Herpetologica, 61(3), 2005, 250259 2005 by The Herpetologists League, Inc.

RELATIONSHIPS AMONG FORAGING VARIABLES, PHYLOGENY, AND FORAGING MODES, WITH NEW DATA FOR NINE NORTH AMERICAN LIZARD SPECIES
WILLIAM E. COOPER, JR.1,4, LAURIE J. VITT 2, JANALEE P. CALDWELL2,
1 2

AND

STANLEY F. FOX3

Department of Biology, Indiana University-Purdue University at Fort Wayne, Fort Wayne, IN 46805, USA Sam Noble Oklahoma Museum of Natural History and Department of Zoology, University of Oklahoma, Norman, OK 73072, USA 3 Department of Zoology, Oklahoma State University, Stillwater, OK 74078, USA ABSTRACT: Complete characterization of lizard foraging behaviors may require information about aspects rarely measured. Most studies record only number of movements per minute (MPM) and/or percent of time moving (PTM), but lizards differ markedly in average speed (AS) and speed while moving (MS) during foraging and in proportion of attacks initiated after detecting prey while the lizard is moving (PAM). We present data on these variables for nine lizard species and on foraging speed for several others, permitting rst assessments of relationships between speed, PAM, and both phylogeny and foraging mode; examination of the effect of body length on foraging speed; and correlative analyses of relationships between foraging variables. Although sprint speed may increase with body size, foraging speed did not, presumably for two reasons. Because search speed is much lower than sprint speed, as is speed of movement between ambush sites, searching efciency and stamina may be more important determinants of foraging speed than is sprint speed. Second, the body size range was small, allowing the possibility that foraging speed may vary with body length over the much larger size range between the smallest and largest species worldwide. Nevertheless, a large majority of lizard species are in the size range tested, suggesting that body length may not strongly affect foraging speed except when extremely short or long species are included in comparative analyses. High PAM, high AS, and low MS were characteristic of autarchoglossans and active foragers, whereas low PAM, low AS and high MS were exhibited by iguanians and ambush foragers. In independent species analyses, signicant correlations were observed between several pairs of foraging variables. In analyses using phylogenetically independent contrasts, the only signicant nding was a strong positive correlation between PAM and PTM. Although these ndings suggest that foraging speed, MPM, and either PTM or PAM may provide independent measures of foraging activity needed to adequately describe interspecic variation, this conclusion is tentative due to the small sample size of limited taxonomic breadth. Key words: Foraging behavior; Foraging mode; Foraging speed; Lizard; Squamata

LIZARD foraging behavior has been intensively investigated since the early 1980s, in large part to test predictions made by Huey and Pianka (1981) and others (e.g., Huey and Bennett, 1986; Regal, 1978; Schoener, 1971; Stamps, 1977; Vitt and Congdon, 1978) about consequences of ambush foraging and active foraging for several aspects of lizard ecology, behavior, sensory capacities, and morphology and relationships between foraging modes, physiology and locomotor performance. These studies have been very fruitful, leading to the foraging mode paradigm that seeks to explain much of the evolutionary and ecological diversity of lizards (Vitt et al., 2003). Although numerous correlates of foraging mode have been discovered, the existence of
4

CORRESPONDENCE: e-mail, cooperw@ipfw.edu 250

the distinct foraging modes has become controversial. Some investigators maintain that lizard foraging behavior exhibits sufcient continuous variation that ambush and active foraging modes are inadequate to characterize it (Perry, 1999; Pietruszka, 1986). Others have continued to recognize the existence of discrete foraging modes either based on bimodality of the distribution of continuous foraging variables (McLaughlin, 1989), the occurrence of two distinct groupings in discriminant analyses (Butler, in press), or convenience of statistical analysis for testing hypotheses about factors related to foraging behavior (Cooper, 1995, 1997). It has been clear for over 40 years that foraging behavior differs greatly among major lizard taxa and is very stable within some large taxa (Cooper, 1994a,b; Evans, 1961; Perry, 1999). As more data have become available for

September 2005]

HERPETOLOGICA

251

TABLE 1.Sample sizes (n), duration of observations (minutes), number of movements per minute (MPM), and percent time moving (PTM) sizes for 9 species of North American lizards.
MPM Family Species n Minutes  x SE Range  x PTM SE Range

Phrynosomatidae Sceloporus arenicolous S. merriami S. poinsettii S. sleveni S. undulatus Polychrotidae Anolis sagrei Tropiduridae Leiocephalus carinatus Teiidae Aspidoscelis exsanguis A. tessellata

5 11 12 16 15 22 24 12 9

32.4 109.9 104.2 144.0 142.2 209.1 230.0 103.7 41.9

0.80 0.24 0.08 0.24 0.31 0.25 0.20 0.75 1.01

0.42 0.06 0.03 0.05 0.10 0.06 0.05 0.08 0.09

0.200.24 0.000.60 0.000.30 0.000.53 0.001.49 0.000.90 0.001.00 0.401.30 0.591.45

1.4 0.9 0.2 0.4 1.0 0.5 0.5 82.4 88.6

0.7 0.3 0.1 0.1 0.4 0.0 0.1 2.9 3.2

0.34.0 0.03.7 0.00.8 0.00.9 0.05.3 0.00.2 0.02.0 67.199.0 71.896.8

an ever-increasing diversity of lizard families, much continuous variation has been discovered within and among taxa, leading to the conclusion that two modes are inadequate to express all of the variation (Perry, 1999, in press). However, data remain unavailable for several important lizard families and the number of species sampled even in large families remains small. More data are needed to assess the extent to which foraging behavior may vary continuously among species or may be divided into categories corresponding to separate modes, each characterized by internal continuous variation. For a large majority of species for which foraging behavior has been quantied, data are restricted to number of movement bouts per minute and/or percent time spent moving. These are important features of searching for prey, but other important information is needed to characterize foraging behavior more fully. Huey and Pianka (1981) reported average speed during focal observations and speed while moving during foraging, but information on foraging speeds has rarely been published since their paper (see Cooper and Whiting, 1999 for an exception). One reason for the rarity of speed data may be that Huey and Pianka (1981) noted that differences in foraging speeds might be attributable to differences in body size, making them difcult to interpret and perhaps requiring adjustment for interspecic differences in body size.

Because ambush and active foraging modes are methods of searching for prey, Cooper and Whiting (1999) proposed that the proportion of attacks on prey initiated after discovering the prey while moving (PAM) could be a superior index of foraging mode. They and Cooper et al. (2001) presented data on this variable for limited numbers of African lacertids and for North American representatives of several families, but more data are needed to assess the utility of PAM and relationships between PAM and other foraging variables. We report new data on all variables discussed above for nine species of North American lizards representing four families and present foraging speeds for additional species for which movements per minute (MPM), percent time moving (PTM), and PAM were reported previously (Cooper et al., 2001). We provide a rst test of the hypothesis that foraging speed varies with body length. By combining our data with previously published data, we assess for the rst time possible differences in foraging speeds and the attackbased index between major lizard taxa and foraging modes. We examine for the rst time correlations among all foraging variables. MATERIAL AND METHODS Lizard species, samples sizes and duration of observations for species observed in 2004 are presented in Table 1. Sceloporus arenicolous

252

HERPETOLOGICA

[Vol. 61, No. 3

were observed on 24 May in the Mescalero Sands in Chaves County, New Mexico; S. merriami on 30 April to 2 May and S. poinsettii on 12 May at Black Gap Wildlife Management Area in Brewster County, Texas; S. sleveni on 19 May and Aspidoscelis (formerly Cnemidophorus) exsanguis on 918 May in the Chiricahua Mountains of Cochise County, Arizona (Herb Martyr area in Coronado National Forest, on the grounds of the American Museum of Natural Historys Southwest Research Station, and in the Chiricahua National Monument); S. undulatus and A. tesselata on 23 May in Eddy County, New Mexico; and Anolis sagrei and Leiocephalus carinatus on 1926 June in Palm Beach, Palm Beach County, Florida. All lizards observed were adults except eight A. exsanguis. Because sex differences in MPM and PTM were not detected in studies of other North American species (Cooper et al., 2001) using similar or larger sample sizes, sexes were not recorded. An observer walked slowly through each area to sight lizards for focal observations using unaided vision and searching with binoculars to detect lizards at greater distances to reduce possible disturbance. Upon detecting a lizard, an observer stopped moving immediately, also to minimize disturbance. The distance between observer and lizards during observations was greater than 5 m in all but a few cases noted below. Many individuals of the larger ambush-foraging species, S. poinsettii and L. carinatus, were observed at distances greater than 30 m. Aspidoscelis sp. typically were observed at relatively short distances as they foraged actively, moving into and through vegetation and other objects that obscured them from view at greater distances. Their movement sometimes required that the observer move to keep them in view, especially for individuals that moved relatively long distances. Fortunately, by moving slowly only while a lizard is moving and maintaining a minimum distance of ve meters between lizard and observer while the lizard is moving, it is possible to observe normal foraging by Aspidoscelis sp. with minimal disturbance (Anderson, 1993; Cooper et al., 2001). As in previous studies of Aspidoscelis (Anderson, 1993; Cooper et al., 2001), some individuals of A. exsanguis and A. tesselata (formerly A. grahami) changed direction and approached

the immobile observer, appearing to be unaware of his presence. Each focal observation lasted 10 min when possible. However, focal observations were sometimes shorter (Table 1), with a minimum of 1.5 min for all but one species because lizards sometimes moved to locations where they could not be observed. Shorter intervals (68 and 72 s) were accepted for two individuals of S. arenicolous because the sample size for this species was very small. Foraging data were recorded on microcassette tapes only for individuals that did not appear to have been disturbed. To minimize possible effects of disturbance to the lizard that could not be readily discerned, an observer waited 2 min after detecting a lizard and stopping before beginning a focal observation. To ensure that data points were independent, individuals were observed only once. To accomplish this, each area was sampled only once and before moving to a new location. For each focal observation, the species, locality, date, time, and food-related behaviors were noted. The behavioral data recorded were time spent moving (to the nearest second), time spent immobile, distances moved (to the nearest 0.1 m), feeding attempts, and whether feeding attempts were made after detecting prey while the lizard was immobile or while it was searching actively. Lizards that are not walking or running often move specic body parts such as the head or tail. Only translational movements, i.e., ones to new locations, were included in estimates of time spent moving and distance moved. Thus, postural changes and movements of specic body parts by lizards not moving the entire body to a new location did not contribute to estimates of time moving or distance moved. Several foraging variables were calculated from the movement and attack data. PTM and MPM (e.g., Cooper et al., 1997, 2001; Huey and Pianka, 1981; Perry, 1999, in press), the most widely used indices of lizard foraging behavior, were calculated. Ambush foragers have lower PTM than active foragers, but MPM is more variable within foraging modes. The other foraging movement variables were average speed (AS), i.e., distance moved divided by observation time in seconds, and speed while moving (MS), which is distance divided by time moving.

September 2005]

HERPETOLOGICA

253

TABLE 2.Average speed (m/s), speed while moving (m/s), and sample sizes for North American lizard species studied by Cooper et al. (2001).
Average speed Family Species n  x SE Range n  x Moving speed SE Range

Crotaphytidae Crotaphytus collaris Phrynosomatidae Cophosaurus texanus Holbrookia propinqua Sceloporus clarkii S. jarrovii S. virgatus Urosaurus ornatus Scincidae Eumeces laticeps Teiidae Aspidoscelis deppii A. sexlineata A. sonorae A. uniparens

39 14 14 20 46 19 20 25 39 12 9 29

0.003 0.002 0.006 0.000 0.003 0.003 0.005 0.030 0.030 0.063 0.050 0.058

0.001 0.011 0.002 0.001 0.001 0.001 0.001 0.004 0.003 0.009 0.009 0.006

0.0000.036 0.0000.153 0.0000.027 0.0000.021 0.0000.016 0.0000.015 0.0000.027 0.0030.082 0.0020.072 0.0330.113 0.0280.076 0.0180.158

14 13 12 14 37 19 24 25 39 12 9 29

0.689 0.612 0.477 0.457 0.351 0.186 0.225 0.042 0.049 0.092 0.051 0.070

0.168 0.191 0.132 0.142 0.057 0.027 0.035 0.006 0.004 0.008 0.009 0.007

0.0502.286 0.0722.316 0.1071.676 0.0041.829 0.0211.625 0.0250.432 0.0250.596 0.0080.134 0.0130.107 0.0430.145 0.0310.077 0.0200.197

The nal foraging variable was PAM, calculated as number of attacks on prey discovered while moving divided by total attacks by both moving and immobile lizards. Movements by ambushers needed to attack prey detected by lizards while immobile must be excluded because they are part of the attack rather than search for prey. PAM provides a more direct measure of the outcomes of sit-and-wait versus active search than PTM or MPM and requires no information about time spent moving, number of movements, or speed. PAM can supplement these latter variables and eventually may provide an excellent comparative measure of search by ambush versus while active. It also might help to assess relationships among the other measures of foraging activity. However, PAM data are available for fewer species than other foraging variables because data collection is much more time-consuming for PAM than the other variables. PAM is especially difcult to estimate for ambush foragers having low feeding rates. Environmental conditions may affect lizard activity strongly. In particular, heliothermic species spend more time basking and less foraging at lower temperatures; at high temperatures, time spent in thermoregulatory cooling by shuttling between sunny and shaded sites or resting in shade increases with temperature (e.g., Vitt et al., 1993 for A. deppii). To ensure

that estimates of foraging activity are comparable for between species, it is necessary to make observations under conditions that favor normal foraging activity. Species observed in the present study are all diurnal, and all had free access to sunlit and shaded sites, permitting normal thermoregulation. Focal observations were made only on sunny days when morning basking had been completed and lizards were active. The data represent foraging activity by lizards at or close to preferred body temperatures for each species because body temperatures are maintained within a narrow range while heliothermic species are active. Preferred body temperatures vary among species, but foraging is usually conducted by each species in its own range of preferred body temperature. In comparative analyses involving species not studied in 2004, published data were used for all variables for ve lacertid species (Cooper and Whiting, 1999) and for PAM, PTM and MPM for a crotaphytid, six phrynosomatids, a skink, and four teiids (Cooper et al., 2001). Previously unpublished AS and MS for species in Cooper et al. (2001) are presented in Table 2. Maximum speed may vary with body length, but whether foraging speed does is unknown. To determine whether speed variables should be adjusted for body length, we conducted regressions of AS and MS on snoutvent length

254

HERPETOLOGICA

[Vol. 61, No. 3

TABLE 3.Average speed (m/s), speed while moving (m/s), and sample sizes for nine species of American lizards.
Average speed Family Species n  x SE Range n  x Moving speed SE Range

Phrynosomatidae Sceloporus arenicolous S. merriami S. poinsettii S. sleveni S. undulatus Polychrotidae Anolis sagrei Tropiduridae Leiocephalus carinatus Teiidae Aspidoscelis exsanguis A. tesselata

5 11 11 16 15 15 24 12 12

0.002 0.001 0.001 0.002 0.005 0.001 0.003 0.029 0.030

0.001 0.000 0.000 0.001 0.002 0.000 0.000 0.005 0.003

0.0000.008 0.0000.004 0.0000.007 0.0000.005 0.0000.023 0.0000.004 0.0000.011 0.0090.047 0.0190.045

5 8 5 12 12 14 18 12 9

0.165 0.226 0.301 0.441 0.390 0.259 0.543 0.032 0.035

0.037 0.086 0.091 0.079 0.057 0.026 0.082 0.005 0.005

0.0150.262 0.0550.806 0.1270.635 0.1521.075 0.0760.740 0.1190.521 0.0501.130 0.0120.075 0.0200.063

(SVL) for 19 species in the phylogeny presented below. SVLs were obtained from Vitt et al. (1993), Branch (1998), Conant and Collins (1998), and Stebbins (2003). Differences in PAM between iguanian and scleroglossan lizards were examined for signicance using a Mann-Whitney U-test for limited data available from the present study, Cooper et al. (2001) for several families of American lizards, and Cooper and Whiting (1999) for ve species of African lacertids. Estimates of PAM based on less than four observations for a single species were excluded. For Mann-Whitney U-tests of phylogenetic differences in speed variables, present data were combined with data on African lacertids for which n ! 5 from Cooper and Whiting (1999). Phylogenetic differences in MPM and PTM were not tested here because they have been examined by Perry (1999) for a much wider range of species. Relationships among PAM, MPM, PTM, and foraging speeds were examined using simple linear regression and correlation (Zar, 1996) as well as the method of independent contrasts (Felsenstein, 1985). The traditional nonphylogentetic (tips) approach was useful for indicating the extent to which two variables apparently indicate the same degree of foraging activity, and independent contrasts were needed to determine whether correlated evolution has occurred between variables. For correlations using independent contrasts, we used phylogenies from Reeder (1995), Reeder and Wiens (1996), Wiens and Reeder (1997), and Flores-Villela et al. (2000) for phrynoso-

matids; Schulte et al. (1998) for iguanians; Wright (1992) and Reeder et al. (2002) for teiids; Fu (2000) for lacertids; and Estes et al. (1988) for autarchoglossans. Independent contrasts were computed using COMPARE 4.6. Branch lengths in Garland (1994) were used, but were unavailable for some branches whose lengths were estimated by assuming early divergence among families. The phylogeny and branch lengths used were: (((Leiocephalus carinatus:151, Crotaphytus collaris:151):10, ((Cophosuarus texanus:3, Holbrookia propinqua:3):62, (Urosaurus ornatus:26.3, ((Sceloporus jarrovii:1, S. clarkii:1, S. virgatus:2):24.3):38.7):96):10, ( Eumeces laticeps:155, ((Pedioplanis namaquensis:2, P. undata:2):6, (Heliobolus lugubris:5, (Meroles knoxii :2, M. ctenodactylus :2):3):3):138, (Aspidoscelis deppii:6, (A. sexlineata:3, (A. exsanguis:2, (A. uniparens:1, A. sonorae:1) :1):1):3):140):9):16):1. All tests of signicance were two-tailed with a 5 0.05. Although unadjusted P values are presented in results, reported signicance has been veried by sequential Bonferroni adjustment (Wright, 1992) for the number of regressions. Data in the text are reported as  x 6 SE. RESULTS 2004 Data Although MPM values were variable, PTM, average speed, and speed while moving were similar among all species except the two teiids (Tables 1, 3). The four Sceloporus spp., the

September 2005]

HERPETOLOGICA

255

polychrotid, and the tropidurid exhibited movement patterns typical of ambush foragers. MPM was low (,0.33) in most of these iguanians, the only exception being S. arenicolous for which sample size was small. PTM was ,1.5% in all iguanians, indicating that they spent a large majority of the time immobile, hunting by ambush. Combined with previous data for other of Sceloporus and Anolis, present data conrm low PTM in these taxa. For eight species of Sceloporus, PTM is 1.36 6 0.66 with range 0.185.84. For 19 species of Anolis (sensu lato), PTM is 1.63 6 0.47 (range 0.087.20). The two teiids are typical active foragers. In both species of Aspidoscelis, MPM was substantially higher than in iguanians other than S. arenicolous. Their PTM values were typical for the genus and much higher than those of iguanians. PAM data were obtained only for four species and were limited in all cases. Sceloporus merriami attacked once, Anolis sagrei twice, and L. carinatus four times from ambush. Because none of these species was observed to attack during active search, PAM 5 0.00 for each species. Aspidoscelis exsanguis attacked eight times, all as a result of discovering prey during active search, giving PAM 5 1.00. Foraging speeds differed dramatically between iguanians and teiids. Iguanians had AS 5 0.002 6 0.001 m/s (range 0.0010.005), far lower than AS for the two teiids, which were 0.030 6 0.001 m/s, with range 0.0290.030 m/s. Iguanians, however, had much faster speeds during movements, MS 5 0.332 6 0.0500 m/s with range 0.1650.543 m/s, compared to the teiids, for which speed while foraging was only MS 5 0.034 6 0.002 m/s with range 0.032 0.035 m/s. Phylogenetic Differences in Foraging Speed and PAM For the 19 species used in regressions and phylogenetic analyses below, foraging speeds were not signicantly correlated with body length (AS 5 0.19 SVL 0.05; F 5 0.64; df 5 1, 17; P . 0.43; MS 5 0.14 SVL 0.13; F 5 0.36; df 5 1, 17; P . 0.56). Therefore, foraging speeds were not adjusted for differences in body size in the following tests. Average speed of the seven iguanian species observed in 2004 was far less than that of the

two teiid and ve African lacertid species, which was AS 5 0.036 6 0.007 m/s (range 0.0070.059) even though one of the lacertids (Meroles knoxii) is an ambush forager. The average speed of the lacertids and teiids was signicantly greater than that of the iguanians (U 5 0.0; n 5 7, 7; P , 0.001 each). Average speed of the six species of active foragers (0.041 6 0.006 m/s) was signicantly greater than that of the eight species of ambush foragers (0.002 6 0.001 m/s; U 5 0.0; P , 0.002). Speed while moving was signicantly greater for seven iguanians than for the teiids and lacertids, for which MS 5 0.091 6 0.034 m/s, range 0.0320.159 m/s (U 5 0.0; n 5 7, 7; P , 0.001). Moving speed of eight ambush foragers (0.311 6 0.110) was signicantly greater than that of the six active foragers (0.079 6 0.032; U 5 0.0; n 5 6, 8; P , 0.002). Adding present data, PAM estimates are now available for eight iguanian, one scincid, ve teiid and ve lacertid species. PAM 5 0.00 for all of the iguanians, and one lacertid that is an ambush forager. For all of the autarchoglossans, PAM 5 0.83 6 0.10, with range 0.00 1.00. For all of the actively foraging species (all but the one species of the autarchoglossans), PAM was 0.92 6 0.07, with range 0.331.0. PAM was signicantly lower for representatives of the iguanian families Phrynosomatidae, Tropiduridae, and Crotaphytidae than for representatives of the autarchoglossan families Lacertidae, Teiidae, and Scincidae (U 5 4; n 5 8, 11; P , 0.001). It was signicantly lower for ambush foragers than active foragers (U 5 0.00; n 5 9, 10; P , 0.001). Relationships between Foraging Variables In independent species analyses, PAM was strongly and signicantly related to PTM (PTM 5 0.97 PAM 2.09; F 5 231.84; df 5 1, 17; P , 1 3 106), but was not signicantly related to MPM (F 5 3.78; df 5 1, 17; P , 0.069). MPM and PTM were not signicantly related to each other (F 5 2.66; df 5 1, 17; P , 0.13). Average speed and moving speed exhibited a negative linear relationship to each other having the equation AS 5 0.63 MS 0.05 (F 5 11.14; df 5 1, 17; P , 0.004). Both speeds were signicantly related to PAM, average speed positively (AS 5 0.87 PAM 0.01; F 5 54.09; df 5 1, 17; P 5 2.0 3 106) and moving speed negatively (MS 5 0.78

256

HERPETOLOGICA

[Vol. 61, No. 3

PAM 0.40; F 5 25.84; df 5 1, 17; P 5 9.2 3 105). Average speed had a positive, but nonsignicant, linear relationship with MPM (AS 5 0.44 MPM 0.02; F 5 4.14; df 5 1, 17; P 5 0.058) and a highly signicant relationship with PTM (AS 5 0.87 PTM 0.01; F 5 50.90; df 5 1, 17; P 5 2.0 3 106). Moving speed had a negative linear relationship with both MPM (MS 5 0.53 MPM 0.37; F 5 6.70; df 5 1, 17; P , 0.019, marginally nonsignicant after Bonferroni adjustment) and PTM (MS 5 0.78 PTM 0.40; F 5 26.58; df 5 1, 17; P 5 7.9 3 105). Most of these relationships were not significant in the phylogenetic analyses using independent contrasts (Table 4). PAM and PTM were the only variables that were signicantly correlated (r 5 0.84; df 5 1, 17; P , 0.001). The correlation between PAM and average speed was suggestive, but not signicant after Bonferroni adjustment (r 5 0.50; df 5 1, 17; P , 0.032). No other correlations approached signicance in the independent contrasts analyses. DISCUSSION Foraging Differences between Clades and Foraging Modes All iguanians in this study exhibited the low PTM characteristic of ambush foragers, whereas the two species of Aspidoscelis exhibited much higher PTM values typical for active foragers. MPM is more variable than PTM within foraging modes among lizards because some ambush foragers move briey yet fairly frequently and some active foragers spend such a high percentage of the time in continuous movement that their MPM values are necessarily low (Cooper et al., 2001). In the present study, the number and diversity of species sampled did not represent the full range of foraging diversity for MPM within either Iguania or Autarchoglossa. However, the lower MPM for iguanians (all but one species) and the higher values for the two teiids were typical of North American species studied previously (Cooper et al., 2001). That foraging speeds were unrelated to body size in the 19 species in the regression analyses was not surprising. Although maximum sprint speed increases with body size (Van Damme and Vanhooydonck, 2001), lizards forage at

TABLE 4.Correlations between phylogenetically independent contrasts for pairs of foraging variables for 19 species. * Signicant at P , 0.001.
PTM MPM AS MS

PAM PTM MPM AS

0.84*

0.35 0.34

0.50 0.34 0.28

0.17 0.20 0.06 0.03

much lower than maximum speeds. Ambush foragers necessarily have much lower average than maximum speeds because they are motionless much of the time. Even when attacking prey and moving from one ambush post to another, maximum speeds are not needed and might interfere with ability to capture slow-moving prey. For active foragers that spend much of the time moving, high speed would interfere with searching ability, especially by species that rely on tongueicking to locate chemical cues from hidden prey or that visually hunt for prey in litter or other complex substrates. Foraging speed in active foragers must be slow enough to avoid exhaustion. Thus, foraging speed may be affected by foraging modes and stamina and must be much lower than maximal burst speed. For lizards in the size range of this study (maximum SVL 51143 mm), body length did not affect foraging speed. However, the size range here includes only a small portion of the size range for lizards worldwide, which exhibit 50-fold variation in SVL between tiny geckos and the largest varanids. Foraging speeds might be affected by body length if a broader range of body size is considered. Nevertheless, the lack of relationship between foraging speed and body length in the size range used here is important because it applies to a large majority of lizards in many regions. For example, in the geographical area covered by Conant and Collins (1998), none of 72 species indigenous to the USA are smaller than species in the analysis and only ve anguids and a single phrynosomatid (Sceloporus serrifer) are larger. Thus, 92% of species are in the size range of species in this study. The larger anguids move slowly, suggesting that body size may not have a strong effect for them, either. Many of the largest species are herbivorous (e.g., Cooper and Vitt, 2002; Pough, 1973), including iguanas, and thus are excluded from studies of

September 2005]

HERPETOLOGICA

257

search modes in carnivores. Thus, the ndings suggest that effects of body size on foraging speed are likely to be negligible except when very small geckos or very large active foragers such as varanids, the largest teiids and skinks are compared with other lizards. Foraging speeds differed dramatically between lizard clades. The two teiids studied had average speeds 15 times greater than the iguanians, but their speed while moving was only about one tenth that of the iguanians. In the comparison adding ve species of African lacertids, the difference in average speed between iguanians and autarchoglossans was slightly greater, but the difference in speed while moving was reduced, autarchoglossans having a speed while moving slightly greater than one fourth that of iguanians. Differences in foraging modes may account better for differences in foraging speed than phylogeny alone, but too few data are available to conduct a meaningful test for an effect of foraging mode independent of phylogeny. However, differences between analyses by phylogenetic group and foraging mode are attributable to inclusion of one ambush forager, Meroles knoxii (PTM 5 0.07), one of several ambushers in a family composed primarily of active foragers (Arnold, 1990; Perry et al., 1990). PAM differed greatly between both taxonomic groups and foraging modes. Although this variable was conceived as a possibly reliable way to determine differences among species in methods of searching for prey (Cooper and Whiting, 1999), present evidence is inadequate to assess the effect of foraging mode independent of phylogeny on PAM. Nevertheless, the single species for which a transition from ancestral active foraging to ambush foraging has occurred, PAM has decreased. In another lacertid that has atypically low PTM, yet is an active forager, PAM is also atypically low. To assess the importance of foraging modes to PAM, data on PAM, foraging speeds, and the other foraging variables are needed for multiple taxa representing lineages in which evolutionary shifts in foraging mode have occurred. Relationships between Foraging Variables Relationships involving PAM and foraging speeds are of interest because they have not been reported previously for speed and data for

PAM have included only North American species (Cooper et al., 2001). As reported previously for a smaller sample of species, PAM and PTM are very highly correlated when species are considered independent. Although the previous study (Cooper et al., 2001) found no signicant relationship between PAM and PTM in an analysis accounting for phylogenetic relationships, these variables were strongly related in this study using a larger and taxonomically broader sample. In part, a strong relationship between PAM and PTM is inevitable because a predator that spends a greater percentage of its time moving is expected by chance to locate and attack prey a greater proportion of the time while moving. For the limited data, it appears that active foragers may discover and attack prey while moving at a higher percentage of the time than they spend moving and that ambush foragers may attack prey while moving at a lower percentage of the time than they spend moving. Foraging movements by ambushers are usually attacks on prey or shifts to new ambush posts, whereas movements and search are combined by active foragers. In the independent species analysis, average speed was negatively related to speed while moving and positively to PAM and PTM, whereas speed while moving was negatively related to both PAM and PTM and had a marginal negative relationship with MPM. These relationships conform to predictions based on differences in searching behavior, but were not signicant in the analyses using independent contrasts. However, the correlation between average speed and PAM was suggestive. The relationships of foraging speeds to other foraging variables merit further investigation in comparative studies with a larger sample size and representing several major shifts in foraging behavior within lineages. Foraging speed variables may be important for characterizing lizard foraging behavior and clarifying the existence of discrete foraging modes. MPM and PTM were not signicantly correlated in either tips or independent contrast analyses. This contrasts with a signicant tips correlation (r 5 0.73) reported by Cooper et al. (2001). Present results are more reliable because the sample is more than 50% larger and more taxonomically diverse than in

258

HERPETOLOGICA

[Vol. 61, No. 3

previous work. PTM and MPM need not be related because a species that moves a xed percentage of the time may divide movements into bouts of variable duration and number. Independence of PTM and MPM suggests the importance of considering both variables simultaneously when characterizing lizard foraging behavior (Butler, in press).
Acknowledgments.We are grateful to E. Martins for help with use of COMPARE. WEC is grateful to J. Dixon for valuable advice about eld sites in Texas and to T. Vanzant, S. Lerich, and M. Pittman for permission to do research and hospitality at the Black Gap and Elephant Mountain Wildlife Management Areas in Texas. WEC thanks D. Wilson of the American Museum of Natural Historys Southwest Field Station for hospitality during eld work and R. Cox for suggesting eld sites in the Chiricahua Mountains. We thank R. Van Loben Sels for hospitality during led work in the Chiricahuas. SFF thanks the Wichita Mountains National Wildlife refuge for permission to conduct observations under permit 56841.

LITERATURE CITED
ANDERSON, R. A. 1993. An analysis of foraging in the lizard, Cnemidophorus tigris. Pp. 83116. In J. W. Wright and L. J. Vitt (Eds.), Biology of Whiptail Lizards (Genus Cnemidophorus). Oklahoma Museum of Natural History, Norman, Oklahoma, U.S.A. ARNOLD, E. N. 1990. Why do morphological phylogenies vary in quality? An investigation based on the comparative history of lizard clades. Proceedings of the Royal Society of London, Series B 240:135172. BRANCH, B. 1998. Field Guide to Snakes and other Reptiles of Southern Africa. Ralph Curtis Books, Sanibel Island, Florida, U.S.A. BUTLER, M. A. In press. Foraging mode of an old-world chameleon, Brdypodion pumilum: a sit-and-wait or active forager? Biological Journal of the Linnaen Society. CONANT, R., AND J. T. COLLINS. 1998. A Field Guide to Reptiles and Amphibians: Eastern/Central North America. Houghton Mifin, New York, New York, U.S.A. COOPER, W. E., JR. 1994a. Chemical discrimination by tongue-icking in lizards: a review with hypotheses on its origin and its ecological and phylogenetic relationships. Journal of Chemical Ecology 20:439487. . 1994b. Prey chemical discrimination, foraging mode, and phylogeny. Pp. 95116. In L. J. Vitt and E. R. Pianka (Eds.), Lizard Ecology: Historical and Experimental Perspectives. Princeton University Press, Princeton, New Jersey, U.S.A. . 1995. Foraging mode, prey chemical discrimination, and phylogeny in lizards. Animal Behaviour 50:973985. . 1997. Correlated evolution of prey chemical discrimination with foraging, lingual morphology, and vomeronasal chemoreceptor abundance in lizards. Behavioral Ecology and Sociobiology 41:257265. COOPER, W. E., JR., M. J. WHITING, AND J. H. VAN WYK. 1997. Foraging modes of cordyliform lizards. South African Journal of Zoology 32:913.

COOPER, W. E., JR., AND M. J. WHITING. 1999. Foraging modes in lacertid lizards from southern Africa. Amphibia-Reptilia 20:299311. COOPER, W. E., JR., L. J. VITT, J. P. CALDWELL, AND S. F. FOX. 2001. Foraging modes of some American lizards: relationships among measurement variables and discreteness of modes. Herpetologica 57:6576. COOPER, W. E., JR., AND L. J. VITT. 2002. Distribution, extent, and evolution of plant consumption by lizards. Journal of Zoology 257:487517. ESTES, R., K. DE QUEIROZ, AND J. GAUTHIER. 1988. Phylogenetic relationships within Squamata. Pp. 119 281. In R. Estes and G. Pregill (Eds.), Phylogenetic Relationships of the Lizard Families. Stanford University Press, Stanford, California, U.S.A. EVANS, L. T. 1961. Structure as related to behavior in the organization of populations of reptiles. Pp. 148178. In W. F. Blair (Ed.), Vertebrate Speciation. University of Texas Press, Houston, Texas, U.S.A. FELSENSTEIN, J. 1985. Phylogenies and the comparative method. American Naturalist 125:115. FLORES-VILLELA, O., K. M. KJER, M. BENABIB, AND J. W. J. SITES. 2000. Multiple data sets, congruence, and hypothesis testing for the phylogeny of basal groups of the lizard Genus Sceloporus (Squamata, Phrynosomatidae). Systematic Biology 49:713739. FU, J. 2000. Toward the phylogeny of the family Lacertidaewhy 4708 base pairs of mtDNA sequences cannot draw the picture. Biological Journal of the Linnaean Society 71:203217. GARLAND, T., JR. 1994. Phylogenetic analysis of lizard endurance capacity in relation to body size and body temperature. Pp. 237259. In L. J. Vitt and E. R. Pianka (Eds.), Lizard Ecology: Historical and Experimental Perspectives. Princeton University Press, Princeton, New Jersey, U.S.A. HUEY, R. B., AND E. R. PIANKA. 1981. Ecological consequences of foraging mode. Ecology 62:991999. HUEY, R. B., AND A. F. BENNETT. 1986. A comparative approach to eld and laboratory studies in evolutionary biology. Pp. 8298. In M. E. Feder and G. V. Lauder (Eds.), Predator-Prey Relationships: Perspectives and Approaches from the Study of Lower Vertebrates. University of Chicago Press, Chicago, Illinois, U.S.A. MCLAUGHLIN, R. L. 1989. Search modes of birds and lizards: evidence for alternative movement patterns. American Naturalist 133:654670. PERRY, G. 1999. The evolution of search modes: ecological versus phylogenetic perspectives. American Naturalist 153:99109. PERRY, G. In press. Foraging behavior in lizards. In S. Reilly, L. McBrayer, and D. B. Miles (Eds.), Foraging Behavior in Lizards. Cambridge University Press, Cambridge, UK. PERRY, G., I. LAMPL, A. LERNER, D. ROTHENSTEIN, E. SHANI, N. SIVAN, AND Y. L. WERNER. 1990. Foraging mode in lacertid lizards: variation and correlates. Amphibia-Reptilia 11:373384. PIETRUSZKA, R. D. 1986. Search tactics of desert lizards: how polarized are they? Animal Behaviour 34:17421758. POUGH, F. H. 1973. Lizard energetics and diet. Ecology 54:837844. REEDER, T. W. 1995. Phylogenetic relationships among phrynosomatid lizards as inferred from mitochondrial ribosomal DNA sequences: substitutional bias and

September 2005]

HERPETOLOGICA

259

information content of transitions relative to transversions. Molecular Phylogenetics and Evolution 4:203222. REEDER, T. W., AND J. J. WIENS. 1996. Evolution of the lizard family Phrynosomatidae as inferred from diverse types of data. Herpetological Monographs 10:4384. REEDER, T. W., C. J. COLE, AND H. C. DESSAUER. 2002. Phylogenetic relationships of whiptail lizards of the genus Cnemidophorus (Squamata: Teiidae): a test of monophyly, reevaluation of karyotypic evolution, and review of hybrid origins. American Museum Novitates 3365:161. REGAL, P. J. 1978. Behavioral differences between reptiles and mammals: an analysis of activity and mental capabilities. Pp. 183202. In N. Greenberg (Ed.), Behavior and Neurology of Lizards. National Institute of Mental Health, Rockville, Maryland, U.S.A. SCHOENER, T. W. 1971. Theory of feeding strategies. Annual Review of Ecology and Systematics 2:369404. SCHULTE, J. A., II., J. R. MACEY, A. LARSON, AND T. J. PAPENFUSS. 1998. Molecular tests of phylogenetic taxonomies: a general procedure and example using four subfamilies of the lizard family Iguanidae. Molecular Phylogenetics and Evolution 10:367376. STAMPS, J. A. 1977. Social behavior and spacing patterns in lizards. Pp. 265334. In C. Gans and D. W. Tinkle (Eds.), Biology of the Reptilia, Vol. 7. Ecology and Behaviour A. Academic Press, London, U.K. STEBBINS, R. C. 2003. A Field Guide to Western Reptiles and Amphibians, 3rd ed. Houghton Mifin, Boston, Massachusetts, U.S.A.

VAN DAMME, R., AND B. VANHOOYDONCK. 2001. Origins of interspecic variation in lizard sprint capacity. Functional Ecology 15:186202. VITT, L. J., AND J. D. CONGDON. 1978. Body shape, reproductive effort, and relative clutch mass in lizards: resolution of a paradox. American Naturalist 112: 595608. VITT, L. J., P. A. ZANI, J. P. CALDWELL, AND R. D. DURTSCHE. 1993. Ecology of the whiptail lizard Cnemidophorus deppii on a tropical beach. Canadian Journal of Zoology 71:23912400. VITT, L. J., E. R. PIANKA, W. E. COOPER, JR., AND K. SCHWENK. 2003. History and the global ecology of squamate reptiles. American Naturalist 162:4460. WIENS, J. J., AND T. W. REEDER. 1997. Phylogeny of the spiny lizards (Sceloporus) based on molecular and morphological evidence. Herpetological Monographs 11:1101. WRIGHT, J. W. 1992. Evolution of the lizards of the genus Cnemidophorus. Pp. 2781. In J. W. Wright and L. J. Vitt (Eds.), Biology of Whiptail Lizards (genus Cnemidophorus). Oklahoma Museum of Natural History, Norman, Oklahoma, U.S.A. ZAR, J. H. 1996. Biostatistical Analysis. Prentice-Hall, Upper Saddle River, New Jersey, U.S.A. Accepted: 7 March 2005 Associate Editor: Michael Dorcas

Herpetologica, 61(3), 2005, 259267 2005 by The Herpetologists League, Inc.

THE ILIOSACRAL ARTICULATION IN PSEUDINAE (ANURA: HYLIDAE)

ADRIANA S. MANZANO1,3
1

AND

NICA BARG2 MO

CICyTTP-CONICET, Matteri y Espan a, (3105), Diamante, Entre Ros, Argentina 2 CONICET, Fac. de Biologa, UNMDP

We analyzed the iliosacral joint and iliolumbaris muscle anatomy of the species of Pseudinae, based on the two types of articulations dened by Emerson. Unusual characters should be expected in the pseudine iliosacral articulation structure, given their aquatic habits and evolutionary history. In fact, the presence of a ligamentous cap joining the ilium with the sacral diapophysis has not been described previously, and may be unique to pseudines. This particular group lacks a single pattern of iliosacral articulation for the whole subfamily, and the articulations are not strictly referable to any of the types or subtypes described by Emerson. The iliosacral articulation is interspecically variable within Pseudinae and is intermediate between the articulations of Type IIA and IIB. Key words: Anura; Hylidae; Iliosacral articulation; Ligamentous cap; Pseudinae

THE PELVIS of anurans generally is formed by three pairs of elements (ilium, ischium and
3

CORRESPONDENCE: e-mail, cidmanzano@infoaire.com.ar

pubis) that show some modications within modern anurans. Within Neobatrachia, the pelvis has been described as a stable structure formed by the ilium and the ischium, leaving the pubis as a reduced, nonossied cartilage,

260

HERPETOLOGICA

[Vol. 61, No. 3

with different degrees of mineralization (Trueb, 1973) in various taxa. There are some variations in the anuran pelvis that have been described before; for example, the presence of pre-pubic bones in Leiopelmatids, some Pipids (Cannatella and Trueb, 1988; Trueb, 1973), and Discoglossus sardus (Clarke, 1988; gener and Maglia, 1997), and post-pubic Pu elements in Ascaphus truei. Additional elements can be mineralized or remain cartilaginous. These variations, and some related to the shape of pelvic bones, have been used in phylogenetic analyses (e.g., by Heyer, 1975; Inger, 1972; Lynch, 1973). The expansion of the sacral diapophyses and their ventral joint with the ilium is a characteristic that makes modern anurans unique among the existing vertebrates (Jenkins and Shubin, 1998; Trueb, 1973). The arrangement of the iliosacral joint elements were categorized by Emerson (1979), who demonstrated that the type of articulation and its possible movements are related to the different forms of anuran locomotion. Emerson dened two types of articulations (I and II) based on the extension of the sacral diapophysis, the position and shape of the sesamoids, and the origin and insertion of the articular ligaments in extant anurans. The iliosacral articulation forms a functional complex with the muscle Iliolumbaris that is responsible for the movement of the pelvis during locomotion (Emerson, 1979; Whiting, 1961). Herein, we describe and analyze the iliosacral functional complex within some species of the two genera of the hylid subfamily Pseudinae (Duellman, 2001). Pseudinae is an essentially aquatic group of South American anurans with characters that lead to controversial interpretations, such as the presence of intercalary elements and an opposable thumb of the manus (Cei, 1980; Manzano, 1996). Functionally, opposable thumbs are associated more closely with an arboreal than an aquatic lifestyle, and are characters that are shared with other anuran groups, such as some Phyllomedusinae, Rhacophoridae, and Mantellidae. Although Pseudinae seem to be a monophyletic group, it is difcult to separate pseudids from the Hylidae and Centrolenidae (da Silva, 1998; Ford and Cannatella, 1993). Da Silva (1998) undertook a phylogenetic analysis

including Allophryne, Centrolenidae, Hylidae, and Pseudidae, and presented evidence that Pseudidae belongs within Hylidae, as a sister group of Hylinae. Subsequently, Duellman (2001) included Pseudinae as a subfamily of Hylidae. We accept Duellmans (2001) classication of Pseudinae. Nevertheless, there is still much to understand about the morphology and evolutionary relationships of this group, which includes variations of morphological novelties, (e.g., those of the iliosacral joint). In this study, we focus our interest on the analysis of the iliosacral anatomy of the species of Pseudinae, based on the two types of articulations dened by Emerson (1979, 1982). Given their aquatic habits and shared evolutionary history, unusual characters could be expected in the pseudine iliosacral articulation structure. MATERIALS AND METHODS We dissected and observed the iliosacral articulation and associated musculature of 25 specimens of Pseudinae, ve specimens of Centrolenidae, and six specimens of other Hylidae (see Appendix I). The specimens were subjected to differential staining methods to study musculature based on a modication of Wassersugs (1976) technique. The modication involves almost the same steps required for staining bones and cartilages, with differences in the time that the specimens remain in the solutions, depending on their size. Specimens previously skinned and preserved in 10% formaldehyde, were washed in tap water for a period of two hours before introducing them in a solution of Alcian Blue in Absolute Alcohol and Acetic Acid. After 12 days in the dye solution, the cartilage in a medium size frog, would take on a blue color; the specimens were then xed in Absolute Alcohol (3 h). In the next step, the material was introduced into a solution of Potassium Hydroxide for approximately 2 minutes to neutralize the acid of the Alcian Blue solution. After that, specimens were kept for 12 days in a solution of Red alizarin in Potassium Hydroxide. Once the specimen bones were dyed red, the specimen was preserved in 70% ethanol, avoiding muscle maceration. With this technique, bones stain red, cartilage stains blue, and muscles remain unstained and can be observed intact.

September 2005]

HERPETOLOGICA

261

FIG. 1.Pseudis minuta (FML:03676-1). (A) Photograph of dorsal view of the iliosacral articulation; (B) drawing of Figure 1A, dorsal view of the iliosacral articulation. dc, dorsal crest; il, iliac shaft; lc, ligamentous cap; sd, sacral diapophysis; ses, sesamoid; u, urostyle. Scales 5 1 mm.

To study the articulation in detail, we embedded two specimens in parafn for histological sections (Anderson and Bancroft, 2002), one of Pseudis minuta (FML03676) and one of Lysapsus limellus (DIAM 019), and stained them with hematoxilin-eosin according to the standard staining procedure for parafn sections (Wilson and Gamble, 2002). The specimens were examined with the aid of a Olympus stereo microscope equipped with a camera lucida and photographs were taken with a camera attached to the microscope. Pseudis minuta was used as a reference species, with descriptions of the articulation and morphology of its iliolumbaris muscle. The remaining species of Pseudinae and its subspecies were compared to P. minuta. We follow the terminology of Trueb (1973) for osteology and Gaupp (1896) for musculature. Two specimens of Pseudis minuta were videorecorded with a GR-VL505 JVC camera to allow analysis of the movements of the pelvic girdle during swimming and resting. A specimen of Xenopus laevis also was video-recorded for comparison.

RESULTS The Iliosacral Articulation Pseudis minuta (Fig. 1).The sacral diapophyses are narrow, dorsoventrally compressed (with an elliptic transversal section) and oriented posterolaterally. The distal margin of the sacral diapophysis is almost completely mineralized and not expanded. An oval, well-developed sesamoid is present on the lateral edge of the sacral diapophysis. The shaft of each ilium extends anteriorly well beyond the sacral diapophysis, reaching the mid-level of the Presacral Vertebra VIII. An expanded ligament arises from the anterior end on both margins of the iliac crest, and covers dorsally the anterior end of the iliac shaft. Laterally, the ligament encloses the terminus of each sacral diapophysis, forming a ligamentous cap and incorporating the sesamoid to this structure. This ligamentous cap inserts on the dorsal surface of the sacral diapophysis, including the anterior and posterior edges of the lateral terminus of this bony structure. The M. iliolumbaris originates as

262

HERPETOLOGICA

[Vol. 61, No. 3

FIG. 2.Pseudis paradoxa platensis (FML: 004661-1). (A) Photograph of dorsal view of the iliosacral articulation; (B) drawing of Fig. 2A, dorsal view of the iliosacral articulation. cd and dc, dorsal crest; il, iliac shaft; lc, ligamentous cap; sd, sacral diapophysis; ses, sesamoid; u, urostyle. Scales 5 1 mm.

a tendon on the ligamentous cap, from the anterior region of the iliac crest. Lateral processes of presacral Vertebrae VII and VIII are oriented anterolaterally and as long as the sacral diapophysis. The lateral processes of the resting presacral vertebrae are oriented posterolaterally. Other species of Pseudis.The subspecies of P. paradoxa (P. p. paradoxa, P. p. occidentalis and P. p. platensis) that we describe show an arrangement of the elements that conforms to the articulation described for P. minuta (Fig. 1). Nevertheless, the form and size of the sesamoid within P. paradoxa is more rounded than oval and relatively larger, being almost as long as the width of the lateral end of the sacral diapophysis (Fig. 2). In this species of Pseudis, the distal tips of each sacral diapophysis is at least partially mineralized. Within the subspecies of P. paradoxa, the articular ligament originally described by Emerson (1979, 1982) is not evident, but a ligamentus cap is present. The articular ligament of P. cardosoi and P. bolbodactyla, like that of P. minuta, has an insertion on the mid-dorsal region of each sacral diapophysis and is included in the ligamentous cap. Except for P. bolbodactyla, the

length of the iliac shafts of Pseudis species is very homogeneous, extending anteriorly to the sacral diapophysis and reaching the mid-level of presacral vertebra VIII. In P. bolbodactyla, the iliac shafts extend to the anterolateral border of the sacral diapophysis. There is an iliac crest present in all species studied, but with variation in size and length. In the histological sections of the iliosacral joint, we observed a combination between Type IIA and IIB articulation for the species of Pseudis (Fig. 3). The tip of the sacral diapophysis is curved ventrally, with a wedge-like sesamoid, located laterally (as in a Type IIA articulation). Although, in the histological transversal section, the form of the ilium and the diapophysis with its small cartilaginous tip, correspond to Type IIB articulation. The ligamentous cap envelopes the ilium ventrolaterally and the sesamoid laterally, and inserts on the mid-dorsal surface of the sacral diapophysis. Lysapsus limellus.The condition of Lysapsus limellus is intermediate between Types IIA and IIB (Emerson 1979, 1982). The sacral diapophyses are expanded laterally with a cartilaginous distal border that is larger

September 2005]

HERPETOLOGICA

263

FIG. 3.Lysapsus limellus (FML: 00725-1). (A) Photograph of dorsal view of the iliosacral articulation; (B) drawing of Fig. 3A, dorsal view of the iliosacral articulation. dc, dorsal crest; il, iliac shaft; illm, M. Iliolumbaris; la, articular ligament; lc, ligamentous cap; sd, sacral diapophysis; ses, sesamoid; u, urostyle. Scales 5 1 mm.

than that of Pseudis (see Figs. 3, 5). The ligamentous cap envelopes the lateral border of each sacral diapophysis, the anterior extreme of the iliac shaft, incorporating the sesamoid, and a very differentiated internal articular ligament. The articular ligament inserts on the mid-dorsal region of the sacral diapophysis and the insertion is not close to the vertebral body as in Emersons Type IIA iliosacral articulation. The lateral processes of

Presacral Vertebrae VIII is as long as the sacral diapophysis (as in Emersons Type IIB articulation) and is oriented strongly anterolaterally. The oval sesamoids are smaller than those of Pseudis, and cover only a third of the anterolateral edge of the sacral diapophysis (Fig. 4). The iliac shaft extends anteriorly as far as the posterior border of the Presacral Vertebra VII and has a small, short longitudinal crest. Histological sections of the pelvic girdle in Lysapsus, show a Type IIA iliosacral articulation (Fig. 5), and like Pseudis minuta, the tip of the sacral diapophysis is curved ventrally and the sesamoid is positioned laterally. The form and position of each sesamoid with respect to the ilium and the insertion of the ligament also is similar to the condition seen in Pseudis minuta (Figs. 3, 5). Ilium tips are completely enveloped by the ligamentous cap. The Muscle Iliolumbaris Pseudis minuta.In Pseudis minuta (Fig. 6), the Iliolumbaris muscle has a wide origin by a tendon on the ligamentous cap, positioned on the anterolateral region of the shaft of the ilium and the anterolateral portion of the sacral diapophysis. This muscle inserts on the lateral processes of Presacral Vertebrae VVIII, and

FIG. 4.Pseudis minuta (FML03676-2). Histological section of the right iliosacral articulation, transversal view. il, ilium; lc, ligamentous cap; sd, sacral diapophysis; ses, sesamoid; arrows show the insertion point of the ligamentous cap. Scale 5 0.5 mm.

264

HERPETOLOGICA

[Vol. 61, No. 3

FIG. 5.Lysapsus limellus (DIAM 019-2). Histological section of the left iliosacral articulation, transversal view. il, ilium; lc, ligamentous cap; sd, sacral diapophysis; ses, sesamoid; arrows show the insertion point of the ligamentous cap. Scale 5 0.5 mm.

lies parallel to the Longissimus Dorsi muscle, covering most of the external surface of the lateral processes of Presacral Vertebrae VI VIII. It is a thick, wide muscle with transverse tendinous inscriptions by which it inserts to the tip of each of the lateral processes of Presacral Vertebrae VVIII. An incipient division of the muscle into two branches at the origin was observed, with the lateral branch being more developed and thicker than the medial one. Other Pseudines.In Pseudis paradoxa platensis and P. p. occidentalis, the Iliolumbaris muscle originates on the ligamentous cap by a short tendon, as was observed in P. minuta. It is present as single muscle and, although two bundles of muscle bers are distinguishable, no division is observed. This muscle inserts on the lateral process of the Presacral Vertebrae IVVIII. Lysapsus limellus.In Lysapsus limellus, the Iliolumbaris muscle is undivided and has a narrow origin on the anterolateral portion of the iliac shaft. It inserts on the tips of each lateral process of Presacral Vertebrae IVVIII. Monitoring Locomotion in Pseudis minuta In the video-recording, we observed that Pseudis minuta swims by moving the posterior limbs without evident rotation of the iliosacral articulation. The leg motion is practically limited to exions and extensions of the tibiobula and

FIG. 6.Pseudis minuta. Dorsal view of the iliosacral articulation. m.il.ext., M. Iliacus Externus; m. iliolum.l., M. Iliolumbaris Lateralis; m. iliolum.m., M. Iliolumbaris Medialis; lc, ligamentous cap; ses, sesamoid; ssc, suprascapulae. Scale 5 2 mm.

foot (i.e., tarsometatarsus). When the animals foot is in contact with the substrate (a submerged rock or the aquarium oor), the swimming movement is propelled by a kick. Most of the time specimens of this species remain oating with their heads out of the water and their webbed feet expanded at the same level as the head, partially out of the water. The back is curved dorsally in a concave arc and the iliosacral articulation appears to form an angle smaller than 1708, compared with the iliosacral articulation of Rana catesbiana in a resting position (shown in Jenkins and Shubin, 1998: Figs. 9A, 10A). DISCUSSION The morphology of the iliosacral articulation in pseudines does not coincide completely with any of the types described by Emerson (1979, 1982). Although some characteristics would indicate a Type IIB iliosacral articulation in some pseudines, important variations make it inappropriate to place them in this category. Also, in some species, there is a combination of characteristics from both types of articulation (IIA and IIB).

September 2005]

HERPETOLOGICA

265

The ligamentous cap described herein has not been described previously, and may be unique to pseudines. In the case of Pseudis, the non expanded, cylindrical shape of the sacral diapophysis, the lateral processes of the presacral vertebrae that are as long as the sacral diapophysis, and the presence of an iliac crest would indicate a Type IIB iliosacral articulation (Emerson, 1979). However, the insertion of the ligamentous cap displaced from the lateral edges of the sacral diapophysis, as well as the absence of a well-dened ligament (Emerson, 1982), suggest variations in this pre-established pattern. Only in Lysapsus limellus was a well-dened ligament clearly differentiated from the ligamentous cap that generally is observed; the articular ligament was difcult to identify in Pseudis species, where it is included as a part of the ligamentous cap. In Pseudis paradoxa there is no evidence of the presence of the articular ligament. The insertion of the M. Iliolumbaris on the lateral processes of Presecral Vertebrae IV VIII only, instead of all the pre-sacral vertebrae, and their wide origins on the ligamentous cap (except in Lysapsus), at the same level as the lateral extremes of the sacral diapophysis, are other observed variations. These differences are more evident in Lysapsus limellus, where the sacral diapophysis are expanded, the sesamoid is small, and the insertion of the ligamentous cap is closer to the mid-region of the sacral vertebra than in Pseudis. This would suggest a Type IIA articulation, but the presence of the iliac crest and the long lateral presacral processes (such as in Pseudis) are characteristic of Type IIB articulation. Nevertheless, except for the presence of a dened ligament in Pseudis cardosoi, little variation in the morphology of the iliosacral articulation in Pseudis was observed, especially among the subspecies of Pseudis paradoxa. Bigalke (1927) mentioned a proximal bifurcation of the M. Iliolumbaris in Rana (R. esculenta, R. fusca, and R. arvalis) and describes an almost complete separation in two branches in Bufo bufo (as Bufo vulgaris) with or without tendinous origins. In pseudines, the branches of the M. Iliolumbaris have various degrees of differentiation; the medial branch is present only in species of Pseudis, and although in P. paradoxa the

medial branch is not separated from the lateral branch, it is identied easily. In Lysapsus limellus the muscle shows no branch separation. In all pseudines observed, the origin of the muscle is on the ligamentous cap. In the remaining hylids analyzed, the division of the muscle varies from an incipient bifurcation (Scinax acuminata) to a double muscle (Hyla pulchella and H. andina). In Centrolenidae the muscle is single and originates from the iliac shaft. All centrolenids exhibit a Type IIA articulation, but also with some variations. The sacral diapophysis are expanded, there is no crest in the iliac shaft, they have long ovate-shaped sesamoids (widest at the base), and the articulary ligament inserts on the distal edge of each sacral diapophysis (as in Type IIB). The ilium reaches the diapophysis anteriorly, not overlapping it. Emerson (1979) considers that the pattern of iliosacral articulation is family specic, except in Hylidae, Microhylidae, and Discoglossidae. Additionally, she recognize the existence of exceptions within Atelopus, Dendrobates, and Leptopelis, and associated them to the differences in locomotion of those species (Emerson, 1982, 1988). Similarly, the subfamily Pseudinae, recently classied within Hylidae (Duellman, 2001), not only lacks a single pattern for the whole subfamily, but also the articulations are not strictly referable to any of those types and subtypes described by Emerson. The iliosacral articulation in Pseudinae is intermediate between the articulation Types IIA and IIB, and shows interspecic variations. The ability to jump is present in some form in all anurans (Duellman and Trueb, 1986), but is limited by the exibility of the trunk and iliosacral articulation (Jenkins and Shubin, 1998; Trueb, 1996). The anurans that are good jumpers have a very exible iliosacral articulation (Green, 1931) with anteroposterior and rotational movements that, together with the sacral-urostilic articulation, align the vertebral column with the pelvic girdle during jumping (Jenkins and Shubin, 1998; Kargo et al., 2002). Some authors consider that aquatic anurans have a movement and design similar to those of saltatorial specialist jumpers, propelling themselves with a kick (Emerson, 1982; Gans and Parsons, 1965); although, others (Abourachid and Green, 1999) state that these could

266

HERPETOLOGICA

[Vol. 61, No. 3

be independently derived locomotory trends. Kargo et al. (2002) demonstrated that the iliosacral articulation is a gliding joint, in which the trunk translation and rotation are independent of one another. Emerson and de Jongh (1980) hypothesized that this translation of the trunk could be important during swimming, and is observed in frogs that are specialized for swimming. In pseudines, the movement during swimming is propelled mainly by the exion of the tibiobula and the foot (tarsometatarsus), and exion of the knee seems to be minimal or nonexistent. Except when the animal rests on the waters surface, it swims by a jumping-style movement of the legs. This kind of movement, which mainly involves the distal elements of the legs (tibiabula and tarsometatarsus), was described previously for Hymenochirus boettgeri (Gal and Blake, 1988), but for the adjustment of direction rather than propulsion. In a common jumping frog like Rana pipiens, the involvement of the distal elements of the hind limbs could increase the jump distance (Kargo et al., 2002). Interestingly, Pseudis also is a very good jumper during an explosive escape response (A.S. Manzano and M. Barg, personal observation). The iliosacral articulation acts as the main mechanical axis during jumping, transmitting the impulse from the legs towards the vertebral column. In this movement, the vertebral column aligns with the pelvic girdle and the angle of the iliosacral articulation rotates, from 1208 (in resting position), to 1708 (in jumping positions) (Callow and Alexander, 1973). During swimming, the iliosacral articulation seems to be less important comparing with the movements of the legs. In specimens of Pseudis that are resting and oating, the angle formed by the iliosacral joint seems to increase from the 1708 angle during swimming as a consequence of the dorsally curved, concave position of the back. In other resting, aquatic anurans, the vertebral column remains aligned with the pelvic girdle (e.g., Xenopus laevis, A.S. Manzano and M. Barg, personal observation). It is known that in Xenopus laevis the iliosacral joint is fused and akinetic (Jenkins and Shubin, 1998). Although, a broad, cuff-like ligament joining the ilium with the sacral diapophisis has been described in pipids (Trueb, 1996), the mor-

phology of iliosacral articulation in pseudines is quiet different. In pipids cuff-like ligament covers all the sacral vertebrae between the ilia. Among pipids, the greatly expanded sacral diapophyses and cuff-like ligaments restrict the movements of the iliosacral articulation to an anterior to posterior sliding movement. A combination of characters of Type IIA and IIB iliosacral articulation in Lysapsus limellus, like the expanded sacral diapophyses and long lateral processes of the Presacral Vertebras could reduce considerably the lateral rotating movements of the iliosacral articulation, but in this case (Lysapsus limellus), the lateral processes of the Presacral Vertebras are strongly anterolaterally oriented thus the lateral rotating movement are not limited. The presence of the ligamentous cap in the iliosacral articulation of the species of Pseudis could increase the range and angle articulation during the transition from resting to swimming positions, acting as a hinge and adding freedom to the movement of rotation. The variations in the morphology of the iliosacral articulation may or may not imply mechanical variations in its movement, but can contribute to the hyperextension of the back during rest.
Acknowledgments.We are very grateful to T. Burton, J. Faivovich, G. Perotti and I. Vaquilla for critically reading our manuscript. We thank V. Casco and F. Izaguirre for the histological sections and E. Lavilla for helping us with the photographs. We also thank the referees of Herpetologica for the valuable suggestions to improve the manuscript. Supported by PICT 12418 and PEI 6155.

LITERATURE CITED
ABOURACHID, A., AND D. GREEN. 1999. Origin of the frogkick? alternate-leg swimming in primitive frogs, families Leiopelmatidae and Ascaphidae. Journal of Herpetology 33:657663. ANDERSON, G., AND J. BANCROFT. 2002. Tissue processing and microtomy. Pp. 85107. In J. Bancroft and M. Gamble (Eds.), Theory and Practice of Histological Techniques, 5th ed. Churchill Livingston, Elservier Science Limited, London, U.K. te (Bufo BIGALKE, R. 1927. Zur Myologie der Erdkro vulgaris, Laurenti). Pp. 236353. In J. F. Bergmann and J. Springer (Eds.), Zeitschrift Fu r Anatomie und Entwicklungsgeschichte. Mu nchen, Berlin, Germany. CALLOW, L., AND R. MCNEAL ALEXANDER. 1973. A mechanical analysis of hindleg of a frog (Rana temporaria). Zoological Journal of London 171:293321. CLARKE, B. T. 1988. Evolutionary Relationships of Discoglossids FrogsOsteological Evidence. Ph.D Dissertation, City of London Politechnic and British Museum (Natural History), London, U.K.

September 2005]

HERPETOLOGICA

267

CANNATELLA, D., AND L. TRUEB. 1988. Evolution of Pipoid frogs: intergeneric relationship of the aquatic frog family Pipidae (Anura). Zoological Journal of the Linnean Society 94:138. CEI, J. M. 1980. Amphibians of Argentina. Monitore Zoologico Italiano (N.S.) Monograa 2:1609. DA SILVA, H. 1998. Phylogenetic Relationships of the Family Hylidae with Emphasis on the Relationships within the Subfamily Hylinae (Amphibia: Anura). Ph.D. Thesis, The University of Kansas, Lawrence, Kansas, U.S.A. DUELLMAN, W. 2001. The Hylid Forgs of Middle America, Vol. 2. Society for the Study of Amphibians and Reptiles. Ithaca, New York, U.S.A. DUELLMAN, W., AND L. TRUEB. 1986. Biology of Amphibians. McGraw-Hill, New York, New York, U.S.A. EMERSON, S. 1979. The iliosacral articulation in frogs: form and function. Biological Journal of the Linnean Society 11:153168. . 1982. Frog postcranial morphology: identication of a functional complex. Copeia 1982:603613. . 1988. Convergence and morphological constraints in frogs: variation in postcranial morphology. Fieldiana: Zoology 43:119. EMERSON, S., AND H. J. DE JONGH. 1980. Muscle activity at the iliosacral articulation on frogs. Journal of Morphology 166:129144. FORD, L., AND D. CANNATELLA. 1993. The major clades of frogs. Herpetological Monographs 7:49117. GAL, J. M., AND R. W. BLAKE. 1988. Biomechanics of frog swimming II. Mechanics of the limb-beat cycle in Hymenochirus boettgeri. Journal of Experimental Biology 138:413429. GANS, C., AND T. S. PARSONS. 1965. On the origin of the jumping mechanism in frogs. Evolution 20:9299. GAUPP, E. 1896. A. Eckers und R. Wiedersheims anatomie des frosches. Braunschweg Friedrich: Friedrich Vieweg und Sohn 2:1961. GREEN, T. L. 1931. On the pelvis of the Anura: a studying adaptation and recapitulation. Proceedings of the Zoological Society of London 1931:12591290. HEYER, W. R. 1975. A preliminary analysis of the intergeneric relationships of the frog family Leptodactylidae. Smithsonian Contribution to Zoology 199:155. INGER, R. 1972. Bufo of Eurasia. Pp. 102111. In W. F. Blair (Ed.), Evolution in the Genus Bufo. University of Texas Press, Austin, Texas, U.S.A. JENKINS, F., AND N. SHUBIN. 1998. Prosalirus bitis and the anuran caudopelvic mechanism. Journal of Vertebrate Paleontology, 18:495510. KARGO, W. J., F. NELSON, AND L. C. ROME. 2002. Jumping in frogs: assessing the designs of the skeletal system by anatomically realistic modeling and forward dynamic simulation. The Journal of Experimental Biology 205: 16831702. LYNCH, J. D. 1973. The transition from archaic to advanced frogs. Pp. 133182. In J. L. Vial (Ed.), Evolutionary Biology of the Anurans. University of Missouri Press, Columbia, Missouri, U.S.A. lisis de la musculatura de la MANZANO, A. S. 1996. Ana familia Pseudidae (Amphibia: Anura). Tesis Doctoral. n, Tucuma n, Universidad Nacional de Tucuma Argentina.

GENER, L. A., AND A. M. MAGLIA. 1997. Osteology and PU skeletal development of Discoglossus sardus (Anura: Discoglossidae). Journal of Morphology 233:267286. TRUEB, L. 1973. Bones, frogs and evolution. Pp. 65132. In J. L. Vial (Ed.), Evolutionary Biology of Anurans. University Missouri Press, Columbia, Missouri, U.S.A. TRUEB, L. 1996. Historical Constraints and morphological novelties in the evolution of the skeletal system of pipids frogs (Anura: Pipidae). Pp. 349377. In R. C. Tinsley and H. Kobel (Eds.), The Biology of Xenopus. The Zoological Society of London, Clarendon Press, Oxford, U.K. WASSERSUG, R. 1976. A procedure for differential staining of cartilage and bone in whole formaline-xed vertebrates. Staining Technology 51:131134. WHITTING, H. 1961. Pelvic girdle in amphibian locomotion. Symposium of the Zoological Society of London 5:4357. WILSON, I., AND M. GAMBLE. 2002. The hematoxylins and eosins. Pp. 125137. In J. Bancroft and M. Gamble (Eds.), Theory and Practice of Histological Techniques, 5th ed. Churchill Livingston, Elservier Science Limited, London, U.K.

Accepted: 27 March 2005 Associate Editor: Christopher Sheil

APPENDIX I
Specimens Examined Specimens belong to the Herpetological Collections of Museu de Historia Natural, Campinas, Sao Paulo, UNICAMP (ZUEC); Museu de Ciencias e Tecnologia da PUCRS, Brazil (MCP); Carnegie Museum (Carn. Museum); Museo Argentino de Ciencias Naturales (MACN); Funda n Miguel Lillo (FML); Universidad de Corrientes cio os, (UNNEC); CICyTTP-CONICET Diamante, Entre R n La Salle de Ciencias Argentina (DIAM); Fundacio Naturales, Venezuela (EF). Family Hylidae: Subfamily PseudinaePseudis bolbodactylus ZUEC 11800, ZUEC 11801; P. caribensis EF-112, 13554; P. cardosoi MCP 3375, MCP 3775; P. paradoxa paradoxa Carn. Museum 49512; P. paradoxa occidentalis MACN 37698, MACN 37699, FML00708 (2 specimens); P. minuta MACN 37700, MACN 37701, MACN 37702, FML03676 (1 specimen for histological sections, 1 for photograph); P. paradoxa platensis UNNEC 03455, FML00936 (2 specimens); P. paradoxa FML04661 (2 specimens, photographs); Lysapsus limellus FML 00791 (photograph), FML 00725, DIAM 019 (2 specimens, one for histological sections). Two living specimens of Pseudis minuta, male and female, were video-recorded and photographed; they remain in an aquarium in M. Barg possession at the University of Mar del Plata, Argentina. Subfamily PhyllomedusinaePhyllomedusa hypochondrialis FML 04286. Subfamily HylinaeScinax nasicum DIAM 023; S. squalirostris DIAM021; Phrynohyas venulosa DIAM 024; Hyla andina DIAM 022; H. pulchella DIAM 038. Family CentrolenidaeHyalinobatrachium aureogutattum DIAM 055; Centrolene robledoi DIAM 056; C. geckoideum DIAM 075; C. grandisone DIAM 076; Cochranella ignota DIAM 057.