Carnegiea Gigantea, Cactaceae) Growth Rate Over Its American Range and the Link to Summer Precipitation

THE SOUTHWESTERN NATURALIST 50(1):65–106 MARCH 2005

NOTES

SAGUARO (CARNEGIEA GIGANTEA, CACTACEAE) GROWTH RATE OVER ITS AMERICAN RANGE AND THE LINK TO SUMMER PRECIPITATION

TALY DAWN DREZNER*

Department of Geography, Bolton 410, P.O. Box 413, University of Wisconsin–Milwaukee, Milwaukee, WI 53201-0413 *Correspondent: [email protected]

ABSTRACT Using an index of saguaro (Carnegiea gigantea) growth rate, relative growth was compared among multiple saguaro populations and mapped across the American range of the species. Relative growth at 10 sites was strongly linked to summer precipitation. This is the ﬁrst study of relative growth and the ﬁrst to link growth rate to environmental variables in multiple populations.

RESUMEN Usando un´ndice ı de la tasa de crecimiento del saguaro (Carnegiea gigantea), se com- paro´ el crecimiento relativo entre poblaciones mu´ltiples de saguaro y se ilustro´ a trave´s del rango americano de la especie. El crecimiento relativo en 10 sitios fue fuertemente ligado a la precipi- tacio´n de verano. Este es el primer estudio de crecimiento relativo y el primero en ligar la tasa de crecimiento a variables ambientales en poblaciones mu´ltiples.

Saguaro (Carnegiea gigantea) growth rates are The first objective of this study was to com- generally consistent relative to height (e.g., pare saguaro growth rates across multiple lo- slow growth at youth, highest growth rates be- cations that are geographically distant. In 2000, tween 3 and 4 m in height, and declining I used a telescoping leveling rod to collect data growth rates following) (Steenbergh and on saguaro trunk height for 16 to 21 saguaros Lowe, 1983; Turner, 1990; Niklas and Buch- of varying heights and number of branches at man, 1994). Actual growth, however, varies by several locales, all in flat areas (growth rate locale (Steenbergh and Lowe, 1983). Drezner varies by slope and aspect; Pierson and Turner, (2003a) recently developed a new method to 1998); washes were excluded (McAuliffe and estimate saguaro growth rate and age in any Janzen, 1986; Drezner, 2003b). I revisited 2 of population. By assuming a fixed growth pat- these locales in 2002 (see also Drezner and tern by height, a local growth factor can be Balling, 2002), and I revisited 5 others in 2003. computed from repeat sampling of individuals. I calculated the factor for the 7 resampled This value serves as an index, and growth rates locales by the procedure outlined in Drezner can then be compared between populations (2003a). The Drezner (2003a) technique is (Drezner, 2003a). The value of the factor at based on a single general formula that models Saguaro National Park, East was set to 1.0, by saguaro growth over its life, with a multiplica- definition, and values were calculated for Sa- tive adjustment factor for relative growth at guaro National Park, West (factor ϭ 0.743) each site. This local adjustment factor (or in- and Organ Pipe Cactus National Monument dex value) can be calculated through repeat (factor ϭ 0.617) (Drezner, 2003a). Thus, for sampling of individuals and fitting height and example, a saguaro that grows 10 cm per year growth at that height to the general growth at a given height at Saguaro National Park, curve. Height and growth are fitted to the East will only grow 7.43 cm at that same height growth curve by minimizing the sum of at Saguaro National Park, West and 6.17 cm at squares, the same principle used for fitting a Organ Pipe Cactus National Monument. regression line (Drezner, 2003a). 66 The Southwestern Naturalist vol. 50, no. 1

For the first time, saguaro growth in differ- able, stepwise regression yielded one variable, ent populations (with unique population struc- mean July precipitation (P Ͻ 0.001; r2 ϭ 0.84; tures) can be quantified and compared across Fig. 2). The resulting formula based on these the range of the species. A preliminary map of 10 locales is: growth factor ϭ 0.2705 ϩ 0.2713 saguaro growth rates over its Arizona range is (mean July precipitation). shown in Fig. 1. Saguaros generally exhibit rel- One could indirectly estimate the growth atively fast growth rates in the southeast, and factor if mean July precipitation is available. It growth rates decline northward and westward is important to note, however, that the growth across their American range. factor is an estimate, and relying on summer A second objective of this study was to ex- precipitation data to deduce growth rate might plore the relationship between saguaro growth compound the error and should be discour- rate and climate over its geographic range in aged. The growth factor should be calculated Arizona. I calculated 62 monthly climate vari- from direct field observation. It is clear that ables. These were based on 30 years of data where July rain is high, growth rates are high, (1968 through 1997) from 47 first-order me- and areas that receive less July rainfall support teorological stations and the cooperative ob- slower growing saguaro populations. The pop- serving network (National Climate Data Cen- ulations sampled in this study exhibited sub- ter, Summary of the Day database) ranging stantial variations in growth rate (Fig. 1). For Њ from the USA–Mexico border to 34.5 N lati- example, a 5-m tall saguaro at the westernmost tude and from the Arizona–California border site (Fig. 1) with a factor of 0.41 is estimated Њ to 110 W longitude and 4 stations in eastern- to be 135 years old (e.g., Drezner and Balling, most California. Although summer precipita- 2002), while at Saguaro National Park, East, a tion is highly variable over the short term (e.g., 5-m tall saguaro is estimated to be only 59 years 1 rain event), the data for precipitation are re- old (Steenbergh and Lowe, 1983). Thus, if no liable when calculated for a 30-year period. field data are available, mean July precipitation The 62 variables used were lowest maximum will likely provide a better estimate than the monthly temperature and lowest minimum data of Steenbergh and Lowe (1983) from Sa- temperature in December, in January, in Feb- guaro National Park (which receives large ruary, and in all 3 months combined (8 vari- quantities of summer rain); this is most notably ables); mean minimum and mean maximum true for sites far from Saguaro National Park monthly temperature for these same months and the high summer rainfall belt. Nonethe- (8 variables); mean maximum monthly tem- less, extreme caution should be used in inter- perature in May through September (5 vari- preting any results based on mean July precip- ables); mean minimum monthly temperature in May through September (5 variables); high- itation data rather than direct field observa- est maximum monthly temperature in May tion of populations. through September (5 variables); highest min- Others have observed, through intensive di- imum monthly temperature in May through rect measurement over a summer season, that September (5 variables); highest maximum saguaros grow in response to rainfall and dur- temperature observed from June to August (1 ing the summer (Hastings, 1959–1960; Has- variable); maximum monthly precipitation and tings and Alcorn, 1961; Steenbergh and Lowe, average monthly precipitation (24 variables); 1983). These observations have all been made and mean annual precipitation (1 variable). at Saguaro National Park. The present study These were interpolated to the 10 sampled lo- quantified growth rate at multiple locales over cales based on the methods described in Drez- the northern Sonoran Desert. For the first ner (2003b), with adjustments from Steen- time, a quantitative comparison of growth rate bergh and Lowe (1983). A complete list of the in different populations in different parts of meteorological stations and their locations is the range of the species has been conducted, provided in Drezner and Garrity (2003). Using and a preliminary trend map is shown that de- the Kolmogorov-Smirnov normality test, all var- picts saguaro growth patterns over the north- iables were normally distributed at P Ͼ 0.01, ern portion of their range. Finally, after an and most were normal at P Ͼ 0.05. analysis that included many possible environ- With growth factor as the dependent vari- mental variables that might be linked with March 2005 Notes 67

FIG.1 The location and growth factor of the 10 populations of saguaro (Carnegiea gigantea) sampled, including Saguaro National Park East (SNP-E), Saguaro National Park West (SNP-W), and Organ Pipe Cactus National Monument (OPCNM). The growth factor values for SNP-W, SNP-E, and OPCNM were derived using the data of Steenbergh and Lowe (1983) (Drezner, 2003a). The saguaro range is based on Turner et al. (1995). 68 The Southwestern Naturalist vol. 50, no. 1

FIG.2. The regression line for the growth factor of saguaros (Carnegiea gigantea) in 10 northern Sonoran Desert plots and mean July precipitation (cm).

growth rate, I linked summer precipitation HASTINGS, J. R. 1959–1960. Precipitation and saguaro with growth over Arizona. growth. University of Arizona Arid Lands Collo- quia 1959–1960/1960–1961:30–38. This project was supported in part by the National HASTINGS,J.R.,AND S. M. ALCORN. 1961. Physical Science Foundation (NSF#9987612). I thank A. Ellis, determinations of growth and age in the giant Arizona State Climatologist, for climate data, R. cactus. Journal of the Arizona Academy of Sci- Turner for helpful suggestions, M. Weesner, Saguaro ence 2:32–39. National Park Science and Resource Management MCAULIFFE,J.R.,AND F. J. JANZEN. 1986. Effects of Chief for information about climate data, and B. intraspecific crowding on water uptake, water Trapido-Lurie for help in the creation of the map. I storage, apical growth, and reproductive poten- thank R. Tenorio for translating the abstract into tial in the sahuaro cactus, Carnegiea gigantea. Bo- Spanish, and 2 anonymous reviewers for their help- tanical Gazette 147:334–341. ful suggestions. NIKLAS, K. J., AND S. L. BUCHMAN. 1994. The allom- etry of saguaro height. American Journal of Bot- any 81:1161–1168. LITERATURE CITED PIERSON,E.A.,AND R. M. TURNER. 1998. An 85-year DREZNER, T. D. 2003a. Saguaro (Carnegiea gigantea, study of saguaro (Carnegiea gigantea) demogra- Cactaceae) age-height relationships and growth: phy. Ecology 79:2676–2693. the development of a general growth curve. STEENBERGH,W.F.,AND C. H. LOWE. 1983. Ecology American Journal of Botany 90:911–914. of the saguaro, III: growth and demography. Na- DREZNER, T. D. 2003b. Revisiting Bergmann’s rule for tional Park Service Scientific Monograph Series saguaros (Carnegiea gigantea (Engelm.) Britt. and Number 17. United States National Park Service, Rose): stem diameter patterns over space. Jour- Washington D.C. nal of Biogeography 30:353–359. TURNER, R. M. 1990. Long-term vegetation change at DREZNER, T. D., AND R. C. BALLING,JR. 2002. Climatic a fully protected Sonoran Desert site. Ecology 71: controls of saguaro (Carnegiea gigantea) regener- 464–477. ation: a potential link with El Ninõ. Physical Ge- TURNER, R. M., J. E. BOWERS, AND T. L. BURGESS. 1995. ography 23:465–475. Sonoran Desert plants: an ecological atlas. Uni- DREZNER, T. D., AND C. M. GARRITY. 2003. Saguaro versity of Arizona Press, Tucson. distribution under nurse plants in Arizona’s Son- oran Desert: directional and microclimate influ- Submitted 13 May 2003. Accepted 23 May 2004. ences. Professional Geographer 55:505–512. Associate Editor was David Wester. March 2005 Notes 69

PROBABLE HYBRIDIZATION IN THE ACACIA CONSTRICTA SPECIES GROUP (FABACEAE: MIMOSIDEAE)

H. DAVID CLARKE,JOHN E. EBINGER,* AND DAVID S. SEIGLER

Department of Biology, University of North Carolina-Asheville, Asheville, NC 28804 (HDC) Emeritus Professor of Botany, Eastern Illinois University, Charleston, IL 61920 (JEE) Department of Plant Biology, University of Illinois, Urbana, IL 61801 (DSS) *Correspondent: [email protected]

ABSTRACT Principal components analyses (PCA) suggest that Acacia constricta and A. neovernicosa rarely hybridize, 1 lone probable hybrid being reported from Durango, Mexico. In contrast, hybrids of A. neovernicosa and A. schottii seem to be more common. Probable hybrids involving these 2 species were found in Brewster and Presidio counties, Texas, all within the known geographic range of A. schottii.

RESUMEN Ana´lisis de componentes principales (PCA) sugiere que Acacia constricta y A. neovernicosa raramente sufren hibridizacio´n, solo se registra 1 probable hı´brido de Durango, Me´xico. En contraste, hı´bridos de A. neovernicosa y A. schottii parecen ser ma´s frecuentes. Hı´bridos probables de estas dos especies se encontraron en los condados de Brewster y Presidio, Texas, todos adentro del rango conocido de A. schottii.

The 5 members of the Acacia constricta spe- both a geographical and morphological basis, cies group occur in North America from south- as well as specimens that other investigators western Texas and southern New Mexico and had considered to be hybrids (based on field Arizona, south through the deserts of northern or annotation labels), were included in these and central Mexico. Shrubs, small trees, or analyses. Because some multistate characters rarely prostrate shrubs, these taxa have stipular were included in the original analyses, PCAs spines, small bipinnately compound leaves with also were run using only continuous measure- 1 to 8 pairs of pinnae, leaflets mostly less than ment characters. To avoid use of multistate 6 mm long, and globose inflorescences with characters in these analyses, a resemblance co- small yellow flowers with numerous separate efficient for mixed data types was calculated stamens. Members of this group are separated (Legendre and Legendre, 1983; Podani, 1999) from other members of Acacia subgenus Acacia and the resulting matrix subjected to principal by the presence of involucral bracts near the coordinates analyses (PCoA). middle of the peduncles, a chartaceous peri- A total of 62 specimens was selected for anal- carpic strip lining the inner fruit valves, and ysis. Abbreviations used here for herbaria fol- oblong mottled seeds with small oval to U- low Holmgren et al. (1990). Each specimen shaped pleurograms. Included in this group was scored for 13 vegetative and 4 fruit char- are A. biaciculata, A. constricta, A. glandulifera, acters (Table 1). Measurement characters were A. neovernicosa, and A. schottii (Clarke et al., determined for each specimen (3 or more 1990). measurements) and plotted to confirm that During recent studies, occasional specimens gaps indeed exist in the character states em- were encountered that did not fit the expected ployed. Only primary leaves, that is, those as- range of variation for some of the taxa includ- sociated with nodes on elongated branches ed in this species group. These individuals were used for measurement of leaf and leaflet were scored for many of the characters used in characters. Many individuals of these species the principal components analysis (PCA) of develop numerous leaves on short shoots (spur the original study of this species group by branches) in the axes of primary leaves. These Clarke et al. (1990) (Table 1). Data from spec- leaves, designated as short-shoot leaves, are imens of the putative parental taxa, selected on smaller, have fewer pinna pairs, few leaflets per 70 The Southwestern Naturalist vol. 50, no. 1

TABLE 1—Characters scored for a principal components analysis involving Acacia constricta, A. neovernicosa, A. schottii, and the probable hybrids of these taxa, and their means and standard deviations.

Character Mean SD 1 Twig coating (TWI) (1 ϭ not glutinous, 2 ϭ glutinous) 1.40 0.49 2 Petiole length (PTL) (mm) 6.68 1.84 3 Petiole pubescence (PTP) (1 ϭ puberulent, 2 ϭ glabrous or nearly so) 1.48 0.50 4 Petiole gland diameter (GLA) (mm) 0.36 0.17 5 Rachis length (RAC) (mm) 9.23 9.52 6 Pinna pairs per leaf (PRS) 2.60 1.48 7 Pinna length (PLN) (mm) 11.30 4.68 8 Leaflet pairs per pinna (LPR) 6.71 2.23 9 Leaflet arrangement (LAR) (1 ϭ opposite to subopposite, 2 ϭ alternate) 1.32 0.47 10 Leaflet length (LLN) (mm) 2.80 1.07 11 Leaflet width (LWH) (mm) 0.73 0.24 12 Leaflet cross section (CS) (1 ϭ flat, 2 ϭ subterete) 1.29 0.46 13 Distance between leaflets (LDS) (mm) 1.65 0.97 14 Fruit width (FWH) (mm) 4.60 1.43 15 Fruit coating (FCO) (1 ϭ not glutinous, 2 ϭ glutinous) 1.73 0.45 16 Fruit glands (FGI) (1 ϭ absent, 2 ϭ sessile, 3 ϭ stalked) 1.73 0.45 17 Fruit constrictions (FCT) (1 ϭ constrictions, 2 ϭ rarely constricted) 1.19 0.40 pinna, and smaller leaflets than leaves associ- brids (not shown) and including A. constricta ated with elongated stems. and A. neovernicosa with the respective putative For principal components analyses (PCA), hybrid (not shown), all suggested that hybrid- the data were first standardized and a corre- ization occurs between these species, based on lation matrix, eigenvalues, and eigenvectors intermediacy of purported hybrid operational calculated using NTSYS-pc version 2.1 (Rohlf, taxonomic units in the analyses. This overall 2000). Eigenvectors were scaled by the square intermediacy of presumed hybrid individuals root of ␭. The axes were rotated and the re- occurred despite considerable variation and sulting loading values plotted as both 2-dimen- lack of intermediacy in some of the individual sional and 3-dimensional plots (Fig. 1). characters observed. Moreover, the positions To carry out PCoA, Gower’s resemblance co- of these individuals were generally more vari- efficients were calculated (Legendre and Le- able than those of the putative parents in the gendre, 1983; Podani, 1999; Dickinson, 2000). various analyses conducted. The nature of each character was designated In the PCA with only measurement charac- (binary, multistate, and quantitative descrip- ters (Fig. 1a, b), the first 2 characters account- tors) and all characters were weighted equally ed for 80% of the total variance. Leaf width (Dickinson, 2000). The data matrix derived and distance between leaflets (characters 11 was transformed by the DCENTER algorithm and 13) were most important for determining using distances squared, and eigenvectors and the component score on the first axis, petiole eigenvalues calculated with NTSYS-pc version and leaflet length (characters 2 and 10) were 2.1 (Rohlf, 2000). Eigenvectors were scaled by most important for the second axis, and the the square root of ␭. The resulting loading val- number of leaflet pairs per pinna (character ues were plotted as both 2-dimensional and 3- 8) for the third axis (Table 1). In a biplot with dimensional plots (Fig. 2). the corresponding vectors, individuals of A. A PCA using only measurement characters schottii (S) were differentiated from others by (Fig. 1), a PCA based on all 17 characters (not the pinna and leaflet length, fruit width, and shown), a PCoA based on Gower’s similarity distance between leaflets. The petiole length coefficients for A. constricta, A. neovernicosa, A. differentiated individuals of A. schottii and A. schottii, and all of the possible and proposed constricta (C) from those of A. neovernicosa (N). hybrids (Fig. 2), as well as PCAs including A. Individuals of A. constricta were distinct from neovernicosa and A. schottii with the putative hy- others in the petiolar gland diameter, rachis March 2005 Notes 71

FIG.1 Three-dimensional plot of the ﬁrst 3 axes (top) and a biplot (bottom) with a 2-dimensional plot of axis 1 vs. 2 from a principal components analyses (PCA) using 10 measurement variables for 24 specimens of Acacia neovernicosa, 16 specimens of A. constricta, 16 specimens of A. schottii, and 6 probable hybrids and backcrosses, and a plot of the eigenvectors superimposed. Axis 1 (49), axis 2 (31), and axis 3 (6) (not shown) correspond to a total of 86% of the total variance, in contrast to 29, 19, and 14%, respectively, as predicted by the broken-stick model. 72 The Southwestern Naturalist vol. 50, no. 1

FIG.2 Three-dimensional plot of the first 3 axes (top) and 2-dimensional plot (bottom) of axis 1 vs. 2 from a principal coordinates analysis (PCoA) employing Gower’s similarity coefficients based on 17 variables for 24 specimens of Acacia neovernicosa, 16 specimens of A. constricta, 16 specimens of A. schottii, and 6 probable hybrids and backcrosses. Axis 1 (21), axis 2 (17), and axis 3 (6) (not shown) correspond to a total of 44% of the total variance, in contrast to 8, 9, and 5%, respectively, as predicted by the broken-stick model. March 2005 Notes 73 length, number of leaflet pairs per pinna, and long), petioles more than 10 mm long, and leaflet width. fruits without glands, all traits of typical A. con- For the PCA based on 17 characters, the first stricta. It was similar to A. neovernicosa in having 2 principal components accounted for 79% of a glutinous covering on the stem and fruits, the total variance. Leaf width, leaflet arrange- and mostly only 1 pair of pinnae per leaf. The ment, and distance between leaflets (charac- label information for this specimen is: MEXI- ters 11, 9, and 13) were most important for CO: Durango: Camp O, Mezquital, ca. 0.5 mile determining the component score of the first SW of town along Rıó Mezquital, elevation axis, whereas petiole pubescence, presence or 1,460 m, 15 November 1970, O. H. Soule 2059 absence of a glutinous twig coating, and peti- (MO). ole length (characters 3, 1, and 2) were most Acacia neovernicosa ϫ A. schottii: According to important for determining the second axis Isely (1969), A. schottii shows close affinities to (Table 1). A. constricta and A. neovernicosa. More recently, To avoid the problem of using both contin- Clarke et al. (1990) found some specimens uous and multistate or binary characters in from Brewster and Presidio counties, Texas, PCA, a PCoA using Gower’s resemblance co- that were intermediate in leaflet characteristics efficients was performed (Legendre and Le- between A. schottii and A. neovernicosa and were gendre, 1983; Podani, 1999; Dickinson, 2000). presumed to be of hybrid origin. The resulting loading values are presented in Usually A. schottii is easily distinguished from a 2-dimensional and 3-dimensional plots (Fig. A. constricta and A. neovernicosa by its linear, su- 2). The parental species and putative hybrids bulate, alternate, and widely spaced leaflets were more clearly defined by this approach that give the foliage a feathery appearance. than in the plot based on only measurement This species is apparently quite successful in its characters. restricted range, where A. schottii is a common Acacia constricta ϫ A. neovernicosa: Acacia neov- species, often occurring in large, relatively ernicosa is now rarely confused with A. constric- pure thickets (Turner, 1959). These areas of ta, but originally was considered a part of that thorn-scrub vegetation are located on lime- taxon (Standley, 1919; Benson, 1943). Easily stone or gypsum derived soils, where A. constric- separated, A. neovernicosa has glutinous leaves ta and A. neovernicosa are common. Acacia schot- and twigs, is mostly glabrous throughout, and tii is particularly common in the southern por- has 1 to 2 (rarely 3) pairs of pinnae, whereas tion of the Big Bend region of Texas and pos- A. constricta lacks glutinous leaves and twigs, sibly occurs to the south in adjacent Coahuila, generally is lightly pubescent, and has leaves Mexico, although the authors have seen no with 3 to 8 pairs of pinnae. Specimens of A. constricta are consistently cyanogenic, those of specimens of this taxon from that region. To A. neovernicosa only rarely so (Seigler et al., our knowledge, the chromosome number for 1976; Clarke et al., 1990). In addition, the A. schottii has not been published. chromosome numbers differ, 2n ϭ 52 for A. Both PCA and PCoA analyses suggested that ϫ constricta and 2n ϭ 26 for A. neovernicosa (Turn- the hybrid A. neovernicosa A. schottii is occa- er and Fearing, 1960). Acacia neovernicosa is sionally encountered in Brewster and Presidio usually found on gypseous soil and generally counties, Texas, within the known geographic flowers 1 to 2 weeks before A. constricta (Turn- range of A. schottii. Most hybrid individuals in- er, 1959). Isely (1969) mentioned that these cluded in the PCA were probably F1 hybrids, species are not separated geographically, or ap- although others (H02, H07, and H08) could parently ecologically, but he found little evi- be possible backcrosses to A. schottii (Figs. 1 dence of hybrids. The PCA and PCoA analyses and 2). Most of the putative hybrids were in- of this study also suggested that hybrids occur termediate in leaflet length and width, the only rarely between these 2 species (Fig. 1). Of number of leaflet pairs per pinna, the distance the 10 specimens that were suggested to be of between leaflets along the pinna rachis, and in hybrid origin by information on the label or fruit width (Table 1). In general, most individ- on annotation labels, only 1 specimen fell out- uals were similar to A. neovernicosa in having side the cluster of either species. This speci- glutinous material on some parts of the plant men (N05) had large leaflets (2 to 3.5 mm and flat leaflets, and similar to A. schottii in hav- 74 The Southwestern Naturalist vol. 50, no. 1 ing puberulent petioles and rachises and alter- CLARKE, H. D., D. S. SEIGLER, AND J. E. EBINGER. 1990. nately arranged leaflets. Acacia constricta (Fabaceae: Mimosoideae) and re- The probable hybrids are listed below, along lated species from the southwestern U.S. and with label information. Three specimens (H02, Mexico. American Journal of Botany 77:305–315. DICKINSON, T. A. 2000. Program Gower6, BASIC soft- H07, and H08) might represent backcrosses ware for calculation of Gower’s coefficients. and, although similar to other probable hy- Made available by the author. brids, had leaves with longer rachises and pin- HOLMGREN, P. K., N. H. HOLMGREN, AND L. C. BAR- nae, more typical of A. schottii. UNITED NETT, editors. 1990. Index Herbariorum. Part I: STATES: Texas: Brewster County: clay-slate the herbaria of the world, eighth edition. New hills, near Lajitas, 27 June 1978, A. M. Powell York Botanical Garden, Bronx, New York. Updat- 3333 (SRSC, TEX); limestone soil 2 miles E of ed in: http://www.nybg.org/bsci/ih/ih.html. Packsaddle Mountain, elevation 1,070 m, 12 ISELY, D. 1969. Legumes of the United States: I. Na- June 1949, B. L. Turner 1072 (SRSC); lime- tive Acacia. Sida 3:365–386. stone hills at Boquillas, Big Bend National LEGENDRE, L., AND P. L EGENDRE. 1983. Numerical ecology. Elsevier Scientific Publishing, Amster- Park, elevation 985 m, 5 August 1966, B. H. dam, The Netherlands. Warnock 20896 (SRSC); sandy soil near Adobe PODANI, J. 1999. Extending Gower’s general coeffi- Wall Spring, elevation 1,020 m, 19 June 1949, cient of similarity to ordinal characters. Taxon B. H. Warnock, B. L. Turner, and J. O. Parks 48:331–340. 1134 (SRSC). Presidio County: gravel hills, ca. ROHLF, F. J. 2000. Numerical Taxonomy and Multi- 5 miles N of Lajitas, 7 July 1984, A. M. Powell variate Analysis Systems (NTSYSpc), version 2.1. and M. L. Powell 4406 (SRSC). Exeter Software, Setauket, New York. SEIGLER, D. S., J. E. DUNN, AND E. E. CONN. 1976. We wish to thank an anonymous referee and T. A. Acacipetalin in Acacia constricta from North Dickinson who provided valuable suggestions and a America. Phytochemistry 15:219–200. BASIC program used to calculate Gower’s resem- STANDLEY, P. C. 1919. New Mimosaceae. Contribu- blance coefficients. In addition, we thank the cura- tion from the United States National Herbarium tors of the Missouri Botanical Garden, University of 20:184–191. Texas, and Sul Ross State University for the loan of TURNER, B. L. 1959. The legumes of Texas. Univer- specimens used in this study. Financial support was sity of Texas Press, Austin. provided through grant NSF DEB 04-5803. TURNER, B. L., AND O. S. FEARING. 1960. Chromo- some numbers in the Leguminosae. III. Species LITERATURE CITED of the southwestern United States and Mexico. American Journal of Botany 47:603–608. BENSON, L. 1943. Revisions of status of southwestern desert trees and shrubs. American Journal of Bot- Submitted 4 October 2002. Accepted 19 April 2004. any 30:230–240. Associate Editor was Chris Lauver.

TEMPORAL ASSESSMENT OF A WEST TEXAS STREAM FISH ASSEMBLAGE

TIMOTHY H. BONNER,* CHAD THOMAS,CASEY S. WILLIAMS, AND JOHN P. K ARGES

Department of Biology/Aquatic Station, Texas State University-San Marcos, San Marcos, TX 78666 (THB, CT, CSW) The Nature Conservancy, P.O. Box 2078, Fort Davis, TX 79734 (JPK) *Correspondent: [email protected]

ABSTRACT We compared ﬁsh survey data across 3 collection periods (1952 to 1968, 1976 to 1994, and 2001 to 2002) from Independence Creek (Rio Grande drainage, Terrell County, Texas) to assess long-term changes in ﬁsh assemblage structure. The number of species collected declined from 28 in the 1952 to 1968 collection period, to 26 (plus 1 hybrid) in the 1974 to 1994 collection period, and to 23 (plus 1 hybrid) in the 2001 to 2002 collection period. Species loss included 5 March 2005 Notes 75

native ﬁshes that currently persist in the adjoining Pecos River and 4 species that were non-native to the lower Pecos River. Species additions included species native to the Pecos River that rarely inhabit smaller tributary streams and species that were introduced into the Pecos River. Twenty species were collected consistently from 1952 through 2002. Unweighted averages of abundant species across collection periods were Dionda episcopa (32%), Gambusia spp. (23%), Cyprinella proserpina (11%), C. lutrensis (10%), and Notropis amabilis (9%). Although apparent extirpations and additions of several rare species (Ͻ1% in relative abundance each) were noted, the Independence Creek ﬁsh assemblage remained similar during the last 50 years, in contrast to the general trend of native species reductions in much of the Rio Grande drainage.

RESUMEN Comparamos los datos de estudios pesqueros en el arroyo Independencia (cuenca del rıó Grande, condado de Terrell, Texas), a lo largo de tres perıódos de colecta (1952 a 1968, 1976 a 1994 y 2001 a 2002), para evaluar los cambios a largo plazo en la estructura del ensamblaje de peces. El nu´mero de especies colectadas declino´ de 28 en el periodo 1952 a 1968, a 26 (ma´s 1hı´brido) en el periodo 1974 a 1994 y a 23 (ma´s 1 hı´brido) en el periodo 2001 a 2002. La pe´rdida de especies incluyo´ a 5 especies de peces nativas que actualmente persisten en el adyacente rıó Pecos y 4 especies no nativas de la parte baja del mismo rıó. El aumento de especies incluyo´ especies nativas del rıó Pecos que rara vez habitan en los pequenõs arroyos tributarios y especies que fueron introducidas al rıó Pecos. Veinte especies fueron colectadas consistentemente, de 1952 a 2002. Los promedios no ajustados de las especies abundantes a lo largo de los periodos de colecta fueron Dionda episcopa (32%), Gambusia spp. (23%), Cyprinella proserpina (11%), C. lutrensis (10%), y Notropis amabilis (9%). Auń cuando hubo erradicaciones y aumentos aparentes de varias especies raras (Ͻ1% de abundancia relativa de cada una), el ensamblaje de peces del arroyo Independencia se mantuvo similar durante los u´ltimos 50 anõs, en contraste con la tendencia general en la reduccioń de especies nativas en la mayor parte de la cuenca del rıó Grande.

In the lower Pecos River (Carlsbad, New es in the fish assemblage of Independence Mexico, to the Rio Grande), the number of Creek. native fishes has been reduced by half because Independence Creek drains 1,935 km2 of of anthropogenic modifications, including in- limestone-based Ector Rock Outcrop soils com- troduced species, reduced water quality, and posed of stony loams, stony clay loams, and diminished spring and stream flows (Hoags- rock outcrops (Turner and Fox, 1974). The pe- trom, 2003). However, the largest tributary of rennial portion of Independence Creek is the lower Pecos River, Independence Creek, about 16 km long and is sustained largely by has been minimally impacted by human activ- Vanderbeek Spring and Caroline Spring, with ities and continues to support many of the fish- several peripheral springs contributing lesser es native to the lower Pecos River and Rio flow (Brune, 1981). Throughout its course, In- Grande drainages (Karges, 2003). The Inde- dependence Creek is dominated by runs and pendence Creek fish assemblage includes 2 pools, with few riffles and backwater areas. state-listed (Texas) threatened species (Cypri- Stream width ranges from 6 to 17 m, current nella proserpina and Etheostoma grahami), 1 spe- velocity ranges from 5 to 45 cm/s, and depth cies (Notropis jemezanus) considered threatened ranges from 18 to 48 cm. Dominant substrates in Texas by Hubbs et al. (1991), 1 species (Ic- are cobble and gravel, with a few reaches of talurus lupus) considered of special concern in bedrock and cemented gravel. Dominant Texas by Hubbs et al. (1991), and 2 species streamside plant associations are classified as (Dionda episcopa and Notropis amabilis) endemic Walnut-Desert Willow, Saltcedar, and Sawgrass- to the Edwards Plateau of Texas and Rio Willow (Webster, 1950). Grande drainages. In 1998, The Nature Con- Information on the fish fauna of Indepen- servancy purchased 8,000 ha that included 13 dence Creek was obtained from Tulane Muse- km of Independence Creek and Caroline um of Natural History (1952 to 1976); L. Spring, a primary contributor of freshwater to Campbell (1959, Basic survey and inventory of the creek. The purpose of our study was to species present in the Pecos River of Texas, compare historical fish survey data with that Texas Game and Fish Commission, Austin); N. from recent surveys to assess long-term chang- Valdez (1994, Composition and structure of 76 The Southwestern Naturalist vol. 50, no. 1

fish assemblages of Chandler Independence TABLE 1—Relative abundance (%) of species from Creek Preserve, Texas A&M University, unpub- 3 collection periods in Independence Creek, Texas. lished report); G. W. Linam and L. J. Kleinsas- The letter ‘‘I’’ denotes nonnative species in Inde- ser (1996, Relationship between fishes and wa- pendence Creek. The letter ‘‘P’’ denotes fishes that were present but not counted. ter quality in the Pecos River, Texas, Texas Parks and Wildlife Department, Austin); G. P. Relative abundance Garrett (1997, Chihuahuan Desert fishes status survey, Texas Parks and Wildlife Department, 1952– 1974– 2001– Ingram); and G. W. Linam, L. J. Kleinsasser, Species 1968 1994 2002 and K. B. Mayes (2002, Regionalization of the Lepisosteus osseus 0.02 index of biotic integrity for Texas streams, Tex- Dorosoma cepedianum 0.07 P as Parks and Wildlife Department, Austin). Campostoma anomalum (I) 0.04 Collectively, these records were from 1952 Cyprinella lutrensis 21.3 7.0 0.5 through 1994. Specimens were collected with Cyprinella proserpina 7.1 12.7 11.7 Cyprinella venusta (I) 0.01 0.05 0.2 seines or electrofishing gear, but effort and Cyprinus carpio (I) 0.1 P habitat types sampled were not always record- Dionda episcopa 36.1 30.8 29.7 ed. We assumed that species composition and Macrhybopsis aestivalis 0.02 0.4 abundances obtained during these collections Notropis amabilis 4.0 5.2 19.1 accurately represented the Independence Notropis braytoni 0.5 8.7 Creek fish assemblage at the time and location Notropis jemezanus 0.9 of sampling. Notropis stramineus 0.02 Our collections were made from December Pimephales vigilax 0.2 0.7 2001 through July 2002 at 7 sites on Indepen- Carpiodes carpio 0.1 0.02 Moxostoma congestum 0.05 0.1 0.4 dence Creek from its headwaters at Vander- Astyanax mexicanus 3.6 1.6 1.7 beek Spring to its confluence with the Pecos Ictalurus sp. 0.01 River. Fishes were collected with seines and a Ictalurus lupus 2.9 0.5 0.8 backpack electrofisher (Smith-Root Model 12- Ictalurus punctatus 0.5 P B), identified, enumerated, and released, ex- Pylodictis olivaris 0.04 0.05 cept for voucher specimens. Voucher speci- Cyprinodon variegatus ϫ mens were anesthetized in tricaine methane C. pecosensis (I) 0.05 0.5 sulphonate and preserved in 10% formalin. Fundulus zebrinus 5.1 0.02 0.5 Lucania parva 0.2 0.4 0.5 Field identification of 3 previously reported Gambusia spp. 13.2 26.8 28.7 Gambusia (G. affinis, G. geiseri, and G. speciosa) Menidia beryllina (I) 0.02 was complicated by hybridization and nomen- Lepomis auritus (I) 0.1 0.4 2.5 clatorial confusion between G. affinis and G. Lepomis cyanellus 0.9 0.09 0.1 speciosa (G. P. Garrett, pers. comm.). There- Lepomis macrochirus 0.04 0.02 0.07 fore, all Gambusia were treated as 1 taxon. Lepomis megalotis 1.2 1.1 0.5 Assemblage composition and relative abun- Micropterus salmoides 0.2 0.2 0.3 dances were compared among 3 collection pe- Pomoxis annularis (I) 0.02 Etheostoma grahami 0.5 0.4 1.2 riods: 1952 to 1968, 1974 to 1994, and 2001 to Cichlasoma cyanoguttatum 2.3 1.3 1.0 2002. Shannon-Weiner index (H; natural base) for species diversity, evenness (Buzas and Gib- Number of individuals 8,388 5,514 4,062 son’s E; Hayek and Buzas, 1997), and Renko- Species richness 28 27 24 Species diversity (H) 1.97 1.97 1.79 nen similarity index (Krebs, 1989) were used Evenness (E) 0.25 0.26 0.25 to quantify and compare assemblages across time. Thirty-three species and 1 hybrid were taken 1 hybrid) in the 2001 to 2002 collection peri- from Independence Creek from 1952 through od. Species diversity was 1.97 from the 1952 to 2002 (Table 1). The number of species col- 1968 and 1974 to 1994 collection periods, and lected declined from 28 in the 1952 to 1968 slightly lower (1.79) for the 2001 to 2002 col- collection period, to 26 (plus 1 hybrid) in the lection period. Evenness values (range 0.25 to 1974 to 1994 collection period, and to 23 (plus 0.26) were not substantially different among March 2005 Notes 77 collection periods. Assemblage similarity was lutrensis (10%), and N. amabilis (9%) for the 68% between 1952 to 1968 and 1974 to 1994 most common fishes. Abundant species with collection periods, and 79% between 1974 to the greatest change in relative abundance 1994 and 2001 to 2002 collection periods. Al- through time were C. lutrensis (decreased from though assemblages were similar among collec- 21 to 0.5%), N. amabilis (increased from 4% to tion periods, some notable changes in assem- 19%), and Gambusia (increased from 13 to blage composition and species abundance 29%), although theses changes might be par- were apparent. tially attributed to previous collections taken Of 8 species not taken during 2001 to 2002, mainly from lower reaches, whereas 2001 to 3(Lepomis microlophus, Pomoxis annularis, and 2002 collections were evenly distributed possibly Campostoma anomalum) are not native throughout Independence Creek. In 2001 to to the lower Pecos River (Hubbs et al, 1991; 2002 collections, C. lutrensis was abundant only Hoagstrom, 2003), and 5 native fishes (Lepisos- in the lower reach of Independence Creek. teus osseus, Macrhybopsis aestivalis, Notropis stra- Among species of conservation concern, C. mineus, N. braytoni, and Pimephales vigilax) per- proserpina, E. grahami, and I. lupus were taken sist in the lower Pecos River near Indepen- during each collection period. Cyprinella proser- dence Creek (Hoagstrom, 2003). These native pina was the most abundant state-listed threat- fishes normally inhabit larger rivers with silt ened species in Independence Creek and was and sand substrate, such as the lower Pecos taken from all sampling locations in 2001 to River, and marginally or temporarily inhabit 2002 collections. Etheostoma grahami and I. lu- smaller tributaries with cobble and gravel sub- pus were relatively uncommon (Ͻ3%) but per- strate, such as Independence Creek (Pflieger, sistent in Independence Creek. Among habi- 1975; Robison and Buchanan, 1988; Edwards tats, E. grahami was abundant in riffles and I. and Contreras-Balderas, 1991; Page and Burr, lupus was abundant in deep runs in the upper 1991; Rhodes and Hubbs, 1992). Thus, re- reaches of Independence Creek. Notropis jeme- duced species richness in the Independence zanus has not been collected in Independence Creek fish assemblage through time was not Creek since 1991 or in the lower Pecos River attributed to native species extirpations. since 1987 (Hoagstrom, 2003). Dionda episcopa Six species (Dorosoma cepedianum, N. jemezan- and N. amabilis, species endemic to the Ed- us, I. punctatus, Pylodictis olivaris, Cyprinodon var- wards Plateau and Rio Grande drainage, com- iegatus ϫ C. pecosensis, and Menidia beryllina) posed 49% of the current fish assemblage in not taken during 1952 to 1968 were found dur- Independence Creek. ing 1974 to 2002. Four species (D. cepedianum, Although apparent extirpations and addi- N. jemezanus, I. punctatus, and P. olivaris) are tions of several uncommon species (Ͻ1% in native to the lower Pecos River, although I. relative abundance each) were noted, the In- punctatus might have derived from individuals dependence Creek fish assemblage remained stocked for sportfishing purposes in Caroline similar during the last 50 years. This is in con- Spring (Karges, 2003). Cyprinodon variegatus trast to other lotic environments in the Rio (and its subsequent hybridization with C. peco- Grande and Gulf slope drainages of Texas, sensis) and M. beryllina occurrences are attri- where 80% of 129 fish assemblages were sub- buted to incidental stockings in the late 1950s stantially changed during a 30-year period (An- or early 1960s in the lower Pecos River (Minck- derson et al., 1995). Reservoir construction, ley, 1965; Stevenson and Buchanan, 1973). dredging, pollution, salination, and exotic in- Eighteen species were present during each troductions were responsible for changes in collection period. Two species (Cyprinus carpio 60% of the assemblages characterized as al- and Carpiodes carpio) were not collected during tered by Anderson et al. (1995). Although hab- 1964 through 1974. Among these 20 species, 4 itats of Independence Creek have been modi- (Cyprinella venusta, Cyprinus carpio, L. auritus, fied by water diversions for irrigation, im- and 1 of the Gambusia) are not native to the poundment of Caroline Spring, and nonnative lower Pecos River (Hubbs et al., 1991; Hoags- fish introductions, fish composition and abun- trom, 2003). Average percent abundances dance have not substantially changed. The per- across all 3 periods were D. episcopa (32%), sistence of a largely intact native fish assem- Gambusia (23%), Cyprinella proserpina (11%), C. blage in Independence Creek is a contrast to 78 The Southwestern Naturalist vol. 50, no. 1

the general trend of native species reductions HUBBS, C., R. J. EDWARDS, AND G. P. GARRETT. 1991. in much of the Rio Grande drainage (Edwards An annotated checklist of the freshwater fishes of and Contreas-Bladeras, 1991; Platania, 1991; Texas, with keys to identification of species. Texas Edwards et al., 2003; Hoagstrom, 2003). Journal of Science, Supplement 43:1–56. KARGES, J. 2003. Aquatic conservation and The Na- Texas Parks and Wildlife Department, The Nature ture Conservancy of West Texas. In: G. P. Garrett Conservancy, and Texas State University-San Marcos, and N. L. Allen, editors. Aquatic fauna of the Department of Biology, provided funding for this Northern Chihuahuan Desert. Special Publica- project. J. C. Wrinkle (Lower Pecos Project Director, tions Number 46, Museum, Texas Tech Univer- The Nature Conservancy) provided access on Inde- sity, Lubbock. Pp. 141–150. pendence Creek. N. G. Martinez, J. M. Watson, C. A. KREBS, C. J. 1989. Ecological methodology. Hooker, and C. Hartl assisted with fieldwork. G. W. HarperCollins, New York. Linam, G. P. Garrett, and an anonymous reviewer MINCKLEY, W. L. 1965. Records of atherinid fishes at provided helpful comments that improved this man- inland localities in Texas and northern Mexico. uscript. D. G. Solı´s provided resumen translation. Great Basin Naturalist 25:73–76. PAGE, L. M., AND B. M. BURR. 1991. A field guide to freshwater fishes of North America north of Mex- LITERATURE CITED ico. Houghton Mifflin Company, Boston, Massa- ANDERSON, A. A., C. HUBBS,K.G.WINEMILLER, AND chusetts. R. J. EDWARDS. 1995. Texas freshwater fish assem- PFLIEGER, W. L. 1975. The fishes of Missouri. Missou- blage following three decades of environmental ri Department of Conservation, Jefferson City. change. Southwestern Naturalist 40:314–321. PLATANIA, S. P. 1991. Fishes of the Rio Chama and BRUNE, G. M. 1981. Springs of Texas, volume 1. upper Rio Grande, New Mexico, with prelimi- Branch-Smith, Inc. Fort Worth, Texas. nary comments on their longitudinal distribu- EDWARDS,R.J.,AND S. CONTRERAS-BALDERAS. 1991. tion. Southwestern Naturalist 36:186–193. Historical changes in the ichthyofauna of the low- RHODES,K.,AND C. HUBBS. 1992. Recovery of Pecos er Rio Grande (Rio Bravo Del Norte), Texas and River fishes from a red tide fish kill. Southwest- Mexico. Southwestern Naturalist 36:201–212. ern Naturalist 37:178–187. EDWARDS, R. J., G. P. GARRETT, AND E. MARSH-MAT- ROBISON,H.W.,AND T. M. BUCHANAN. 1988. Fishes THEWS. 2003. Fish assemblages of the Rio Con- of Arkansas. University of Arkansas Press, Fayette- chos basin, Mexico, with emphasis on their con- ville. servation and status. In: G. P. Garrett and N. L. STEVENSON, M. M., AND T. M. BUCHANAN. 1973. An Allen, editors. Aquatic fauna of the Northern analysis of hybridization between the cyprino- Chihuahuan Desert. Special Publications Num- dont fishes Cyprinodon variegatus and C. elegans. ber 46, Museum, Texas Tech University, Lub- Copeia 1973:682–692. bock. Pp. 75–89. TURNER,A.J.,AND R. E. FOX. 1974. Soil survey of HAYEK,L.C.,AND M. A. BUZAS. 1997. Surveying nat- Terrell County, Texas. United States Department ural populations. Columbia University Press, New of Agriculture, Soil Conservation Service. York. WEBSTER, G. L. 1950. Observations on the vegetation HOAGSTROM, C. W. 2003. Historical and recent fish and summer flora of the Stockton Plateau in fauna of the lower Pecos River. In: G. P. Garrett northeastern Terrell County, Texas. Texas Jour- and N. L. Allen, editors. Aquatic fauna of the nal of Science 2:234–242. Northern Chihuahuan Desert. Special Publica- tions Number 46, Museum, Texas Tech Univer- Submitted 21 January 2003. Accepted 10 May 2004. sity, Lubbock. Pp. 91–109. Associate Editor was David Propst. March 2005 Notes 79

STATUS AND DISTRIBUTION OF THE ALLIGATOR SNAPPING TURTLE, MACROCHELYS TEMMINCKII, IN OKLAHOMA

J. DAREN RIEDLE,* PAUL A. SHIPMAN,STANLEY F. FOX, AND DAVID M. LESLIE,JR.

Oklahoma Cooperative Fish and Wildlife Research Unit, Department of Zoology, Oklahoma State University, Stillwater, OK 74078 (JDR, PAS) Department of Zoology, Oklahoma State University, Stillwater, OK 74078 (SFF) United States Geological Survey, Oklahoma Cooperative Fish and Wildlife Research Unit, Oklahoma State University, Stillwater, OK 74078 (DML) Present address of JDR: Arizona Game and Fish Department, Nongame Branch, 2221 West Greenway Road, Phoenix, AZ 85023 Present address of PAS: Department of Biological Sciences, Rochester Institute of Technology, 85 Lomb Memorial Drive, Rochester, NY 14623 *Correspondent: [email protected]

ABSTRACT Although historic records of the alligator snapping turtle, Macrochelys temminckii, show a past distribution throughout eastern Oklahoma, little is known about the current status and distribution of this species in the state. In 1997, surveys were initiated to identify extant populations of M. temminckii and assess their relative densities and viability. We surveyed 67 sites in 15 counties, with a total effort of 1,085 net nights. A total of 63 M. temminckii was captured at 11 sites, which are only in the southeastern quarter of the state and occur only in protected or isolated locations. Because of this apparent decrease of the range of this species in Oklahoma and because so few sites exhibited capture rates high enough to suggest possible healthy populations, we conclude that dramatic population declines of M. temminckii have taken place in Oklahoma. Possible reasons for these declines include overharvest and habitat alteration.

RESUMEN Aunque los registros histo´ricos de la tortuga Macrochelys temminckii demuestran una distribucioń anterior en todo el este de Oklahoma, no se sabe mucho del estatus y distribucioń actual de la especie en el estado. En 1997, se iniciaron muestras para identificar poblaciones actuales de M. temminckii y para evaluar sus abundancias relativas y viabilidad. Muestreamos 67 sitios en 15 condados con un esfuerzo total de 1,085 red-noches. Un total de 63 M. temminckii fue capturado en 11 sitios, que se ubican solamente en el cuarto sureste del estado y se encuentran solamente en localidades protegidas o aisladas. Debido a esta aparente disminucioń de la distri- bucioń de esta especie en Oklahoma y debido a que so´lo unos cuantos sitios mostraron tasas de captura tan altas para sugerir poblaciones saludables, concluimos que declines drama´ticos de M. temminckii han ocurrido en Oklahoma. Las razones posibles para los declines incluyen la caserıá excesiva y la alteracioń del ha´bitat.

The alligator snapping turtle, Macrochelys temminckii for listing as a threatened species. temminckii, is the largest freshwater turtle in However, the request for listing was precluded North America, with males attaining a cara- due to a lack of ecological information about pace length of 80 cm and a live mass of 113 kg the species. The status of the species was re- (Pritchard, 1989). Adults exhibit sexual dimor- viewed again in 1991, but no further actions phism; females reach a maximum live mass of were taken (United States Fish and Wildlife only 35 kg (Pritchard, 1989). Little informa- Service, 1991). At the state level, M. temminckii tion exists on the biology of M. temminckii. Prit- is afforded some protection in all states in chard (1989) and Ernst et al. (1994) suggested which it occurs, except Louisiana (Roman and that M. temminckii populations have declined Bowen, 2000). Macrochelys temminckii currently throughout its range. Overharvesting and hab- is listed as a species of special concern in itat alteration have been listed as the primary Oklahoma. causes (Reed et al., 2002). In 1984, the United Macrochelys temminckii is conﬁned to river sys- States Fish and Wildlife Service proposed M. tems that drain into the Gulf of Mexico. The 80 The Southwestern Naturalist vol. 50, no. 1

FIG.1 Historic distribution of the alligator snapping turtle, Macrochelys temminckii, in Oklahoma.

species occurs as far north as Kansas and Illi- of 4 hoops (1.05 m in diameter) covered with nois (Galbreath, 1961; Clarke, 1981) and from 2.5-cm square-mesh net. Nets were set up- the Florida Panhandle to eastern Texas and stream from submerged structures, such as Oklahoma (Conant and Collins, 1991). Histor- trees and log jams, and were baited with fresh ically, M. temminckii occurred throughout the ﬁsh suspended by a piece of twine on the hoop eastern one-third of Oklahoma (Glass, 1949; farthest from the opening of the trap. We set Webb, 1970; Black, 1982; Carpenter and Kru- nets in the late afternoon or evening and pa, 1989; and Heck, 1999; Fig. 1). Historical checked them the following morning. accounts of M. temminckii are based on single We surveyed 67 sites in 15 counties (Fig. 2). individuals, so information on distribution and Some sites were surveyed more than once due demography of M. temminckii in Oklahoma is to the presence of M. temminckii or if seemingly meager. Our objectives were to: 1) identify ex- good habitat was present. Our total trapping tant populations of M. temminckii in Oklahoma; effort was 1,085 net nights (1 net per night ϭ 2) assess their relative densities and viability; 1 net night). From 1997 to 1999, we made 69 and 3) capture, permanently mark, and release captures of 63 individuals of M. temminckii all specimens for subsequent population mon- (plus 8 more captures in July 2000 from Se- itoring. quoyah National Wildlife Refuge) at 11 sites We sampled sites throughout the eastern (Table 1; Fig. 3). one-third of Oklahoma from May through Au- Macrochelys temminckii was once distributed gust 1997 to 1999, with supplemental sampling throughout all the major river systems in east- of 2 sites in July 2000. Many of those sites were ern Oklahoma but was captured in our study at or near historic sites of occurrence for the at only 11 of the 67 sites sampled within that species in Oklahoma. We surveyed a variety of historic range. This suggests a dramatic decline habitats to adequately sample all possible hab- in numbers of M. temminckii in the state. Cur- itats in which M. temminckii might occur. rently known populations seem to be restricted We sampled sites using commercial hoop to a few remote or protected locations in the nets that were 2.1 m in length and constructed southeastern one-quarter of Oklahoma. Of March 2005 Notes 81

FIG.2 Sites sampled for Macrochelys temminckii in Oklahoma between 1997 and 1999. Some points represent more than 1 site due to close proximity of sample sites.

those populations, only the sites at Eufala Res- one-quarter of the state. The possible reasons ervoir (Mill Creek and Dutchess Creek) and for this decline are habitat alteration and deg- Sequoyah National Wildlife Refuge (Big Vian radation and historical incidental and illegal Creek, Little Vian Creek, Dirty Creek, Horton harvest. Slough, and Hezekiah Creek) yielded capture There are several forms of habitat alteration rates high enough to suggest viable popula- that might have a negative effect on M. tem- tions (Table 1). Macrochelys temminckii seems to minckii in Oklahoma. The Verdigris River has have been extirpated from the northeastern been channelized for navigation throughout

TABLE 1—Alligator snapping turtle (Macrochelys temminckii) capture rates in Oklahoma by site.

Capture rate (number of Number of Net turtles per Site County catpures nights net night) Little River McCurtain 3 167 0.018 Kiamichi River* Pushmataha 2 34 0.059 Dirty Creek** Muskogee 7 37 0.120 Hezekiah Creek** Sequoyah 3 17 0.180 Big Vian Creek*** Sequoyah 24 126 0.200 Little Vian Creek Sequoyah 26 64 0.410 Dutchess Creek McIntosh 4 9 0.444 Mill Creek McIntosh 8 13 0.620 * Represents 1 site on the Kiamichi River and 1 site on its tributary, Mill Creek. ** Resampled July 2000, bringing total number of M. temminckii captures to 77. *** Represents 2 sites on Big Vian Creek and 1 site on Horton Slough. 82 The Southwestern Naturalist vol. 50, no. 1

FIG.3 Distribution of Macrochelys temminckii in Oklahoma based on the 1997 to 1999 survey. Some points represent more than 1 site due to close proximity of sample sites.

much of Oklahoma. This manipulation of the managed as a coldwater stream for trout fish- river channel turns a low-energy, meandering, ing. Summer water temperatures taken during aquatic system with high habitat diversity into the study varied between 17Њ and 21ЊC. Little a higher energy system with low habitat diver- work has been done with the thermal require- sity that is vastly different from the habitat pre- ments of M. temminckii, but Allen and Neill ferred by M. temminckii (Shipman, 1993; Moll (1950) noted that they refuse food at temper- and Moll, 2000). atures Ͻ18ЊC. Based on our observations, the Macrochelys temminckii is exclusively aquatic, thermal environment in rivers such as the except for females during egg laying (Prit- Mountain Fork is suboptimal for M. temminckii. chard, 1989). An impoundment, such as a dam A 36.4-kg M. temminckii was captured on the or a lock, would block movement of individuals Mountain Fork River in 1993 by anglers (Ship- upstream or downstream of the structure. The man, Fox, and Riedle, pers. obser.), but no in- Arkansas, Caney, Verdigris, and Neosho rivers dividuals were captured on the Mountain Fork seem to be the major dispersal pathways for M. during our survey. Heck (1998) reported a de- temminckii throughout the central and north- cline in the number of M. temminckii observed ern parts of its range in Oklahoma. The series on the Mountain Fork River since the con- of locks and dams along the Arkansas, Caney, struction of the Broken Bow Dam in 1969; his and Verdigris rivers might be the main imped- last M. temminckii reported from the Mountain iment to the dispersal of individuals into the Fork River was from 1995. northern reaches of Oklahoma rivers and Heck (1998) listed several sources of pollu- streams. tion on the Little River that might have con- Thermal alteration of aquatic environments, tributed to the decline of M. temminckii over such as hypolimnetic release of cold water, also the last 30 y. Sources include sewage discharge, might be responsible for the decrease in M. runoff from chicken farms, wastewater dis- temminckii abundance in Oklahoma. The charge from chicken processing plants, chem- Mountain Fork River in McCurtain County is ical runoff, and soil erosion from commercial March 2005 Notes 83 timber harvest. Large amounts of wastewater and downstream from current populations discharge were noted on the Little River south might be nonexistent. We recommend that the of Broken Bow, McCurtain County, in 1998. future management of this large aquatic turtle Turtle captures at both the Little River and species in Oklahoma include: 1) continued Mountain Fork River were not only especially monitoring of the species in the state, 2) con- low for M. temminckii, but for all aquatic turtles tinued protection of known populations from native to that area of Oklahoma (Riedle, incidental and illegal take, and 3) reestablish- 2001). ment of extirpated populations. The primary forms of harvest of M. temminckii include incidental and illegal capture. The project was funded by the Federal Aid, En- Most incidental captures are those on trotlines dangered Species Act under Project E-40 of the and limb lines set by fishermen for catfish. Oklahoma Department of Wildlife Conservation Shipman and Riedle (1994) identified unat- and Oklahoma State University, with additional fi- tended limb lines and trotlines as a primary nancial and logistical support from Sequoyah Na- tional Wildlife Refuge and the Oklahoma Coopera- threat to turtles on the Saint Francis River in tive Fish and Wildlife Research Unit (Oklahoma southeastern Missouri. Shipman et al. (1991) State University, Oklahoma Department of Wildlife reported a specimen caught on a limb line 32 Conservation, United States Geological Survey Bio- km north of the Oklahoma border on a trib- logical Resources Division, and the Wildlife Manage- utary of the Verdigris River in Kansas. Heck ment Institute cooperating). We also want to thank (1998) listed several accounts of M. temminckii our field assistants and volunteers K. Cole, C. Finch- captures on limb lines and trotlines in Mc- er, C. O’Melia, M. O’Melia, R. Stark, T. Talbot, and Curtain County, Oklahoma. M. Stangland-Willis for all their help in ensuring the Due to its large adult size, M. temminckii has success of this project. been harvested historically throughout its range as a source of meat for personal and LITERATURE CITED commercial use (Pritchard, 1989). Sloan et al. (1995) reported 17,117 kg live-weight of M. ALLEN,E.R.,AND W. T. NEILL. 1950. The alligator snapping turtle, Macroclemys temminckii, in Flori- temminckii purchased by a single buyer in Lou- da. Special Publication Number 4, Ross Allen’s isiana between 1984 and 1986. Turtles were Reptile Institute, Silver Springs, Florida. historically taken for this market from Florida, BLACK, J. H. 1982. An annotated bibliography to ar- Georgia, Mississippi, Arkansas, Texas, and pos- ticles, notes and photographs on reptiles and am- sibly Oklahoma (Pritchard, 1989). The major- phibians appearing in Oklahoma Game and Fish ity of historical records for Oklahoma (Glass, News, Oklahoma Wildlife, and Outdoor 1949; Webb, 1970; Black, 1982; Carpenter and Oklahoma. Oklahoma Herpetological Society, Krupa, 1989; Heck, 1998) were based on indi- Special Publication Number 2. viduals taken by fishermen, and all were kept CARPENTER,C.C.,AND J. J. KRUPA. 1989. Oklahoma by the fishermen themselves or donated to pri- herpetology: an annotated bibliography. Univer- sity of Oklahoma Press, Norman. vate or public collections. Commercial harvest CLARKE, R. F. 1981. A record of the alligator snap- in Louisiana is still ongoing, even though M. ping turtle, Macroclemys temminckii (Testudines: temminckii is protected in surrounding states. Chelydridae), in Kansas. Transactions of the Kan- Although little information is available on sas Academy of Science 84:59–60. the historical status of M. temminckii, it seems CONANT, R., AND J. T. COLLINS. 1998. A field guide to that the species has suffered a severe decline reptiles and amphibians: eastern and central in Oklahoma. Current populations occur in ar- North America, third edition. Houghton Mifflin eas that are difficult to access and are afforded Company, Boston, Massachusetts. some protection from harvest. In Missouri, ERNST, C. H., J. E. LOVICH, AND R. W. BARBOUR. 1994. Shipman and Riedle (1994) found that M. tem- Turtles of the United States and Canada. Smith- sonian Institution Press, Washington, D.C. minckii was absent or rare from sites that had GALBREATH, E. C. 1961. Two alligator snappers, Ma- seemingly suitable habitat but were in close croclemys temminckii, from southern Illinois. Trans- proximity to historic and current turtle meat actions of the Illinois State Academy of Science markets. This also seems to be true in 54:134–135. Oklahoma. Due to the many river impound- GLASS, P. B. 1949. Macroclemys temminckii in ments in Oklahoma, recruitment upstream Oklahoma. Copeia 1949:138–141. 84 The Southwestern Naturalist vol. 50, no. 1

HECK, B. A. 1998. The alligator snapping turtle (Ma- purchased in the south-eastern United States of croclemys temminckii) in southeast Oklahoma. Pro- America. Animal Conservation 3:61–65. ceedings of the Oklahoma Academy of Science SHIPMAN, P. A. 1993. Alligator snapping turtle, habi- 78:53–58. tat selection, movements, and natural history in MOLL,E.O.,AND D. MOLL. 2000. Conservation of southeast Kansas. Unpublished M.S. thesis, Em- river turtles. In: M. W. Klemens, editor. Turtle poria State University, Emporia, Kansas. conservation. Smithsonian Institution Press, SHIPMAN, P. A., D. R. EDDS, AND D. BLEX. 1991. Re- Washington, D.C. Pp. 126–155. port on the recapture of an alligator snapping PRITCHARD, P. C. H. 1989. The alligator snapping tur- turtle (Macroclemys temminckii) in Kansas. Kansas tle: biology and conservation. Milwaukee Public Herpetological Society Newsletter 85:8–9. Museum, Milwaukee, Wisconsin. SHIPMAN,P.A.,AND D. RIEDLE. 1994. Alligator snap- REED, R. N., J. CONGDON, AND J. W. GIBBONS. 2002. ping turtle, Macroclemys temminckii, trap, mark and The alligator snapping turtle [Macrochelys (Macro- release project 1994. Final Report, Missouri De- clemys) temminckii]: a review of ecology, life his- partment of Conservation, Jefferson City. tory, and conservation, with demographic analy- SLOAN, K. N., K. A. BUHLMANN, AND J. E. LOVICH. ses of the sustainability of take from wild popu- 1995. Stomach contents of commercially harvest- lations. Report, Division of Scientiﬁc Authority, ed adult alligator snapping turtles, Macroclemys United States Fish and Wildlife Service, Aiken, temminckii. Chelonian Conservation and Biology South Carolina. 2:96–99. RIEDLE, J. D. 2001. The ecology of the alligator snap- WEBB, R. G. 1970. Reptiles of Oklahoma. University ping turtle, Macrochelys temminckii, in Oklahoma. of Oklahoma Press, Norman. Unpublished M.S. thesis, Oklahoma State Uni- versity, Stillwater. ROMAN, J., AND B. W. BOWEN. 2000. The mock turtle Submitted 8 January 2004. Accepted 17 May 2004. syndrome: genetic identiﬁcation of turtle meat Associate Editor was Geoffrey C. Carpenter.

BLACK-NECKED STILT (HIMANTOPUS MEXICANUS) BREEDING RANGE EXTENSION IN MEXICO

ARMANDO J. CONTRERAS-BALDERAS*

Laboratorio de Ornitologıá, Facultad de Ciencias Biolo´gicas, Universidad Autońoma de Nuevo Leoń, A.P. 425, San Nicola´s de los Garza, N. L., Me´xico 66450 *Correspondent: [email protected]

ABSTRACT This report documents a breeding range extension of the black-necked stilt (Himan- topus mexicanus) in Coahuila, Mexico.

RESUMEN Se amplia la distribucio´n del a´rea de reproduccio´n del candelero americano (Hi- mantopus mexicanus) en el estado de Coahuila, Me´xico.

On 6 May 2002, I found a pair of black- I found no other nests. Friedmann et al. necked stilts (Himantopus mexicanus) nesting at (1950) and Urban (1959) did not list this spe- Venustiano Carranza (Don Martin) Reservoir cies for Coahuila. The American Ornitholo- on the Rio Sabinas (27Њ31Ј13ЉN, 100Њ37Ј57ЉW). gists’ Union (1998) and Howell and Webb The nest was on the ground between rocks (1995) showed the species as a non-breeding (Fig. 1). On 6 May, it had 3 eggs, and the next visitor in southern Coahuila. Therefore, these day had 4 eggs. The 4 eggs averaged 42.7 ϫ observations document a northern extension 31.6 mm. Six other pairs of black-necked stilts of the known breeding range in the interior of also were observed in the immediate area, but northern Mexico and begin to ﬁll the gap in March 2005 Notes 85

the known breeding range between Mexico and United States (Robinson et al., 1999).

LITERATURE CITED

AMERICAN ORNITHOLOGISTS’UNION. 1998. Check-list of North American birds. American Ornitholo- gists’ Union, Washington, D.C. FRIEDMANN, R., L. GRISCOM, AND R. T. MOORE. 1950. Distributional check-list of the birds of Me´xico, part I. Paciﬁc Coast Avifauna 29:1–202. HOWELL,S.N.G.,AND S. WEBB. 1995. A guide to the birds of Mexico and northern Central America. Oxford University Press, New York. ROBINSON, J. A., J. M. REED,J.P.SKORUPA, AND L. W. ORING. 1999. Black-necked stilt (Himantopus mexicanus). In: A. Poole and F. Gill, editors. The birds of North America, number 449. Birds of North America, Inc., Philadelphia, Pennsylvania. URBAN, E. K. 1959. Birds from Coahuila, Mexico. University of Kansas, Publications of the Museum of Natural History 11(8):443–516

FIG.1 Photos of black-necked stilt nest at Don Martin Reservoir, Coahuila, Mexico: top, 6 May Submitted 9 December 2003. Accepted 22 March 2004. 2002; bottom, 7 May 2002. Associate Editor was Timothy Brush.

ATY PICAL PAIR-BONDING BEHAVIOR AMONG GOLDEN-FRONTED WOODPECKERS (MELANERPES AURIFRONS)

MICHAEL S. HUSAK*

Department of Biology, Angelo State University, San Angelo, TX 76909 Present Address: Department of Biological Sciences, Cameron University, 2800 W Gore Boulevard, Lawton, OK 73505 *Correspondent: [email protected]

ABSTRACT While conducting a study of golden-fronted woodpecker (Melanerpes aurifrons) territorial behavior and site ﬁdelity in west-central Texas in 1996, I observed a color-banded male interacting with 2 females in a manner consistent with pair-bonding. This behavior, which was previously unreported for this species, was observed late in the breeding season and continued through the following winter. No nesting attempts were observed, but the male and females did exhibit typical mate behavior, including the exchanging of mate-greeting calls and displays. The females occupied separate, adjacent territories, both of which the male regularly occupied. The chronology of events and potential factors inﬂuencing such behavior are discussed.

RESUMEN Durante un estudio sobre el comportamiento territorial y fidelidad al sitio del car- pintero cheje (Melanerpes aurifrons) en la parte central del oeste de Texas en 1996, observe´la asociacioń de un macho marcado con anillos de color con 2 hembras en una manera consistente con la formacioń de la pareja. Este comportamiento, que no se habıá registrado previamente en esta especie, se observo´ tarde en la temporada de apareamiento y continuo´ hasta el siguiente 86 The Southwestern Naturalist vol. 50, no. 1

invierno. No se observaron tentativas de nidificar, pero el macho y las hembras mostraron comportamiento tı´pico de parejas, incluyendo el intercambio de vocalizaciones de reconocimiento y cortejo. Las hembras ocuparon territorios separados y adyacentes, los cuales fueron tambień ocu- pados regularmente por el macho. Se discuten la cronologıá de eventos y los factores probables que influencian dicho comportamiento.

As a family, woodpeckers are overwhelming- The 410-ha site was located along the North ly monogamous breeders (Short, 1982), often Concho River and included the North Concho maintaining individual pair bonds between River Day Use Area, Bald Eagle Creek Camp- breeding seasons (Winkler and Christie, 2002). ground, and Wildlife Management Area 8. Veg- Even among cooperatively breeding species, etation of the site was predominantly mesquite extra-pair copulations are rare ( Jackson, (Prosopis glandulosa) brushland of varying den- 1994). The scattered records of polygamy in sity, associated with narrow bands of riparian the family have predominantly been of classical woodlands dominated by pecan (Carya illinoien- polyandry (e.g., Willimont et al., 1991; Kotaka, sis) along 3 waterways that pass through the 1998; Wiebe, 2002). Ligon (1993) suggested area. For details of the study area, see Husak that male participation was essential for nest (2000). Approximately 70% of all individuals on success among woodpeckers, and thus the lack the study site were color-banded. Two bands of classical polygny. However, recent observa- were placed on each leg of adult and juvenile tions suggest that polygyny is possible (e.g., woodpeckers that were subsequently known by Wiktander et al., 2000; Conner et al., 2001). 4-letter codes representing their color combi- Woodpeckers can exhibit considerable sea- nation. Resident individuals that were not band- sonal variation in social interactions and tol- ed could readily be identified by unique varia- erances (Short, 1982). Members of the New tions in the color of the nape and nasal tufts World genus Melanerpes, in particular, exhibit a and the unique pattern of white bands on the diversity of social interactions, ranging from outer rectrices. seasonal solitary to communal and cooperative Male RSRS was banded as an after-hatch-year behavior (Short, 1982; Winkler and Christie, bird in October 1995. He remained on the 2002). study site and attempted to breed with a color- Golden-fronted woodpeckers (Melanerpes au- banded female (YGYG) in spring 1996. After rifrons) are the least studied melanerpine their nest failed on 17 June, the pair began woodpecker in the United States, but as far as exploring new cavities and renewed courtship is known, they are monogamous and aggres- behavior, including mutual tapping, mounting, sively defend breeding territories against con- reverse mounting, and cavity excavation. How- specifics (Husak and Maxwell, 1998). During ever, YGYG abandoned the study area during the non-breeding seasons, social interactions the last week of June before a second nest are variable. Individuals can remain with mates could be initiated. on a territory, occupy individual territories, Within a week of the disappearance of wander, make use of communal foraging YGYG, an unbanded female of unknown origin grounds in which conspecifics are mildly tol- was observed foraging without incident in the erated in times of concentrated food abun- territory of RSRS. By the middle of July, RSRS dance, or a combination of these behaviors and the new female were observed moving (Husak, 2000). Here I report on a previously about the territory as a pair. The pair regularly undescribed behavior for golden-fronted foraged in the same tree, actively defended woodpeckers: a male sharing a foraging terri- boundaries together, and maintained vocal tory and maintaining pair bonds with multiple contact when apart. They also regularly ex- females during the late summer, fall, and early changed mate-greeting displays, including winter of 1996. head-swinging, bill-pointing, and bill-waving All observations were made during studies of (see Husak, 1996, for descriptions of displays). golden-fronted woodpecker territorial behavior Neither copulation nor nesting was observed. and site fidelity at San Angelo State Park, Tom In late August, female YGYG returned to the Green County, Texas (31Њ31ЈN, 100Њ33ЈW), con- study site for the first time since June. Follow- ducted from January 1995 through May 1998. ing her return, YGYG and the unbanded fe- March 2005 Notes 87 male were observed fighting near the center of served; however, the novelty of such observa- the territory of the male. By 3 September, tions warrants documentation and invites fu- YGYG occupied a territory adjacent to the east- ture consideration. During the course of field- ern edge of the territory of the unbanded fe- work, this was the only case of such interac- male, where a service road served as a bound- tions observed, suggesting it is rare among ary between the 2 females. For the following 6 golden-fronted woodpeckers. However, it un- weeks, the 2 females maintained well-defined derscores the plasticity of behavior and social territories, and the male divided his time be- interaction within this poorly known species. tween the 2 territories. The male regularly ex- changed vocalizations and mate-greeting dis- I thank J. Husak, C. Adkins, and C. Weaver for plays with both females, and actively assisted assistance with banding of golden-fronted wood- both females in defending their territory peckers. R. Conner, A. Husak, T. Maxwell, U. Wik- boundaries from other intruding conspecifics. tander, and an anonymous reviewer provided nu- However, the male aided neither when the fe- merous helpful comments that improved earlier versions of this manuscript. I am most grateful to I. males engaged in aggressive encounters with Vilchez Ramirez for Spanish translation of the ab- one another. stract. From mid October until mid November, territorial behavior greatly decreased on the study LITERATURE CITED site as large numbers of golden-fronted woodpeckers from adjacent areas began foraging CONNER, R. N., J. R. MCCORMICK,R.R.SCHAEFER,D. along waterways, exploiting the ripening pe- SAENZ, AND D. C. RUDOLPH. 2001. A red-cockaded cans (Husak, 2000). RSRS, YGYG, and the un- woodpecker group with two simultaneous nest banded female occupied home ranges consis- trees. Wilson Bulletin 113:101–104. tent with previous territories described above. HUSAK, M. S. 1996. Breeding season displays of the The unbanded female was last seen in the area golden-fronted woodpecker. Southwestern Natu- on 30 October; the reasons for her disappear- ralist 41:441–442. ance are unknown. By 3 November, another HUSAK, M. S. 2000. Seasonal variation in territorial behavior of the golden-fronted woodpecker in female (YSSS) was observed occupying much west-central Texas. Southwestern Naturalist 45: of the former territory of the absent unbanded 30–38. female. This female was submissive to RSRS, HUSAK, M. S., AND T. C. MAXWELL. 1998. Golden- and within several days seemed to be accepted fronted woodpecker (Melanerpes aurifrons). In: A. by him and behaved much as a mate would, Poole and F. Gill, editors. The birds of North exchanging calls and displays, and foraging in America, number 373. Birds of North America, close proximity to one another. Once com- Inc., Philadelphia, Pennsylvania. munal foraging came to an end in mid Novem- JACKSON, J. A. 1994. Red-cockaded woodpecker (Pi- ber, females YGYG and YSSS began defending coides borealis). In: A. Poole and F. Gill, editors. their respective territories against one another, The Birds of North America, number 85. Acad- emy of Natural Sciences, Philadelphia, Pennsyl- as well as other conspecifics, and continued ex- vania, and American Ornithologists’ Union, changing mate displays with RSRS. None of Washington, D.C. these individuals remained on the study site by KOTAKA, N. 1998. Classical polyandry in the great the end of winter, preventing further observa- spotted woodpecker Dendrocopos major. Ibis 140: tion. 335–336. Although nesting attempts were not ob- LIGON, J. D. 1993. The role of phylogenetic history served, this does not necessarily mean that in- in the evolution of contemporary avian mating dividuals were not exhibiting pair-bonding be- and parental care systems. In: D. M. Powers, ed- havior for future reproductive potential. Gold- itor. Current ornithology, volume 10. Plenum en-fronted woodpecker pair-bonding behavior Press, New York. Pp. 1–46. SHORT, L. L. 1982. Woodpeckers of the world. Del- does intensify during late February and March, aware Museum of Natural History Monograph but pair formation has been observed year- Series, Number 4. round (Husak and Maxwell, 1998). Too little is WIEBE, K. L. 2002. First reported case of classical known about the earlier history and ultimate polyandry in a North American woodpecker, the fate of these individuals to allow conclusive northern flicker. Wilson Bulletin 114:401–403. statements about why such behavior was ob- WIKTANDER, U. O., O. OLSSON, AND S. G. NILSSON. 88 The Southwestern Naturalist vol. 50, no. 1

2000. Parental care and social mating system in dae (woodpeckers). In: J. del Hoyo, A. Elliott, the lesser spotted woodpecker Dendrocopos minor. and J. Sargatal, editors. Handbook of birds of the Journal of Avian Biology 31:447–456. world, volume 7. Jacamars to woodpeckers. Lynx WILLIMONT, L. A., J. A. JACKSON, AND B. J. S. JACKSON. Edicions, Barcelona, Spain. Pp. 296–555. 1991. Classical polyandry in the West Indian woodpecker on Abaco, Bahamas. Wilson Bulletin 103:124–125. Submitted 14 October 2003. Accepted 29 June 2004. WINKLER, H., AND D. A. CHRISTIE. 2002. Family Pici- Associate Editor was Timothy Brush.

ECTOPARASITES AND FOOD HABITS OF ELLIOT’S SHORT-TAILED SHREW, BLARINA HYLOPHAGA

CHRISTOPHER M. RITZI,* BRIAN C. BARTELS, AND DALE W. SPARKS

Department of Ecology and Organismal Biology, Indiana State University, Terre Haute, IN 47809 (CMR, DWS) Department of Biological Sciences and Sternberg Museum of Natural History, Fort Hays State University, Hays, KS 67601 (BCB) Present address of CMR: Department of Biology, Sul Ross State University, Alpine, TX 79832 *Correspondent: [email protected]

ABSTRACT The natural history of Elliot’s short-tailed shrew (Blarina hylophaga) has remained virtually unknown since it was recognized as a distinct species in 1981. For this study, we examined 26 specimens of B. hylophaga for ectoparasites. We found 5 species of ﬂeas, 1 species of tick, and 17 species of mites. We also examined stomach contents of 30 specimens of B. hylophaga and discovered that insects (particularly beetles), slugs, and spiders were the most common food items. To determine the degree of similarity between B. hylophaga and the 2 other widely distributed short-tailed shrews, we compared these data to the published ectoparasite and food habits data for B. carolinensis and B. brevicauda. We found that both the ectoparasitic fauna and the diet of B. hylophaga broadly overlap those of its congeners. We also noted, however, that each species of Blarina had parasites that were more closely associated with it than with the other 2 species of Blarina. In addition, the diet of B. hylophaga consists of harder-bodied foods, containing more beetles, fewer earthworms, and less fungi than that of the other 2 species.

RESUMEN La historia natural de la musaranã Blarina hylophaga ha quedado virtualmente des- conocida desde que fue reconocida como una especie distinta en 1981. En este estudio examinamos los ectopara´sitos de 26 B. hylophaga. Encontramos 5 especies de pulga, 1 especie de gar- rapata, y 17 especies de aćaro. Tambień examinamos los contenidos estomacales de 30 especı´- menes de B. hylophaga y se descubrio´ que los insectos (particularmente escarabajos), babosas y aranãs fueron la comida ma´s comuń. Para determinar el grado de semejanza entre B. hylophaga y las otras dos musaranãs ampliamente distribuidas, comparamos estos datos con los publicados de ectopara´sitos y ha´bitos alimenticios de B. carolinensis y B. brevicauda. Encontramos que ambos la fauna de ectopara´sitos y la dieta de B. hylophaga se sobrelapan ampliamente con las de sus con- geńeres. Sin embargo, notamos tambień que cada especie de Blarina tuvo para´sitos que estań cercanamente asociados con la misma que con las otras 2 especies de Blarina. Adema´s, la dieta de B. hylophaga consiste de comida ma´s dura, conteniendo ma´s escarabajos, menos lombrices y menos hongos que la de las otras 2 especies.

Taxonomic studies of Blarina during the past George et al., 1981; George et al., 1982; Han- 3 decades (Genoways and Choate, 1972; dley and Varn, 1994) have led to a split of the March 2005 Notes 89 northern short-tailed shrew, B. brevicauda, into then were dried and examined for ectopara- 3 species. As revised, the northern short-tailed sites using a 7–30ϫ stereomicroscope, with all shrew ranges from New Brunswick west to Sas- parasites tabulated by group. Representative katchewan and south along the Appalachians voucher specimens were slide-mounted in PVA to Georgia in the east and to southern Ne- medium and identified using a combination of braska on the Great Plains. The southern keys, including Krantz (1986), Whitaker short-tailed shrew, B. carolinensis, occurs pri- (1982), Lukoschus et al. (1988), and Fain et al. marily on the Coastal Plain in the Southeast, (1991). Finally, shrews were examined under a and Elliot’s short-tailed shrew, B. hylophaga, is stereomicroscope to determine if any parasites distributed from southern Iowa through west- remained attached after the wash. Although no ern Kansas and south to Texas. additional ectoparasite species were recovered Unfortunately, this increase in the number at this time, the value of microscopy in locating of species recognized has not led to detailed parasites embedded or firmly attached to the natural history studies of the re-elevated spe- host is well documented (Whitaker et al., 1993; cies (Genoways and Choate, 1998). Most of the Ritzi and Sparks, 2002). Voucher specimens of available natural history data is based on the both parasites and shrews were deposited at most widespread species, B. brevicauda, with both MHP and the Indiana State University some information also available for B. caroli- Vertebrate Collection. nensis (Mumford and Whitaker, 1982; Geno- Parasites obtained from B. hylophaga were ways and Choate, 1998). The purposes of this compared with data from B. brevicauda (Whi- study were to conduct the first detailed study taker and Mumford, 1972; Mumford and Whi- of the diet and ectoparasites of B. hylophaga, taker, 1982; Pascal, 1984) and B. carolinensis and to compare these results with what is (Pascal, 1984; Whitaker et al., 1994) by using known about B. brevicauda and B. carolinensis. Chi-square analysis of 2 ϫ 3 contingency tables We obtained B. hylophaga for this study from (Zar, 1996). Parasitological terms follow those 2 sources. Seven shrews were unprepared spec- suggested in Bush et al. (1997); prevalence imens stored in the freezers of the Sternberg pertains to the percentage of hosts infested Museum of Natural History (formerly the Mu- with a particular parasite, and mean intensity seum of the High Plains, MHP) since as early refers to the average number of a particular as 1993; the other 23 specimens were collected parasite species per infested host. in summer 2002. All shrews were collected dur- The diet of 30 shrews was studied using the ing the months of May through September same techniques as Whitaker et al. (1994). from north-central Kansas using line transects Briefly, we removed the stomachs and opened of both Sherman live traps and Museum Spe- them into watch glasses. Stomach contents cials, plus pitfalls set against drift fences. Upon were removed, suspended in 75% ethanol, and capture, shrews were killed, individually examined using a 7–30ϫ stereomicroscope. bagged, and stored as frozen carcasses at MHP The percentage of each food item in each until examination. All shrews were taken in ac- stomach was then estimated visually. Data were cordance with the American Society of Mam- summarized in terms of percent volume (the malogists guidelines for animal care and use average percent of the diet of each shrew) and (Animal Care and Use Committee, 1998). percent frequency (percent of shrews that had Twenty-six shrews were processed for ecto- eaten a particular food item). We then used a parasites using the washing technique outlined MANOVA followed by a series of Student-New- by Ritzi and Whitaker (2003). Hosts were man-Keuls multiple range tests to compare the placed in a glass jar with soapy water to cover, diet of B. hylophaga in this study to data col- shaken vigorously for 1 to 2 minutes, and re- lected previously for B. brevicauda (Mumford moved from the wash after being rinsed with and Whitaker, 1982) and B. carolinensis (Whi- 70% ethanol. The wash was then vacuum fil- taker et al., 1994). tered using filter paper and a Buchner funnel. Ectoparasites Only 2 species of parasite are While the wash was filtering, the jar and lid known for B. hylophaga, a laelapid mite, Eulae- were rinsed (into the funnel) with ethanol to laps stabularis, from Kansas ( Jameson, 1947), maximize ectoparasite recovery and prevent and a flea, Stenoponia americana, from Kansas contamination between hosts. The filter papers (Poorbaugh and Gier, 1961). Both records 90 The Southwestern Naturalist vol. 50, no. 1

TABLE 1—Ectoparasites from Blarina hylophaga from north-central Kansas, with comparative prevalence (Prev) and mean intensities (MI) from B. brevicauda and B. carolinensis. Data on B. brevicauda compiled from Mumford and Whitaker (1982) and Pascal (1984), and data on B. carolinensis compiled from Pascal (1984) and Whitaker et al. (1994).

B. hylophaga B. brevicauda B. carolinensis (n ϭ 26) (n ϭ 131) (n ϭ 126) Ectoparasite n Preva MIb Prev MI Prev MI Insecta Siphonaptera Corrodopsylla curvata (Rothschild)c 58 42.3 5.3 6.9 1.7 Ctenopthalmus pseudogyrtes Bakerc 4 7.7 2.0 17.6 1.9 7.1 3 Doratopsylla blarinae Foxc 5 15.4 1.3 6.1 2.4 19.0 1.5 Epitedia wenmanni (Rothschild)c 1 3.9 1.0 3.8 1.4 Orchopeas leucopus (Baker)c 2 7.7 1.0 Acarina Laelapidae Androlaelaps fahrenholzi (Berlese)c 14 19.2 2.8 27.5 2.6 35.7 1.7 Echinonyssus blarinae (Herrin)c 107 53.9 7.6 1.5 1.0 25.4 6.8 Eulaelaps stabularis (Koch)d 3 7.7 1.5 6.9 1.0 10.3 1.2 Haemogamasus liponyssoides Ewingc 20 34.6 2.2 16.0 2.0 29.4 2.6 Glycyphagidae Glycyphagus hypudaei (Koch)c 37 26.9 5.3 2.3 3 9.5 1.0 Orycteroxenus soricis (Oudemans)c 2,503 92.3 104.0 34.4 34.0 47.6 6.3 Listorphoridae Asiochirus blarina Fain and Hylandc 462 69.2 25.6 12.2 163.3 67.4 9.4 Geomylichus texana Fain et al.c 17 11.5 5.7 Myobiidae Blarinobia simplex (Ewing)c 112 50.0 8.6 14.5 8.3 12.7 3.1 Protomyobia blarinae Lukoschus et al.c 114 61.6 7.1 9.9 8.3 27.8 6.5 Pygmeophoridae Bakerdaniac 2 3.9 2.0 4.6 1.2 30.2 2.6 Pygmephorus whitakeri Mahunkac 1 3.9 1.0 2.3 2.3 7.1 1.2 Trombiculidae Euschoengastia diversa Loomisc 1 3.9 1.0 Neotrombicula ﬁtchi (Loomis)c 8 15.4 2.0 Cyrtolaelapidae Cyrtolaelapsc 1 3.9 1.0 1.5 10.5 25.4 2.5 Myocoptidae Myocoptes musculinus (Koch)c 3 3.9 3.0 Rhodacaridae Rhodacarusc 2 3.9 2.0 Ixodidae Dermacentor variabilis (Say)c 1 3.9 1.0 11.1 4.2 a Prevalence reported as a percentage. b Mean intensity as average number of parasites per infected host. c New host record. d New state host locality record.

have been attributed to B. carolinensis, but the (Table 1). This included a total of 22 new host hosts were actually B. hylophaga (Genoways and records for B. hylophaga (i.e., the first report of Choate, 1998). Our examination of 26 B. hylo- a particular parasite species on this host). Also, phaga yielded 3,478 ectoparasites of 23 species E. stabularis previously had not been reported March 2005 Notes 91 from B. hylophaga from Kansas and, thus, is a TABLE 2—Food habits of 25 Elliot’s short-tailed new host-locality record for the state. We did shrews (Blarina hylophaga) from north-central Kan- not recover S. americana. sas. All animals (Metazoa) are adults unless other- The 24 species of ectoparasites now known wise indicated. Volume (Vol.) is the average percentage of the stomach contents for each item; from B. hylophaga contrast with 44 species as- frequency (Freq.) is the percentage of shrews whose sociated with B. carolinensis and 120 species as- stomachs contained a food item. sociated with B. brevicauda. Of these 24 parasites, B. hylophaga shares 16 with B. carolinensis Vol. Freq. and 21 with B. brevicauda. Fifteen species of Food (%) (%) parasites are shared by all 3 host species. The only parasites recovered exclusively from B. hy- Metazoa (total) 97.0 100 lophaga were Geomylichus texana, Neotrombicula Arthropoda (total) 71.8 100 Insecta (total) 61.6 96 fitchi, and Myocoptes musculinus. Both N. fitchi Insecta (unknown) 4.6 16 and M. musculinus commonly are associated Coleoptera (total) 50.3 96 with other hosts native to the Great Plains, and Coleoptera (larvae) 17.0 32 in our opinion, they represent accidental oc- Coleoptera (unknown) 3.6 16 currences. The 17 specimens of G. texana col- Carabidae 14.1 36 lected from 3 shrews captured in Rooks Coun- Scarabidae (total) 14.0 52 ty, Kansas, are more difficult to explain. Listro- Scarabidae (unknown) 12.6 52 phorid mites are highly co-evolved with their Cotinis 0.2 4 hosts, and host switching typically results in Phanaeus 1.2 4 speciation events (Fain, 1994). The specimens Curculionidae 1.6 8 Lepidoptera trace 4 of G. texana recovered during this study are Diptera 0.4 8 morphologically identical to museum speci- Isoptera 2.4 4 mens obtained from Ord’s kangaroo rat (Di- Orthoptera (Gryllidae) 2.4 8 podomys ordii). Resolving this relationship Homoptera (Cicadellidae) should be a goal of future studies. Draeculaecephalis 0.6 4 Although the 3 species of Blarina share many Hymenoptera (Formicidae) 0.8 4 parasites, there is interspecific variation in the Arachnida (Araneae) 7.3 32 composition of these communities. In partic- Isopoda 3.0 4 ular, each species of Blarina tends to be infest- Mollusca (Gastropoda) 21.2 24 ed with different flea species: Ctenophthalmus Chordata (Rodentia) ␹2 ϭ Peromyscus maniculatus? 4.0 4 pseudagyrtes on B. brevicauda ( (dfϭ2) 39.7, P Ͻ Plantae (total) 1.0 12 0.001), Doratopsylla blarinae on B. carolinensis Forbs (unknown) 0.6 8 ␹2 ϭ ϭ ( (dfϭ2) 6.5, P 0.05), and Corrodopsylla cur- Poaceae 0.4 8 ␹2 ϭ Ͻ vata on B. hylophaga ( (dfϭ2) 459.7, P Fungi (total) 2.0 12 0.001). Laelapid mites also provide evidence of Basidiomycota (unknown) 1.2 8 interspecific variation. Echinonyssus blarinae was Zygomycota (cf. Endogone) 0.8 8 ␹2 more often associated with B. hylophaga ( (dfϭ2) ϭ 386.2, P Ͻ 0.001) than the other 2 species ␹2 of Blarina, whereas H. liponyssoides was found ga, and low numbers on B. carolinensis ( (dfϭ2) more often on B. carolinensis than expected ϭ 928.4, P Ͻ 0.001). The other mites and in- ␹2 ϭ Ͻ ( (dfϭ2) 24.0, P 0.001). The glycyphagid sects found on all 3 hosts were distributed mites Glycyphagus hypudaei and Orycteroxenus sor- equally across the 3 species. icis were both significantly more common on Diet Twenty-five stomachs contained food, ␹2 ϭ Ͻ B. hylophaga ( (dfϭ2) 273.5, P 0.001 and whereas 5 were empty and excluded from fur- ␹2 ϭ Ͻ (dfϭ2) 12,598.2, P 0.001, respectively) ther analysis (Table 2). Insects (61.6% volume, than on the other 2 hosts, as were the myobiid 100% occurrence), especially beetles (50.3% ␹2 ϭ Ͻ mites, Blarinobia simplex ( (dfϭ2) 302.5, P and 96%, respectively), were the most com- ␹2 ϭ 0.001) and Protomyobia blarinae ( (dfϭ2) 1,843, mon food, followed by slugs (21.2% and 24%, P Ͻ 0.001). Finally, the listrophorid mite Asi- respectively) and spiders (7.3% and 32%, re- ochirus blarina was found in large numbers on spectively). B. brevicauda, moderate numbers on B. hylopha- The diet of B. hylophaga was similar in most 92 The Southwestern Naturalist vol. 50, no. 1

TABLE 3—Summarized food habits for 3 species of items, to B. brevicauda, which consumes rela- Blarina reported as untransformed mean percent tively soft food items. Studies of shrews in ge- volumes of each food group. Data for this summary nus Sorex from the Paciﬁc Northwest (Carraway are available in Mumford and Whitaker (1982) for and Verts, 1994; Carraway et al., 1996) have re- B. brevicauda, Whitaker et al. (1994) for B. carolinen- vealed relationships between the diets of these sis, and Table 2 for B. hylophaga. Signiﬁcantly distinct groups (Student-Newman-Keuls multiple range tests, shrews and their ability to generate bite force. ␣ϭ0.05) are denoted by letters (a, b, c) following Thus, we suggest mensural data collected dur- percent volumes. Percent values with multiple letters ing earlier taxonomic studies of Blarina be re- represent cases where a value was indistinguishable examined in light of the functional morphol- from more than 1 other species. ogy of the jaws.

Volume (%) We thank Fort Hays State University, the Stern- berg Museum of Natural History, and Indiana State B. brevi- B. caro- B. hylo- University for enabling this research. J. R. Choate, J. cauda linensis phaga R. Thomasson, and E. Gillock provided logistical ϭ ϭ ϭ Food item (n 125) (n 41) (n 25) support and access to specimens. Thanks go to G. Coleoptera Liggett and C. Liggett for housing CMR and DWS Larva 5.6 a 8.2 ab 17.0 b while conducting this study at the Sternberg Muse- Adult 8.7 a 12.0 a 33.3 b um of Natural History. Thanks to J. R Choate, J. O. Lepidoptera Whitaker, Jr., and an anonymous reviewer for im- Larva 8.2 3.7 0.0 proving earlier versions of this manuscript, and to J. Orthoptera 8.1 2.1 2.4 H. Magers for aid in composing the Spanish resu- Other Arthropoda 17.8 22.7 19.1 men. This study was funded primarily by a grant to Annelida 35.7 a 14.8 b 0.0 c DWS and CMR from the Indiana Academy of Sci- Gastropoda 8.6 18.0 21.2 ence. Additional funding was provided to CMR from Vertebrata 0.1 1.0 4.0 the Indiana State University Ofﬁce of Sponsored Plantae 3.6 1.2 1.0 Programs Student Research Scholarship, and the In- Fungi 3.6 a 16.3 b 2.0 a diana State University Graduate Research Fund. DWS was supported during this work by a grant from the Indianapolis International Airport.

respects to that of the other species of Blarina, LITERATURE CITED consisting primarily of animal matter, although the shrews occasionally take small amounts of ANIMAL CARE AND USE COMMITTEE. 1998. Guidelines fungi and higher plants. To make direct com- for the capture, handling, and care of mammals parisons among and between the 3 species of as approved by the American Society of Mam- shrews, we summarized the data on food habits malogists. Journal of Mammalogy 79:1416–1431. for the 3 species into 10 categories (Table 3). BUSH, A. O., K. D. LAFFERTY,J.M.LOTZ, AND A. W. A MANOVA revealed that the overall diet of SHOSTAK. 1997. Parasitology meets ecology on its own terms: Margolis et al. revisited. Journal of the 3 shrews varied significantly between spe- Parasitology 83:575–583. cies (␭ϭ0.581, P Ͻ 0.001). A series of Student- CARRAWAY, L. J., AND B. J. VERTS. 1994. Relationship Newman-Keuls multiple range tests indicated of mandibular morphology to relative bite-force ␣ϭ that the 3 species all varied significantly ( in some western Sorex. Carnegie Museum of Nat- 0.05) from one another in terms of the per- ural History Special Publications 18:201–210. cent volume of earthworm in the diet. Earth- CARRAWAY, L. J., B. J. VERTS,M.L.JONES, AND J. O. worms were the most common individual food WHITAKER,JR. 1996. A search for age-related item for B. brevicauda in Indiana, while we changes in bite force and diet in shrews. Ameri- failed to record any in the diet of B. hylophaga can Midland Naturalist 135:231–240. from Kansas. Blarina hylophaga consumed more FAIN, A. 1994. Adaptation, specificity and host-parasite coevolution in mites (Acari). International adult beetles than the other species, and more Journal of Parasitology 24:1273–1283. beetle larvae than B. brevicauda. Blarina caroli- FAIN, A., J. O. WHITAKER,JR., AND H. H. THOMAS. nensis consumed more fungi than the other 1991. Two new fur-mites of the genus Geomylichus species. These observations suggest that hard- Fain, 1970 (Acari, Listrophoridae) from kanga- ness of the diet of these species range from B. roo rats, Dipodomys spp., from the U.S.A. Inter- hylophaga, which focuses on hard-bodied food national Journal of Acarology 17:175–180. March 2005 Notes 93

GENOWAYS, H. H., AND J. R. CHOATE. 1972. A multi- PASCAL, D. D., JR. 1984. A taxonomic study of Mid- variate analysis of systematic relationships among western short-tailed shrews (genus Blarina) with populations of the short-tailed shrew (genus Blar- emphasis upon ectoparasites. Unpublished Ph.D. ina) in Nebraska. Systematic Zoology 21:106–116. dissertation, Indiana State University, Terre GENOWAYS, H. H., AND J. R. CHOATE. 1998. Natural Haute. history of the southern short-tailed shrew, Blarina POORBAUGH,J.H.,AND H. T. GIER. 1961. Fleas (Si- carolinensis. Occasional Papers of the Museum of phonaptera) of small mammals in Kansas. Jour- Southwestern Biology 8:1–43. nal of the Kansas Entomological Society 34:198– GEORGE, S. B., J. R. CHOATE, AND H. H. GENOWAYS. 204. 1981. Distribution and taxonomic status of Blari- RITZI, C. M., AND D. W. SPARKS. 2002. New ectopar- na hylophaga Elliot (Insectivora: Soricidae). An- asite records for the rock squirrel, Spermophilus nals of the Carnegie Museum, 50:493–513. variegatus grammurus, in Socorro County, New GEORGE, S. B., H. H. GENOWAYS,J.R.CHOATE, AND Mexico. Southwestern Entomologist 27:257–262. R. J. BAKER. 1982. Karyotypic relationships within RITZI, C. M., AND J. O. WHITAKER,JR. 2003. Ectopar- the short-tailed shrews, genus Blarina. Journal of asites of small mammals from the Newport Mammalogy, 63:639–645. Chemical Depot, Vermillion County, Indiana. HANDLEY, C. O., JR., AND M. VARN. 1994. Identiﬁca- Northeastern Naturalist 10:149–158. tion of the Carolinian shrews of Bachman 1837. WHITAKER, J. O., JR. 1982. Ectoparasites of mammals In: J. F. Merritt, G. L. Kirkland, Jr., and R. K. of Indiana. Monograph Number 4, Indiana Acad- Rose, editors. Advances in the biology of shrews. emy of Science. Special Publication 18, Carnegie Museum of Nat- WHITAKER, J. O., JR., G. D. HARTMAN, AND R. HEIN. ural History, Pittsburgh, Pennsylvania. Pp. 393– 1994. Food and ectoparasites of the southern 406. short-tailed shrew, Blarina carolinensis (Mamma- JAMESON, E. W., JR. 1947. Natural history of the prai- lia: Soricidae), from South Carolina. Brimleyana rie vole (mammalian genus Microtus). University 21:97–105. of Kansas Publication, Museum of Natural His- WHITAKER, J. O., JR., AND R. E. MUMFORD. 1972. Food tory 1:125–151. and ectoparasites of Indiana shrews. Journal of KRANTZ, G. W. 1986. A manual of acarology, second Mammalogy 53:329–335. edition. Oregon State University Book Stores, WHITAKER, J. O., JR., W. J. WRENN, AND R. E. LEWIS. Inc., Corvallis. 1993. Parasites. In: H. H. Genoways and J. H. LUKOSCHUS, F. S., G. J. JEUCKEN, AND J. O. WHITAKER, Brown, editors. Biology of the Heteromyidae. JR. 1988. A review of the Protomyobia americana Special Publication Number 10, American Soci- group (Acarina: Prostigmata: Myobiidae) with descriptions of Protomyobia panamensis n. sp. and Pro- ety of Mammalogists. Pp. 386–478. tomyobia blarinae n. sp. Journal of Parasitology 74: ZAR, J. H. 1996. Biostatistical analysis, third edition. 305–316. Prentice Hall, Upper Saddle River, New Jersey. MUMFORD,R.E.,AND WHITAKER, J. O., JR. 1982. Mam- mals of Indiana. Indiana University Press, Bloo- Submitted 11 November 2003. Accepted 2 April 2004. mington. Associate Editor was Cody W. Edwards. 94 The Southwestern Naturalist vol. 50, no. 1

NEST RECORD OF SOREX MILLERI FROM MADERAS DEL CARMEN, ME´XICO

JONA´ S A. DELGADILLO VILLALOBOS,BONNIE REYNOLDS MCKINNEY,* FELICIANAO HEREDIA PINEDA, AND SANTIAGO GIBERT ISERN

Proyecto El Carmen, Cemex Central S.A. de C.V., Ave. Constitucı´on 444 Pte. 64000 Monterrey, Nuevo Leo´n, Me´xico (JDV, BRM, FHP, SGI) Present address of BRM: Proyecto El Carmen, P.O. Box 420608, Del Rio, TX 78842 *Correspondent: [email protected]

ABSTRACT We report on the ﬁrst documented nest of Miller’s shrew (Sorex milleri) from the Maderas del Carmen, Coahuila, Me´xico. We located 7 nests in mesic montane forest at 2,400 to 2,700 m elevation during October and November 2003.

RESUMEN Reportamos el primer nido documentado de la musaran˜a (Sorex milleri), de Maderas del Carmen, Coahuila, Me´xico. Se localizo´ un total de 7 nidos en las mesetas boscosas hu´medas de 2,400 hasta 2,700 m de elevacio´n durante octubre y noviembre del 2003.

We report on the first documented nest of in western Texas and 165 km northwest of Miller’s shrew (Sorex milleri) and provide addi- Muzquiz, Coahuila, Mexico (29Њ04Ј06ЉN, tional vouchered records of the occurrence of 102Њ37Ј17ЉW). Maderas del Carmen is a typical this species in Maderas del Carmen, Coahuila, sky island surrounded by Chihuahuan Desert Me´xico. Little information on the life history with a range in elevation from 900 m to 2,700 of this species is known for northern Coahuila. m. Dominant trees above 2,000 m are fir, pine- While collecting mammals during October and oak, oak-pine, and pine associations. November 1940, F. W. Miller obtained 3 spec- In October 2003, while we were conducting imens of Miller’s shrew, including the type small mammal surveys as part of the baseline specimen from the Maderas del Carmen at inventory of the Maderas del Carmen, we Campo Madera, an abandoned lumber camp found an active Miller’s shrew nest at Cuadro in a mesic montane forest dominated by Doug- Pelota (29Њ56Ј94ЉN, 102Њ35Ј14ЉW). The nest las-fir (Pseudotsuga menziesii), Coahuila fir (Abies was located in a forest opening with decaying coahuilensis), and southwestern white pine (Pi- tree trunks and logs surrounded by mesic nus strobiformis) ( Jackson, 1947; Baker, 1956). montane forest at an elevation of 2,400 m. Baker (1956) captured 23 Miller’s shrews in Ground cover consisted of grasses, including southeastern Coahuila about 20 km east of San poverty oat grass (Danthonia spicata), tobosa Antonio de las Alazanas at an elevation of grass (Hilaria mutica), bull muhly (Muhlenbergia 2,805 m in an association of Douglas-fir, pine, emersleyi), and Pringle needlegrass (Piptochae- and aspen (Populus tremuloides). Findley (1955) tum pringlei), with scattered pines, Douglas-fir, compared the Miller’s shrew that Jackson col- and silverleaf oak (Quercus hypoleucoides). Lo- lected to other shrews and considered the Mill- cated at the entrance of a Botta’s pocket go- er’s shrew to be a relict population of S. ciner- pher (Thomomys bottae) burrow under a rotting eus isolated in the Sierra Madre Oriental in log 25 cm in diameter, nest measurement from northeastern Mexico. Miller’s shrew has been the outside edges was 10 cm in diameter across documented from Cerro Potosı´ in the adjacent the top, which was open and flush with the state of Nuevo Leoń, which is part of the Sierra underside of the log. The nest showed visible Madre Oriental (Baker, 1956). Miller’s shrew is signs of wear and was a cup-like, loosely con- currently listed as ‘‘protected and endemic’’ in structed mass of grass blades. Pine needles, bits the NORMA Oficial Mexicana (NOM, 2002). of bark, and oak leaves were around the bot- Maderas del Carmen is a mountain range lo- tom of the nest. The nest contained a side en- cated 60 km south of Big Bend National Park trance hole that was 19 mm in diameter. March 2005 Notes 95

The nest contained 1 adult Miller’s shrew, habits the lower desert elevations to the higher which fled into the Botta’s pocket gopher tun- grasslands that are interspersed with beaked nel, and 6 young that scrambled from the nest yucca (Yucca rostrata), sotol (Dasylirion leiophyl- when disturbed. We hand-captured the young, lum), and junipers (Juniperus) from 1,000 to photographed them, and retained 2 for vouch- 1,415 m, and the Miller’s shrew inhabits the er specimens (Proyecto El Carmen, Registro mesic montane forest from 1,900 m to 2,700 m de Coleccioń Cientifica, Clave: NL-MAM-139- in the Maderas del Carmen in northern Coa- 11-02; 046, 047). The adult Miller’s shrew was huila, Me´xico. brownish gray, with an elongated nose, and was larger than the young. The young had their This research project was funded by Cemex and eyes closed, their elongated noses were fleshy conducted as part of the Proyecto El Carmen within pink, the pelage was short, soft, and chocolate the Greater Maderas Del Carmen Ecosystem. We brown, the tail was short and hairless. When greatly appreciate the unwavering support of Ce- mex, which has made a long-term commitment to the vouchers were prepared, we noted the the conservation of wildlife in Me´xico. We thank L. skulls of the 2 specimens were not ossified. The Harveson and R. Valdez for reviewing the manu- total length, tail, hindfoot, and ear of the 2 script. We especially thank R. Baker for his insightful voucher specimens were: 82, 83 mm; 28, 29 suggestions and inspiration, and in recognition of mm; 9, 9 mm; and 4, 4 mm, respectively. Other his studies on the mammals of Coahuila. small mammals collected in the same locality were Botta’s pocket gopher, yellow-nosed cot- LITERATURE CITED ton rat (Sigmodon ochrognathus ochrognathus), brush mouse (Peromyscus boylii), and western BAKER, R. H. 1956. Mammals of Coahuila, Me´xico. harvest mouse (Reithrodontomys megalotis). University of Kansas Publication, Museum of Nat- ural History 9:125–335. On the same day, 2 additional nests were Њ Ј Љ FINDLEY, J. S. 1955. Taxonomy and distribution of found at Mesa Bonita (23 00 67 N, some American shrews. University of Kansas Pub- Њ Ј Љ 102 36 76 W), 12 km north of Cuadro Pelota lication, Museum of Natural History 7:613–618. at 2,700 m in similar habitat. Four additional JACKSON, H. H. T. 1947. A new shrew (genus Sorex) unoccupied nests were located on 8 November from Coahuila. Proceedings of the Biological So- 2003 at Mesa Bonita. These 6 nests were con- ciety of Washington 60:131–132. structed in the same manner with the same NOM. 2002. NORMA Oficial Mexicana. NOM-059- nest materials, and all were under decaying ECOL-2001, Proteccioń ambiental nativas de logs in old Botta’s pocket gopher tunnels. Me´xico de flora y fauna silvestres-Categorıás de riegso y especificaciones para su inclusioń, exclu- These nests were in various stages of disarray sioń o cambio-Lista de especies en riesgo. Segun- from use, and 1 nest contained several small da Seccion, Secretaria de Medio Ambiente y Re- feathers. cursos Naturales 2:1–81. Intense baseline inventory for 2 years (2001 to 2003) in the Maderas del Carmen revealed Submitted 9 December 2003. Accepted 30 April 2004. that the desert shrew (Notiosorex crawfordi) in- Associate Editor was Cody W. Edwards. 96 The Southwestern Naturalist vol. 50, no. 1

THREE NEW RECORDS OF BATS FROM EL SALVADOR

JAMES G. OWEN

Universidad Salvadoren˜a ‘‘Alberto Masferrer,’’ A.P. 2053, San Salvador, El Salvador, Central America *Correspondent: [email protected]

ABSTRACT I report distributional records that document the presence of 3 additional species of bats (Carollia sowelli, Lasiurus ega, and Eptesicus furinalis) for the fauna of El Salvador, Central America.

RESUMEN Se reportan registros de distribucio´n que documentan la presencia de 3 especies adicionales de murcie´lagos (Carollia sowelli, Lasiurus ega, y Eptesicus furinalis) para la fauna de El Salvador, Centroame´rica.

Geographical range maps in Hall (1981) cluding El Salvador, were previously referred and Reid (1997) tacitly predict the occurrence to C. brevicauda (Pine, 1972). Central American of Carollia sowelli (Sowell’s short-tailed bat), populations of C. brevicauda are now thought Lasiurus ega (southern yellow bat), and Eptesi- to be restricted to eastern Panama (Wright et cus furinalis (Argentine brown bat) in El Sal- al., 1999; Baker et al., 2002). Carollia sowelli is vador. Although each of these species is widely represented by 6 males (CM 114837–114841, distributed in Central America (Hall, 1981; 114843) and 6 females (CM 114836, 114842, Reid, 1997), none has actually been docu- 114844–144847) from northwestern El Salva- mented from this country. I report the first dor, 3.4 km S La Palma, Department of Chal- Њ Ј Њ Ј specimens of these bats from El Salvador. atenango (14 17 N, 89 09 W; 980 m elevation; These 3 records bring the total number of spe- Subtropical Very Moist Forest). The sample cies of bats reported for El Salvador to 62 (Ta- size, mean, and range for length of forearm ble 1), representing about 50% of the total and greatest length of skull, respectively, are: n ϭ ϭ ϭ mammalian fauna of the country (Felten, 12, mean 40.5 mm, range 39.8 to 41.4 ϭ ϭ ϭ 1955, 1956a, 1956b, 1956c, 1957; Burt and Stir- mm; n 12, mean 22.5 mm, range 22.3 ton, 1961; Davis, 1968; Jones and Bleier, 1974; to 22.9 mm. These measurements, as well as distinctive characteristics of the pelage, match Hellebuyck et al., 1985; Owen et al., 1991, those presented by Pine (1972) and Owen et 1993; Engstrom et al., 1994). al. (1984) for Middle American C. sowelli, I examined specimens cited in this report, which they recorded as C. brevicauda. and they are now deposited in the following At the latitude of northwestern El Salvador, institutions: Carnegie Museum of Natural His- C. sowelli has been collected from the Atlantic tory (CM), Museo de Historia Natural de El versant (Nueva Ocotepeque, Honduras, 840 m Salvador (MUHNES), Texas Cooperative Wild- elevation) down to the Caribbean lowlands life Collection (TCWC), and the Museum of (Pine, 1972; Hall, 1981). This report docu- Texas Tech University (TTU). Capitalized life ments its occurrence at about the same lati- zones are those of Holdridge (1975). Eleva- tude as Nueva Ocotepeque from the Pacific tions and geographical coordinates are those versant as well. This species might occur across recorded by the original collectors, when avail- much of northern El Salvador, where it probable, otherwise I estimated them using topo- ably is restricted to humid montane localities. graphical maps at a scale of 1 to 50,000, pub- In the area of the Gulf of Fonseca, C. sowelli is lished by the Salvadoran Ministry of Public known from the Pacific lowlands of Honduras, Works. 6 km east of El Amatillo, El Salvador, 60 m A collection of bats of the genus Carollia elevation (Pine, 1972; Hall, 1981). I predict from El Salvador includes the species C. perspi- that this species also occurs in the Pacific low- cillata, C. sowelli, and C. subrufa. Specimens of lands of El Salvador, in the eastern Depart- C. sowelli from western Panama northward, in- ment of La Unioń. March 2005 Notes 97

TABLE 1—Species of bats collected in El Salvador. TABLE 1—Continued.

Taxon1,2,3 Common name4 Taxon1,2,3 Common name4 Emballonuridae C. villosum Hairy big-eyed bat Balantiopteryx plicata Gray sac-winged bag Dermanura azteca Aztec fruit-eating bat Peropteryx macrotis Lesser dog-like bat D. phaeotis Pygmy fruit-eating bat Rhynchonycteris naso Proboscis bat D. tolteca Toltec fruit-eating bat Saccopteryx bilineata Greater sac-winged bat Enchisthenes hartii Velvety fruit-eating bat S. leptura Lesser sac-winged bat Platyrrhinus helleri Heller’s broad-nosed bat Sturnira lilium Little yellow-shouldered Noctilionidae bat Noctilio albiventris Lesser bulldog bat S. ludovici Highland yellow-shoul- N. leporinus Greater bulldog bat dered bat Mormoopidae Uroderma bilobatum Tent-making bat Mormoops megalophylla Ghost-faced bat U. magnirostrum Brown tent-making bat Pteronotus davyi Davy’s naked-backed bat Natalidae P. gymnonotus Big naked-backed bat Natalus stramineus Mexican funnel-eared P. parnellii Parnell’s mustached bat bat P. personatus Wagner’s mustached bat Vespertilionidae Phyllostomidae Eptesicus furinalis Argentine brown bat Micronycterinae E. fuscus Big brown bat Micronycteris microtis Little big-eared bat Lasiurus blossevillii Western red bat Desmodontinae L. ega Southern yellow bat Desmodus rotundus Vampire bat L. intermedius Northern yellow bat Diaemus youngi White-winged vampire Myotis elegans Elegant myotis bat M. keaysi Hairy-legged myotis Diphylla ecaudata Hairy-legged vampire bat M. nigricans Black myotis Lonchorhininae M. velifer Cave myotis Lonchorhina aurita Tomes’ sword-nosed bat Rhogeessa tumida Black-winged little yellow Phyllostominae bat Chrotopterus auritus Big-eared woolly bat Macrophyllum macro- Molossidae phyllum Long-legged bat Eumops auripendulus Black bonneted bat Phyllostomus discolor Pale spear-nosed bat E. underwoodi Underwood’s bonneted Trachops cirrhosus Fringe-lipped bat bat Glossophaginae Molossus rufus Black mastiff bat Anoura geoffroyi Geoffroy’s tailless bat M. molossus Pallas’ mastiff bat Choeroniscus godmani Godman’s long-tailed bat Nyctinomops laticauda- Glossophaga commissar- Commissariss’ long- tus Broad-eared bat isi tongued bat 1 Subfamilies of the family Phyllostomidae are G. leachii Gray long-tongued bat those proposed by Baker et al. (2003). G. soricina Pallas’ long-tongued bat 2 Myotis fortidens, the cinnamon myotis, has been Leptonycteris curasoae Southern long-nosed bat collected from the Paciﬁc lowlands of Guatemala, Carolliinae about 23 kilometers from the border of El Salvador Carollia perspicillata Seba’s short-tailed bat (Dolan and Carter, 1979) and by Dickerman et al. C. sowelli Sowell’s short-tailed bat (1981) at Montufar, Guatemala, near the southwest- C. subrufa Gray short-tailed bat ern border of El Salvador. Based on geographical Stenodermatinae and habitat considerations, M. fortidens probably also Artibeus inopinatus Honduran fruit-eating occurs in the adjacent lowlands of El Salvador. bat 3 Molecular data suggest that A. intermedius might A. intermedius Intermediate fruit-eating not be speciﬁcally distinct from A. lituratus (Lim et bat al., 2004). A. jamaicensis Jamaican fruit-eating bat 4 Common names mostly follow Wilson and Cole A. lituratus Great fruit-eating bat (2000). Centurio senex Wrinkle-faced bat Chiroderma salvini Salvin’s big-eyed bat 98 The Southwestern Naturalist vol. 50, no. 1

On 27 February 1967, a female Lasiurus ega vation; Subtropical Moist Forest); and 1 female (TCWC 19749) was collected from north-cen- (TTU 64080) 4.5 km NW San Luı´s, Depart- tral El Salvador 20 km W Chalatenango, De- ment of La Paz (13Њ29ЈN, 89Њ07ЈW; 76 m ele- partment of Chalatenango (14Њ06ЈN, 89Њ06ЈW; vation; Subtropical Moist Forest). 250 m elevation; Subtropical Moist Forest). The subspecies from El Salvador, E. f. gau- The capture of this specimen in winter sug- meri, is widely distributed in Central America gests that L. ega has migratory movements, at (Dickerman et al., 1981; Hall, 1981; McCarthy, least on a local geographical scale, similar to 1987) and is considered to be a lowland form, those of other North American lasiurines. Las- mainly occurring at elevations Ͻ1,000 m (Da- iurus ega resembles L. intermedius, but in areas vis, 1966). These data suggest that E. furinalis of geographical overlap, L. ega is smaller (Hall is distributed throughout El Salvador from the and Jones, 1961). Length of forearm of this Pacific lowlands up to elevations of about 1,000 specimen is 47.97 mm, which is well below the m. mean and range of 2 specimens of L. intermedius (1 from western El Salvador and 1 syntop- I thank D. Aguilar, R. J. Baker, D. Schlitter, and J. ically with this specimen of L. ega) n ϭ 2, mean R. Wible for permission to examine specimens un- ϭ 54.95 mm, range ϭ 53.04 and 56.86 mm. der their care. D. Wilson and an anonymous review- Lasiurus ega occurs from lowlands (Handley, er made comments that improved this paper. Spec- imens, herein cited as new records for El Salvador, 1976) to elevations of up to about 2,300 m were originally collected in the field by R. J. Baker, (Genoways and Jones, 1968). Based on the W. J. Bleier, R. K. LaVal, and J. G. Owen. Collecting range of elevations from which specimens of permits for Owen were provided by A. Sańchez, Na- this species have been collected and based on tional Parks and Wildlife, Salvadoran Ministry of Ag- its wide geographical distribution in Central riculture. America (Dickerman et al., 1981; Hall, 1981; Dinerstein, 1985; McCarthy, 1987), I predict LITERATURE CITED that L. ega occurs throughout El Salvador, except in the highest cloud forest. BAKER, R. J., S. R. HOOFER,C.A.PORTER, AND R. A. VAN DEN BUSSCHE. 2003. Diversification among Specimens of Eptesicus furinalis have been New World leaf-nosed bats: an evolutionary hy- collected at 8 scattered localities in central El pothesis and classification inferred from digen- Salvador as follows: 1 male (TTU 13371) and omic congruence of DNA sequence. Occasional 3 females (TTU 13372, 13373, 13374) 1.9 km Papers, Museum of Texas Tech University 230:1– W Suchitoto, Department of Cuscatlań 32. (13Њ56ЈN, 89Њ03ЈW; 420 m elevation; Subtropi- BAKER, R. J., S. SOLARI, AND F. G. HOFFMAN. 2002. A cal Moist Forest); 2 males (TTU 13370, 13381) new Central American species from the Carollia and 1 female (TTU 13369) 13.5 km NW Coloń, brevicauda complex. Occasional Papers, Museum Department of La Libertad (13Њ48ЈN, 89Њ24ЈW; of Texas Tech University 217:1–11. 460 m elevation; Subtropical Moist Forest); 2 BURT, W. H., AND R. A. STIRTON. 1961. The mammals of El Salvador. Miscellaneous Publications, Mu- females (TTU 17128, 17129) 3.2 km E Usulu- seum of Zoology, University of Michigan 117:1– tań, Department of Usulutań (13Њ20ЈN, Њ Ј 69. 88 24 W; 80 m elevation; Subtropical Moist DAVIS, W. B. 1966. Review of South American bats of Forest); 1 female (CM 115632) 10 km S Zaca- the genus Eptesicus. Southwestern Naturalist 11: tecoluca, Hacienda Escuintla, Department of 245–274. La Paz (13Њ24ЈN, 88Њ54ЈW; 5 m elevation; Sub- DAVIS, W. B. 1968. Review of the genus Uroderma tropical Moist Forest); 1 male (MUHNES (Chiroptera). Journal of Mammalogy 49:676– 1123) and 1 female (MUHNES 1124) Nancu- 698. chiname National Park, Department of Usu- DICKERMAN, R. W., K. F. KOOPMAN, AND C. SEYMOUR. lutań (13Њ22ЈN, 88Њ43ЈW; 8 m elevation; Sub- 1981. Notes on bats from the Pacific lowlands of Guatemala. Journal of Mammalogy 62:406–411. tropical Moist Forest); 1 male (TCWC 19733) DINERSTEIN, E. 1985. First records of Lasiurus casta- 20 km W Chalatenango, Department of Chal- Њ Ј Њ Ј neus and Antrozous dubiaquercus from Costa Rica. atenango (14 06 N, 89 06 W; 250 m elevation; Journal of Mammalogy 66:411–412. Subtropical Moist Forest); 2 males (TTU 64078 DOLAND, P. G., AND D. C. CARTER. 1979. Distribution 64079) Deininger National Park, Department notes and records for Middle American Chirop- of La Libertad (13Њ29ЈN, 89Њ16ЈW; 20 m ele- tera. Journal of Mammalogy 60:644–649. March 2005 Notes 99

ENGSTROM, M. D., B. K. LIM, AND F. A. REID. 1994. LIM, B. K., M. D. ENGSTROM,T.E.LEE,JR., J. C. PAT- Two small mammals new to the fauna of El Sal- TON, AND J. W. BICKHAM. 2004. Molecular differ- vador. Southwestern Naturalist 39:281–306. entiation of large species of fruit-eating bats (Ar- FELTEN, H. 1955. Fledermaüse (Mammalia, Chirop- tibeus) and phylogenetic relationships based on tera) aus El Salvador. Teil 1. Senckenbergiana the cytochrome b gene. Acta Chiropterologica 6: Biologica 36:271–85. 1–12. FELTEN, H. 1956a. Fledermaüse (Mammalia, Chirop- MCCARTHY, T. J. 1987. Distributional records of bats tera) aus El Salvador. Teil 2. Senckenbergiana from the Caribbean lowlands of Belize and adja- Biologica 37:69–86. cent Guatemala and Mexico. Fieldiana Zoology, FELTEN, H. 1956b. Fledermaüse (Mammalia, Chirop- New Series 39:137–162. tera) aus El Salvador. Teil 3. Senckenbergiana OWEN, J. G., J. ARROYO-CABRALES, AND J. K. JONES,JR. Biologica 37:179–212. 1993. First record of Noctilio albiventris (Chirop- FELTEN, H. 1956c. Fledermaüse (Mammalia, Chirop- tera, Noctilionidae) in El Salvador. Texas Journal tera) aus El Salvador. Teil 4. Senckenbergiana of Science 45:273–274. Biologica 37:341–367. OWEN, J. G., J. K. JONES,JR., AND R. J. BAKER. 1991. FELTEN, H. 1957. Fledermaüse (Mammalia, Chirop- Annotated checklist of land mammals of El Sal- tera) aus El Salvador. Teil 5. Senckenbergiana vador. Occasional Papers, Museum of Texas Tech Biologica 38:1–22. University 139:1–17. GENOWAYS, H. H., AND J. K. JONES,JR. 1968. Notes on OWEN, J. G., D. J. SCHMIDLY, AND W. B. DAVIS. 1984. bats from the Mexican State of Zacatecas. Journal A morphometric analysis of three species of Car- of Mammalogy 49:743–745. ollia (Chiroptera, Glossophaginae) from Middle HALL, E. R. 1981. The mammals of North America, America. Mammalia 48:85–93. second edition. John Wiley and Sons, New York. PINE, R. H. 1972. The bats of the genus Carollia. HALL,E.R.,AND J. K. JONES,JR. 1961. North Ameri- Technical Monograph, Texas Agricultural Exper- can yellow bats, ‘‘Dasypterus,’’ and a list of the iment Station, Texas A&M University, 8:1–125. named kinds of the genus Lasiurus Gray. Univer- REID, F. A. 1997. A field guide to the mammals of sity of Kansas Publications, Museum of Natural Central America & southeast Mexico. Oxford History 14:73–94. University Press, New York. HANDLEY, C. O., JR. 1976. Mammals of the Smithson- WILSON,D.E.,AND F. R. COLE. 2000. Common names ian Venezuelan Project. Brigham Young Univer- of mammals of the world. Smithsonian Institu- sity Science Bulletin, Biological Series 20:1–91. tion Press, Washington. HELLEBUYCK, V., J. R. TAMSITT, AND J. G. HARTMAN. WRIGHT, A. J., R. A. VAN DEN BUSSCHE,B.K.LIM,M. 1985. Records of bats new to El Salvador. Journal D. ENGSTROM, AND R. J. BAKER. 1999. Systematics of Mammalogy 66:783–788. of the genera Carollia and Rhinophylla based on HOLDRIDGE, L. R. 1975. Mapa ecolo´gica de El Salva- dor: memoria explicativa. Ministerio de Agricul- the cytochrome-b gene. Journal of Mammalogy tura y Ganaderıá, San Salvador, El Salvador. 80:1202–1213. JONES,J.K.,JR., AND W. J. BLEIER. 1974. Sanborn’s long-tongued bat, Leptonycteris sanborni, in El Sal- Submitted 25 February 2003. Accepted 5 April 2004. vador. Mammalia 38:144–145. Associate Editor was Cheri A. Jones. 100 The Southwestern Naturalist vol. 50, no. 1

RANGE EXPANSION OF RED FOXES IN NORTHWESTERN TEXAS AND NORTHEASTERN NEW MEXICO

JAN F. KAMLER,WARREN B. BALLARD,* ROBERT L. HARRISON, AND C. GREGORY SCHMITT

Department of Range, Wildlife, and Fisheries Management, Box 42125, Texas Tech University, Lubbock, TX 79409 (JFK, WBB) Department of Biology, University of New Mexico, Albuquerque, NM 87131 (RLH) New Mexico Department of Game and Fish, Box 25112, Santa Fe, NM 87504 (CGS) Present address of JFK: Wildlife Conservation Research Unit, Department of Zoology, South Parks Road, Oxford, 0X1 3PS, United Kingdom *Correspondent: [email protected]

ABSTRACT We recently collected specimens of red foxes (Vulpes vulpes) in northwestern Texas and northeastern New Mexico that extend the known range of this species in both states. In Texas, the new record likely represents a western range expansion of introduced red foxes that had spread across the state during the past 100 years. In New Mexico, the new record likely represents a westward expansion from Texas, or possibly an eastward or northward expansion from previously documented populations in the state.

RESUMEN Recientemente colectamos especı´menes de zorra roja (Vulpes vulpes) en el noroeste de Texas y noreste de Nuevo Me´xico, con lo que se extiende la distribucioń conocida de esta especie en ambos estados. El nuevo registro en Texas probablemente representa la expansioń hacia el oeste de zorras rojas introducidas que se han extendido a trave´s del estado en los u´ltimos 100 anõs. El nuevo registro en Nuevo Me´xico probablemente representa la expansioń hacia el oeste desde Texas, o posiblemente la expansioń hacia el este o norte de poblaciones previamente documentadas en el estado.

Red foxes (Vulpes vulpes) are common fauna (Packard and Bowers, 1970; Davis, 1974). In in most areas of North America. However, red northwestern Texas, the western limit of red foxes are absent from many semi-arid regions, foxes was reported as the central Panhandle including lowland areas of the southwestern (Davis and Schmidly, 1994; Choate, 1997). United States (Hall and Kelson, 1959; Samuel In September 2000, we collected remains of and Nelson, 1982; Voigt, 1987). In Texas and a road-killed red fox in Stratford, in western New Mexico, red foxes reportedly occur in Sherman County, Texas (TTU #85507, The most areas of both states, except for large por- Museum, Texas Tech University, Lubbock). tions of the shortgrass prairie region in ex- This record extends the known range of red treme western Texas and eastern New Mexico foxes approximately 75 km west than previous- (Findley et al., 1975; Davis and Schmidly, 1994; ly reported in the northern Texas Panhandle Choate, 1997). (Davis and Schmidly, 1994). We also observed In Texas, red foxes are considered nonna- red foxes in some towns of adjacent counties tive, as they did not historically occur in the to the west and south of Sherman County, sug- state (Davis and Schmidly, 1994). Starting in gesting red foxes likely occur throughout the 1895, red foxes (probably from states to the entire western part of the Texas Panhandle. east) were introduced for sport in the central In New Mexico, all records and specimens and eastern portions of Texas, where they es- of red foxes (n Ͼ 30) before 1950 were from tablished populations by the 1940s (Strecker, montane habitats and valleys throughout cen- 1926; Taylor and Davis, 1947; Davis and tral and western New Mexico (Townsend, Schmidly, 1994). By the 1960s, red foxes 1893; Bailey, 1932; Halloran, 1946; Clothier, spread north and west in Texas, and were re- 1957; Hall and Kelson, 1959; Findley et al., ported as far north as Swisher County, and as 1975), indicating native red foxes historically far west as Hale, Lubbock, and Upton counties occupied the western two-thirds of the state. March 2005 Notes 101

For example, red foxes were reported from the LITERATURE CITED San Juan and Sangre de Cristo mountains across northern New Mexico (Bailey, 1932; Fin- BAILEY, V. 1932. Mammals of New Mexico. North dley et al., 1975), southwards as far as the San American Fauna 53:1–412. Andres (Halloran, 1946; Hall and Kelson, CHOATE, L. L. 1997. The mammals of the Llano Es- 1959; Findley et al., 1975) and Mogollon (Bai- tacado. Special Publications of The Museum, ley, 1932) mountains in southern and western Texas Tech University, Number 40. New Mexico, respectively. Historically, red fox- CLOTHIER, R. 1957. Distribution of the mammals of es were rare or absent from the plains of east- the Sandia and Manzano mountains, New Mexi- ern New Mexico. A red fox captured by a trap- co. Unpublished Ph.D. dissertation, University of per in Chaves County in 1958 was the ﬁrst re- New Mexico, Albuquerque. cord from eastern New Mexico (Findley et al., DAVIS, W. B. 1974. The mammals of Texas. Texas 1975). The trapper had never seen a red fox Parks and Wildlife Press, Austin. DAVIS, W. B., AND D. J. SCHMIDLY. 1994. The mam- in the region and was not sure of its identity mals of Texas. Texas Parks and Wildlife Press, (Findley et al., 1975). No additional records Austin. were reported in eastern New Mexico until 2 FINDLEY, J. S., A. H. HARRIS,D.E.WILSON, AND C. red fox specimens were collected in the late JONES. 1975. Mammals of New Mexico. University 1980s near Portales, Roosevelt County (#9707 of New Mexico Press, Albuquerque. and #10533; Natural History Museum, Eastern HALL,E.R.,AND K. R. KELSON. 1959. The mammals New Mexico University, Portales). of North America. Ronald Press Company, New In May 1998, we collected a red fox speci- York. men near Clayton, Union County, in north- HALLORAN, A. 1946. The carnivores of the San An- eastern New Mexico (Museum of Southwest- dres Mountains, New Mexico. Journal of Mam- ern Biology, University of New Mexico, Albu- malogy 27:154–161. querque; specimen number not yet assigned). PACKARD, R. L., AND J. H. BOWERS. 1970. Distribution- This record is approximately 200 km east of al notes on some foxes from western Texas and previous records from the mountains of northeastern New Mexico. Southwestern Naturalist 14: ern New Mexico (Hall and Kelson, 1959; Fin- 450–451. SAMUEL,D.E.,AND B. B. NELSON. 1982. Foxes. In: J. dley et al., 1975; Samuel and Nelson, 1982). A. Chapman and G. A. Feldhamer, editors. Wild This specimen also is 250 km north of red fox- mammals of North America: biology, manage- es collected from the plains of east-central New ment, and economics. Johns Hopkins University Mexico (Roosevelt County). The specimen we Press, Baltimore, Maryland. Pp. 475–490. collected might represent a westward expan- STRECKER, J. K. 1926. A check-list of the mammals of sion from Texas, an eastward expansion from Texas. Baylor Bulletin (Baylor University) 29:1– existing populations in the mountains of 48. northern New Mexico, or a northward expan- TAYLOR,W.P.,AND W. B. DAVIS. 1947. The mammals sion from red foxes in east-central New Mexico of Texas. Bulletin of Texas Game, Fish and Oyster (which had an unknown source). Red foxes Commission 27:1–79. have been harvested by trappers in all counties TOWNSEND, C. H. T. 1893. On the life zones of the of eastern New Mexico during the past 20 years Organ Mountains and adjacent region in south- (unpublished harvest data, New Mexico De- ern New Mexico, with notes on the fauna of the partment of Game and Fish, Albuquerque), range. Science 22:313–315. suggesting red foxes now occur throughout VOIGT, D. R. 1987. Red fox. In: M. Novak, J. A. Baker, eastern New Mexico. M. E. Obbard, and B. Mallock, editors. Wild fur- bearer management and conservation in North America. Ontario Ministry of Natural Resources, We thank R. Gilliland, D. Anderson, C. Cosper, Toronto, Canada. Pp. 379–392. and F. Pronger for providing information concern- ing red foxes. This is Texas Tech University, College of Agricultural Sciences and Natural Resources pub- Submitted 15 May 2003. Accepted 2 April 2004. lication T-9-847. Associate Editor was Cody W. Edwards. 102 The Southwestern Naturalist vol. 50, no. 1

OCCURRENCE OF JAGUAR (PANTHERA ONCA) IN SINALOA, MEXICO

CARLOS J. NAVARRO-SERMENT,CARLOS A. LO´ PEZ-GONZA´ LEZ, AND JUAN-PABLO GALLO-REYNOSO*

Centro de Investigacioń en Alimentacioń y Desarrollo, A.C. Unidad Guaymas, Carretera a Varadero Nacional, Km. 6.6. Col. Las Playitas, Guaymas, Sonora 85480, Me´xico (CJNS, JPGR) Escuela de Biologıá, Universidad Autońoma de Quere´taro, Cerro de las Campanas s/n, Quere´taro, 76010, Me´xico (CALG) *Correspondent: [email protected]

ABSTRACT Little is known about the distribution of jaguar (Panthera onca) in Sinaloa. We provide current information about the distribution of this species in this state. Different areas of the state were visited, and people were interviewed from May 2000 to May 2002. We obtained 57 records, of which 41 are new; records were most abundant in the southern half of the state. The extensive areas still covered by tropical deciduous forest in good condition are important for the conservation of the jaguar. The density of prey species for jaguar seems to be high according to local residents. Free-ranging cattle also are distributed in all areas of the state, and predation on cattle is common. Ranchers regularly kill jaguars using poisons (strychnine), dogs, steel traps, and shooting. The abundance of records suggests that the jaguar still exists in Sinaloa, especially in the mountainous areas, but there is an urgent need to conduct additional studies to determine its actual status.

RESUMEN Se sabe poco de la distribucioń del jaguar (Panthera onca) en Sinaloa. Proveemos informacioń actual de la distribucioń de esta especie en este estado. Se visitaron diferentes a´reas del estado y se entrevisto´ a los lugarenõs de mayo del 2000 al mayo del 2002. Se presentan 57 registros de los cuales 41 son nuevos registros; los registros fueron ma´s abundantes en la mitad sur del estado. Las extensas a´reas todavıá cubiertas por bosque tropical caducifolio en buena condicioń son importantes para la conservacioń del jaguar. La densidad de las especies que con- forman las presas del jaguar son aparentemente altas de acuerdo con las personas entrevistadas. El ganado vacuno tambień se encuentra distribuido por todo el estado, y la depredacioń del ganado parece ser comuń. Los rancheros regularmente matan a los jaguares usando venenos (estricnina), perros, trampas de acero y disparańdoles. La abundancia de registros sugiere que el jaguar auń existe en Sinaloa, especialmente en la zona serrana, pero hay una necesidad urgente de conducir ma´s estudios para determinar su estado actual.

The largest felid in the New World, the jag- survey [of Mexican wildlife] were along the uar (Panthera onca) historically ranged from heavily forested flatlands and foothills of south- the southwestern United States to northern Ar- ern Sinaloa, the swamps of coastal Nayarit, the gentina; however, habitat destruction and remaining uncut forests along the Gulf coast hunting have reduced its former range (Sey- as far east as central Campeche, and the great mour, 1989) to less than 37% of its extent in rain forests of northern Chiapas.’’ It is a wide- ca. 1900 (Sanderson et al., 2002). In north- spread belief that the range of the jaguar in western Mexico, it still occurs in the state of Mexico has been reduced so much that it is Sonora, where at least 3 sub-populations have only possible to find important populations in been identified in recent years (Lo´pez-Gonza´- the southeastern states (e.g., Te´llez-Giroń and lez and Brown, 2002). Just south of Sonora and Lo´pez-Forment, 1995; Aranda, 1996). Howev- 500 km south of the USA-Mexico border, the er, that belief possibly reflects the lack of re- state of Sinaloa constitutes the link between search and available information rather than populations in Sonora and more southerly the lack of jaguars, as was discussed during the populations near the Mexican Pacific coast. Le- last attempt to gather information on current opold (1959:529) reported that ‘‘the highest distribution in 1999 during the ‘‘Jaguar in the densities of jaguars noted in the course of this New Millennium’’ workshop (Sanderson et al., March 2005 Notes 103

We obtained 57 records of individual jaguars, of which 41 are first reported here (Ap- pendix 1). Jaguar records were abundant in the southern half of the state. The local people considered the species a common and regular member of the local fauna throughout the mountainous sierra. The jaguar seems to be most common in the San Ignacio and Concor- dia municipalities; however, in those 2 municipalities the search effort was more extensive. More research is needed in northern Sinaloa. During the course of this survey, only 1 (possibly 2) records were found from higher elevations in oak-pine (Pinus-Quercus) forest and 1 record from riparian vegetation. Most occurrences were from the tropical deciduous forest that originally occupied most of the lowlands FIG.1 Approximate locations of the recent records of jaguar (Panthera onca) obtained by us (black in the state and still covers much of the sierra. dots) and cities where interviews were made (open Tropical deciduous forest currently covers circles). 40.09% of the area in the state, whereas oak and pine forests cover 14.71% of Sinaloa (INE- GI, 2000). Today, most of the coastal plain 2002). It was shown at that workshop that in a (34.72% of Sinaloa) is being transformed for large area of western Mexico, of which Sinaloa agriculture, aquaculture, or human settlement represented a considerable portion, the status (INEGI, 2000), and few adequate habitat of the jaguar remained largely unknown (San- patches remain for jaguars there. However, 2 derson et al., 2002). The tropical deciduous recent records occurred close to the coast, ap- forests that cover much of the sierras of Sina- proximately 80 km north of the city of Maza- loa are part of the 18% of the historic range tlań; one was within the Mesa de Cacaxtla Nat- of the jaguar on which its status remains un- ural Protected Area. Although much impacted known, and where the development of surveys by aquaculture operations or human settle- is considered a priority (Sanderson et al., ments, mangrove swamps still represent 7.36% 2002). To better understand the current situ- of the area in the state (INEGI, 2000). ation of the jaguar in Sinaloa, we undertook a Predation on livestock by jaguars seems to be series of field trips, reviewed literature ac- a common event. Cattle traditionally are al- counts of the species, and summarized pub- lowed to roam over vast expanses of land. Lo- lished information (Appendix 1). cal ranchers kill jaguars regularly using differ- We conducted 60 interviews opportunistical- ent methods, including poison (strychnine), ly from the summer of 2000 to May 2002. In- dogs, and steel traps, or by shooting whenever terviewees included local ranchers, local live- they have a chance encounter. Steel traps for stock association officials, hunting club mem- large cats can be purchased readily (prices cur- bers, and employees of local tanneries from rently range from 400 to 600 pesos, about those areas where jaguars were known or sus- US$40 to US$60) at many local hardware pected to exist, especially in mountainous stores, despite the fact that jaguars and moun- country. We tried to obtain physical records tain lions (Felis concolor) have been legally pro- (photographs, skins, or bones) of hunted jag- tected in Mexico since 1986 and are included uars, as well as basic information regarding kill in the ‘‘Norma Oficial Mexicana’’ (NOM-059- sites (such as location, date, and type of vege- ECOL-1994) (SEMARNAP, 1994), the federal tation). We visited 8 towns and cities (El Fuer- list of protected animals and plants. In some te, Badiraguato, San Ignacio, Concordia, Chu- areas, landowners pay bounties of up to 5,000 paderos, Copala, Mazatlań, and Teacapań; Fig. pesos (about US$500) for killing a trouble- 1) and summarized previously published rec- some jaguar in cattle areas or, in some places, ords from Sinaloa. for any jaguar. To determine whether a jaguar 104 The Southwestern Naturalist vol. 50, no. 1 or a mountain lion killed cattle, ranchers observe where the predator started to eat the cow. Mountain lions kill cattle by biting the throat, producing asphyxia; they begin to eat the hindquarters and cover the carcass with branches and dirt. Jaguars kill by biting the cervical vertebrae or the skull. Jaguars drag their prey by the muzzle, which causes the tongue to protrude. Jaguars begin eating the tongue, the muscles of the chest, or the muscles of the face, and do not cover the carcass. Both felines remove the stomach and intes- tines. In 2 interviews, ranchers saw a jaguar eating a dead cow. These methods to identify which feline killed cattle also are used in So- nora (Brown and Lo´pez-Gonza´lez, 2001). Whenever a ‘‘tigre’’ (jaguar) is killed, it is FIG.2 Jaguar (Panthera onca) hunted near Co- skinned and the hide normally is sold, usually pala, in the municipality of Concordia, in the early to people from larger cities, such as Mazatlań, 1990s. This specimen might be 1 of the 2 ‘‘tigre pin- Concordia, Culiacań, and Escuinapa, or to ta menuda’’ reported north of South America. tourists. Ranchers usually sell the skins for 1,500 to 3,000 pesos (about US$150 to 300). No evidence of current sport hunting of jaguar (Brown and Lo´pez-Gonza´lez, 2001). This uars was found during this survey, but some specimen and the 1 described by Alessio-Ro- might still occur. bles (2002) from Campeche are the only 2 Extensive areas of tropical deciduous forest such specimens reported north of South Amer- remain along the Sinaloan sierras. Densities of ica. jaguar prey, such as armadillo (Dasypus novem- The abundance of recent records suggests cinctus), coatimundi (Nasua narica), collared that a jaguar population still exists in Sinaloa, peccary (Tayassu tajacu), and white-tailed deer especially throughout the sierra, and more sur- (Odocoileus virginianus), seemed to be high, ac- veys urgently are needed to better understand cording to most people, and free-ranging do- its current status. mestic animals were widespread. A new, undescribed population of European wild boar or We acknowledge the help of S. Vizcarra, S. Vizcar- feral hog (Sus scrofa) might be established in ra, Jr., A. van der Heiden, H. Plascencia, A. van der the Concordia municipality, as suggested by a Heiden, D. Castro, J. Calderoń, E. Castro, A. Ruı´z- set of tusks from the area. European wild boars Luna, and C. Puente, all of whom provided valuable information and assistance in the field. V. Gehr- were introduced into the buffer zone of La mann created the map. C. J. Navarro took the pho- Michilıá Biosphere Reserve, Durango, in 1989 tograph of the ‘‘tigre pinta menuda’’ (Fig. 2). Part (Weber, 1995), about 160 km from Concordia. of this study was supported by grants from the Hogs might provide jaguars another prey item, Northern Jaguar Project and the Minority Interna- because jaguars frequently prey on several pig tional Research Training University of California species, both wild and domestic (Leopold, Santa Cruz Centro de Investigacioń en Alimenta- 1959; Seymour, 1989; Brown and Lo´pez-Gon- cioń y Desarrollo, Unidad Guaymas. The manuscript za´lez, 2000). The widespread practice of drug improved thanks to the help of D. Valenzuela. harvest and traffic during recent years might have benefited jaguars, because fewer sport LITERATURE CITED hunters and foreigners visit the area. ALESSIO-ROBLES, M. 2002. Un jaguar en el Silencio. A specimen (Fig. 2) hunted near Copala, in Estafeta. Impresos Naucalpan. San Andre´s Atoto the municipality of Concordia, in the early Numero 12. Colonia San Andre´s Atoto. Me´xico 1990s, has unusually small, broken rosettes D. F., 53500, Me´xico. without interior spots. This variation is known ARANDA, M. 1996. Distribucioń y abundancia del jag- as ‘‘tigre pinta menuda,’’ or small-spotted jaguar, Panthera onca (Carnivora: Felidae) en el es- March 2005 Notes 105

tado de Chiapas, Me´xico. Acta Zoolo´gica Mexi- SEYMOUR, K. L. 1989. Panthera onca. Mammalian Spe- cana 68:45–52. cies 340:1–9. ARMSTRONG, D. M., J. K. JONES,JR., AND E. C. BIRNEY. TE´LLEZ-GIRO´ N, G., AND W. LO´ PEZ-FORMENT. 1995. 1972. Mammals from the Mexican state of Sina- Panthera onca veracrucis (Carnivora: Felidae) en loa. III. Carnivora and Artiodactyla. Journal of Quere´taro, Me´xico. Revista Mexicana de Masto- Mammalogy 53:48–61. zoologıá 1:73–75. BROWN,D.E.,AND C. A. LO´ PEZ-GONZA´ LEZ. 2000. WEBER, M. 1995. La introduccioń del jabalı´ europeo Notes on the occurrences of jaguars in Arizona en la Reserva de la Biosfera La Michilıá, Duran- and New Mexico. Southwestern Naturalist 45: go: implicaciones ecolo´gicas y epidemiolo´gicas. 537–546. Revista Mexicana de Mastozoologıá 1:69–73. BROWN,D.E.,AND C. A. LO´ PEZ-GONZA´ LEZ. 2001. Bor- derland jaguars. University of Utah Press, Salt Submitted 7 March 2003. Accepted 5 July 2004. Lake City. Associate Editor was Cheri A. Jones. INSTITUTO NACIONAL DE ESTAD´ıSTICA GEOGRAF´ıAEIN- FORMA´ TICA (INEGI). 2000. Anuario estadı´stico del APPENDIX 1 Records of jaguar from Sinaloa, Mex- estado de Sinaloa, Me´xico. www.inegi.gob.mx ico, in reverse chronological order (1857 through (Accessed: July 2000). 2002). We have included the date of the kill, loca- LEOPOLD, A. S. 1959. Wildlife of Mexico. University tion (geographic coordinates are approximate), bi- of California Press, Berkeley. otic community, sex (if known), documentation, and LO´ PEZ-GONZA´ LEZ,C.A.,AND D. E. BROWN. 2002. Dis- citation, if the record was previously published. tribucioń y estado de conservacioń actuales del JULY 2002, Concordia municipality, tropical decid- jaguar en el noroeste de Me´xico. In: R. A. Me- uous forest, female and cub (sex unknown), hunted dellıń, C. Equihua, C. L. B. Chetkiewicz, P. G. by local ranchers (A. van der Heiden, pers. comm., Crawshaw, A. Rabinowitz, K. H. Redford, J. G. 2002). MAY 2002, between Rancho Coyote and Ma- Robinson, E. Sanderson, and A. Taber, compilers. gistral, Concordia municipality (23Њ21Ј49ЉN, Њ Ј Љ El jaguar en el nuevo milenio. Fondo de Cultura 106 59 45 W), tropical deciduous forest, sex un- Econo´mica Universidad Nacional Autońoma de known, dead livestock seen by local ranchers and Me´xico Wildlife Conservation Society. Me´xico, positively identified as jaguar kill. APRIL 2002, ‘‘Los D. F. Pp. 379–391. Llanitos,’’ ca. 2 km from El Ma´rmol, Mazatlań mu- Њ Ј Љ Њ Ј Љ LO´ PEZ-GONZA´ LEZ, C. A., D. E. BROWN, AND G. LOR- nicipality (23 32 19 N, 106 35 58 W), tropical decid- ENZANA. 2000. El jaguar en Sonora, ¿desapare- uous forest, sex unknown, hunted by local people ciendo o solamente desconocido? Especies. Re- (A. Ruı´z-Luna, pers. comm., May 2002). APRIL 2002, Њ Ј Љ vista sobre Conservacioń y Biodiversidad. 9(3): ca. San Juan, San Ignacio municipality (23 56 43 N, Њ Ј Љ 19–23. 106 20 15 W), tropical deciduous forest, sex unknown, dead livestock seen by local ranchers and MCCURDY, R. 1981. Life of the greatest guide; hound stories and others of Dale Lee. Blue River Graph- positively identified as jaguar kill. MARCH 2002, ‘‘Dos Arroyos,’’ Badiraguato municipality (25Њ16Ј32ЉN, ics, Phoenix, Arizona. 107Њ24Ј47ЉW), tropical deciduous forest, sex un- SANDERSON, E. W., C. L. B. CHETKIEWICZ,R.A.MED- known, hunted by local rancher. MARCH 2002, Ma- ELL´ıN,A.RABINOWITZ,K.H.REDFORD,J.G.ROB- gistral, Concordia municipality (23Њ21Ј41ЉN, INSON, AND A. B. TABER. 2002. Prioridades geo- 106Њ59Ј16ЉW), tropical deciduous forest, young in- gra´ficas para la conservacioń del jaguar. In: R. A. dividual stoned to death by local ranchers, sex un- Medellıń, C. Equihua, C. L. B. Chetkiewicz, P. G. known. MARCH 2002, El Habal, ca. Copala, Concor- Crawshaw, A. Rabinowitz, K. H. Redford, J. G. dia municipality (23Њ23Ј58ЉN, 105Њ56Ј07ЉW), tropical Robinson, E. Sanderson, and A. Taber, compilers. deciduous forest, sex unknown, hunted by local El jaguar en el nuevo milenio. Fondo de Cultura rancher. Econo´mica Universidad Nacional Autońoma de 1998–2002, Concordia municipality, tropical de- Me´xico Wildlife Conservation Society. Me´xico, ciduous forest, skins of 4 individuals seen by us at D. F. Pp. 601–628. Mazatlań, sexes unknown. SECRETAR´ıADELMEDIO AMBIENTE RECURSOS NATURA- SUMMER 2001, San Isidro Ranch, Concordia munic- LES Y PESCA (SEMARNAP). 1994. Norma Oficial ipality (23Њ21Ј51ЉN, 105Њ59Ј59ЉW), tropical deciduous Mexicana NOM-059-ECOL-1994, que determina forest, male (?), footprints seen by local rancher. DE- las especies y subespecies de Flora y Fauna silves- CEMBER 2001, ‘‘La Laguna,’’ ca. Durango Ranch, Con- tres terrestres y acua´ticas en peligro de extincioń, cordia municipality (23Њ20Ј57ЉN, 105Њ54Ј14ЉW), trop- amenazadas, raras y las sujetas a proteccioń es- ical deciduous forest, female (?), poisoned by local pecial y que establece especificaciones para su ranchers. NOVEMBER 2001, Rancho Coyote, Concordia proteccioń. Diario Oficial de la Federacioń. municipality (23Њ21Ј54ЉN, 105Њ59Ј58ЉW), tropical de- www.semarnat.gob.mx (Accessed: July 2000). ciduous forest, sex unknown, heard by local ranchers. 106 The Southwestern Naturalist vol. 50, no. 1

Killed a goat at the ranch that night. FALL 2001, 1980s, ca. Los Angeles ranch, Escuinapa municipality ‘‘southern Sinaloa,’’ male and female, captive cubs in (22Њ39Ј56ЉN, 105Њ48Ј14ЉW), tropical deciduous forest- private home (A. van der Heiden, pers. comm., mangrove swamps, sexes unknown, 3 individuals 2002). 2001, Coacoyol, San Ignacio municipality hunted by local people, skins seen by us. 1970s. Playa (23Њ57Ј42ЉN, 106Њ29Ј34ЉW), tropical deciduous forest, Brujas, ca. 23 km NW of Mazatlań, Mazatlań munici- sex unknown, shot by local rancher as livestock-killer. pality, tropical deciduous forest, 1 individual of un- 2001, Mesa de Cacaxtla, San Ignacio municipality known sex seen walking on the beach. 1962, Sierra (23Њ39Ј38ЉN, 106Њ42Ј59ЉW), tropical deciduous forest, Madre NE of Matantań near the Durango–Sinaloa sex unknown, hunted by local ranchers. 2001, Sahui- boundary, oak-pine forest (?), sex unknown, skull of tapa, ca. San Ignacio, San Ignacio municipality individual shot purchased in Rosario (Armstrong et (24Њ00Ј00ЉN, 106Њ29Ј47ЉW), tropical deciduous forest, al., 1972). 1960’s, Sierra de Surutato, Badiraguato mu- 2 individuals shot by local ranchers, sexes unknown. nicipality (25Њ48Ј37ЉN, 107Њ33Ј28ЉW), oak-pine forest, SUMMER 2000, ‘‘Cerro El Pirame,’’ Concordia mu- male, dead individual seen by the president of the nicipality (23Њ21Ј50ЉN, 105Њ59Ј56ЉW), tropical decid- Badiraguato local livestock association. 1940s, NE of uous forest, male shot by local rancher as livestock- San Ignacio, San Ignacio municipality, tropical decid- killer. Skull collected and deposited at Centro de In- uous forest, male shot by hunting party (McCurdy, vestigacioń en Alimentacioń y Desarrollo, Guaymas 1981). collection (CIAD-070502-1). SUMMER 2000, El Tule, PRIOR TO 1959, near Santiago, San Ignacio munic- San Ignacio municipality (23Њ49Ј36ЉN, 106Њ25Ј32ЉW), ipality, tropical deciduous forest, male (?), shot by tropical deciduous forest, 2 individuals separately shot sport hunters (Leopold, 1959). PRIOR TO 1959, vicin- by local ranchers while feeding on dead cows, 1 male ity of San Ignacio, San Ignacio municipality, tropical and an individual of unknown sex. LATE 2000, ca. deciduous forest, sex unknown (Leopold, 1959). Coyotitań, San Ignacio municipality (23Њ47Ј53ЉN, 1937 or 1938, Mesa de Pla´tanos, San Ignacio munic- 106Њ34Ј58ЉW), tropical deciduous forest, sex un- ipality, tropical deciduous forest, male and female, 2 known, killed by car while crossing the road at night, individuals shot by hunting party (McCurdy, 1981). skin in possession of the former mayor of San Ignacio. SUMMER 1935 or 1936, Los Frailes, San Ignacio mu- 2000, Coacoyol, San Ignacio municipality nicipality, tropical deciduous forest, male shot by (23Њ57Ј42ЉN, 106Њ29Ј34ЉW), tropical deciduous forest, hunting party (McCurdy, 1981). 1935 or 1936, San sex unknown, shot by local rancher. Juan, 10 miles S of San Ignacio, San Ignacio munic- 1999, ‘‘Cerro El Elefante,’’ Concordia municipal- ipality, tropical deciduous forest, 2 females shot by ity (23Њ19Ј00ЉN, 105Њ59Ј58ЉW), tropical deciduous hunting party (McCurdy, 1981). PRIOR TO 1933, Es- forest, sexes unknown, 2 cubs captured alive by local cuinapa, Escuinapa municipality, tropical deciduous rancher. 1992, ca. Mesillas, Concordia municipality forest, sex unknown (Nelson and Goldman, 1933, as (23Њ17Ј44ЉN, 106Њ04Ј13ЉW), tropical deciduous for- reported by Armstrong et al. 1972). est, male, attracted by commercial predator-call tape 1905–1906, Escuinapa, Escuinapa municipality, and shot as livestock killer, skin seen by us in Con- tropical deciduous forest, 1 male, 3 females, and 1 cordia. EARLY 1990s, ca. Copala, Concordia munici- immature female ( J. H. Batty, as reported by Brown pality (23Њ23 58ЉN, 105Њ56Ј00ЉW), tropical deciduous and Lo´pez-Gonza´lez, 2001). 1901, Escuinapa, Escui- forest, sex unknown, skin seen by us in Copala. 1990, napa municipality, tropical deciduous forest, male Baroteń, ca. 3.5 Km S of El Fuerte, El Fuerte munic- ( J. H. Batty, as reported by Brown and Lo´pez-Gon- ipality (26Њ23Ј45ЉN, 108Њ35Ј55ЉW), riparian vegeta- za´lez, 2001). Prior to 1901, Cacalotań, Rosario mu- tion along El Fuerte river, sex unknown, hunted by nicipality, tropical deciduous forest, sex unknown local people and seen by local veterinarian. (Mearns, 1901, as reported by Armstrong et al. LATE 1980s, EARLY 1990s. El Tule, ca. Escuinapa, Es- 1972). cuinapa municipality, tropical deciduous forest, sexes 1857, Mazatlań, tropical deciduous forest, sex un- unknown, 3 individuals seen on different occasions known, type specimen of Panthera onca hernandesii quietly following people along forest trails at night. (Gray, 1858, as reported by Armstrong et al. 1972).