humans) in a natural setting in northern Chile. Because it may be difficult to predict a priori whether autotomized individuals should respond more or less rapidly to predation pressure, we hypothesized simply that these two groups should respond differently to predation threat. We also estimated body size of animals (small, medium, large) as a surrogate for relative ages. We conducted our study at Parque Nacional Bosque Fray Jorge, located about 450 km north of Santiago, on the coast of northern Chile (30⬚41⬘N, 71⬚40⬘W). Covering just less than 10,000 ha, this park constitutes the largest patch of intact thorn scrub habitat in northern Chile, classified as coastal and forested steppe chaparral (Gajardo, 1993). Emerging from the chaparral is a range of hills that supports a series of relict patches of temperate rain forest and dense shrubs. One road traverses the ridge of these hills to provide access to these forests by tourists. On five days in the Austral summers of 2000 and 2001, we walked along this road slowly, watching for lizards that bask in the sun at the edge of the road. When a lizard was spotted, one of us would steadily approach the animal until it began to flee. We recorded the distance at which the animal began to move (‘‘fleeing distance’’), the status of the tail (intact vs autotomized, using binoculars where


323

TABLE 1. Results of a chi-square contingency table analysis on the distribution of tailed and autotomized lizards (Liolaemus nigromaculatus) from Fray Jorge, Chile (␹2 ⫽ 1.56, P ⬎ 0.10).

Tailed Autotomized Proportion Sum

Small

Medium

Large

Sum

31 11 (0.35)

50 22 (0.44)

8 11 (0.73)

92 41

42

72

19

133

needed), and we estimated size as small (SVL ⬍ approximately 11 cm), medium (SVL approximately 11– 15 cm), or large (SVL ⬎ 15 cm). Because some animals may have been observed more than once we interpret our results cautiously. Additional sources of bias include variations in environmental conditions (e.g., ambient temperature, wind speed, time of day), and the number of people walking on the road (generally one, but occasionally three; most tourists simply drove slowly on the road, and we only saw two tourists on foot during our study periods). Because we made observations during the middle of the day and only on sunny, fog-free days, we believe that we limited these biases. Any remaining bias should serve principally to increase background noise in our dataset, such that significant results should be relatively robust. Some animals labeled as having intact tails may have regrown their tails from previous losses; we were unable to quantify this variable, and so our comparisons focus on individuals who have recently lost their tails, versus individuals with complete or regrown tails. The distribution of observations across size and tail classes was compared with a chi-square test of proportions (Zar, 1999). We analyzed fleeing distance across all size classes and tail-loss categories (intact vs autotomized) with two-way analysis of variance (ANOVA). We tested for differences within each size class with simple t-tests (Zar 1999). Although there was a tendency toward greater tail loss in larger size classes (Table 1), this was not significant (␹2 ⫽ 1.56, df ⫽ 2, P ⬎ 0.10), in contrast to work on North American skinks (Eumeces; Vitt and Cooper, 1986). Vitt and Cooper (1986) suggested that the positive relationship between size class and the percentage of individuals autotomized reflected the longer time at risk of older (larger) animals. By contrast, Jaksic and Fuentes (1980) noted that body size did not correlate with frequency of tail loss across 12 species of Liolaemus in Chile; our study adds a 13th species to this list. Fleeing distance did not differ across size classes or between animals with intact vs. autotomized tails (body size, df ⫽ 2, MS ⫽ 15.53, F ⫽ 1.80, P ⬎ 0.17; tail condition, df ⫽ 1, MS ⫽ 21.76, F ⫽ 2.52, P ⬎ 0.10), but the interaction between these factors was significant (df ⫽ 2, MS ⫽ 29.08, F ⫽ 3.36, P ⬍ 0.05; Fig. 1). Individual t-tests within size classes document that both small- and medium-sized animals exhibited similar fleeing distances for both tail classes but that large autotomized animals were significantly more tolerant of our approach than were large animals with intact tails.

FIG. 1. Bar diagram showing the mean (⫾ 1 SD) ‘‘Fleeing distance’’ for tailed and autotomized lizards (Liolaemus nigromaculatus) in three size classes. Statistics are t-tests assuming equal variances for medium sized animals and assuming unequal variances for small and large size classes.

Formanowicz et al. (1990) also reported that Scincilla lateralis with intact tails responded to a model predator at significantly greater distances than did autotomized animals, although they did not evaluate the role of age or body size. These results were somewhat unexpected because these animals rely on their tail for locomotion and are susceptible to snake predation (Dial and Fitzpatrick, 1984). Formanowicz et al. (1990) noted that the tail is an important storage organ in this species and that autotomized animals may have greater incentive to conserve remaining resources, thereby remaining cryptic when approached by a predator. Additionally, they argued that, if the tail is important in locomotion, tailless animals might compensate for reduced running speed by adopting cryptic behavior. Thus, autotomized individuals may tolerate greater proximity by predators either to reduce further energy loss or because of a reduced probability of escape. Within Liolaemus, the caloric value of lipids in the tail is known to differ; Medel (1992) reported that the energetic content of the tail was greater in a high elevation species (Liolaemus altissimus) than in a species at lower elevations (Liolaemus fuscus). In the absence of caloric values for tail lipids in L. nigromaculatus, the low elevation of our study provides little support for a strategy of excessive energy storage in the tail. Although males of this species are brightly colored (Donoso-Barros, 1966; Fuentes and Cancino, 1979), the speckled pattern on these animals may be cryptic to some predators. Further investigation into the response of this species to different predator types should be informative in this regard. We know of no other studies that simultaneously evaluated both the frequency of tail autotomy and fleeing distance in lizard species while controlling for body size. The size-specific differences are intriguing, and we can think of three potential explanations. First, older animals may learn to respond differently to predators than do young animals. The only way to evaluate this would be to survey marked individuals over time, either in the field or in controlled laboratory conditions. In the absence of such data this explanation remains speculative.

324


Second, animals of different body sizes may be subjected to different types of predators. If predators on small individuals were swifter or more mobile than those predating large individuals then larger animals would be expected to respond less rapidly to a potential predator. The habitat at our study site is densely vegetated with low shrubs, a habitat in which snakes might have a substantial advantage over mammalian or avian predators. Although both of the snake species in the park are known to prey on lizards (Jaksic et al., 1981), their ability to capture and consume large L. nigromaculatus is not clear, although only the largest individuals of these snakes consumed small mammals and only occasionally (Jaksic et al., 1981). It would appear likely, therefore, that large L. nigromaculatus might also be at the limits of the prey spectrum for these snakes. Finally, the consequences of tail autotomy for animal mobility or energetic stores may be proportionally greater for large individuals than for smaller body size classes. Above a certain body size, this species appears to add mass without increasing in length; that is, it becomes bulkier and somewhat less streamlined. Thus, each body size class may be responding behaviorally to maximize its chances of escaping a potential predation event. In this scenario, predatortolerance may be the better tactic for large animals, whereas predator-avoidance (greater fleeing distance) might be more suitable for smaller species. To our knowledge, all other studies on predator responses by autotomized animals have been conducted in laboratory settings where ancillary variables could be controlled. We studied animals subjected to a background of white noise (e.g., variations in temperature, wind, etc.) that we could not control. Although we were able to control some of these factors to some degree (e.g., we sampled at the same time of day on days without fog or heavy wind), the remaining noise likely reduces our ability to perceive underlying relationships, so that the significance of our results therefore should be rather robust. Liolaemus are distributed across 45⬚ of latitude and most habitats in temperate South America from sea level to more than 5000 m in elevation, and include a very large number of species (Schulte et al., 2000). Although the taxonomy of the group is in dire need of revision (see Etheridge, 1995; Schulte et al., 2000), this genus provides an ideal group to compare against studies on the diverse skinks (Eumeces) and iguanids (especially Sceloporus) in North America. With growing resolution of the phylogenetic relations within this genus the stage is set for ecological studies on one of the most ecologically diverse reptile taxa in South America. Acknowledgments.—We thank F. Jaksic and an anonymous reviewer for insightful comments that helped to improve the manuscript, and the Corporacioń Nacional Forestal (CONAF) for permission to conduct research at Parque Nacional Bosque Fray Jorge. This work was supported by the National Science Foundation (DEB 9873708 to P. L. Meserve and D. A. Kelt) and by FONDECYT (1970576 to J. Gutierrez).

LITERATURE CITED ARNOLD, E. N. 1984. Evolutionary aspects of tail shedding in lizards and their relatives. Journal of Natural History 18:127–169.

. 1988. Caudal autotomy as a defense. In C. Gans and R. B. Huey (eds.), Biology of the Reptilia. Vol. 16, pp. 233–273. Alan R. Liss, Inc., New York. COOPER JR., W. E. 1997. Escape by a refuging prey, the broad-headed skink (Eumeces laticeps). Canadian Journal of Zoology 75:943–947. DIAL, B. E., AND L. C. FITZPATRICK. 1984. Predator escape success in tailed versus tailless Scincella lateralis (Sauria: Scincidae). Animal Behaviour 32: 301–303. DONOSO-BARROS, R. 1966. Reptiles de Chile. Ediciones de la Universidad de Chile, Santiago. ETHERIDGE, R. 1995. Redescription of Ctenoblepharys adspersa Tschudi, 1845, and the taxonomy of Liolaeminae (Reptilia: Squamata: Tropiduridae). American Museum Novitates 3142:1–34. FORMANOWICZ JR., D. R., E. D. BRODIE JR., AND P. J. BRADLEY. 1990. Behavioural compensation for tail loss in the ground skink, Scincella lateralis. Animal Behaviour 40:782–784. FUENTES, E. R., AND J. CANCINO. 1979. Rock-ground patchiness in a simple Liolaemus lizard community (Reptilia, Lacertilia, Iguanidae). Journal of Herpetology 13:343–350. GAJARDO, R. 1993. La vegetacioń natural de Chile. Clasificacioń y distribucioń geogra´fica. Editorial Universitaria, Santiago, Chile. JAKSIC, F. 1997. Ecologıá de los vertebrados de Chile. Ediciones Universidad Cato´lica de Chile, Santiago, Chile. JAKSIC, F., AND E. R. FUENTES. 1980. Correlates of tail losses in twelve species of Liolaemus lizards. Journal of Herpetology 14:137–141. JAKSIC, F., H. W. GREENE, AND J. L. YAN˜EZ. 1981. The guild structure of a community of predatory vertebrates in central Chile. Oecologia 49:21–28. MART´ıN, J., AND R. A. AVERY. 1998. Effects of tail loss on the movement patterns of the lizard, Psammodromus algirus. Functional Ecology 12:794–802. MART´ıN, J., AND P. LO´PEZ. 1995. Escape behaviour of juvenile Psammodromus algirus lizards: constrain of or compensation for limitations in body size? Behaviour 132:181–192. . 2000. Fleeing to unsafe refuges: effects of conspicuousness and refuge safety on the escape decisions of the lizard Psammodromus algirus. Canadian Journal of Zoology 78:265–270. MEDEL, R. G. 1992. Costs and benefits of tail loss: assessing economy of autotomy in two lizard species of central Chile. Revista Chilena de Historia Natural 65:357–361. NUN˜EZ, H., AND J. L. YAN˜EZ. 1984. Colas de lagartijas de geńero Liolaemus: autotomıá e influencia en la predacioń. Studies on Neotropical Fauna and Environment 19:1–8. ROCHA, C. F. D., AND H. G. BERGALLO. 1990. Thermal biology and flight distance of Tropidurus oreadicus (Sauria, Iguanidae) in an area of Amazonian Brazil. Ethology, Ecology, and Evolution 2:263–268. SCHULTE II, J. A., J. R. MACEY, R. E. ESPINOZA, AND A. LARSON. 2000. Phylogenetic relationships in the iguanid lizard Liolaemus: multiple origins of viviparous reproduction and evidence for recurring Andean vicariance and dispersal. Biological Journal of the Linnean Society 69:75–102. SMITH, G. R. 1996. Correlates of approach distance in

SHORTER COMMUNICATIONS the striped plateau lizard (Sceloporus virgatus). Herpetological Journal 6:56–58. VITT, L. J., AND W. E. COOPER JR. 1986. Tail loss, tail color, and predator escape in Eumeces (Lacertilia: Scincidae): age-specific differences in costs and benefits. Canadian Journal of Zoology 64:583–592. WILSON, R. S., AND D. T. BOOTH. 1998. Effect of tail loss on reproductive output and its ecological significance in the skink Eulamprus quoylii. Journal of Herpetology 32:128–131. YDENBERG, R. C., AND L. M. DILL. 1986. The economics of fleeing from predators. Advances in the Study of Behavior 16:229–249. ZANI, P. 1996. Patterns of caudal-autotomy evolution in lizards. Journal of Zoology (London) 240:201– 220. ZAR, J. H. 1999. Biostatistical Analysis. 4th ed. Prentice-Hall, Englewood Cliffs, NJ. Accepted: 8 September 2001.


The Identity of Hyla ehrhardti Mu¨ller, 1924 (Anura, Hylidae) JULIAN FAIVOVICH,1 Herpetology, Division of Vertebrate Zoology, American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024, USA; and CERC/Columbia University; E-mail: [email protected] CARLOS A. G. DA CRUZ, Museu Nacional, Universidade Federal do Rio de Janeiro, Quinta da Boa Vista, 20940-040, Rio de Janeiro, Rio de Janeiro, Brazil; E-mail: cagcruz@ uol.com.br OSWALDO L. PEIXOTO, Departamento de Biologia Animal, Instituto de Biologia, Universidade Federal Rural do Rio de Janeiro, 23851-970 Serope´dica, Rio de Janeiro, Brazil; Email: [email protected] The Neotropical genus Scinax is composed of approximately 84 recognized species (Frost, 2000) distributed from Mexico to eastern Argentina. Most of the species were originally described as Hyla, and successive revisions by various authors (e.g., Fouquette and Delahoussaye, 1977; Duellman and Wiens, 1992) resulted in their present association with Scinax. Most of these generic assignments were made after examination of specimens, but Scinax ehrhardti (Mu¨ller, 1924) was reassigned in the absence of direct examination of specimens. Hyla ehrhardti was originally described from a single male collected in Humboldt (Basin of the Rio Novo), State of Santa Catarina, Brazil, and deposited in the Zoologisches Staatssammlung, Munich, Germany (ZSM). Until 1973, only Bokermann (1966) mentioned this species (though misspelling it as Hyla erhardti) and gave the current name of the type locality as Corupa´. Lutz (1973) included H. ehrhardti as a low1

Corresponding Author.

325

TABLE 1. Measurements (in millimeters) of the holotypes of Hyla ehrhardti and Hyla arianae (both adult males).

SVL Head length Head width Left tympanic diameter Left eye diameter Left eye–nostril distance Internostril distance Interocular distance Tibia length

Holotype Hyla ehrhardti

Holotype Hyla arianae

33.7 12.98 10.38 2.21 2.92 3.66 2.46 4.27 17.27

35.01 12.13 11.21 2.29 3.06 3.48 2.25 4.11 17.17

land species allied to the (then) Hyla catharinae group, together with a translation of the original description, and stated in error (see below) that ‘‘the types’’ were destroyed during WWII. She also stated (p. 198) ‘‘I have not been able to identify this species, though I may have seen specimens of it abroad,’’ which indicates that, regardless of whether or not she examined the holotype, it was not required for her conclusion that H. ehrhardti belonged to the H. catharinae group. Hyla ehrhardti subsequently was associated with the (then) Hyla rubra group by most successive authors (the H. rubra group historically comprised two large groups, the H. rubra and the H. catharinae groups). When Fouquette and Delahoussaye (1977) transferred the H. rubra group (sensu lato) to Ololygon, they also transferred H. ehrhardti, without associating it with a species group. This was followed without comment by Harding (1982), Frost (1985), and Duellman and Wiens (1992). The last authors applied the senior synonym, Scinax, to Ololygon, making the new combination Scinax ehrhardti. Pombal (in Frost, 1999, 2000) stated that this name has not been associated with any known population. An examination of Mu¨ller’s (1924) original description of H. ehrhardti, as translated by Lutz (1973), suggests that there were only a few elements (e.g., the reduction of webbing between first and second toes, concave loreal region) to associate this species with the (then) H. catharinae group, and it even contains some evidence against this association (‘‘vomerine teeth in two long, very slightly curved patches . . . first finger with a strongly marked pollex rudiment forming a semicircular border’’). Contrary to Lutz (1973) and Duellman (1977), the type specimen was not destroyed during WWII; it is currently housed at the ZSM and is in a fairly good state of preservation. Examination of the holotype in question (Figs. 1A, 2A; for a list of specimens examined, see Appendix 1) reveals that it does not belong to Scinax, but the similarities with the Hyla albomarginata group are immediately evident. A closer examination and comparison with the holotype of Hyla arianae Cruz and Peixoto, 1985, MZUSP 58652 (Figs. 1B, 2B) indicates that H. ehrhardti is a senior synonym of this species. The measurements of both holotypes (Table 1) are similar, although the head of the holotype of H. ehr-

326


FIG. 1. Head profiles of the holotypes of (A) Hyla ehrhardti ZSM 80/1921 and (B) Hyla arianae MZUSP 58652, not to same scale.

hardti is proportionally somewhat larger. Morphologically, both specimens are very similar and share all the diagnostic characters of H. arianae established by Cruz and Peixoto (1985). The main differences are that the holotype of H. ehrhardti has slightly better developed webbing on hands (left hand webbing formula II 1 2/3–3⫺ III 2⫹–2⫹ IV vs II 1 2/3–3⫺ III 2 1/3–2⫹ IV in the holotype of H. arianae) and foot (left foot webbing formula I 1 1/2–2⫹ II 1⫹– 2 1/2 III 1 1/3–21/2 IV 2⫹–1⫹ V vs I 1 2/3–2⫹ II 1 1/3–2 2/3 III 2⫺–3⫺ IV 2⫹–1⫹ V in the holotype of H. arianae; formula follows

Savage and Heyer, 1967, as modified by Myers and Duellman, 1982), has more dark spots on the dorsum, and lacks the scattered white spots present in the holotype of H. arianae (likely an artifact of preservation). The webbing formula of the foot of the holotype of H. arianae is slightly different from that given by Cruz and Peixoto (1985), but determining webbing formulae sometimes has an element of subjectivity, as discussed by Myers and Duellman (1982), and we do not regard this difference as a problem. The type locality of H. arianae is Rio dos Cedros,

FIG. 2. Dorsal views of the holotypes of (A) Hyla ehrhardti and (B) Hyla arianae, not to same scale. Note that the head is deflected in A, but snout shape is actually the same in both specimens.

SHORTER COMMUNICATIONS Saõ Bernardo, Santa Catarina, Brazil, and the paratypes are from Novo Horizonte and Saõ Bento do Sul, both localities in Santa Catarina. Rio dos Cedros is about 32 km from the type locality (Corupa´) of H. ehrhardti. The synonymy of this species is: Hyla ehrhardti Mu¨ller, 1924 Hyla ehrhardti Mu¨ller, 1924. Original description. Hyla erhardti: Bokermann, 1966. Unjustified emendation or subsequent misspelling. Ololygon ehrhardti: Fouquette and Delahoussaye, 1977. First combination with Ololygon. Hyla arianae Cruz and Peixoto, 1985. New synonymy. Scinax ehrhardti: Duellman and Wiens, 1992. First combination with Scinax. Acknowledgments.—We are indebted to F. Glaw (ZSM) and M. T. Rodrigues (MZUSP) for the loan of the holotypes of H. ehrhardti and H. arianae, respectively. T. Grant, D. R. Frost, and C. W. Myers read and criticized the manuscript. W. L. Smith kindly took the photographs. M. Weksler helped with the localities. JF thanks the American Museum of Natural History for funding. CdC is a Bolsista do Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´gico (CNPq).

LITERATURE CITED BOKERMANN, W. C. A. 1966. Lista Anotada das Localidades Tipo de Anfı´bios Brasileiros. Serviço de Documentaçaõ, Reitoria da Univ. de Saõ Paulo, Saõ Paulo, Brasil. CRUZ, C. A. G., AND O. L. PEIXOTO. 1985. Espećies verdes de Hyla: o complexo ‘‘albofrenata’’ (Amphibia, Anura, Hylidae). Arquivos da Universidade Federal Rural de Rio de Janeiro 8:59–70. DUELLMAN, W. E. 1977. ‘‘1976.’’ Liste der rezenten amphibien und reptilien. Hylidae, Centrolenidae, Pseudidae. Das Tierreich 95:1–225. DUELLMAN, W. E., AND J. J. WIENS. 1992. The status of the hylid frog genus Ololygon and the recognition of Scinax Wagler, 1830. Occasional Papers of the Museum of Natural History, University of Kansas 15:1–23.

327

HARDING, K. A. 1983. Catalogue of New World Amphibians. Pergamon Press Ltd., New York. FOUQUETTE, M. J., AND A. J. DELAHOUSSAYE. 1977. Sperm morphology in the Hyla rubra group (Amphibia, Anura, Hylidae), and its bearing on generic status. Journal of Herpetology 11:387–396 FROST, D. R. (ED.). 1985. Amphibian Species of the World: A Taxonomic and Geographical Reference. Allen Press, Inc., Lawrence, KS. . 1999. Amphibian Species of the World: An Online Reference. V2.1. Available at http://research.amnh.org.herpetology/amphibia/index.html . 2000. Amphibian Species of the World: An Online Reference. V2.20. Available at http://research.amnh.org/herpetology/amphibia/index.html. LUTZ, B. 1973. Brazilian Species of Hyla. Univ. of Texas Press, Austin. MU¨LLER, L. 1924. Neue laubfro¨sche aus dem state Santa Catharina, S. O. Brasilien. Zoologischen Anzeiger 59:233–238. MYERS, C. W., AND W. E. DUELLMAN. 1982. A new species of Hyla from Cerro Colorado, and other tree frog records and geographical notes from western Panama. American Museum Novitates 2752:1–32. SAVAGE, J. M., AND W. R. HEYER. 1967. Variation and distribution in the tree-frog genus Phyllomedusa in Costa Rica, Central America. Beitrage zur Neotropischen Fauna 5:111–131. Accepted: 8 September 2001. APPENDIX 1 List of specimens examined (MZUSP, Museu de Zoologia, Universidade de Saõ Paulo, Brasil. ZSM, Zoologisches Staatssammlung Mu¨nich, Germany). Hyla ehrhardti: MZUSP 58652 (Holotype of Hyla arianae): Brazil: Santa Catarina: Rio dos Cedros to Saõ Bernardo. 9 January 1982. W. R. Heyer coll. ZSM 80/ 1921 (Holotype): Brazil: Santa Catarina: Humboldt (basin of the Rio Novo). September 1918. W. Erhardt coll.