Scientific Papers

Natural History Museum The University of Kansas

24 July 1998 Number 8:1-12

Geographic Variation in Bufo valliceps (Anura: Bufonidae), a Widespread Toad in the United States and Middle Amexica

By JosePH R. MENDELSON III’

Natural History Museum and Department of Systematics and Ecology, The University of Kansas, Lawrence, Kansas 66045-2454, USA

CONTENTS JNTBISSIRI RUAN Co RO ee oe oe Cree eee 1 LIN pI UEX © TP) U8) THT IN Fea cece sete evsoe ec ine eden ieee onsesee tases D PNGKN@WIFEDXG MIEN Sree esett race serrer sc oat essere ee oa ere or SE TR Ren SEER E SENS 3 MATERIALS AND METHODS ..........ccccecccesccescecsscceseeecceseeesseenaeenaes 3 [UES COATING). 5 eal rR ERS Eg eRe PR Fa Sen 4 (CEOGRAPHIGAVARTATIONUINIGIZE |< -sserccossnccsosersnce caseoerts coectennecccenceeseneee 4 @WATITATIVE) COMPARISONS \secccecsses crese ses san tesctesconvaoeiuccen waceneasesmneeee 6 IB) ISG@WSSIO Nip serra sce tsee reseed ss sera eae Me 8 BODY SIZEFANDYARIDIMY. Ssesoesexn es cer ee a dae Beek eee 8 GRESTHEIRIGHIVANDIETEVATION Gs: .cccsoceesscessconcs escsssecoccesecesscscsccsees tes 9 RIVA ONOMMUGEISSUIES Hose ase tees eoarc a ete n canta ete pees Sevan arate tee tboseetosene 10 ET NG ES RR/ANSTHUG) FRIES) TAT NID) reese eee ccc t esses cnccsseceass cos cesecestocsseccsescesticees 11 FATT Tes INY IB) [DXGpemersnte aetna eB NR eo Fee ses cats seace seiconsecaheoseehenencerenss® 11

ABSTRACT The common lowland toad Bufo valliceps has a large distribution in the southern United States, Mexico, and most of Central America; this ample distribution across diverse temperate and tropical habitats is unusual among frogs. Geographic variation in size, shape, skin texture, and color pattern among populations of this species was reviewed. Although there are great differences between extreme northern and southern popula-

‘Present address: Department of Biology, Utah State University, Logan, Utah 84322-5305, USA.

SCIENTIFIC PAPERS, NATURAL History Museum, THE UNIVERSITY OF KANSAS

tions, I did not find smooth clinal variation for any character examined. Nor is there a discrete break among these continuous morphological variables that separate the the northern and southern mor- phs. Variation in Bufo valliceps is characterized by a high degree of inter— and intrapopulational varia- tion that cannot be attributed to simple trends associated with latitude, elevation, or climate. An analysis the relationship between body size and aridity, along a precipiation gradient on the Yucatan Penin- sula, found no consistent covariation. Previous claims for a postive relationship between elevation and degree of development of the cranial crests were found not to be accurate. Although the several relatively distinct populations are referrable to the taxa Bufo nebulifer and Bufo valliceps wilsoni, | do not propose recognition of these taxa, pending further research.

Key words: Bufonidae; Bufo valliceps; Geographic variation; Middle America; Systematics.

INTRODUCTION

Bufo valliceps is a ubiquitous species that occurs in open habitats from extreme southwestern Mississippi, across Texas to the Big Bend region, southward along the Atlan- tic coast to extreme northeastern Costa Rica, and across the Isthmus of Tehuantepec and southeastward along the Pacific coasts of Chiapas and Guatemala, at elevations from sea level to 1700 m (Fig. 1). This distribution includes an enormous variety of habitats and climatic regimes—e.g., relatively aseasonal rainforests of eastern Nicaragua; dry and seasonally cold Chihuahuan desert of Val Verde County, Texas; wet and seasonally cool swamps of south- ern Louisiana; cool and perpetually wet Atlantic slopes of the Chiapas highlands in Mexico; and hot and seasonally dry thorn forest in the upper Grijalva River Basin of Huehuetenango, Guatemala. However, B. valliceps seems to be ubiquitous in open habitats, whether the grasslands are natural (e.g., the Texas Gulf Coast) or the secondary growth and pastures resulting from human activities in areas of former rainforest in eastern Guatemala (Mendelson, 1994). The habitat preference of B. valliceps resembles that of the well-known human commensal Bufo marinus (Zug and Zug, 1979) in that the toad is often abun- dant in villages and agricultural areas.

Porter (1962, 1963, 1964, 1970) reviewed the taxonomic status, distribution, and geographic variation of Bufo valliceps, with special reference to Mexican populations. He acknowledged a great amount of morphological varia- tion across its geographic range, especially with respect to overall size, skin texture, and development of cranial crests. Nevertheless, Porter (1970) chose not to recognize any of the subspecies that had been proposed—viz., B. valliceps macrocristatus Firschein and Smith from the Atlantic rainforests of Oaxaca; B. valliceps wilsoni Baylor and Stuart from the Grijalva Valley of Chiapas and adjacent Guate- mala; and B. valliceps microtis Werner from Honduras. Sub- sequent to Porter’s papers, more specimens from remote regions of southern Mexico have been collected; these specimens allowed for resolution of taxonomic problems pertaining to certain populations that were referred by Porter and others variously as B. cavifrons, B. cristatus, and B. valliceps. Mendelson (1994, 1997a,b) reviewed much of this newer material and recognized or described B.

macrocristatus, B. campbelli, B. spiculatus, B. tutelarius, all of which had previously been confused, at least partially, with B. valliceps. Mendelson (1998) referred the taxon B. valliceps microtis to the synonymy of B. coccifer. Although it is clear that some of the morphological variation observed by Por- ter is attributable to species-level variation that occurs among these newly recognized species, the fact remains that B. valliceps has a large geographic distribution and exhibits remarkable morphological variation.

In the northern areas of its range, B. valliceps are larger than are conspecifics in southern Mexico and Central America (Porter, 1970). Blair (1963, 1972) commented gen- erally on differences in dorsal color, throat color in males, and size among populations from northern and southern extremes of the range. Porter (1964) summarized geo- graphic variation in snout-vent length (SVL) and charac- teristics of the advertisement calls of males for populations from Texas through southern Mexico, and Porter (1962) summarized other morphological variation in these same populations. The remarkable level of variation in color patterns among individuals of B. valliceps from Nicaragua and the Yucatan Peninsula were discussed by Villa (1972) and Lee (1993), respectively. Finally, several authors (Por- ter, 1963, 1964; Blair, 1966, 1972; Branson, 1995) have re- ported that the height of the cranial crests is directly corre- lated with the elevation at which a particular population occurs; this claim warrants specific attention and is dis- cussed in more detail.

Herein, I summarize geographic variation in Bufo valliceps, and replicate part of Lee’s (1993) analysis of varia- tion in anuran body size along the precipitation gradient on the Yucatan Peninsula. I also address other apparent trends in morphological variation among populations of B. valliceps, and I discuss the remaining taxonomic issues of the status of B. valliceps wilsoni and Bufo nebulifer Girard. This study is a complement to those of K. R. Porter (1962, 1963, 1964, 1970), but differs by considering the entire range of the species, excluding several populations that are now known to represent different species (Mendelson, 1997b, 1998), and by being based on more specimens than were available to Porter.

GEOGRAPHIC VARIATION IN BUFO VALLICEPS 3)

Gulf of Mexico

Pacific Ocean

Fig. 1. Geographic distribution of Bufo valliceps, modified from Por- ter (1970) and excluding erroneous records from El Salvador and west- ern Nicaragua (Frost, 1985:52, 63). Bufo valliceps is absent in all areas above 1700 m of elevation.

ACKNOWLEDGMENTS

I am grateful to the curators and collection managers of the following institutions for allowing me access to materials under their care: American Musuem of Natural History, California Academy of Sciences, Carnegie Mu- seum of Natural History, Field Museum of Natural His- tory, Berlin Museum ftir Naturkunde, UNAM Museo de Zoologia, University of Colorado, University of Illinois Natural History Museum, United States National Museum, University of Michigan Museum of Zoology, University of Kansas Natural History Museum, University of Texas Arlington Collection of Vertebrates, University of Texas at El Paso, University of Texas Natural History Collection, Texas A & M Cooperative Wildlife Collection. The follow- ing persons provided every variety of advice, discussion,

Gulf of Mexico

Fig. 2. Map showing the location of the 19 samples of Bufo valliceps studied in the range-wide morphometric analyses: (1) Baton Rouge, East Baton Rouge Parish, Louisiana; (2) Walker County, Texas; (3) Tarrant and Dallas counties, Texas; (4) Austin, Travis County, Texas; (5) Liverpool, Brazoria County, Texas; (6) Brownsville, Cameron County, Texas; (7) Val Verde County, Texas; (8) Don Martin Dam, Coahuila; (9) Huejutla, Hidalgo; (10) Cuautlapam, Veracruz; (11) Acayucan, Veracruz; (12) Tuxtla Guitierrez, Chiapas; (13) Chinaja, Alta Verapaz, Guatemala; (14) Escarcega, Campeche; (15) Pueblo Nuevo X-Can, Quintana Roo; (16) Tucuru, Alta Verapaz, Guatemala; (17) Champerico, Retalhuleu, Guate- mala; (18) San Pedro Sula, Cortés, Honduras; (19) Condega, Esteli, Nica- ragua. Inset shows the mean snout-vent length and 95% confidence in- tervals for populations. Sample sizes for each population are shown in Table 4.

and assistance: J. A. Campbell, D. C. Canatella, H. R. da Silva, W. E. Duellman, A. Graybeal, A. M. Maglia, D. J. Meinhardt, J. Pramuk, A. Nieto-Montes de Oca, C. Sheil, N. Slade, A. Sproston, and L. Trueb. Comments on this manuscript were provided by W. E. Duellman, A. Maglia, S. Ron, and L. Trueb.

MATERIALS AND METHODS

I examined specimens of Bufo valliceps from all areas of its range between Louisiana and Esteli, Nicaragua; I have not seen specimens from Mississippi, Arkansas, or Costa Rica, but rely on reports by others (Conant and Collins, 1991; J. M. Savage, pers. comm.) for the accuracy of these records. Note that for localities from the USA or Mexico only state or county designations are provided. The mor-

phometric survey included only adult males from 19 popu- lations (Fig. 2); sexual maturity was assessed by presence of nuptial excrescences and vocal slits. The following mor- phometric features were measured in a manner following that of Duellman (1970:fig. 2): snout-vent length (SVL); tibia length (TIB); foot length (FTL); head length (HL); head width (HW); tympanum width (TYMP). The following

4 SCIENTIFIC PAPERS, NATURAL History Museum, THE UNIVERSITY OF KANSAS

Table 1. Results of the first four principal components explaining 89% of the variation in a PCA (covariance matrix) performed on 15 log-trans-

each variable are also shown.

formed morphometric variables measured on male Bufo valliceps representing 19 populations from throughout the range of the species. Loadings of

Variable PC 1 PE? PE3 PC 4 Eigenvector 0.057243 0.003921 0.002469 0.001992 Proportion of variance 0.780 0.053 0.034 0.027 Cumulative proportion 0.780 0.833 0.867 0.894 SVL -0.258 0.023 0.078 0.113 FML -0.292 -0.066 0.248 0.042 ML -0.252 0.081 0.058 -0.074 RUL -0.234 0.134 0.177 -0.016 HDL -0.246 0.032 0.102 -0.042 TIB -0.252 0.102 0.389 -0.569 FTL -0.246 0.017 0.147 -0.128 HL -0.258 0.026 0.120 0.072 HW -0.284 0.089 0.106 0.055 ORB -0.188 0.041 0.091 0.088 END -0.234 0.085 -0.122 0.462 TYMP -0.263 -0.135 0.083 0.577 SPTYMP -0.270 0.660 -0.608 -0.109 PARL -0.308 -0.418 0.460 -0.279 -0.267 -0.560 -0.277 -0.091

PARW

features were measured in a manner following that of Mendelson (1994): orbit diameter (ORB); eye—nostril dis- tance (END); length of supratympanic crest (SPTYMP); length of parotoid gland (PARL); width of parotoid gland (PARL). The following features were measured in a man- ner following that of Lee and Crump (1981): femur length (FML); metatarsal length (ML); radioulna length (RUL);

hand length (HDL). All measurements were taken with digital calipers, rounded to the nearest 0.1 mm, and log- transformed. Principal components analysis (PCA; cova- riance matrix), One-Way ANOVA, and Tukey’s Method for Unplanned Comparisons were performed using MINITAB (Macintosh ver. 10.5; Minitab Statistical Software, 1995). Stepwise discriminant function analysis (DFA) was per- formed using the BMDP computer program.

RESULTS

GEOGRAPHIC VARIATION IN SIZE

Bufo valliceps from the northern part of the range are substantially larger than most individuals of the species from southern Mexico and Central America. Differences in mean SVL among the 19 populations are shown in Fig- ure 2; ANOVA indicated significant differences among these populations (F = 42.29; df = 18; P < 0.0001).

Range-wide comparisons.—The loadings resulting from a PCA of the 15 morphometric variables measured on adult males representing these 19 populations are displayed in Table 1. The first four principal components (PC) accounted for 89.4% of the variation. All of the loadings on the first PC were of the same sign and of similar magnitude and, therefore, seem to represent overall size and size-correlated variation among all variables. The remaining PCs repre- sented nonsize-correlated variation within the sample. The PC I has a strong positive loading for SPTYMP and high negative loadings for PARL and PARW. The PC III also had similar high loadings for SPTYMP (negative loading) and PARL (positive loading), as well as relatively high negative loadings for PARW, FML, and TIB. The PC IV had a high negative loading for TIB and high positive load- ings for END, and TYMP. Despite these strong loadings, plots of individual scores on PC I-IV showed little disper- sion among the representatives of the populations; PC II-

IV explain relatively little variation. Therefore, I directed my efforts toward comparison of overall size using PC I as an indicator.

An ANOVA of scores on PC I revealed significant dif- ferences in overall size among the 19 populations between Texas and Esteli, Nicaragua (F = 37.95; df = 18; P < 0.0001). Tukey’s Method identified homogenous sets of these popu- lations based on overall size (Fig. 3) and demonstrated a general trend of size-increase with increased latitude. How- ever, this trend is not entirely consistent. For example, the sample from Brazoria County, on the Gulf Coast of Texas, is contained within sets that include samples from Guate- mala and Nicaragua; likewise, the sample from Louisiana is contained within sets including populations from south- ern Mexico and Guatemala. The largest toads in this analy- sis form a small set of two samples from central and west- ern Texas. The sets containing the next overall largest toads included samples from most of the northern populations (excluding Louisiana and Brazoria County, Texas), as well as samples from Huejutla, Hidalgo (south of the Tropic of Cancer) and from Pueblo Nuevo X-Can, Quintana Roo, at the northeastern tip of the Yucatan Peninsula. A mid-sized group of sets includes a variety of disjunct localities in- cluding Louisiana, southern Texas, Cuautlapam (Veracruz), and two samples from near the base of the Yucatan Penin-

GEOGRAPHIC VARIATION IN BUFO VALLICEPS 5

Table 2. Results of the first four principal components explaining 81% of the variation in a PCA (covariance matrix) performed on 15 log-trans- formed morphometric variables measured on male Bufo valliceps representing nine populations from across the length of the Yucatan Peninsula. Loadings of each variable are also shown.

Variable Pel PC 2 PC3 PC 4

Eigenvector 0.017310 0.003854 0.002758 0.002265 Proportion of variance 0.537 0.120 0.086 0.070 Cumulative proportion 0.537 0.657 0.742 0.813 SVL -0.237 -0.089 0.017 0.034 FML -0.308 0.931 0.151 0.053 ML -0.247 -0.048 -0.010 -0.070 RUL -0.215 -0.095 -0.025 0.071 HDL -0.257 -0.118 0.027 0.110 TIB -0.182 -0.079 0.006 0.025 FTL -0.193 -0.146 -0.014 0.010 HL -0.235 -0.122 0.099 -0.071 HW -0.300 -0.121 -0.083 0.756 ORB -0.317 0.043 -0.767 -0.435 END -0.315 -0.171 0.608 -0.451 TYMP -0.261 -0.079 0.017 0.003 SPTYMP -0.276 -0.050 0.004 0.059 PARL -0.243 -0.043 -0.007 0.012 PARW -0.239 -0.040 0.007 -0.031

sula. The sets containing the smallest toads included samples from northern (Brazoria County, Texas), interme- diate (Acayucan, Veracruz, and the Grijalva Valley of Chiapas and adjacent Guatemala), and southern localities (Nicaragua and the Pacific Coast of Guatemala). These re- sults suggest that in the northern area of their range, B. valliceps are larger than those in the southern areas, but there is no smooth north-south cline in overall size.

These results provide conflicting evidence with respect to the hypothesis that anurans from more xeric environ- ments are larger than those from more mesic areas. For example, the largest toads are from areas that receive little mean annual rainfall: 855 mm (Dallas, Texas; 1961-1990), 810 mm (Austin, Travis County, Texas; 1961-1990), and 472 mm (Del Rio, Val Verde County, Texas;1961-1990); a fe- male (UTA A-7228) collected by J. A. Campbell near Fort Worth, Texas, has an SVL = 121.3 mm. However, toads from areas receiving similarly scant amounts of rainfall were much smaller (Figs. 2, 3)—e.g., for example: 675 mm, Brownsville, southern Texas; 832 mm, Cintalapa, Chiapas (= Grijalva Valley; rainfall data from Johnson, 1990). The smallest toads did come from rather wet areas: 1220 mm (San Pedro Sula, Depto. Cortes, Honduras; 1944-1982); 2500 mm (Tapachula, Chiapas, [near Champerico, Guate- mala]; Shelford, 1963); and 1740 mm (Veracruz, Veracruz [near Acayucan]; 1971-1980). However, Chinaja, Alta Verapaz, Guatemala, receives about 4000 mm of rain each year (Duellman, 1963), and the sample of toads from this locality was not among the smallest in this study (Figs. 2, 3). Size variation along a precipitation gradient.—The load- ings resulting from PCA of the 15 morphometric variables from nine populations over the length of the Yucatan Pen-

insula are displayed in Table 2. The first four PCs accounted for 81.3% of the variation among the specimens included in the analysis. As with the overall analysis, all of the load- ings on PC [are of the same sign and of similar magnitude and, therefore, seem to represent overall size and size-cor- related variation among all variables. The remaining PCs represent nonsize-correlated variation within the sample. The second PC has a high positive loading for FML, PC III has high negative loadings for ORB and END, and PC IV has a high positive loading for HW. Despite these strong loadings, plots of individual scores on PC I-IV showed little dispersion among the representatives of the nine populations. However, there are significant differences in overall size among the samples of B. valliceps, as indicated by the results of an ANOVA on PCI scores (F = 6.07; df = 8; P < 0.0001). The mean score for PC I and standard devia- tion for each population, and the homogeneous subsets of populations identified by Tukey’s Method are shown in Figure 4. Toads from the more xeric northwestern portion of the peninsula (e.g., Mérida, Chichén Itza) are not sig- nificantly larger than those from the rainforests of Chinaja, Guatemala (Fig. 4). In fact, the two populations that com- pose a distinct subset in this analysis are from opposite ends of the peninsula— La Libertad, El Petén, Guatemala, and Pueblo Nuevo X-Can, Quintana Roo. La Libertad lies ina savanna habitat having a relatively high, but seasonal, rainfall, whereas Pueblo Nuevo X-Can lies in a region that has unusually high rainfall compared to the rest of the northern peninsula (Lee, 1980).

Discriminant Function Analysis.—In order to identify which morphometric measurements most effected sepa- ration among groups and to determine degree of

6 SCIENTIFIC PAPERS, NATURAL History Museum, THE UNIVERSITY OF KANSAS

16 |

19 |

| | | | | | | -5.0 -48 -4.6 -4.4 -4.2 -40 -3.8 PC | Scores Large < > Small Fig. 3. Mean scores with 95% confidence intervals on Principal

Component I for each sample population of Bufo valliceps (Fig. 2), ranked by magnitude. The vertical bars to the right indicate homogeneous sub- sets of samples identified by Tukey’s Method. Sample sizes for each popu- lation are shown in Table 4.

morphomeric distinctiveness among samples, I performed an initial stepwise DFA of the 15 morphometric variables from the 19 sample populations of (Fig. 2). All variables except TIB varied significantly among groups (F < 0.05) in the initial analysis; a second analysis that specified the 14- variable model for the canonical discriminant analysis was performed. Group means were different (F < 0.001) at each step in the 14-variable model and 100 % of the variation was displayed on five canonical axes; the first two axes displayed 99 % of the variation (CAN I, CAN II; Fig. 5). Review of the standardized (pooled within-group vari- ances) coefficients for the canonical variables (Table 3) re- veals that dispersion on CAN I is primarily because of variance among groups in the following variables: snout- vent length; foot length; head width; orbit diameter; and eye-nostril distance. Dispersion on CAN II also was caused by variance in snout-vent length, foot length, and head width, but this axis also displayed dispersion caused by variance among groups in head length and parotoid width.

None of the 19 sample populations had 100 % correct classifications in the jackknifed classification matrix of the

Sample

On ao fF ow DN OO

———

45 43 41 40 PC | Scores Large <—————_> Small

Fig. 4. Map of the Yucatan Peninsula showing location of nine samples of Bufo valliceps used in the morphometric analysis along the Yucatan precipitation gradient [sample sizes in brackets]: (1) Chinaja, Alta Verapaz, Guatemala [19]; (2) La Libertad, El Petén, Guatemala [4]; (3) Tikal, El Petén, Guatemala [9]; (4) Escarcega, Campeche [19]; (5) Champoton, Campeche [7]; (6) Dzibalchen, Campeche [4]; (7) Mérida, Yucatan [6]; (8) Chichen Itza, Yucatan [10]; (9) Pueblo Nuevo X-Can, Quintana Roo [12]. Inset shows mean scores and 95% confidence inter- vals for each sample on Principal Component I, ranked by magnitude. The vertical bars to the right indicate homogeneous subsets of samples identified by Tukey’s Method.

14-variable model (Table 4). The samples with the highest percentage of correct classifications were Sample 11 (92.3%), Sample 5 (91.7%), Sample 1 (87.5%), and Sample 12 (82.1%). The samples with the lowest percentage of cor- rect classifications were Sample 2 (13.3%); Samples 7 and 8 (25.0%), and Sample 3 (35.7%). The pattern of misclassifications (Table 4) indicates a slight geographic trend among the samples. Northern toads (Samples 1-8) tended to be misclassified as members of other northern samples. Toads from more southerly samples (Samples 6, 9) were primarily misclassified as members of either more northern or more southern samples, and southern toads (Samples 10-19) tended to be misclassified usually as mem- bers of other southern populations, but also as members of any but the largest northern samples (Samples 3, 4, 7, 8).

QUALITATIVE COMPARISONS

There is considerable variation in external features of Bufo valliceps. As with the morphometric variation de- scribed above, there are some geographic trends, but in- ter-individual variation in dorsal pattern and skin texture obfuscate discrete characterizations of the different popu- lations.

Toads from northern localities (e.g., USA, northern Mexico) are remarkably consistent in having a sharply

GEOGRAPHIC VARIATION IN BUFO VALLICEPS 7

Canonical Axis ||

Ou Oe 4 eae ees

Canonical Axis |

Fig.5. Plot of mean canonical discriminant scores for 19 samples of Bufo valliceps. Circles represent approximate 95% confidence intervals around each mean score. For purposes of clarity, prediction intervals are not shown, but these have a diameter of 2.45 units on each axis and, therefore, would indicate substantial overlap among many groups. Sample sizes for each population are shown in Table 4.

granular dorsal skin texture, granular ventral skin, rela- tively small, ovoid or subtriangular parotoid glands, an indistinct lateral descending row of tubercles, and a rela- tively wide, distinct, middorsal pale stripe (Fig. 6). The dorsal pattern lateral to the middorsal stripe may be strongly marbled, with either dark brown or tan-yellow shades predominating, mostly pallid, or nearly uniform dull brown; usually, there is a distinct dorsolateral pale area, along the dorsal border of the lateral descending row of tubercles. The venter is always immaculate and may appear (in preservative) pale yellow, whitish, or dull tan. The lateral descending row of tubercles is bordered ven- trally by a dark brown stripe that becomes thin or broken on the flanks.

The tubercles on all dorsal surfaces of the body are large, densely distributed, sharply pointed, and kerati- nized. Usually, there is a single keratinized apex sur- rounded by a granular patch of keratin and often these are coalesced into a keratinous blotch from which a pointed center arises. These tubercles give the toad a rough, granu- lar texture. The ventral skin texture is granular with many tiny, pointed tubercles, each bearing a single keratinized apex. Males and females also have heavy layers of keratin on the peripheral surfaces of all cranial and facial crests. The parotoid glands may be relatively large and ovoid, as in individuals from the Grijalva Valley, in Chiapas and adjacent areas of Huehuetenango, Guatemala (discussed

Table 3. Standardized (pooled within-group variances) coefficients for the canonical variables on the first two canonical axes; cumulative pro- portion of dispersion (variance) displayed on each axis shown in paren- theses.

Variable Canonical Axis I Canonical Axis II

(0.987) (0.995) SVL -0.79626 0.65282 FML 0.29343 -0.01370 ML 0.36157 0.04695 RUL 0.36908 -0.29779 HDL 0.04474 -0.25055 FT -0.85926 0.80126 HL 0.40648 1.13339 HW -0.61490 -1.64387 ORB 0.72146 -0.20203 END 0.61903 -0.22798 TYMP 0.20178 0.30574 SPTYMP -0.33749 0.31473, PARL 0.34510 0.20732 PARW -0.48930

-0.76280

below under B. valliceps wilsoni), or relatively small and distinctly triangular, as in some individuals from the Yucatan Peninsula. Most individuals, regardless of origin, have ovoid or subtriangular parotoid glands.

The color patterns among Bufo valliceps from southern Mexico and Central America are extraordinarily variable, even among individuals from a single locality. In some, the dorsal pattern is uniformly dark brown, or nearly so, with or without a thin, middorsal pale stripe, whereas in others, it is nearly uniformly tan and lacks dorsal mark- ings. Other individuals are pale brown with distinct black or dark brown, paired markings, with or without a dis- tinct interorbital bar, or pale brown with a complex marbled pattern of small dark brown blotches. Nearly every inter- mediate pattern and combination of the above extremes may be found among B. valliceps from Veracruz to Nicara- gua. Laterally, all individuals have a dark area below the lateral descending row of tubercles. This dark area is wide, dark brown or pale gray, and extends from the tympanic area, or the level of the insertion of the arm, onto the flank; in some individuals, the dark area becomes diffuse and disappears anterior to the flank. The ventral pattern usu- ally is dull cream with any number of diffuse or distinct dark markings that sometimes are expressed as a uniform dull gray area on the throat and pectoral area; few indi- viduals have immaculate venters.

Southern toads have a smoother skin texture than do those from the northern areas. The tubercles on the dor- sum of the body of southern toads are less numerous, low, round and bear a small patch of granular keratin; few in- dividual tubercles are pointed, and some specimens have little keratin atop the dorsal tubercles. The ventral skin texture is smoother than in the northern toads, with many low, round tubercles; some southern individuals have

8 SCIENTIFIC PAPERS, NATURAL History Museum, THE UNIVERSITY OF KANSAS

Table 4. Jackknifed classification matrix from the 14-variable model generated by the stepwise DFA of 19 sample populations of Bufo valliceps. The bold figures indicate the number of specimens correctly classified; the number in parentheses indicates the number of specimens in that sample.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

1 |7 (8) 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 2 O 2(15) 2 2 2 0 4 2 0 0 0 1 0 0 0 0 0 0 0 3 0 2 5 (14) 4 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 4 0 0 2 8 (13) 0 0 2 0 0 0 0 0 0 0 1 0 0 0 0 5 0 0 0 O 11(12) 1 0 0 0 0 0 0 0 0 0 0 0 0 0 6 2 Z 1 0 O 9 (18) 0 0 0 0 0 Z2 0 0 0 0 1 1 0 7 0 2 1 2 0 0 2 (8) 0 0 0 0 0 0 0 1 0 0 0 0 8 0 1 0 2 0 0 1 2 (8) 1 0 0 0 0 0 0 0 0 0 0 9 0 1 0 0 0 0 0 1 1119) 2 0 72 1 0 0 0 0 0 0 10 0 0 0 0 0 0 0 1 1 6 (15) 2 1 1 0 0 2 0 0 1 11 0 0 0 0 0 0 0 0 0 O 12 (13) 0 0 0 0 1 0 0 0 12 1 0 0 0 0 0 0 0 1 0 O 23 (28) 2 0 0 0 0 0 1 13 1 0 0 0 0 0 0 0 0 0 0 2 11(19) 2 2 0 1 1 0 14 0 1 0 0 0 0 22 0 0 0 1 0 1 14 (19) 1 0 0 0 0 15 0 0 0 0 0 0 0 0 0 1 0 0 0 3 6 (12) 0 0 0 0 16 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 71) 1 1 0 17 | 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 4 1(22) 1 0 18 0 0 0 0 0 0 0 0 0 1 0 1 1 0 0 1 9 9 (15) 2. 19 0 0 0 0 1 0 0 0 0 0 0 0 2 0 0 2 1 1 6 (14)

pointed and unkeratinized tubercles on the venter and, therefore, have a relatively granular ventral texture simi- lar to that of the northern toads. The tubercles forming the lateral descending row may be indistinct or distinct, and low and round, or high and pointed. Few adults of either sex have keratin overlying the peripheral margins of the cranial crests; when present, the keratin is thin and brown, rather than black.

Northern and southern Bufo valliceps differ in overall size, color pattern, and skin texture. However, I have not identified a clear disjunction in the distribution that corre- sponds with these characteristic differences. Samples from northern Veracruz and southern Tamaulipas are intermedi- ate in size (Figs. 2, 3; Sample 9) and have a mixture of color patterns and skin textures (among and within individuals) typical of both more northern and southern samples.

DISCUSSION

It is clear that there is considerable morphological variation in B. valliceps across its extensive range, and that there are few apparent trends that correspond with geo- graphical parameters. These toads vary extremely in size, color pattern, parotoid gland shape, and skin texture. The inconsistent variation has caused many specimens to be misidentified and resulted in general taxonomic confusion with respect to several populations of crested toads in Mexico and Central America (Mendelson, 1994, 1997 a,b, 1998). The range of B. valliceps, as considered here (Fig. 1), includes both Central America and a substantial area of North America. Across this range, and indeed east-to-west from Louisiana to western Texas, populations of B. valliceps are subject to diverse environmental regimes of seasonal rainfall and freezing temperatures. Moreover, this toad occurs from sea level to at least 1700 m, and these popula- tions are subject to different local conditions of rainfall and cloud cover. Few clear morphological trends seem to be associated with these environmental variables.

Bopy SIZE AND ARIDITY

Lee (1993) reviewed the theoretical literature that pre- dicts that amphibians, because of their high rates of cuta- neous evaporative water loss, presumably should adapt to arid environments by evolving larger body size in order to achieve a favorable surface-to-volume ratio. He also pointed out that no rigorous test of this prediction had been carried out prior to his work with the ubiquitous tropical treefrog Smilisca baudinii. Lee (1993) compared morpho- metric variables among samples of S. baudinii along a sharp precipitation gradient that exists on the Yucatan Peninsula. He found, contrary to the theoretical predictions, that there was no direct correlation between body size and aridity in this species and that, in fact, frogs from the more mesic base of the peninsula were larger than those from the more arid coastal tip. However, Lee (1993) did find that frogs from the drier, more seasonal areas have slightly smaller appendages—a finding consistent with the surface-volume desiccation argument. My analysis of overall size among

GEOGRAPHIC VARIATION IN BUFO VALLICEPS 9

Fig. 6.

Comparison of typical adult males from Alta Verapaz, Guatemala (left; KU 65551; SVL = 60.8 mm) and Coahuila, Mexico (right; KU

47011; SVL = 86.3 mm). Note differences in overall size and the distinctive middorsal stripe of the male from Coahuila.

samples of B. valliceps across the same precipitation gradi- ent on the Yucatan Peninsula revealed that although there are differences in overall size among populations of B. valliceps in this area, these differences do not seem to be correlated with the precipitation gradient. Similarly, among samples from throughout the range of B. valliceps, there is considerable variation in overall size, and the largest toads were from relatively arid areas in Texas, Coahuila, and Nuevo Leon. However, toads from other arid areas, such as the Grijalva Valley in Chiapas, were quite small. Con- versely, although the smallest toads did come from rela- tively wet areas, not all samples from wet areas were par- ticularly small.

Lee (1993) discussed the inherent difficulty in ad- equately testing a seemingly simple hypothesis such as correlation between body size and aridity. The elegance of his study lies in the fact that he identified the Yucatan Peninsula as a perfect testing ground for such a study; it is a relatively small area and is nearly uniform with respect to elevation and temperature profile. The environments from which my samples across Middle America originated differ substantially in nearly every environmental aspect—

not the least of which includes relatively severe winters (with snowfall) at the northern periphery of the range. These climates are so different that I deem them incompa- rable. The great amount of morphological variation in B. valliceps, and the lack of obvious trends with respect to geography and climate, lead me to concur with Lee (1993) that simple predictions of correlation between morphol- ogy and climate are easier to generate than to demonstrate; such correlations do not seem to pertain to Bufo valliceps.

Crest HEIGHT AND ELEVATION

Aseries of papers (Porter, 1963, 1964; Blair, 1966, 1972; Branson, 1995) reported that there is a direct correlation between the height of the cranial crests and elevation, and they used the population near Cuautlapam, Veracruz (near 1000 m in the Sierra Madre Oriental) as an example. These claims, presented without supporting data, seem to have their only basis in a figure presented in Porter’s (1962; fig. 26) dissertation. Because the original data used to support this apparent correlation are not widely available, and be- cause such a correlation between morphology and eleva- tion would be interesting, if true, I review Porter’s (1962) treatment of the matter.

10 ScIENTIFIC PAPERS, NATURAL History Museum, THE UNIVERSITY OF KANSAS

0.08

0.06

0.05

0.04

0.03

0.02

Maximum crest height/Snout-vent length

0.01

0.00

200 400 600 800 Elevation at collection site (m)

1000 1200 1400

Fig. 7. A precise representation of Porter’s (1962) Figure 26 show- ing the relationship between crest height and elevation at collection site. Porter’s original caption: “The relationship between maximum crest height and the elevation of the collecting site for B. valliceps. The black line indicates the mean cranial crest height, the [light] gray area one stan- dard deviation either side of the mean, the [dark gray] area two standard errors either side of the mean. All collecting site elevations were rounded- off to the nearest 100 meters.”

Porter’s original figure (1962; fig. 26) is represented in Fig. 7; several points require attention. First, Porter’s con- cept of Bufo valliceps included all specimens referable to B. macrocristatus (Mendelson, 1997a); this is a sexually dimor- phic, montane species in which females have dramatically enlarged crests. Other specimens that Porter referred to B. valliceps subsequently have been referred to other species such as Bufo spiculatus, which have larger cranial crests than B. valliceps (Mendelson, 1997b). Because Porter did not provide a list of the specimens included in his analysis, it is impossible to determine which, if any, of these other crested toads were mixed in with his samples of B. valliceps. Second, Porter apparently did not separate the sexes for his morphometric summaries; this is particularly unset- tling because females are larger than males in overall size and may differ somewhat in certain proportions, especially in the case of specimens now referred to B. macrocristatus. Third, Porter (1962:5) did not describe adequately how he measured crest height; I abandoned my own attempts to measure this feature because I could not develop a repeat- able measurement based on homologous landmarks. In summary, Porter’s analysis of correlation between crest height and elevation is suspect because it seems likely that he included among his samples individuals referable to

other species, including upland species with greatly en- larged crests; morecver, he did not account for sexual di- morphism in size, and he did not describe adequately his measurement techniques.

Pursuant to my concerns with Porter’s methodology, it is interesting to note that, despite how the results have been presented in subsequent literature, there seems to be no actual direct correlation between crest height and el- evation (Fig. 7). Comparison of the mean crest height/SVL ratios from below 200 m and near 1400 m indicates less than 1% difference, and the slope of the line between these two points probably does not differ significantly from zero; Porter did not perform a regression analysis on these data. What is evident from this plot is a remarkable amount of variation in specimens from an elevation of about 1000 m; most specimens of B. macrocristatus have been collected from near 1000 m of elevation (Mendelson, 1997b).

TAXONOMIC IssUES

Ina series of papers (Mendelson, 1994; 1997a,b,1998), I have resolved much of the taxonomic confusion sur- rounding populations of crested toads that previously (and inconsistently) were referred to B. cavifrons, B. cristatus, and B. valliceps. 1 removed the taxon B. valliceps macrocristatus Firschein and Smith from the synonymy of B. valliceps, rec- ognizing it as a full species, and transferred the taxon B. valliceps microtis Werner to the synonymy of Bufo coccifer. Two taxa remain to be considered—B. valliceps wilsoni Baylor and Stuart and Bufo nebulifer Girard.

Baylor and Stuart (1961) described the subspecies B. valliceps wilsoni based on a series of specimens collected by L. C. Stuart at Jacaltenango, Huehuetenango, Guate- mala. This locality is in the upper Grijalva Valley, at the base of the Sierra de los Cuchumatanes. The region is quite dry and thorn scrub forest prevails. The key diagnostic features of this taxon are the presence of relatively large parotoid glands and short supratympanic crests. The au- thors were careful to compare their specimens with B. valliceps from many other areas, and chose a comparative specimen from El Petén, Guatemala to illustrate the dis- tinctive nature of B. valliceps wilsoni (Baylor and Stuart, 1961:figs. 1, 2); they also cautiously pointed out that rec- ognition of this new taxon was not meant to imply that all other populations referred to B. valliceps were conspecific. Their choice of a comparative specimen from El Petén ex- aggerates the distinctness of B. valliceps wilsoni, because some specimens from the base of the Yucatan Peninsula have unusually small and triangular parotoid glands, as does the specimen in their figure. When specimens from the Grijalva Valley are compared with specimens from other areas of Mexico, the differences in the size and shape of the parotoid glands and the length of the supratympanic crests are less apparent. Nonetheless, L. C. Stuart’s careful eye, particularly with respect to the stout nature of the supratympanic crests, prevails in this case. Individuals from the Grijalva Valley are consistently different from B. valliceps from other areas. However, I do not recommend recognition of the taxon B. wilsoni at this time.

GEOGRAPHIC VARIATION IN BUFO VALLICEPS 11

Baird and Girard (1852) described Bufo granulosus based ona specimen collected “between Indianola and San Antonio,” Texas, during the U.S.-Mexico boundary sur- vey. Inasmuch as this name was preoccupied by Bufo granulosus Spix 1824, the replacement name Bufo nebulifer Girard 1854 was proposed. Peters (1863) placed B. nebulifer in the synonymy of B. valliceps Wiegmann 1833. The north-

ern B. valliceps are larger and have a distinctive pattern and skin texture compared to B. valliceps from Central America; these differences perhaps warrant recognition of B. nebulifer. However, in light of the general variation in these continuous characters among and within populations of B. valliceps, and the lack of a discrete character by which to diagnose the northern and southern populations, I do not propose recognition of B. nebulifer at this time.

LITERATURE CITED

Baird, S. F, and C. Girard. 1852. Characteristics of some new reptiles in the museum of the Smithsonian Institution. Proceedings of the Acad- emy of Natural Science of Philadelphia 6:173.

Baylor, E., R., and L. C. Stuart. 1961. A new race of Bufo valliceps from Guatemala. Proceedings of the Biological Society of Washington 74:195-202.

Blair, W. F. 1963. Evolutionary relationships of North American toads of the genus Bufo. A progress report. Evolution 17:1-16.

Blair, W. F. 1966. Genetic compatibility in the Bufo valliceps and closely related groups of toads. Texas Journal of Science 18:333-351.

Blair, W. F. 1972. Bufo of North and Central America. Pp. 93-101 in W. F. Blair (ed.), Evolution in the genus Bufo. Austin: University of Texas Press. viii + 459 pp.

Branson, B. A. 1995. The Gulf Coast Toad: a distribution master. Reptile and Amphibian Magazine, Nov/Dec 1995:59-63.

Conant, R., and J. T. Collins.1991. A Field Guide to Reptiles and Amphibians: Eastern and Central North America. Boston: Houghton Mifflin Co. xiv + 450 pp.

Duellman, W. E. 1963. Amphibians and reptiles of the rainforests of south- ern El] Petén, Guatemala. University of Kansas Publications, Museum of Natural History 15:205-249.

Duellman, W. E. 1970. The hylid frogs of Middle America. University of Kansas Museum of Natural History Monograph 1:1-753.

Frost, D. R. 1985. Amphibian Species of the World. Lawrence, Kansas: Allen Press, Inc. and Association of Systematics Collections, vi + 732 pp.

Johnson, J. D. 1989. A biogeographic analysis of the herpetofauna of north- western nuclear Central America. Contributions in Biology and Ge- ology, The Milwaukee Public Museum 76:1-66.

Lee, J. C. 1980. An ecogeographic analysis of the herpetofauna of the Yucatan Peninsula. University of Kansas Museum of Natural His- tory Miscellaneous Publication 671-75.

Lee, J. C. 1993. Geographic variation in size and shape of Neotropical frogs: A precipitation gradient analysis. Occasional Papers of the Museum of Natural History, Univeristy of Kansas 163:1-20.

Lee, J.C., and M. L. Crump. 1981. Morphological correlates of male mat- ing success in Triprion petastatus and Hyla marmorata (Anura: Hylidae). Oecologia 50:153-157.

Mendelson, J. R., III. 1994. A new species of toad (Anura: Bufonidae) from the lowlands of eastern Guatemala. Occasional Papers of the Mu- seum of Natural History, Univeristy of Kansas 166:1-21.

Mendelson, J. R., Ill. 1997a. A new species of Bufo (Anura: Bufonidae) from the Pacific Highlands of Guatemala and southern Mexico, with comments on the status of Bufo valliceps macrocristatus. Herpetologica 53:14-30.

Mendelson, J. R., III. 1997b. A new species of toad (Anura: Bufonidae) from Oaxaca, Mexico, with comments on the status of Bufo cavifrons and Bufo cristatus. Herpetologica 53:59-70.

Mendelson, J. R., III. 1998. A review of the Guatemalan toad Bufo ibarrai (Anura: Bufonidae), with distributional and taxonomic comments of Bufo valliceps and Bufo coccifer. In J. Johnson and O. Flores-Villela (eds.), Herpetology of Mesoamerica: A Modern Perspective. El Paso: Western Texas Publication (in press).

Peters, W. C. H. 1863. Bemerkungen tiber verschiende Batrachier, namentlich uber die Original-exemplare er von Schneider und Wiegmann beschriebenen Arten des zoologischen Museums zu Ber- lin. Monatsberichte der Preussischen Akademie der Wissenschaften zu Berlin, 1863 (Februaur):76-82.

Porter, K. R. 1962. Evolutionary relationships of the Bufo valliceps Group in Mexico. PhD. Dissertation, University of Texas, Austin. viii + 149

PP- Porter, K. R. 1963. Distribution and taxonomic status of seven species of Mexican Bufo. Herpetologica 19:229-247. Porter, K. R. 1964. Morphological and mating call comparisons in the Bufo valliceps complex. American Midland Naturalist 71:232-245. Porter, K. R. 1970. Bufo valliceps. Catalogue of American Amphibians and Reptiles 94:14.

Shelford, V. E. 1963. The Ecology of North America. Urbana: University of Illinois Press, 610p.

Villa, J. 1972. Anfibios de Nicaragua. Managua: Instituto Geografico Nacional and Banco Central Nicaragua, 216 pp.

Zug, G. R., and P. B. Zug. 1979. The marine toad, Bufo marinus: a natural history resumé of native populations. Smithsonian Contributions in Zoology 284:1-58.

APPENDIX SPECIMENS OF BUFO VALLICEPS EXAMINED

GUATEMALA: Atta Verapaz: 16.4 km W Tucurt (UTA A-7418-19, 7421-28); Finca Los Alpes (KU 65551). CHiquimuLa: Esquipulas (UMMZ 106794). EL Peten: 8 km NNW Chinaja (KU 55873-74, 58376, 55878-81); 10 km NNW Chinaja (KU 55887); 11 km NNW Chinaja (KU 55891-92, 5589496); 16 km NNW Chinaja, Rio San Roman (KU 55900, 55911-12, 55913 —-16); 15 km NNW Chinaja (KU 55918); 8.6 mi W El Cruce (KU 156414-15); 1.9 miS La Libertad (KU 156409-13); ca 4 mi N Poptun (KU 156396403); 3 mi S Tikal (156416-20); Tikal (LSUMZ 28138-39); Toocog, 15 km S La Libertad (KU 55920); Uaxactun (KU 156390-95). HUEHUETENANGO: Cuilco, Carretera Cuilco—Canibal, 1105 m (UTA A-47564); Jacaltenango (UMMZ 119371-74, 119380, 126298). IzaBaL: Quirigua, Puebla Ranch, United Fruit Company (CAS 70828-35, 70837-38); 1.8 km km SW Morales turnoff on Hwy CA-9, 85 m (KU 190101); 4 km ENE Morales turnoff on Hwy CA-9, 85 m (KU 190102); Rio Blanco,120 m (KU 190098), 99; 3.2 km SW Puerto Santo Tomas, 12 m (KU 190132); Aldea

Vista Hermosa, Los Amates, 700 m (KU 190141-42); Aldea Vista Hermosa, Los Amates, 135 m (KU 19014647); El Estor, Club Sechoc (UTA A-34048); Puerto Libre Hotel, at road fork between Puerto Santo Tomas and Puerto Barrios (UTA A-21677-78); N slope Sierra de las Minas, Finca Pueblo Viejo (UTA A-28869-71, 28874, 28876, 28878-79, 28885, 28894, 28898); Nickel Mine Airstrip at El Estor (KU 7429); 1.7 mi W El Estor, Las Dantas (UTA A-7430-31); Montanas del Mico, 1.4 km WSW Puerto Santo Tomas, near Las Escobas (UTA A-24738-39); Sierra de Santa Cruz, 10.0 km W Finca Semuc headquarters, Semococh (UTA A-24932-33, 24942); Montanas del Mico, 5.1 rd km WSW Puerto Santo Tomas, Las Escobas (UTA A-33046). EscurNTLa: circa 26 air km SE Escuintla, Finca El Caobanal, 100m (UTA A- 28957-58, 28961); 7.7 km SSW Santa Lucia Cotzumalguapa, on road to Las Playas (UTA A-29009-20); [Finca] El Salto, near water tank by the falls (MVZ 88352); Finca El Salto, ca 2 km E Escuintla, 1000 ft [304 m] (MVZ 104375); Rio Guacalate, near Masagua (USNM 12524045). JaLara:

6.9 km SE Jalapa on Rn 19 (TNHC 31345, 31500-08). JuTiara: Finca La Trinidad, near Casa Grande (UMMZ 107818). RETALHULEU: Hacienda Casa Blanca, 1-3 km N and NW Casa Grande (UMMZ 107820); 3.2 km N Champerico (UTA A-25849-64). SACATEPEQUEZ: 1.4 km SSE San Antonio (TNHC 31492-96). SAN Marcos: road between La Blanca and Tilapa, 3m (UTA A-47538). SANTA Rosa: [11.9 km] W Chiquimulilla, 49 m (KU 97704— 12). SUCHITEPEQUEZ: Mazatenango, Finca El Horizonte (LSU 9323, 9328); Rio Nahualate, 9 mi NNW Tiquasate (Depto. Escuintla) (USNM 125307).

HONDURAS: ATLANTIDA: Corozal, ca. 15 km E La Ceiba (LSUMZ 21611, 21614). CHoLUTECA: 28.8 mi S Sabana Grande (LSUMZ 33625). CoLon: Puerto Casilla (LSUMZ 22473); Rio Grande (LSUMZ 33626); Trujillo (LSUMZ 22491, 27743). Copan: 4.3 mi SW Santa Rosa de Copan (LSUMZ 22457); 9 km S La Entrada (LSUMZ 22588-89, 22591-92, 22597). Cortes: W of San Pedro, Hacienda Santa Ana (FMNH 4617-20, 4624, 4626, 4629-30); 12 km E San Pedro, Lake Ticamaya (FMNH 4632); Copan (FMNH 28513); 3.2 km NE San Pedro Sula (KU 97713-22); Cerro Cusuco, 1520 m (KU 209270); Quebrada de Colorado, ca Buenos Aires (KU 194223). Gracias A Dios: Tancin, 15 km NW Puerto Lempira (LSUMZ 21600). OLancuo: 0.5 km WNW Catamacas (LSUMZ 21590, 21592-94); Esculea Nacional de Agricultura, 4.5 km SE Catamacas (LSUMZ 21597); 3.4 km N San Esteban, 510 m (KU 209271); 15.7 km S San Esteban, 480 m (KU 209272); 5.6 km S San Esteban, 450 m (KU 209273); 4.6 km S San Esteban, 440 m (KU 209274-75). SANTA BARBARA: W side Lago de Yojoa, 775 m (KU 65544); San Jose de los Andes, 1610 m (KU 209279-80). Yoro: 2 km S Coyoles on Rio Aguan, 120 m (KU 101179); Rancho San Lorenzo, 25 km WSW Coyoles (LSUMZ 21606); 0.5 km N Coyoles (LSUMZ 21607); Santa Rita (KU 192295); Subirana Valley (FMNH 21789-21792).

MEXICO: CampecHeE: 5 km S Champoton (KU 70991-71003); 3 km N Hopelchen (KU 75231-33); Dzibalchen (KU 75234-37); 7.5 km W Escarcega (KU 71004-71021); 1 km W Escarcega (KU 71022-33). CHIAPAS: 26 km N Ocozocoautla (UTEP 5817-19, 5823); 1 km N Ocozocoautla (UTEP 5821, 5824, 5831); 20 km N Ocozocoautla (UTEP 5822); 23-24 km N Ocozocoautla (UTEP 5825-27); 2.4-5.3 km W Ciudad Cauhuatemoc (KU 97723-26); 1 km N Tuxtla Gutierrez (UTEP 5828-29, 5833); 54.5 km S Pueblo Nuevo Solistahuacan (TNHC 27029-32); 41 km S Pueblo Nuevo Solistahuacan (TNHC 27046, 27048, 27052); San Fernando (TNHC 25233- 34, 25237, 25239-40); 38 mi W Cintalapa (TNHC 27057-59); 3 mi E Cintalapa (TNHC 27054-55, 27358); 10 mi E Cintalapa (TNHC 27359). CoanulLa: spillway canal below Don Martin dam (KU 128778-86); 2-6 mi W Sacramento (KU 47010-13); Rio Salado de los Nadadores, El Carino

2 SCIENTIFIC PAPERS, NATURAL History Museum, THE UNIVERSITY OF KANSAS

(KU 80310-15). Hipatco: 12.5-38.5 km SW Huejutla (UTA A-13110, 13113- 19, 13121-31). Nuevo LEOn: La Huasteca Canon (KU 192507-19). Oaxaca: 6 km N Palomares (KU 58333-60). QuUINTANA Roo: Pueblo Nuevo X-Can, 10 m (KU 71036-50). Veracruz: 12.8 km N Acayucan (KU 97672-85); Cuautlapan (KU 97686, 105521—26); 16 km NE Fortin de las Flores (KU 9768796); Portero Viejo (KU 2583645, 25847-51, 26720-24); Cuautlapan (TNHC 27014-19, 17, 27123-28). YucATAN: Chichen Itza (KU 71059-61; FMNH 26956-58); 12 km E Chichen Itza (KU 71062); 17 km N Piste on rd to Tzitsin (KU 751950); Xocchel (KU 156435-37); 12.3-16.8 mi E Izamal (KU 156439-41); Mérida (FMNH 40653-58, TNHC 33224).

NICARAGUA: EsteLi: 7 km N, 16 km E Condega, 1200 m, (KU 85253-— 57); 5 km N, 14 km E Condega, 40 m (KU 85251). Matacatra: 10.5 km N, 9 km E Matagalpa, 960 m (KU 85258-62). NuEvo Sgcovia: 5 km N, 2.5 km E Jalapa, 680 m (KU 112722-26); 1.5 km N, 1 km E Jalapa, 660 m (KU 112727-39). ZELAYA: Bonanza (KU 85263-67, 101178); El Recreo, S side Rio Mico, 25 m (KU 112708-21); 11 mi W Muelle de los Bueyes, Hwy 7 (KU 136410).

USA: Louisiana: East BaToN RoucEe Par.: 4 mi S Baton Rouge (KU 22519-26; 22527) 1.6 mi E jct River Road and Brightside Dr. (KU 145509); ca 2 miS LSU campus on River Road (KU 145510); 0.9 miS East Feliciana (KU 145511-16). LivincsTon Pakr.: ca. jet rtes 16 and 1026 (KU 145507). Sr. CHARLES Par.: 3.6 mi N Norco (KU 145508). Texas: Atascosa Co.: Benton (KU 11997). Bexar Co.: Somerset (KU 20045—-46); Helotes (KU 11590). Brazoria Co.: 5 mi E Liverpool (KU 44884908); 1 miS Freeport (KU 44909- 15). CAMERON Co.: Brownsville (KU 11591-608, 14100-01, 14309-32). DaL- Las Co.: Dallas (KU 33556); South Cockrell Hill Road (WTA A-581-83); Oak Cliff (UTA A-1861). PARKER Co.: 2.0 mi NW Wheatland (UTA A-7228). Starr Co.: Rio Grande (KU 11998-99). Tarrant Co.: 8.0 mi W Fort Worth (UTA A-486); Arlington (UTA A-384, 9144, 17450-52, 41964); Benbrook— Aledo Road (UTA A-1223); E of US Hwy 377, jct of Mary’s Creek and Vickery Blvd (UTA A-7390). UvaLbE Co.: FM 1022, 2.6 miS Rte 90 (UTA A-37382). Travis Co.: Austin (TNHC 15643-47, 15649-53, 1565457). VaL VERDE Co.: Moose Canyon at jct Pecos River (KU 195052); Route 90, 1.4 mi E loop 25 (UTA A-37378-37380); State Hwy 163, 67.6 km S jct Inter- state Hwy 10 (UTA A-17453); 18 mi NE Comstock (TNHC 32355-57). WaLkKER Co.: Park Road 40 at entrance Huntsville State Park (UTA A- 37279-80); 0.25 mi down Fish Hatchery Rd (UTA A-41586); New Waverly (UTA A-40957); Texas Dept. of Corrections, Ellis Unit (UTA A-41696-98); Pritchett Field (UTA A-42327-48). Wess Co.: 4 mi N Laredo (KU 23383).