'^j-.; ;. J i -! '''^'■'<' ■;'■';. i , '■ ;; ■- ■■*'! .!' I ■ ., ' ■ 1 -■■■ . t ',■ ■ : . ■ . n i-Mi .■' • '',■;' -I U: V- *1 .''''..■■■■.. '•■•i'lV'.' ■ -r;-'*.- '.- ; •>rK. ;!'■•> "Vj; ■ '■. i-e\'. >•• A 1-U ■5i 1*. i' 1 - - • ;• '' ' '' I ' Ok HARVARD UNIVERSITY mm Ernst Mayr Library of the Museum of Comparative Zoo logy TULANE STUDIES IN ZOOLOGY AND BOTANY Volume 30, Number 1 January 20, 1995 WOODY VEGETATION OF TROPICAL LOWLAND DECIDUOUS FORESTS AND MAYAN RUINS IN THE NORTH-CENTRAL YUCATAN PENINSULA, MEXICO DAVID A. WHITE Department of Biological Sciences Loyola University, Neiu Orleans, l^\ 701 18 AND STEVTN P. DARWIN Department of Ecology, Evolution, and Organismal Biology Tulane University, Neiu Orleans, lA 70118 POPUU\TION STRUCTURE AND CLONAL GROWTH IN BROMELIA P/NG^/iV L.(BROMELIACEAE) IN DRY FORESTS OF COASTAL VERACRUZ, MEXICO JOSE G. GARCiA-FRANCO AND VICTOR RICO-GRAY Instituto deEcologia, A. C. Apdo. Postal 63, Xalapa, Ver., 91000 Mexico THE CYCLOPOID COPEPOD (CRUSTACEA) FAUNA OF NON-PLANKTONIC CONTINENTAL HABITATS IN LOUISIANA AND MISSISSIPPI JANET W. REID Department of Invertebrate Zoology, MRC-163 National Museum of Natural History, Smithsonian Institution Washington, DC 20560 AND GERALD G. MARTEN Neu) Orleans Mosquito Control Board 6601 Lakeshore Drive, Neio Orleans, LA 70126 and Defjartment of Tropical Medicine Tulane University, Neiv Orleans, lA 70112 Tulane University New Orleans TULANE STUDIES IN ZOOLOGY AND BOTANY ISSN 0082-6782 Department of Ecology, Evolution, and Organismal Biology Tulane University, New Orleans, Louisiana 70118 Steven P. Darwin, Editor INFORMATION FOR AUTHORS Manuscripts dealing with all aspects of ecology, evolution, and systematic biology are encouraged. All manuscripts are reviewed. Page charges are assessed at the rate of $45 per page. Some financial assistance from the George Meade Memorial Fund is available if authors are unable to meet these charges. Style of manuscript should follow guidelines appearing in the Council of Biology Editors Style Manual. Manuscript (with abstract) must be typed on good quality 8.5 x 1 1 inch paper, dou- ble-spaced throughout (including Literature Cited, figure captions, footnotes, and tables). All pages must have 1 inch margins on all sides. Dot-matrix printing is not acceptable; do not justify right-hand margins. Do not hyphenate words at ends of lines. Last name(s) of author (s) should appear on each page with page number; this includes tables and figure captions. Lettering used in figures must be easily read after reduction to our page size. Pho- tographs (halftones) should be made for no or little reduction. The original and three copies of the manuscript, along with figures, should be sub- mitted. Prospective authors should request a copy of the detailed guidelines for manu- script preparation (including footnotes, typefaces, literature and specimen cita- tion, and dimensions of figures) from the Editor at the above address. EXCHANGES AND STANDING ORDERS Tulane Studies in Zoology and Botany is available for exchange with institutions publishing comparable series. Otherwise, standing orders are accepted, with issties priced individually (foreign postage extra). A price list for back issues is available on request. Remittances should be made payable to Tulane University. WOODY \TGETATION OF TROPICAL LOWLAND DECIDUOUS FORESTS AND MAYAN RUINS IN THE NORTH-CENTRAL YUCATAN PENINSULA, MEXICO Da\ ID A. White DepartmenI of Bioloirical Sciences Loyola Universily, Nnu Orleans, LA lOlIH AND Steven P. Darwin Department nfEroloir\, Kvolnt'mn, and Orgauisnial Biolo/ry Tulaiie Lhiiiiersiy, AWc Orleans, LA 71)1 IS Abstract Quadrats were sampled at five dry-forest and two Mayan ruin sites in ilie north-central Yucatan Peninsula of Mexico. A total of 149 woody species were found, and data for the 81 most important species were analyzed (Importance Values >L0; DBH >3.() cm). Their distribution correlates with dry season precip- itation patterns (polar ordination). Average basal area is 20.2 sq. m per ha for forest sites and 28.2 sq. m per ha for ruin sites. Average density is 3344 stems per ha in forests and 1432 stems per ha on ruins. The Shannon-Weaver Diversity Index varies from 1.76 to 2.25 for forests and 0.94 to 1.25 for the two ruins. Important forest species are Bursera simaruba, Thouinia paucidentata, Piscidia pis- cipula, Gymnopodium florihuudum, and Caesalpinia gaunieri. Legimiinoiis species are more common in the drier, western forests than in the eastern forests, but are rare on the ruins. The two Mayan ruin sites have a vegetation that is distinct from that of the central peninsula and is dominated by Brosimum alicastrum. Introduction According to the Holdridge (1967) system of vegetation classification, the Yucatan Peninsula of Mexico is dominated by dry tropical forests (Holdridge et al., 1971). Until the middle 1900s, most botanical work on the peninsula was flo- ristic and stuAey-type, conducted bv Lundell (1934, 1937, 1938), Millspaugh (1895, 1896, 1898), Standley (1930), and others with ties to the Carnegie Insti- tute in Washington, D.C. Miranda (1958) divided the Yucatan forests into five types based upon woody species presence, canopy height, and deciduousness. The boundaries of his forest types correspond precisely with isohyets of increas- ing precipitation across the peninsula from northwest to southeast (Garcfa, 1964; Isphording, 1975). The first quantitative ecological stud\ of a forest on the peninsula was by Thien et al. (1982) near and on the ruins of D/ibik haltun in the very northwest of the peninsula. Others have quantified the wooch \egeta- tion of old-growth forest in the north-central peninsula (Rico-Gray et al., 1988) and have described the relationships of the local [x-ople, descciuhuiis of the Maya, to the present vegetation of tlie region (Rico-Ciray et al., 1990; Rico-Ciray and Garcia-Franco, 1991, 1992). Whigham et al. (1990, 1991) studied a north- Tulane Studies in Zoolog) and Botany .^0: l-2ri. 1993 2 Tulane Studies in Zoology and Botany [Vol. 30 eastern forest, quantifying woody plant growth, litter production, and impacts of hurricanes. No quantitative vegetation study ofthe central peninsular forests has been published, although this is a large region under intense human pressure (Brown and Lugo, 1990; Sabogal, 1992). Most scientific studies conducted in the region have focused on the archaeol- ogy of Mayan sites, and have spawned some ancillary descriptions of vegetation associated with ruins (Lundell, 1937; Folan et al., 1979; Orejuela, 1980; Lambert and Arnason, 1982; Peters, 1983). A paradox exists as to why the vegetation of ruin sites is often different from neighboring forest vegetation. This study describes and quantifies extant old natural forest vegetation, compares it to the vegetation on two collapsed ruins, and then compares it in general to dry tropi- cal vegetation of the world. The variation in the peninsula's forest vegetation is then related to known climatological patterns in Yucatan. Materials and Methods STUDY REGION. The northern Yucatan Peninsula of Mexico (states of Campeche, Yucatan, Quintana Roo) consists of limestone bedrock of Tertiary age, flat and without surface runoff (streams are absent). The average annual temperature is 27°C, with a July average of 29°C and a January average of 24°C. The lowest recorded temperature is 6°C. Total annual precipitation across the peninsula averages 1100 mm, with a dry season from October to May (average total = 300 mm) and a wet season from Jime to September (average total - 800-900 mm) due to adiabatic heating. The northern peninsula has no signifi- cant topographic features to create unusual localized rainfall or temperature conditions. There is a noticeable west to east, dry to wet gradient over the penin- sula (Figure 1), the prevailing winds being from the Caribbean Sea in the east. For several thousand years, an extensive Mayan civilization with enormous cities occupied the region (Morley and Brainerd, 1983). The current inhabitants of the peninsula are Mayan subsistence farmers who practice a type of slash-and-burn agriculture principally in the northern and central portion of the State of Yucatan. Near and almost parallel to the southwestern border ofthe state of Yucatan is a small but topographically significant anticlinal hill range (the Puuc, 100 m maximiun elevation; Isphording, 1975), which emerges from the flat landscape of the north. This range begins about 50 km from the western coastline and its fault-line is continuous for 160 km before disappearing near the southeastern border ofthe state of Yucatan and the state of Quintana Roo. Because it is easier for the inhabitants to cultivate the surrounding flatter land, some of the least disturbed vegetation in the north-central part of the peninsula are found on the slopes of these hills. By local estimates, forests on some slopes have grown for 50 to more than 100 years without disturbance from cultivation. This is likely a sig- nificant amotmt of time for recovery from distiubance, since it has been esti- mated that dry tropical forests generally require only 150 years for total recovery (Opler et al., 1977), and species composition may not change over that period (Miranda, 1958). STUDY SITES. Three forest study sites 40 to 75 years old were established along the Puuc range, one at each end and one near the middle. The Calcehtok site was the western site (20°33'N, 89°54'W) near the village of Calcehtok (Figure 1). The eastern site (20°02'N, 89°04'W) was just south ofthe town of Tzucacab and 1994] Woody Vegetation of Yucatan Figure 1. Map of the northern Yucatan Peninsula, Mexico, showing state boiuidaries, principal cities, archaeological sites (solid circles), location of the seven study sites (stars), three studv sites cited in the literature (open circles), and isohyets (mm) adapted from Garcia (1965). V..\L = Calcehtok forest; TIC = Ticul forest; TZU = Tzucacab forest; SAN = San Mateo forest; OTH = Othon Blanco forest; L IT = I'itiz- ina ruin; MIR = Mirador ruin. given that name. The third site was located about equidistant between the other two (20°22'N, 89°34'W), southwest of Ticul. In addidon to those sites within the Piuic, two others were studied because of their unusually old age. The Othon Blanco site (19°37'N, 89°00'W), is of older growth (more than 150 years), with a closed forest canopy and by appearance representative of the extensive forest in the east-central part of the peninsida. It is located near the village of Othon Blanco in western Quintana Roo, about the same distance (50 km) from the Tzucacab site as the distance separating the other forest sites. The San Mateo site (20°24'N, 88°49'W) was chosen for study because Rico-Gray et al. (1988) found that forests in this region show almost no evidence of human occupation or distmbance. These last two sites are on limestone substrate and a consider- able distance from the Piuic area. Two other sites were chosen becatise they were the least disturbed closed can- opy forests found growing on the rubble of collapsed Mayan ruins. At each site the clearly visible rubble occupies an area less than 0.20 sq. km and forms small hills up to 20 m high in the otherwise flat landscape. The Uiti/.ina ruin site (19°52'N, 88°58'W) is located 15 km southeast of the town of Tzucacab. The for- 4 Tulane Studies in Zoology and Botany [Vol. 30 ests growing on the ruin are striking in their lushness, contrasting with the sur- rounding cattle ranches of savanna-like fields. The Mirador ruin site is approximately 5 km south from the Othon Blanco site and a considerable dis- tance from any village or cultivated land. The structure of the closed-canopy for- est around it is similar to that of the Othon Blanco site. All seven sites were identified by local people as the oldest forests within their region. At all but the San Mateo and Mirador sites, culling of trees occurs for firewood and house construction with increasing frequency from east to west. \TGETATION COMPOSITION AND STRUCTURE. At each site, the vegetadon was sampled using sets of three 10 m x 20 m plots placed along 50 m transects. The total number of transects, and thus of plots, varied at each site from three to nine depending upon species-area curve calculations. The transects were spaced at least 25 m apart on flat land and were oriented east-west; on land with relief, the transects were oriented along the cardinal direction most parallel to the slope. The long axis of each plot was perpendicular to the transect. Within the plots, all woody stems (including coppiced stems)>3.0 cm DBH were meastired and identified with the Mayan vernacular name supplied by one or several of six remarkably knowledgeable local field assistants. To be sure of their abilities and consistency in applying names, we repeatedly collected multiple samples of ster- ile specimens identified by the Mayans to compare with vouchers. Very rarely were mistakes made. Voucher collections were made for each species, most of which were eventu- ally found with flowers or fruits. Identification to species was tedious because the flora of the peninsula is not well known. Collections were identified at Tulane University (NO), with questionable and luinamed specimens taken for scrutiny to the Field Museum (F), or the Missouri Botanical Garden (MO). Nomencla- ture follows the floras of Panama (Woodson and Schery, 1943-1980), Guatemala (Standley and Williams, 1946-1977), and Mexico (Standley, 1920-1926, 1930; Gomez-Pompa, 1978-; Cano et al., 1982; Sosa et al., 1985), or recent mono- graphic studies. All vouchers have been deposited in the Tulane University Her- barium, with duplicates sent to other institutions. For each site. Importance Values were calculated using relativized values of density, total basal area, and frequency of occurrence in the plots (Mueller-Dombois and Ellenberg, 1974). The vegetation data were analyzed using the BCORD program on the PC-ORD system (McCune, 1987). This pro- gram performs the Bray-Curtis (polar) ordination. The Euclidean dissimilarity measurement was used with endpoints selected by the variance-regression method (Beals, 1985); all other options were defaulted. The polar ordination technique was employed instead of detrended correspondence analysis and other techniques following Causton (1988); with the small number of sites, all multivariate techniques would likely give insignificantly different results. Species diversity was estimated as Richness (S), Evenness (E), and Diversity (H'). S is defined as the total number of species present, and E as H'/ln S (Pielou, 1969), where H' is X(pilogpi) (Shannon and Weaver, 1949). Results General Vegetational Characteristics A total of 123 plots were sampled, 99 at the five forest sites and 24 at the two ruin sites (Table 1), giving a sample area of 1.98 ha of forest and 0.48 ha on 1994] Woody Vegetation of Yucatan CS ^ u * z o X X Tp ^ ^ ^^ ^, ^ (J5 ry; =2 CT) 4 2; ^ O) cy^ : ■ O cc, !0' -C — ' O C^J CO ^ X 1^ X «3 tT CO CTl Jg 5^ 5vi CTi X Z fs) H - ■ J a 5 ^' 06 § ^ ^ ^'^- O SM CM CO r^ m ^ ^; c^ to ■ ^t t^ gc ^o CM _■ • CM eo Tf »n • CO t^ c^j o — CO cvi — c^i _■ X- _, Tf-CO-^f^^mcO^firo ^ ^ ^ X S S 00 ^ S 15 I 00 — c^j r-CM d"^ X CM lO O^ C-. s s^ ^- g s :::: ^ s ?:' ^ 00^ °^ ;: ^^ CM X CO CT) '- :;•, - =2 „■ 3 O' O — or. to _. X cri 02 ^' ^ Tf in ^ ^ ^^ c^i ^ : o >n i> O O CM cr. S y^ =<^ 9 H S r^ in c^i _; <^ '^ ^ X S-' t^ ^ ,-{; 05 . (O <0 to ^ Xi ^ 1; 4-1 c/5 ?^ ^ 1^ 03 \ c X u T3 C > .1^ 5-g oi = £2 w L 1; fS J-* y. ■<-' z h= 02 p; r ~ ?3 aj ;i -: j= .r .^ C/D D- C/5 c/5 C/D s a c o a 3 bo a- u T3 2 u O in U bo u 3.0 cm DBH were identified in one or more sites. A summary- of important vegetation parameters for the seven sites is pre- sented in Table 1. The mean Richness of the forest sites is 65 species, whereas the ruin sites have an average of 28 species. The average Richness in each forest quadrat varies from 13.8 species at Calcehtok to 21.2 at Tzucacab, with an aver- age of 17 for each plot over the five sites. Both of the ruin sites have fewer spe- cies in each plot (S = 8.6), and little intersite variabilit)' (alpha-diversit)')- The Shannon-Weaver Diversity Index is usuallv much higher for the forest sites (H' >2.00) than for the ruin sites (H' ca. 1.00). The density in the forest sites is much greater (3344 stems per ha) than in the ruin sites (1432 stems per ha). However, stem diameters are on average greater (15.9 cm) in ruin sites than in forest sites (9.3 cm). This translates into an average of 28.2 sq. m per ha total basal area for the two ruin sites and 20.2 sq. m per ha total basal area for the five forest sites. Of the 149 species identified, 81 may be considered ecologicallv common because their Importance \alues are >1.00 (out of a total of 100.00) in at least one of the seven studv sites. The 68 remaining species are ecologically rare and were not included in these data analyses. In Appendix I, we present the fre- quency, density, and basal area data for the 81 common species at the seven sites; these data represent 93% of the total tallied stems. Appendix II is a list of the 68 rare species by site. CO UJ u LU a. co 100 TZU 80 ^ CAL OTH 60 /^^ SAN UO ^^^" MIR 20 /^^^^^ UIT 20 24 28 QUADRAT NO Figure 2. Species-area curves for the seven study sites in the northern \ucatan Peninsula, Mexico. The X-axis is the number of 200 sq. m quadrats sampled. The y-axis is the cumulative number of species measured. CAL = Calcehtok forest; TIC = Ticul forest; TZU = Tzucacab forest; S.-VX = San Mateo forest; OTH = Othon Blanco forest; LIT = Uitizina ruin; MIR = Mirador ruin. 1994] Woody Vegetation of Yucatan The ordination (Figure 3) indicates the distinctiveness of the ruin sites. The spatial position of the forest sites in the two-axis ordination corresponds exactly to a northwest to southeast transect paralleling the northern boimdarv of the state of Campeche (Figine 1), with C'alcehtok the westernmost site and Othon Blanco the easternmost site, and the San Mateo site close to the Tzucacab site. The first two axes explain 78% of the variation in the data. No new relationships are revealed if the ruin sites are eliminated and only the five forest sites ordi- nated; fotn- sites are positioned as endpoints, with the San Mateo site in the cen- ter of the two-axis plot. Forests of Central Yucatan Calcehtok forest— \n the Calcehtok forest, Bursera simaruba and Gsmnopodium Jloribundum are the most frequently found woody plants; they were in HSSt and 25 - CVJ CO X < I o Q. PO-AXIS I Figure 3. Polar ordination of five forest and two Mavan ruin sites in northern '\iicatan Peninsula. Mexico. CAL = Calcehtok forest; TIC = Ticul forest; TZU = T/iicacab forest; S.W = San Mateo forest; OTH = Othon Blanco forest; UIT = Uitizina ruin; MIR = Mirador ruin. 8 Tulane Studies in Zoology and Botany [Vol. 30 75% of the plots, respectively (Appendix I). Gymnopodiuni floribundurn, Machao- nia lindeniana, and Bursera simaruba are most abundant, accounting for 49% of the stems. Caesalpinia gaumeri, Guettarda elliptica, Lysiloma bahamense, and Thouinia paucidentata also have high densities, totalling 20% of the stems. They are also frequent species, found in about 70% of the plots. Together these seven species contribute 69% of the total density, and in some portions of this forest G. floribundum is found in large, nearly monotypic stands. Gymnopodiuni floribundurn is the most abundant species and represents 26% of the density, but it contrib- utes only a third of the total basal area, a reflection of the small average stem diameter. Bursera simaruba is the largest tree, with a basal area of 6.70 sq. m per ha (32% of the total). Those two species, plus C. gaumeri and L. bahamense, con- tribute 13.9 sq. m per ha, or 67% of the basal area. The average DBH at the site is a small 8.7 cm. The Calcehtok site is the least diverse of the forest sites (H' = 1.76), but has a relatively large total number of species (64). The canopy height varies from 8 to 15 m. Bursera simaruba, Lysiloma bahamense, and Vitex gaumeri are usually the tree species of any size in this forest. Ticul forest- In the Ticul forest, 650 stems were measured, and these have the lowest average diameter (6.5 cm) of any site (Table 1). Bursera simaruba and Thouinia paucidentata were found in every plot, while Guettarda elliptica and Cae- salpinia gaumeri were relatively abundant, present in eight and six of the nine plots, respectively (Appendix I). Gymnopodiuni floribundum is clumped in this for- est, found in four plots. These five species are the most abundant at the site, comprising 54% of the total number of stems (Table 1). Each is found at about the same density. As at the Calcehtok site, B. simaruba is, by a factor of three, the largest contributor (3.13 sq. m per ha or 26%) to the total basal area of 11.9 sq. m per ha. Caesalpinia gaumeri, Piscidia piscipula, and Thouinia paucidentata con- tribute 34% to the basal area. Piscidia piscipula is the largest tree at this site with a mean DBH of 12.9 cm. As in forests to the east, the Diversity Index (H' = 2.00) is significantly greater than at the Calcehtok site. The average canopy height is from 8 to 15 m, but broken with taller individual trees oi Lysiloma bahamense and Spondias mombin. Tzucacab forest- The Tzucacab forest has the largest number of recorded spe- cies (93) and the greatest Diversity (H' = 2.25; Table 1). The six most frequent species are Bursera simaruba, Piscidia piscipula, Diospyros cuneata, Croton reflexifo- lius, Lonchocarpus rugosus, and Coccoloba spicata (Appendix I). Bursera simaruba is the most dense species with 574 stems per ha; Croton reflexifolius and Diospyros cuneata are also very abundant, yet they comprise only 29% of total density. Other relatively dense species include Eugenia axillaris, Lonchocarpus rugosus, Lysiloma bahamense, Neea choriophylla, Piscidia piscipula, and Vitex gaumeri, with a combined density of 1204 stems per ha, or 25% of the total. As in the other for- ests, Bursera simaruba is the largest contributor (3.39 sq. m per ha or 19%) to the total basal area. Lysiloma bahamense and Vitex gaumeri contribute 4.51 sq. m per ha, or 26% of the total. The ten species just mentioned make up 65% of the basal area in this forest; the other 84 species found in the Tzucacab forest con- tribute 35%. Average stem diameter at this site is 6.8 cm, slightly larger than in the Ticul forest. Canopy height is 15 to 25 m, broken and very plot-dependent, with no single or group of species consistently tall. San Mateo forest- Trees of the San Mateo forest have a mean stem diameter of 9.6 cm. This site has a high Diversity Index (H' = 2.23) , very similar to that of the Tzucacab forest, and a relatively low total density of 2513 stems per ha (Table 1). 1994] Woody Vegetation of Yucatan 9 Thouinia paucidentata is most frequent (92%; Appendix I), wliilc Bursera simaruha has a frequency of 83%, and Diphysa carthagenensis, Eugeriia buxifolia, and Pisridia piscipula each with a frequency of 75%. Gymnopodium Jloribundum is the most commonly encountered species with 288 stems per ha, but it only accounts for 12% of the total number of stems. The other relatively abundant species, Thouinia paucidentata, Malpighia punicijolia, Diphysa larthagoicusis, and Eugenia buxifolia contribute an additional 30% to the density. The total basal area of 18.4 sq. m per ha is not very different from that of the previously described sites. Burs- era simaruha and Caesalpinia gaumeri are the largest trees and contribiue 24% of total basal area. Thouinia paucidentata, Piscidia piscipula, Gymnopodium poribun- dum, Lysiloma latisiliquum, and Vitex gaumeri -Ava also significant species and aboiu equal in their basal area contributions, a combined 34% of the total. The tallest trees in this forest reach 20 m, with Caesalpinia gaumeri, Bursera simaruha, Talisia olivaeformis, and Vitex gaumeri the usual species with an average diameter of approximately 18 cm. An occasional large crown oi Lysiloma latisiliquum (32.5 cm average DBH) protrudes through the inieven canopy. Othon Blanco forest- Of all the forest sites, the Othon Blanco forest has the largest total basal area (32.5 sq. m per ha; Table 1); Manilkara zapota is the larg- est contributor (13%). Bursera simaruha, Gymnanthes lucida, Caesalpinia gaumeri, and Metopium brownei are the other major contribiuors to basal area (ca. 7% each; Appendix I). Six other species, Drypetes laterifolia, Coccoloba spicata, Gymnop- odium floribundum, Lysiloma latisiliquum, Psidium sartorianum, and Thouinia pauci- dentata are also important in this forest, with a combined basal area of 24%. Gymnanthes lucida, Drypetes laterifolia, and Manilkara zapota are strikingly abun- dant and contribute 45% of the total number of stems (Table 1). Here, the two most widespread species are Manilkara zapota, found in 93% of the plots, and Drypetes laterifolia, found in 89% of the plots. Gym)ianthes lucida, Gymnopodium Jlo- ribundum, and Cosmocalyx spectabilis are also frequently encountered. Bursera simaruha is not so freqtient as in the more western forests, here found in 44% of the plots. The Othon Blanco forest has twice the average stem basal area (1.40 sq. dm per stem, or DBH = 13.4 cm; Table 1) of any of the other forests studied. This is the only forest with a persistent, well-defined canop\' and luiderstory. The canopy height is uniformly 30 to 35 m with Lysiloma latisiliquum (45.6 cm DBH), Bursera simaruha (28.7 cm DBH), Metopium brownei (30.3 cm DBH), and Spondias mombin (31.3 DBH) the largest, most common, and tallest trees. Some individu- als of Guettarda combsii, Luehea speciosa, Piscidia piscipula, Vitex gaumeri, and Z.uela- nia guidonia are at canopy level also. Forests on Ruins The two ruin-associated forests are remarkable in their similar differences from the five other forest sites (Table 1). Both exhibit low Diversity, low Rich- ness, low total density, high basal area coverage, and tall canopies (30 m). They are also similar in species composition, sharing nine species, particularly the dominant Brosimum alicastrum in the overstory, and Croton lundelii in the under- story (Appendix I). Brosimum alicastrum is the most frequent, most dense, and largest tree in the ruin sites; it was found in all plots at both sites, accounting for 32% of the total stems encountered at the Uiuzina site and 43% at the Mirador site. Trees of Brosimum alicastrum comprise 69% of the basal area at the IMti/ina site and 50% at the Mirador, site, with average diameters of 23.9 cm and 26.9 cm. 10 Tulane Studies in Zoology and Botany [Vol. 30 respectively. In both forests, Croton lundelii is the second most abundant species (Uitizina, 19%; Mirador, 17% of total density), but only in the understory. It is frequently encountered, found in 100% of the Uitizina and 75% of the Mirador plots. Both ruin-site forests are tniique in the presence of Urera baccifera in the understory. It is common at the Uitizina site (16% of total density) and small, with an average diameter of 3.5 cm. Bursera simaruba and Thouinia paucidentata are important overstory species at both sites, both species being larger at Uitiz- ina. The sites differ in other overstory species. The Uitizina forest has the over- story species Ehretia tinifolia and Talisia olivaeformis, while the Mirador overstory is much richer and inckides Karwinskia humboldtiana, Chlorophora tinctoria, Spon- dias mombin, Piscidia piscipula, and Ficus cotinifolia, all with average diameters between 38 cm and 22 cm. In addition, there are several unidentified species of large Ficus growing over the rubble of this ruin. Forest Phytosociology In Table 2 the Importance Values (I.V.) for the 81 most common species are listed by forest; to show patterns of distribiuions, the species are separated into eight groups. By sttidying only the ecologically common species (those with I.V. >1.0), intersite phytosociological relationships become clear. The elimination of the 68 rare species shotild not affect the robustness of the data; the combined Importance Values for the rare species averaged 6% across all seven forests (Table 2, bottom), and was never greater than 14%. Of the 81 more important species in the central Yucatan forests, 13 are espe- cially common; three {Bursera simaruba, Thouinia paucidentata, Bunchosia swartzi- ana) are present in all seven forests, and one, Spondias mombin, is found in five forests but is likely present in all (Group I, Table 2). Of the 13 species, Bursera simaruba is the most important tree contributing 11% to the total Importance Value for the five forest sites and 5% of the Importance Value for the two ruin sites. Its Importance Value is highest in the westernmost forests. Eleven species were found only in the forest sites (Group II, Table 2); two of these are particu- larly noteworthy, Gymnopodium floribundum. and Caesalpinia gaumeri having rela- tively high average Importance Values. Ten species are present only in the four western forest sites at Calcehtok, Ticul, Tzucacab, and San Mateo; of them, Guettarda elliptica, Lysiloma bahamense, and Neea choriophylla are most important. Nine species are present only in the three eastern forest sites (Tzucacab, San Mateo, and Othon Blanco); two species of Coccoloba (C. coziimelensis and C. spicata) are the most important of these. The Tzucacab forest is a site of species distribution overlap. The three eastern forests share eight species with one or both of the ruin sites (Group V, Table 2). Three species, Manilkara zapota, Cosmocalyx spectabilis, and Gymnanthes lucida, are present at one ruin site (Mirador) and are especially important in the very easternmost forest site (Othon Blanco). Legumes comprise exactly one-third of the 24 species in the first two groups (Table 2). The western forests have a greater ecological Importance Value in legumes. Four of the ten Group III species (Western Species) are legumes, whereas there is only one, Pithecellobium albicans, in all the eastern groups (Groups IV, V, VI). In Group VII, legumes are more important in the western sites. No other distributional pattern is evident for the Group VII species; maybe each should be considered only locally important, or maybe they were present 1994] Woody Vegetation of Yucatan 11 -yi >^ n (U u 5J c t« !1/ ri r. p: ^ a- C r« 1 ^^ C/1 T3 C II T3 C < (« iJ t/j V ^ U t« n u- re > ^ x: :^ u :/: n n n u iM a; ^ ,,-^ ■a o T3 rr> OJ II Al P OJ N d: 03 H Q U ;j- ■y. cr hr 1/ OJ c U t/: U -i __ D. |2 V] o >~ X H ■u T! o ^ II ■^ c ^ 1^ . 0 =" ^1 n3 r^ w C II 's ^ II -a -J x: u x: ^ '^ - - " ~ ■-'"- w i K I £ w "S 1^ ^ Ij-c '^ ~~ — °^ ^;^ 1^ ■- y I s ^i 'x c n C i: > T! a; 0; C !1) x: "2 ^^ n o o u a r^ V 0 s ^J t/: l; a; '■" /^ (^ > ^ u o A! Q. o > F --^^ C c o u p X V c s 1; -s V 0 2 O II X r» 1j ^ H r^ ^ C C P ^ -vf (« x: ■ w 1; C^l -T* u s W c; pa <; H 3 N] H U H PJ ^ — C^J 0^ yD lo t^ iri ■<*" ff4 « 05 --< tri 05 o eo c» ■* !> CO iri to O CM CO o 00 IN m ti^ CO Tt- Tf — . Oj CM m oomoo — c^jCMr^Tt^inor^— ' eq p ■* C OC C^ CO s\i — Tf- o o in d I- C^l CO O X C: O) 05 5^ X 'x; f 1^ iR IR CO CM Ol — CO O tX! (O CO CM 05 I> I> ^ ^ oo r^ -H CM ■* r— iR ■* Tf X O) tT O- 'X 05 O) CO X CO (X 1— CO t^ in in x Tt-X^Xt^XCM — COTft^ in o CM o c^i — — o 3 CO CO ^- ■ — xt^Oiinxxi^oi'*X'+i>-rr 0)-- — c^j ov Tf o oi CM CM r^ CO, tt CO o m X CO X in — ~ CM -^ C^J — t^ in CO CO -r ^ — xO'-occd — — ~ c X -^ ^ ^ "^ s « a r- -S 'J~. S I ^ 5 G :? c c ■j~. • 2 « ? :: ;<-^ ^~ .^ ~ c ^ ■~ ^ "S a X lii =: ;j :<; -; = ^ •§-■=; 'i -^ S c ■= i; s ^ s § c ^ s •:^ s^^ '• 2 ■?■ ^ c ■^•i: f^- .~ 12 Tulane Studies in Zoology and Botany [Vol. 30 H X H o Z < c/5 P H 1; O O oi w S CD CD r^ CM 00 X oi in O CO CM CT) CM CTi C^ ■* 00 1> ^_ CO O "* csi ^ ^ G^i _' rt Ttiooi>^oooi>(ri Tt<(:oTj<(£5o^Lncooq COOCMUDCMOOCMi— I OOtD-HCOCMOO-*!— I cvood^oocMO Tf O lO i> in CM 00 o in cc to -H s^j CO o r^ CT) '^^r^CMin — CMCTit^ O CO — "' O O ^' CM o ^ in CO CM CM CM ^^ CO CM 00 CTi ^ ■-<■ O O CO CM -"t 00 in ooocMincM-^Lni> o C^JCT)CM-^CMCOcooodod* — ^ Tf eocMi>'-i(OCMcr!a5^ d-^CM^-n'od^d 00 tr> O CD y: CM in 1-H to ^ Tj< r^ in ^O* OCMOTfO u CO CO q CM ^ to r^ CO CO ^ CO 00 r^ CO y3 CM CM H d — < c4 ^ CJ CM d CM -^ d f-H oocor^in — Tt--.xc2 r» c, i> ^- 05 oo ■^' 5^i — c"! or, — '^ IT) -^ ^ ■— < CO O O o iT, 5^J O"' X 1^ 0-- o in OO 05 ^ o c^ CM t-^ O CO 1— I o t-^' ■-< o CO o c; t^ in in m -^ CM CO ^ y^ ^ r-' C^i rt ^o in cv o ^o in ^ !> CO CO CO O 1— I .— ( CO vc X -* ocoinc^ x^ OCMCV ^-O"— <^ C5 iT 05 --^ -! ^ ^ ^ -^ r^ d to Oh (J h S e c Mill 1 1 r J 1 1 * 1 1 ,M f 1 i A s 1^1? I|.| .i ■= 14 Tulane Studies in Zoology and Botany [Vol. 30 but unrecorded in some sites because of their rarity or clumped distribiUion. The two ruin sites have eleven non-leguminous species not usually present at any of the forest sites. Brosimum alicastrum was the most important of these ruin-species; at both sites it makes up over 33% of the Importance Value. As stated, Croton lundelii and Urera baccifera are also important species at the ruin sites. Discussion Forests within the northern (above 20°N latitude) Yucatan Peninsula are clearly dry-tropical by the Holdridge (1967) classification. More specifically, the five forest sites are within fotir different tropical vegetation types described by Miranda (1958), and corroborate quantitatively his height- and deciduousness-based classification system. The Calcehtok and Ticul sites are, in the "low-medium deciduous" forest zone of the western and central portions of the state of Yucatan (Figure 1), and have the predicted canopy height of 15 to 20 m. The San Mateo site is a "medium transition" forest with about 25 m canopy height. Adjacent to and east of this zone is Miranda's zone of "tall-medium sub- deciduous" forest with a canopy of 25 to 30 m; the Tzucacab site is located within this zone. The Othon Blanco forest is near the western edge of Miranda's "tall subevergreen" forest, which covers the central and southern portions of Quin- tana Roo. The increase in canopy height from northwest to southeast across the north- ern peninstila follows the change in species composition seen across the forest sites as identified by the ordination (Figure 3). These trends in canopy height and species composition mirror annual precipitation patterns for the peninsula presented by Garcia (1965; Figure 1). She gives no precipitation breakdown by wet-dry seasons, but recent official maps prepared by SPP (Secretaria de Progra- macion y Prestiptiesto) do give isohyets across the peninsula for both seasons from the years 1921 to 1975. Although the five forest sites are in the same isohyet (900 mm) for the wet season (May-Oct), the amount of rain in the dry sea- son (Nov-Apr) differs considerably, from 150 mm in the west (Calcehtok forest region) to more than 300 mm in the east (Othon Blanco forest region), a result of the prevailing westerly winds from the Caribbean Sea. It is likely that during those drier months the difference in precipitation of more than 150 mm affects plant growth and produces the observed gradient in canopy height. Past quantitative vegetation research in other areas of the peninsula supports this conclusion. Thien et al. (1982) presented data from the dry northwestern region of the peninsula, a forest site around the ruins of Dzibilchaltun (Figure 1) within Miranda's zone of "low deciduous" forest where annual rainfall is less than 700 mm. Here, canopy height ranges from 6 to 8 m. Whigham et al. (1990) described a forest on the east coast of the peninsula near Puerto Morelos, within Miranda's "tall subevergreen" forest zone with a recorded annual precipitation of about 1100 mm, although quite variable. That area has an upper canopy height of 25 m. Rico-Gray et al. (1988) worked in an old-growth forest at San Pedro, in the same area as the San Mateo forest. They recorded an annual pre- cipitation ofjtist imder 1000 mm and a canopy height of 20 m. The T/P (annual mean temperature/annual total precipitation) ratios for those three forests, and for forests reported on here, range from 3.7 (Dzibilchal- tun) in the northwest to 2.3 (Othon Blanco) in the southeastern part of the 1994] Woody Vegetation of Yucatan 15 study area, and thus are in the worldwide range for dry tropical forests as described by Murphy and Lugo (19S6a). Canojn heights in the Yucatan Penin- sula are typical for world dry tropical forest, varying from 8 ni (D/ibilchaltuu) to 35 m (Othon Blanco), and likewise woody plant Richness (DBH >ca. 2.5 cm), with a range from about 50 species (Dzibilchaltun) to 100 species (Othon Blanco). The broken nature of the canopy is a pronounced characteristic of the more western forests (Calcehtok, Ticul, Tzucacab, San Mateo), and this also was reported for the Dzibilchaltun, San Pedro, and the Puerto Morelos forests. Over the entire northern Yucatan Peninsula the forests are characterized by several widespread species. Bursera simaruba, Gymnopodium Jloribundum, Piscidia piscipula, and Thouinia pnucidentata are usually particularlv important. As in the present study, Rico-Grav et al. (1988) and Thien el al. (1982) found the first three species to be common and dominant. There is a species compcxsiiion dif- ference corresponding to a dry-season precipitation difference of more than 150 mm across the peninsula. In the west, Casealapinia gaumeri, Pithecellobium leu- cospermum. Acacia gaumeri, Bunchosia sioartziana, Diospyros cuneata, and Vitex gaumeri are important species in all the studied forests. The drv-season wetter forests in the eastern and northern region (within Miranda's "tall subevergreen" forest zone) have five species not usually found in the west: Brosimiim alicastrum, Drypetes laterifolia, Gymnanthes lucida, Manilkara zapota, and Talisia olivaeformis (this study and Whigham et al. (1990)). Murphy and Lugo (198f)b) reported that Gymnanthes lucida was, the most common tree in a Puerto Rican dr\ troj:)ical forest and it has been recognized as an important component of the Antillean flora (Miranda, 1958). Miranda (1958) identified Vitex gaumeri as the "abundant" indicator tree in his "tall to mediimi subdeciduous" vegetation zone, and an important compo- nent of his "mediimi transition" forest just to the west. From the results of this study, the indicator status for Vitexis in doubt. Even though Vitex gaumeri has the highest Importance Values at the two sites (Tzucacab and San Mateo, Table 2) within the two vegetation zones just mentioned, it cannot be singled out as any- thing but a general "forest species." However, this study does corroborate Manilkara {Achras) zapota as an indicator species of the most eastern vegetation zone of "tall subperennial" species: the species was found only at the two sites (Othon Blanco and Mirador ruins) in that zone. Additional quantitative sam- pling of the peninsula's vegetation needs to be done to confirm other Miranda conclusions. Could all published accounts of the northern Yucatan forests be describing aberrant vegetation so disrupted by lengthy Maya habitation that little resem- blance to a native type exists? Possibly, but this is a worldwide concern in dry-tropical forests (Murphy and Lugo, 1986a). Great care was taken in this study to find the least disturbed sites in the central peninsula, but it is difficult to assess the true age of any of the peninsular forests because of incessant culling practices and coppicing (Rico-Gray and Garcia-Franco, 1992). A few emeigent trees scattered among the canopy cover may be a false indication of old age. Still, in this study, all Mayan field assistants were certain that no site had been severely cut for farming during the past 40 years, and, as noted above, some indi- vidual plots, even entire forest sites, are more than 100 years old. Miranda (1958) predicted that forest species richness and diversity would not be signih- cantly different in the absence of lumians, onl\ that the indi\idual tiees would be larger. 16 Tulane Studies in Zoology and Botany [Vol.30 Rico-Gray et al. (1988) and Thien et al. (1982) named several species as indi- cators of the oldest quadrats studied if found in large size and high density. These include Bursera simaruba, Vitex gaumeri, Caesalpinia gaumeri, Piscidia piscip- ula, and Apoplanesia paniculata. Along with Lysiloma bahamense, these were among the largest trees in the five forest sites analyzed for the present study. It is certain that the forests growing over the ruins were the oldest studied; the Maya viewed them with a special significance and unequivocally stated that little culling had occurred within them for many years. The San Mateo forest of this study, and the San Pedro forest (Rico-Gray et al., 1988), are less than five km apart. Of the 51 species reported in the San Pedro forest and the 45 species in the San Mateo forest, 25 are common to both, a reflection of tropical diversity; using those 25 species to characterize the forests of this region seemed reasonable, and in total they contributed heavily to the Importance Values: 75% in the San Pedro forest and 66% in the San Mateo for- est. As for many other studies (Bartlett, 1935; Lundell, 1937; Folan et al., 1979; Lambert and Arnason, 1978, 1982), this study clearly shows that the woody vege- tation growing on Mayan ruins of the Yucatan Peninsula is different from the vegetation of even adjacent forests (Figure 3). Eleven species are associated with the two ruin sites (Table 2); they contribvUe 55% of the Mirador site and 75% of the Uitizina site total Importance Values. Of particular significance is Brosimum alicastrum, which makes up more than a third of the total Importance Value at each site. It has been thought (Lundell, 1937; Bardett, 1935), but recently chal- lenged (Lambert and Arnason, 1978, 1982), that Brosimum alicastrum, among other species, is more common around ruins because it is a relict from the ancient days of Mayan cultivation. However, such species may be associated with ruins because conditions for growth on rubble are satisfactory (Miranda, 1958) and free from competition from other forest species; i.e., conditions are optimal for growth and reproduction (Lambert and Arnason, 1982). For Brosimum, it might be that the bat-dispersal of seeds creates high potential for establishment on ruins, which are bat roosting sites (Peters, 1983). The Uitizina site is unique. Nine of the 13 widespread species (Table 2) are absent from the site. Manilkara zapota, historically a species associated with ruins (Lundell, 1937), is also absent. Oddly, this species is not important at the Mira- dor site (I.V = 1.47%). Talisia olivaeformis was identified as another ruin-associated species by Lambert and Arnason (1978). In our study, both of the ruin sites, as well as the three most eastern forest sites (Table 2), record it in low Importance. Conclusions This first quantitative study of the north-central Yucatan Peninsula's dry-tropical forests corroborates Miranda's qualitative description of forest types, and clearly documents a change in species composition and height across the peninsula, with these differences likely caused by small and local differences in dry-season rainfall. Certain species {Bursera simaruba, Thouinia paucidentata, Piscidia piscipula, Gymnopodium floribundum, Caesalpinia gaumeri) are widespread, quandtatively important, and therefore characteric of the forests in this part of the world. 1994] Woodv Vegetation of Yucatan 17 The average total basal area for all five forest sites (20.2 sq. m per ha) is low, but within the range of 17 to 40 sq. m per ha given by Murphy and Lugo (1986a) for dry-tropical forests. Likewise, the forests studied here fall within parameters from other dry-tropical forests with respect to T/P ratios. Richness, Diversity, canopy height, and structure. The abandoned Maya ruins of the northern peninsula have a distinct vegeta- tion dominated by Brosimiim alicastrunu this likeh due to the boulderv substrate. ACKNOWO^EDGMENTS The Mesoamerican Ecolog\' Institute of Tulane University under the direc- tion of Arthur Welden was our principle source of support. We thank Lovola University for giving D.A.W. a sabbadcal leave and faculty research grants for travel to the Missouri Botanical Garden and to Yucatan during the later stages of the study. Ron Leisner and personnel at the Missouri Botanical Garden were helpful with final species determinations. We thank the staff at the Field Museum in Chicago for help in the earliest stages of plant identification. Leonard Thien ancl Victor Rico-Gray supported the effort throughout and made valuable suggestions on the manuscript. A large number of Loyola and Tulane students helped with the fieldwork; their efforts under less than ideal conditions warrant a special thanks. The fieldwork would not have been possible without the aid and knowledge of several Yucatecans: Manuel Ken Cabellero, his sons Reyes and Pedro, and son-in-law Anatolio; Clotilde Can Quijano and friend Benito. Literature Cited Bartlett, H. H. 1935. A method of procedure for field work in tropica! .\nierican phytogeography based upon a botanical reconnaissance in parts of Britisii Hondmas forests of Guateniala. Carnegie Inst. Washington Piibl. No. 461, 25 pp. Be.\LS, E. W. 1985. Bray-Curtis ordinations: an effective strateg)' for analysis of multivariate ecological data. Adv. Ecol. Res. 14: 1-55. Brown, S. and A. E. Lugo. 1982. The storage and production of organic matter in tropical forests and their role in the global carbon cycle. Biotropica 14: 161-187. Cano, E. C, M. Sousa S., and O. Tei.I.EZ Valdes. 1982. Imagenes de la Flora Quintanarroense. CIORO, Puerto Morelos, Q. Roo, Mexico, 224 pp. Causton, D. R. 1988. An Introduction to Vegetation Analysis: Principles, Practice and Interpretation. Unwin Hyman Ltd., London, 342 pp. FOLAN, W. J., L. A. Fletcher, and E. R. Kintz. 1979. Fruit, fiber, bark, and resin: social organi/.uioii of a Mayan urban center. Science 204: 697-701. Garcia, E. 1964. Modificaciones a! sistema de classificacion climatica dc Koppcn (para adapterlo a kis condiciones de la Republica Mexicana). Offset Larios S.A., Mexico, 71 pp. GarcIa, E. 1965. Distribucion de la precipitacion en la Republica Mexicana. Publicaciones del Instituto de Geografia, U.N.A.M. 1: 173-191. GOMEZ-POiVlPA, A. (ed.). 1978-. Flora de Veracruz. I.N.I.R.E.B., Xalapa, Mexico. Holdridge, L. R. 1967. Life Zone Ecology. Tropical Science Center, San Jose, Costa Rica, 206 pp. Hoidridge, L. R., W. C. Grenke, W. H. HATHEWA^■, T. Lianc;, and J. A. Tosi, Jr. 1971. Forest Environ- ments in Tropical Life Zones. Pergamon. NewA'ork. 747 pp. Isphording, W. C. 1975. The physical geology of Yucatan. Trans. Cluif Cloast .Assoc. Cieol. Soc. 25: 231-262. Lambert, J. D. H. and J. T. ARNASON. 1978. Distribution of vegetation on .Mayan ruins and its relation- ship to ancient land-use at Lamanai, Belize. Turrialba 28: .33-41. Lambert, J. D. H. and J. T. Arnason. 1982. Ramon and Mavan ruins: an ecological, not an economic, relation. Science 216: 298-299. Lundell, C. L. 1934. PreliminaiT sketch of the phytogeography of the Yucatan Peninsula. Carnegie Inst. Washington Publ. No. 436: 257-.321. 18 Tulane Studies in Zoology and Botany [Vol. 30 LUNDELL, C. L. 1937. The vegetation of Peten. Carnegie Inst. Washington Publ. No. 478: 1-244. LUNDELL, C. L. 1938. The 1938 botanical expedition to Yucatan and Quintana Roo, Mexico. Carnegie Inst. Washington Yearb. 37: 143-147. McCUNE, B. 1987. Multivariate analysis on the PC-ORD system: a software documentation report. Hol- comb Research Institute Report No. 75, Butler University. MiLLSP.\UGH, C. F. 1895. Contribution to the flora of Yucatan. Chicago Field Mus. Bot. Sen 1: 1-56. MiLLSPAUGH, C. F. 1896. Second contribudon to the coastal and plain flora of Yucatan. Chicago Field Mus. Bot. Sen 1:282-339. MlLLSP.\UGH, C. F. 1898. Third contribution to the coastal and plain flora of Yucatan. Chicago Field Mus. Bot. Sen 1:345-510. MlR,4ND.\, F. 1958. Estudios acerca de la vegetacion. In: E. Beltran (ed.), Los recursos naturales del sur- este y su aprovechamiento. 2: 213-272. Inst. Mexicano de Recursos Naturales Renovables, Mexico City, Mexico. MORLEY, S. G. .\ND G. W. Br.\INERD. 1983. The 7\ncient Maya (4th ed.). Stanford Univ Press, 708 pp. MUELLER-DOMBOIS, D. AND H. Ellenberg. 1974. Aims and Methods of Vegetation Ecology. John Wiley & Sons, New York, 547 pp. MURPH\', P. G. AND A. E. Lugo. 1986a. Ecology of dry tropical forest. Ann. Rev. Ecol. Syst. 17: 67-88. . MURPH\', P. G. AND A. E. Lugo. 1986b. Structure and biomass of a subtropical dry forest in Puerto Rico. Biotropica 18: 89-96. Opler, p. a., H. G. B.\KER, AND G. F. Frankie. 1977. Recovery of tropical lowland forest ecosystems, pp. 379-421. In:]. Gairns.Jr., K. L. Dicrson, .AND E. E. Herrigks (eds.), Recoveiy and Restoradon of Damaged Ecosystems. Univ. Press, Charlottesville, Virginia, 531 pp. OREJUEL.A, J. E. 1980. Niche relationships between turquoise-browed and blue-crowned motmots in the Yucatan Peninsula, Mexico. Wilson Bull. 92: 229-244. Peters, C. M. 1983. Observations on Maya subsistence and the ecology of a tropical tree. Amen Antiq- uity 48: 610-615. PlELOU, E. C. 1969. An Introduction to Mathemadcal Ecology. John Wiley & Sons, New York, 286 pp. Rico-Gray, V., J. G. Gargi a- Franco, A. Puch, and P. SimA. 1988. Composition and structure of a tropi- cal diy forest in Yucatan, Mexico. Internat. J. Ecol. Environm. Sci. 14: 21-29. Rico-Gray, V., J. G. GarcIa-FRANCO, A. Chemas, A. Puch, and P. Sima. 1990. Species composition, simi- larity, and structure of Mayan homegardens in Tixpeual and Tixcacaltuyub, Yucatan, Mexico. Econ. Bot. 44: 470-487. Rico-Gr,\Y, V. AND J. G. GarcIa- Franco. 1991. The Maya and the vegetadon of the Yucatan Peninsula. J. Ethnobiol. 11: 135-142. Rico-Gray, V. and J. G. Garcia-Franco. 1992. Vegetadon and soil seed bank of successional stages in tropical lowland deciduous forest. J. Vegetat. Sci. 3: 617-624. S.\BOG.\L, C. 1992. Regeneration of tropical dry forests in Central America, with examples from Nicara- gua. J. Vegetat. Sci"^ 3: 407-416. Srannon, C. E. and W. Wea\'ER. 1949. The Mathematical Theory of Communication. Univ. Illinois Press, Urbana, 117 pp. SOSA, v., J. S. Flores, V. Ric:0-Gr.\y, R. Lir.a., and J. J. Ortiz. 1985. Lista Flon'stica y Sinonimia Maya. In: V. SoSA (ed.), Etnoflora Yucatanense 1: 1-225. INIREB, Xalapa, Mexico. StandleY, p. C. 1920-1926. Trees and shrubs of Mexico. Contn U. S. Nat. Herb. Vol. 23. Standley, p. C. 1930. Flora of Yucatan. Field Mus. Bot. Sen 3: 157-429. Standley, p. C. and L. O. Williams. 1946-1977. Flora of Guatemala. Fieldiana: Botany. Thien, L. B., a. S. Bradburn, and A. L. Welden. 1982. The woody vegetation of Dzibilchaltun: a Mayan archaeological site in northwest Yucatan, Mexico. Middle Amen Res. Inst. Occas. Pap. 5: 1-18. Whigham, D. F., p. Zugast\-Towle, E. Cabrera-Cano, J. O. Neill, and E. Ley. 1990. The effect of annual variation in precipitation on growth and litter production in a tropical dry forest in the Yucatan of Mexico. Trop. Ecol. 31: 23-34. Whigham, D. F., I. Olmsted, E. C. Cano, .\ND M. E. Harman. 1991. The impact of Hurricane Gilbert on trees, litterfall, and woody debris in a dry tropical forest in the northeastern Yucatan Peninsula. Bio- tropica 23: 434-441. 1994] Woody Vegetation of Yucatan 19 in ,1 PS ~ S £ aj c Q. X U E - c V c^ n. X ^ c -= ,_^ w I ^ C II ^< 2| £ r* r^ P "^ 0 C u u II a. W Ob §^ ^ II jf Q I t^ w o c _o a '5 "a- JJ aj — -^ 0- — -^ 15 C c« ■£ Al f2 |i| C ^ 1- t''' r- ^ Ji § :: C/J i« o a2 z UJ C 5 to t^ CM C: Oi in — I^ e<~, t^ X — ' — ir. ir. o 'J: X ?J (MCOCCCC CT) «Oir;tCCTi'rfvr, CCI^X-^OOtCO CM (N X X cc cc i^ c^i V. X i^ ir, in o s^i :^i O O X C", — T CC O '-^ 5^1 5^1 in -p 3". 3^ r-^ o X -t- o oc p4 'X: '^ — c^i in J^i -C t^ X to en in i^ CM in X s in CO X CM d o (M CM r^ •05 — '-^ — CO o -f c^i in '-c — X CO — o r^ en CO — CM X X in in c^i in CO X in CM 3^ ai X CO Tf X 'f -^ CO m CO c^j — ■ X CO oi in ^ I-- in •^ --c X be — p ^^U 1—1 M = ^ 9-w be c 1- s: ~ s "s :s = :j~ ^ 1; _0 CM H o I— X •:= ?2 ?■ fc C . biD .y O ■5 oQ — < , iS ^ J P3 « ^-' %^ a -s ■« . T3 cr =: CA) 2 J? tl. o s ■« -a. c -< -< CQ ■):^ ■S s i 4 c s^ ^ ^1 05 o S s s 'C o o- 9 -S S ^ D- — jri ,^. ^ > S 03 = I S = ■^ 'S X its "5- -S -S 5 s s X — 32 ^ g cr u "^ >; r -« _ 2 ■ji:. rs -2 « i'^:-i ^ ^ :^ i; q: ^ to CM — in CM in CO r-H to d CO oi to to --I cr> — to X X '-C' — CM to o r^ — CT' o in 'JD (T. d CM O — ■* ,-H 'f to 5" X CO 1^ 1>l X CM X CM to d to CO ■*" J r^ c <^^ X F 1 o^ I g >f 1 cl X •J-, 5 •C j^ "a, "y. "*: '=>: X — 5 ir. '» I^ — X C^ Z CM C^J 1^ o t^ — in --; Tt- r-^ ^ in CO CO X X in d d C^J CO CO CO t^ c>i X X X t^ in o X X — 3": o o q -• CM in q ^ CM to to -^ r~ d -^ c^i c-i c^i i--^ 'T to i> i^ ,— ^H CO t~- c^i in in c^i in co c^i X c in in r^ o m in CM c^i — X r- in to CO CO CM X c-i c^j — in CO CO — X — c-i I- — CM t^ CM '-C CM to ?1 — ■^ CO X — ' CM -^f -^ — C^l TftocMCMCMX-^— -mtoi^ C: — X bo c iS u — : c o J2 u u 3 N J^S %. ■« s -t "s I '-^ ^ I ¥ ■ = 5 S ~ '^ 5 £ ■c -J I 5 cr X r ^ d. =^ ~ ca r- ^, s>ccxc£ . '= -S —• i< J< -. oi r Sx2:5ti5t.i-:vi CQ a; -, ~ P s U Q uq it; -5 ^ 2: -^ :^ ^ a; s ^ — -t: t; 20 Tulane Studies in Zoology and Botany [Vol. 30 r- cr- oc CM cr- i> Z w CM oc CO CM CO X CO o H c/5 0/ ■o u o r ^ ^^ X S z w c/5 Z H c/) u. ir, i> J> t^ Tt< '-C O O i> CM r^ ^ O^i iC 'O^j * I^ X 'i^ no CO -H Tj< lo in ^H CO GM Tfy^CM-^-^tCcoO Tj< CM ^ „ rt 05 in 00 in X m CM (£3 CO in i>r~cooc^a2j><^ tc '.^ CO o CM CO tr> X to in CM to (o' 00 «5 iri "* ^ CO CO CO O - . CO in CM 00 -H X (O in ^ I> t^ t^ — CO CO CO 'f .— --■ to c^i o to CM to CMC^I— I — -^ '* CO en X r-H X CMcotom-^cMcoxcMcoint^to-* mo^ ^ c ^ "-^ =: -g g ■~ ~ s ? is be C '^ £ ? ^ !^ 3 e S CQ cq o U cu ^ ^^ s ■2 r- c c C/5 0 ^ K ^ 0; 03 J ^ ^ o e — ' .^ s g « s € C ^^ ■« c/5 ^ ^ i* -j ^ « -S "s s ^ « t! 1= I I" I .1 § '^ 6 ^^ ^' c o S s' a; -^ O ^ fi; a; Dh ~ • ~ -a. £c 2 tO' to X 'tt X CO r^ ^ CO in CM ^ ■* CO CO ■* cc^ir^co-^to^o^CMtor^x^o^ ^ CO r^ CO CO o to ^ 00 r^' cri Tf< ^ ^ c^i Tf- -^ -H .-H X> CO CM CM X o t^ o t^ C^J to " -H to CM in ^ ^_ to CM in ^ — < 5 o o " 00 to t^ r^ in to to c^i in to c^i CO to ^ o o o — ' X in O O O ^ X ^ ^ ^ to to o in in in in CTi in 'J" ^^ CTi CM X c^i X CM to ^- CO CO ^^ to 0\ ^H i> ^ -^ coTt^dco.— 'CMCM-^Tfcoi— iin tj 3 N CO ^ ^ bD C .o OJ be . i? id C/3 cn a; oi c2 X <: e 2 "S -S 5d g r; S fee .^ ■s; "- « <^ « *- :; ^ ?^, f- -a ts b '^^ X 5 1; '^ H3 = Si. w 2 X ii -^ S =Q c :x -5 oa tu ^ i ^ f i e g tT ■r^ -:i -r^ t^ c ^ 2 -5 05 03 03 O O 3 c. ■< -J •< -2 ic s: c^ s o "o- P c3 _0 "H. 0 fcu CQ < N X CM CM in X o to CM tO C^l o X 05 (05 in 1 — 1 CO CM CO CO tT d to CO Tf d d d 00 to' 1> CO P to CM r^ O' CM m o CO in X in 00 to CM in 00 CO to o X d CM CM tji in 00 CO d CM 00 X d d CM Tt^ T— 1 CO in o o CM o X -H in -H to CM to CO in 05 CO in Tf' c^i CM O CM ■■^i rt Tt< -H Tti r^ C O OJ t; ^ ^ CQ X ^ • S J- E ts ■-^ K ^ S =5 p •a cr ^H c w o — ^ c« t« ^ C^ CTl r-; 3M C^. — -^ "T ■* 04 X r^ (N — cc iC J^ sC CC — ■ X ir: oc c-' t^ rp i ~' c^i ^c r^ 3 X s^i c^. oi ?4 '-c cc 1^ f t^ X o X w c^. c^. ^ ?i ir X C-. — 1^ i^ -r c^. r ■-C -r ~ :<". ir, c^. — ir. r: 71 1^ o M — '.c -r X — — c^j — — 1^ -r •-C CTi 00 O X tC u: — ■ ~ — — -X: — ?l CM — . tC t^ ^ CM ui X >n — ; — _ ir, tt ir, oi iri oi cc ^ oi c^i TT c>i — ^ — ^ iR r-^ 00 inooi^ coo — ^cr. ^^o Oi— 1 CO t^ I^ C: O CO to — O IT. X O Tf IC SM „ in t-» — 3 O I^ I^ CO O I^ CO CO O I^ I^ O CO CO ic. ic — '~c CO o — CO CO ir, — — ir, CO X r-^ t--^ c^ — x o oi X CO CM Tt 3-; t-^ x o CO CO t^ a-/ lO in CM CO — ic t^ X cm CM a~; CO CM — CO t^ X ■* '.o CO CO X o CO— '-^ r^— ' o CO in — CM -^ CO X CO CT. C75 CO CM X C^l t^ C^l X CO CO CO — 71 lO •* CM CO — — -f CI c^^ 05 — — ' in o^ — ' tt in CI 3~' CO >n CO cm co t-^ x cm rf t^ CT) cr> c~. '.C co — -r — co i^ ~ ci co s ^ = ^ Si j5 c u s g=: C: ^ ^ 2; S; e o o -i: s 1; ■5 O c m o ^ ^ J ■^^ j^ « s -a, •S =^ ^ = ^ £ Di ^ s: a: s e -1 E, F S ^ -fc ■ S o ^ -ft- s g -0 ■q- 6 o CM X5 s ft -J — '^ s s cq 03 s •2 br. o be J « e c c 3 ■2 ^ ^ 8 u u 3 ^ 5 '^ £ ~ f* S ^ P ^ p^, 1= I •c = X :: 5 &: n; r °3 & 5 — ; X' c-2 -J "I" ft £; OUQiiiti02;-i ^ < i, a, a; 1/5 ~ X '-T; CI CD CO X in CI T — t^ cri •—' O y^XCOcD'-H CTi — -HCMCMCMJ>y3cq tL. J^O^DOCMOicOOC^'^'OTt'COO^ O CM Ol tC p I> tt iri o r-^ CM — — CM cr- in cTi CD -^ CD tC CM Tt' CD oi — — in CM CD 'D O t^ I^ CD in 'D — o CM in — ' CO CD CI CM — CD t^ 'D D -r t^ — C: CD '-D in CM ci cri X t^ i^ Tt- Tj< o O ^ x> in — ' CMXt^in-^cDCM'tco — — X V Jt (^ - J= u c/5 -a 1; ■XT cQ _• ~ 2 -C •;= CQ ? — q3 .2 a- 5 ^ 2 .= S It bo "" ^ .i 3 ~ S '^ c S 5: X -^ 4. ■a. .S -c ^ ^ ~ Cfl I X £ •S: S Oj S « I ■5 « Sic I s e ■- ■-J --J ^ 5 ^ cc s; V5 r-< s u X t3 C be. oi • c^ =^ J-;^CQ •C E 0; •a ca ■£ 2 ^ c: "i ■^ =^ s e " s S •2 -S i £ ■? 'w -2 '^ .>; ^. C "7= ~ " ' 'i: "^ cq CQ C Si u T3 C 3 X X K ^ is 3 S -^ ^ ^ I? ^' X 3" 3 c £i Di -3 ^ a. cQ -u S ~ r^ ;■= s -s CO = = ii 2 Z 3 CQ -a t I §11 ;j o C! s e ■p? t£ i a c = q Lq tS o ^:^ 5 5 ■ ^ ST ^ c ■- c: ■~ ►J ^ ^ a, 0, 22 Tulane Studies in Zoology and Botany [Vol. 30 Z Q c/: < C s I> CO o O «D CM q 00 q CO CO 00 1> in q CM CO m ^ O) I— 1 CM 00 CO ^ o I— H o^i CO 00 in "* t-H ^ 00 d i-H CM ^^ iri d CM T-H in CO d CM i-H d d iri CM d iri r— 1 T^ ^ 05 in ^ 00 00 o o CO q o o to in en to in in 00 Th (£5 tn in in 00 00 rr CO CD r^ o ^ ^ d GM CM r-; CM r-^ ^ r^ CO iri CM CO CO CO iri CM iri r^ d t£) CM in CM t^ iri d ^ o ID '■^ ^ CO CM Tf lO - o C^I CM CM CO 00 00 CO - CM '^ CO - in ■* Tf< en CM CO 03 fed --- c ^^ c £ ■ 5 S "^ I .« bo E^tS 00 -H 00 o — H o in '^ in (D 00 IT: z d in d d CM Q CM t/5 s CM CM in w in H c/) in 'H CO 1; ^ .i ^ ^ S -c ;5 ;= o2 ^^ S c 5: s "S c/) ^ "^ ■^ — ' II ■2 1 ii 1 ■c .2 « -S .s S ■ ^ 5 53 S 5 ■^ in in CM'*cM'-— I ^H CO a; J2 OJ OJ ■O S -I So :? = « O s t2 c^S N s s s c u be S-c c« — & S tJ ~i ^ to § ^ - ts .!=• re; ^ 03 cq O lis Vi CQ X « s cr -J 2 ^ c -J s S "a ^ S ■i -^ -a S S S e be 2; ^ c s c75 2 s •5 ■= pa IS S ^ 2 « o S n t« X! u u ■c D u. ^ cr ^ 0 'c5 ^ 0 ^ t; g ^ ^ ^ g ^ -c "« 1 c r^ ^ J ■|^ c OJ '- ^ s ■ cr to 00 1 ^ s js s o ooooo-*^ cct^coooininaiOTfoooo-^cocM^^CMO 2!?, ooooinini> ^-iHOCMCM-HLncoinxinq-^inqi>r^'* ^"^ dcMocodd ddcMj>o6cMddo6o6diri^coco-H-^r—( toqj ir C-l r^cOCO^CM tO<0 ^CMt^CO'^QOO CO CM r-i CM CM --^ inotocMin ini>CMtoxr^oi>tO'— iCM^^r^tocot^'-^in '~! ^^ oor^incM-H oooincMTfor^cocMXin^qinqqoqco '^. ^ ^co iri CM06 ^'Tt'ood— ■■^codd-^od-HTj-inCTi^'*-^ TrT— ■ i>->+i t^^HincocM cMini— i^-"^HOMincMCMto ^^i— 1 --H 00 to ^HintOinc^Jcr)CMoo^^J>ooinoint^QOoO'— I -^iin Oi fad CQ II £ -^ c & u .ts o • — fe ci- ts s go H ^ ^ s: cc C/5 f2 S 1994] Woody Vegetation of Yucatan 23 CO «2 OC CD ^ I> Tf o CO CM ^ OC 00 ctj f in X -^ — f CM ^ in OC CM o t^ cc in l> CM in CM O) CM o ^^ in t^ ■^ o o o o CO ^tc(OinO'—ininOrtino CO 'J' r^ ^ CO -^ CM CM Ol ■^ 1> CM CTl CM CO -H S CO 2 -^ cci>t^r~-r^r^ccr^t^ OC ^^ — ' --^ CO CO' cc -^ CO ■*o6-^-* ^-i CM ^ OC ■q-^- CMCC — I^-^'-^CM^-^in^-cfCl c^i — < CM in r- ■"f Tf CM ^ cc ■* — rr c^i bo u C a u X! a. .r (« — ■ O -c C/5 c rs u a. ^ S 5 05 oa 5 -^ X i: .5 CO g ■a. g .2 S .= Sf S p . 3 S 9--S -S Q. ^ "S Sr s .2 t£: a: :«: c a; >-- ^ O c - J^ > X J J « e -o t/3 CC CO -* OC o r^ OC in ^- — cr. CO CM CO X CO CM O CO in 0-. X tT in x CM CO X in X t^ o^ in cc r^ CO in r^ r^ r-- o o CO ^- in in --^ tT CM CM O t^ -r — ■ --■ CM cc o — OC cc in ^ CO — ■>% OC CC .I' X) XI S to ■S -J 1-1 ^— c ^ 1) X) u. t>C -r — X ^ — Cli '-^ CC = 5 cr ^ '= "5 ^ "^ ca ■ ~ ;^ 4^ . ^ "^ -^ <3 '3' JS c 5 -= '~ •** ^ ^ ^ ,^ 03 U • 12 o ■!- S i s u vj ^ r ^ < — if. — I"^ ci 2- 2- 2? OC Tf OC X •o m OC OC 5C x. E^ ^ ^ 1^ X r-. ^ ^ ^ X in -r I- X 3-. o in ~ OC 'T CC - u. — X ^ CO — in CD ^ o^ Q^ CTi r^ in r- co o (^ X in o in o ~ ■>l — OC CM S=J ^ gi ^ 2^ X t^ X *^ c^ 3*- CO -O CO — Tf in c^i CM OC g? ^ ^ 6S CD X in X CO in t^ i^ -r '*r ■:::^ ^ T in in — r — - 1/ yj *H r*^ ■^ ^ ^ ai c^ - ■>. 7 I ;j !- 24 Tulane Studies in Zoology and Botany [Vol. 30 Appendix II Presence of 68 rarer species (Importance Values <1.0 in all sites) at five forest sites and two ruin sites in the north-central Yucatan Peninsula, Mexico. CAL = Calcehtok forest; TIC = Ticul forest; TZU = Tzucacab forest; SAN = San Mateo forest; OTH = Othon Blanco forest; MIR = Mirador ruin forest; UIT = Uitizina ruin forest. CAL TIC TZU SAN OTH MIR UIT Acacia collinsii Safford Acacia dolichostachya Blake Acacia pennatula (Schlecht. & Cham.) Benth. Adelia barbinervis Schlecht. & Cham. Agonandra obtusifolia Standi. Albizia tomentosa (Micheli) Standi. AUophylliis cominea (L.) Swartz Bauhinia ungulala L. Caesalpinia yacatanensis Greenm. Callicarpa acuminata H.B.K. Casearia randioides Lundell Casimiroa tetrameria Millsp. Cecropia obtusifolia Bertol. Cedrela odorata L. Ceiba aesculijolia (H.B.K.) Britt. Sc Baker Chiococca alba (L.) Hitch. Chrysophyllum mexicanum Brand, ex Standi. Coccoloba acopulcensis Standi. Cordia dodecandra DC. Cornutia pyrarnidata L. Cupania glabra Sw. Dalbergia glabra (Mill.) Standi. Diospyros verae-crucis Standi. Diospyros yatesiana Standi. Dipholis salicifolia (L.) A. DC. Duranta erect a L. Enterolobium cyclocarpum (Jacq.) Griseb. Erythrina americana Mill. Eugenia laevisvar. gaumeri (Standi.) McVaugh Exostema mexicanum Gray GuazLima ulmifolia Lam. Helicteres baru^nsis ]2iC(\. Heliocarpus americanus L. Jacquinia macrocarpa Cav. Koanophyllon albicaule (Sch. Bip. ex Klatt) K.&R. Leucaena leucocephala (Lam.) deWit Malpighia glabra L. Neomillspaughia emarginata (Gross.) Blake Nissolia fructicosa Jacq. Parmentiera aculeata (H.B.K.) Seem. PhyUostylon rhamnoides (Poisson) Taubert Piper sempendrens (Trel.) Lundell Pisonia aculeata L. Platymiscium yucatanum Standi. Plumeria alba L. 1994] Woody Vegetation of Yucatan 25 Appendix II (ContimR-d) (Al IK. I /.I SAN (VIM MIR IIT Fliinima ohliisa l„ Pouteria campechitma (H.B.K.) Bat'lini Pouleria unilocularis (Donn. Sm.) Baehni Pseudobombax elliptic urn (H.B.K.) Dugand Ranitid loiiiriliih/i Hemsl. Randia tmncata Greenm. & Thompson Rehdera tiinernis (Blake) Moldenke Sabal mexicana Martius Senna atomaria (L.) Irwin &: Barnebv Senna racemosa (Mill.) Irwin ic Barnebv Sideroxylon obtusifolium (R.&S.) Penn. Simarouba glauca DC. Spondias purpurea L. Tabebuia rosea (Bertol.) DC. Thevetia gaumeri Hemsl. Trichilia arborea C. DC. Trichilia glabra L. Ximenia americana L. Xylosma flexuosum (H.B.K.) Hemsley Zanthoxylum caribaeum Lam. Zemenia frutescens (Mill.) Blake Unknown No. 1 Unknown No. 2 • • Totals 20 8 41 8 17 2 \ POPULATION STRUCTURE AND CLONAL (iROWTH IN BROmiJA PINGUINL. (BROMELIACEAL) IN DRY FORESTS OF COASTAL VTRACRUZ, MEXK:0 Jose G. Garcia-Fr^nco AND Victor Rico-Gray Insliliiln dc Kiiiloi^ifi, A.C. Apdo. Postal 63, Xalapa, Ver., 9W00 Mexico ABSTR.ACT We Studied the distribution, population structure, and clonal growth ol Bw- melia pinguin in two areas (young and old selva) of the tropical deciduous forest at Estacion Biologica La Mancha and related those parameters to soil character- istics and seed and seedling predation. Bromelia pinguin inhabits both voung and old selva, but characteristics of the young selva allow populations of the species to produce more ramets and cover a greater area than do Bromelia populations in the old selva. Environmental conditions in both areas are not homogeneous, differing in soil characteristics and the size of canopy gaps. The Bromelia popula- tion is composed of adult and young individuals that reproduce sexually and asexually by clones. Recruitment is based on clone formation. More than 30% of the ramets in each population were mother plants. The percentage of mother plants in old selva populations was significantly higher than in young selva. This difference suggests that conditions in old selva are more favorable for reproduc- tion. The mean number of stolons in most populations was slightly higher that one, suggesting that maintenance and increase of the populations is through clonal growth. INTRODUCTION In the tropical deciduous forest of La Mancha Coastal Research Station. \'cr- acruz, Mexico, several factors influence plant distribution and abundance. This commiuiity grows on sandy soil of different ages: fossil dimes with a possible Late Glacial age, and recent sand with a possible age of less than a few centuries (see Kellman and Roulet, 1990, for detailed characteristics). Soil characteristics influence plant species distribution and ecological importance values. For exam- ple, on the fossil dime, Bursera simaruba (L.) Sarg. and Brosimum alirastntm Sw. are prominent ("old selva"); in the recent sand, Enterolobium cyclocarpum (Jacq.) Griseb., Cedrela odorataL., and Coccoloba barbadejisis jacq. are prominent ("young selva"; Kellman and Roulet, 1990; Ramirez and Garcia, 1990; Blain and Kellman, 1991). In addition, root proliferation and fine-root biomass are larger in recent sand (Kellman, 1990). Predation is another factor that influences species distribution and abun- dance in the tropical deciduous forest at La Mancha. The red land crab {Gecarci- nus lateralis Frem.) is abundant and consumes seeds and seedlings of several tree species, e.g., B. alicastrum and E. cyclocarpum (Blain, 1988; Delfosee, 1990), Tulane Studies in Zoolog)' and Botany 30: 27-37. 1995 27 28 Tulane Studies in Zoology and Botany [Vol. 30 including the clonal herb Bromelia pinguin L. (Garcia-Franco et al., 1991). Crab predation affects diversity, abundance, and recruitment in those plant species (Delfosee, 1990; Garcia-Franco et al., 1991). Here, we report on a study of the distribution, population structure, and clonal growth of Bromelia pinguin at La Mancha Station, and discuss the effect of soil and predation on those parameters, and on population maintenance. Materials and Methods STUDY SPECIES. Bromelia pinguin L. (Bromeliaceae) is a terrestrial plant with indeterminate growth, inhabiting tropical lowland deciduous and dry forests (sea level to 780 m) from Mexico throughout Central America to Venezuela and the Caribbean Islands (Hallwachs, 1983; Smith and Downs, 1979). The species presents serial monocarpy or sympodial dichotomy (plants produce a series of monocarpic axes; Benzing, 1980), and also reproduces by clonal growth. In La Mancha Coastal Research Center, B. pinguin forms dense patches or populations (demes) in the tropical deciduous forest. Flowers are present for only a few days, usually in March and April during the dry season. Butterflies and hummingbirds {Amazilia beryllina and Cyranthus latirostris, Trochilidae) are the main flower visi- tors. Abimdant fruits and seeds are produced each year (Garcia-Franco et al., 1991). However, few genets are recruited because seed and seedling predation by land crabs, Gecarcinus lateralis, is very high, sometimes 100% of the seeds and seedlings being eaten (Garcia-Franco et al., 1991). The number of genets in each Bromelia population is unknown, but it is possible to distinguish clones based on physical separadon. In the present study, we consider each unit (clone) as an individual (sensu Jackson et al., 1985), and the largest plant in each unit as the "mother plant" for the clone. STUDY SITE. Field work was conducted in lowland tropical deciduous forest at the La Mancha Coastal Research Center (Centro de Investigaciones Costeras La Mancha, CICOLMA) in the state of Veracruz, Mexico (19"36'N, 96"22'W; alti- tude <100 m). Annual precipitadon is 1100-1300 mm, with most rain falling June-September. Mean temperature is 22"-26"C, with a minimum of 10"C and a maximum of 38"'C (Moreno-Casasola, 1982). Several types of vegetadon are present in the area of the field station: tropical deciduous forest, tropical dry forest, sand dune matorral, and mangrove forest (Novelo, 1978; Moreno-Casasola et al., 1982; Rico-Gray and Lot, 1983). The tropical deciduous forest has a relatively simple structure and composition with tree species such as Bursera simaruba (Burseraceae), Brosimum alicastrum (Moraceae), Enterolobium cyclocarpum (Leguminosae), Cedrela odorata (Meliaceae), and Ficus cotinifolia H.B.K. (Moraceae). Common shrub species include Nectandra coriacea (Sw.) Griseb. (Lauraceae), and Coccoloba barbadensis (Polygonaceae) (Ramirez and Garcia, 1990; Blain and Kellman, 1991). Soils on recent dune (young selva) and fossil dune (old selva) are described by Kellman (1990). Sand pardcles comprise >95% of the soil at the recent sand study site, but this is reduced to 70-80% on the fossil dune site, where consider- able clay formadon has occurred. Soil pH is alkaline (7.8-8.5) in recent sand, but approximately neutral (6.9-7.1) in the fossil dune soil. Cation exchange capacity varies from 3 to 12 mmol/lOOg in recent sand, depending upon organic carbon content, and 20 to 33 mmol/lOOg in the fossil dune soil (correlated with the clay 1994] Population Structure o{ Bromelia pinguin in Veracruz 29 content there, r~=0.05, P<0.05). Total nitrogen in the soil increases throughout the succession, and is closclv correlated with organic carbon (r'=0.94, P<().()r)). Extractable cations are dominated b\ C:a in the recent sand, and Mg and K are substantially higher in the fossil dune soil. A total element analysis of Ca, Mg, K, Na, and P indicated that soil on the fossil dune has experienced a net loss of Ca, Mg and P dining weathering, biu no net loss of K and P. (Kellman and Roulet 1990). Soil water infiltration rates on both recent and fossil dune soils are high (Kellman and Roulet, 1990). POPUL^\TION STRUCTURE. Data were obtained between June 1989 and June 1990. All subpopulations of Bromelia pinguin were identified in the old and yoiuig selvas of the stud\ area. All individuals in each subpopulation were coimted, and listed as either non-reproductive or reproductive. For those that were reproductive, we distinguished asexual reproduction (clonal growth, with stolon presence) or sexual reproduction (presence of fruit or flower). To esti- mate plant size, we measiued the length of five mature leaves (base to tip) from the third leaf verticil. CLONAL GROWTH. Clonal growth in Bromelia pinguin was assessed in 10 of the 16 populations studied; five on young selva, and five on old selva. The subpopu- lations were selected from a variety of light conditions, slope variations, and types of surrounding vegetation. Clones of B. pinguin are so intermixed (tan- gled) that observations were sometimes difficult to make (see Table 2 for density data). Developing stolons were divided into two categories: those with a rosette of leaves (rosette), and those without expanded leaves at their tip (stolons). All sto- lons were measured monthly between June 1989 and July 1990. For those scored as "stolons," we measured the distance from the base of the leaf of the mother plant to the stolon tip. For those scored as "rosette," we measured the length of five leaves beginning at the base of the rosette. Growth data were pooled to obtain a mean monthly rate of growth, and correlated with precipitation and temperature data recorded during the study. As stolons developed and leaves expanded, they were then classified as rosettes; therefore the number of obser- vations in each category changed during the study. The mean number of stolons per mother plant was obtained from reproductive-stage ramets present in the populations; nine of these with at least ten ramets with clonal growth were com- pared (Table 2). Results DEMOGRAPHIC STRUCTURE. There was a total of 16 populations of Bromelia pin- guin in the study site, eight in the old selva, and eight in the young selva. The number of individuals per population varied between one and 1096. There were 2235 plants (77.12%) in the young selva subpopulations and 663 plants (22.88%) in the old selva subpopulations (Table 1). Old and voung selva sub- populations had the same size-structine pattern; few plants in the smaller classes, higher frequency in the middle classes, and few in the largest size classes (Figures 1 and 2). Plants could be assigned to three biological stages: non-reproductive (infan- tiles), clonally reproductive (juveniles), and sexually reproductive (adults). 30 Tulane Studies in Zoology and Botany [Vol. 30 3 a. o o. ^ Si ^ i£ n s: 3 T3 o -0 -r <■% u c ,o HH '*^ -a > X! ^-0 • ^ J2 ^ y ^ Si o HJ o bCo S V ■3 a, c .- (U 01 s '^ U CO Qh II 1- Tf CM c/5 c« _aj "s c ■o ■^ti c '> e CO ^ ^ S Of g CO °5 ti -^^ t« Tf c«? C« (M uji (« ^ 0 ^ ^^ crt Z ii • < '-' -^ > c« i« 0 "s +IZ < b Ji ^ j; c ~ J^ "■ ^— ' >-; .3 bb? _aj 3 — ■ ^ QJ 5 N ■ — 'K (t l+H "3 C« ^ U U. — 0 c a. c« II (U * S ^ r- J ^ 0 *^ W '3 -^ U ! ^ .^ caj a. * in « 00 i^ en tt o ■^ ir. cc ^— * CM 00 in o o X to CO 5 00 +1 CM 00 X +1 X CO +1 in in r^ tT X en in in O X CO J> --I o o ^-' in c^( oi o ^^ ^ ^ -^ ^-- CM Tf CO CM CO CM in CO in o CM CO X +1 CM +1 X en +i C^( in CM +i o in X «o to in CO (O in in CM CO t^ 1-1 o — ' --^ o in r^ to ^-' in ^^ ^ CO to t^ t^ X in X CO X X en o rr ^H o o to '^ CO +1 CM +1 CO +1 ?? +1 +i CT) -* o o o r- O) en ^ CO in o ^ CO in r^ CM in -* oincMin^CMCMr^ CMCOCTl^-^^^CM^^^-- ^-^ CM ^-^ ^^ o en C^J m in o o CO in CO en CM +1 m CO Tl CD CM +1 o 1 en CM +1 C^i +1 CO in +1 to o CO CO C^l CO CO in ^ CO X r^ in CO CO o CO ^ to CM X CO CO -^ CO l>l CM X ^ ^ to l> en CT) — en Tf X to -^ ^ '^ — ^^ cn — ^^ — ^ CM in o ■— CO ^H X ^^ — ' ^^ •^^ ■ ' C^I to o t~- en o m in o o X o CO +1 X +1 CO in X CO +1 m en CO +1 o o CO in CM o CO X CO CM in en CM ^COCMCMXXinCM X ^ o to o CTj r~-' to 00 en ■*' CM CO tT —I CO +1 +1 +1 +1 +1 ^ to to o ^ X J> CM CM ^ +1 +1 in CT! X c^j _ _ - CO CM CO CO C^l CO CO .. +1 CO CT) o o r m o -4 W O z .C — c^icoTt-mtot^x X. OTO— 'CMcOTfinto Q o in CO in CM +1 o to in '■^ iri CM +1 m CM +1 in to 00 CO +1 X to C^l 1994] Population Structure of Bromelia pinguin in Veracruz 31 Table 2. Number of mother plants and number of stolons produced in populations of Bronu'lia /yiiiffuin with more than 10 clones. Area occupied, and density oi plants and stolons ol populations \siih similar total niunber of plants (*ANOVA F=8.66, Fqooi (8io40)=3.27;Tukey test qo.ooi, 1040, 9=^-^^3). Plant Stolon Total Mother Stolons per Density Density Sub- Mother Plants with Mother Plant (Plants per (Stolons population Plants Stolons (%) x±s.d. Area (.sq. m) scj.m) pel" scj. m) Young Selva 13 1096 410 (37.0) 1.185±0.48 1849.91 0.59 0.22 6 586 201 (34.3) 1.149±0.39 939.85 0.62 0.21 5 290 97 (33. .5) 1.072+0.26 638.30 0.45 0.15 3 163 58 (35.6) 1.224±().49 223.56 0.73 0.26 4 72 23 (31.9) 1.174±0.39 128.92 0.56 0.18 Old Selva 10 388 148 (38.5) 1.169±0.47 512.48 0.75 0.29 11* 163* 71 (43.6)* 1.690±0.89 128.01 1 .27 0.55 14 46 19 (41.3) 1.053+0.23 36.16 1.27 0.53 9 42 22 (52.4) 1.136±0.35 33.26 1.26 0.66 Infantiles were present in the smaller leaf-size classes (89.6-137.3 cm), adults were present in the intermediate leaf-size classes (125.0-185.0 cm), and Juveniles were present over a wider range of leaf-size classes (119.9-185.0 cm) (Figures 1 and 2; Table 1). The young and old selva subpopulations were significantly dif- ferent in mean size of all individuals and in mean size within the biological stages (One Way ANOVA, p<0.001; Table 1). Mean leaf size of all individuals in young selva populadons (Kruskal-Wallis, H=125.65 p<0.05) were significaiuh differ- ent, and old selva populadons (Kiuskal-Wallis, H=9.52 p<0.05) were also signifi- cantly different. Individuals with sexual reproducdon can be separated into two subgroups: those with sexual reproduction as well as clonal growth, and those with only sex- ual reproduction. Mean leaf size and inflorescence size of individuals with only sexual reproduction were smaller than for individuals with both sexual and clonal reproduction (t-Student; t=6.19 d.f.=73, and t=4.23 d.f.=73, p<0.05, respectively). Mean size of individuals within populations in young selva was significantly different (One Way ANOVA, F=41.15, p<0.05). On the contrary, there was no dif- ference in mean size of individuals in old selva (One Way ANOVA, F=2.15, p>0.05). Mean size of individuals for each biological stage surveyed (infant, juve- nile, adult) for populadons in young selva was significantly different (One Way ANOVA, F=29.25, F=32.19, and F= 13.46, p<0.05, respectively). In old selva, the mean size of individuals was significantly different onlv for juveniles and adults (One Way ANOVA, F=13.20,' F=3.60, p<0.05, respectively), and not among infants (F=l. 5, p>0. 05). Subpopulations with similar total numbers of individuals in young and in old selvas were compared. For example, young selva population number 5 had a larger mean size for all individuals, juveniles, matures, and adults than did old selva populadon number 10 (t-Student; t=6.93, d.f.=289; t=3.69, d.f.= 172; t=4.62. 32 Tulane Studies in Zoology and Botany [Vol. 30 d.f.=29; and t=9.30, d.f.=96, p<0.05; respectively). On the other hand, sizes of individuals in old selva population number 11 were larger then in young selva population number 3 (t-Student; t=5.79, d.f.=162; t=4.18 d.f.=91; t=2.52, d.f.=12; t=4.82, d.f.=57; p<0.05; in the same order). CLONAL GROWTH. Individuals of Bromelia pinguin produce ramets before flow- ering, usually by the time they reach a mean leaf length of 151.4 (±25.65) cm (n=1060). Clonal growth starts in the axillary buds of basal rosette leaves nearest to the ground. The stolons, which are covered by bracts, grow plagiotropically or diagonally upwards over the mother plant. The first leaves are produced when the stolons reach a mean length of 78.2 (±18.33) cm (n=35). Mother plants can abort stolon development at any time prior to stolon leaf production (five cases were observed). Abortion is evidently not related to insect attack or disease. Mother plants can produce new stolons at a later time. Stem production and growth occur throughout the year but more frequently during the rainy season, when water availability and temperature are higher. Sto- lons grow faster during this period because of higher precipitation (Figure 2) (r=0.48, p<0.05; Spearman rank correlation coefficient). There was no statistical correlation with temperature. Growth of rosettes was significantly correlated with temperature (r=0.54, p<0.05) and with precipitation (r=0.47, p<0.05). Rosettes grow rapidly and can produce roots when a mean size of 102.17 (±14.76) cm (n=28) is attained. Such clones can produce new stolons before they flower, such that at times as many as three generations of clones are inter- connected, corresponding to the "guerrilla" type of reproduction (Lovett-Doust, 1981). The size of the rosettes that produce new stolons (117.6 (±5.73) cm, n=106) is very similar to the size of juveniles within the whole population at the study site (Table 1). Mother plants of Bromelia pinguin produce one to five clones during their life- time, but 1.20 (±0.49) is the average number of stolons per adult (n=1060). However, mother plants of subpopulation 1 1 produced a higher number of sto- lons per adult ramet (Table 2). Two pairs of populations with similar numbers of individuals were compared. Populations 9 (n=42) and 14 (n=46) were not differ- ent in mean number of stolons (p>0.1; Table 2), but populations 3 and 11, with equal numbers of individuals (163), were significantly different in the number of stolons produced (p<0.001; Table 2). Discussion Bromelia pinguin inhabits old and yoimg areas of deciduous forest in the study site, but the young selva populations have many more ramets and cover a greater area than do populations in old selva (Figures 1 and 2; Tables 1 and 2). Environ- mental conditions in young and old areas are not homogeneous. Bromeliad populations in young selva differ significantly in mean size when all plants are pooled or when categorized by life stages. In contrast, pooled individuals from populations in old selva differ significantly in size only when compared by life stages. Individuals in one area of the forest were larger than individuals in other areas, without a clear dominance pattern between areas. This suggests that soil characteristics and size of canopy gaps could be a factor determining ramet development within populations. 1994] Population Structure of Bromelia pinguin in Veracruz 33 20 number oi individuals 15- 10- 5- all subpopulatlon iDfantlles ClZi adultf ^^M juveniles I 20 40 60 80 100 120 140 160 180 200 220 240 a) 20 40 60 80 100 120 140 160 180 200 220 240 leal size class (cm) b) Figure 1. Leaf size class distriliution for individuals of Bwtnrlid piui^iiiii liom iwo |)()|)ulalions in young selva at La Mancha Station, Mexico. Data from Table 1: a = population nuiiilKM 1; h = po|>u!aiion number 5. 34 Tulane Studies in Zoology and Botany [Vol. 30 100 number of individuals 80- 60 40 20- all subpopulatlon Infantlles EH] adults ^^ juveniles 11 I m I ^ ■I i i i .1 a) 350 300- 250- 200- 150 100 60- 20 40 60 80 100 120 140 160 180 200 220 240 20 40 60 80 100 120 140 160 leaf size class (cm) 180 200 220 240 b) Figure 2. Leaf size class distribution for individuals of Bromelia pinguin from two populations in old selva at La Mancha Station, Mexico. Data from Table 1: a = population number 10; b = population num- ber 13. 1994] Population Structure of Bromelia pinguin in Veracruz 35 Each Bromelia piyiguin population is formed by adult and yoinig individuals that reproduce both sexually and asexually (by clones). Hij^h seed and seedling predation means that recruitment is based on clone formation (Garc fa-Franco et al., 1991). More than 30% of the ramets in each jjojiulation were mother jilants (Table 2). Nevertheless, the percent of mother plants in old selva populations was significantly higher than in young selva (t-Student; t=3.41, p<0.()r), percent values transformed to arcsin; Sokal and Rohlf, 1969). This difference suggests that conditions in old selva are more favorable for clonal reproduction. The mean number of stolons in most populations is slightly higher that one, suggest- ing that replacement is only of the mother plant. There is strong seasonality in availability of water and nutrients in the study site (Kellman and Roulet, 1990; Blain and Kellman, 1991). Biomass increases during the rainy season (Kellman, 1990); humidity strongly affects clonal growth of rosettes and stolons. Clonal growth dining wetter periods allows each ramet to cover a greater area with increasing light reception. Integration (-a plnsical and physiological interdependence among the ramets of a clone, sensu Jackson et al., 1985) can be particularly beneficial in environments where resources are scarce or are distributed over space and time, such as on sand dimes or rocks (Grime, 1979). Integration may well inhibit the establishment of other plant spe- cies within populations, such as in colonies of the terrestrial bromeliad Arhmea magdalenae (Andre) Andre ex Baker (Brokaw, 1983). growth (cm)/temperature (oC) precipitation (mm) 350 Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul I 1989 i 1990 I time Figure 3. Growth oi Bromelia pinguin at La Mancha Station, Mexico, relative to monthly temperature and precipitation. r=growth of rosettes in cm. s=growtli in leiiirth of stolons in cm. t=icmp(Tature. p=precipitation. 36 Tulane Studies in Zoology and Botany [Vol. 30 Individuals of Bromelia pinguin usually begin reproduction by clonal growth and reproduce sexually after or during stolon development. Some individuals reproduce sexually before producing stolons, and our data suggest that individu- als reproducing both clonally and sexually are more vigorous, larger, and pro- duce larger inflorescences. Clonal structure apparently benefits these plants by making more resources available for sexual reproduction. Vegetative reprodtiction can be advantageous for plants inhabiting environ- ments of great stress, or when ramets outlive seedlings (Fenner, 1985). The abundance of crabs and their effects on seed predation and dispersal can influ- ence the structtire of plant communities (Louda and Zedler, 1985; O'Dowd and Lake, 1990). Also, selective seedling herbivory has been shown to occur (O'Dowd and Lake 1990). Seed predation, mainly by Gecarcinus lateralis, affects diversity and abundance of species such as Brosimum alicastrum, Enterolobium cyclo- carpiim, and Cedrela odorata in deciduous forest at our study site (Blain, 1988; Del- fosee, 1990; Garcia-Franco et al., 1991). The similarity in the structure of Bromelia pinguin populations suggests that they likewise may be stibject to the same seed and seedling predation pressures at La Mancha. Acknowledgments We are grateful to personnel at the La Mancha Coastal Research Center for their hospitality and for providing climatic data. We also thank two anonymous reviewers for their comments and suggestions. Literature Cited Benzing, D. H. 1980. The Biology of Bromeliads. Mad River Press, Eureka. 305 pp. Blain, D. 1988. Factors affecting the early stages of regeneration of three tropical tree species in a sea- sonal forest, Veracruz, Mexico. M. Sc. thesis. York University, Ontario, Canada. 155 pp. Blain, D. and M. Kellman. 1991. The effect of water supply on tree seed germination and seedling sur- vival in a tropical seasonal forest in Veracruz, Mexico. J. Trop. Ecol. 7: 69-83. Brokaw, N. V. L. 1983. Groundlayer dominance and apparent inhibition of tree regeneration by Aech- mea magdalenae (Bromeliaceae) in a tropical forest. Trop. Ecol. 24: 194-200. Delfosee, B. 1990. The effect of the red land crab Gecarcinus lateralis (Frem.) on the litter layer nutrient availability and seedling recruitment in a coastal semi-deciduous seasonal tropical forest. M. Sc. the- sis. York University, Ontario, Canada. 176 pp. Fenner, M. 1985. Seed Ecology. Chapman and Hall, London. 151 pp. GARciA-FR.\NCO, J. G., V. RicO-Gray, and O. Zayas. 1991. Seed and seedling predation of Bromelia pin- guin L. by the red land crab Gecarcinus lateralis ¥iem. in Veracruz, Mexico. Biotropica 23: 96-97. Grime, J. P. 1979. Plant Strategies and Vegetadon Processes. John Wilev and Sons. Chichester, New York. 222 pp. Hallw.^cHS, W. 1983. Bromelia pinguin and B. karatas. In: D. H.J.4NZEN (ed.), Costa Rican Natural His- tory. The University of Chicago Press, Chicago, Illinois, pp. 195-197. Jackson, J. B. C, L. W. Buss, and R. E. Cook. 1985. Populadon Biology and Evolution of Clonal Organ- isms. Yale University Press. New Haven and London. 530 pp. Kellman, M. 1990. Root proliferation in recent and weathered sandy soils from Veracruz, Mexico. J. Trop. Ecol. 6: 355-370. Kellman, M. and N. Roulet. 1990. Nutrient flux and retention in a tropical sand-dune succession. J. Ecol. 78: 664-676. Louda, S. M. and P. H. Zedler. 1985. Predation in insular dynamics: an experimental assessment of postdispersal fruit and seedling sumval, Enewetak Atoll, Marshall Islands. Amer.J. Bot. 72: 438-445. Lovett-Doust, L. 1981. Population dynamics and local specialization in a clonal perennial {Ranunculus repens). \. The dynamics of ramets in contrasting habitats. J. Ecol. 69: 743-755. 1994] Population Structure of Bromelia pinguin in Veracruz 37 Moreno-CasASOIA, p. 1982. Ecolo^fa dc la xt'getacion df dunas rosteras: factores fi'sicos. Biotica 7: 577-602. Moreno-Casasola, p., E. v.\n der Maarel, S. Castillo, M. L. Huesca, and I. Pisants'. 1982. Ecologi'a de la vegetacion de duna.s costeras: Estructura v compo.sicion en el Mono de I.a Vlaiu lia. \ci, I. liioi- ica 7: 491-526. NOVELO, R. a. 1978. La vegetacion de la Estacion Biologica El Mono de La Mancha, Veracruz. Biotica .'?: 9-23. O'DOWT), D. L. .\ND P. S. Lake. 1990. Red crab.s in rain forest, Christmas Island: difierential herbivor\ ol seedlings. Oikos 58: 289-292. Ramirez, R. F. ANDJ. Garcia. 1990. Estructura y composicion flon'stica de la sei\a cosiera en La Mam lia, Veracruz. In: SOC.IED.A.D BOTANICA DE MEXICO, A. C. (ed.), \I Congreso Mexicano de Boianica, Pro- gramas y Resumenes. Oaxtepec, Morelos, pp. 188. Rico-Gray, V. hNV> A. Lot. 198.S. Produccion de hojarasca del manglar de la Laguna de La Mancha, Ver- acruz, Mexico. Biotica 3: 29.5-301. Smith, L. B. .and R.J. Downs. 1979. Flora Neotropica, Monograph 14, Part 3 (Bromelioideae). The New York Botanical Garden. New York. pp. 1665-1666. Soiv\i., R. R. .AND F.J. ROHLF. 1969. Biometria. H. Blume, Madrid. 832 pp. Z,\r, J. H. 1984. Biostadstical Analysis. Prendce-Hall, Englewood Cliffs, New Jersey. 718 pp. THE CYCLOPOID C:OPEPOD (CRUSTACEA) FAUNA OF NON-PLANKTONIC CONTINENTAL HABITATS IN LOUISIANA AND MISSISSIPPI Janet W. Reid Depart intent of Invnlebrdte Zoology, MRC-]63 National MuM'uin oj Xalinal History, Smithsonian Institution Washington, DC 20560 AND Ger.\ld G. Marten AVj/' Orh'dns Mosquito Control Board 6601 Lakes hore Drive, Neiu Orleans, L.A 70126 and Department of Tropical Medicine Tulane University, New Orleans, L\ 70112 Abstract Cyclopoid copepod crustaceans were collected in benthic and littoral sam- ples from spatially limited mosquito-breeding habitats including a canal, small perennial and ephemeral ponds, sloughs, swales, rice fields, ditches, puddles, cemetery pits, and old tires in southern Louisiana, principally in greater New Orleans, and in Mississippi. The local fauna is relatively diverse, with 25 species in this collection and a total of 34 recorded from the region. An imexpectedh' large neotropical faunal component included Apocyclops panamensis, Diacyclops bernardi, Mesocydops longisetus, Mesocyclops reidae, Thermocyclops inversus, and Thermocy clops tenuis. The known ranges of D. bernardi, M. reidae, and 7! inversus are extended to the United States. Metacyclops cushae, described from New Orleans, closely resembles neotropical species. Mesocyclops ruttneri, newly recorded from the United States, may have been introduced from tropical East Asia. Introduction During 1989-91, cyclopoid copepods (Crustacea) were collected from mos- quito breeding sites in southern Louisiana and northwestern Mississippi as part of a search for copepod species that would be potentially useful for biological control of larvae o{ Aedes albopictiis (Skuse) (Marten, 1989, 1990a-c). We list the species collected and compare the species richness of the Louisiana-Mississippi assemblage with cyclopoid faunas of other regions of North America, especiallv the south-central United States. We discuss the known geographical distribu- tions and habitats of species of particular interest. Materl\ls and Methods Copepods were collected throughout the year at ca. 200 sites in the New Orleans-Slidell area (soiUheastern Loviisiana). Habitats included canals, Spartina marshes, permanent and ephemeral ponds, ephemeral pools, and ditches. Sev- Tulane Studies in Zoologv' and Botan)' 30: 39-45. 1995. 39 40 Tulane Studies in Zoology and Botany [Vol. 30 eral hundred discarded tires and 40 tree holes were also sampled. Copepods were collected from 32 rice fields in Jefferson Davis Parish (southwestern Louisi- ana) and 10 rice fields in the vicinity of Cleveland (northwestern Mississippi). The copepods were collected with a standard dipper used to collect mosquito larvae. The dipper consisted of a 400 ml plastic cup at the end of a one-meter wooden pole. Copepods were also collected from canals and ponds by dragging a plankton net parallel to the shore. They were collected from tires by removing the water with a hand-held plastic cup, and from tree holes by sucking out the water with a turkey hasten Water dipped from each site was passed through a 200 jim Nitex net to strain out copepods. The contents of the net were suspended in a small amount of water, and copepods were removed with a Pasteur pipette and fixed and preserved in 70% ethanol. For species determination, specimens were mounted temporarily in glycerin or lactic acid media, or permanently in com- mercial poly\'inyl lactophenol medium with a little chlorazol black E added, and examined using a Wild M20 compound microscope. Representative specimens were deposited in the National Museum of Natural History, Smithsonian histitu- tion; the remaining material is in the collection of COM. Results Species inventory and nomenclature. The collections contained a total of 25 cyclopoid species (Table 1). Identifi- cation of most species was straightforward, but explanation is necessary for some of the names employed in this article. We use the old name Eucyclops elegans for the American species that is frequently identified as the European Eucyclops sper- atus (Lilljeborg, 1901), and have listed records of £. speratus by other authors as E. elegans in Table 1. The systematics of Megacyclops is presently confused; speci- mens from New Orleans, reported as ''Megacyclops viridis species group" by Mar- ten (1990b) were eventually identified as M. latipes according to unpublished criteria developed by A. A. Weaver (in litt. toJWR, 1993). Most of the specimens of Mesocyclops longisetus were congruent with M. longisetus subspecies curvatus Dussart, 1987. Mesocyclops ruttnen ■w^s identified with some difficulty because of its resemblance to several Caribbean species, and is redescribed elsewhere (Reid, 1993). Metacyclops cushae \vas earlier listed as M. denticulatus by Marten (1989). Microcy clops rubellus (Lilljeborg, 1901) is considered by Reid (1992) and other authors as a separate species rather than a subspecies of Microcyclops vari- cans (G. O. Sars, 1863). The species of Tropocyclops encountered in this study fit the current diagnoses of T. prasinus sens. str. or T. extensus, as presented by Dus- sart and Fernando (1990) and Reid (1991), but not that of the commonly reported T. prasinus mexicanus (Kiefer, 1938). Species reported from several other general collections in the region are listed in Table 1 under their currently recognized generic names. In the case of the very old records listed by Penn (1947), the species are listed under our best guess as to the taxa meant: Cyclops brevispinosus Herrick, 1884, is possibly a syn- onym of Acanthocyclops vernalis; Cyclops serrulatus Fischer, 1851, is a synonym of E. agilis; and Cyclops fluviatilis Herrick, 1882, is a synonym of Tropocyclops prasinus. Diacyclops navus was reported by Nasci et al. (1987) as Thermocyclops dyboxoskii (Loven, 1890); this record was corrected by Reid et al. (1989). 1994] Cyclopoid Copepods in Louisiana and Mississippi 41 Table 1. Copepoda CXclopoida species records from five sur\eys in the south-central I'niied States. Locales are as follows: present report, southern Louisiana and northwestern Mississi|>pi; l\nn (1947), New Orleans; Nasci et al. (1987), southwestern Louisiana; Harris (1978), northern Missis- sippi; Binford (1978), southern Louisiana. Present Penn Nasci et al. Harris Bintord Suidv (1947) (1987) (1978) (1978) COSMOPOLITAN SPECIES Ectocyclops rubesftns Erady, 1904 Eucyclops agilis (Koch, 1838) Paracydops chiltoui (Thomson, 1883) Paracydops fimbriatus (Fischer, 1853) Macrocydops albidus (Jurine, 1820) Marrocydops fiiscus (Jurine, 1820) NORTH AMERICAN OR CIRCUMBORE.\L SPECIES Halicydnps sp. Eucyclops agilis monlaiuis (Brady, 1878) Eucyclops prionophorus Kiefer, 1931 Eucyclops elegans (Herrick, 1884) Megacydops latipes (Lowndes, 1927) Megacydops luriiiis (Jurine, 1820) Orthocydops modestus (Herrick, 1883) Paracydops poppei (Rehberg, 1880) Tropocydops fxlmsus (Kiefer, 1931) Tropocydops prasinus (Fischer, 1860) Tropocy clops prasinus mexicanus Kieier, 1938 Acanthocydops exilis (Coker, 1934) Acanthocydops venustoides (Coker, 1934) Acanthocydops vernalis (Fischer, 1853) s.l. Diacydops crassicaudis \2ir. brachycercus (Kiefer, 1929) Diacydops navus (Herrick, 1882) Diacydops thomasi (S.A. Forbes, 1882) Mesocydops edax (S.A. Forbes, 1891) Metacydops cushaeReid, 1991 Microcydops rubellus (Lilljeborg, 1901) NORTH AND SOUTH AMERICAN SPECIES Homocyclops ater (Herrick, 1882) NEOTROPICAL SPECIES Apocyclops panamensis (Marsh, 1913) Diacydops bernardi (Petkovski, 1986) Mesocydops longisetus (Thiebaud, 1914) s.l. Mesocydops reidae?etko\?,V.\, 1986 Thermocydops inversus Ydefer, 1936 Therrnocydops tenuis (Marsh, 1909) ASIAN SPECIES [?] Mesocydops ruttneri Kieier, 1981 SPECIES TOT.^ 25 16 17 42 Tulane Studies in Zoology and Botany [Vol. 30 Species-habitat relationships. Large permanent water bodies (canals, golf course ponds) in the New Orleans-Slidell area contained Acanthocyclops vernalis, Eucyclops agilis, Eucyclops elegans, Homocyclops ater, Macrocyclops albidus, Mesocyclops ruttneri, Microcyclops rubel- lus, and Tropocyclops prasinus. Nine species were common in the New Orleans-Slidell area in marshes, roadside ditches, temporary ponds, and tempo- rary pools holding water for extended periods: Acanthocyclops vernalis, Diacyclops navus, Ectocyclops rubescens, Eucyclops agilis, Eucyclops elegans, Macrocyclops albidus, M. rubellus, Thermocy clops inversus, and Tropocyclops extensus. In addition to these, pools in wooded areas contained Megacyclops latipes and Orthocyclops modestus. Only one tree hole contained copepods, and they were Paracyclops poppei. Dis- carded tires in wooded areas contained some of the same species found in the surrounding woods, namely A. vernalis, D. navus, E. rubescens, E. agilis, M. albidus, P. poppei, and O. modestus. Eucyclops elegans, P. chiltoni, and T. prasinus were each collected from a single tire, and a single specimen of M. ruttneri was collected from a marsh. In the New Orleans-Slidell area, pools and small depressions (swales) hold- ing water only for short periods of time contained Acanthocyclops vernalis, Diacy- clops navus, Metacyclops cushae, Microcyclops rubellus, and Thermocyclops tenuis. These species often appeared as soon as there was water. Diacyclops crassicaudis var. brachycercus and Diacyclops bernardi were also collected, but each was fotmd in only one swale. Apocyclops panamensis was found in a freshwater swale that was sometimes flooded with brackish water. The Louisiana rice fields contained Acanthocyclops vernalis, Ectocyclops rubescens, Eucyclops elegans, Macrocyclops albidus, Mesocyclops ruttneri, Mesocyclops reidae, Microcyclops rubellus, Paracyclops chiltoni, Paracyclops poppei, Thermocyclops inversus, T tenuis, and Tropocyclops extensus. The Mississippi rice fields contained A. vernalis, Diacyclops crassicaudis var. brachycercus, Eucyclops agilis, E. elegans, Macrocyclops albidus, Mesocyclops ruttneri, M. reidae, Micro- cyclops rubellus, and Tropocyclops extensus. Either Acanthocyclops vernalis or Mesocy- clops ruttneri occurred in almost every field that was sampled in Louisiana and Mississippi, but never together. Discussion These collections, although limited primarily to small bodies of standing or very slowly flowing water, revealed a species-rich cyclopoid fauna (Table 1 ) . Pre- vious studies in the region collected many fewer species on average. Only six spe- cies were recorded previously from New Orleans proper (Penn, 1947). Binford (1978) collected seven species of cyclopoids during monthly samples over a one-year period in the Atchafalaya River and associated swamps in south-central Louisiana. Harris (1978) listed 17 species from collections in northern Missis- sippi made over a one-year period from a wide variety of lentic and lotic habitats. Nasci et al. (1987) recorded 16 species from similar mosquito-breeding habitats in southwestern Louisiana. The fauna of ephemeral pools reported here is like that occurring in similar waters elsewhere. Nasci et al. (1987) reported that Acanthocyclops vernalis was almost constantly present in three woodland ponds in southwestern Louisiana. In those ponds, two of which frequently dried, Diacyclops navus "was collected less often but also persisted through the year. Nasci et al. (1987) reported a total of 1994] Cyclopoid Copepods in Louisiana and Mississippi 43 6-11 species from each pond. In New Orleans, similar woodland ponds con- tained a total of 1 1 species, most of these the same that were recorded bv Nasci et al. (1987). Taylor and Mahoney (1990) reported an analogous association of Acanthocyclops vernalis and Diacyclops haueri Kiefer, 1931, in an ephemeral bay lake in South Carolina, which supported an assemblage of nine cyclopoid spe- cies. Some 96 species of cyclopoid copepods of the family CVclopidae are pres- ently recorded from North America north of Mexico. Usually, about one-fourth of this nimiber inhabit any particular region. The 34 copepod taxa now reported from Louisiana and Mississippi constittue the most species-rich assemblage so far recorded from North America. Although such comparisons are not well foimded because the few intensive investigations of regional cyclopoid copepod faunas on this continent differed in their objectives and collecting techniques, a review of some other local totals provides perspective. One of the most compre- hensive fatmistic studies in the United States is the eight-year survey reported by Bruiting (1973) for Tennessee, in which 22 species were recorded from 653 sam- ples from 476 localities including a wide variety of habitats. The copepod faima of North Carolina is almost equally well known, thanks to the investigations of Yeatman (summarized in 1944) and subsequent workers. From North Carolina there are 27 species recorded to date (Adams et al., 1992). Smith and Fernando (1978) reported 25 species from lacustrine and non-lacustrine habitats in south- ern Ontario. In tropical regions of comparable extent, cyclopoid copepod spe- cies diversity may be slightly higher. For instance, from Cuba, 35 species and subspecies of cyclopids have been reported, although several records need to be substantiated (Reid, 1990). The 25 species recorded in our collections is a higher total than would be expected from the limited types of habitats sampled. Even though the local list may be slightly inflated because of taxonomic prob- lems, the area total of 34 taxa is surprising. Characterization of the geographical distribution of the cyclopoid species collected in Louisiana and Mississippi (Table 1) reveals that the relatively high number of species is due mainly to a significant neotropical component. South- ern Louisiana lies within the northern ranges of six neotropical species, three of these doctmiented from the United States for the first time in this study. A sev- enth, previously undescribed species with neotropical affinities {Metacyclops cushae) was discovered in New Orleans, and an eighth, probably introduced spe- cies {Mesocyclops ruttneri) is recorded for the first time in the Americas. Apocyclops pancnnensis is a common inhabitant of brackish coastal lagoons and ponds around the Gulf and (Caribbean, and has been recorded from the mid-Atlantic coast of the United States, south as far as the northern coast of South America (Reid, 1990). Mesocyclops longisetus occurs commonly from south- ern South America to the Antilles and Mexico, and sporadically in the southern conterminous United States, where the New Orleans find is only the third con- firmed record, the others being from Texas and Florida (reviewed bv Reid, 1993). The subspecies M. longisetus curvatus may be distributed through Central America, the Andlles, and southern North America, with the nominate species restricted to continental South America (Dussart, 1987). Mesocyclops reidae, an Antillean and Central American species, is common in small bodies of water in Honduras (Marten and Reid, unpublished). The record from Mississippi rice fields extends its known range significantlv northward from tropical Mexico (Reid, 1990, 1993). 44 Tulane Studies in Zoology and Botany [Vol. 30 Thermocyclops tenuis extends throughout the lowland neotropics, with spo- radic records in the southwestern and south-central United States. Many records of this species are from ephemeral waters (Reid, 1989). The record of Thermocyclops inversus was previously reported briefly by Reid (1989). This species is common in southern Mexico, Central America, and Cuba, with a few records from Brazil, but was unknown north of the Tropic of Cancer imtil the New Orleans find. The Louisiana record thus represents a northward range extension of about 900 km. Thermocyclops inversus is a common plankter in large Central American lakes, but has also been recorded from reser- voirs, ponds, wells, and caves (Reid, 1989). The find of Diacyclops bernardi represents an equally surprising northward range extension. It is known elsewhere only from San Andres Island off Nicara- gua, and from Yucatan, Mexico (Reid, 1993). The species was collected from brackish water (salinity up to 30 %o) in a dug well and a shallow cave on San Andres, and in Mexico from a coastal pond. Metacyclops cushae, described from New Orleans, resembles an Argentinian species, M. denticulatus Dussart and Frutos, 1986, and is most likely neotropical in origin. This is the first confirmed published report of a species of Metacyclops in North America, althotigh the genus is widespread in tropical and temperate regions of other continents including Europe and Asia (Reid, 1991). Metacyclops cushae was fotuid at only three localities in greater New Orleans, two grassy sloughs and one ditch, all containing water only seasonally. The eighth species of probable tropical origin is Mesocyclops ruftneri, which is newly recorded for North America in this study. Mesocyclops ruttneri\^2L% originally described from a greenhouse in Austria, now destroyed; no wild populations are known to exist in Europe (Kiefer, 1981). Specimens of M. ruttueri were recently identified by Reid (1993) in samples from wild populations in China, Thailand, and Viet Nam, and the species is morphologically similar to several Asian conge- ners. It seems likely that M. ruttneri was introduced from Asia into both Austria and the southern United States. Specimens of M. ruttneri were collected in abun- dance from a permanent pond-like canal in New Orleans, and in rice fields in Louisiana and Mississippi. The species is obviously well established in the region. This study confirms the supposition of Reid (1992) that the neotropical cyclopoid copepod faunal component in the southern United States is more sig- nificant than previously estimated. However, most of the cyclopoids of neotropi- cal affinities present in southern Louisiana and Mississippi may be near the northern limit of their ranges. Most have been collected only sporadically, and are probably existing alongside rather than directly competing with members of the North American fauna. The only suggestion of competition in the available collection data is the apparent mutual exclusion of Acanthocyclops vernalis and Mesocyclops ruttneri in rice fields. Acknowledgments We wish to thank Edgar Bordes, Mieu Nguyen, and other members of the staff of the New Orleans Mosquito Control Board for their contributions to this study. We thank Drs. Harry C. Yeatman and Andrew A. Weaver for valuable advice on proper identification of Megacyclops. Gerald Marten was supported through a senior research associateship from the National Research Council. The Division of Vector-Borne Infectious Diseases, Centers for Disease Control, provided financial support. 1994] Cyclopoicl Cx)pcp()cls in Louisiana and Mississippi 45 LlTER.\TL'RE CiTED Adams, W. F., J. M. Alderman, D.J. DeMont, J. W. Reid, A. Y. T.\n.OR, and D.J. Wii.i.i.vms. 1992. A Report on the Consenation Status of North Carolina's Freshwater and Terrestrial Crustacean Faiuia. Submitted to the Division of Wildlife Management, North Carolina Wildlife Resouices Commission, Raleigh, N.C., by the Scientific Council on Freshwater and Terrestrial Crustacea, December 1992. BiNEORD, M. W. 1978. Copepoda and Cladocera commimities in a ri\er-swamp svsiem. \'erh. Internal. Verein. Limnol. 20: 2524-2."i3(). Bunting, D. L. 1973. The Cladocera and Ck)pepoda of Tennessee II. Cyclopoid copepods. J. Tennessee Acad. Sci. 48(4): 138-141. DUSSART, B. H. 1987. Sur quelques Mesocyclops (Crustacea, Copepoda) d'Americpu' du Sud. Ania/oniana 10(2): 149-161. DUSS.\RT, B. H. .\ND C. H. Fern.a.ND(). 1990. A review of the taxonomy of Hve Oniaiio genera of freshwa- ter cyclopoid Copepoda (Crustacea). Canadian J. Zool. 68: 2594-2604. Harris, M.J. 1978. Copepoda of northern Mississippi with a description of a new subspecies. Tulane Stud. Zool. Bot. 20: 27-34. KlEFER, F. 1981. Beitrag zur Kenntnis von Morphologic, Taxonomie und geographischer Verbreitimg von Mesocyclops leuckarti auctorum. Arch. Hydrobiol. Suppl. 62 (Monogr Beitr.) 1: 148-190. ]VL\RTEN, G. G. 1989. A suiTev of cvclopoid copepods for control ot Aedes rt//;o/;/r/!/ilanae. Bull. Soc. Vec- tor Ecol. 14: 232-236. Marten, G. G. 1990a. Issues in the development of Cyclops for mosquito control. Arbovirus Research in Australia 5: 159-164. Marten, G. G. 1990b. Evaluadon of cyclopoid copepods for Aedes alhopictus control in tires. J. ,\mer. Mosquito Control Assoc. 6: 681-688. M.\RTEN, G. G. 1990c. Elimination oi Aedes albopictus horn die piles by introducing Macrocyclops albidus (Copepoda, Cyclopidae). J. Amer. Mosquito Control Assoc. 6: 689-693. Nasci, R. S., S. G. F. Rare, and M. Vecchione. 1987. Habitat associaUons of mosquiK^ and copepod spe- cies. J. Amer. Mosquito Control Assoc. 3: 593-600. Penn, G. H. 1947. Branchiopoda and Copepoda of the New Orleans area as recorded by Ed Foster in the early 1900's. Proc. Louisiana Acad. Sci. 10: 189-193. Reid, J. W. 1989. The distribution of species of the genus Thermocyclops ((Copepoda, (]yclopoida) in the western hemisphere, with description of T. parvus, new species. Hydrobiologia 175: 149-174. Reid, J. W. 1990. Condnental and coastal free-li\ing Copepoda (Crustacea) of Mexico, Central .\merica and the Caribbean Region. In D. NUARRO L. AND J. G. ROBINSON, eds. Diversidad Biologica en la Reserva de la Biosfera de Sian Ka'an, Quintana Roo, Mexico, pp. 175-213. Centro de Investigaciones de Quintana Roo (CIQRO) and Program of Studies in Tropical Consenation, Univ. Florida; Chetu- mal, Quintana Roo. Reid, J. W. 1991. The genus Metacyclops (Copepoda: Cyclopoida) present in North America: M. cushae, new species, from Louisiana. J. Crustac. Biol. 1 1 (4): 639-646. Reid, J. W. 1992. Copepoda (Crustacea) from fresh waters of the Florida Everglades, U.S.A., with a description of Eucyclops conrowaen. sp. Trans. Amer. Microsc. Soc. 1 1 1 (3): 229-254. Reid, J. W. 1993. New records and redescriptions of American species of A/«wvr^/^s and (){' Di/irxchps her- nardi (Petkovski, 1986) (Copepoda: Cyclopoida). Bijdr. Dierk. 63(3): 173-191. Reid, J. W., S. G. F. Hare, and R. S. Nasci. 1989. Diacyclops navus (Crustacea: Copepoda) redescribed from Louisiana, U.S.A. Trans. Amei". Microsc. Soc. 108(4): 332-344. Smith, K. .a.nd C. H. Fernando. 1978. A guide to the freshwater calanoid cUid cNclopoid copepod Crusta- cea of Ontario. LIniv. Waterloo Biol. Series 18: 1-74. T.\M.0R, B. E. and D. L. M.\H0NEV. 1990. Zooplankton in Rainbow Bav, a Carolina bav pond: population dynamics in a temporan' habitat. Freshwater Biol. 24: 597-612. Yeatman, H. C. 1944. American cyclopoid copepods of the viridis-vemalis gTOu\>, (including a descrip- tion of Cyclops carolinianus n. sp.). Amer. Midi. Nat. 32(1 ): 1-90. TULANE STUDIES IN ZOOLOGY A^p^ BOTANY LIBRARY f — ". — ' 1 7. !>',.'■ Volume 30, Number 2 '" * ^' '"^^ December 30, 199(5 HARVARD LIFE HISTORY ayPl^HEafPeSQCp^ARTER (PERCIDAE: ETHEOSTOMA RANEYl), A SPECIES ENDEMIC TO NORTH-CENTRAL MISSISSIPPI CAROL E.JOHNSTON AND WENDELL R. HAAG USDA Forest Sniiice, Southern Research Station, Forest Hydrology Laboratory, P.O. Box 947, Oxford, MS 38655 LEAF-MINER DEFENSES IN BROMELIA PINGUINL. (BROMELLVCEAE) IN VERACRUZ, MEXICO VICTOR RICO-GRAY, JOSE G. GARCIA-FRANCO Departamento de Ecologia Vegetal, Instituto deErologi'a, A.C., Apdo. 63, Xalapa, WR 91000, Mexico'^ ANGEL TRIGOS-LANDA Departamento de Qiiimicay Biologia, Universidad de la Americas, Cholula, PUE 72820, Mexico RACHEL MATA, AND PERLA CASTANEDA Departamento de Farmacia, Facultad de Quimica, U.N.A.M., Mexico, DF 04510, Mexico OBSERVATIONS ON SOME FUNGI FROM LOUISIANA AND MISSISSIPPI IN COMPARISON WITH THOSE OF MEXICO GASTON GUZMAN Instituto de Ecologia, Apartado Postal 63, Xalapa, Veracnxz 91000, Mexico SOME REALLOCATIONS OF TYPE LOCAIJTIES OF REPTILES AND AMPHIBIANS DESCRIBED FROM THE MAJOR STEPHEN H. LONG EXPEDITION TO THE ROCKY MOUNTAINS, WITH COMMENTS ON SOME OF THE STATEMENTS MADE IN THE ACCOUNT WRITTEN BY EDWIN JAMES HAROLD A. DUNDEE Tulane University Museum of Natural History, Belle Chasse, LA 70037-3098 Tulane University New Orleans TULANE STUDIES IN ZOOLOGY AND BOTANY ISSN 0082-6782 Department of Ecology, Evolution, and Organismal Biology Tulane University, New Orleans, Louisiana 70118 Steven P. Darwin, Editor INFORMATION FOR AUTHORS Manuscripts dealing with all aspects of ecology, evolution, and systematic biology are encouraged. All manuscripts are reviewed. Page charges are assessed at the rate of $45 per page. Some financial assistance from the George Meade Memorial Fund is available if authors are unable to meet these charges. Style of manuscript should follow guidelines appearing in the Council of Biology Editors Style Manual. Manuscript (with abstract) must be typed on good quality 8.5 x 1 1 inch paper, dou- ble-spaced throughout (including Literature Cited, figure captions, footnotes, and tables). All pages must have 1 inch margins on all sides. Dot-matrix printing is not acceptable; do not justify right-hand margins. Do not hyphenate words at ends of lines. Last name(s) of author(s) should appear on each page with page number; this includes tables and figure captions. Lettering used in figures must be easily read after reduction to our page size. Pho- tographs (halftones) should be made for no or little reduction. The original and three copies of the manuscript, along with figures, should be sub- mitted. Prospective authors should request a copy of the detailed guidelines for manu- script preparation (including footnotes, typefaces, literature and specimen cita- tion, and dimensions of figures) from the Editor at the above address. EXCHANGES AND STANDING ORDERS Tulane Studies in Zoology and Botany is available for exchange with institutions publishing comparable series. Otherwise, standing orders are accepted, with issues priced individually (foreign postage extra) . A price list for back issues is available on request. Remittances should be made payable to Tulane University. LIFE HISTORY OF THE YAZOO DARTER (PERCIDAE: ETHEOSTOMA llWEYf), A SPECIES ENDEMIC TO NORTH-CENTRAL MISSISSIPPI Carol E.Johnston AND Wendell R. Hxw, USDA Forest Srti'ia'. Southern Research Station, Forest Hydrology l.ahoralory, P.O. Box 947, Oxford, MS 38655 ' ' Abstract The Yazoo darter, Etheostoma raneyi, is restricted to tributaries of the Little Tal- lahatchie and Yocona Rivers (Yazoo River drainage) in north-central Mississippi. The species inhabits small, clear streams, many of which are spring-fed, and have a variety of substrate types including silt, clay, sand and gravel. In Morris Creek, a second-order stream in Lafayette County, Mississippi, E. raneyi was found in a diversity of microhabitats. Etheostoma raneyi is short-lived, with very few individu- als living more than two years. Yazoo darters are sexually dimorphic, with males larger and more brightly colored than females. Both sexes become reproduc- tively mature during the first year of life. The sex ratio of the Morris Creek pop- ulation is female-biased. Like other members of subgenus Natiostoma, Yazoo darters use the egg-attaching spawning strategy. Batch fecundity of Yazoo darters is 52, and the average diameter of ova in this size class is 1.05 nun. Based on gonadal condition, the spawning season lasts from March to Jime. Yazoo darters are the most abundant fish in Morris Creek. Introduction Etheostoma raneyi is one of eight species of subgenus Nanosloma (snubnose darters) described since 1991, bringing the number of described species to 20. Additional species await scientific description. Snubnose darters are relatively short-lived, small, sexually dichromatic fishes that use the egg-attaching spawn- ing strategy (Page, 1983). Most species inhabit flowing pools and riffles of small streams (Suttkus and Etnier, 1991). At least three species of subgenus Nanostoma are considered imperiled (Williams et al., 1989), due in part to their limited dis- tributions. The restricted range and lack of life-history information for many species of subgenus Nanostoma have concerned conservationists. The objective of this study is to provide life history informadon necessary for protection of the Yazoo darter and to compare with life histories of other species of Nanostoma. Most snubnose darters occur in southern tributaries to the Ohio River, C.ulf Coastal drainages east of the Mississippi River, and lower Mississijjpi iril)utaries (Suttkus and Etnier, 1991). Many are endemic to single drainages or have lim- ited geographical distributions (Boschung et al., 1992). The Yazoo darter {Etheo- stoma raneyi Suttkus and Bart) is endemic to small streams of the upper Yazoo River drainage, Mississippi (Figure 1). Suttkus et al. (1994) reported the species from 15 sites (five in the Yocona River and ten in the Little Tallahatchie), while Tulane Studies in Zoology and Botany .^0: 47-60. 1996 47 48 Tulane Studies in Zoology and Botany [Vol. 30 arw 8^00 I Hotly Springs Ripley a Mississippi Locator Map. I 89^30 89°00 Figure 1 . Distribution oiEtheostoma raneyi. Study site is indicated by a star (modified from Thompson andMuncy, 1985). in an unpublished status report Thompson and Muncy (1985) reported the Yazoo darter from 23 sites. Six sites listed by Suttkus et al. (1994) were not cited by Thompson and Muncy (1985), suggesting that the species may be found present at 29 sites. Study Site Morris Creek is a clear, second-order stream that drains forested and agricul- tural lands for approximately 6.5 km before draining into the Yocona River. The study area is 3.2 km northeast of Taylor, Mississippi (T9S, R3W, sec. 16, 19 and 20) (Figure 1). At the study site the stream averages 3 m in width and has pool, riffle, and run habitats, and abundant woody debris. Substrates include clay, silty sand, coarse sand, gravel, and cobble. A narrow riparian zone is present, but much of the surrounding land is used for row crops or pasture. Yazoo darters used for spawning observations were collected from two streams: Lee Creek and an unnamed tributary to Bay Springs Branch, both of the Little Tallahatchie drainage, Lafayette County, Mississippi. The tributary to Bay Springs Branch is 1996] Yazoo Darter Life HistoiT 49 within the University of Mississippi Biological Field Station (T7S, R2W, sec. 34); its average width is approximately 1.5 m, and the dominant substrate is sand. Woody debris and submerged aquatic vegetation (primarily Sparganium) are conmion. Lee Creek (T7S, R3W, sec. 1) averages 3 m in width and substrates include clay, silty sand, coarse sand, gravel, and cobble. Woody debris is com- mon, and much of the surrounding land is forested. Materials and Methods Yazoo darters were collected monthly from March 1993 to May 1994; two samples per month were made March-May 1994. All sampling was done using a three-meter-long minnow seine, and an effort was made to sample a wide variety of microhabitats. A general description of habitat type (run, riffle, pool, or undercut bank) was made for all seine hauls. At every location where a Yazoo darter was collected, the following habitat variables were measured: water depth and velocity (average of three measurements of each), substrate composition (percent composition by particle size, estimated by the modified Wentworth scale outlined in Ross et al., 1990), stream width, percent canopy cover, and the percent and make-up of instream cover. The identity and abundance of other species of fishes collected was recorded for all seine hauls. Exceptions are for species of Fundulus and Erimyzon that were not identified to species in the field. For the pinpose of Table 1, species of Lepomis and Micropterus were lumped. Specimens of Yazoo darters were preserved in 5% buffered formalin; other fishes were released. Homogeneity of distribiuion of darters among the five habitat types was eval- uated using R X C G-tests of independence (Sokal and Rohlf, 1981). The rela- tionship between darter density and habitat type was further explored using principal components analysis (PCA) with Varimax rotation, using the S\'STAT software package (Wilkinson, 1990). A scree test indicated the niunber of mean- ingful components. Standardized factor scores for all combinations of compo- nents were plotted and ntmibers of darters were overlaid on the factor score plots. This two-step procedure does not assume linearity between abundance of fishes and environmental variables and is appropriate for these data (Ross et al., 1987; Ross etal., 1990). Sex, standard length (SL), and age of preserved darters were determined in the laboratory. Standard length was measured to the nearest 0.1 nun using dial calipers. Age was determined to the nearest month by coimdng scale annuli. April was used as month zero, since it is near the middle of the spawning season. Scales for age analysis were removed above the lateral line and near the tip of the depressed pectoral fin. Seasonal changes in gonad mass for both sexes were quantified using gona- dosomatic index (GSI). Gonads and eviscerated specimens were dried at 55°C for 24 hours and weighed to the nearest 0.001 gm. Gonadosomatic index was cal- culated by dividing gonad mass by adjusted somatic mass (mass of evi.scerated specimen) and multiplying by 1000. The reproductive condition of males was classified as latent or mature by examination of testes. Latent testes were tiny strands of clear tissue; mature testes were enlarged and opaque. The reproduc- tive condition of females was determined by examination of ovaries and their classification into one of six developmental stages according to Heins and Baker (1993): latent (LA), early maturing (EM), late matining (LM), matiue (MA), ripening (MR), or ripe (RE). Five ova were counted and measured in each of the 50 Tulane Studies in Zoology and Botany [Vol. 30 Table 1. Relative abundance of occurrence of fishes in Morris Creek, Lafayette County, Mississippi, 1993-1994 (n=12 samples). Fishes were identified in the field and released; fishes in the genera Lepomis, Mirropteriis, Fundulus and Erimyzon are lumped for the purpose of this table (see text for explanation) . Etheostoma raneyi Cyprinella camura Percina sciera Fundulus spp. Noturus phaeus Pimephales notatus Notropis rafinesquei Etheostoma lynceum Lepomis spp. .207 .168 .102 .168 .079 .073 .057 .040 .040 Semotilus atromaculatus Erimyzon spp. Etheostoma whipplei Micropterus spp. Hypentelium nigricans Etheostoma proeliare Etheostoma parvipinne Etheostoma nigrum Pimephales vigilax Etheostoma sicaini .036 .021 .018 .012 .010 .010 .003 .002 .001 <.001 three largest size classes for 43 females from seven samples taken during the breeding season (March-June 1993; March-May 1994). These size classes were distinct, and separated by at least 0.30 mm of diameter. The diameter of each ovum was expressed as an average of the smallest and largest diameters to account for irregularity in shape (as typical for many darter species, the eggs were concave in shape). Meastuements were made using a dissecting microscope and ocular micrometer. Reproductive data were analyzed using the SYSTAT soft- ware package (Wilkinson, 1990). Spawning mode was determined by observation of breeding activities in aquaria. Reproductively active males (identified by bright spawning coloration) and females (identified by distended abdomens containing ripe eggs) were col- lected from a tributary to Bay Springs Branch (1993 and 1994) and Lee Creek (1995) and transported to the laboratory. Animals were placed in 84 or 840 liter aquaria and maintained at 19-21°C. Aquaria contained the following substrates known to be used for spawning by other darter species: logs, plants and boulders (used by egg-attachers); gravel and sand substrate (used by egg-buriers); and cavities (tised by egg-clumpers and egg-clusterers) (Page, 1985). Study animals were collected on 17 April 1993 (6 females, 3 males), 10 March 1994 (5 females, 4 males), and 25 March 1995 (2 females, 2 males); spawning ustially occurred within two days of capture. Observations were made several times a day for 10 days following introduction into the aquaria. Observation periods were from 15 minutes to 2 hours. Spawning events were recorded with an 8 mm video camera. A total of seven spawning pairs were observed. The spawning description is based on the observations and from reviewing video tapes of spawning episodes. Five eggs from each spawning event were measured to the nearest 0.01 mm with a dissecting microscope and ocular micrometer. Results Etheostoma raneyi'W3.s the most abtindant fish in Morris Creek (Table 1). Other common fishes included bltmtface shiner {Cyprinella camura), dusky darter {Percina sciera), topminnows {Fundulus notatus and Fundulus olivaceous) , brown madtom {Noturus phaeus), and bltnitnose minnow {Pimephales notatus). Other species of Etheostoma present in the stream included redfin darter {Etheostoma whipplei), cypress darter {Etheostoma proeliare), goldstripe darter {Etheostoma parvipinne) , johnny darter {Etheostoma nigrum), brighteye darter {Etheostoma lyn- ceum), and gulf darter {Etheostoma sioaini) (Table 1). 1996] Yazoo Darter Life History 51 Table 2. OccuriTnce of \';i/o(i darters (Ethcoslonin rain'yi) anioiij; five liahitat rv'pes in Morris Creek, Lafavette (loiinty, Mississippi. Data were pooled lioin 10 sample dates, H)9;VH)94. Habitat \imisi:r of SAMriTs Ximbkr of Darikrs Pfrcf.nt with Darikrs Run 52 23 44% S\vift riffle 74 30 41% Moderate rittle 105 46 44% Undercut bank 100 47 47% Pool 55 37 67% H.\BITAT. Etheostoma raneyi was a habitat generalist. Yazoo darters were e\eiily distributed among runs, swift riffles, moderate riffles, and tmdercut banks, but were slighth more common in pools (Table 2). Overall comparison of all five habitat types showed a significant association between habitat and darter occur- rence (G=11.00, p<0.03, 4 df). However, when pools were dropped from the analysis, the result became nonsignificant (G=0.73, p<0.87, 3 df). Principal com- ponents analysis of habitats associated with darter occurrence showed that dart- ers were found in a wide range of habitat configurations (Figure 2). Similar overlays of darter density were made for all other combinations of factor plots, but no relationships were demonstrated. Three components explained 64% of the total variance (Table 3). Principal component 1 (factor I) ordinated darter occurrences along a gradient from deep sites with fine substrates and low veloc- ity to shallow sites with coarse substrates and high velocity. Principal component II (factor II) ordinated occurrences along a gradient of sites with low instream cover and a high percentage of canopy cover to sites with high instream cover and a low percentage of canopy cover, while component III described an inverse relationship between stream width and water velocity. Demographics. Age and sex composition of the E. raneyi population in Mor- ris Creek is shown in Table 4. Sex ratios were significantly female-biased in the age one class, and for the total sample. The oldest female was 45.5 mm SL and 34 months of age; the oldest male was 52.6 mm SL and 33 months of age. Mean length of males was longer (SL) (36.4 mm, SD=6.1) than females (34.5, SD=5.2) (t=3.3; p<0.001) in the total sample. Growth. Growth of £. mneyiwas rapid during the first months of life, declined between four and ten months of age, and leveled off after ten months (Figine 3). Standard length (mm) was positivelv correlated with age (months) for 129 males (SL=19.484+18.388*logX, r=0.7, p<0.01) and 199 females (SL=1 7.346+1 7.366*logX, r=0.8, p<0.01). By seven to nine months of age, mean length of males was greater than that of females (samples were lumped for these three months due to sample size) (t=6.5, p<0.001; mean SL males=39.46 mm, n=19; females=33.01 mm, n=17), suggesting that males grow faster than females do. Gonadal development. All females >30 mm SL (8 months of age and older) and all males >34 mm SL (8 months of age and older) were reproductively mature in March of 1993 and 1994. Males developed intense breeding colora- don (described by Suttkus et al., 1994) and mature testes in December (Figure 4). Some males had mature testes throughout the spring and early summer, but the testes of all males were latent by July. This prolonged period of reproductive readiness is supported by the gonadosomadc index (GSI) for males, which 52 Tulane Studies in Zoology and Botany [Vol. 30 fine low deep substrate -^ coarse velocity water depth -^ high _^ shallow 3 1 1 1 1 1 A 2 A A A A A • /a ■ \m aaA^ a^aa A A ^ ^ A. A B^A k A cvT 1 ^ ■ ^ A A 0^ o 0 _ • ^ ■'...' :•■ m -1 • A A ■■ Aa ■■ A •« ■ . m • A • A _ • • -■ A ■ A • \"^ -2 ■^ ■ A ^ ▲ — -3 1 A 1 1 1 1 ■5* 5 I k L. :> o o E «^ s_ +^ -3 -2 -1 0 1 FACTOID (1) A one ■ two • >3 o o o c o 5 -5^ Figure 2. Overlay of numbers of Etheostoma raneyi samples on plot of standardized factor scores for Factor 1 and Factor 2. increases during late winter, peaks in early spring, and declines in the summer (Figure 5). Ovaries of females begin maturation in late fall, and some individuals have ripe ovaries by early spring (Figure 6). All ovaries are latent by July, similar to the pattern seen in males. The GSI for females also closely follows the pattern of development in males (Figure 5); the values of GSI are much higher in females than in males, illustrating the higher investment by females in ova. Standard length was positively correlated with the GSI of males (r-=0.183, p<0.006; GSI=5. 455+3. 074 SL) and of females (r-=0.127, p<0.002; GSI=-1. 35 1+0.238 SL) (Figure 7). There was also a relationship between SL and somadc weight for females (r'=0.942, p<0.001; somatic weight=-0. 441+0. 017 SL), 1996] Yazoo Darter Life History 53 Table 3. Variable loadini>;s > 0.4 (loadini; valuf of O.WS was retained) on three principal coniponcni axes, after variniax rotation. Percent variance exi)lainecl by each component is jfiven in |)arentheses. Princii'.m. Component V.\RL\BLE 1(26.0%) 11(19.0%) III (18..'i%) Substrate 0.727 Velocity 0.698 0.398 Depth -0.665 Cover 0.773 Canopy -0.694 Width -0.933 and for males (r-=0.942, p<().001; somatic \veight=-0. 666+0. 023 SL) and males and females had virtually identical relationships between these variables (Figure 8). By June, the number of ova in all size classes decreases (Figure 9). Peak num- bers were seen in early spring, corresponding to peak GSI values and percent- ages of ripe ovarian stages. The mean number and diameter of ova in the three largest size classes for 43 females sampled during the breeding season are, from largest to smallest: 52 (SD=20), 1.05 mm (SD-0.11); 80 (SD=35), 0.74 nun (SD=1.10); 303 (SD=129), 0.38 mm (SD=0.03). Both the number and diameter of ova increased with standard length (r-=0.480, p<0.001; number=64. 732+3. 243 SL; r^=0.376, p<0.001; diameter=0.524+0.014 SL) (Figure 10). SpaW'NING behavior. Male Yazoo darters are nonterritorial, but display aggression toward conspecific males when a male engaged in courtship or spawning is approached by another male. Males react by chasing the intruder, or performing stationary lateral displays with erect fins. No elaborate courtship dis- plays were observed in this species. Prior to spawning, males typically pursue females as they move about the substrate selecting a spawning site. When a female is receptive, the male mounts her back with his caudal peduncle to her side. The pair then vibrate as the female attaches an egg to the spawning sub- strate. Eggs are usually attached singly to the substrate. Of the 29 episodes observed, 18 (62%) eggs were attached to a submerged log, 6 (21%) to plants or plant roots, 3 (10%) to the gravel substrate, and 2 (7%) to large rocks. The aver- age diameter of five eggs removed from an aquarium was 1.28 mm. Typically, a single pair spawned numerous times in succession, but not all mountings Table 4. Age and sex composition of Yazoo darters from Morris Caeek, Lafayette Coimiv, Mississippi, 1992-1993. A(. K (iROlP 0 1 + 2+ 3+ Toi \l Male 71 49 8 1 129 Female 82 106 11 0 199 Total 153 155 19 1 328 r 0.79, p= 0.3 20.90, p<(). 001 0.47, p=0.49 14.94, p<0.0()l Sex ratio 1.2:1 2.2:1 1.4:1 1.5:1 (femalermale) 54 Tulane Studies in Zoology and Botany [Vol. 30 50n 0 • male O female ' I ' I ' — I — ' — I ' — I — ' I — I — I I — I — 1 — I I — I 0 4 8 12 16 20 24 28 32 36 40 Months of Age Figure 3. Growth curves for males and females oi Etheostoma raneyi. Points represent sample means. resulted in the release of an egg. Males and females were promiscuous, and no parental care was provided to the eggs. Discussion The Yazoo darter {Etheostoma raneyi) is one of 20 described species of snub- nose darters, subgenus Nanostoma (Page, 1981). Typical habitat of most adults of subgenus Nanostoma was described by Page and Mayden (1981) as clean pools with moderate current and bedrock, cobble or gravel substrate. Exceptions among upland Nanostoma species are Etheostoma etnieri, which prefers riffles and runs (Bouchard, 1977), Etheostoma zonale, which inhabits rocky, vegetated riffles (Page and Mayden, 1981), and Etheostoma coosae, which occurs in all stream mesohabitats (riffles, runs, pools) over gravel or cobble substrate (O'Neil, 1981). Habitat descriptions of Nanostoma species foimd in lowland streams reflect the physical characteristics of these streams, which usually have sand and gravel substrates and low to moderate gradients. Etheostoma zonistium, a lowland species, occurs predominantly in stream margin habitat with sand or sand and gravel stibstrate, and is found less freqtiently in run and riffle habitat (Carney and Burr, 1989). Etheostoma pyrrhogaster is found in stream margin, pool, run and riffle habitat over sand substrate (Carney and Burr, 1989). Etheostoma raneyi is also found in lowland streams, and like E. coosae (O'Neil, 1981) and E. pyrrhog- aster, is found in all habitats. Our analysis showed that E. raneyi individuals were evenly distribiued among runs, swift riffles, moderate riffles and undercut banks, but were slightly more common in pools. However, higher darter density 1996] Yazoo Darter Life Histoiy 55 CD I (0 CO CO CO Gi o> (y> < "=5 — 3 CO o CM LD CM CO cn I CO CO CO CD I en < I Cvl CO CO O) CD CO CO 05 O) Q. (D en I ■>* CM o o I CM > o o 0) Q t I CJ) -"t Date I CO -p CM 05 O) CD CJ5 CD CD CD CD LL CO CM CO CO < < "^ >^ CO A CO •- T- CM O CM CM N. Figure 4. Monthly changes in testicular development of Etheostoma ran^z males. in any particular habitat was not demonstrated by the overlay of darter density on plots of PCA factor scores. Etheostoma raneyiwas the most coinmon species in Morris Creek, and was found in most seine samples that had fishes. Similarly, Carney and Burr (1989) found that E. zonistium was the most common species found in West Fork Clarks River, Kentucky. Like other species in the snubnose group, Yazoo darters are short-lived, with very few individuals surviving more than two years (33-34 months). Some indi- viduals of £. zonistium, E. pyrrhogaster and E. f oova;^ live more than two years (Car- ney and Burr, 1989; O'Neil, 1981), while the maximum life span of £. simoterum is just 18 months (Page and Mayden, 1981). The longest life span reported for a Nanostoma species is for E. zonale, with individuals that survive more than three years (Lachner et al., 1950). The sex ratio of £. raneyi is female-biased for age group one, and for the total sample. Female bias has been reported for components of pcjpulations off. coo- sae (O'Neil, 1981), E. simoterum (Page and Mayden, 1981), and E. zonistium (Car- ney and Burr, 1989), and may be caused by differential mortalit\ in the brightly colored males (Carney and Burr, 1989). As reflected by higher GSI, female E. raneyi invest more than males do in gonadal tissue. This is typical of most fishes (Moyle and Cech, 1988), and sug- gests that females invest more in reproduction. However, male investment in reproduction could be manifested not only in gonadal tissue, but in sexually selected traits as well. In many species, larger males ha\e a reproductive advan- tage over smaller males, due to intrasexual competition for females or lemale choice for larger males. This size difference could be manifested in greater length (SL) of males or in greater somatic weight. In E. raneyi, males do not invest more than females in somatic weight, but achieve longer lengths (SL), as 56 Tulane Studies in Zoology and Botany [Vol. 30 o o o E 0) 140 120 100 0 ■^ Females ■«■ Males 12 10 R ^ ID (D Q 6 w X L o A o 4 o 0 MAMJJYASONDJ FMMAAMM Figure 5. Monthly changes in mean gonadosomatic index oi Etheosloma raveyi m o o Q I I CD -^ Date C i3 CO 0) ~? ^ c!j CO ■^ CM CO CO CO CM < < I I CM cn en I I CO CO CM t^ Figure 6. Monthly changes in ovarian development of Etheostoma raneyi females. Ovarian stages are latent (la), early maturing (em), late maturing (Im), mature (ma), and ripe (re). 1996] Yazoo Darter Life HistoiT (0 E S o o o CD o 200 150 100 Standard Length (mm) Figure 7. Relationship between standard length (SL) and gonadosomatic index (CiSl) io\ Elheostoma raneyi males and females. o 0.1 E o w o 0.01 .>■ 1^.. I - ■ Females A Males 1 , , , , 1 , > 1 30 35 40 45 Standard Length (mm) 50 Figure 8. Relationship between standard length (SL) and somatic weight for Elheostoma raneyi mdles and females. 58 Tulane Studies in Zoology and Botany [Vol. 30 Small Medium ■ Large CO CO CO CD 03 Q. m S < 1 ^ CO C\J CD CM o CM CO c -p uS CM C55 CJ) cts 2 Q. < Q. < 1 CO CM h- O (M CD I (0 CM CD I CO h~ Date Figure 9. Changes in tiie number of ova in the three largest size classes of Etheostoma ran eyi females during the 1993 and 1994 breeding seasons. in other Nanostoma species (Page and Mayden, 1981; Carney and Burr, 1989). If males expend more energy than females do on courtship and aggression during the breeding season, males would be predicted to lose more somatic weight dur- ing that time. This is not suggested by our data, but our samples were lumped over the duration of the breeding season, and may not reveal such subtle rela- tionships. Males also invest in bright breeding colors that are undoubtedly the result of sexual selection, although this has yet to be tested for darters. Bright colors may be responsible for differential mortality in males due to predation, and this could also be viewed as a cost of reproduction. More research on repro- ductive investment in darters is needed before an apparent differential invest- ment by the sexes is understood. The spawning period for E. raneyi is from March through June. A relatively long spawning season is also found in E. pyrrhogaster and E. zonistium (Carney and Burr, 1989), and in E. rafinesquei (Weddle and Burr, 1991). April has been reported as the peak spawning season for E. barrenense and E. rafinesquei (Stiles, 1974; Page and Burr, 1982), E. coosae (O'Neil, 1981), and E. simoterum (Page and Mayden, 1981). Comparing the fecundity and diameter of ova for fishes is difficult, due to differences in methodology among studies. The mean number of ova in the larg- est size class, or batch fecundity, of female E. raneyi was 52 (n=43 females), and the mean diameter of these ova was 1.05 mm. The mean batch fecundity oi E. rafinesquei is 48.3 (Weddle and Burr, 1991). Etheostoma raneyi shares the egg-attaching spawning strategy with all other species of Nanostoma for which spawning behavior is known (Page, 1985; Keevin et al., 1989). These species exhibit no parental care, and deposit eggs in small 1996] Yazoo Darter Life History 59 1.4 0 •Egg diameter (mm) ■Number of eggs 110 90 70 50 30 3 (D C 3 cr o < 0) 25 30 35 40 Standard Length (mm) 45 50 10 Figure 10. Relationship benveen number and diameter of ova and standard length (SL) of Ethenstntnn raneyi females. numbers. Male E. raneyi did not maintain territories associated with particular spawning substrates, but were aggressive to other males. Such moving territories were also observed in E. ba rrenense a.nd E. rafinesquei (Stiles, 1974), E. simoterum (Page and Mayden, 1981), E. flavum (Keevin et al., 1989), and apparently E. pyr- rhogaster and E. zonistium (Carney and Burr, 1989). Page (1983) proposed that the egg-attaching strategy provided some protection from egg predation. This tendency to place the eggs in several places probably precludes economic defen- sibility of any given spawning substrate and, as a result, territoriality is not seen in these species. Although E. raneyi has a limited distribution, within its range it can be the most common species, and does not appear to be in danger of extinction. How- ever, individual populations off. ra/?^'/ should be carefully monitored, because the small streams that are typical habitat for the species are often targeted for habitat alteration, or are degraded by poor land use. For a species with a limited distribution, the loss of several populations and resulting fragmentation of the species' range could quickly lead to imperilment. With E. raneyi, monitoring and habitat protection coidd alleviate retroactive recovery efforts currently needed for other species of subgenus Nanostoma. ACPLNOWT^EDGM E NTS We would like to thank Reid Adams for analysis of growth data, Mel Warren for assistance with gonadal analysis, and Mark Peterson and Steve Ross for dis- cussion of PCA of habitat variables. We appreciate the assistance of Laura Hub- bard with figures. 60 Tulane Studies in Zoology and Botany Literature Cited BOSCHUNG, H. T., R. L. Mayden, and J. R. TOMELLERI. 1992. Etheostoma rhermocki, a new species of darter (Teleostei: Percidae) from the Black Warrior River drainage of Alabama. Bull. Alabama Mus. Nat. Hist. 13: 11-20. Bouchard, R. W. 1977. Etheostoma etnieii, a new percid fish from the Caney Fork (Cumberland) River system, Tennessee, with a redescription of the subgenus Ulocentra. Tulane Stud. Zool. Bot. 19: 105- 130. Carney, D. A. and B. M. Burr 1989. Life histories of the bandfin darter, Etheostoma zonistium, and the firebelly darter, Etheostoma pyrrhogaster, in Western Kentucky. Illinois Nat. Hist. Surv. Biol. Notes 134 (16 pp.). Heins, D. C. .and J. A. Baker. 1993. Reproductive biolog)' of the brighteye darter, Etheostoma lynceum (Teleostei: Percidae), from the Homochitto River, Mississippi. Ichthyol. Explon Freshwaters 4: 11-20. Keevin, T. M., L. M. P,\GE, and C. E. Johnston. 1989. The spawning behavior of the saffron darter (Etheostoma flavum). Trans. Kentucky Acad. Sci. 50: 55-58. LachneR, E. a., E. F. WestlakE, and p. S. HWDWERK. 1950. Studies on the biology of some percid fishes from western Pennsylvania. Amen Midi. Nat. 43: 92-1 11. Mo\TE, P. B. AND J. J. Cech. 1988. Fishes: An Introduction to Ichthyology. Prentice Hall, Englewood Cliffs, New Jersey (559 pp.). O'Neil, p. E. 1981. Life histor\' of Etheostoma coosae (Pisces: Percidae) in Barbaree Creek, Alabama. Tulane Stud. Zool. Bot. 23: 75-83. Page, L. M. 1981. The genera and subgenera of darters (Percidae, Etheostomatini). Occas. Pap. Mus. Nat. Hist. Univ. Kansas 90 (69 pp.). Page, L. M. 1983. Handbook of Darters. T. F. H. Publications, Inc., Neptune City, New Jersey (271 pp.). Page, L. M. 1985. Evolution of reproductive behaviors in percid fishes. Illinois Nat. Hist. Surv. Bull. 33: 275-295. Page, L. M. and B. M. Burr. 1982. Three new species of darters (Percidae, Etheostoma) of the subgenus Nanostoma from Kentucky and Tennessee. Occas. Pap. Mus. Nat. Hist. Univ. Kansas 101 (20 pp.). Page, L. M. and R. L. M.WDEN. 1981. The life history of the Tennessee snubnose darter, Etheostoma slmot- erum, in Brush Creek, Tennessee. IlHnois Nat. Hist. Surv. Biol. Notes 117 (11 pp.). Ross, S. T., R. H. McMiCHAEL, Jr., and D. L. Rupee. 1987. Seasonal and diel variation in the standing crop of fishes and macroinvertebrates from a Gulf of Mexico surf zone. Estuarine Coastal Shelf Sci. 25: 391-412. Ross, S. T.,J. G. Knight, and S. D. Wilkins. 1990. Longitudinal occurrence of the bayou darter (Per- cidae: Etheostoma rubrum) in Bayou Pierre — response to stream order or habitat availability? Polsk. Arch. Hydrobiol. 37: 221-233. SOKAL, R. R. and F.J. Rohlf. 1981. Biometry (ed. 2). W. H. Freeman and Company, New York (859 pp.). Stiles, R. A. 1974. The reproductive behavior of the Green and Barren River Ulocentra (Osteichthyes: Pericidae: Etheostoma). ASB Bull. 21: 86-87. SUTTKUS, R. D. AND D. A. Etnier 1991. Etheostoma tallapoosae and E. brevirostrum, two new darters, subge- nus Ulocentra, from the Alabama River drainage. Tulane Stud. Zool. Bot. 28: 1-24. SUTTKUS, R. D., R. M. Bailey, and H. L. B,\RT. 1994. Three new species oi Etheostoma, subgenus Ulocentra, from the Gulf Coastal Plain of southeastern United States. Tulane Stud. Zool. Bot. 29: 97-126. Thompson, K. W. and R.J. MuncY'. Undated (submitted in 1985). A. status report on the Yazoo River darter, Etheostoma {Ulocentra) sp. in northern Mississippi. U.S. Fish and Wildlife Service Report, Con- tract 14-16-009-1543. Weddle, G. K. and B. M. Burr 1991. Fecundity and the dynamics of multiple spawning in darters: an in- stream %i\\Ay oi Etheostoma rafinesquei. Copeia 1991: 419-433. Wilkinson, L. 1990. SYSTAT: The System for Statisdcs. SYSTAT, Inc., Evanston, Illinois. WiLLL\Ms, J. E., J. E. Johnson, D. A. Hendrickson, S. Contrer.\s-Balder\s, J. D. Willl\ms, M. Navarro-Mendoza, D. E. McAllister, and J. E. Deacon. 1989. Fishes of North America: endan- gered, threatened, or of special concern. 1989. Fisheries 14: 2-20. LEAF-MINER DEFENSES IN BROMEUA PINGULWL. (BROMELIACEAE) IN \TR.\(:RUZ, MEXICO Victor RkxvGra^, Jose G. Garcia-Franco Di'partnmrnto dp Erolojria Vrirclal. histilitlo dv Ecohi^a, A.C.. Aj)do. 63, Xala/ui. MIR 91000, Mexico Angel Trigos-L\nda Deparliinu-ulo dc Qiiitnica y Biologin, Vnivnsidad de las Americas, C.liohdti. I'll'. 72S20, Mexico Rachel Mata and Perla Castaneda Departamcntn dp Farmacia, Facidtad dp Qiumica, U.N. A.M., Mexico, DF(M^10, Mexico Abstract Leaf-miners are present on leaves of all adult sexually reproducing individu- als of Bromel/a pinguin (Bromeliaceae), but not on leaves of vegetativeh repro- ducing or immature individuals; thus, we hypothesized that there should be differences in defenses between them. No differences were found when we com- pared metabolic contents, humidity, and nitrogen of mined and unmined leaves, but fotmd significant differences in ash, fiber and lipid contents. Mined leaves had more fiber and lipids and less ash than unmined leaves, but differences may exist only during a short period of time and were not detected by oiu^ research. The leaf-miner Melanagromiza sp. (Diptera: Agromizidae) attacks leaves with the highest fiber contents. Cited research is on trees and their seedlings, but results could differ in large clonal plants where metabolites move between ramets. The response of leaf-miners could be to a combined series of factors other than food quality or the amount or type of defenses, such as a plant-herbivore-parasitoid complex. Introduction Plants cover most of the Earth's surface, and even though subject to varying degrees of herbivory by a diversity of organisms, they are not eliminated because, among other things, they ensure their own defense (Crawley, 1983; Howe and Westley, 1988; Thompson, 1982; Weis and Berenbaum, 1989). Plant defenses include mechanical protection on the siuTace of the plant, compU-x polymers or silica crystals that reduce plant digestibilit), and plant toxins that kill or repell herbivores at very low concentrations (Coley and Aide, 1991; Howe and Westley, 1988; Rico-Gray, 1989). There are large differences among plant species in both the amount and type of anti-herbivore defense toxins in their leaves (Coley, 1988). Defenses against herbivory should be allocated in direct proportion to the tissue or plant part that confers the greatest fitness to the indi- vidual plant (Dirzo, 1984). Plant defenses change with time (e.g., leaves toughen as they mature, and sometimes potent toxins in young leaves are replaced by lig- nified or silicified tissues in older leaves and twigs), and virtually all plants and plant shoots become less palatable as they mature (Dirzo. 1984: Howe and West- Tulane Studies in Zo()I()sr\ and Botany 30: 61-()7. I'J'.Ki 61 62 Tulane Studies in Zoology and Botany [Vol. 30 ley, 1988). In general, young leaves are more vulnerable to herbivores than are mature leaves (Coley and Aide, 1991), and leaf-miners have been found to be more abundant on seedlings in comparison with mature plants of the same spe- cies (Godfray, 1985). In a tropical dry forest on the coast of Veracruz, Mexico, the leaf-miner Melanagromiza sp. (Diptera: Agromizidae) was present on leaves of all adult, sex- ually reproducing individuals of Bromelia pinguin L. (Bromeliaceae), whereas, leaves of vegetatively reproducing or immature individuals did not have leaf- miner tunnels. Bromelia pinguin is monocarpic, dying slowly after fruiting, usually leaving a living offshoot (Garcia-Franco and Rico-Gray, 1995). We hypothesized that there should be differences in defenses between dying, sexually reproduc- ing individuals, and younger, non-sexually reproducing individuals. Plants of B. pinguin present many features considered to be protection against herbivores. The straplike leaves are highly fibrous, they have hard pointed tips, with sharp hooked thorns on their margins, and thick cuticles (Garcia-Franco and Rico- Gray, 1995; Hallwachs, 1983). Leaves oi B. pinguin also have flavonoids (pend- uletine, cirsimaritine, casticine), ferulic acid, diterpenoids derivated from filoc- tadone (3-oxofilocladan-16-ol, filocladan-16a-diol, 3-oxopimar-15ene-7p, 8^- diol), sterols (estigmasterol, [3-sitosterol, P-D-glucositosterol), and P-D-glucopira- nose (Chavez-Gallardo, 1993, and references therein). Despite this range of defenses, the plants are subject to attack by different organisms (Table I); in particular a leaf-miner, the larva of a fly, Melanagromiza species. Study Site Field work was conducted in a lowland tropical deciduous forest at Centro de Investigaciones Costeras La Mancha (CICOLMA) on the coast of the state of Veracruz, Mexico (19° 36' N, 96° 22' W; altitude <50 m). Annual precipitation varies greatly (1100-1700 mm), but most falls between June and September, and mean temperature is 22°-26°C (minimum 10°C, maximum 38°C; Garcia-Franco and Rico-Gray, 1995). Bromelia pinguin inhabits the sandy understory of the trop- ical dry and deciduous forests, which have a relatively simple structure and com- position, with tree species such as Bursera simaruba (L.) Sarg. (Burseraceae), Brosimum alicastrum Swartz (Moraceae), Enterolobium cyclocarpum (Jacq.) Griseb. (Leguminosae), Cedrela odorata L. (Meliaceae), Ficus cotinifolia Kunth (Mora- ceae), and shrub species such as Nectandra coriacea (Sw.) Griseb. (Lauraceae), and Coccoloba barbadensis Jacq. (Polygonaceae) (Blain, 1988; Blain and Kellman, 1991). Table 1. Herbivores observed on Bromelia pinguin and the plant structure affected at La Mancha, Ver- acruz, Mexico during the period of April 1989-March 1991. Leaf: (larva) Melanagromiza sp. (Diptera:Agromyzidae) (larva) Mandura quinquemnculata (Lepidoptera:Sphingidae) (adult) Gecarcinus lateralis (Decapoda:Brach)au-a) Spike: (ant-tended) Membracidae (Homoptera) Fruit: (larva) Drosophilidae, Sarcophagidae (Diptera) (larva) Lepidoptera Seed: (adult) Cecarcinus lateralis Seedling: (adult) CJecarcinus lateralis 1996] Leaf-Miner Defenses in Broinrlia 63 Materials and Methods Bwmelia pingiiiu is a terrestrial plant in dry habitats from Mexico to \'ene/u- ela and on the C>aribbean Islands, from sea level to 780 m elevation (Ilallwaehs, 1983; Smith and Downs, 1979). Serial monocarpy or sympodial dichotomv (Ben- zing, 1980), sexual reproduction, and clonal growth are present in its life cycle. In CICOLMA it forms dense patches or subpopnlations (denies) on sandy soils in deciduous forest. Flowers are present lor a few davs during the c\y\ season (March-April). Butterflies and hummingbirds {Amazilia beryUina, Cyraulhus lat- irostris: Trochilidae) are the main flower visitors (Garcia-Franco and Rico-Gray, 1995). Although a high fruit set with abundant seeds is produced every year, seed and seedling predation by red land crabs {Gecarcinus lateralis Frem.) is so high that few genets are recrtiited (Garcia-Franco et al., 1991). Plant material was collected between August and September ol 1989, air dried and sent to the laboratories (5 kg dry weight each of mined and unmined leaves to each laboratory). Collections were made just after peak rainfall, to avoid possible differences due to environmental stress. In general, herbivores use plants because either they are rich in niurients (e.g., nitrogen, humidit\' con- tent) or poor in defenses (physical or chemical); thus two types of anahses were needed. Ash, hvmiidity, fiber, lipids, and nitrogen analyses were conducted at the Departamento de Quimica y Biologia, Universidad de las Americas. Determina- tions were made in duplicate according to the methods in Horwitz (1980) with an enor margin of 0.1-2.0%. Metabolic contents (MeOH extract), to test for alle- lochemicals, were analyzed with chromatography (flPLC) and visible U\' light spectrum at the Departamento de Farmacia, Facultad de Quimica, U.N.A.M. (Chavez-Gallardo, 1993). Insects were collected throughout the study period (April 1989-March 1991). Results No differences in metabolic contents were found when mined and unmined leaves were compared (Figure 1). We did not find significant differences when humidity (ANOVA, F,;,=1.22, p>0.384) and nitrogen (ANOVA, F,,„=9.9, p>0.088) contents were compared between mined and unmined leaves (Figure 2). On the other hand, we found significant differences in the percent of ash (ANOVA, F,,,, =54.366, p<0.018), fiber (ANOVA, F,,., =16.341, p<0.05), and lipid (ANOVA, F,| 2 =133.031, p<0.007) contents, between mined and unmined leaves. Mined leaves had more fiber and lipids, and less ash than did unmined leaves (Figure 2). Discussion Leaf-miners attack leaves on old and decaying Bwmclia p/iii^uin plants: the toughest leaves, the ones with the highest fiber content. We found no difier- ences in nitrogen or water content, or in secondary compounds, between leaves in young and old plants, unless the higher lipid contents of older leaves is attrac- tive to miners. Coley (1987) found that in Cerropia, herbivores show significant preference for leaves from larger plants, suggesting that leaf defenses decrease with plant size or age. Cooke et al. (1984) demonstrated that young leaves of several tree species have more toxins and less nutritive value than do mature leaves. How- 64 Tulane Studies in Zoology and Botany [Vol. 30 CD O C CC SI o CO 2.5 e 1.5 1 0.5 0 — mined -^unmined 200 250 300 350 400 450 500 wavelength nm Figure 1. Ultra\iolet spectra of the methanolic extract of Bromelia pinguin for mined and immined leaves. ever, the response of leaf-miners could also be to a combination of factors other than food quality or the amount or type of defenses. Flowering and fruiting by B. pinguin may act as a mating cue for adults of the Melanagromyza fly. Female flies oviposit on the leaves of decaying plants, whereas their offspring develop during the warm-humid portion of the year. The mating time and ovipositing behavior of adult Mekmagrornyza flies could be a response to an absence of predators or parasitoids. Instead of a direct response to plant phe- nology, Melanagromyza flies might be influenced by a plant-herbivore-parasitoid complex (Price et al., 1980, 1986), or fly behavior may be influenced by wasp activity patterns (Thompson, 1994). 1996] Leaf-Miner Defenses in Bromelia 65 50 40 0) o 1— 0 Q- 30 20 10 0 ash ra humidity EElfiber lipids Unitrogen mined leaves unmined leaves Figure 2. Percent contents of ash, water, fiber, lipids, and nitrogen for mined and unmined leaves oi Bromelia pinguin. In certain situations, a heavy attack by leaf-miners could be considered bene- ficial because it increases the chance of leaf abscision and eHniination of miners (Simberloff and Stiling, 1987). The effect of leaf-miners on B. pingui)i should be negligible because the attack is on dying leaves whose main compoimds already should have been relocated to offshoots, fruits, and seeds (Jackson et al., 1985). Leaf-miners were only found in decaying individuals, but mined and inimined leaves did not differ in secondary metabolites. The question remains: Why are only the decaying individuals attacked? The response of a plant to attack from herbivores or pathogens depends in part on its condition, and in part on its inherited ability to relocate resources. If serious stress diverts energy from chemical defense, plants or even parts of plants weak- 66 Tulane Studies in Zoology and Botany [Vol. 30 ened by shade, disease, poor soil, or recurrent defoliation may be vulnerable to herbivory (Howe and Westley, 1988). In our example, old plants ("mother plants") could be weakened as a result of relocation of metabolic products to off- shoots ("daughters") and fruits (Cook, 1985; Pitelka and Ashmun, 1985). Finally, much work has been done on the effect of certain metabolic products (e.g., tan- nins) and nitrogen levels on insect growth and food preferences, with conflict- ing results (Cooke et al., 1984). Although our results do not show major differences between mined and unmined plants, differences may exist during a short period of time, but were not detected by our research. Also, cited research is on trees and their seedlings, and results could differ in large clonal plants, where metabolites move between ramets. Acknowledgments We thank Vicente Hernandez and Alfonso Pescador who identified the leaf- miner and the moth, respectively. Selene Chavez-Gallardo and Gerardo Franco helped with laboratory analyses. Field and laboratory work were supported in part by CONACYI grants 903579 and 1259-N9204 (to VRG), and Insdtuto de Ecologia, 902-16. LITERATI' RE CiTED Benzing, D. H. 1980. The Biology of Bromeliads. Mad River Press, Eureka (305 pp.). BlAlN, D. 1988. Factors affecting the early stages of regeneration of three tropical tree species in a sea- sonal forest, Veracruz, Mexico. Unpubl. M. Sc. thesis. York University, Ontario, Canada (155 pp.). Blain, D. and M. KELLMAN. 1991. The effect of water supply on tree seed germination and seedling sur- vival in a tropical seasonal forest in Veracruz, Mexico. J. Trop. Ecol. 7: 69-83. COLEY, P. D. 1987. Patrones en las defensas de las plantas: ^porque los herbivoros prefieren ciertas espe- cies? Rev. Biol. Trop. 36 (suppl. 1): 151-164. COLEY, P. D. 1988. Effects of plant growth rate and leaf lifetime on the amount and type of anti-herbi- vore defense. Oecologia 74: 531-536. COLEY, P. D. .4ND T. M. Aide. 1991. Comparison of herbivoiy and plant defenses in temperate and tropi- cal broad-leaved forests. In: P. W. Price, T. M. Lewinsohn, G. Wilson Fernandes, and W. W. Benson (eds.), Plant-Animal Interactions: Evolutionary Ecology in Tropical and Temperate Regions. John Wiley and Sons, Inc., New York (pp. 2.5-49). Chavez-Gallardo, M. S. 1993. Estudio quimico y biologico de Bromelia pinguiu (Bromeliaceae). Unpu- blished B. Sc. thesis, Facultad de Quimica, U.N.A.M. Mexico, D.F. Cook, R. E. 1985. Growth and development in clonal plant populations. In: J. B. C.Jackson, L. W. Buss, and R. E. Cook (eds.). Population Biology and Evolution of Clonal Organisms. Yale University Press, New Haven (pp. 259-296). Cooke, F. P., J. P. Brown, and S. Mole. 1984. Herbivoiy foliar enzyme inhibitors, nitrogen and leaf structure of young and mature leaves in a tropical forest. Biotropica 16: 257-263. Crawler-, M.J. 1983. Herbivory. Blackwell Scientific Publications, Oxford (437 pp.). DiRZO, R. 1984. Herbivoiy: a phytocentric overview. In: R. Dirzo and J. Sarukhan (eds.). Perspectives on Plant Population Ecology. Sinauer Associates, Inc. Sunderland, MA (pp. 141-165). GARCiA-FR.\NC.O, J. G. AND V. Rico-Gray. 1995. Population structure and clonal growth in Bromelia piii- gidn L. (Bromeliaceae) in diY forests of coastal Veracruz, Mexico. Tulane Stud. Zool. Bot. 30: 27-37. Garcia-Franco, J. G., V. Ric.o-Gray, and O. Zayas. 1991. Seed and seedling predadon of BrumeUa pin- guin L. by the red land crab Gecarcinus lateralisFrem. in Veracruz, Mexico. Biotropica 23: 96-97. Godfray, H. C]. J. 1985. The absolute abundance of leaf miners on plants of different succcssional stages. Oikos 45: 17-25. Hallwachs, W. 1983. Bromelia pinguin and B. karatas. In: D. H. Janzen (ed.), Costa Rican Natural His- toiy University of Chicago Press, Chicago (pp. 195-197). HORWITZ, W. (ed.). 1980. Methods of Analysis. Association of Official Analytical Chemists. Washington, D.C. (1018 pp.). Howe, H. F. and L. C. Westley. 1988. Ecological Relationships of Plants and Animals. Oxford Univer- sity Press, New York (273 pp.). 1996] Leaf-Miner Defenses in Bromelia 67 Jackson, J. B. C, L. W. Buss, and R. E. Cook. 1985. Population Biologvaiul Kxoiuiion dl Clonal c:)rjfan- isnis. Yale University Press, New Haven (530 pp.). PlTELK.\, L. F. .\NDj. W. ASHMUN. 1985. Physiology' and integration of ramets in ( lonal plants. In: ). B. C. Jackson, L. W. Buss, and R. E. (^ook (eds.). Population Biol()g\' and Kvoluiion ol (Clonal Organisms. Yale University Press, New Haven (pp. :i99-435). Price, P. W., C. E. Bouton, P. Gross, B. A. M(.Piii:ron, J. N. Thompson, .\nd A. E. Wkis. 1980. Interac- tions among three trophic levels: influence oi plants on interactions ijetwi'cn inse(t lu-ihivores and natural enemies. .\nn. Rev. Ecol. Syst. 1 1: 41-t)5. Price, P. W., M Westobv, B. Rice, P. R. Atsatt, R. S. Fritz, J. N. Thompson, and K. Mobi.b. 1986. Par- asite mediation in ecological interactions. Ami. Rev. Ecol. Svst. 17: 487-505. Rlt;o-GR\V, V. 1989. The importance ot floral and ciicum-floral nectar to ants inhaJMting ch) tropical lowlands. Biol. J. Linnean See. 38: 173-181. SiMBERl.OFF, D. .\ND P. Stiiing. 1987. Larval dispersion and sur\ivorshi|j in a leaf-mining moth. Ecology 68: 1647-1657. Smith, L. B. and R.J. Downs. 1979. Flora Neotropica, Monograph 14, Part 3 (Bromelioideae). The New York Botanical Garden, New York (pp. 1665-1666). Thompson, J. N. 1982. Interaction and Coevolution. John Wiley and Sons, New \brk ( 1 79 pp) ■ Thompson,}. N. 1994. The Coevolutionary Process. University of Chicago Press, Chicago (376 pp.). Weis, a. E. .and M. R. Berenbaum. 1989. Herbivorous insects and green plants. In: V\. G. .\brahamson (ed.), Plant-Animal Interactions. McGraw-Hill Book Company, New York (pp. 123-162). OBSERVATIONS ON SOME FUNC.I FROM LOUISIANA AND MISSISSIPPI IN COMPARISON WITH THOSE OF MEXICO Gaston Guzman Inslitiilo df Ecologia, Aparlado Postal 63, Xalapa, Veracruz 91000, Mexico Abstr.\ct Forty species of fungi (2 Ascomycodna, 37 Basidiomycotina, and I lichen) were collected during several field trips carried out in Louisiana and Mississippi. The discussion of these species emphasizes their distribution in Mexico, mainly in mesophytic forests, a type of vegetation with strong similarities to the eastern deciduous forests of the U.S.A. Psilocybe pseudobullacea is recorded from the U.S.A. for the first time, and P. tampanensis and P. cuhensis are new records from Mississippi or Louisiana. Introduction Many species of fungi found in the eastern U.S.A. also occur in the tropics, subtropics, and high mountains of Mexico (Sharp, 1948; Welden and Lemke, 1961; Guzman, 1973). During a summer foray organized by the Gulf States Mycological Society (12-19 June 1995) in Mississippi, and some independent trips in Louisiana, several fungi were found that are common to both countries. These are discussed. Fungi Considered Table 1 shows the forty species discussed here, of which twenty were gathered by the author and deposited in the herbarium of the Institute of Ecolog)' (XAL) at Xalapa, Veracruz, Mexico. These fungi were found in five localities (Table 2). Two of these are covered by forests, two are deposits of sugarcane bagasse, and the fifth consists of the gardens and streets of New Orleans. Observations on the Species One of the two ascomycetes (Table 1) was Hypomyces lactifluorum. It is a com- mon parasite of russulaceous fungi and distorts the basidiocarp and gives it a reddish orange color. The parasite is common in Quercus, Pinus-Qiiercus, and mesophytic forests of Mexico, and also occurs in the deciduous forests of eastern North America. This host-parasite combination is edible and quite common in Mexico where it is sold in markets under the name of "enchilado" or "hongo enchilado" (i.e., hot mushroom), because of its somewhat spicy taste. Xylaria magnoliae grows only on the fallen fruits of Magnolia, a tree common both to the deciduous forests of eastern U.S.A. and the mesophytic forests of Mexico. San Martin-Gonzalez and Rogers (1989) reported this species from Mexico on fruits of M. schiedeanaSchX. Xylaria magnoliae is closely similar to X persicaria (Schwein.: Tulane Studies in Zoology and Botany 30: 69-74. 1996 69 70 Tulane Studies in Zoology and Botany [Vol. 30 Table 1. Species of fungi considered. The numbers indicate the localities (Table 2). An asterisk indi- cates specimens deposited at XAL Herbarium. ASCOMYCOTINA Hypotnyces lactifluorum (Schwein.: Fr.) Tul. (1) *Xylaria magnoliae]. D. Rogers (1) Basidiomycotina Amanita hemibapha (Berk. &: Broome) Sacc. (1) Amanita rubcscens (Pers.: Fr.) S. F. Gray (1) Amanita vaginata (Bull.: Fr.) Vittad. (1) Bot£tus edulisV>\\\\.: Fr. (1) Boletus pinophilus Pilat & Dermek ( 1 ) Ca/i'«//«f3)«//i{/o//M/5 (Bosc) Morgan (1) Calostoma cinnabarina Desv. ( 1 ) *Cantharellus conjluens (Beik. & M. A. Curtis) Petersen (1) *CanthareUus lateritins (Berk.) Singer (1) * Chlorophyll um molydites (Meyer: Fr.) Mass. (3,4,5) * Conocybe lactea (Lange) Metrod (1) *Craterellus odoratiis (Schwein.) Fr. (1) Cymatoderma caperatum (Berk. & Mont.) Reid (2) Ganodertna curtisii (Berk.) Murrill (1) Hydnopolyporus fimbriatus (Fr.) Reid (1) Lactarius hygrophoroidesBerk. & Cint. (1) Lentinus boryanus (Berk. &; Mont.) Singer (1) *Lepiota americana (Peck) Peck (3,4) * Microporellus obovatus (Juhng.) Ryvarden (1) *Panafolus antillanun (Fr.) Dennis (1) *Pe)r)iniporia phloipliila G\\b. &: Blackwell (5) *Pleurotus pulmonarius (Fr.) Quel. (2) *Pluteus cervinus (Schaeff.: Fr.) Kiunm. (1,4) *Psilocybe cubensis (Earle) Singer (1) *Psilocybe pseitdobullacea (Petch) Pegler (3) *Psilocybe tampanensis Guzman & Pollock (1) Pycnoponis sanguineus (L.: Fr.) Murrill (1) Russula brniipes Peck ( 1 ) Russulajlavida Frost & Peck ( 1 ) *Russula virescens (Schaeff. ex Zanted) Fr. (1) Schizophyllum communeYr. (1) Sparassis spathulata (Schwein.: Fr.) Fr. (1) * Stereum frustulatum'vax. subpileatum (Berk. & M. A. Curtis) Welden (2) *Stereumfasciatum (Schwein.) Fr. (2) Trichaptum biforinis (Fr.) Ryvarden (1) *Volvariella volvacea (Bull.: Fr.) Singer (4) *Xerula chrysopepla (Berk. & M. A. Curtis) Singer (1) Lichens *Chiodeclon sanguineum (Sw.) Vain. Fr.) Berk. & M. A. Curtis, a fungus found exclusively on fallen fruits of Liquidam- bar, another common tree of the above-mentioned U.S.A. forests. Among the basidiomycetes observed were species of Cymatoderma, Sparassis, and Stereum of the thelephoraceous fungi, Cantharellus and Craterellus among the 1996] Mexican and Gulf Coast Fungi 71 Table 2. Localities where timgi were obsened. 1. Old River Wild Life ManageineiU Area, I'earl River C>()iinl\, Mississippi. Subdeciduous fbresi with Qui'iTus, CMvya, Magnolia, Pinus, and meadciws. 2. [ean Lafille National Historical Park, liaralaria I'nil, Jeffersson Parish, Louisiana. .Swampy forest with Taxodium, Acer, Celtis, Qiieniis. 3. Donaldsomille region, .Ascension Parish, Louisiana. Sugarcane bagasse. 4. Near Brusly, West Baton Rouge Parish, Louisiana. Sugarc ane bagasse. 5. New Orleans, Orleans Parish, Louisiana. Streets and parks. chanterelles, and Ganoderma, Hydnopolyporus, Pycnoporus, Perenniporia, and Trichaptum of the polyporaceous fungi. Cantharellus lateritius and C. confluens ■a.re closely related. Corner (1966) con- sidered C. conjluens conspecific with C. odoratus, citing clamp connections on the hyphae of both. Petersen (1979a) wrote about C. latent ins And C. conjlueny. "The colors of pileus and stipe (in C. conjluens) are brighter than those of C. lateritius, and with virtually no pinkish component. Microscopic characters vary little in the two taxa." At first (1979b) Petersen reported C. lateritius as C. odoratus and C. confluens as C. lateritius. In his color plate 11, figures 5-(S, he shows one fimgus (the true C. confluens) with a more orange pileus and a white to yellow hyme- nium as compared with the other (the true C. lateritius), which is paler or more pinkish in both parts. Cantharellus confluens was described from Orizaba, Ver- acruz, Mexico (Berkeley, 1867) as Craterellus cow/7nf/?5 Berkeley & M. A. Courtis. It was next reported by Guzman and Sampieri (1984) as C. odoratus from the region of Huatusco, Veracruz. This fungus is very frequent in both regions and is the most important edible mushroom in the Huatusco market, where it is known as "hongo de encino" (i.e., oak mushroom). It is so popular and excellent in taste among edible species that it won first place among mushroom dishes pre- pared during a Mushroom Exposition at Xalapa in 1983 (Guzman and Sampieri, 1984). This fungus is known today in CcSrdoba and Xalapa regions as "hongo de Huatusco" (i.e., Huatusco's mushroom), because every year it is present in the dishes at the mushroom fairs of Cordoba. Petersen (1979a) states that Cantharel- lus lateritius "is by far the most common cantharelloid fimgus with suppressed hymenial folds in eastern North America." It is a tropical element in the eastern U.S.A. versus C. confluens, which appears to be a rare species. One of the collec- tions of C. odoratus reported by Guzman and Sampieri (1984) from the Cofre de Perote region (Veracruz State) in a color plate (Figure 4) is really C. cibarius; this is also an important edible mushroom quite frequent in Pinus-Quercus forests of Mexico. Cantharellus cibarius \s exported to the U.S.A. and Europe. Cymatodentia caperatum is a tropical element in the eastern U.S.A. and is quite common in southern Louisiana and Mississippi dining warm, wet periods. Spar- assis spathulata is unknown in Mexico, but 5. crispa Wolf: Fr. and S. radicata Weir are more or less common in the Pinus-Quercus forests of the country. Stereumfrus- tulatum var. subpileatum Welden and S. fasciatum {S. ostrea (Blume & Nees: Fr.) Fr.) are common in the same Pinus-Quercus and mesoj)hvtic forests of Mexico. Among the polypores, Ganoderma curtisii, Hydnopolyporus Jimbriatus, Microporellus obovatus, and Trichaptum biformis are of frequent occurrence in Pinus-Quercus and 72 Tulane Studies in Zoology and Botany [Vol. 30 mesophytic forests of Mexico. Perenniporia phloiphila, found on the bark of Qiier- cus virginiana Mill, in New Orleans, is unknown in Mexico. Pycnoporus sanguineus is sometimes considered conspecific with P. cinnabarina (Jacq.: Fr.) Karst. (Cun- ningham, 1995; Castillo and Guzman, 1970). I have examined several collections from the eastern U.S.A., mainly from Michigan, and compared them with several from Mexico, Colombia, and Brazil and could not find any difference between them in either macroscopic or microscopic features. This fungus, with Schizophyl- lum commune, is a very common species in the eastern U.S.A. but in Mexico they grow only in the tropical and subtropical (mesophytic) vegetation; they are absent from Pinus-Quercus forests except in some sites that border on the tropics. These two species, Pycnoporus sanguineiis and Schizophyllum commune, may be used as ecological indicators from tropical sites with a strong human influence (Guzman, 1994). Among the agarics (Table 1), all are more or less common in pine-oak forests or mesophytic forests of Mexico, or both, except Chlorophyllum. molybdites, which is typical of meadows and gardens of tropical regions. Lepiota americana, frequent in the sugarcane bagasse of sites 3 and 4, seems rare in Mexico (Guzman and Guzman-Davalos, 1992). Panaeolus antillarum, also known as P. solidipes (Peck) Sacc, P. campanulatus (L.: Fr.) Quel., or P. phalaenarum (Fr.) Quel, sensu Kiihner & Romagnesi, is not an uncommon meadow mushroom on cow or horse dung in all the world's tropical and subtropical regions, including the eastern U.S.A. Vol- variella volvacea occurs in all tropical regions of the world but is rare in the east- ern U.S.A. and in the mesophytic forests of Mexico. Xerula chrysopepla is a widespread mushroom growing from the tropics to the cold coniferous forests, although rare in the latter and in deciduous and mesophytic forests. The three species of Psilocybe represent new records for Mississippi and Loui- siana. Psilocybe tampanesis was known previously only from the type locality: SE of Badon, near Tampa, Florida, and is based on one specimen (Guzman and Pol- lock, 1978; Guzman, 1983). The Mississippi material, also one specimen, was col- lected in a meadow of sandy soil, a habitat similar to that of the type collection. However, the Mississippi specimen has a conic-subumbonate pileus versus a con- vex and slightly umbilicate pileus in the type, a feature not considered signifi- cant enough to separate the Mississippi specimen into another species. More specimens are needed from the two localities to properly evaluate this feature. The spores and cheilocystidia, as well as the paucity of pleurocystidia, agree in both collections. Spores from the Mississippi material are 9.5-10.5 (-11) x 6.5-7 (- 8) X 6-6.5 (-9) |im, and are subrhombic and thick-walled. The pleurocystidia are rare, 12-16 (-17.5) x (5-) 5.5-6.5 (-9) |im, vesiculose-acuminate or lageniform, and cheilocystidia are (14. 5-) 16-25.5 (-32) x 4-6.5 (-7) )im and lageniform. Psilo- cybe pseudobullacea was known only from eastern Africa, Sri Lanka, New Guinea, Ecuador, Venezuela, and Mexico (Guzman, 1983), where it grows on dung or sugarcane bagasse. The Louisiana specimens, one collection with several basidi- omes on sugarcane bagasse, agree well with my previotis description (Guzman, 1983). This is the first record for the U.S.A. The spores are (8.5-) 9.5-11 x (6-) 6.5-7 x 6-7 ^im, the cheilocystidia (12-) 13.5-25 x 5.5-8 )im; the pleurocysddia are absent. The fungus does not show a bluing reaction, and consequently it does not have hallucinogenic properties. Finally, P. cubensis is a subtropical fimicolous fimgus widely distributed throughout the world, but has not been reported for- mally from the Mississippi region. The single specimen collected has spores (12-) 13-16 (-17.5) |im long, which agree well with those reported earlier (Guzman, 1996] Mexican and Gulf Coast Fungi 73 1983). It is interesting to note that the halhicinogenic P. caerulescens Miurill, described from Montgomery, Alabama in 1923 has not been collected elsewhere in the U.S.A. It is a common species in Mexico and has been reported from Pan- ama and Venezuela (Gu/man, 1983) and from Maitinique in the Caribbean zone (Pegler, 1983b). Pleurotus pidmouarius, also known as P. oslrealus var. jlorida Eger (Guzman et al., 1994), seems to be of frequent occurrence in the eastern U.S.A. and in southern Europe, but unknown in Mexico. By means of a spore print from the Louisiana collection, this species is now being cultined at Xalapa in an attempt to obtain fructifications and for crossing with other strains, especially the Florida strain. Lentinus boryamis, also known as L. dentosa (Fr.) Minrill, is a common edi- ble fungus in subtropical Latin America, where it grows on stimips. It is closely related to the Japanese shi'itake L. edodes (Berk.) Singer. Both species are treated by Pegler (1983a) as members of the genus Lenlinula (Tribe Collybiae). Amanita hemibapha is recognized in eastern U.S.A. imder the name A. caesarea var. americana (Scop.) J.-E. Gilbert, and is confused with A. arkansana Rosen. Amanita hemibapha is edible and is commonly sold in the markets of southeastern Mexico imder the names "tecomate" and "yullo" because of a similarity in the reddish orange color of the pileus to a wild cucurbitaceous fruit called by those names. Calostoma cinnabarina and Calvalia cyanthiformis are often found in the sub- tropics of Mexico, the former in mesophytic forests and the latter in the mead- ows and grasslands of those forests. Calvatia is a popular edible mushroom called "hongo bola" (i.e., ball mushroom). Spores from old basidiomes are used fre- quently to heal wounds. Chiodecton sanguineum, also known as Cryptotheca rubroci- nata (Ehrenb.) Thor, is, with its conspicuous reddish margin, the only lichen treated here. It occurs so regularly in Mexican mesophytic forests that it can be used as an ecological indicator for climax conditions of subtropical humid for- ests in Mexico (Guzman, 1994). Acknowledgments The author thanks his friends of the Gulf States Mycological Society, William Cibula, Toby Feibleman, Dorothea Munchow, Odin Toness, and A. L. Welden for their kind help during and after the foray in Mississippi and the field trips in Louisiana. He also thanks his assistant, Fidel Tapia, for help with microscopic observations. LiTERATl'RE ClTKO Berkeley, M. A. 1867. On some new fungi from Mexico. Jour. Linn. Soc. Bot. 9: 422-42.0. Castillo, J. and G. Guzman. 1970. Estudios sobre los polipon'm-os de Nuevo I.t-on, II. Obsei-\aciones sobre las especies conocidas y discusiones acerca dc su distrit)uc i(')n en Mexico. Boi. Soc. Mcx. :^1: 1- 48. Corner, E. J. H. 1966. A monograph of die Canlhari-Iloid fungi. Oxford I'niversitv Press. London {2m pp. + b pis.). Cunningham, G. H. 196.^>. Polyporaceae of New Zealand. Bull. New Zealand Depart. Sci. Indus. Res. 164: 1-.S04. GUZM4n, G. 1973. Some distributional relati()iisliii)s iHlwetn Mexico and L nited States mycollora. Mycologia 6.5: 1.319-1.3.30. Guzman, G. 1983. The genus Psiloryhi'. Beih. Nova I ledwigia. Cramer. \'aduz (4.39 pp. + 40 pis.). GuzmAn, G. 1994. Algunos aspectos imporlantes en la ecologia de los bongos (en especial de los macro- micetos). Ecologica. 3(2): 1-9. 74 Tulane Studies in Zoology and Botany Guzman, G. and L. Guzman-Davalos. 1992. A checklist of the lepiotaceous fungi. Koeltz Scientific Books, Champaign (216 pp.). Guzman, G., L. Montoya, G. Mata, and D. Salmones. 1994. Studies in the genus Pleurotus. III. The vari- eties of P. ostreatus complex based on interbreeding strains and the study of basidiomata obtained in culture. Mycotaxon 50: 365-378. GUZM\N, G. AND S. H. Pollock. 1978. A new bluing species of Psilocybe from Florida, U.S.A. Mycotaxon 7: 373-376. Guzman, G. and A. Sampieri. 1984. Nuevos datos sobre el bongo comestible Cantharellus odoratiis en Mexico. Bol. Soc. Mex. Mic. 19: 201-205. Pegler, D. N. 1983a. The genus Lentinus. A world monograph. Kew Bull. Add. Ser. 10. H. M. S. O., Lon- don (281 pp.). Pegler, D. N. 1983b. Agaric flora of the Lesser Antilles. Kew Bull. Add. Ser. 9. H. M. S. O., London (668 pp. + 27 pis.). Petersen, R. H. 1979a. Notes on cantharelloid fungi, X. Cantharellus conflitens and C. lateritius, Craterellus odojatus and C. aureus. Sydowia 32: 198-208. Petersen, R. H. 1979b. Notes on cantharelloid fungi, IX. Illustrations of new or poorly understood taxa. NovaHedwigia3I: 1-23. S.\N M.\RTIN-Gonzalez, F. and J. D. Rogers. 1989. A preliminary account of Xylaria of Mexico. Myco- taxon 34: 283-373. Sharp, A. J. 1948. Some fungi common to the highlands of Mexico and Guatemala and eastern United States. Mycologia 40: 499-502. Welden, a. L. and p. a. Lemke. 1961. Distribution of some Mexican fungi in North America. Ainer. MidL Naturalist 65: 111-117. SOME re.\llo(:ations of tvpe localities of reptiles and AMPHIBIANS DESC:RIBED FROM THE MAJOR STEPHEN 1 1. LONC; EXPEDITION TO THE ROCKY MOUNTAINS, WITH COMMENTS ON SOME OF THE STATEMENTS MADE IN THE ACCOUNT WRITTEN nx EDWIN JAMES R\ROLD A. DL'NDEE Tulfint' l'uii'(rsil\ Mtisriim of Xatinri/ f[i\l(in\ HcUr C.hnssc, I A 70037-309S Abstract The Stephen H. Long Expedition to the Rockv Mountains was the first expe- dition to the West that had trained biologists in its party. In this paper that sec- tion of the expedition from Council Bluff to the Rockies and retiun to Cape Girardeau is reviewed, and various type localities and dates of collection of type specimens are corrected or elaborated. These include Elaphe obsoleta. Coluber con- strictor flavwentris, Thamnophis proximus, Thamnophis sirtalis parictal/.s, Sistriirus cat- enatus tergeminus, Masticophis jlagellum testaceus, Cnemklophorus tessellatiis, Biifo cognatiis, Crotaphytiis collaris, and Scincella lateralis. Comments on the genus Bipes and horned lizards, genus Phrynosoma are included. Introduction The Major Stephen H. Long expedition from Pittsburgh to the Rocky Moun- tains in 1819-1820 was a milestone in the early exploration of the American West. For the first time actual scientists were made part of an expedition, these being Thomas Say, zoologist, William Baldwin, botanist, Augustus E. jessup, geologist, and Titian Ramsay Peale, an artist and taxidermist as "assistant natu- ralist." Baldwin, unfortimately, became ill and left the expedition at Franklin, Missouri Territory (Missouri did not become a state undl 1821), and Jessup resigned in the fall of 1819. They were replaced in June 1820, at Council Bluff (site of Omaha, Nebraska and Coiuicil Bluffs, Iowa) by Edwin James, a phvsician, botanist, and geologist. To allow the reader an appreciation of the problems posed by the account of the expedition, a brief history seems in order. Stroud (1992) suggested that "Although Say, as senior scientist, most likely should have assumed the task of compiling his writings with those of Long, Bald- win, and James, James took over instead. Probably Say was too involved with wi it- ing American Entomology and editing and writing for the Journal ol the Academy of Natural Sciences. Nevertheless he was much occupied with editing the expedi- don's account." Furthermore, "James had apparently very much wanted the job of writing the expedition's account, most likely because he was virtuallv penni- less at the time. But he told his brother that he felt 'whollv incompetent to it. It has been as you know a matter of my own seeking and if I shall herealter be thought to have failed in it, on my own head be the mischief.'" Stroud (op. cit.) also says that James indicated he would probably be "able to shift a part of the responsibility from myself by the proper construction of the dde," which he did: Tulane Studies in Zoology and Botany 30: 75-89. 1996 75 76 Tulane Studies in Zoology and Botany [Vol. 30 the title page states that the account is "compiled from the notes of Major Long, Mr. T. Say, and other gentlemen of the party." The account of the expedition was published in two places, one (James, 1823) was published by Carey and Lea in Philadelphia, and with some differ- ences, a little later (1823) in London by Longman, Hurst, Rees, Orme, and Brown (persons citing one or the other should note the minor differences in the titles). The London edition contains on page 36 a statement referring to the Philadelphia edition, thus the Philadelphia edition is the earlier printing. Although both accounts are dated 1823, the Philadelphia edition, at least, actu- ally had an 1822 printing of an Atlas (I have not seen the Atlas, which contains four pages of text plus plates and figures) and, in fact, the 1823 account had a claim of propriety registered with the Chief of the Eastern District of Pennsylva- nia in 1822 (Thwaites, 1905). Another account, much abbreviated and with little of the natural history, was written by Captain J. R. Bell (Bell's official report apparently was lost or suppressed by the War Department but by a fluke his diary was discovered by Harlin M. Fuller, in 1932, whose college roommate had men- tioned an old diary passed on within his family). Bell's account has important value for dates because he noted some event, no matter how trivial, for each day of the expedition. Fuller used the diary as the basis for a master's degree at Stan- ford University and later, with Leroy R. Hafen, published an edited version (Fuller and Hafen, 1957). Goodman and Lawson (1995) also mention that Bell's dates are more reliable than some in James' account. And herein begins some of the problem with the accoinit. Returning from Colorado, the party divided into two groups, one led by Long, with James in accompaniment, to find the headwaters of the Red River, but which mistakenly followed the Canadian River, and another group with Bell and Say to follow the Arkansas River downstream to Fort Belle (Fort Smith), Arkansas Territory. Say's manuscripts and descriptions of species were stolen on the night of August 30, 1820, by three deserting soldiers who took the horses that carried Say's saddlebags. At that time the party was located about ten miles south of Tulsa, Oklahoma, near the Arkansas River. Where Say's notes made sub- sequent to the August 30 theft might be is unknown. To complicate matters, the collected specimens were placed in the Charles Willson Peale Museum in Phila- delphia, a precedent set earlier by Thomas Jefferson, who ordered that speci- mens from the Lewis and Clark Expedition be deposited there. The Peale Museum went bankrupt in 1843 and six years later the museum items were bought at auction by P. T. Barnum and Moses Kimball (Stroud, 1992). Barnum's material was destroyed by fire in 1851, but most of the Kimball part eventually went to the Museum of Comparative Zoology at Harvard; Jose Rosado of that Museum says (in litt.) that they have no specimens identifiable with the Long Expedition. The lost Barnum holdings probably included the types of species described by Say. The account written by James thus is partially done from mem- ory because of the theft of Say's notes. Yet the detailed descriptions in the account, e.g., scale counts of snakes, suggest that the specimens Say collected had tags with data on them. Stroud (1992) says "Baldwin recounted that Mr. Peale has painted most of Say's fishes and amphibia," but this was before they left Pittsburgh. Stroud (op. cit.) also says "It was fortunate that Say had such a good memory — probably trained by the numerous literary quotations he had once committed to it — for he was able to reconstruct most of the information stolen with his saddlebags. Luckily he still had the notes he had taken up to and 1996] Long Expccliiion I ype Localities 77 including his stay at Engineer Cantonment" (campsite used in winter of 1819- 1820). In Thwaites (1905, vol. 17), Long is quoted as sa\ing that "Most of the collec- tions made on the expedition have arrived at Philadc'li)hia and are in good pres- ervation . . ." Long's account is mainly a report on the topograph) and Indians (he was attached to the Topographical Engineers of the army); but Long men- tions mammals and a casual reference to abimdance of reptiles and says that Say's specimens were shipped from New Orleans. Long also reports that, "The sketches executed by Mr. Peale amounted to one hundred and twenty-two. Of these, twenty-one only were finished; the residue being merely outlines of quadrupeds, birds, insects, &c." Dates are sometimes different, but Bell's diary apparently has the accurate dating because he had made an entry for every day of the trip. Chittenden (1902, vol. 2, p. 578) says of the itinerary, "It would scarcely be possible to find in anv narrative of Western History so careless an itinerary, and in a scientific report like that of Dr. James it is inexcusable." Typically, when an animal was obtained by Say he immediately would take notes on its appearance. However, where descriptions of new species are presented by James, they may not necessarily be at the point in the chronological account where the animal was taken. James, in fact, bitterly commented about Major Long, whom he apparently disliked; in a letter to his brother from Cape Girardeau, he stated that Long allowed him neither time to examine and collect, nor means to trans- port plants or minerals. "We were hurried through the country as if our sole object had been, as it had been expressed in the orders which we received at starting 'to bring the expedition to as speedy a termination as possible'" (Stroud, 1992: 123). One must suppose that Long pushed as rapidly on the entire expedition, thus descriptions may have been delayed until adequate dme allowed for a presentation, or that details from memory could not be made in clear association with dates or places. For the present writing I have mostly used the microcard version of the Phila- delphia edition of James' account, and in place of the London edition I have used the version of Thwaites (1905). My interest in the Long expedition resulted from the publication of Goodman and Lawson (1995), which, though primarily concerned with the botany of the trip, has proved most fruitful in establishing the probable routes taken and dates of encampment. The text of James' account is difficult to follow because one is not sure whose notes are being utilized, even though the writing styles are different. The Philadelphia edition is more com- plete, and also has an index. Goodman and Lawson deal only with the trip from Council Bluff, and emphasize the botany, nothing of the animals, but they do give dates, maps, and photographs to verify the illustrations made by the expedi- tion artist Samuel Seymour. I do not deal with that part of the expedition from Pittsburgh to St. Louis, only with the St. Louis area, the trek westward, and the return to Cape Girardeau, Missoiui, an area visited previously during the St. Louis stopover. T\TV. Localities Ehij)he obsoleta This is the first herpetological species described from the trek westward from St. Louis. Dowling (1952) gives the statement "on the Mi-ssouri River from the vicinity of Isle au Vache to Council Bluff," this from the Philadelphia account. 78 Tulane Studies in Zoology and Botany [Vol. 30 Dowling then says, "Cow Island is near Leavenworth Comity, Kansas, and is somewhat more than 100 miles downriver from Council Bluffs, Iowa. At present it appears unnecessary to restrict the type locality further. Should this be neces- sary at some later date, it could reasonably be restricted to the vicinity of Cow Island, where Say carried on his investigations for some time." Schmidt (1953) gives the type locality as "Isle au Vache to Council Bluffs on the Missouri River; rest, to Council Bluffs on the Missouri River." Stejneger and Barbour (1943) say for Elaphe obsoleta obsoleta, "Isle au Vache [sic] (the "sic" is S. and B.'s wording) to Council Bluffs on the Missouri River." On pages 139-140 (vol. 1) of James (1823) in the Philadelphia edition, which is an accoimt of the area around Isle au Vache, James reported (obviously from Say's notes) "We have seen at Bellefontain [an old military post near St. Louis], as well as at several other points on this river, a pretty species of sparrow, which is altogether new to us;* and several specimens of a serpent have occurred, which has considerable affinity with the pine snake of the Southern States or bull snake of Bartram [Bartram, 1791].*" The terms pine snake and bull snake would refer to genus Pituophis. The asterisks indicate descriptions at the bottoms of pages 139 and 140. On page 139 the description of the sparrow Fringilla grammacea Say begins and states "Shot at Bellefontain on the Missouri," and on page 140, foot- note 163 refers to the description of Coluber obsoletus. At this point the accoimt is referring to animals seen at Bellefontain. The actual description is of a black snake of "anterior half with a series of dull-red large circles, formed upon the skin between the scales on the side," and the animal is said to resemble Coluber constrictor '"hut the scales are decidedly smaller, and the number of its plates and scales approach it still more closely to that imcertain species C. ovivorus " [ C. ovi- vorus Merrem 1820: 134, but the spelling there is ouiourus and a footnote indi- cates Linn S.N. I, p. 385]. I think the reference is to Tomus I of the 12th edition of Linnaeus' Systema Naturae where ovivorus appesirs on page 385. The reference also says "C.N. 203+73" [the ventral scute count of ouiourus is much below that of obsoleta but would fit Elaphe vulpina, whose range currently reaches just north of St. Louis north of the Missoin i River, and the subcaudal coimt of 73 exceeds the 71 maximum reported by Powell (1990) for Elaphe obsoletus]. "It is not an uncom- mon species on the Missouri from the vicinity of Isle au Vache to Council Bluff." That the sparrow was described much later than it was collected is a clear indica- tion that descriptions are not necessarily associated with point of collection, and, of course, might also be an indication that Coluber obsoletus yN-aiS taken from Belle- fontain. The statement "It is not an uncommon species. . . ." sounds like an after- thought and makes Bellefontain seem a logical type locality. What was Say seeing at Bellefontain? Terms such as pine snake and bull snake would apply to blotched specimens, hence possibly Elaphe guttata emoryi at the edge of its currently known range, or juvenile Elaphe obsoleta. Eampropeltis c. calli- gaster and Elaphe vulpina are ruled out because their highest ventral counts are below the minimum count for E. obsoleta. Anderson's (1965) map shows the northern range of E. guttatta as being a little south of Bellefontain. Bartram's (1791) pine snake, however, would be the Florida variety, whose blotches are obscure on the anterior half of the body but the Floridian version of Pituophis melanoleucus does not occur anywhere near St. Louis, and P. m. sayi, the species that does occur near St. Louis, is vastly different in pattern. At this point I must accept the "Isle au Vache" type locality, but the puzzle is what Say was seeing at Bellefontain where they stayed for several days. That some 1996] Long Expedition Type Localities 79 degree of blotching was evident in the Elaphe at Isle an Vache could be a confir- mation of what Say saw at Bellefontain and assumably Say would have captured some of the snakes seen at Bellefontain. Coluber constrictor flaviventris, Tliamnophis sirtalis parietalis, Thamnophis proximus The expedition spent the winter of 1819-1820 near present-day Omaha, Nebraska, and Council Bluffs, Iowa. The winter quarters (James vol. 1, p. 146 of the Philadelphia edition) "was on the west bank ol the Missouri, about a half mile above Fort Lisa, five miles below Council Bluff, and three miles above the mouth of Boyer's River." The site, called Engineer Cantonment, was about 100 yards northwest of a limestone quarry. Sometime between December 12 and December 30, 1819 (both the Philadelphia and London editions describe events of December 12, but without any other accoimting; the next date shown is December 30, so the exact date cannot be determined), laborers at the quarry opened fissures in which a number of snakes were hibernating. Included were three kinds considered to be new species: Coluber parietalis {Thamnophis sirtalis parietalis), C. flaviventris (C. constrictor flaviventris) and C. proximus {Thamnophis proximus). Setting the site of Engineer Cantonment as the type locality seems easy enough. The latest statements of type locality are, for Thamnophis sirtalis parietalis, "Stone quarry on west side of Missouri River three miles above mouth of Boyer's River, Washington County, Nebraska . . . collected by Thomas Say 1819" (Rossman, 1980); for Coluber constrictor flaviventris Wihon (1978) says, "west bank of the Missouri . . . three miles above the mouth of Boyer's River; restated by Rossman (1963: 109) as 'approximately 3 miles ENE Fort Calhoun, Washington County, Nebraska'." Wilson adds, "collected by T. Say on 12 Decem- ber 1819." Wilson's literature citation for the James account says Peary and Lea (sic), an obvious error for Carey and Lea. For Thamnophis proximus, Rossman (1970) gives ''Coluber proximus Sny 1823: 339 . . . Stone quarry on west side of the Missouri River, 3 miles above the mouth of Boyer's River," restated as "approxi- mately 3 miles ENE Fort Calhoun, Washington County, Nebraska" by Rossman (1963: 109). Rossman (1963) actually says, "The specimen was collected in Nebraska at a stone quarry on the west side of the Missouri River, 3 miles above the mouth of Boyer's River (Iowa) by Long's expedition to the Rocky Mountains in 1819 or 1820. The type locality lies approximately 3 miles ENE Fort Calhoun, Washington County, Nebraska." Rossman (1963) also indicated that Smith and Taylor (1945) listed the holotype as lost, but it was probably deposited in the Academy of Natural Sciences of Philadelphia, though this is not a certainty (H. M. Smith, personal communication to Rossman). As stated above, we know that the expedition's specimens were to be deposited in the Charles Willson Peale Museum. Goodman and Lawson (1995, p. 3, footnote 4), say that the mouth of Boyer River once entered the Missouri about 3/4 mile below the Washington-Douglas county line but has now been shifted via modern feats of engineering upstream five miles. Thus, to allow for the original Boyer River situation, the type locality for all three species described from the quarry, and with respect to the map and statements on page five of Goodman and Lawson, I hereby restate the type local- ities as about five miles south of Ft. Calhoun, Section 28, T17N, R13E in Wash- ington County, Nebraska. 80 Tulane Studies in Zoology and Botany [Vol. 30 Sistrurus catenatus tergeminus According to Minton (1983), the type locality of Sistrurus catenatus tergeminus, described by Say as Crotalus tergeminus, is "indefinite; restricted to Winfield, Cow- ley County, Kansas [Smith and Taylor, 1950]. No holotype designated." In my estimate, this assessment is erroneous. Grant that Cowley County is well within the range of that subspecies, but the expedition account of August 17, perhaps August 18, 1820, contains no mention of rattlesnakes. The actual description that appears on page 499 (vol. 1) of the Philadelphia edition is for the date of July 4, 1820, when the party was in Adams County, Colorado. Stejneger and Bar- bour (1939) did not recognize 5. c. tergeminus. Schmidt (1953) says, "Locality not stated; designated as headwaters of Arkansas River." Gloyd (1940) says "Type locality indefinite; no type specimens designated," and Klauber (1956) says, "Type locality: between the Mississippi River and the Rocky Mountains; type specimen: none designated. Type locality subsequently restricted to Winfield, Cowley County, Kansas, by Smith and Taylor . . ." But many references to Crotalus tergeminus appear at earlier dates in the text, and such a clear statement of encounters with it is given that I believe I can give a more specific locality. In volume one, page 344 of the Philadelphia edition appears the following: "On the 12th [April 1820] Lt. Graham, Lt. Talcott of Camp Missouri, Mr. Sey- mour and I, accompanied by a soldier, departed in our small row boat, for the purpose of ascending Boyer Creek [Boyer River], and ascertaining the point at which that stream discharges from the Bluffs." James, we must note, did not arrive at the Council Bluff winter encampment until June, and he remarks in a footnote (p. 200) that succeeding notes, especially about the Omawhaws [Omaha Indians], and the last chapter of volume one and the first five chapters of volume two were from Say's notes; the statement of the trip up the Boyer starts off chapter six, but clearly the "and I" refers to Say. The writing style, the lack of specificity about plants, and the detailing of animals clearly constitute Say's writing style. The account says for April 14: "We saw numbers of the smaller species of rattle snake, [p. 375, lists reptiles seen at Engineer Cantonment, the campsite of the winter of 1819-1820, or at other indicated places on their jour- ney. That list includes Crotalus horridus, a much larger species and C. durissus (probably a misnomer for some specimens of C. horridus)] which had, no doubt, but lately left their winter dwelling." On April 17 they returned to their tent where a soldier had killed three rattlesnakes while pitching the tent (the tent was erected April 15 and the party had to put blankets on their backs to travel because of the bluffs on the river) . The opportunity was ripe for securing and preserving specimens; thtis, a possible type locality is along the Boyer River in Crawford or Harrison County, Iowa, most likely Harrison County because the account suggests that they did not get too far upriver. But as I will show below, a more likely type locality is in Colorado. In Nebraska today, the range of Sistrurus catenatus tergeminus is restricted to a small section of the extreme southeastern part of the state (Filmore, Lancaster coundes, and Gage and Nemaha coundes in 1891 according to Hudson, 1942). The James account gives several other references to the species in Nebraska and also associates them with prairie dog towns. As indicated below, I am surprised that they did not indicate that they were encountering or also encountered the prairie rattlesnake, Crotalus viridis, which usually is the rattlesnake associated with prairie dog towns. To further back up the expedition's report that Sistrurus 1996] Long Expedition Type Localities 81 occurred in prairie dog towns is the statement at the top of vohime two, page 48 of die Philadelphia edition; it leads, "We also captured a rattlesnake*, which like the tergeminus, we have found to inhabit a barren soil, and to frequent the villages of the Arctomvs of the prairie [= prairie dog], but its range appeared to us conhned chieflv to the vicinity of the Rocky Mountains . . ." The asterisk refers lo the description of Crotalus confluent us [C. viridis] on page 48. Admittedly the failure to encounter C. viridis in Nebraska is puzzling, but perhaps that is because all of Say's notes covering that region were lost with the stolen saddlebags and the account from memon' mav not ha\'e been as good as one would like. What few prairie dogs {Cynomys ludovicinnus) remain in Nebraska occur relatively far east, biu do not overlap the ciurent range of Sistrunis catenatus tergeminus. As late as 1900 a prairie dog colony was located near Fontanelle in Washington (bounty, about 25 miles northwest of Omaha and the records show also that the prairie dog reached Jeffer- son Countv (Jones, 1964). But Nebraska is a heavily agricultural state and the tran- sition of prairie to farmland may perhaps have eliminated Sistrurus from the bulk of the state. This is not an easy conclusion to reach because Sistrurus catenatus has survived agricultural activity in Illinois, Kansas, and Oklahoma. On June 28 (vol. one, p. 485 of the Philadelphia edition), when the party was near Sterling, Logan County, Colorado, the accoimt says, "About some sandy ridges, which we passed in the middle of the day, several miliary rattle snakes were seen, two of which were killed. These had been occasionally met with all along the Platte, but were by no means numerous." The term "miliary" is refer- ence to Sistrurus miliarius, the pigmy rattlesnake of the southeastern states. On July 4 (vol. one, pp. 499-500 of the Philadelphia edidon), at which time the party was somewhere between Platteville, W^eld County, Colorado and just soiuh of Brighton, Adams County, the account reads: "Rattle snakes of a particular spe- cies* [the * indicates the description of Crotalus tergeminus on p. 499] are some- times seen in these villages [piairie dog villages]. They are found between the Mississippi and the Rocky Mountains, and appear to prefer unproductive soil, where their sluggish gait ma\ not be retarded by the opposing obstacles of grass and weeds. W^hilst exploring Boyer Creek, of the Missouri, in the spring of 1820, our party met with six or eight of them during one day's march on the prairie, and on our subsequent journey to the Rocky Mountains we several times encountered equal or even greater numbers, in the same space of time. This is the species of serpent which travellers have observed to frequent the villages of the prairie dogs, and to which they have attributed the unnatural habit of \olun- tary domiciliation with that interesting animal. . . . From the disparity in the number of plates and scales, and the greater size of the vertebral spots in this species than in the C. miliarius we have been induced to consider this a distinct species. Specimens are in the Philadelphia museimi." Because they were still referring to "miliary ratUesnake" on June 28, and the July 4 description seems to be applied to snakes seen at that time, I believe that this is the most appiopriate association of a date, place, and description to define the type locality of Crotalus tergeminus, notwithstanding the fact that opportiniity to collect the snakes occurred first on the Boyer Creek trek in Iowa and that the localitv is north of present-day distribiuion. Masticophis flagellum testaceus Mention of this snake is made in different places in the two versions of James' account and differ somewhat. The Philadelphia edition mendons the dis- 82 Tulane Studies in Zoology and Botany [Vol. 30 covery and description (vol. one, p. 48), the description immediately following that of Crotalus confluentus, whereas the London edition, as seen in Thwaites' presentation, reports Coluber testaceus in volume three, page 236 (p. 39 of Thwaites), and Crotalus confluentus appears on page 40 of Thwaites. From the reading of the James account, one might glean that this snake was taken on July 17, possibly July 18, but Bell, who was exacting in covering every day no matter how trivial a report, says clearly that hunters caught a large red snake on July 16. Wilson (1973) gives the type locality as "headwaters of the Arkansas near the Rocky Mountains = junction of Turkey Creek with the Arkansas River, 12 miles W Pueblo, Pueblo County, Colorado. Holotype lost." Other than the minor date detail, this seems adequately accurate when compared with the different accounts that mention the place where the species was discovered. I note that Wilson (1973) erroneously lists Carey and Lea as Peary and Lea as the publisher and fails to indicate that the type locality, as relating to Pueblo, was more specifi- cally defined by Maslin (1953). A casual sketch of a snake's head made by Titian Ramsay Peale on 16 July 1820, which appears to be that of a Masticophis flagellum, is in the collection of Peale's drawings at the American Philosophical Society. Cnemidophorus tesselatus In the Philadelphia account by James this species is described as "Genus Ameiva. A. Tesselata Say. Tessellated Lizard." The London edition, published slightly later, says "Genus Ameiva — A. tessellata. Say Tesselated lizard." The holotype is lost. Price (1986) cited the Philadelphia account; he indicates that literature up to 1949 used either spelling and also that Burt (1931) treated C. tigris and C. graha- m?iasjunior synonyms of'testellatus'' (Burt had emended the name to "tessellatus" based on the adjectival spelling and justified by article 39 of the International Code of Zoological Nomenclature). Whether the Philadelphia edition has a typo- graphical error probably is a moot question, and article 32 of the International Code of Zoological Nomenclature (1985) probably would allow the single "1" spelling to prevail. Price (op. cit.) says that '"tesselatus is derived from tessella (L., "litde square stone")." Checklists, e.g., Stejneger and Barbour (1943) or Schmidt (1953) use "" tessellatus," obviously considering the single "1" to be an error. The entire matter of the nomenclature takes on a new perspective in a paper by Walker et al. (in press). What has been called Cnemidophorus tesselatus involves both diploid and parthenogenedc lizards, and Walker et al. show that the origi- nal type locality today apparently has no lizards that match the type description; hence they establish a neotype and new type locality for tesselatus and give the population at the original type locality a new name. James' account is not clear on where the lizard was collected. Various authors have discussed the type locality, (Smith and Burger, 1949; Smith and Taylor, 1950; Maslin, 1950; Milstead, 1953) the problem being the names of creeks encountered en route, but the restriction of the type locality to Pueblo, Colo- rado by Maslin (1959) seems reasonable in terms of the general area through which they were passing. Maslin ascertained that the type locality should be restricted to "the junction of Fountain Creek and the Arkansas River, Pueblo Co., Colorado." Goodman and Lawson (1995) confirm the idendty of the creek. Bufo cognatus Krupa (1990) gives the type locality as "The alluvial fans of the [Arkansas] River" in Prowers County, Colorado. The actual statement (James, vol. two, p. 1996] Long Expedition Type Localities 83 190, Philadelphia edition) is "The alluvial margins of the river are gradually dilating as we descend . . ." The date of collection is July 29, 1820, and likely the specimen was taken aroinid the camp in the evening because heavy rain had fallen on the 28th. From Bell's account of July 29 (Fuller and Hafen. pp. 204- 205) the partv apparently was a few miles west of Holly, Prowers (>ouni\, which I here define as the type locality: Approximately 3 miles west of Holly, Prowers County, Colorado. Various checklists (e.g., Schmidt, 1953) have defined it as Powers County, but the correct name is Prowers (>ounty, the name used by Mas- lin (1959), Smith (1934), and Stejneger and Barbour (1943), among others. Crolaphytus collaris This species, described as Agama collaris in voliune two, page 252 of the Phila- delphia version of James, is mentioned in the September 5, 1820, narrative: "A beautiful species of lizard, (agama) is occasionally met with in this territory." Axtell (1989) discussed the type locality, which is an enigma. Axtell pointed out that James indicated (p. 251) that the party [the Bell party that followed the Arkansas River] had arrived at Mr. Glen's trading house near the Verdigris River about a mile above its confluence with the Arkansas. Stejneger (1890, p. 103) expressed the first opinion on the type locality: "The type of Say's Agama collaris came from the Verdigris River, near its junction with the Neosho River, Creek Nation, Indian Territory." Stejneger gave no explanation for his restriction. Stej- neger and Barbour (1917, 1923,^^1933, 1939, 1943) gave the restricted locality as "Verdigris River near its union with the Arkansas River, Oklahoma." Axtell goes on to say that most workers have accepted that type locality, or changed it only slightly, but Webb (1970) gives the type locality as "near Colonel Hugh Glenn's [both the London edition of James' account and Bell's account spelled it Glenn] Trading Post on the east bank of the Verdigris River about two miles above its confluence with the Arkansas River." According to Webb, this site is in Wagoner County, Oklahoma, about four miles ENE of Ft. Gibson (Muskogee County) near the Muskogee-Wagoner County Line. Note, however, that Ft. Gibson is east of the Grand River (= Neosho River) and the site indicated by Webb would be in either Cherokee or Muskogee County. More correctly he should have said WNW of Fort Gibson. Axtell (1989) states that he visited the vicinity of Webb's inferred type local- ity, where the Verdigris has an exposed bluff where Glenn's Trading Post might have been, but no one living in the vicinity remembered anything about the post or its location. At the nearby town of Okay, Oklahoma, the science teacher and other residents Axtell questioned knew the lizard by the Oklahoma vernacular name of "mountain boomer," but agreed that they had not seen or heard of them in the vicinity of Okay. Axtell thus chose not to accept Webb's version because of the still unanswered site questions (personal connniuiication) and provisionally accepted the Stejneger and Barboin- (1917) type localitv. I con- tacted Webb, who said that he well-remembered discussing the matter with Dr. A. O. Weese, an ecologist at the University of Oklahoma, who had edited the journal for that part of the expedition chronicled by Titian Ramsay Peale and suggested the type locality that Webb used, but that so far as Webb knew, sub- stantiating information has not been published. In anv case, I note that Webb (1970) showed records for Crotaphytus collaris in both Wagoner and Muskogee counties. The Bell party arrived at the trading post at 10 A.M. and left that morn- 84 Tulane Studies in Zoology and Botany [Vol. 30 ing, hardly time enough for Say to have done any collecting, and their subse- quent trek was through low country, hardly a suitable habitat for Crotaphytus. However, as noted for September 3 and 4, the account tells of passing rocky- topped hills. The map shown in Fuller and Hafen (1957), apparently con- structed by the Bell party, shows the campsites. When the latitudes and longi- tudes of the map are compared with those of a modern-day map, the campsites of September 3 and 4 are NW of Muskogee, Oklahoma, or perhaps at Muskogee. Both sites are south of the Arkansas River, and definitely not along the Verdigris River. Although the map coordinates shown by Bell are inaccurate, the account is clear in relation to the Verdigris River. In July, 1996, I searched around Tulla- hassee and Porter, Wagoner County, areas just west of Okay, areas that I suspect may have been the type locality; I can verify that suitable habitat for Crotaphytus exists west of Okay. The crux of the type locality problem, I believe, lies in how one reads the statement "A beautiful species of lizard (agama), is occasionally met with in this territory." My feeling is that the term "territory" really refers to the fact that Oklahoma, at that time, was part of Arkansas Territory. That the lizard was met with occasionally does not mean that it was necessarily seen and captured near Okay. I believe that I can infer that the lizard was likely taken in some of the rug- ged terrain west of Okay, possibly even as far northwest as Tulsa. But because the party reached the summit of the hills that divide the drainages of the Arkansas and Verdigris, and with the large hills near the Grand (Neosho) River in sight, I would guess that the lizard was secured perhaps somewhere southwest or north- west of Tullahassee, Wagoner County, Oklahoma on 4 September, 1820. In view of the ambiguities associated with the narrative, the Stejneger and Barbour (1917) statement seems the best that can be said, except that I would add that the type locality should read, "Verdegris River . . . Oklahoma, probably in Wag- oner County." Scincella lateralis In volume two, page 324 of the Philadelphia account (no mention of the spe- cies is given in the London account presented by Thwaites, 1905) this species was described as Scincus lateralis. Brooks (1975) gives for the type locality "banks of the Mississippi River below Cape Girardeau [Cape Girardeau County] (the brackets are those of Brooks) Missouri." Brooks then goes on to say: "Holotype, U.S. Nad. Mus. 3152, collected by S. W. Woodhouse (not examined by author)." The expedition had been to Cape Girardeau on the way west; they returned to Cape Girardeau October 10, 1820, and James (Philadelphia edition, p. 324) says "about November 1, 1820 Mssrs. Say, Graham, and Seymour had so far recovered their health, as to venture on a voyage to New Orleans, on their way home." Each of those men had apparently contracted malaria. "They left Cape Girardeau in a small boat, which they exchanged at the mouth of the Ohio River for passage on a steam boat [Riverboat Yankee (Stroud, 1992: 124)] about to descend. Mr. Peale, who had escaped the prevailing sickness, accompanied them. On his way down the Mississippi, Mr. Say observed the new animal described in the subjoined note.*" A footnote follows, continuing onto page 325, in which Scincus lateralis is described. The account (p. 325) says of Scincus lateralis: "occurred during our passage down the Mississippi to New Orleans, and proved to be specifically identical with specimens which 1 formerly collected in Florida and East Georgia." No mention is made of Woodhouse or of any land- 1996] Long Expedition 1 ypc Localities 85 ings on the banks of the Mississippi. Furthermore Samuel W. Woodhouse was born in 1821, after the expedition was completed, and could not have collected the skink. Also, we must recall that specimens collected bv the expedition were to be deposited in the Charles Willson Peale Museinn in Philadeli)hia, not in Washington. Stejneger and Barbour (1917, 1923) called the animal Leiolopisma lalerale (Say), but later (1939) called it Leiolopisma unicolor (Harlan). That change, Stej- neger (1934) explained, was because Scincus lateralis Say was preoccupied, hav- ing been used by Daudin (1803). Lonnberg (1896) showed that the specimen was actually a Mediterranean species, Gongylus occulatus (Forskal, 1775), and that the available name for the North American species was Scincus unicolor Harlan (1827). The publication date for Harlan's name was corrected by Stejneger and Barboiu- (1939) to 1825, but they used the name Leiolopisma unicolor, giving the type locality as "Southern states." Schmidt (1953), however, gives the credit to Say, 1823, in Long's expedition and as volume two, page 324, with the site as "Banks of the Mississippi below Cape Girardeau, Missouri." I contacted Dr. George Zug at the U.S. National Museum of National History, who said that Doris (Doris Cochran, 1961) listed USNM 3152 as a type, but that she had listed a fair number of specimens that are not types. Further, he said, her specimen locality data far exceed those in the catalog. The catalog entry is ''Plestiodon Arkansas River Dr. Woodhouse 21 1." Zug said, "reidentified, pencil entry in Doris' handwriting: Mo added to locality, also in pencil by Doris to note date when specimen was entered in the catalog." Zug thinks the specimen imlikely the type, although we have no way of tracking its origin beyond the t)riginal cata- log entry. Thus the original version in James' account must stand, and the type locality should read "Likely just below Cape Girardeau, Cape Girardeau County, Missouri" with credit given to Say, 1823. I searched a nimiber of habitats that seemed suitable for Scincella just below Cape Girardeau in July, 1996, but did not encounter the species. Miscellaneous Commentary One of the most remarkable things mentioned in the Long Expedition account appears in volume one, page 484 of the Philadelphia version for Jinie 27, 1820. At that time, the party apparently was in what we now know as Logan Comity, Colorado, between Crook and Sterling (Goodman and Lawson, 1994) near the Platte River: "We observed in repeated instances, several individuals of a singular genus of reptiles (Chirotes. Cuv.) which, in form, resemble short ser- pents, but are more closely allied to the lizards, but being ftunished with two feet. They were so active that it was not without some difficulty that we suc- ceeded in obtaining a specimen. Of this (as was our imiform custom, when any apparently new animal was presented) we immediately drew oiu a descrijjtion. But as the specimen was unfortimately lost, and the description formed part of the zoological notes and observations, which were carried off by our deserters, we are reduced to the necessity of merely indicating the probability of the exist- ence of the Chirotes lumbricoides of nMnvdWsis, within the territory of the United States." Modern nomenclature would place (Ihirotes in the genus Bipes, a lizard genus known only from western Mexico and Baja (California. A justification of this remarkable observation was suggested by Campbell (1980). Support for the validity of the observation was presented by Dimdee (1980), who was told by a farm boy in the western Nebraska sandhills, and apparently familiar with all of 86 Tulane Studies in Zoology and Botany [Vol. 30 the lizards on his home area, that two-legged lizards were occasionally found there, especially when digging post holes or pulling tip old fence posts. But Gans and Papenfuss (1980), who worked extensively with amphisbaenians, to which group Bipes belongs, report that all species they have encountered were rather sltiggish and easily caught and doubted very much that the Long party had encountered such an animal. Supplemental to the Long Expedition accotuit are Taylor's (1938) remarks strongly suggesting that Bipes might occur in Arizona. But how do we explain the observation? Say was noted for being meticulotis in his observations and the statement "We encountered" stiggests that indeed other members of the party saw the animal. Clearly some very intensive collect- ing in the Logan County area and the Nebraska sandhills is needed to try to ver- ify the expedition's narrative. Campbell (1980) did suggest that possibly the species may have been relict in the Nebraska-Colorado area and perhaps close to extinction. Even if the original notes had not been stolen, the lack of a voucher specimen would still place the observation in doubt. A remarkable oversight of the Long Expedition seems to me to be Say's fail- ure to describe a species of Phrynosoma; a horned lizard certainly is an animal not easily forgotten and the party obviously saw Phrynosoma because, on several occasions, the account refers to orbicular lizards, the term used in those days for horned lizards. On or about July 15, 1820, probably near Fountain Creek, per- haps two miles south of Fotmtain, El Paso Comity, Colorado (Philadelphia edi- tion, vol. 2, p. 35) the account says, "Orbictilar lizards were found about this camp, and had been once or twice noticed near the base of the motintain." In the Philadelphia edition (vol. two, p. 51) is the statement, "We had proceeded eight or ten miles from our camp. . . . Here the barren cedar ridges are suc- ceeded by still more desolate plains, with scarce a green, or living thing upon them, except here and there a tuft of grass, an orbicular lizard, basking on the scorching sand. . . ." This was July 19, probably close to Pueblo, Colorado. Ear- lier, on August 6, 1819, Say and a party journeyed westward up the Kansas River (Kaw River) and in the Philadelphia edition (vol. one, p. 138) the account says: "In ascending the Konzas river [Kansas River], one hundred, or one hundred and twenty miles from the Missouri, you discover numerous indications, both in the soil, and its animal and vegetable productions of an approach of that Great Sandy Desert, which stretches eastward from the base of the Rocky Mountains. You meet there with the orbicular lizard or 'horned frog', an inhabitant of the arid plains of New Mexico. "James' and Long's party, on August 1, 1820, after the parties had split, were on Ute Creek east of Yates in Harding County, New Mex- ico and volume two, pages 89-90 of the Philadelphia edition say, "Several rattle- snakes were seen and many orbicular lizards. These are evidently of two distinct species, differing from each other in the length of the spines and position of the nostrils. Scarce any two of either species are precisely similar in colour, but the markings are permanent. Both species possess in a slight degree, the power of varying the shades of colour. W'e could find no conspicuous difference marking the different sexes in the species with long spines; the other we have not have sufficient opportunity to examine." Jtidging by today's horned lizard distribu- tion, they most likely were seeing Phrynosoma cornutum and P. douglassi. From so many mentions of orbicular lizard, one would think that stirely Say would have described at least one, possibly two species, the latter if James' and Long's party had collected specimens. Coues (1895, vol. 2, p. 431) says that Zeb- ulon M. Pike wrote in his journal for October 24, 1806: "and strange as it may 1996] Long Expedition Type Localities 87 appear, I have seen the wishtonwish, the rattlesnake, the horn frog [Phrynosoma douglasi, with which the prairie abounds . . .] (The brackets are by Coues)." Coues is wrong about Phrynosoma douglnsi (sic) because on October 24 Pike was between Larned and Great Bend, Kansas; donirJassi does not occur in Kiuisas, only P. cornutum. Say quite likely had seen the paper by Barton (1806), who asso- ciated a specimen taken by the Lewis and Clark Expedition with Lacerta orbicu- laris (Linnaeus, 1758). Price (1990) has pointed out that L. orbicularis is a senior synonym oi Phrynosoma orbiculare, and that Barton (1806) called his specimen "Lacerta Tapajaxin" but deferred a full description to another paper. That paper (Barton, 1809) is a description of a salamander, Ambystoma maculatum. Thus Bar- ton's name L. Tapajaxin is a nomen nudum for Phrynosoma cornutum, and credit for P. cornutum goes to Harlan (1825). Most likely Say thought that the orbicular lizards encountered were a species that Barton had already described. Had the two different forms mentioned by the Long and James party in New Mexico come into Sav's hands, quite likely Say would have described at least one horned lizard. All in all, a reader of the Long Expedition can only curse the deserters who fled with Say's saddlebags; otherwise an accurate zoological account would have been a significant contribution from the expedition team. Acknowledgments My appreciadon is extended to Drs. George Zug and Robert G. Webb for pro- viding useful information on some of the more puzzling species accounts. Dr. Ralph W. Axtell for reading over parts of the manuscript, and to Patricia Tyson Stroud for clarifying some of the historical parts of the account. I also thank Mr. Eric Wedig of the Tulane University Howard-Tilton Library for allowing me to borrow^ the Philadelphia version of the James account for office use. Others who kindlv checked for specimens or missing notes include Scott DeHaven of the American Philosophical Society, Carol Spawn and Ted Daschle of the Academy of Natural Sciences, and Jose Rosado of the Museum of Comparative Zoolog)', Harvard University. Literature Cited Anderson, P. 1965. The Reptiles of Missouri. Univ. Missouri Press, Columbia. Axtell, R. W. 1989. Interpretive Atlas of Texas Lizards No. 8. Crotnphylus rollnris. Prixatelv printed, Edwardsville, Illinois. Barton, B. S. 1806. Miscellaneous facts and observations. Natural history'. Philadelphia Med. Phys. Jour. 2, pt. 1, 1st suppl. 2nd sec: 6.5-7.S. Barton, B. S. 1809. Some account of a new species of North American lizard. Trans. Amer. Philos. Soc. (o.s.) 6: 108-112. Bartrwi, W. 1 791 . Travels through North and South Carolina, Georgia, East and West Florida . . .James and Johnson, Philadelphia. Brooks, G. R. 1975. Scincella lateralis. Cat. Amer. Amph. Rept. 169.1-169.4. Burt, C. E. 19.'^1. A studv of the teiid lizards of the genus Cnt'wiflofi/ionis whb special reference to ilu-ir phylogenedc relationships. U.S. Nat. Mus. Bull. (154). viii + 286 pp. Campbell, J. H. 1980. Edwin James's report on /?//;« reconsidered. Jour Herpetol. 1 1: (>7. Chittenden, H. M. 1902. The .\merican Fur Trade of the Far West . . . vol. 2. F. P. Harper, New York. Cochran, D. 1961. Type specimens of reptiles and amphibians in the United States Nadonal Museum. U.S. Nad. Mus. Bull. No. 220. xv + 291 pp. CoUES, E. 1895. The Expeditions of Zebulon MontgoiueiT Pike to Headwaters of the Mi.ssissippi Ri\c-r . . . During the Years 180,5-6-7. A new edition. Francis P. Harper, New York. Dundee, H. A. 1980. A comment on J. Howard Campbell's Bipes report. Herpetol. Rev. 1 1 : 74, 76. 88 Tulane Studies in Zoology and Botany [Vol. 30 DOWLING, H. G. 1952. A taxonomic study of the ratsnakes, genus £/a/»/?/' Fitzinger. IV. A check list of the American forms. Occ. Pap. Mus. Zool. Univ. Michigan 541: 1-12. FORSKAl, p. 1775. Descriptiones Animalium . . . Orienti. Hauniae [Copenhagen]. Fuller, H. M. ,\ND L. R. R\FEN (eds.). 1957. The Journal of Captain John R. Bell: Official Journalist for the Stephen H. Long Expedidon to the Rocky Mountains, 1820. The Arthur H. Clark Co., Glendale, California. Cans, C. .\ND T. P,\PENFUSS. 1980. There is no e\ddence that B i pes occuia in the U.S. Herpetol. Rev. 11: 74. GLCiT), H. K. 1940. The rattlesnakes, genera Sislrurus and Crotalus. Chicago Acad. Sci. Spec. Publ. 4: 1- 266. Goodman, G.J. and C. A. Lawson. 1995. Retracing Major Stephen H. Long's 1820 Expedidon. The Itin- eraiT and Botany. Univ. Oklahoma Press, Norman and London. Harlan, R. 1824 (1825). Descripdon of two new species of Agama. jour. Acad. Nat. Sci. Philadelphia 4: 296-304 + 2 pi. R\RL.\N, R. 1825. Descripdon of a variety of the Coluber /(//tvh.v Linn., a new species of Sciticus, and two new species of Salamandra. Jour Acad. Nat. Sci. Philadelphia 5: 154-158. Hudson, G. E. 1942. The Amphibians and Reptiles of Nebraska. Univ. Nebraska Conserv. and Surv. Div., Nebraska Conserv. Bull. No. 24. iv + 146 pp. lNTERNATlON.\L CODE OF ZOOLOGICAL NOMENCLATURE (3rd ed.). 1985. Internauonal Trust for Zoologi- cal Nomenclature, London. J.AMES, E. 1823. Accoimt of an Expedition from Pittsburgh to the Rocky Mountains, Performed in the Years 1819, '20 . . . Under the Command of Major Stephen H. Long. 2 vols. H. C. Carey and I. Lea, Philadelphia. James, E. 1823. Account of an Expedidon from Pittsburgh to the Rocky Mountains, Performed in the Years 1819, 1820 . . . Under the Command of Maj. S. H. Long, of the U.S. Top. Engineers ... 3 vols. Longman, Hurst, Rees, Orme, and Brown, London. Jones, J. K., Jr. 1964. Distribution and taxonomy of mammals of Nebraska. Univ. Kansas Publ. Mus. Nat. Hist 16: 1-356 + 4 pi. Klauber, L. M. 1956. Ratdesnakes: Their Habits, Life Histories, and Influence on Mankind. Vol. 1. Univ. California Press, Berkeley and Los Angeles. Krupa, J. J. 1990. Bufo cogiuiius. Cat. Ainer. Ainphib. Rept. 457.1-457.8. Linnaeus, C. 1758. Systema Naturae per Regna Tria Naturae . . . Edido decima, reformata. Laurentii Salvii, Holmiae [Stockholm]. Linnaeus, C. 1766. Systema Naturae per Regna Tria Naturae . . . Editio duodecima, reformata. Tomus 1. Laurentii Salvii, Holmiae [Stockholm]. Lonnberg, E. 1896. Linnean t\pe-specimens of birds, repdles, batrachians and fishes in the Zoological Museum of R. University' in Upsala 22, sec. 4(1): 1-42. M.\SLIN, T. P. 1950. Herpetological notes and records from Colorado. Herpetologica 6: 89-95. M.A.SLIN, T. P. 1959. An annotated check list of the amphibians and repdles of Colorado. Univ. Colorado Studies, Ser Biol. 6: 1-98. Merrem, B. 1820. Testamen Systematis Amphiborum. J. C. Kriegeri, Marbvrgi [Marburg]. MlLSTE.\D, W. W. 1953. Ecological distribudon of the lizards of the La Mota Mountain region of Trans- Pecos Texas. Texas Jour. Sci. 5: 403-415. MiNTON, S. A. 1983. Sistrurus catenatus. Cat. Amer. Amphib. Rept. 332.1-332.2. POW'ELL, R. 1990. Elaphcvulpina. Cat. Amer Amphib. Rept. 470.1-470.3. Price, A. H. 1986. Cnemidoplwrus tesselatus. Cat. Amer. Amphib. Rept. 398.1-398.2. Price, A. H. 1990. Phrynosoma comutum. Cat. Amer. Amphib. Rept. 469.1-469.7. ROSSMAN, D. A. 1963. The colubrid snake genus Thamnophis: A revision of the 5fl!/n7M5 group. Bull. Flor- ida State Mus. Biol. Sci. 7: 99-178. RosSMAN, D. A. 1970. Thamnophis proximus. Cat. Amer. Atnphib. Rept. 98.1-98.3. R0SSM.\N, D. A. 1980. Thomiwphis sirialis. Cat. Amer. .\mphib. Rept. 270.1-270.4. Schmidt, K. P. 1953. A Check List of North American Amphibians and Reptiles (6th ed.). Amer. Soc. Ich- thyol. Herpetol. Univ. Chicago Press, Chicago. Smith, H. M. 1934. The amphibians of Kansas. Ainer Midi. Nat. 15: 377-528. Smith, H. M. and W. L. Burger 1949. The identit)' of Ameina lesselnta. Bull. Chicago Acad. Sci. 8: 277- 284. Smith, H. M. and E. H. Tayi.OR 1950. Type localities of Mexican reptiles and amphibians. Univ. Kansas Sci. Bull. 33, pt. II.: 313-380. Stejneger, L. 1890. Annotated list of repdles and batrachians collected by Dr. C. Hart Merriam and Ver- non Bailey on the San Francisco Mountain Plateau and Desert of the Litde Colorado, Arizona, with descriptions of new species. N. Amer. Fauna 3: 103-118 + 2 pi. 1996] Long Expedition Type Localities 89 Stejneger, L. 1934. Sf incus lalnalisSay, preoccupied. Copeia \\KV\: 1.S2. StejNEC.ER, L. and T. B.\RB()l'R. 1917. A Clu'ck List of North Aim-rican Ajnphibians and Rci)tilcs. Har- vard Univ. Press, Cambridge, Massachusetts. Stejneger, L. .\ND T. BarboL'R 1923. A Check List ot North American Amphibians and Reptiles (2nd ed.). Hanard Univ. Press. Cambridge, Massachusetts. Stejneger, L. .\ND T. B.\RB0LR 1933. A Check List of North .\merican Ampliibians and Reptiles (3rd ed.). Hanard Univ. Press, Cambridge, Massachusetts. Stejneger, L. .\ND T. B.ARBOLR 1939. A Check List of North American Amphibians and Reptiles (4th ed.). Harvard Univ. Press, Cambridge, Massacluisetts. Stejneger, L. and T. B.\RB0UR. 1943. A Check List of North American Amphibians and Reptiles (5th ed.). Bull. Mus. Comp. Zool., Harvard College. Cambridge, Massachusetts. Stroud, P. T. 1992. Thomas Say: New World Naturalist. Univ. Pennsylvania Press, Philadelphia. Taylor, E. H. 1938. Does the amphisbaenid genus Bipes occur in the L'nited States? Copeia 1938: 202. Thwaites, R. B. (ed.). 1905. Early Western Travels. Vols. 14-17. Account of an Expedition from Pitts- burgh to the Rocky Mountains Performed in the Years 1819, 1820. . . . The Ailhur H. Clark Co., Cleveland, Ohio. Webb, R. G. 1970. Repules of Oklahoma. Univ. Oklahoma Press, Norman. W,\1.K.ER, J. M.,J. E. Cordes, .\ND H. L. Ta\'IX)R. (in press). Parthenogenetic Cnemidophorus tesselatus com- plex (Sauria:Teiidae): A neotype for diploid C. tesselatus (Say, 1823), redescription of the taxon, and description of a new triploid species. Herpetologica 1997 (2). Wilson, L. D. 1973. Masticophis Jlagellum. Cat. Amer. Amphib. Rept. 145.1-145.4. Wilson, L. D. 1978. Coluber constrictor. Cat. Amer. Amphib. Rept. 218.1-218.4. ^62 3 2044 093 361 103