ii,>'K ■Ti}' 4^; ■* m^ HARVARD UNIVERSITY \M/ Ernst Mayr Library of the Museum of Comparative Zoology ISSN 0082-6782 TULAME STOPSLs) m^ mcz I iBRARY ;0©LOGY AMD BOTAMY OCT 0 3 1989 i lARVARD Volume 27, Number 1 $7.00 IJrJliV€.lfE3Pf9f' THE DISTRIBUTION AND STATUS OF THE FLATTENED MUSK TURTLE, STERNOTHERUS DEPRESSUS (TESTUDINES: KINOSTERNIDAE) CARL H. ERNST DEPT. OF BIOLOGY, GEORGE MASON UNIVERSITY, FAIRFAX, VIRGINIA 22030 WILLIAM A. COX DEPT. OF BIOLOGY, MARION MILITARY INSTITUTE, MARION, ALABAMA 36756 KEN R. MARION DEPT. OF BIOLOGY, UNIVERSITY OF ALABAMA AT BIRMINGHAM BIRMINGHAM, ALABAMA 35294 THE PREVALENCE AND INTENSITY OF PERKINSUS MARINUS FROM THE MID NORTHERN GULF OF MEXICO, WITH COMMENTS ON THE RELATIONSHIP OF THE OYSTER PARASITE TO TEMPERATURE AND SALINITY THOMAS M. SONIAT DEPARTMENT OF BIOLOGICAL SCIENCES AND CENTER FOR BIO-ORGANIC STUDIES UNIVERSITY OF NEW ORLEANS, LAKEFRONT, NEW ORLEANS, LA. 70148 AND JULIE D. GAUTHIER DEPARTMENT OF BIOLOGICAL SCIENCES, UNIVERSITY OF NEW ORLEANS LAKEFRONT, NEW ORLEANS, LA. 70148 Tulane University New Orleans TULANE STUDIES IN ZOOLOGY AND BOTANY, a fully refereed publica- tion of the Biology Department of Tulane University, is devoted mainly to pa- pers containing original studies in ecology and systematics, without geographical constraints. Reviews, bibliographies, and symposia results are occasionally pub- lished. Normally two numbers plus an index and table of contents are issued annually. Preferred citation of the journal is Tulane Stud. Zool. and Bot. INFORMATION FOR AUTHORS. Manuscripts should be double-spaced throughout, including Literature Cited and legends to tables and figures, on 8 1/2 X 11 paper, with margins no less than % inch from top, sides, and bottom. Do not use dot matrix printing. Avoid footnotes; if a footnote is necessary, place it on a separate sheet of paper. Do not use hyphenated words on line endings. Manuscripts should conform to the style of recent issues of Tulane Studies, and figures should be proportioned for one or two columns corresponding to our printed page. Lettering used in illustrative material should be easily read after reduction to our page size. Guidelines for letter and other extraneous markings should be done with a red pencil, not with the traditional blue. We recommend conformance to the CBE Style Manual. Two additional copies of the manuscript should accompany the original. Legends for figures should appear on a separate page and in sequence. An abstract, briefly stating the results of the research, and not to exceed three percent of the length of the article, must accompany the manuscript. Page charges of $45.00/page will be assessed. Authors are urged to provide for page charges through grants or institutional support. Financial assistance from the George Meade Memorial Fund is available if needed. EXCHANGES, SUBSCRIPTIONS, ORDERS FOR INDIVIDUAL COPIES. Exchanges are invited from institutions publishing comparable series. A price list of back issues is available on request. Remittances should be made payable to "Tulane University". Subscription rates are $12.50 per volume domestic, $13.50 per volume foreign. COMMUNICATIONS: Address inquiries about exchanges to Head, Science and Engineering Division, Howard-Tilton Library, Tulane University, New Or- leans, LA 70 11 8. Editorial and all other matters should be addressed to Editor, TSZ&B, Department of Biology, at the same address. Alfred E. Smalley, Editor Printed by E. O. Painter Printing Company, P. O. Box 877, DeLeon Springs, Florida 32130. THE DISTRIBUTION AND STATUS OF IHE FLATTENED MUSK TUR I LE, STERNOTHERUS DEPRESSUS (TESTUDINES: KINOSTERNIDAE) CARL H. ERNST DEPr. OF BIOLOGY, GEORGE MASON UNIVERSITY, FAIRFAX, VIRGINIA 22030 WILLIAM A. COX DEPT. OF BIOLOGY, MARION MILITARY INSTITUTE, MARION, ALABAMA 36756 KEN R. MARION DEPT. OF BIOLOGY, UNIVERSITY OF ALABAMA AT BIRMINGHAM, BIRMINGHAM, AIABAMA 35294 Abstract A survey of the distribution and local popula- tion densities of the flattened musk turtle, Sler- notherus depressm, was conducted in the Black War- rior River drainage of Alabama in 1983. A total of 577 S. depressus was collected at 36 (53%) of the 68 sites sampled; 18 (26%) sites had moderate to high density populations, and 8 (12%) other sites supported populations regarded as between low to moderate in density. One 1.6 km stretch in the Sipsey Fork, Bankhead National Forest, Winston County, supported the highest density population yet recorded: 237 S. depressus were collected in 3 days. Several of the sites from which the animal previously was reported to have been extirpated or absent produced specimens during the present study. Extensive sand deposits were a major phys- ical component of most of the high density sites, contrary to what has been considered optimal S. depressus habitat. Major threats to the species were determined to be agricultural influence, surface mining, forestry practices, commercial collecting, and municipal and industrial pollution. State legislation (1984) now prohibits collection other than by permit for scientific study. Because popu- lation numbers are low or appear to be declining across large portions of the species' range (includ- ing habitats formerly occupied), we conclude that the status of S. depressus should be monitored. However, we believe that the available evidence, including the recent discovery of several high den- sity populations in the northern part of the range, does not support a threatened or endangered list- ing at the time of this writing. Additional observa- tions on habitat, daily activity patterns, predation and health are also presented. Introduction The flattened musk turtle, Ster- notherus depressus, was described by Tinkle and Webb (1955) on the basis of specimens collected "in the Mulberry Fork of the Black Warrior River, 9 miles east of Jasper, Walker County, Alabama, near the bridge crossing of U.S. highway 78 — a sluggish tributary of the Black Warrior River." Since then S. depressus has been collected by scien- tists only sporadically, and its geog- raphical range has been determined to be permanent streams of the Black War- rior River system above the Fall Line (Iverson, 1977). Thus, its entire range lies within Alabama and, more impor- tantly, within the Warrior Coal Basin. The secretive nocturnal nature, re- stricted range, reported commercial exploitation in recent years, and general lack of information regarding this turtle have led some conservationists and her- petologists to question the overall popu- lation status of the turtle and its chances of continued survival (Iverson, 1977; Mount, 1975, 1981). Within the range of the flattened musk turtle extensive surface mining, forestry, and agricul- ture occur, and several of the major streams have been impounded. Iverson (1977) thought that elimination of fast water environments by impounding was detrimental, and Mount (1981) sug- gested that acid runoff from mines and siltation from mining, logging and ag- riculture have been harmful to the tur- tle. Unfortunately, little is known re- garding the basic biology of i\ depressus or the status of existing populations. Studies on the species' biology and dis- 1 Tulane Studies in Zoology and Botany Vol. 27 tribution are limited to triose of 1 inicle and Webb (1955), Tinkle (1958), Es- tridge (1970), Close (1982) and Mount (1981). The Office of Endangered Species, United States Fish and Wildlife Service, in 1981 contracted Dr. Robert H. Mount, Auburn University, to deter- mine the status of the flattened rnusk turtle. In his report, Mount (1981:80- 81) concluded that "(1) depressus has probably been extirpated from about 220 miles of its former stream habitat; (2) about 101 miies of habitat have been degraded to the extent that the proba- bility oi depressus occurring ana/or being able to maintain viable populations within it is very low; (3) about 349 miles of habitat are marginal in supportive ca- pacity, and that while aepressus may now occur, its ability to survive under the current conditions ranges from ques- tionable to doubtful; and (4) about 142 miies of free-flowing stream habitat are of sufficiently high quality to continue to support healthy, viable populations of depressus." According to Mount (1981:81), the single greatest threat is siltation, and he placed the major blame for this on surface coal mining. On the basis of Mount's report, a federal pro- cess to list 5. depressus as '"threatened" was begun. The present study was in- itiated in an attempt to provide addi- tional data regarding the current status of the species and to gain some knowl- edge of the waterways in whicii it lives. It should be noted that Seidel et al. (1986) have recently proposed placing the genus Sternotherus in the synonomy of Kinosternon. For purposes of general familiarity, however, this paper retains the taxon Sternotherus. Methods and Materials Field work extended from 15 June 1983 until 9 September 1983. Initially (15 June-ca. 15 July), stream conditions were poor for collecting because of high water levels. Rainfall during the first half of the year within the Warrior Basin averaged 22.5-25.0 cm above nor- mal. 1 hereafter, precipitation de- creasea aramatically and water levels subsequently declined. Collecting condi- tions had greatly improved by 15 July and generally remained conducive for trapping and/or wading for the remain- der of the study period. Several methods were utilized in col- lecting specimens of 5. depressus. These are in order of decreasing effectiveness: trapping, nocturnal wading with a lan- tern, diurnal wading, seining, raking and seining through weed beds, and snorkeling. Traps used were very similar to those previously utilized and described by Mount (1981; after Iverson, 1979) and differed only in that 18 mm mesh rather than 25 mm mesh chicken wire was used. Several designs of traps con- structed trom corrugated plastic pipe (20 crn and 30 cm diameters) were tested following discussions with several persons wno recommended traps with dark interiors. These were later dis- carded, in favor of the lighter and more easily manipulated wire traps, when re- sults indicated no significant differences in success Detween the two types. Traps were set in a wide variety of microhabitats between 1500 hours and 1900 hours and recovered early the fol- lowing morning. Depending upon logis- tics, between 30 and 60 traps were set per night (normally 15-20 traps per site). In most instances, traps were spread over a 300-500 m stretch of stream and left undisturbed for 12 to 18 hours. Successful settings ranged in depth from 25 cm to 450 cm and in- cluded areas of extremely fast current to pools with virtually no detectable cur- No. 1 Distribution and Status of Sternotherus depressus rent. Most traps, however, were set at depths ranging from 45 cm to 150 cm in areas with slow to moderate current. We observed that traps placed with one of the long sides flush to the bank, a large boulder, a crevice in the bedrock or a submerged log were, on the aver- age, more effective. A variety of baits were tested early in the study to determine if one bait type was more effective than others. Chicken parts, chicken livers, various types of canned cat food, canned sar- dines, canned herring, and native mol- luscs (snails and mussels ground or sliced and placed in a porous bait pouch) were compared. Canned her- ring and native molluscs appeared to be the least effective attractants. Because there were no observable differences in trap-capture rates between the other types of baits, canned sardines were used predominantly during the major- ity of the study because of convenience, cost and relative effectiveness. Several references were used to as- sist in the preliminary selection of sam- pling sites. These included: Alabama County Highway Maps; Geological Sur- vey of Alabama, Atlas Series 12, "Drain- age Areas for the Upper Black Warrior River Basin, Alabama (1978)"; U.S.G.S. 7 1/2 minute quadrangle maps; U.S.G.S. 1:250,000 Base Maps (Birmingham and Gadsden); county surface mining maps provided by the Alabama Surface Min- ing Commission; and the USFWS publi- cation on the status of 5. depressus by Mount (1981). Final selection of sampling sites (Fig. 1) was based on two major factors: Location within the Warrior Basin — The need to provide the broadest pos- sible coverage of known or poten- tial 5. depressus habitats was consid- ered. Although practical consider- ations prevented a true random distribution oi sites, the overall "scatter" of sampling sites throughout the basin yielded an approximate random distribution. Specific site selection — Determination of the precise locality for sampling involved consideration of several factors, such as access to a stream, size of the stream, permission of landowner and time limitations. Isolated areas, such as those where access was limited or where access could be gained only by crossing private land, were preferred. Bridge sites were, of necessity, utilized in many instances. Of these, bridge access sites that re- ceived minimal disturbance from human recreational activities were normally preferred. Access to sev- eral sites could be gained only by boat. Fifty-seven percent of the streams examined were sampled at more than one site, for one or more of the follow- ing reasons: a) The length of the stream or size of the impoundment necessitated multiple sampling sites in order to compile as accurate an assess- ment as possible (e.g., Locust Fork, Mulberry Fork, Smith Lake). b) The presence of particularly good 5. depressus populations along selected stretches of a stream made it advisable to sam- ple multiple sites. In this way, dif- ferences in habitat quality could be compared to population esti- mates at specific sites within the same stream (e.g., Blackburn Fork of the Little Warrior River, Blackwater Creek). c) The absence of good 5. depressus populations within some streams demanded the sampling of more Tulane Studies in Zoology and Botany Vol. 27 than one site to minimize the chances of deriving a population estimate which was artificially low (e.g.. Clear Creek, Lost Creek, North River, Gurley Creek). In several instances, a particular loc- ality was investigated more than once. Most were areas which appeared to meet the habitat requirements for flat- tened musk turtles, but at which trap- ping results were initially low, primarily due to rain-related stream conditions. At some sites, subsequent efforts re- vealed the presence of a larger popula- tion than would have been suspected after only one attempt (e.g., Turkey Creek, Blackburn Fork, Blackwater Creek). At others, low population den- sity estimates or the apparent absence of the species was supported (e.g., Gur- ley Creek. Clear Creek, Duck Creek, Lo- cust Fork, Mulberry Fork). Some sites yielded excellent results immediately, and subsequent efforts were made to more accurately estimate population density and population numbers (e.g., Sipsey Fork, Brushy Creek). A general description of each site was prepared for future reference, and pertinent habitat characteristics noted. These included dominant terrestrial and aquatic vegetation present, charac- teristics of banks, width of stream, odors, vertebrate and invertebrate as- sociates, water clarity estimates, bottom characteristics, column profile, stream current, dissolved oxygen content (ppm), air and water temperatures, and selected water quality parameters. A modified Wentworth Scale was employed to determine bottom particle size as follows: boulders = 256 mm or larger; cobble = 64-256 mm; pebble (includes both pebble and granule categories) = 2-64 mm; sand = indi- vidual particles visible to the unaided eye up to 2 mm; silt and clay = fine particles individually invisible to the un- aided eye. The preceding scale was used to quantify stream bottom characteris- tics by estimating the percentage of the total for each particle size category pres- ent. Water depth is also an important habitat characteristic for 5. depressus, as it fluctuates considerably within a stream, both geographically and tem- porally. Therefore, a column profile was constructed at each sight. The col- umn profile is a proportional represen- tation of various depth categories within the water column. Each of these is ex- pressed as a percentage of the total within the area sampled. The following categories were established: a) riffles = shallow areas with fast current and pos- sessing a roiling surface turbulence; b) <15 cm depth; c) 15-60 cm depth; d) 60-150 cm depth; e) 150-240 cm depth; and f) >240 cm depth. The exact location, description, and above physical characteristics of each site have been compiled into an Appen- dix which will be made available to legitimate researchers by the authors upon request. Figure 1 depicts the ap- proximate location of the sites sampled. Captured flattened musk turtles were taken to the laboratory, assigned an individual identification number by notching the marginal scutes, meas- ured, weighed (to the nearest 0.1 g), photographed and released within 24 hours at the capture site. Previous dam- age to or erosion of other marginal scutes was noted. Linear measurements (to the nearest 0.1 mm) made with dial calipers were: carapace length, carapace width, plastron length, anterior plastron width, and shell height. Body mass was obtained using a Sartorius top-loading electronic balance (accuracy = ± 0.003 g). Photographs of each captured indi- vidual were taken from a position di- rectly above the animal. No. 1 Distribution and Status of Sternotherus depressiLs 5 10 15 20 ^.,^3 0 5 10 15 20 25 30 Kilometers Fig. 1. Map of the Black Warrior River drainage basin showing approximate location of sites examined during 1983. Sites are identified by number. Exact locations and descriptions of sites are available Irom the authors upon request. Town abbreviations: ALBV = Albertville; ARAB = Arab; BHM = Birming- ham; BZ = Boaz; CLM = Cullman; HLV = Haleyville; JSP = Jasper; ON = Oneonta; ISC = Tuscaloosa. Additional data recorded for each turtle included the presence of old or recent injuries or lesions; sex; the pre- sence of algae, leeches or other epizoics; the presence of oviductal eggs (deter- mined by palpation); whether or not fecal samples were obtained; and the date, site and method of capture. Interviews with landowners, local re- sidents or others familiar with a particu- lar area or stream often enhanced our understanding of local conditions. On five occasions, we received comments from people who possessed knowledge of important information spanning 20- 30 years. Results and Discussion A total of 577 Sternotherus depressus was collected (Table 1) and approxi- mately 20 others were sighted but es- caped. Of those collected, 509 were caught in traps in 20,169.7 trap-hours Tulane Studies in Zoology and Botany Vol. 27 u a re ^ w O JD be •- OJ ■b T3 {/3 (1 i! '^ < s: "O be u c a. fe LI "O 3 « S Q- ?= -a C/5 a, O ^ X) -r; 3 u c re QJ .bf) bbO ^ 3 1; C c'3- o . CO 00 'c« C5 ^ r^ l~^ u o re u >, n 1- u:. m E H > 3 bf) C/5 c/) ■-< x: u hf) -J c pa u < H ■o re re <*- h u CO c o o CT5 O o o o CO t: be be c c re -J ^ (u Q ^^ 2 + re U c?5 — ' '^ CO Ol — c —I in If) tr> CO OJ CO CO o o o »n CO mooocM-Hootoomoio-Hooco-^oo— 'O— 'Or^io lOirjtri— '-^r^ooocTiooioc^jcMtocD— 'CMco— icoo — ooocM (mco-^cx3-*-hoo^ cvjtMinTtiG^i^-^co-Hc^icri-Hinc^j — -* CO (TJ — CM m m CM — • — — CO 00 CO CO m CM CM Tp C^ ^ ■^ CO to in 00 o to 00 CO CO O O CT) — . CM — r^ m OM o QO ■^ in to i> to to CO in CO CM oi in CM ~ O CO CT> CO +1 + 1 +1 +1 +1 in to CO to CO en CO rf CO O^ 00 ci t^ 00 ^ CM in t^ in o 00 CT5 in r^ r^ O I> CO to CM 00 in CO CM Ol t^ -^ -" in CO CO + 1 +1 +1 +1 CM in 00 TT ^ O) 00 o o a^ CO 00 ai CD 1 en in 00 CT5 +1 +1 CT) Tf 00 00 CO 05 to in o o CO in in — CM CO 00 r^ to CM CO CO CM C^J CM CM — — -H ^ CM — CM mcMTf to in-*co cO'^ — cm CM M in CM O' o o -1 co 21 00 CM 00 CO +1 +1 '— o Oi t^ CM ^ rt in — CO Tj> — t^ -"t !> CO CM o ■* tt cn Tf o in CO — o^ -H -. Tj- OOtT — OC7>OtO 00 CT) ■^oor^TfooCT>coooo CM in o o o r^ a^ CM o ^1 " — 1 »— 4 in in en o> in 52- CO 2- o to to OI CTi — — o CM ■^ 00 CM + 1 +1 +1 +1 CM o r^ — < 00 IT- en 00 00 CM CT) en CM o Tj- o o o CM ^ — o — CMcoTtooeno — CMcoTfi intor^ooCTio — CMcoTt^incot^ooaio — CMco-^in No. 1 Distribution and Status of Sternotherus depressus in CO 05 ■<*< un 00 (£5 CO CO CM in CO ocMOin'^CT)Oc^dO-*oinococo-*int^ — — ''-Oin — CM^£>tD— 'CO ocotDincotccn— 'Ot^'^r^a;OCMtD-*m'^i>'f<^ooinoot^ocMa; (MCO — C^CMin-^ — CTl— '— — OJ — — — (MCMtC'*(MC0 04 — — ^OCCW — ' 00 oc CM — CM — CM ^ '^ to 'S' CM CM to 00 d CO d d s>- t>- (J) to CO CO to to in l> 00 CO CO 00 to CO ""-^ ' — ^ ^-^ >—*' ^-^ ' — ■ S_^ .,_• >_• s..^, ^^^- -«— ^ >— .^ ■*■ in c^i -^ o q c — ; q 'O -^ r-. r^ X in 00 iC OC' CM CM CO •-f iri CM -^ to -^ to Ol — ~- CM — ^- -M -M — +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 CT; C-i CTi — CTi 00 0(D CM O Tf CM CM — r^ o •t q tT — CO CM 00 d '^ Tt" in CO d d — CM oi to iri 05 o 00 to to 00 00 i> CTi O) t^ 1^ O) r- X 0-- X ^ — OCMO~ — cOOtOOOOOCMcOOOeOC^I — tO'OCOcO — ooo — — ~ — ^CM— — inx— -r tor^xo^o — CMco-^intot^xoo — CMcoTintot^xoio — CMco cocococoT}"-^-*'^T}Tf'^-<}<'^inininininininininintototO'0 "^ m to t- X to to to to to 8 Tulane Studies in Zoology and Botany Vol. 27 during 1250 trap-nights (mean = 16.1 trap-hours/night) for a rate of 0.41 tur- tles/trap-night. An additional 68 speci- mens were hand-captured while wading 265.75 man-hours (0.26 turtles/man- hour). Our rate of capture was not sig- nificantly different from that of Mount (1981), who had a yield of 0.4 turtles/ trap-night (272 trap-nights) and 0.23 turtles/man-hour wading (44.4 hours). Of the 68 sites sampled, 36 (53%) yielded S. depressus (Table 1). Mount col- lected them at 26 (38%) of 68 sites (ap- proximately half of these sites corres- pond to those sampled in the present study). Several of the 32 sites that did not yield S. depressus in our study had been previously listed as degraded, and the presence of the species considered unlikely by Mount. However, a number of sites where we found turtles were lo- cations where Mount found none (see discussion of waterway status below). Distribution and Status Table 2 assesses the status of 5. depre- ssus at each site. Because no historical population density data are available for the species, the categories in Table 2 are relative estimates of population density. Placement of sites into specific categories was based on 1) collecting re- sults, 2) effort expended, and 3) both qualitative and quantitative habitat evaluation. The following discussion of distribu- tion and status is presented by drainage from the southwesternmost to the northeasternmost. Refer to Fig. 1 and Table 1 for site locations and collecting results. A. Black Warrior River (beloiu junction of Locust and Mulberry Forks) and its tributaries. — In Bankhead Lake (Site 1), flattened musk turtles are fairly com- mon around the scattered rocky bluffs, TABLE 2. Status of the flattened musk turtle at sites examined during 1983. Refer to Fig. 1 for approximate site locations. Exact locations and descriptions of sites are available from the authors upon request. Status determinations are relative estimates of density based on 1) collecting results, 2) effort expended, and 3) both qualitative and quantitative habitat assess- ment. Population Assessment Sites Absent Absent to Low Low Low to Moderate Moderate Moderate to High High 2,21,60,63,64 7.8, 11, 13, 16, 17,20,25, 26, 3L 32, 33, 35, 36, 38, 40, 4L 46, 47, 48, 49, 52, 53,65,68 10, 15, 19,23,24,34,37,39, 55,59,66,67 1.9, 14, 18,43,50,54,61 3,4, 12,27,28 22,29,30,42,44,45,51,58 5,6,56,57,62 but are low in numbers or absent from the numerous mud flats. The overall as- sessment is that of a low to moderately dense population. However, access to suitable trapping sites is limited, and trapping at additional sites could reveal a larger population than we estimate. This agrees essentially with Mount's (1981) conclusions that 5. depressus is lo- cally common in Bankhead Lake. Hurricane Creek (Site 2), an eastern tributary, was the most degraded water- way sampled. No 5. depressus were col- lected, silt deposits were deep, coal fines were abundant, and the pH was 4.2. The only direct evidence of current pol- lution from surface mining runoff de- tected during the study was here, where nearby "abandoned mines" (according to ASMC inspectors) were still leaching into the creek. We agree with Mount's (1981) conclusion that essentially no 5. depressus survive in this stream. Healthy populations of moderate density were found at two sites on the No. 1 Distribution and Status of Stcrtwt/wrus dcprcssus North River (Sites 3, 4). However, runoff from agricultural areas and silta- tion are evident. Blue Creek is a stream in which Mount (1981) found no turtles, but he concluded that conditions there were acceptable, and thought 5. depressus probably well-represented. The popula- tions at Sites 5 and 6 (approximately 1 .6 km apart) are healthy, reproducing, and of high density. However, the stream does not fit the criteria given by Mount (1981) for optimal habitat. The stream bed is extremely sandy, with some boul- ders and rock crevices present. The original substrate is largely obscured by sand, up to 60 cm deep in places. Blue Creek is of particular interest since it is apparently recovering from severe past pollution as a result of nearby surface mining. Local residents told us that the turtle and fish faunas were "completely wiped out" over 20 years ago when the contents of a sedimentation basin flooded into it. Mining operations ceased about 10 years ago. The pre- sence of a high density population today apparently indicates that the species can naturally repopulate areas from which it has been eradicated. The turtles ap- parently immigrated either from areas above those affected by mining and/or the Black Warrior River, of which this stream is a direct tributary. The musk turtles taken from Blue Creek seemed intermediate in head pattern and shell depth between 5. depressus and 5. minor peltifer, but whether this represents hyb- ridization or merely intergradation is not known. Big Yellow Creek (Site 7) is a small, intermittent, silty stream that may in- frequently support 5. depressus near its mouth. Agriculture is prevalent in the area, and there are several old surface mines 5-6 km upstream. S. odoratus is ■^mmon in Big Yellow Creek and may outcompete S. depressus. Mount (1981) considered S. depressus to be scarce or absent in this waterway, and we concur. Mud Creek (Site 8), a heavily- sedimented tributary of Valley Creek, has historically been influenced by ag- riculture, surface mining and coal-re- lated industries. High water levels at the time of sampling may have contributed to our collecting no flattened musk un- ties there, but probably at best only a low density population exists at this site. B. Locust Fork and its tributaries. — Mount (1981) considered the ecological condition of the Locust Fork to be poor, and we concur. Overall, it is degraded; the upper portions are heavily impacted by agriculture and to a lesser degree by surface mining. In our opinion, agricul- ture has probably had the greater im- pact in the upper Locust Fork. Exten- sive algal growths cover the rock sub- strate in many areas, possibly indicating excess nutrient influx. Sediments in many places are also typical of those as- sociated with runoff from cultivated areas. Such sediments typically consist of fine particulates possessing a floccu- lar appearance. The possibility exists that pesticide/herbicide influence in these (and other) areas has had a de- bilitating effect. Coal fines and other evidence of mining-related activities are less obvious. The lower portions, how- ever, have been more severely affected by surface mining, and industrial and municipal pollution. In some areas, old surface mines abut the river's edge. Habitats in the extreme lower portion have also been modified by reservoir backup. All 12 sites sampled during the 1983 study resulted in below average collections, and all of these sites showed some evidence of degradation. The data presently available suggest that S. depre- ssus populations in the Locust Fork have 10 Tulane Studies in Zoology and Botany Vol. 27 declined. There are still low to moder- Turkey Creek (Sites 22, 23) has been ate density, reproducing populations at heavily influenced by surface mining Sites 9, 14 and 19. Populations at other and, to a lesser degree, by agriculture sites, however, such as 10, 11, 13, 16, 17 and other human activities (e.g., road- and 20, are very low in density or have building, construction, urbanization, been eradicated. Those at 12, 15 and 18 etc.). Mount (1981) considered the pre- are low density and may not be success- sence of S. depressus unlikely. Turkey fully reproducing. Previous scientific Creek at Site 23 is small and considera- coUecting on the Locust Fork indicated bly degraded, with a heavy silt load and moderate to high density populations at obvious coal fines; a large inactive sur- some localities (Mount, 1981; Cox, per- face mine is located upstream. Only a sonal experience; James V. Peavy, per- low density S. depressus population is sonal communication). In addition to present. Site 22, downstream from 23, habitat degradation, collection for the is also degraded with silt and coal fines, pet trade appears to have been a major but in all other respects appears to be factor in the decline at certain sites. The favorable habitat, having large boul- populations at easily accessible sites, ders, riffles and weed beds. There are such as 16 and 18, have been especially extensive sand deposits on the bottom; affected by collectors. Mount (1981) snails are abundant and Corbicula are provided comments on Site 18. At Site present. The 5. depressus population at 16, Cox collected 15 5. depressus in 10 this site is one of moderate density. If man-hours wading during 1977, and this population is left undisturbed, it has Peavy reports having seen numerous in- the potential for naturally restocking dividuals at this site during the period those sections of the stream that retain 1966-1976. We caught no 5. depressus at suitable habitat characteristics, this location during our survey, despite We trapped but did not wade three ideal collecting conditions. Since gas- sites on Gurley Creek (Sites 24, 25, 26) tropods and Corbicula (important food and caught no S. depressus. Historically, items in the diet of adult S. depressus) it has been affected by surface mining were usually abundant at most sites on and municipal pollution. Some seg- the Locust Fork during our study and ments of the stream are still degraded, the general physical characteristics (cur- while others appear able to support rent, rock substrate, refugia) appear ac- these turtles. Mount (1981) arrived at ceptable, the lack of success in obtaining essentially the same conclusions. Possi- flattened musk turtles at these and bly a low density population of 5. depre- other Locust Fork localities was startl- ssus is present, as potential food items ing. {Corbicula, snails, and aquatic insect lar- Village Creek (Site 21) has been vae) are moderately abundant and the highly polluted by both industrial and physical habitat appears acceptable, municipal effluents in the past, and ap- The Blackburn Fork (Sites 27, 28, pears degraded and presently eut- 29, 30) contains favorable habitat for 5. rophic. No snails or clams were ob- depressus, and the present overall density served, and it is unlikely that S. depressus is moderate to high. Populations are still exists in the creek. However, our more dense in the upper reaches. How- one collecting attempt was during a ever, the population appears to have de- period of high water. The fish fauna clined in recent years, especially at Site seems abundant and varied, and 28 (Cox and Marion, past experience), perhaps the stream is recovering. and this decline is probably attributable No. 1 Distribution and Status of Stemotherus depressvs 11 to commercial collecting. Site 28 has been extensively collected during the past ten years by amateurs and commer- cial collectors and is well known throughout the southeast by those in- terested in flattened musk turtles. Dis- cussions with several people who are familiar with the pet trade (and compar- ing figures provided by them) yields a conservative estimate of 300-400 turtles removed from this stretch during the last ten years. If recent state legislation prohibiting commercial collecting is en- forced, however, the Blackburn Fork overall should continue to support a large breeding population of 5. depre- ssus; the water quality is good and food is abundant. No flattened musk turtles were col- lected at two sites (31, 32) on Inland Lake, despite setting 40 traps for two consecutive days. It is a large, deep, cold, rocky reservoir, with water depths dropping off sharply near the shoreline. No molluscs or gastropods and few other potential S. depressvs food sources were observed, even while scuba diving. The species may no longer be present in the lake. Calvert Prong (Sites 33, 34) has been heavily influenced by agricultural runoff and possibly surface mining (Mount, 1981), and appears degraded. A low density 5. depressus population is present. The possibility of additional natural restocking exists, as there is a moderate to high density population in the Blackburn Fork upstream from the confluence with Calvert Prong. A fisherman, when shown a flattened musk turtle, said he was familiar with the animal, and that he had seen them at Site 33. We caught none there in traps, but the habitat appeared suitable. Slab Creek (Site 35) is small, with a silt-sand bottom. Sedimentation is heavy, and fine sediments typical of nearby agricultural utilization un- iformly cover all rock substrate to a depth of 1.5 cm. Corbiada are abundant, but no snails or aquatic insect larvae were observed. Although we collected no 5. depressus while wading one hour (1200-1300), a low density population may occur. C. Mulberry Fork and tributaries. — The Mulberry Fork (Sites 36, 37, 38, 39) has been considerably degraded by agricul- tural siltation (and possibly pesticides/ herbicides) and perhaps by effluents from several municipalities. Surface mining has had little effect except in the lower reaches, where commercial col- lecting has also occurred. Mount (1981) thought this stream supported a low density population of 5. depressus, and this is also our assessment. The popula- tion is probably declining; according to one source, commercial collectors found it more difficult to obtain specimens in 1982-83 than in previous years. Lost Creek (Sites 40, 41, 42) has some areas of suitable habitat remaining in the upper stretches. A moderate den- sity, reproducing population of 5. depre- ssus exists near Pocahontas (Mount, 1981), and we found a moderate to high density reproducing population near Cedrum (Site 42). We trapped 1 1 speci- mens there after heavy overnight rains made collecting conditions suboptimal. This site contains numerous shelves of eroded stratified rock; snails, Corbicula, and aquatic insects are abundant. The lower and middle stretches, however, have been strongly influenced by past and present surface mining, agriculture and impoundment (Sites 40, 41), and collecting efforts were unsuccessful. At best, a low density population may still exist in the lower reaches of the stream. Mount (1981 considered the S. depre- ssus population in Blackwater Creek (Sites 43, 44, 45, 46, 47) to be the most extensive and important in this portion 12 Tulane Studies in Zoology and Botany Vol. 27 of the species' range. This is especially true of the midreaches of the stream. We also found these areas to be favora- ble habitat and still supporting a moder- ate to high density population. How- ever, many sites are readily accessible to human disturbance. The upper reaches are degraded with silt and some coal fines; agriculture and mining have both affected these localities. The confluence of Blackwater Creek with the Black Warrior River (Site 43) is the type local- ity of 5. depressus (Tinkle and Webb, 1955), and a low to moderate density, reproducing population still occurs there (possibly reinforced by turtles en- tering from the Black Warrior River). This site has apparently not changed appreciably since it was described as the type locality (J. Whitfield Gibbons, per- sonal communication following on-site inspection). Duck Creek (Sites 48, 49) is small, but Corbicula are abundant and the physical characteristics are well within the limits of favorable habitat observed elsewhere. Possibly a low density popu- lation of 5. depressus occurs there, al- though repeated efforts (Mount, 1981, and present study) have failed to pro- duce specimens. The heavy concentra- tion of agriculture along this stream (and, indeed, all of those within Cullman County) has probably ad- versely affected the species. Fine sedimentation is moderate to heavy in some locations and extensive algal growths indicate probable organic en- richment. Although there are no his- torical records indicating dense popula- tions in this and other streams within Cullman County, the habitats appear suitable, and adequate routes for dis- persal from populated drainages have historically been present. These facts strongly suggest that 5. depressus was present in these streams in the past. D. Smith Lake. — We sampled four sites (50, 51, 52, 53) on this large reser- voir (8,550 ha). Low to moderate den- sity S. depressus populations occur at the mouths of the Sipsey Fork and Brushy Creek in Winston County (Sites 50, 51), but only low density populations are present at the confluences of Crooked Creek and Ryan Creek with Smith Lake in Cullman County (Sites 52, 53). Many areas of suitable physical habitat are found in the lake, with rock shelves and boulders abundant, and probably low to moderate density populations of turtles occur at these scattered sites. However, the steep bluffs and scarcity of shallow water habitat in the lower reaches may hinder or prevent turtles from inhabit- ing or colonizing the main portions of the lake. Further, populations of poten- tial food items may be low in these re- gions. Mount (1981) sampled four sites in Winston County and believed a heal- thy, but low density, population inha- bited the lake. A massive effort would be required to fully assess the popula- tion density of flattened musk turtles along the extensive shoreline of Smith Lake. The average population density for Smith Lake in its entirety is probably low. It does, however, possess the poten- tial for supporting a large total number of S. depressus, especially if isolated pop- ulations are considered in aggregate. E. Streams tributary to Smith Lake and their tributaries (major tributaries arranged from zvest to east, and their tributaries from north to south). — Clear Creek lies west of the Bankhead National Forest and traverses an area of intensive agriculture. We sampled two sites. The lower site (54) has sufficient rock crevices for cover and a moderate supply of snails for food. Sand deposits are extensive and and the site apparently supports a low density, and probably reproducing. No. 1 Distribution and Status of Sternothcrus depressus 13 population of S. depressus. The upper site (55) is a less suitable habitat, with few rock crevices, a very sandy bottom, and little niolluscan food, but an occa- sional flattened musk turtle can still be found. Generally, the lower reaches of the creek seem to be better habitat than those areas upstream. Agriculture ap- pears to be the major influence along much of the stream, as there are areas of occasional brown surface "foam" and extensive algal growths on the sub- merged rocks, both indicative of organic enrichment. Mount (1981) felt that Clear Creek had acceptable physical habitat and that 5. depressus was moder- ately common. He suggested that low food availability was the most important limiting factor at Camp McDowell. The Sipsey Fork was sampled at four sites above Smith Lake. Site 50 (at the confluence with Smith Lake) supports a low to moderate density population of S. depressus. Upstream, Site 58 on the West Sipsey Fork at Cranal Road (Ban- khead National Forest) was sampled by us and also by Mount (1981). The site has some deep pools and rock crevices, but also has several sandy stretches. We found both snails and Corbicula present, and our data indicate that a moderate to high density, reproducing population occurs there. However, this is a recrea- tion area, and human disturbance may adversely influence the turtle popula- tion. Mount (1981) considered the habitat relatively good and S. depressus "moderately common", but that popula- tion levels are probably limited by low food availability. Sites 56 and 57 are separated from each other by about 1 .6 km and are downstream from Site 58. Although boulders and some rock cre- vices are present, the bottom is ex- tremely sandy. Snails, Corbicula and in- sects are common to abundant. The 5. depressus population at Sites 56 and 57 is one of extremely high density and is healthy and reproducing. In fact, this portion of the Sipsey Fork supports the densest population of S. depressus re- corded to date (Table 1). Borden Creek, a shallow tributary of the Sipsey Fork, was trapped at one site (59). No turtles were collected, but some areas appeared to be suitable habitat. Corbicula were common and some snails were present. The stream widens and deepens below the bridge access, and the presence of ^'. depressus between that point and the confluence with the Sip- sey Fork, approximately 1.6 km down- stream, is considered probable. How- ever, due to the short length of this stream segment and general character of the habitat, only a low density popu- lation should be expected. Braziel Creek (Site 60), a small tributary of Borden Creek, contains marginal habitat, and few if any flattened musk turtles occur there. Brushy Creek was sampled at four sites (51, 61, 62 and 63). We found a low to moderate density, reproducing population at U.S. 278 (Site 61), but the habitat was somewhat degraded by log- ging operations. A high density, repro- ducing population of 5. depressus is pres- ent at Site 62, which is easily accessible by a Forest Service road. Corbicula and snails are common, and the stream drainage is densely forested. At Brushy Creek Lake, a shallow, swampy im- poundment on the upper reaches of the stream (Site 63), we caught no .S. depre- ssus and the physical habitat was unac- ceptable. There is, however, a sizeable population of S. udoratus in the lake. The confluence of Brushy Creek and Smith Lake at County Road 63 (Site 51) supports a moderately dense popula- tion. A wide size-class range (46.1 mm to 93.6 mm CL) was captured, and the presence of smaller animals indicates a 14 Tulane Studies in Zoology and Botany Vol. 27 reproducing population. Brushy Creek arm below this site probably continues to support a moderately dense popula- tion, at least in areas of suitable habitat. Limited trapping in Beech Creek (Site 64) produced no 5. depressus, and our data and observations of habitat characteristics suggest they are absent. It is a small tributary of Brushy Creek, with poor habitat and low food availabil- ity. Another Brushy Creek tributary, Rush Creek (Site 65), may have at best a low density population of S. depressus in the lower reaches. It is shallow and the habitat questionable; few large rocks or hiding places are present. This site was waded once, and collecting condi- tions were good. Lower water levels later in the season alter habitat condi- tions, and the stream probably does not retain sufficient water depth for safe permanent occupancy by musk turtles. Rock Creek (Site 66) supports a low density population of 5. depressus. There are heavy sand deposits and evidence of organic enrichment, probably due to ag- riculture. Trapping was minimally suc- cessful. Corbicula are fairly abundant, and several defecated piles of Corbicula fragments were also observed. Mount (1981) examined Rock Creek at Smith Lake where it is a narrow, deep channel, and felt that the presence of 5. depressus was unlikely at that locality. Crooked Creek in Cullman County (Sites 52, 67) supports only a low density population of 5. depressus. The stream is marginal habitat at best; it is small and shallow, with little rock substrate and low food availability. Ryan Creek (Site 68) is agriculturally silt-degraded; Corbicula are present, but no snails. No S. depressus were detected and, if the species is present, the creek supports only a low density population. Habitat Mount (1981:23), based on his col- lections, proposed the optimal habitat for S. depressus to be "a segment of a free-flowing large creek or small river having the following characteristics: (1) drainage area between 50 and 500 mi^, (2) depth averaging 2 ft., with vegetated shallows alternating with pools at least 3 1/2-4 ft. deep, (3) pools with detectable current, (4) an abundance of sub- merged rocks with crevices, overlapping flat rocks, or accumulations of boulders, (5) abundant molluscan fauna, (6) low silt load and minimal silt deposits, (7) relatively low nutrient content and bacterial count, (8) moderate tempera- tures (max. 85°F), and (9) minimal pol- lution by synthetic chemicals and toxic inorganic materials." We generally agree with these habitat requirements and can further substantiate the impor- tance of certain of these characteristics. Over the stream section surveyed at each site, we were able to quantify aver- age depth, per cent bottom cover, and amount of silt coverage on the stream bottom. We also qualitatively estimated stream size, current, and molluscan population levels on a scale of 1 (smal- lest or lowest) to 3 (largest or highest). In order to relate turtle population size to these stream characteristics, negative scores were given to a site for any of the following (based on comparative site differences): if an average depth of two feet did not occur in at least 50% of the area surveyed; if boulders, rock sub- strate and cobble did not exceed or equal 40% of the stream bottom area; if fine silt covered more than 15% of the bottom; and if any of the qualitatively- judged characters scored a " 1 ". A Spear- man's rank correlation test (SAS Insti- tute) demonstrated a significant nega- No. 1 Distribution and Status uf Sternotherus depressus 15 tive relationship between population size at a site (using the seven categories in Table 2) and the total number of negative stream characteristics, using the six we were able to measure or esti- mate (rs = -0.929; P = 0.003). Indeed, of the 18 sites ranked with moderate or greater 5. depressus levels, only one had more than one negative character, and the five high density population sites had no negative characteristics. However, many of the sites we examined, and especially those in which we found the highest density popula- tions, varied somewhat from these criteria. Particularly notable is that we found sand to be a much more ubiquit- ous characteristic of 5. depressus habitat than has been recognized in the past. Sandy habitats, when coupled with (a) adequate exposed bedrock, boulder, or cobble substrate, and (b) a sufficient food supply (snails, Corbicula, and aqua- tic insects), often support dense popula- tions of flattened musk turtles. In fact, the five sites judged to support the high- est density populations (Sites 5, 6, 56, 57 and 62) all had extensive sand de- posits (>35% coverage of stream bot- tom), sometimes as deep as 60 cm and obscuring as much as 75% of the rock substrate. We have seen 5. depressus bur- row into the sand and have collected others that had already buried them- selves. The specific nature of sandy sed- iments also plays a role. For example, the species seems to avoid areas in which the sand is firm or hard-packed, unless there are other suitable refugia nearby (i.e., boulders, bedrock crevices, holes in banks, etc.). Apparently, such areas are suitable for foraging, as we have cap- tured specimens in traps set there. We have not, however, observed 5. depressus burrowing into hard-packed sand. For the most part, turtles seem to burrow in loose sand or in a combination of sand- silt and sandy-clay substrate. Whereas sandy habitats may support good populations of 5. depressus, finer sediments (silt and clay), on the other hand, do appear to be a decisive factor in determining habitat quality for the animal, in that areas with heavy deposits of fine silt and clay sediments do not support large populations. A Spear- man's rank correlation test comparing the number of sites at each population assessment category with the occurrence of a high silt load (>15% coverage of stream bottom) revealed a significant negative relationship between popula- tion size and stream silt load (rs = -0.929; P = 0.003). Adequate bottom cover at a site was also significantly correlated to 5. de- pressus population levels. Using a criter- ion of >40% coverage of the stream bot- tom by bedrock, boulders, and cobble as adequate bottom cover, we found a positive relationship between popula- tion size and bottom cover (rs = 0.857; P = 0.014). Finally, though food was not quan- titatively measured during our study, each site was qualitatively judged as to whether bivalves and snails, primary 5. depressus food items, were in low, medium, or high numbers. Food abun- dance appeared strongly linked to turtle abundance, as only 2 of the 18 sites ranked as having moderate-sized or greater 5. depressus populations were judged to be deficient in food. By con- trast, at many sites where S. depressus populations were ranked as low or ab- sent, food levels were also low. Daily Activity Patterns Estridge (1970) and Mount (1981) both commented that juveniles have a 16 Tulane Studies in Zoology and Botany Vol. 27 greater daily activity span than do adults, which are predominantly noc- turnal. We agree essentially with this, and can provide additional observations made while wading. Fewer active adults than juveniles (7 versus 9) were hand- collected during the daylight (non- twilight hours), and the juveniles col- lected then were usually of the smaller size classes (<50 mm CL). Juveniles be- came more active at twilight, with most captures occurring during this period, and as darkness increased, so did activ- ity involving all size classes. Though precise records were not kept, there seemed to be a correlation between the time of day and the size of active turtles. In early evening we observed more juveniles foraging, but as night progres- sed more young. adults emerged, fol- lowed by the very large adults after 2200 hours. Predation We found only two dead predated 5. depressus. Both had apparently been killed by raccoons, whose tracks were found nearby. Several other live indi- viduals were missing limbs or tails. How- ever, flattened musk turtles are small, and many (especially juveniles) can be swallowed whole by large predators, while even adults could easily be carried off to be eaten elsewhere. Predation on juveniles is probably high, and nesting females are more vulnerable while on land. Health We found no evidence of diseased 5. depressus, although they were heavily parasitized by the leech Placobdella parasitica, which carries the protozoan agents {Haemogregarina and Haemop- roteus) of turtle malaria (Ernst and Ernst, 1979). Estridge (1970) found Haemogregarina in three of seven S. de- pressus he examined. Mount (1981) re- ported that two of five specimens cap- tured while basking at a Jefferson County site had abscesses presumably caused by bacteria. He thought that such basking behavior was abnormal and possibly related to their diseased condition. Basking probably occurs much more frequently in this species than has been previously reported. Healthy individuals we have kept in the laboratory readily climb out of the water when provided with basking platforms and heat lamps. We have also seen bask- ing in the wild, and J.V. Peavy (pers. comm.) reports observing extensive basking by 5. depressus along some streams while canoeing. Conclusions The present study demonstrates that 5. depressus is more prevalent within its range than previously thought. Eigh- teen (26%) of the 68 sites sampled had moderate to high density populations, and an additional 8 sites (12%) had pop- ulation densities which we regarded as between low and moderate (Table 2). Unfortunately, some sites could be vis- ited only once, due to the constraints of time and logistics. As we have men- tioned, initial collecting efforts at some localities failed to produce specimens, but subsequent attempts revealed the \ presence of good populations (e.g., Turkey Creek, Blackburn Fork, Black- water Creek). The results of a single at- tempt, even with adequate trap num- bers, can therefore yield an underesti- mation of population density. Mount (1981) inferred from his and other existing data that discreet popula- tions of 5. depressus may have substan- tially fewer juveniles today than prior to 1970. Table 3 compares the size-class distributions obtained by: (a) pooling re- No. 1 Distribution and Status of Sternotherus depressus 17 suits for all pre- 1970 studies, (b) Mount ssus <60 mm CL. In the present study, (1981), and (c) the present study. The 34 juveniles were collected during al- pre-1970 data are based on several col- most 250 hours wading. Catch-per-unit- lections made by different collectors effort calculations are virtually identical (during different years and at different in both studies, with each juvenile re- periods within the collecting season) quiring about 7.5 hours effort. Table 1 using various collecting techniques. Lit- indicates that juveniles <60 mm CL erature records suggest that wading and were captured at 9 (56%) of the 16 collecting by hand were the primary col- waded localities at which we caught tur- lecting methods employed in these ear- ties, while only two were caught in traps Her studies. In the present study and during the entire study. Published that undertaken by Mount (1981), the studies on turtles indicate that in some major emphasis was placed on trapping, species juveniles apparently comprise a In our experience, smaller juveniles are low percentage of the population (see only occasionally collected by trapping, Harless and Morlock, 1979, for sum- due probably to the mesh size of the mary), although to what extent sampl- traps (1.8-2.5 cm), through which small ing bias (similar to that seen in 5. (/g- individuals can move with ease. Differ- pressus) has played a role is unknown. In ences in dietary preferences in large vs. addition, in almost all turtle populations small turtles may also influence trap- juveniles are more difficult to detect, ping results. The percentage of smaller The size-class distributions obtained by individuals (<60 mm Carapace Length) Mount (1981) and in the present study is much greater in samples obtained by (Table 3) are remarkably similar to wading and collecting by hand (under those obtained in a northern Florida suitable habitat conditions). Therefore, spring for S. minor (Cox, 1978), in which the otherwise most efficient collecting snorkeling and trapping were the major technique (trapping) appears to intro- collecting techniques employed under duce a selective bias favoring the cap- almost ideal conditions. Also, the per- ture of larger individuals (>60 mm CL). centage of juveniles collected by Mount Mount (1981) waded a total of 44.4 (1981) and in this study are not particu- hours and collected 6 juvenile S. depre- larly low when compared to findings in TABLE 3. Comparison of size-class distributions (mm carapace length) of flattened musk turtles col- lected prior to 1970, by Mount (1981), and during the present study (1983). Specimens considered intermediate between Sternolherns depressus and .S'. minor peltifer are not included. Numbers collected are indicated below each size class. The percentage of the total for each size class is given in parenth- eses. Size Class (mm CL) 30-39 40-49 50-59 60-69 70-79 80-89 90-99 100 -h Total Combined pre- 1970 Samples (adapted from Mount, 1981, and Estridge, 1970) 18 (26%) 6 (9%) 7 (10%) 7 (10%) 8 (11%) 9 (13%) 12 (17%) 3 (4%) 70 Mount, 1981 1 (1%) 1 (1%) 4 (4%) 8 (8%) 13 (13%) 29 (29%) 35 (35%) 10 (10%) 101 Present Study, 1983 9 (2%) 14 (3%) 11 (2%) 44 (9%) 107 (22%) 164 (32%) 123 (25%) 26 (5%) 498 18 Tulane Studies in Zoology and Botany Vol. 27 populations of various other turtle examined. They were particularly species (Legler, 1960; Ernst, 1971, noticeable in Cullman County, and else- 1976; Brown, 1974; and Vogt, 1980). where, both the Mulberry Fork and Lo- Although the size-class distribution in cust Fork have been adversely impacted. Table 3 for pre- 1970 studies might re- Sand sediments do not appear to be par- fleet real population structure, it is most ticularly harmful, and many of our remarkable for its bimodal distribution, highest density populations were as- rather than for its similarity to results sociated with extensive (moderate to from other studies. This suggests that heavy) sand deposits. There is, however, pre- 1970 investigators (with emphasis an inverse correlation between the on hand-collecting) inadvertently biased amount of silt and clay sedimentation their samples in favor of smaller indi- and the population density of 5. depre- viduals. Thus, we conclude that the evi- ssus. No cause and effect relationship, dence is not compelling enough to war- however, has yet been established, rant an a priori acceptance of Mount's Early in this study, observations at (1981) hypothesis regarding a decline in various localities suggested that certain the percentage of juvenile 5. depressus. trends regarding land use practices cor- One of the problems that has faced related with habitat degradation. Al- 5. depressus has been commercial coUec- though neither formally stated nor tion. Although we do not know just how rigorously tested, our subsequent ex- extensive this has been, or to what ex- periences and qualitative observations tent this may have reduced specific pop- have strengthened this perception. We ulation levels, serious overcollecting conclude that the two greatest influ- without regard for population dynamics ences affecting habitat quality, and has been reported to us by several therefore affecting populations of flat- sources who have contacts within the pet tened musk turtles, are surface mining trade. Selected areas along the Mul- and agriculture. There is no doubt that berry, Locust and Blackburn Forks have surface mining can have a drastic effect, reportedly received much attention especially upon local populations that from commercial collectors. State legis- are near active or abandoned mines that lation (1984) now prohibits the taking are still leaching into the stream. The of 5. depressus except for valid scientific devastating effects of a massive influx purposes and requires a special permit, of mining residues upon the Blue Creek Hopefully, this will serve to reduce col- population (Sites 5, 6) have been well lecting pressure on the species. How- documented. However, this population ever, recent unconfirmed reports dur- has recovered remarkably well in only a ing 1985 indicated that some commer- 10-20 year period. Long-term problems cial collecting may still be occurring. can also occur, though, especially when We have also found the question of abandoned mines are not reclaimed sedimentation to be more complex than (i.e.. Hurricane Creek, Site 2). The ef- previously realized. First, although sed- fects of mining, therefore, are often im- iments from surface mining were pres- mediate and obvious, but in some cases ent at many sites, the major source of the damage is reversible. The streams sedimentation appears to be agriculture of Cullman County, and the upper Mul- throughout most of the Black Warrior berry and Locust Forks, however, pres- River drainage. We found agricultural ent an entirely different picture. In sediments in almost all of the waterways these areas of intensive agriculture, the No. 1 Distribution and Status of Sternotherus depressus 19 effects of organic enrichment, siltation, and the possible presence of pesticide/ herbicide residues suggest a slow, long- term and insidious degradation of habitat. Just how serious this is, or will be in the future, is not known. No study to date has addressed the potential problems presented by agricultural in- fluence on these waterways and upon their attendant aquatic biota. Mount (1981) concluded that S. de- pressus meets the criteria for inclusion on the federal threatened species list, and he placed the major blame for the species' decline on siltation from surface mining. There is, unfortunately, little historical data from which to make pop- ulation comparisons. The present study reveals 5. depressus to be somewhat more numerous than previously thought, with high density populations at Sites 5 and 6 (intergradient?), 56, 57 and 62, and moderate to high density popula- tions at Sites 22, 29, 30, 42, 44, 45, 51 and 58. We conclude that 5. depressus is a case for special concern, but that its candidacy for inclusion on the federal threatened list is inadvisable at the pres- ent time. We do feel, however, that con- tinued monitoring of its status is jus- tified. In addition, the present lack of knowledge regarding populations in the major impoundments (Bankhead Lake, Smith Lake, Lake Tuscaloosa) precludes a complete understanding of the species' status. Finally, four streams are crucial to our assessment regarding the present status of 5. depressus. These are Blackwater Creek, Blackburn Fork, Sip- sey Fork and Brushy Creek. Should any one of these populations decline ap- preciably or should illegal commercial exploitation continue in these streams, federal protection may be warranted. Acknowledgments We would like to thank several per- sons and organizations for their aid dur- ing this study. Charles Dixon, Britt Mor- ris, Bruce Thompson, and Steve Thompson assisted us in the field and laboratory. Without these student assis- tants we could not possibly have gained such results as we achieved. Dick Darby provided use of a boat during our lake studies. Dr. J. Whitfield Gibbons, Savan- nah River Ecological Laboratory, ac- companied us in the field on one occa- sion and contributed many useful sug- gestions on collection methods and habitat assessment. James V. Peavy pro- vided historical collecting data and per- sonal observations on selected popula- tions of 5. depressus. Perry Hubbard acted as liaison between the Field Direc- tors and Sponsors, and we are grateful for his assistance. Drs. R. Christian Jones and Donald P. Kelso, George Mason University, discussed with us the water quality parameters to be meas- ured, and Drs. W. Mike Howell and Robert Stiles, Samford University, pro- vided information on the aquatic ecol- ogy of the Warrior Basin, as well as en- couragement. We also thank various property owners for allowing us to col- lect on their lands and for information pertaining to local history of the water- ways involved. Samford University pro- vided laboratory space and equipment during the project, and George Mason University liberal leave for Ernst. Fi- nally, we acknowledge the Alabama Coal Association for project funding, logistical assistance, and cooperation in our conducting the research free of in- terference. LrfERATURE Cited Brown, W. S. 1974. Ecology of the aquatic box turtle, Terrapene coahuila (Chelonia, Emydidae) in northern Mexico. Bull. Florida St. Mus. Biol. Sci. 19: 1-67 20 Tulane Studies in Zoology and Botany Vol. 27 Close, D. K. 1982. The reproductive cycle oiSter- nolherus minor depressus. Unpubl. M.S. Thesis, Univ. Alabama at Birmingham, Birmingham, Alabama. 101 pp. Cox, W. A. 1978. Reproduction, growth, and life history strategies in the musk turtle {Ster- notherus minor) in north Florida. Unpubl. Ph.D. dissertation, Univ. Alabama at Birmingham, Birmingham, Alabama. 137 pp. Ernst, C. H. 1971. Population dynamics and ac- tivity cycles of Chrysemys picta in southeastern Pennsylvania. J. Herpetol. 5:151-160. . 1976. Ecology of the spotted turtle, Clemmys guttata (Reptilia, Testudines, Tes- tudinidae), in southeastern Pennsylvania. J. Herpetol. 10: 25-33. and E. M. Ernst. 1979. Synopsis of protozoans parasitic in native turtles of the United States. Bull. Maryland Herpetol. Soc. 15: 1-15. Estridge, R. E. 1970. The taxonomic status of Sternotherus depressus (Testudinata, Kinoster- nidae) with observations on its ecology. Un- publ. M.S. Thesis, Auburn Univ., Auburn, Alabama. 49 pp. Harless, M. and H. Morlock. 1979. Turtles: perspectives and research. John Wiley and Sons, N.Y. 695 pp. Iverson, J. B. 1977. Sternotherus depressus. Catalog. Amer. Amphib. Rept. 194: 1-2. . 1979. Another inexpensive turtle trap. Herp. Review 10: 55. Legler, J. M. 1960. Natural history of the ornate box turtle, Terrapene ornata ornata Agassiz. Univ. Kansas Publ. Mus. Natur. Hist. 11: 527- 669. Mount, R. 1975. The reptiles and amphibians of Alabama. Alabama Agr. Exp. Sta., Auburn Univ., Auburn, Alabama. 347 pp. . 1981. The status of the flattened musk turtle, Sternotherus minor depressus Tinkle and Webb. Final Report, U.S. Fish and Wild- life Serv. Contract 14-16-0004-80-096. 1 19 pp. SAS Institute. 1982. SAS user's guide: statis- tics. SAS Institute, Cary, North Carolina. 352 pp. Seidel, M. E., J. B. Iverson, and M. D. Adkins. 1986. Biochemical comparisons and phylogenetic relationships in the family Kinos- ternidae Testudines). Copeia 1986(2): 285- 294. Tinkle, D. W. 1958. The systematics and ecology of the Sternotherus carinatus complex (Tes- tudinata, Chelydridae). Tulane Stud. Zool. 6: 1-56. and R. C. Webb. 1955. A new species of Sternotherus with a discussion of the Sternotherus carinatus complex (Chelonia, Kinosternidae). Tulane Stud. Zool. 3: 52-67. VoGT, R. C. 1980. Natural history of the map turtles Graptemys pseudogeographica and G. ouachitensis in Wisconsin. Tulane Stud. Zool. BoL 22: 17-48. THE PREVALENCE AND INTENSITY OF PERKINSUS MARINUS FROM THE MID NORTHERN GULF OF MEXICO, WITH COMMENTS ON THE RELATIONSHIP OF THE OYSTER PARASITE TO TEMPERATURE AND SALINITY THOMAS M. SONIAT DEPARTMENT OF BIOLOGICAL SCIENCES AND CENTER EOR BIO-ORGANIC STUDIES UNIVERSITY OF NEW ORLEANS, LAKEFRONT, NEW ORLEANS, LA. 70148 and JULIE D. GAUTHIER DEPARTMENT OF BIOLOGICAL SCIENCES UNIVERSITY OF NEW ORLEANS, LAKEFRONT, NEW ORLEANS, LA. 701 48 Abstract Twenty-three samples of oysters were col- lected from 19 sites along the mid Northern Gulf of Mexico and assayed for the prevalence and in- tensity of the parasite Perkinsus marinus. Preva- lence of infection (measured as percent infection, PI) ranged from 0 to 100%, whereas intensity of infection (measured as weighted incidence, WI = sum of disease code numbers/number of oysters) ranged form 0.00 to 3.06. WI was correlated with water salinity (Kendall Tau Beta correlation coef- ficient = 0.331, P < 0.05). but not with water temperature. Introduction Perkinsus marinus (Mackin, Owen, and Collier) is a protistan parasite of the American oyster, Crassostrea virginica (Gmelin). It was first described as Der- mocystidium marinum by Mackin, Owen, and Collier (1950) because of its close similarity to the freshwater parasitic fungus Dermocystidium salmonis Davis. Later observations indicated that D. marinum gave rise to gliding cells on "mucoid tracks" similar to those of slime molds and was reclassified as Labyrin- thomyxa marina (Mackin and Ray 1966). Although ultrastructural studies by Per- kins (1969) revealed a likeness of the parasite to the fungi, no cytoplasmic ex- tensions or rhizoids were observed. Levine (1978) renamed the parasite Per- kinsus marinus on the basis of electron microscopic work by Perkins (1976) which revealed the presence of an apical complex in a motile zoospore stage (Table 1). P. marinus is a major cause of oyster mortality along the Gulf and Atlantic coasts (Mackin et al. 1950, Ray and Mac- kin 1955, Andrews 1965). Histological studies of oysters infected with the para- site show areas of inflamation and tissue destruction (Mackin 1951). Significant weight loss of diseased oysters has been documented (Ray et al. 1953), and dis- ease intensity and levels of stress in oys- ters have been positively correlated Soniat and Koenig 1982). Numerous studies (e.g., Mackin et al. 1950, Ray et al. 1953, Mackin 1955, Quick and Mackin 1971, Ogle and Flurry 1980, Soniat 1985) have shown that P. marinus is most prevalent during the warm months in high salinity areas. Oysters are more susceptible to the dis- ease during spawning season which, along the Gulf Coast, continues throughout the summer (Mackin et al. 1950). However, high weighted inci- dence values (WI, a measure of disease intensity) have been shown to be directly related to high temperatures (Hewatt and Andrews 1955, Quick 1971). Al- though positive correlations have been made between salinity and WI, low sa- linity apparently imposes no physiologi- 21 22 Tulane Studies in Zoology and Botany Vol. 27 TABLE 1. The various names and suggested taxonomic affinities of the oyster parasite. Name Affinities Reference Democystidium marinum Labyrinthomyxa marina Perkinsus marinus Protista Entomophthorales (Fungi) Haplosporidia (Protozoa) Chytridiales (Fungi) Saprolegniales (Fungi) Labyrinthulaies (Slime mold) Apicomplexa (Protozoa) Mackinetal. (1950) Sparrow (in Ray, 1954a) Sprague (1954) Mackin&Boswell(1956) Perkins &Menzel (1967) Mackin& Ray (1966) Levine(1978) cal handicap on the development of P. marinus (Ray 1954a); freshwater influx may simply dilute infective elements causing lower WI values (Mackin 1962). Stronger correlations have been found between WI and the product tempera- ture and salinity (TxS) than between temperature or salinity alone (Soniat 1985). As part of studies of oyster ecology and genetics, as well as effects of oil and gas operations upon oysters, we have had the opportunity to collect oysters and sample parasites from numerous lo- cations in Texas, Louisiana, and Missis- sippi. This paper further documents a positive correlation between WI, salin- ity, and TxS values. It also provides cur- rent information on distribution, preva- lence, and intensity of P. marinus along the Mid-Gulf Coast region. Material and Methods Oysters were collected from August 1985 to January 1987 at 13 sites along the mid Gulf Coast (Figure 1). Temper- ature (mercury thermometer) and salin- ity (American Optics refractometer) were recorded at each site. Small pieces of mantle tissues (~4mm2) from 9 to 30 oysters of approximately commercial sizes (6.0-12.5 cm umbo-to-bill distance) were examined for parasitism by P. •■l««tl«i lav Fig. L A map of the mid northern Gulf of Mexico showing the sample sites. CO = Confederate Reef, AF = April Fool Reef, BE = Beasley's Reef (Galveston Bay); LC = Lake Calcasieu; VB = Vermilion Bay; TB = Terrebonne Bay; CC = Cocodrie; HB = Hackberry Bay (Barataria Bay area); AB = Adams Bay; BB = Black Bay; CB = Couchon Bay; BL = Bayou Linge; JR = Lake Jean Robin; CQ = Lake Coquille; LB = Lake Borgne; BX = Biloxi. No. 1 Prevalence and Intensity of Perkinsus marinus 23 marinus using the Ray (1966) technique. The tissues were placed for one week in culture tubes containing a fluid thiog- lycollate medium and antibiotics (Chloromycetin and mycostatin). Micro- scopic examination of the tissues re- vealed enlarged P. marinus presporan- gia upon treatment with iodine. The level of infection of each oyster was rep- resented by a disease code number (Quick and Mackin 1971) which ranged as an integer from 0 (uninfected) to 6 (heavily infected). Percent infection (PI, a measure of disease prevalence) and weighted incidence were calculated for each site. Weighted incidence (Mackin 1962) was determined as follows: WI = sum of disease code numbers total number of oysters The temperature/salinity interaction term (TxS) was also calculated for each site (Soniat 1985). Non-parametric correlations (Ken- dall Tau Beta correlations) between temperature, salinity, TxS, WI, and PI were determined using the Statistical Analysis System package (Helwig and Council 1979) installed on a VAX 8600 computer at the University of New Or- leans. Results anl3 Discussion Temperatures varied from 10.5°C at Confederate Reef (COl, Table 2) to 30.4°C at Terrebonne Bay (TBI); how- ever, the average value was 21.1°C and most of the samples were taken when water temperatures were above 20°C. Salinity values varied more broadly. They ranged from 4 ppt at Vermilion Bay (VBl) to 32 ppt at Confederate Reef (C02) with an overall mean of 16 ppt. The TxS was used to measure the potential of the environment to produce and sustain heavy levels of infection. The factor was developed by Soniat TABLE 2. Data for temperature (°C), salinity (ppt), weighted incidence (WI), percent infection (PI), the product of temperature and salinity (TxS), and the number of samples (N) collected from the various sites. Site names and locations are given in Figure 1. Site Date Temperature Salinity WI PI TxS N ABl 8/18/86 30.0 20 1.70 90 600.0 10 AFl 8/24/85 27.4 13 2.90 100 356.2 30 BBl 1 1/7/85 18.0 11 1.33 53 198.0 30 BEl 8/24/85 27.5 8 0.03 3 220.0 30 BLl 12/11/86 12.8 16 0.90 60 204.8 10 BXl 1 1/8/86 25.0 20 2.20 100 500.0 10 BX2 10/14/86 24.2 20 2.27 9 484.0 11 BX3 9/2/86 28.0 25 2.00 93 700.0 15 BX4 9/25/85 27.4 5 2.07 100 137.0 30 CBl 10/7/86 27.5 19 0.60 30 522.5 10 CCl 10/16/86 21.1 12 2.89 89 253.2 9 COl 1/27/87 10.5 16 0.00 0 168.0 10 C02 8/7/86 30.0 32 1.80 93 960.0 30 COS 8/23/85 27.0 30 3.06 100 810.0 30 CQl 12/11/86 12.8 16 1.10 70 204.8 10 CQ2 12/3/86 14.0 14 1.90 90 196.0 20 CQ3 12/11/86 13.0 16 1.90 90 208.0 10 HBl 11/19/85 25.0 9 0.00 0 225.0 30 JRi 12/11/86 12.8 16 1.60 90 204.8 10 LBl 10/8/85 10.9 10 0.17 13 209.0 30 LCI 12/17/85 11.0 11 1.20 60 121.0 30 TBI 9/12/85 30.4 16 1.27 53 486.4 30 VBl 11/30/85 22.8 4 0.00 0 91.2 30 24 Tulane Studies in Zoology and Botany Vol. 27 (1985) as a simple method to account for the combined efforts of temperature and salinity since the parasite is more prevalent when high temperatures and high salinities co-occur (Mackin et al. 1950). Soniat (1985) showed a closer correlation between WI and TxS than between WI and salinity (temperature and WI were not correlated). Further- more, salinity and temperature are sometimes inversely correlated (Soniat 1985), which confounds the interpreta- tion of WI data in relation to these en- vironmental variables. In the present study temperature and salinity were not correlated (Table 3). TxS values ranged from 91.2 at Vermilion Bay (VBl) to 960.0 from Confederate Reef (C02). The highest WI value (3.06) was found at Confederate Reef (C03), which is also where the second highest TxS value (810.0) was found. No infected oysters were found at Confederate Reef (COl on 1/27/87), Hackberry Bay (HBl), or Vermilion Bay (VBl). All oysters were found infected (PI = 100%) at April Fool Reef (AFl), Biloxi (BXl on 11/8/ 86), and Confederate Reef (C03 on 8/ 23/85). The lowest WI values (0.00-0.03) were found at sites with salinities that ranged from 4 ppt (VBl) to 9 ppt (HBl) — with two exceptions. The first TABLE 3. Kendall Tau Beta correlation coeffi- cients and probability values for correlations between temperature (TEMP), salinity (SAL), weighted incidence (WI), percent infection (PI), and the product of temperature and sa- linity (TxS). N = 23. SAL WI PI TxS TEMP 0.200 0.202 0.204 0.564 0.199 0.185 0.191 <0.001 SAL — 0.331 0.354 0.653 <0.050 <0.050 <0.001 WI — — 0.707 <0.00l 0.321 <0.050 PI — — — 0.314 <0.050 exception was found at Confederate Reef (COl) where the WI was 0.00 and the salinity was 16 ppt; however, the temperature was 10.5°C — the lowest re- corded (Table 2). The low temperature could explain the low WI value; how- ever, Ray (personal communication) found a WI of 1.62 on 1/6/87 (12°C, 14 ppt) and a WI of 0.84 on 2/17/87 (13°C, 20 ppt) from Confederate Reef oysters. The second exception was a Biloxi sam- ple (BX4) which had a relatively high WI (2.07) and a low sahnity (5 ppt). This sample was taken shortly after a freshet event and the intensity of infection probably did not have time to decrease from its typically high level. The typical WI pattern of decreas- ing parasitism with decreasing salinity was found at three reefs in Galveston Bay. Low (Beasley's), moderate (April Food), and high salinity (Confederate) reefs were sampled within a week of each other (the temperature difference was only 0.4°C). Beasley's Reef (BEl) had a WI of 0.03, April Fool Reef (AFl) had a WI of 2.90, and Confederate Reef (C03) had a WI of 3.06 (Table 2). De- tails of the sample sites and information on P. marinus in Galveston Bay are pro- vided by Hofstetter (1977). Ray (1987) finds consistently high levels of infec- tion at Confederate Reef. The reef may serve as a refuge for the parasite in the Galveston Bay complex. Confederate Reef may also be a permanent source of infective agents (Ray, personal com- munication) since oysters there are not subject to freshets which would help control the parasite. Studies of oyster spat on Confederate and a nearby reef (Ray, 1987) have, in fact, caused Ray to revise earlier contentions (Mackin 1951, Ray 1954b) that young oysters (espe- cially those less than a year old) are not susceptible to infection by P. marinus. Oysters may need a threshold inoculum of infective agents. The threshold is not No. 1 Prevalence and Intensity of Perkinsus mariniLS 25 only a function of age (or the time the oysters are exposed to infective agents), but also a function of the number of in- fective agents in the environment (or the proximity of oysters to the focus of infection). If oysters do become infected at earlier ages in areas near a focus of infection, then WI and PI data from oysters by size or age classes would be useful in locating disease "hot spots". Ray's work (Ray, 1987) confirms the re- sults of Hofstetter (1971, 1977) who found P. marinus infections in spat from reefs in central Galveston Bay. Craig and Powell (1986) and Powell (personal communication) sampled oysters at 50 sites from Brownsville, Texas to the Florida Everglades, including 4 sites in Galveston Bay. Prevalence of the para- site was never less than 50% from sites in Texas; however, Craig and Powell did not sample any low-salinity reefs in Trinity Bay (from Beasley's Reef we found a PI of 3% and a WI of 0.03). Mackin (1962) sampled oysters from more than 100 bays and bayous in Louisiana and found prevalences rang- ing from 0 to 100%. Some of his stations are near those sampled in the present study. Although P. marinus was discov- ered in Louisiana, where most of the early work was conducted, little recent information exists on the occurrence and distribution of the parasite there. Craig and Powell (1986) and Powell (personal communication) sampled nine Louisiana stations; Ray (1982) reports disease data from Lake Calcasieu; and Turner (1985) talks about P. marinus, but did not assay for the parasite. The Lake Calcasieu site was in West Cove in the immediate vicinity of Ray's (1982) Station T-5 and Turner's (1985) Station W. Ray's (1982) samples from the area (taken on 1 1/7/82) showed higher WI values (2.1 vs. 1.2), and higher PI levels (83% vs 60%), but were taken at a higher temperature (14.4°C vs 1 rC) and a higher salinity (20.1 ppt vs 1 1 ppt). The Vermilion Bay location (VBl) had the lowest salinity, lowest TxS value, and a WI of 0 (Table 2). Mackin (1962) sampled the adjacent Atchafalaya Bay and likewise found no evidence of parasitism. Freshwater discharge from the Atchafalaya River maintains low salinities in both the Vermilion and Atchafalaya Bays, and P. marinus is not prevalent in either. At the two sites established in the Terrebonne Bay area, WI values were 1.27 (TBI) and 2.89 (CCI). These levels of infection are similar to those found by Mackin (1962) who sampled numer- ous bays and bayous in the Terrebonne/ Timbalier Bay area. No infection (Table 2) was recorded in oysters from Hackberry Bay (HBl, NW section of the Barataria Bay com- plex). This bay is a relatively low salinity area (9 ppt on 1 1/19/85), where Mackin (1962) also did not detect the parasite. In contrast, the lower portions of Barataria Bay often have severe prob- lems with P. marinus (Dugas, personal communication); Mackin (1962), for example, found a WI of 4.38 from gap- ers (dead oysters with the shells still at- tached and the meats intact) at Bayou Rigaud (Grand Isle). Adams Bay was sampled numerous times by Mackin (1962); WI values ranged from 0 to 1.00. In the present study (ABl, Table 2) a WI of 1.70 was found. The seven samples taken im- mediately east of the Mississippi River (BBl, BLl, CBl, CQ1-CQ3, JRl) had WI values from 0.90 to 1 .90. These val- ues were higher than those of Mackin (1962), who found a range of 0.00 to 1.77 in his samples taken east of the Mis- sissippi River. With the present data it is not possible to definitively state that the area is experiencing increased prob- 26 Tulane Studies in Zoology and Botany Vol. 27 lems with P. rnarinus; however, with an increase in salinity in this area (due to the intrusion of salt water since the con- struction of the Mississippi River Gulf Outlet) one could, a priori, expect in- creased problems with the parasite. The Lake Borgne (LBl) samples were taken from an area that supports a few low salinity reefs (see Mackin and Hopkins 1962). Salinity, at the time of the sample, was 10 ppt and WI was low— 0.17. A single station in Mississippi was sampled four times (BX1-BX4). WI ranged from 2.00 to 2.27; salinity varied from 5 to 25 ppt, and temperature val- ues were 24.2°C to 28.0°C (Table 2). Thus, temperature and WI values were relatively uniform, although salinity varied greatly. However, as mentioned previously, the BX4 sample (9/25/85) was taken immediately after a freshet, and almost certainly represents a WI typical of higher salinities. Ogle and Flurry (1980) sampled four reefs in Mis- sissippi Sotmd over a 25 month period and found a low prevalance and inten- sity of the parasite. The greatest WI was 0.88 — substantially lower than that of the present study. Temperature was not correlated with salinity, WI, or PI (Table 3). Soniat (1985) found a significant inverse corre- lation between temperature and salinity from a single reef (April Fool Reef, Gal- veston Bay), which was also sampled for P. rnarinus. No correlation was found be- tween temperature and WI and PI in the present study because temperatures did not vary greatly and most of the samples were taken at temperatures greater than 20°C. When temperatures are consis- tently high and salinity varies, salinity appears to be the controlling factor of the disease. Likewise, when salinities are relatively constant and temperatures are more variable, temperature is likely to be correlated with WI. In the present study and in Soniat (1985), salinity is correlated with WI and PI. These re- sults are exactly opposite to those of Quick and Mackin (1971) who found no relationship between WI and salinity, and a significant relationship between WI and temperature. WI and PI are correlated with each other and both are correlated with the TxS interaction term (Table 3). In contrast to the work of Soniat (1985), temperature (in the present study) did not add to the signifi- cance of the TxS term. In fact, the TxS term is not as strongly correlated with WI as is salinity alone. Nonetheless, the TxS interaction term is a useful method of measuring the potential of the envi- ronment to sustain high levels of infec- tion by P. rnarinus — especially when the sampling design includes seasonal sam- pling, and sampling of broad geog- raphical areas, and when temperature and salinity are correlated with one another. ACKNOW^LEDGMENTS We are grateful to Roy Ary, Marc Bieaux, Rick Howell, Livingston Lanclos, Martin Mayer, Al Segura, Lorene Smith, Andy Terpak, and Susan Vreeland for their assistance in the col- lection of oysters. The authors are also grateful to Janet Forbes for typing, and Jerilyn Green for proof reading. The project was primarily funded by the U.S. Fish and Wildlife Service and the UNO Fisheries Initiative Program. Literature Cited Andrews, J. D. 1965. Infection experiments in nature with Deimocystidium marinum in Chesapeaiie Bay. Chesap. Sci. 6: 60-67. Craig, M. A. and E. N. Powell. 1986. A survey of Perkinsus mariniis in Gulf of Mexico. Pre- sented at th 78th Annual Natl. Shellfish. Assoc, meeting, June 22-26, Seattle, Washington. No. 1 Prevalence and Iniensity of Perkinsus rnarinus 27 Helwig, J. 1". and K. A. Council. 1979. SAS user's guide. SAS Insiiiute Inc., Caiy, N.C 494 P- Hewatt, W. G. and J. D. Andrews. 1955. Tem- perature coiurol expeiiiiieiiis on Uie fungus disease, Dermucysiidium martnum, ot oysters. Proc. Natl. Shellfish. Assoc. 46: Iii9-1 -53. HOFSIEIIER, P. 1971. Galveston Bav oyster studies: 1971. Texas Parks anci Wildl. Dept. Coastal Fish Proj. Rept. 1971, 107-124. HOFSTEiiER, R. P. 1977. Trends in population levels ot the American oyster, Crussosirea vir- ginica Gmelin on public reets in Galveston Bay, Texas. Texas Parks and Wildl. Dept. Tech. Ser. No. 24. 99 p. Levine, N. D. 1978. Ferkinsus gen. n and other new taxa in tiie Protozoan phylum Apicomp- lexa. J. Parasitol. 64: 549. Mackin, J. G. 1951. Histopatfiology ot infection of Crassostrea virginica (Gmelin) bv Democys- tidiurn martnum Makin, Owen and Collier. Bull Mar. Sci. Gult and Caribb. 1: 15-30. Mackin, J. G. 1955. Dermonsiidium mannum and salinity. Proc. Nad. Shelltlsh Assoc. 46: 116- 128. Mackin, J. G. 1962. Oyster disease caused by D^r- mocystidium martnum arm other mici oorganisms in Louisiana. Publ. Inst. Mar. Sci. Univ. Texas 7: 132-299. Mackin, J. G. and J. L. Boswell. 1956. The life cycle and relationships of Dermocysiidium mannum. Proc. Natl. Shellfish Assoc. 46: 112- 115. M.ackin, J. G. and S. H. Hopkins. 1962. Studies on oyster mortality in relation to natural envi- ronments and to oil fields in Louisiana. Publ. Inst. Mar. Sci. Univ. Texas 7: 1-131. Mackin, J. G. and S. M. Ray. 1966. 1 he taxonornic relationship of Dermocysiidium marinum. Mackin, Owen and Collie J. Invert. Pathol. 8: 544-545. Mackin, J. G. H. M. Owen, and Collier. 1950. Preliminary note on the occurrence of a new protisian parasite, Derrnocyscidtum mannum n. sp. in Crassusirea virgituca (Gmelin). Science 111: 328-329. Ogle, J. and K. Flurry. 1980. Occurrence and seasonality of Ferkinsus marinus (Pro- tozoa:Apicompiexa) in Mississippi oysters. Gulf Res. Rept. 6: 423-425. Perkins, F. 1969. Ultrastructure of vegetative stages in Labyrinthomyxa manna ( = Dennucys- tidium marinum), a commercially significant oys- ter pathogen J. Invert. Pathol. 13: 199-222. Perkins, F. O. 1976. Zoospores of the oyster pathogen, Dermocysiidium martnum. 1. Fine structure of the conoid and otrier spoiozoan- like organelles. J. Parasitol. 62: 959-974. Perkins, F. O. and R. W. Menzel. 1967. Ultras- tructure ot spot ulation in the oyster pathogen Dermocysiidium marinum J. Invert. Pathol. 9: 205-229. Quick, j., Jr. 1971. Patfiological and parasitolog- ical effects of elevated temperatures t>n the oyster Crassoslrea viiginica with emphasis on the pathogen Labnniiiumyxa viarina. In]. A. Quick, Jr. ed. A pieliminaiy investigation: the effect ot elevateu tempeiaiuie on the American oys- ter Crassoslrea virgmica (Gmelin). p. 105-171. Fla. Dept. Nat. Resources. Mar. Res. Lab. Prof. Paper Ser. No. 15. Quick, J. A, Jr. and J. G. Mackin. 1971. Oyster parasitism by Labyrinthomyxa marina in Florida. Fla. Dept. Nat. Rcsour. Mar. Res. Lab. Prof. Paper. Ser. No. 13. 1-55. Ray, S. M. 1954a. Bioit)gical studies oi Dennocys- tidium marinum a tuiigus parasite of oysters. Rice Inst. Pamphlet Moiiogr. Biol. Spec. Issue, 114 p. Ray, S. M. 1954b. Studies on the occurrence of Dermocysiidium marinum in young oysters. Proc. Nad. Shellfish Assoc. 44: 80-92. Ray, S. M. 1966. A review of the culture method for detecting Dermocysiidium marinum with suggested modifications and precautions. Proc. Natl. Shelltlsh. Assoc. 54: 55-69. Ray, S. M. 1982. Summary report on monitoring program in Lake Calcasieu, Louisiana Ray Biological Company, Galveston, Texas. Ray, S. M. 1987. Salinity requirments of the American oyster, Crassoslrea virginica. In A. J. Mueller and G. A. Matthews, eds. Freshwater inflow needs ot the Matagorda Bay system with focus on the neeas of penaeid shrimp p. El- E29. NOAA Technical Memorandum MNFS- SEFC. Ray, S. M. and J. G. Mackin. 1955. Studies of pathogenesis of Dermocysiidium marinum. Proc. Nad. Shelltlsh Assoc. 45: 164-167. Ray, S. M., J. Mackin, and J. L. Boswell. 1953. Quantitative measurement of the effect on oys- ters of disease caused by Dermocysiidium mamium BuH Mar. Sci. Guff and Caribb. 3: 6- 33. SONIAT, T. M. 1985. Changes in levets of infec- tion of oysters Dy terfcinsus marinus, with special reterences to the interaction of temperature and salinity upon parasitism. Northeast Gult Sci. 7: 171-174. SONIAT, T. M. and M. L. KOENIG. 1982. The effects of parasitism by Fekinsus marinus on the free amino acid coiiiposili(jn of Cra.ssoslrea vir- ginica mamfe tissue. J. Shellfish Res. 2: 25-28. Sprague, V. 1954. Protozoa. In P. S. Galtsoff., ed.. Gulf of Mexico: its origin, waters and marine life. U.S. Fish Wildlife Ser. Fish Bull. Vol. 89. p. 243-256. Turner, H. M. 1985. Parasites of eastern oysters from subiiuai reets in a Louisiana estuary with a note on tiieir use as indicators of water qual- ity. Estuaries 8: 323-325. 4 iddiN uy}nz-o/oz TULAME STUDIES U :OOLOGY AMD EOTAMY Volume 27, Number 2 $7.00 February 20, 1990 MCZ LIBRARY MAR 0 5 1990 SOME WEEVILS FROM WI^^^flZm^^ielMJ: AND OTHER HOSTS FROM NLBWVHEBIJDNIA G. KUSCHEL PLANT PROTECTION, DSIR, PRIVATE BAG, AUCKLAND, NEW ZEALAND NOTROPIS AMMOPHILUS, A NEW CYPRINID FISH FROM SOUTHEASTERN UNITED STATES ROYAL D. SUTTKUS MUSEUM OF NATURAL HISTORY, TULANE UNTVERSITY BELLE CHASSE. LOUISIANA 70037 AND HERBERT T. BOSCHUNG DEPARTMENT OF BIOLOGY, THE UNTVERSITY OF ALABAMA TUSCALOOSA, ALABAMA 35487-0344 Tulane University New Orleans TULANE STUDIES IN ZOOLOGY AND BOTANY, a fully refereed publica- tion of the Biology Department of Tulane University, is devoted mainly to pa- pers containing original studies in ecology and systematics, without geographical constraints. Reviews, bibliographies, and symposia results are occasionally pub- lished. Normally two numbers plus an index and table of contents are issued annually. Preferred citation of the journal is Tulane Stud. Zool. and Bot. INFORMATION FOR AUTHORS. Manuscripts should be double-spaced throughout, including Literature Cited and legends to tables and figures, on 8 V2 X II paper, with margins no less than Va inch from top, sides, and bottom. Do not use dot matrix printing. Avoid footnotes; if a footnote is necessary, place it on a separate sheet of paper. Do not use hyphenated words on line endings. Manuscripts should conform to the style of recent issues of Tulane Studies, and figures should be proportioned for one or two columns corresponding to our printed page. Lettering used in illustrative material should be easily read after reduction to our page size. Guidelines for letter and other extraneous markings should be done with a red pencil, not with the traditional blue. We recommend conformance to the CBE Style Manual. Two additional copies of the manuscript should accompany the original. Legends for figures should appear on a separate page and in sequence. An abstract, briefly stating the results of the research, and not to exceed three percent of the length of the article, must accompany the manuscript. Page charges of $45.00/page will be assessed. Authors are urged to provide for page charges through grants or institutional support. Financial assistance from the George Meade Memorial Fund is available if needed. EXCHANGES, SUBSCRIPTIONS, ORDERS FOR INDIVIDUAL COPIES. Exchanges are invited from institutions publishing comparable series. A price list of back issues is available on request. Remittances should be made payable to "Tulane University". Subscription rates are $12.50 per volume domestic, $13.50 per volume foreign. COMMUNICATIONS: Address inquiries about exchanges to Head, Science and Engineering Division, Howard-Tilton Library, Tulane University, New Or- leans, LA 701 18. Editorial and all other matters should be addressed to Editor, TSZ&B, Department of Biology, at the same address. Alfred E. Smalley, Editor Printed by E. O. Painter Printing Company, P. O. Box 877, DeLeon Springs, Florida 32130. Some Weevils from Winteraceae and Other Hosts FROM New Caledonia G. KUSCHEL Plant Protection, DSIR, Private Bag, Auckland, New Zealand Abstract Eleven species of Curculionidae from Win- teraceae and other plant families are described from New Caledonia and the Loyalty Islands. Palontus new genus seems to be adelphic to Aneuma Pascoe from New Zealand; it is placed in the tribe Storeini of the subfamily Curculioninae. There is no evidence to suggest that these weevils occupy a position within Curculionidae compara- ble to that of Winteraceae in Angiospermae. The alphabetical sequence of the new species, with the type localities and the hosts, when known, in parenthesis, is: Palontus bellioli (Tchamba Valley, on Belliolum), P. dispar (Houailou), P. exospermi (Mt Panic, on Exospermum, Belliolum, Bubbia and Zygogynum), P. fidelis (Mare I., Loyalty Is.), P. flebiiis (Pic d'Amoa, on Nothofagus), P. iodes (Poindimie), P. leucopogonis (Plum, on Leocopogon), P. nothofagi (Mt Do, on Nothofagus), P. paniensis (Mt Panic, on Belliolum), P. turritus (Table Unio) and P. varius (Yahoue). Introduction The small family Winteraceae (Mag- noliales) comprises fewer than a hundred species of trees and shrubs oc- curring in Central and South America, eastern Australia, southeast Asia and Madagascar, New Caledonia, Pacific Is- lands and New Zealand. It is often re- garded as the most archaic family of flowering plants or angiosperms (Cron- quist, 1981; Friis, Chaloner and Crane, 1987). As most botanists seem to concur with this view the weevil species found to be associated with flowers and young growth of Winteraceae in New Cal- dedonia and elsewhere were studied to find out if they showed predominantly ancestral or derived features. The earliest fossil record of Win- teraceae, identified from the distinctive pollen grains characterising the family, dates from the Upper Aptian-Lower Al- bian of Israel, some 105 to 110 million years ago, coinciding with late Lower Cretaceous (Walker, Brenner and Walker 1983). Seed plants of early Mesozoic periods still existing today tend to hold a weevil fauna consisting of species that exhibit a conservative habitus rendering them rather uniform. This fauna possesses relatively high numbers of ancestral character-states or plesiomorphies and shows host associa- tions of a narrow range. This pattern is not evident from the faunas thus far known on the Winteraceae of New Caledonia, New Zealand and Chile, the three areas where the species of this plant family have been searched rather intensively. It appears that the weevils now associated with Winteraceae have little to indicate that they are closer to the groundplan of Curculionidae than others occurring on plant families re- garded as evolutionarily more ad- vanced. Pollen is presumed to have been and continues to be a rich food source for many insects, not least among Coleop- tera. Pollen may have been the ancestral diet of adult Cerambycidae, Megalo- podidae, Orsodacnidae, Nemonych- idae, Belidae, Oxycorynidae and some families of Cucujoidea, and also the diet of the larvae of some of the earliest groups, for instance of Nitidulidae, 29 30 Tulane Studies in Zoology and History Vol. 27 Megalopodidae, Nemonychidae (Kus- chel, 1983, and in press). A team of floral biologists from the Department of Biology, Tulane Univer- sity, New Orleans, Louisiana and the De- partment of Ecology and Evolution, State University of New York, Stony Brook, New York have visited New Caledonia since 1979 to observe the pollination of Winteraceae. The Weevil Fauna OF THE Winteraceae of New Caledonia, New Zealand AND Chile A brief survey of the rather scarce weevil fauna associated with Winteraceae in New Caledonia, New Zealand and Chile, the only areas of the world where these plants have been sufficiently explored, is offered here in response to the interest that this plant family has for systematists. Although the number of winteraceous species is rather high in New Caledonia (cf. Guillaumin, 1948) only Exospermum stipitatum, Zygogynum baillonii, Belliolum rivulare and Bubbia pauciflora have been closely examined by the biologists observing their pollination. These plants have produced three weevil species, all referable to the same genus described below as Palontus. Because these three species are very closely related to one or two others associated with Nothofagus and a further species occur- ring on Epacridaceae is only slightly more divergent, the host range of Palontus ap- pears to be too wide for it to claim a relict pattern. In New Zealand, the first fossil Win- teraceae occur in the Upper Cretaceous (Mildenhall, 1980), but the current flora of the country has only three rather closely related species of Pseudowintera (Allan 1961). Vink (1988) in a recent cladistic analysis, considers this genus to be adelphic (= of sister-group relation- ship) to the Zygogynum-com^Xex. of New Caledonia.The only curculionine weevil known from Pseudowintera is Peristoreus di- lucidus (Broun), a rare species whose larva develops in the fruit (May 1987). Al- though distinctive, the species fits easily into Peristoreus, a genus that has about 60 further species associated with plants of such diverse families as Polygonaceae, Elaeocarpaceae, Malvaceae, Epacrida- ceae, Fabaceae, Cornaceae (Griselinia), Oleaceae, Asteraceae and Liliaceae. Chile has two species of Winteraceae, Drimys winteri on the mainland and D. con- fertifolia on the Juan Fernandez Ar- chipelago. The mainland Drimys has two weevil species associated with young growth, flowers and fruits, one belonging to Berberidicola, a genus having several species on Berberis (Berberidaceae) and one each on Ribes (Grossulariaceae) and Nothofagus (Fagaceae), the other belong- ing to a still undescribed genus somewhat related to Berberidicola. The two weevil species belong to the subfamily Molytinae sensu Kuschel (1987), neither exhibiting morphological features or biological characteristics that could be construed as primitive. The same Drimys winteri has a further weevil species, Aegorhinus luteolus Kuschel whose larva ring-barks live trees usually close to the base of the trunk. This particular species is more closely re- lated to A. nodipennis (Hope) associated with Proteaceae, Eucryphiaceae and others than to any of the eight or so Ater- pinae known to be confined strictly, as adult and larva, to the Nothofagus species of Chile. The Fernandezian Drimys confertifolia holds one weevil species on Robinson Crusoe Island (or Masatierra) and another on Alejandro Selkirk Island (or Masafuera). They belong to the trypeti- dine genus Platynanus (Molytinae) whose species have successfully colonised almost every fern of suitable size and almost all the dicot genera and also the p^hnjuania No. 2 Weevih from New Caledonia 31 australis amongst the monocots. Strangely enough Platynanus has no species on Lac- toris femandeziana, sole species of the Lac- toridaceae, an endemic family to Robin- son Crusoe Island that Cronquist (1981) and other botanists have tentatively as- signed to Magnoliales. Platynanus is re- lated to Trypetidini not from the Neot- ropical Region but from the southwest Pacific, to Pogonorhinus of ferns and Arecoplmga of palms. The few depositories for the speci- mens dealt with in the paper are indi- cated by their standard abbreviations as given in Arnett and Samuelson (1986). These are: BPBM: Bernice P. Bishop Museum, Honolulu, Hawaii. ISNB: Institut Royal des Sciences Naturelles de Belgique, 29 rue Vautier, Brussels, Belgium. MNHN: Museum National d'His- toire Naturelle, 45 Rue de Buffon, Paris, France. NZAC: New Zealand Arthropod Collection, Plant Protection, DSIR, Private Bag, Auckland, New Zealand. Acknowledgments I wish to thank G. McPherson, O. Pellmyr, and L. B. Thien, members of the team of biologists who provided the specimens from the Winteraceae of New Caledonia, but in particular L. B. Thien for inducing me to write the paper. I am grateful for the loan of specimens to G. A. Samuelson, Bernice P. Bishop Museum, Honolulu and L. Baert, Institut Royal des Sciences Naturelles de Bel- gique, Brussels, and also for assistance in one way or another to R. C. Craw, J. S. Dugdale, D. W. Helmore, G. M. Herbert, B. A. HoUoway, K. H. Jack and J. C. Watt. Scientists at the Service des Forets et du Patrimoine Naturel and O.R.S.T.O.M. graciously provided facilities and their time to help collect in- sects. The Erna and Victor Hasselblad Foundation contributed funds for the habitus drawing of Palontus exospermi (Fig. 1). Palontus new genus (Curculioninae Storeini) (palontus: from the Greek pale = dust, pollen and on, ontos = being, creature; gender masculine) Eyes transverse, flat or nearly so. Funicle 7-segmented; club segments gradually increasing in length. Prothorax truncate at base, constricted at apex to form distinct collar. Scutel- lum visible, usually conically raised above level of elytra. Elytra parallel at basal half, considerably wider than prothorax at shoulders, with vertical or slightly proclinate base, 10-striate, striae distinctly sulcate also at declivity, 1 0th finely margined outward for most of length. Hind wings functional. Proster- num with canal, sidewalls of canal con- tinuous, not interrupted or articulated at apical constriction, lacking pit on out- side at constriction; fore coxae contigu- ous or subcontiguous. Femora with or without tooth. Tibiae uncinate or muc- ronate. Claws divaricate, laminate. Stridulatory Files. Present in male as long, narrow iridescent file next to suture or sutural flange on underside at apex of each elytron and 3 (rarely 2) pairs of spiculate plectral granules ar- ranged in 2 parallel rows rather close to median line of tergite 7, Fig. 2; files ab- sent in female. Male. Sternite 8 as in Fig. 3, nar- rowly pigmented at apical margin, pig- mented area of blade divided into 2 con- tiguous or disjunct, strongly transverse plates having row of submarginal setae. Sternite 9 as in Fig. 3, with broad, angu- late, asymmetrical arms, with robust, 32 Tulane Studies in Zoology and History Vol. 27 s (J _o o X c S Oh No. 2 Weevils from New Caledonia 33 towards the apex strongly curved apodeme. Tegmen with complete ring but no parameres. Aedeagus robust, long, extending into metathoracic cav- ity, apodemes articulated laterobasally, usually shorter than aedeagal body; in- ternal sac contained inside aedeagal body, unless basal sclerite transformed to long flagellum. Female. Sternite 8 for all species much as in Fig. 6. Hemisternites elon- gate, broad at base, tapering toward apex, weakly pigmented, undivided, with pores, without distinct setae; styli well developed, rather long. Vagina and bursa short, very broad, bursa occasion- ally heavily pigmented to greater or les- ser extent. Spermatheca variable in shape, specifically distinctive, with small gland; duct extremely variable in thick- ness and length, usually thicker at bur- sal end, insertion on bursa specifically characteristic, ventral, apical or dorsal. Type Species. Palontus exospermi n.sp. HoSTPLANTS. Winteraceae, Fagaceae, Epacridaceae. Distribution. New Caledonia and Loyalty Islands. Remarks. Palontus agrees exter- nally with Aneuma from New Zealand but the latter has rather small and coarse stridulatory files, no distinct plectral granules and the internal sac of the aedeagus is long, exposed a good deal between the apodemes although a flagellum is lacking. The larvae of Palontus are not known. The larvae of the probably quite closely related genus Aneuma are confined to Pittosporum (Pit- tosporaceae), those of A. fulvipes Pascoe developing in the flower buds, A. rub- ricale (Broun) living in the open flowers, and A. fasciatum (Broun) mining the leaves, the larvae of all three species dropping to the ground to pupate in the soil. The Relationships of Palontus The three species of Palontus from Winteraceae, the one from Nothofagm and a further species from an unknown host share an aedeagus that has dis- tincdy sinuous sides and a basal sclerite that is unusually large and is contained in the aedeagal body in the sense that it does not show up between the apodemes but may be extended beyond the ostium. These features have not been observed elsewhere and con- sequendy are taken to be derived and indicative of monophyly. A flagellum of the size and type shown for the last four species described below is a further de- rived attribute not seen in possible out- groups. Finally, a conically raised scutel- lum is a rare occurrence in Cur- culionidae as a whole, and yet is found in eight out of the eleven species of Palontus and nowhere else in Storeini and nearby tribes. Clark, Whitehead and Warner (1977) classified a group of curculionine weevils, traditionally called Tychiinae, into four tribes. The authors discussed the interrelationships of the Palearctic, Nearctic and Neotropical genera and the tribe Endaeini, left in limbo since Burke (1976) had excluded it from Kiss- inger's (1964) placement in the An- thonominae. It is yet too early to be positive about the phylogenetic relation- ships of these and hundreds of other genera now incorporated in Cur- culioninae (Kuschel 1982). Palontus species have divaricate claws, a male sternite 9 with broad, angulate, asym- metrical arms, and elytra with stridulat- ory files on the underside next to the sutural flange at the tip in males only. These attributes are those oi Storeus and most Storeini (Lacordaire 1863), a tribe well represented in the Australian Re- gion. The genera that differ from Sto- reini by having stridulatory files in both 34 Tulane Studies in Zoology and History Vol. 27 sexes are grouped in Ochyromerini (Voss 1935) and Dinorhopalini (Voss 1936). The storeine fauna of New Caledonia is rich in genera and species but few are described. Aporotaxus and Gryphosternus are endemic genera hold- ing the largest known species of the tribe, up to 13 mm long in the case of Aporotaxus kanalensis Perroud whose larva develops in the fruit of Ficus species. The New Zealand fauna is simi- larly diverse and rich. A number of storeine genera have a prosternal canal, as Palontus has, but then the sidewalls of the canal are distinctly articulated at the apical constriction and a deep pit is showing on the outside by the articula- tion. Palontus has these sidewalls con- tinuous all the way from the fore coxae to the apex and lacks the excavation on the outside, sharing these characters with Aneuma from New Zealand and Rhinidotasia from Queensland. Al- though Palontus and Aneuma resemble each other closely and are likely closely interrelated, it is not possible to be sure that there is an adelphic relationship be- tween them without a more thorough study than the present one, not only of the Storeini, but also of the Ochyrome- rini, Endaeini, Derelomini and other re- lated tribes. Rhinidotasia differs consid- erably from Palontus and Aneuma in being rather glabrous and smooth, of rhomboidal, continuous outlines, in having narrowly separated eyes, broadly separated fore and middle coxae, long, slender, almost lineal femora and dis- tinctly striate tibiae. Stridulatory files on the underside of the elytra seem to have appeared rather late in the superfamily consider- ing that the fossil evidence traces weevils back to the Jurassic (Arnoldi 1977) and, according to Zherichin (1977), even to the Triassic, and that the files are absent in the families of extant Orthoceri. These are absent also in Brachycerinae, Ithycerinae, Erirhininae, Rhyn- chophorinae, Entiminae and a whole series of "subfamilies" more closely re- lated to the broad-nosed weevils (Adelognatha) than to the long-nosed weevils (Phanerognatha). Files are a common occurrence in the latter group. The adults of the Curculionidae, again sensu Kuschel (1982), have the files, if present, confined to the male sex, ex- cept for Ochyromerini and Dinorhopa- lini; the other groups have them, again if present, almost always in both sexes. Elytral files appear to be a derived attribute in Curculioninae. The Chilean and Australian Derelomini associated with the male strobili of Araucariaceae have, in the male, a 9th sternite with a large median plate detached from the symmetrical arms, and a tegmen with distinct parameres, and a lack of elytral files. The files are missing also from all other Curculioninae of Chile, including a dozen or so associated with Nothofagus, and from the dozen species occurring on Nothofagus in New Zealand whilst nearly all of the close to a hundred other Zelandic Curculioninae associated with flowering plants other than Nothofagus have them. A rather com- mon weevil on the Nothofagus species of New Caledonia agrees generically well with Neomycta species of New Zealand, and like all other species of the genus, lacks the elytral files, too. The presence of files in Palontus indicates that the genus would not easily qualify as a relict within Storeini. It is virtually certain that the Cur- culioninae of flowering plants of Chile are not related to the Storeini of New Zealand and New Caledonia in the sense of an adelphic relationship. New Caledonia and New Zealand are be- lieved to have separated from Austro- Antarctica on a common continental block 78 to 56 million years ago creating No. 2 Weevils from New Caledonia 35 the Tasman Sea (Coleman, 1980) and presumably taking with them some Sto- - reini. Later on, "towards the end of Eocene times, about 37 million years 2. ago, came the parting of the ways for New Caledonia and New Zealand" (Ste- vens 1981). As already stated, fossil evidence shows that Winteraceae have existed at least since early Middle Cretaceous 3. times. Although biogeographic patterns of the extant weevil fauna would suggest that representatives of the sub- family Curculioninae should have been around since late Upper Cretaceous, none of the Mesozoic fossils thus far has been recognized as belonging to a species of the long-nosed weevils and only one of the broad-nosed weevils is - known (Kuschel 1959). It was described from an area north of Magellan Strait from a Maestrichtian deposit of very late Upper Cretaceous. The recently discovered fossiliferous Orapa deposits of Botswana dated as of Middle to Upper Cretaceous times (McKay &: 4. Rayner 1986) contain rather well pre- served impressions of weevils but those that are identifiable with some confi- - dence from the photographic copies made available (R. J. Rayner, pers. comm.) belong to families of the 5. evolutionarily less advanced Orthoceri, apparently, to Attelabidae and the fam- ily-complex Brentidae/Apionidae. From the above evidence, the species of Palontus now found on Winteraceae in New Caledonia and, similarly, those of Elleschodes of the same tribe Storeini - pollinating Eupomatiaceae in Australia, show little that might suggest they are relicts within Curculioninae or its tribe Storeini. Key to Species of Palontus 1. Metasternum with prominent tubercle at hind angle. 1.9-2.6 mm 1. exospermi Metasternum lacking tubercle at hind angle 2 Metepisterntmi with 1 row ol hairs or scales on posterior half 3 Metepisternum with 2 or more rows of hairs or scales on pos- terior half 4 Dark markings on basal quarter of elytra consisting of more or less coalescent dots or short stripes on interstriate 3, 5 and 7, one on 5 slighdy farther back than others. Elytral vesti- ture consisting mostly of ellip- tic or oval scales. 2.0- 2.5 mm 2. bellioli Dark markings on basal quarter of elytra coalescent, forming anteriorly concave arc or ex- tending to interstriae 5. Elytral vestiture consisting mostly of hairs or lineal scales. 1.7 mm. 3. paniensis Scutellum raised conically high above level of elytra. Femora with strong tooth 5 Scuttelum flat or only slighdy raised, if raised then middle and hind femora unarmed. . 6 Sides of prothorax pubescent. Elytra variegated with pale reddish and dark brown, with large dark triangular marking in scutellar area and postme- dian band across elytra. 2.0- 2.5 mm 4. turritus Sides of prothorax white, cov- ered in dense broad scales. Elytra rusty brown but usually with reddish brown, broad sutural stripe involving first three interstriae from basal quarter to apex, this stripe traversed by darkish band in middle. 2.20-2.45 mm 5. nothofagi 36 Tulane Studies in Zoology and History Vol. 27 6. All femora with distinct tooth. 7 — All or some femora lacking tooth 8 7. Elytra squamose on reddish brown integument. Interstriae not granulose. 1 .9- 2.4 mm 6. flebilis — Elytra pubescent on rusty brown integument. Interstriae finely granulose. 2.7-3.0 mm. 11. iodes 8. Scuttelum large, flat, shiny. 1.9-2.5 mm 7. varius — Scuttelum small, raised, dull. 9 9. Elytra with tubercle on intrestria 3. Pronotum humped on mid- dle in lateral view. 2.5- 2.8 mm 8. dispar — Elytra lacking tubercle on in- terstria 3. Pronotum convex on middle in lateral view 10 10. Elytra distinctly variegated with white, brown and black mark- ings. Rostrum only faindy sad- dled at base. 2.3- 2.6 mm 9. fidelis — Elytra not variegated, but with suf- fused bands across dorsum. Ros- trum distinctly saddled at base. 2.6-2.7 mm 10. leuco- pogonis 1. Palontus exospermi n.sp. Figs. 1-8 Integument pale or dark reddish brown. Vestiture forming nebulous de- signs (Fig. 1), consisting of dark dots ac- ross basal third of elytra, loose band ac- ross dorsum just beyond middle, usually better defined band across top of decliv- ity, and 1 or 2 dots on each side halfway down declivity. Head dull, shallowly punctate, usually squamose above and behind eyes. Ros- trum in male rather robust, gently curved, 3 x as long as broad at apex, 0.70- 0.75 length of prothorax, dull, with 3 rows of very fine puncta on each side, the rows separated by very fine carinules; prerostrum shiny, finely punctate, broader than long. Rostrum in female more slender, slightly more curved, 4 x as long as broad at apex, 0.88-0.90 length of prothorax, lacking carinules; preros- trum 1.5 X as long as broad. Prothorax broader than long, weakly convex, den- sely granulose-punctate. Scutellum small, dark, convex but not distinctly raised above level of elytra, finely punctate, sparsely pubescent. Elytra weakly convex across middle and from middle to apex. Fore coxae narrowly separate, pubes- cence on top slightly longer and ruffled in males. Metasternum with prominent tubercle at hind angle. Fig. 1 ; metepister- num usually with 2 rows of squamiferous puncta up to middle or beyond. Ventrite 5 in female at most with small and shallow median fovea. Femora with sharp tooth, lower margin fringed with erect hairs in male; tibiae sinuous on lower edge, fore tibiae incurved in male; fore and middle tibiae uncinate, others mucronate, distal corbels of middle tibiae obliquely ascend- ing. Male. Tergite 7 as in Fig. 2. Sternites 8 and 9 as in Fig. 3, plates of sternite 8 contiguous, suffused in middle. Aedeagus as in Figs. 4, 5, constricted be- fore apex; basal sclerite robust, with tubu- lar extension. Female. Sternite 8 as in Fig. 6. Tip of hemisternites as in Fig. 7. Spermatheca and part of bursa as in Fig. 8; sper- mathecal duct short, very broad. Length: 1.9-2.6 mm; width: 0.9-1.3 mm. Distribution: New Caledonia. Mt Panic, 280 m, 6 Dec 1979, 29 specimens on Exospermum stipitatum flowers, McPher- son; Mt Panic, 300 m, 2 Nov 1983, 23 specimens on E. stipitatum flowers, L. Thien; Mt Panic, 300 m, 27 Nov 1983, 1 No. 2 Weevils from New Caledonia 37 specimen on E. stipitatwn flowers, L. Thien; Mt Panic, 440 m, 27 Nov 1983, 1 specimen, Z. Mackeel; Mt Panic, 295 m, 30 Oct 1986, 3 specimens on E. stipitatum flowers, O. Pellmyr & R. Brown; Mt Panic, 700 m, 30 Oct 1986, 1 specimen on Bubbia pauciflora, O. Pellmyr & R. Brown; Mt Do, 800 m, 5 Jul 1980, 6 specimens on Zygogynum flowers, McPherson; Mt Me Ori, 500 m, 21 Nov 1979, 2 specimens on E. stipitatum flowers, McPherson; Mt Me Ori, 13-14 Dec 1983, 8 specimens on E. stipitatum flowers, L. Thien; Mt Dzumac, 800 m, 28-29 Oct 1983, 1 specimen on Zygogynum baillonii flower, L. Thien; Riviere Bleu, 150 m, 24 Jul 1980, 7 specimens on Zygogynum, McPherson. 82 specimens in all. Holotype male, 2.5 x 1.2 mm, Mt Panic, 300 m, 2 Nov 1983, illustrated in Fig. 1, NZAC. Paratypes at BPBM, ISNB, MNHN, NZAC. HosTPLANTS. Adults feeding on pollen of Exospermum stipitatum, Bubbia pauciflora and Zygogynum baillonii, all Winteraceae. Vink (1988) prefers to treat the three genera as one and uses the name Zygogynum. Remarks. Palontus exospermi is rec- ognizable instantly by the prominent tubercle on the hind angles of the metasternum shown in Fig. 1 just in front of the hind femora. 2. Palontus bellioli n.sp. Figs. 9-11 Integument reddish brown, sterna usually darkened. Vestiture with mark- ings similar to those shown for P. exos- permi in Fig. 1. Head dull, finely punctate, pubes- cent or squamose in part. Rostrum in male slender, gently curved, 4.5 x lon- ger than broad at apex, about as long as prothorax; postrostrum dull, with rows of fine puncta, the areas between the rows slightly raised to very fine carinules; prerostrum shiny, finely punctate, about as long as broad. Ros- trum in female more slender, slightly more curved, 6.5 X longer than broad at apex, about 1.5 x lengdi of prothorax; postrostrum very finely punctate, lack- ing carinules; prerostrum 2 or 3 x longer than broad. Pronotum densely and coarsely punctate. Scuttelum small, conically raised slightly above level ol elytra, dull. Elytra moderately convex across middle, distinctly convex in lat- eral view between basal third and apex. Fore coxae contiguous or nearly so. Metasternum lacking tubercle on hind angles; metepisternum with single row of squamiferous puncta for most of length. Ventrite 5 with rather large and deep median fovea at apex in female. Femora armed with small tooth, occa- sionally very small, hardly noticeable on hind femora. Tibiae straight on lower edge; fore tibiae straight in male; fore and middle tibiae uncinate, others muc- ronate; distal corbels of middle tibiae not distinctly ascending. Male. Aedeagus as in Figs. 9, 10, apical half shaped like long, truncate spatula; basal sclerite large, robust, with dorsal tubular extension. Female. Spermatheca as in Fig. 11, duct long and fine. Length: 2.0-2.5 mm; width: 1.0- 1.15 mm. Distribution: New Caledonia. Tchamba Valley, 40 m, 1 Nov 1983, 6 specimens on Belliolum rivulare, L. Thien; Thy Valley, 28 Sep 1979, 2 specimens on Belliolum cf. crassifolium, McPherson. Holotype male, 2.1 x 1.0 mm, Tchamba Valley, NZAC. Paratypes at BPBM and NZAC. HoSTPLANTS. Belliolum rivulare, B. cf. crassifolium. Remarks. A single row of scales on the metepisternum distinguishes P. bel- lioli and P. paniensis from all others; these two species are compared more closely in the following description. 38 Tulane Studies in Zoology and History Vol. 27 3. Pa/owfM5 paniensis n.sp. Figs. 12, 13 Hardly distinguishable from P. bel- lioli on external characters except perhaps for smaller size and for design patterns on elytra. Markings consisting of suffused darkish arc extending from basal quarter of suture forward to in- terstria 5, suffused band across dorsum just beyond middle and very faint dar- kening on each side at top of declivity. Rostrum slightly more curved, rostral carinules less distinct, elytral vestiture finer by consisting mostly of hairs or lineal scales. Male. Aedeagus as in Figs. 12, 13, distinctly different in shape and arma- ture from all other species; basal sclerite lacking tubular extension. Female. Unknown Length: 1.7 mm; width: 0.8 mm Distribution: New Caledonia. Mt Panic, 450 m, 20 Jul 1980, 1 male on Belliolum sp., McPherson. Holotype male, data as above, NZAC. Remarks. P. paniensis is defined by the aedeagal features and the characters indicated in the key to species. HOSTPLANT. Belliolum sp. 4. Palontus turritus n.sp. Figs. 14, 15 Dark or reddish brown, but patterns variable, elytra variegated but rather suffused, with large triangular dark patch in scutellar area and postmedian band across dorsum. Vestiture consist- ing of pubescence and scaling, latter on paler areas. Head dull, slightly asperate. Ros- trum robust, gently curved, 3.5 x longer than broad at apex, 0.75 the length of prothorax, carinulate; preros- trum strongly transverse, nearly twice as broad as long. Prothorax as long as broad, weakly rounded on sides, de- nsely and coarsely punctate, each punctum with raised margin on post- erior half. Scutellum conically protrud- ing well above level of elytra. Fore coxae subcontiguous. Metepisternum with 2 ill-defined rows of puncta. All femora with rather long sharp tooth. Fore tibiae straight but lower edge slightly sinuous, all tibiae with rather small mucro. Male. Sternite 8 with contiguous plates. Aedeagus as in Figs. 14, 15, shape and armature highly characteris- tic, the ostial margins finely and sparsely ciliate. Female. Unknown. Length: 2.0-2.5 mm; width: 1.0- 1.15 mm. Distribution: New Caledonia. Table Unio, 550-1000 m, 16 Oct 1978, 2 males, J. C. Watt; Mt Rembai, 700 m, 18 Oct 1978, 1 male, J. C. Watt; Mt Do, 1000 m, 31 Oct 1978, 1 male, J. C. Watt; Noumea, 1 male. Coll. Fauvel. Holotype male, 2.4 x 1.15 mm, Table Unio, NZAC. Paratypes at BPBM, ISNB, NZAC. Hostplant. Unknown. Remarks. P. turritus resembles closely P. bellioli but may be distin- guished readily by a highly towered up ( = turritus) scutellum, strongly dentate femora and a very distinctive aedeagus. 5. Palontus nothofagi n.sp. Figs. 16-19 Integument reddish brown on ros- trum, prothorax, first 3 interstriae from basal quarter of elytra, underside and legs; rusty brown on head capsule and remainder of elytra. Vestiture pubes- cent except for white scales on sides of head behind eyes, sides of prothorax, elytral suture at basal quarter, pleurites and sides of metasternum. Head dull, shallowly punctate, slightly asperate. Rostrum similar in both sexes, dull, nearly straight, at least 3 X longer than broad at apex, 0.75 length of prothorax, not carinulate, rows of hairs or lineal scales slightly more distinct in male. Prothorax, about No. 2 Weevils from New Caledonia 89 as long as broad, weakly rounded on sides, densely and coarsely punctate, puncta rimmed at posterior half. Scutel- lum raised well above level of elytra. Ventrite 5 with shallow median fovea near apex in female. All femora with strong, sharp tooth; tibiae slightly sinu- ous on lower edge, fore and middle tibiae uncinate, others mucronate, mid- dle tibiae with obliquely ascending distal comb. Male. Tergite 7 with 2 pairs of dis- tinct granules. Sternite 8 with contigu- ous plates. Aedeagus as in Figs. 16, 17, distinctive in dorsal and lateral views. Female. Hemisternal apex as in Fig. 18, stylus unusually long. Bursa with partly folded sclerite. Fig. 19. Sper- matheca as in same Fig. 19, duct very fine at spermathecal end, very wide at bursal end. Length: 2.2-2.6 mm; width: 1.05- 1.20 mm. Distribution: New Caledonia. Mt Do, 900-1030 m, 22 Oct 1978, 2 pairs on Nothofagus codonandra, G. Kuschel; Pic du Pin, 6 Oct 1978, 1 female by beat- ing, J. C. Watt. Holotype male, 2.2 x 1.05 mm, Mt Do, NZAC. Paratypes at BPBM and NZAC. HOSTPLANTS. Nothofagus codon- andra and, presumably, N. aequilateralis, the only southern beech growing at Pic du Pin. Remarks. P. nothofagi stands out in colour and vestiture from all others of the genus. 6. Palontus flebilis n.sp. Fig. 20 Integument reddish brown. Vesti- ture consisting mostly of small yellowish scales on dorsal surface, of larger white scales on underside, elytra crossed by faint band of dots on basal third, band just behind middle and one at top of declivity. Head dull, alutaceous, very shal- lowly, indistinctly punctate, not asper- ate; pubescent or scaly. Rostrum moder- ately shiny, weakly curved, slightly more than 3 x longer than broad at apex, about 0.93 length of prothorax, carinu- late between rows of puncta; preros- trum half as long again as broad. Prothorax slightly transverse, 1.10-1.12 X broader than long, moderately con- vex in lateral view, densely punctate. Scutellum small, dull, raised to low cone. Fore coxae subcontiguous. Fem- ora with strong, sharp tooth on fore and middle pairs, with smaller tooth on hind pair; fore and middle tibiae uncinate, hind ones mucronate. Male. Unknown. Female. Spermatheca as in Fig. 20, with large, wrinkled extension toward very small gland; duct fine throughout, short, inserted terminally on bursa. Length: 1.9-2.4 mm; width: 0.95- 1.15 mm. DiSTRiBUi ion: New Caledonia. Pic d'Amoa, 450-500 m, nr Poindimie, 26 Oct 1978, 2 females on Nothofagus aequilateralis, G. Kuschel. Holotype female, 2.4 x L15 mm and paratype at NZAC. HoSTPLANT. Nothofagus aequilateralis. Remarks. Although the sper- matheca of P. flebilis differs considera- bly from that of the other species, indi- cations are that P. flebilis is related to the preceding P. nothofagi more closely than to the following species. All efforts and time spent to secure more speci- mens were in vain, hence the species name flebilis, Latin for pitiable. 7. Palontus varius n.sp. Figs. 21, 22, 27. Integument reddish brown but elytra always, and other parts often, darkened to dark or blackish brown, es- pecially head and sterna. Vestiture con- sisting of pubescence and scaling, but scales usually dominant on elytra, form- 40 Tulane Studies in Zoology and History Vol. 27 ing variegated patterns on elytra, white or yellowish scales mainly at base of in- terstria 3, on humeral area and on api- cal third where usually forming rather broad band across top of declivity. Head dull, alutaceous, indistinctly, and very shallowly punctate, not asper- ate, pubescent except on frons. Rostrum slender, curved; in male dull, 4.5 x longer than broad at apex, at least l.lOx length of prothorax, finely carinulate between rows of puncta; pre- rostrum nearly twice as long as wide in male, fully twice as long as wide in female. Prothorax distinctly transverse, 1.2-1.3 X wider than long. Scutellum large, flat, shiny. Fore coxae contiguous. Femora unarmed or tooth on front ones appearing as small granule. Tibiae straight, fore and middle pair with small mucro, hind pair indistinctly mucro- nate. Male. Sternite 8 narrowly divided into 2 plates. Aedeagus as in Figs. 21, 22, of characteristic shape in dorsal and lateral views, body elongate, hardly tap- ering toward apex; lacking distinct basal sclerite but internal sac with rather strik- ing longitudinal folds. Female. Bursa and spermatheca as in Fig. 27, bursa with large sclerite, sper- mathecal duct short, fine at both ends. Length: 1.9-2.5 mm; width: 1.0- 1.3 mm. Distribution: New Caledonia. Yahoue, 22 Jan 1963, 1 male, G. Kus- chel; Yahoue, 11 Oct 1978, 1 male by sweeping, J. S. Dugdale; between Noumea and Bourail, 2 pairs. Coll. Fauvel. Holotype male, 2.05 x 1.05 mm, Yahoue, NZAC. Paratypes at ISNB and NZAC. HOSTPLANT. Unknown. Remarks. Distinguishable from all other species of the genus by having a large, flat, shiny scutellum. The lack of a distinct basal sclerite is thus far unique in the genus. 8. Palontus dispar n.sp. Figs. 23, 24 Integument dark reddish brown. Vestiture consisting of pubescence and some pale scales, scales mainly on base of interstria 3, behind humeral callus, before middle of elytra and on under- side. Head dull, densely but very shal- lowly punctate, squamose. Rostrum moderately robust, dull or slightly shiny, curved, about 4 x as long as broad at apex, slightly shorter than prothorax, rather coarsely punctate, areas between rows of puncta carinu- late; prerostrum as long as broad. Prothorax 1.1 x wider than long, strongly rounded on sides, so strongly convex as to appear humped on middle in lateral view, densely and coarsely punctate. Scutellum small, dull, slightly to distinctly raised anteriorly. Elytra rather broad, convex throughout in lat- eral view, interstria 3 on middle with elongated tubercle topped with black vestiture. Fore coxae contiguous. Fore femora with small tooth, others at most asperate. Fore and middle tibiae with small mucro, hind tibiae with very small mucro. Male. Sternite 8 with plates rather broadly separated. Aedeagus as in Figs. 23, 24, basal sclerite modified to flagel- lum as long as entire aedeagus. Female. Unknown. Length: 2.6-2.8 mm; width: 1.35- 1.60 mm. Distribution: New Caledonia. Houailou, 30 Oct 1978, 1 male beaten at night, J. C. Watt; Bale du Prony, 1 male. Coll. Fauvel; Plaine des Lacs, 1959, 1 male, N. L. H. Krauss. Holotype male, 2.6 x 1.35 mm. No. 2 Weevils from New Caledonia 41 Houailou, NZAC. Paratypes at BPBM, and ISNB. HOSTPLANT. Unknown. Remarks. Readily separated from all other species of the genus by the pre- sence of a hump on the pronotum and tubercle on the elytra giving the weevil a rather disparate facies expressed with the Latin adjective 'dispar'. 9. Palontus fidelis n.sp. Figs. 25, 26, 28 Colour and patterns much as in two preceding species. Head dull, shallowly punctate, squamose. Rostrum slender, curved, in male slightly more than 3.5 x longer than broad at apex, 0.90 length of prothorax, carinulate, apical part (pre- rostrum) about as long as wide; rostrum in female slightly more than 4 x as long as wide, about 0.95 length of prothorax, also carinulate between rows of puncta, apical part about twice as long as wide. Prothorax 1.15-1.20 x broader than long, strongly rounded on sides, dis- tinctly convex in lateral view. Scutellum small, dull, raised to low point. Fore coxae contiguous. All femora unarmed. Mucro small on fore and middle tibiae, indistinct on hind tibiae. Male. Aedeagus as in Figs. 25, 26, slightly asymmetrical at apex, with long flagellum. Female. Bursa and spermatheca as in Fig. 28, duct very long, thicker at bur- sal end, inserted on bursal apex. Length: 2.3-2.6 mm; width: 1.20- 1.35 mm. Distribution: Loyalty Islands. Mare I., La Roche, March 1959, 1 male, 3 females, N. L. H. Krauss. Holotype male, 2.5 x 1.3 mm, data as above, BPBM. Paratypes at BPBM and NZAC. HoSTPLANT. Unknown. Remarks. Similar externally and in- ternally to P. dispar but lacking the humped prothorax and tuberculate elytra of the latter. The species name Tidelis' is Latin for loyal. 10. Palontus leucopogonis n.sp. Figs. 29,30 Integument uniformly reddish brown. Vestiture consisting mainly of sil- very grey scales with few subdued darker markings, forming 3 stripes on pro- notum, median band across middle of elytra, small area at suture on top of de- clivity and a sutural stripe at basal third. Head dull, densely covered in white and brown scales. Rostrum slender, 4.2- 4.3 X as long as broad at apex, 0.90-0.95 length of prothorax, distinctly saddled against frons, curved, dull, carinulate be- tween rows of puncta; prerostrum only slightly longer than broad. Prothorax as long as broad, gendy rounded on sides, moderately convex in lateral view. Scutel- lum small, dull, raised slightly above level of elytra. All femora unarmed or fore femora alone with minute tubercle. All tibiae distinctly mucronate but mucro on hind pair smaller. Male. Sternite 8 with broadly sepa- rate plates. Aedeagus as in Figs. 29, 30, with flagelliform basal sclerite, the flagel- lum hardly longer than aedeagal body. Female. Unknown. Length: 2.6-2.7 mm; width: 1.3 mm Distribution: New Caledonia. Plum, 8 Oct 1978, 2 males on Leucopogon (Cyathodes on label), G. Kuschel. Holotype male, 2.6 x 1.3 mm, data as above, and paratype at NZAC. HoSTPLANT. Leucopogon sp., with sil- very white leaves on the underside (Epac- ridaceae). Remarks. P. leucopogonis is charac- terised by its frontorostral area being dis- tincdy depressed as a saddle seat and the silvery white scaling. 42 Tulane Studies in Zoology and History Vol. 27 Fig. 2-8 Palontus exospermi n.sp., Mt Panie, 2. tergite 7 of male, dorsal; 3. sternites 8 and 9, ventral; 4. aedeagus, dorsal; 5. aedeagus, lateral; 6. sternite 8 of female, ventral; 7. hemisternal apex; 8. sper- matheca and part of bursa. Scales = 0.2 mm. Weevils from New Caledonia Figs. 9-13. Palontus bellioli n.sp., Tchamba Valley: 9. aedeagus, dorsal; 10. aedeagus, lateral; 1 1. sper- matheca with full length of duct. P. paniensis n.sp., Mt Panie: 12. aedeagus, dorsal; Vi. aedeagus, lateral. Scale = 0.2 mm. 11. Palontus iodes n.sp. Figs. 31-33. Integument reddish brown, sterna usually darker, discretely pubescent with fine, usually whitish hairs, these slightly coarser on sides of pronotum, at base of interstria 6, across top of de- clivity there forming narrow pale band. Head dull, densely, shallowly punctate. Rostrum robust, gently curved, 2.5-2.6 X as long as broad at apex, about 0.80 length of prothorax, not carinulate. Prothorax nearly 1.10 x broader than long, gently rounded on sides, almost flat dorsally in lateral view, 44 Tulane Studies in Zoology and History Vol. 27 Figs. 14-20. Palontus turritus n.sp., Mt Do: 14. aedeagus, dorsal; 15. aedeagus, lateral. P. nothofagi n.sp., Mt Do: 16. aedeagus, dorsal; 17. aedeagus, lateral; 18. hemisternal apex; 19. spermatheca and part of bursa. P. flebilis n.sp., Pic d'Amoa: 20. spermatheca and bursa. Scale = 0.2 mm. No. 2 Weevils from New Caledonia 45 Figs. 21-28. Palontus varius n.sp., Noumea/Bourail: 21. aedeagus, donsal; 22. aedeagus, lateral; 27. spermatheca and bursa. P. dispar n.sp., Houailou: 2.^. aedeagus, dorsal; 24. aedeagus, lateral. P. fidelis n.sp., I. Mare: 25. aedeagus, dorsal; 26. aedeagus, lateral; 28. spermatheca and bursa. Scales = 0.2 mm. 46 Tulane Studies in Zoology and History Vol. 27 30 Figs. 29-33. Palontus leocopogonis n.sp., Plum: 29. aedeagus, dorsal; 30. aedeagus, lateral. P. iodes n.sp., Poindimie: 31. aedeagus, lateral; 32. aedeagus, dorsal; 33. spermatheca and bursa. Scales = 0.2 mm. densely punctate with raised intervals, rendering it rugose. Scutellum small, dull, flat or nearly so. Elytra weakly con- vex across basal half, dorsum straight for most part in lateral view. Fore coxae contiguous or nearly so. All femora with strong, sharp tooth. Tibiae slightly sinu- ous on lower edge, fore and middle pairs uncinate, hind pair mucronate. Male. Sternite 8 with contiguous plates. Aedeagus as in Figs. 31, 32; apodemes twice as long as aedeagal body; internal sac with very long flagel- lum, this longer than entire aedeagus. No. 2 Weevils from New Caledonia 47 Female. Bursa and spermatheca as in Fig. 33, spermathecal duct wider at bursal end and inserted dorsally on bursa. Length: 2.7-3.0 mm; width: 1.30- 1.45 mm. Distribution: New Caledonia. Poindimie, 11 Feb 1963, 8 specimens, C. M. Yoshimoto and G. Kuschel. Holotype male, 2.7 x 1.3 mm data as above, NZAC. Paratypes at BPBM and NZAC. HoSTPLANT. Unknown. Remarks. P. iodes is distinguished by having a much more robust rostrum than any of the other species of the genus and by its distinctive colour. It differs considerably also by being the only species that has an aedeagus whose body is shorter than the apodemes. The specific epithet 'iodes' is Greek for rusty. Literature Cited Allan, H. H. 1961. Flora of New Zealand 1: LIV+ 1085 pp. R. E. Owen, Wellington. Arnett, R. H. and G. A. Samuelson. 1986. The insect and spider collections of the world. USA. 220 pp. Arnoldi, L. V. 1977. Rhynchophora (in part), Mesozoic beetles. Trudy Paleontological Insti- tute 161: 142-176, 196-198. Burke H. R. 1976. Bionomics of the an- thonomine weevils. Annual Review of En- tomology 21: 283-303. Clark, W. E., D. R. Whitehead and R. E. Warner. 1977. Classification of the weevil subfamily Tychiinae, with a new genus and species, new combinations, and new synonymy in Lignyodini (Coleoptera:Curculionidae). The Coleopterists Bulletin 31: 1-18. Coleman, P. G. 1980. Plate tectonics background to biogeographic development in the south- west Pacific over the last 100 million years. Palaeogeography, Palaeoclimatology, Palaeo- ecology 31: 105-121. Cronquist, a. 1981. An integrated system of classification of flowering plants. Columbia University Press. XVIII+ 1262 pp. Frhs, E. M., W. G. Chaloner and P. R. Crane. 1987. Introduction to Angiosperms. In: The Origins of Angiosperms and their Biological Consequences, Ed. E. M. Friis, W. G. Chaloner and P. R. Crane, pp. 1-15. Cambridge Univer- sity Press, NY. 358 pp. GuiLLAUMiN, A. 1948. Flore analytique et synop- tique de la Nouvelle-Caledonie: Phanero- games. Flore de la France d'Outre-Mer, Paris. 369.pp. Kissinger, D. G. 1964. Curculionidae of America north of Mexico: a key to the genera. V+ 143 pp. S. Lancaster, MA. Kusc:hel, G. 1959. Un curculionido del Cretaceo Superior, primer insecto fosil de Chile. Inves- tigaciones Zoologicas Chilenas 5: 49-54. Kuschel, G. 1982. Apionidae and Curculionidae (Coleoptera) from the Poor Knights Islands, New Zealand. Journal of the Royal Society of New Zealand 12: 273-282. Kuschel, G. 1983. Past and present of the relict family Nemonychidae. Geojournal 7.6: 499- 504. Kuschel, G. 1987. The subfamily Molytinae (Coleoptera:Curculionidae): General notes and descriptions of new taxa from New Zea- land and Chile. New Zealand Entomologist 9: 11-29. Kuschel, G. (in press). Palophaginae, a new sub- family for leaf-beedes feeding as adult and larva on araucarian pollen in Australia (Col- eoptera:Megalopodidae). Invertebrate Tax- onomy. Lacordaire, T. 1963. Genera des Coleopteres 6: 1-637. Paris. McKay, I.J. and R.J. Rayner, 1986. Cretaceous fossil insects from Orapa, Botswana. Journal of the Entomological Society of South Africa 49: 7-17. May, B. M. 1987. Immature stages of Cur- culionoidea (Coleoptera): Rearing records 1964-1986. New Zealand Entomologist 9: 44-56. Mildenhall, D. C. 1980. New Zealand Late Cre- taceous and Cenozoic plant biogeography: a contribution. Palaeogeography, Palaeoclima- tology, Palaeoecology 31: 197-233. Stevens, G. R. 1981. New Zealand adrifi. A. H. & A. W. Reed Ltd., Auckland. XXI 1 -1- 442 pp. Vink, W. 1988. Taxonomy in Winteraceae. Faxon 37: 691-698. Voss, E. 1935. Funf Russler aus den Tribus Prionomerini und Anthonomini von Java (Col. Cure.) Entomologische Blatter 31: 228-233. Voss, E. 1936. Uber die Tribus Scolopterini sowie einige verwandte Gattungen und Gruppen von Neu-Seeland. Arbeiten iiber mor- phologische und taxonomische Entomologie aus Berlin-Dahlem 3: 60-62, 110-121. Walker, J. W., G. J. Brenner and R. G. Walker. 1983. Winteraceous pollen in the Lower Cretaceous of Israel: early evidence of a magnolialean angiosperm family. Science 220: 1273-1275. Zherichin, V. V. 1977. Rhynchophora (in part), Mesozoic beetles. Trudy Paleontological Insti- tute 161: 176-182, 198. NOTROPIS AMMOFHILUS, A NeW CyPRINID FlSH FROM SOUTH- EASTERN Unh ED States Royal D. Sin tkus Museum of Natural History, Tulane University Belle Chasse, Louisiana 70037 Herbert T. Boschung Department of Biology, The University of Alabama Tuscaloosa, Alabama 35487-0344 Abstract Notropis ammophilus, a new shiner (Cyp- rinidae), is described from southeastern United States. It is related to Notropis longirostris. N. sabinae, and another undescribed form. The dis- tribution of A', ammophilus, the orangefin shiner, is primarily below the Fall Line in the Mobile Bay basin in Alabama and Mississippi, in Yellow Creek, a tributary to the Tennessee River in northeastern Mississippi, in the Hatchie River sys- tem in southwestern Tennessee and northern Mis- sissippi, and in the Skuna River drainage of the Yazoo River system in northern Mississippi. The orangefin shiner is more similar to N. lon- girostris in nuptial coloration and tuberculation but more similar to A^. sabinae in general body shape. These three species plus an additional unde- scribed form are considered to comprise a species complex by themselves. Introduction Existence of the orangefin shiner, herein described, has been known to ichthyologists for many years. One of the earliest references to this species ap- peared in Cook (1959). Most authors, in- cluding Smith- Vaniz (1968), assumed this form to be an allopatric replacement of Notropis longirostris in the Mobile basin. We now know that this form occurs in the headwaters of the Hatchie River sys- tem in Mississippi and Tennessee, in tributaries to the south bend of the Ten- nessee River system, specifically Yellow Creek in northeastern Mississippi, and in the upper Skuna River of the Yazoo River system in northern Mississippi. The primary purpose of this paper is to for- mally describe the species and to sum- marize our field observations. Notropis longirostris (Hay 1881) was placed in the genus Alburnops at the time of its original description and sub- sequently in Hybopsis, as was Notropis sabiriae ]o\d2in and Gilbert 1886, by Jor- dan, Evermann and Clark (1930). The type species of Alburnops is Alburnops blen- nius Girard (1856: 194) by subsequent designation of Jordan and Gilbert in 1877 (Gilbert, 1978). The type species of Hybopsis is Hybopsis amblops (Rafinesque 1820: 51) by subsequent designation of Jordan and Gilbert (1877). We do not be- lieve that Notropis longirostris is closely re- lated to N. blennius. The senior author does not believe that N. longirostris is closely related to any of the typical mem- bers of the genus Hybopsis (amblops, win- chelli, amnis, lineapunctata, rubrifrons, hyp- sinotus) contrary to Mayden's (1989) re- cent placement of these forms in his grouping of the genus Hybopsis. At pres- ent we place the new form, together with Notropis longirostris, N. sabinae and another undescribed species, in a species complex, longirostris species group, by themselves. An analysis of the longirostris species group will appear in a forthcoming paper. The four species of this group are 49 50 Tulane Studies in Zoology and History Vol. 27 characterized by: A pronounced elongate swelling of the rami of the lower jaw that bears tubercles in breeding males; distal portion of anterior rays and interradial membranes of anal fin of breeding females opaque and more or less (de- pending on species) flexed dorsally; sub- terete body shape, dorsal contour arched more than ventral; long, decurved snout; large, inferior, horizontal mouth; typi- cally 8 dorsal fin rays, 7 anal rays, 19 caudal rays, 8 — 8 pelvic rays, and 12 caudal peduncle scale rows; silvery peritoneum; and psammophilous life style. The following abbreviations are used: SL = Standard length, CU - Cornell University; GSA = Geological Survey of Alabama; MSU = Mississippi State Uni- versity; TU = Tulane University, Museum of Natural History; UAIC = University of Alabama Ichthyology Col- lection; UF = University of Florida, Florida State Museum; UMMZ = Uni- versity of Michigan, Museum of Zoology; USNM = National Museum of Natural History; and UT = University of Ten- nessee. We follow Robins, et al. (1980) with regards to the gender of the genus Lepomis and generic groupings of the family Cyprinidae. Notropis ammophilus, a new species Orangefin Shiner Figs. 1 — 4 Notropis sp. Cook, 1959 (reference to undescribed species in upper Tombigbee and eastward). Smith- Vaniz, 1968 (key, allopatric with N. longirostns, Mobile basin endemic). Boschung, 1973, 1984, 1987, 1989 (distribution and abundance in upper Tombigbee). Jenkins, 1976 (unde- scribed species, distribution). Gilbert, 1980 (distribution). Gilbert and Burgess, 1980 (distribution). Heins et al., 1980 (re- productive biology and habitat). Tim- mons, 1982 (impoundment of upper Tombigbee). Ramsey, 1984 (distribution and conservation status). Pierson and Schultz, 1984 (distribution and relative abundance in Bull Mountain Creek). Wiley and Mayden, 1985 (distribution). Swift et al., 1986 (distribution). Pierson et al., 1986 (distribution in Buttahatchee River). Mettee et al., 1987 (distribution and abundance in lower Tombigbee). Mettee et al., 1989 (distribution in Black Warrior River system). Pierson et al., 1989 (distribution in Cahaba River sys- tem). Notropis ammophilus. Hubbard, 1987: {nomen nudum). Material. — The type material consists of 3289 specimens measuring from 17.7 to 48.5 mm in standard length, collected from Chilatchee Creek, a tributary to the Alabama River. Other material examined consists of 41,081 specimens from 1061 collections throughout the range of the species (Fig. 1). Holotype.— TU 151630, an adult male 48.5 mm standard length, from Chilatchee Creek, trib. Alabama River, 0.3 mi. N Alberta, State Hwy. 5 at Dal- las-Wilcox county line, Alabama (T15N, R7E, Sec. 30), on 9 April 1988, collected by Royal D. Suttkus. Paratypes. — The 124 specimens (TU 151631) collected with the holotype and 14 other series collected at the same site on the following dates: UAIC 265.04 (74), August 30, 1956; UAIC 2362.13 (116), October 11, 1966; TU 60871 (1147), December 18, 1969; UAIC 5899.04 (172), July 23, 1980; TU 121386 (129), May 15, 1981; TU 125388 (292), May 15, 1982; GSA 6865.07 (84), October 19, 1983; TU 140934, May 24, 1985 distributed as fol- lows: TU 140934 (207), CU 71712 (20), UF 78362 (20), UMMZ 214869 (20), UT 44.4504 (20) and USNM 301164 (20); No. 2 New Notropis from Southeastern U.S. 51 Fig. 1. Notropis ammophilus. Lateral view of a paratype, UAIC 5899.04, male, 44 mm in SL, collected by M. F. Mettee and P. E. O'Neil on July 23, 1980. Photo by P. E. O'Neil. TU 145519 (117), May 20, 1986; UT 44.4341 (19), May 16, 1988; TU 153083 (98), August 2, 1988; TU 153185 (296), August 3, 1988; TU 154399 (25), Feb- ruary 3, 1989 and TU 154734, May 9, 1989 distributed as follows: TU 154734 (19), UMMZ 215189 (100). Additional material examined for counts, measurements, pigmentation and/or used for distribution map, total- ing 41,081 specimens, in 1,061 collec- tions, are as follows (in parentheses, number of collections followed by number of specimens): Lower Tombigbee River drainage. ALABAMA. Choctaw County (194 : 16,399), Clarke County (45 : 1000), Marengo County (42 : 4,837), Sumter County (31 : 1,613), Washington County (1 : 99). MISSISSIPPI. Kemper County (3 : 100), Lauderdale County (2 : 94). Upper Tombigbee River drainage (pre- Tennessee-Tombigbee impoundment). ALABAMA. Fayette County (2 : 14), Franklin County (2 : 34), Greene County (10 : 714), Lamar County (37 : 368), Marion County, (1 1 : 208), Pickens County (16 : 267), Sumter County (11: 124), Tuscaloosa County (6 : 30). MIS- SISSIPPI. Chickasaw County (6 : 1,079), Clay County (11 : 579), Itawamba County (32 : 186), Lee County (7 : 754), Lowndes County (81 : 518), Monroe County (39 : 557), Noxubee County (4 : 41), Oktibbeha County (8 : 224), Pon- totoc County (1 : 309), Prentiss County (6 : 161), Tishomingo County (6 : 28), Union County (2 : 2), Webster County (1 : 110), Winston County (5 : 20). Black Warrior River drainage. ALABAMA. Greene County (4 : 155), Hale County (15 : 560), Tuscaloosa County (28 : 150), Walker County (1 : 1). Alabama River drainage. ALABAMA. Autauga County (12 : 174), Butler County, (1 : 3), Chilton County (10 : 337), Clarke County (8 : 368), Dallas County (21 : 884), Dallas-Lowndes counties (1 : 63), Elmore County (1 : 1), Lowndes County (2 : 13), Marengo 52 Tulane Studies in Zoology and History Vol. 27 1000000 10 20 30 40 „,|j5 10 0 10 20 30 10 50 '"""ictcrs 35° -34° 33° 32° 31 88° 87° 86° 85° Fig. 2. Distribution by record stations oi Notropis ammuphilus (solid dots), type locality (star in circle) and Notropis longirostris in Mobile Bay drainage basin only (solid triangles). County (3 : 50), Marengo-Dallas coun- ties (1 : 147), Monroe County (33 : 348), Perry County (4 : 41), Wilcox County (137 : 2,183). Cahaba River drainage. ALABAMA. Bibb County (21 : 476), Dallas County (15 : 302), Perry County (29 : 681). Coosa River drainage. ALABAMA. El- more County (4 : 64). Tallapoosa River drainage. ALA- BAMA. Bullock County (2 : 65), Elmore County (5 : 43), Lee County (5 : 43), Macon County (21 : 1,662), Montgom- ery County (3 : 18), Tallapoosa County (1 : 3). Tennessee River drainage. MISSIS- SIPPL Alcorn County (1 : 12), Tishomingo County (15 : 157). No. 2 New Notropis from Southeastern U.S. 53 Fig. 3. Notropis ammophihus. Lateral view of a specimen from UMMZ 1 1 1 160, 41 mm in SL. Photo by W. L. Brudon. Hatchie River drainage. Ml^SlSSl^Vl. Yazoo River drainage. MISSISSIPPI. Alcorn County (4 : 21), Prentiss County Calhoun County (2 : 109), Pontotoc (2 : 7), Tippah County (14 : 967), Union County (3 : 55). County (2 : 16), TENNESSEE. Harde- The locality data for A^. longirostris man County (6 : 121), McNairy County and N. sabinae that were used for counts, (2 : 3). measurements and description of pig- Fig. 4. Notropis ammophilus. Anterolateral view of head of specimen in Fig. 3. Photo by W. L. Brudon. 54 Tulane Studies in Zoology and History Vol. 27 mentation are given in the text and in Tables 2 and 3. Collection sites for ele- ven lots of N. longirostris from the Mobile basin as shown on Fig 2. There are six lots from three sites on Little River in Baldwin, Escambia and Monroe counties, Alabama: TU 32554 (1), TU 44400 (6), TU 44414 (14), TU 99939 (15), TU 153958 (7), and UAIC 6528.09 (75). There is one lot from Majors Creek, Baldwin County, Alabama: UAIC 426.02 (6). In addition to these lots from the lower part of the Mobile basin, there are four lots from two sites on the upper Etowah River in Cherokee County, Georgia: UT 44.1940 (8) and UT 44.2157 (6); UT 44.2118 (3) and UT 44.2233 (4). Diagnosis. — A species of Notropis with 4 — 4 pharyngeal teeth and 7 anal rays. Dorsal fm origin directly over pel- vic origin; both dorsal and pelvic fms equal distance between tip of snout and base of caudal fm. Anal fin length of males greater than that of A^. longirostris and N. sabinae. Differs in body propor- tions from other members of the species complex as seen in Tables 1 and 4. Nup- tial males have bright orange fins, orange along lips and on iris (Figure 1). Description. — In addition to the characters given in the diagnosis, counts and measurements are given to com- pare Notropis ammophilus with A^. longiros- tris and A'^. sabinae (Tables 1-4). The pharyneal teeth are moderately hooked, the edges of the well developed grind- ing surfaces are entire or weakly cre- nate, and, m 70 specimens examined, the formula is 4 — 4. The anal fm ray count is typically 7 (in 96 percent of sample), occasionally 6 (1 percent) or 8 (3 percent), where N = 100. The dorsal fm ray count is almost always 8 (98 per- cent of sample), rarely 9 (2 percent), where N = 100. Pelvic fm rays typically number 8 — 8 (92 percent of sample), oc- casionally 7 — 8 (2 percent), 8 — 7 (3 per- cent) or 8 — 9 (3 percent), where N = 100. The principal caudal rays number 19 (94 percent of sample), 18 ( 1 per- cent), and 20 (5 percent), where N = 100. Caudal peduncle scale rows, num- bering 12 in the entire sample of 100 specimens, consist of 5 rows each above and below the lateral line. Other scale counts are given in Table 2. The Webe- rian apparatus, consisting of 4 verteb- rae, and the single urostylar vertebra were included in the total vertebral count. Notropis ammophilus usually has 34 or 35 vertebrae (Table 3). Measurements of the holotype are detailed in Table 1 and its fin ray and scale counts are as follows: D 8, A 7, P 14, V 8, C 19, lateral line scales 34, caudal peduncle scales 12, body cir- cumferential scales 24, and predorsal diagonal scale rows 14. The subterete body form, with a de- cidedly arched dorsal profile and only slightly curved venter, is well adapted to the benthic habitat. The body is mod- erately robust anteriorly but tapers rather abruptly posterior to the dorsal fm. The mouth is inferior and nearly horizontal to slightly angled. The dorsal fm is moderately elevated and its an- terior rays are usually longer than the posterior ones of the depressed fm; the posterior margin is nearly straight, or it may describe a shallow sigmoid curve. The anal fm is less elevated than the dorsal, particularly that of the females, and its posterior margin also describes a sigmoid curve. The anterior rays of the anal fm, even the flexed rays of females, are usually longer than the posterior ones of the depressed fm. Other fms are typical in shape and size for small shiners. Usually the pectorals are more expansive in the males than in females. Only males have well developed No. 2 New Notropis from Soulheastern U.S. 55 c be « Di < be bC -a. be c c re c/5 -a c h UJ rs -1 CO < u. 5: re re be 4J U. o be re re be Q. C X ■oo OTP-^OltO— 'TftOO o o in in — j^ to 00 c^i CM C7) in ininmcoTfcocM— 'C^i — c^j — — 'ocootooo — — 'C^l^^ — — — Tf t^ Tf • 00 CM X I^ ^_^ ^ m C^l ^ o CO t^ in in a~; o tT 00 CM CO 0-. OJ -^ CO — ' ^- OJ C^J — c^i 00 i^ 00 — ' in !>• r^ m t-~ X CO o to .cjicofxcox-^incooiCMOto— "oo — CMXCMTf c^jxcoCTir^toxco-^Oco-^cot^xtoi^ajtomt^co in •* in o-i Tj" CO — — ' c^j — c^i ^^ — — • — ■ c^i — • ^^ in ai in Tf I I 06 {r in — to CO — o X o". inco-^cocM — CM- CM CO c^j — o oi c^j X CM X to "^ X C^J CO O". in c^i -rr CO • a-i CM in CO c; X c tt CO o t^ c^i -r o to t^ '— — 'XXtOXC^I — c^i or~rrTf-xi---roc^i— ' CO CO CM i^ r^ in i^ X in n" c^i — — — ' — c^i c^i — X -V in CM "iOiCTieoTfco— r^r^OiO-^-^OTTX— 'into-^tot^ in o — 0"i — CO in i^ CO o r^ i^ to CTi (jv to CT) — to to i^ ■^ ■^ in in c^4 in CO c^x ^^ c^i ^- c^j — * ^^ c^-i • — < C'i • — ^ — . X in to t^ cr. to X in in o~. CM i^ — -^ c^j CTi in — cocr. c^i-rtox-^ — i^xt^TTCT". Oincor^xxin mmcMincocM-^c^i — c^i— ' — O". a-r^oiCM — CM— '-^ I I I I I I I I I I I I I I I I I I I I I T}< — t-^ — i^ o 1^5 cr. ^, cr, r^ lo in X to X CM in c^. o CTi CT5 ^^ r^ o — tt to c^i O". to to in X X to X o m -r to CO -* in c^i in CO c^i — c^i c^i — — ■ c^i — c^i — — ^. O-^ t^ to CT-. CO O CM 0"j O' tt in m CM in a--ooc^iCMcot^OMOtoCTiCMCMOr^ocM ■* CM ■* CO — r^ to in en CTj to CT. c^j a-, to X in CO C^J — ' S^i — C^l — — • C^l — C^J — — ^. O) oi in to CO to X o i^ CO to CTi m — CO o c^i to CO -^ -r — X t^ in TT in " ' ' ' ' CO to — o a~. o i^ CO — ' X o". m in CO m CO c^j — ' c^i — > c^i — — ' c~. c t^ iDCM'^COCM — C^l — C^l — — t^ C^l X c. ^: m a-, in ^ 'T in in c^i m 3~. m in c. to in c^i c- t^ c^j in CO o to to in x c. C^J — C^l — C^J — — X in i^ — ' — Tf CO to X o X m to -^ CM — C-I — — X to to ~ — in to to X o o in to Tf CM — C^J — — •r O — re V5 -;3 -5 JS V -"C "C "C "C "O ^ re re v^ 5 C ^ oi < c£ ."^ _ W flj ■ .< — ^ . _, .« ^« L« ^ f^ - , "-^ u- re 2 56 Tulane Studies in Zoology and History Vol. 27 -a c •So c 3 O O 03 (J CA) T3 C CS >-. ra c "re u O u u ,o c o 3 u C u 3 cr u u u -J h u c CD CD CO CO CO CO CO (M CO a; u y^ 'SI- a. C/5 o o r- :^ in o 00 i> 00 00 r-; r^ Oi •* CO CO CO CO CO in CO CO CO CO CO CO CO CO o o o o o o o o CO CO CO a-. in in o (M « — CM CO in in CO CD — CM (D CO CD CD -^ Tj- — — rt CT) o^ o CT) 00 in CO CD CM CM CM CD CO CM in Oi CD O t^ CM CM — in a~; CM 00 00 r^ in CM — ' ^ — en CM in (D .>; oi ^ Di a; en CM m 00 CD CM CM CD CO r^ r^ -^ CM -^ Tf O CM ■5P CD -^ ^ Tf Tf Tf Tt" ^ o o o o o O CO CO CO CT~; o in O in CO CD CM 00 = Si < H X >. ^ > S 5 > n c o be C T3 OJ >- CLh crt in _u CM "w CJ C/1 „_ .2 CM c a; ^ a; C4-1 e CO CM 3 (J u u >~ CM -0 CM 0 CQ CD CM CM o CM cn ■a c :« p 0 CO CO CO CD 0 i> 00 in 00 in ■^ CO CO 00 CO ^ CM 00000 0 CO CO CO O) 0 in 0 in r^ CM Tt< CM in CO CM i> in i> CD -^ — — O Ol ■* CT) O CM CO CO —I CM CO ■* O CM — ■* 00 in i> — ' -^ CM Tf in CD — i>- a-- « CM ■* r^ -^ CM ■^ — -H CO > 5 -a, fa S CT3 OJ ? ii o ^ t^i N U ™ 03 CD CO CM CD O CM yD t^ CO 0 r^ ^ CM CD 0 CO — 0 0 CM — CO ^ CM CM CM CM CM CM 0 CM CO CM 00000 0 CO CO CO O) 0 in 0 in CM CM o 00 in in o o CM — > Tt __ „ S- fU > > ^ OS OS ? 'S Sr 0 • SiS T- •■■ -c n 0 U4 e c/D No. 2 Neiv Not topis from Sonlheaslern U.S. 57 TABLE 3. Number of Vertebrae in NotropLs amtnnphilus, N. longirostris and sabinae. Species and system Ninnber ol Ver tebrae N 33 34 35 36 37 X N. (imvH)philus Alabama River 1 51 54 3 109 34.54 Tennessee River 6 15 2 23 34.83 Hatchie River 1 17 17 1 36 34.50 2 74 86 6 168 34.57 N. longirostris Pascagoula River 6 47 21 74 36.20 A^. sabinae Sabine River 5 71 40 1 2 119 34.36 nuptial tubercles. Some females have a few minute tubercles on the top of the head but the remainder of the body is devoid of tubercles. The largest tuber- cles are scattered over the dorsal surface of the head where they are double in size from the internarial area to the oc- ciput. There is a single row of large tubercles, more or less evenly arranged, over the supraorbital rim. Minute tuber- cles cover the snout and are scattered between the larger ones on top of the head. Some males have minute tuber- cles extending to the lachrymal area and to the upper parts of the opercular re- gion. The head tuberculation, as we have described it above, agrees reasona- bly well with the description of Heins et al. (1980) for specimens from Uphapee Creek. Multiple rows of small tubercles are present on the pectoral fin rays, and their number diminish toward the post- erior rays. Nuptial males have tubercles on the first seven to nine pectoral rays. Some nuptial males taken in May have tubercles on the leading edge of the dorsal fin, the anal fin, and on the upper and lower caudal fin margins. In addition, a few males have tubercles on the predorsal and anterolateral scales. The typical intermediate tubercular pat- tern shows tuberculate ridges on the lower jaw, a few tubercles on the leading edge of the dorsal fin, the upper and lower margin of the caudal fin, and the lower two or three branched caudal rays. The lateral line system is usually com- plete. Rarely is the last scale of the lat- eral line unpored. The lateral line is nearly straight, with only a slight dip an- teriorly; above the pelvic fin insertion, it gradually angles dorsally and then levels at the midlateral area of the caudal peduncle. COLORATIC3N. — The orangefin shin- er, when not in nuptial attire, is a pale, very lightly pigmented tish. The preva- lent pigmentation of preserved speci- mens is the dark spots above and below each lateral line pore. The spots are prominent anteriorly but often blend with the lateral stripe on the caudal peduncle. The short lateral stripe be- gins on the lateral area above the inser- tion of the anal fin, and it extends post- eriorly to the base of the caudal fin. The pigmentation of the stripe is somewhat imbedded and is just ventral to the lat- eral line pigmentation. In most speci- mens, the lateral line pigment spots are superimposed on the upper margin of the lateral stripe. Some specimens have an additional row of paired spots on the scale row above the lateral line, thus giv- ing the appearance of a double lateral line on the anterior part of the body. There is a thin submarginal line of pig- 58 Tulane Studies in Zoology and History Vol. 27 ■So s s c o (/) 'u a B o U CO < m j> tr> O) TT — . ai f^ o ^ O (>4 CD j> lO 2- o tc 00 CM o in " CM — £i- f-H i~-i > — ^ — — ^-.^ ^.^ "§■? ly to moderately de- loped on end of snout d lachrymal area. w oi r- Tt- tri •^i CO in i^ u is (M r^ 00 c£> Oi O "^ ^^ Cf^ i> in O flj c^ c/5 ' j_ oj 00 in 1 00 op o CO 1 1 '* Listers on ( achrymal t/3 c 73. 3 — CM tn Tf CT) O) cc 00 o CM in 'T "a; OJ .S 2 'e E g S .£f u "15 ° 2 o ,0 o u O 4; Longest Moderate OJ u fa 2 E 4J O C OJ be c G 3 u s :^ a. T3 C 03 en 13 ■§- 2 3 C U re E 3 :2 c > . n u u C bC '^ -n u C O. ^ ■=. re _ E g t ^ -0 CO ai CM 0 CD ^ 0 cn <— 1 CM ^ 0 CO ■^ CO t^ 00 0 r^ CT5 CD CO en "O Oi Si in CO 0 00 " OJ .E re 0 Tf< ^ CO 0 r^ CM TT 0 CD '^ u^ >-. 0! "^ — 00 00 CD 00 CM ^ 0 0 4-^ re T3 00 0 i^ Tt" 1 06 r^ • c 0 en a ;^ 3 O) CO CO CM i> i> m :^ 00 in bD CM 0 .0 E 0 E CM c _0 V > c 03 0 k. ^0 0 t(5 4^ 0 en in In tn en Q "re 3 ■^ bc-C "u OJ 5 _aj o; 1) W w re • « u 0 u "re E en ^ ;- U 4; t^ re 0 4^ be c < i_ 0 re ^ u 0 0 0 0 XJ 0 3 re u c75 re c75 OJ ra ra J zz CO c/5 c75 en 3 3 CM CO t^ CM 0 CD CJl CM CM 0 CM 0 0 — • I> CD in O) Oi CD CT) CM CT> in C en c "" Si C ^M — ■^ — ' "^-^ — ' Si ^H ^^ en •^ ' ^-^ "O c t^ CO ;d Ol CO CD rj a^ s QJ en -H 00 t^ in Tt< CT) 0 t^ CO re in X OJ W ^ CM -. ^^ Ol ? r~ cn CM CM (— <— ' L*— B o; ^ CJ re Si 06 -^ C^ ^ 1 CM 06 in r^ rli ^ CO 0 flj t« U en 0 CD ;d ■^ 00 O) CD 00 0 00 QJ rt- Cl -C 3 X! >^ re -a re "re -a 3 "re -a 3 re (J 0 -H CM — 4-> re CM be 0 en re _o be c re 0 u > 0 •e s 'e re u OJ V 4J OJ '-5 «J u OJ re 4J 0 T3 -a en re re re u £ u OJ re re u re i- en OJ be C 5 3 en re _>, 1) OJ re OJ c re re (-> u 0; -0 0 -a 0 0 "2 -a 0 3" JO 3 0 "re 3 re Z Z ZQS SSS c sss ^0 c o c o -o o c o u c o o re E c*- 1; i § be"? •r u ^ a. 0 cu re OJ x; C4-. o re re u w jr -c o Q J Q ^ XI en V c en re JJ "re IS u O I- u c "1 r c re 3 O w c >- en V e>< e^ O O J= JZ re ^ u^ 3 ii c -a 't: r.a. °a ^ _ -S — ^ be Sd be S — c c c J; 2 OJ OJ QJ o c _1 J J Q <; c d 03 u re > C 11 < (Ih 6 £ be o "3 o be C -5 V V u ca XI 3 u - E o c _0 3 X O T3 re u c o V c o Z u c o Z c u re -^ "O'c re -^ en (J re nj i« t. 3; ^2 S T3 u -" U (U u ^ IJ — V 'X ii 1) Q. a. o- c E ° £ Z "*» c c o Z T u t< ^ii 3 2 ^'o 0^ "re u u (J CJ J= en en 0 en re U OJ u Oh ca XI X re OJ x: OJ Q. E re >, w en be KM c re >-^ c re No. 2 New Notropis from Southeastern U.S. 59 ment on the scales of the lateral line and at least on two rows above the lateral line. Most of the dorsal and dorsolateral scales also have the thin submarginal line of pigment but other melanophores are clustered just anterior to the sub- marginal "string" and tend to obscure the fine submarginal line, particularly on the middorsal scales. Ventral to the lateral line, the melanophores of the scales' submarginal line become more disconnected and disappear completely on the belly. Most scales, except for a few in the middorsal area, have a non- pigmented anterior central area, thus emphasizing the submarginal pigmenta- tion. With magnification, a thin band (lat- eral stripe) of deeply imbedded melanophores is discernable from the central area of the scale row above the anterior part of the lateral line post- eriorly in a straight line to near the base of the caudal fin. Because of the an- terior dip of the lateral line, the thin band is separated from the lateral line by a distance approximately equal to the depth of a nearby scale, but it is just above the lateral line on the caudal peduncle. There is a thin middorsal stripe before and after the dorsal fin. There are some melanophores along the base of the anal fin that continue posteriorly as a faint midventral caudal peduncle stripe. The dark pigmentation on top of the head diminishes in intensity from the occiput to the internarial area and snout. There is a moderate amount of pigmentation on the upper part of the opercle, and on the preorbital area where it forms a thin band bordering the suborbital rim. The lips, lower cheek, opercle, gular area, breast, and belly are devoid of pigmentation. The dorsal fin rays, some anal rays of some individuals, and caudal fin rays are more or less margined with melano- phores. The pectoral and pelvic fins are essentially immaculate. The basal part of the caudal rays are margined with melanophores, but distally there are few or none. Melanophores at the base of the caudal rays tend to form a dusky spot which is not readily visible with the unaided eye. Breeding coloration is the basis for the suggested common name, orangefin shiner. Males in "high" nuptial condi- tion have bright orange fins. In some males the fins are reddish orange. The leading edges of the paired fins, the dorsal and anal fins, and the upper and lower margins of the caudal fin are milky white. The snout is pinkish orange and the upper and lower lips are bright orange. The iris is orange, par- ticularly its dorsal part. Females, pallid compared with males, lack coloration on the fins, but have weak yellow-orange color on the snout, lips and upper part of the iris. Brightly colored male speci- mens were observed in early April through early August. Some male speci- mens collected in late May had brightly colored fins but were only slightly tuber- culate, whereas, others of the same color intensity had fully developed tubercles. Etymology. — This fish, as stated above, has been known for many years and ammophilus was a manuscript name applied in 1939 by the late Carl L. Hubbs. The name is derived from the Greek ammos, sand, and philo, to love, in reference to the preferred habitat Comparisons. — Characteristics com- mon to Notropis ammophilus, N. Umgirostris and N. sabinae are given in the diagnosis of the species complex. The proportions in Table 4 reflect the shorter body of N. ammophilus and N. sairinae versus the elon- gate body of N. Umgirostm. The general body shape of A^. ammophilus is similar to that of N. sabinae. Notropis Unigirostris is less 60 Tulane Studies in Zoology and History Vol. 27 robust anteriorly and its dorsal profile tap- of clear water and a substrate of clean ers less abruptly than that of the other two sand. It does occupy substrates of silt species. Notropis ammophilus and A^. langiros- covered sand, mixtures of sand and fine tris are similar in head tuberculation and gravel, or hard clay. Chilatchee Creek, breeding colors. Notropis ammophilus and N. the type locality, is typical habitat. The sabinae have the same pharyngeal tooth stream varies in width from 6 to 1 5 me- formula of 4 — 4 whereas N. longirostris is ters or greater, and depths from 0.1 to variable in number. The tooth in die 0.45 meters, depending on the season, minor row of iV. longirostris, when present, We have observed, as did Heins et al. is usuaUy very slender, fragile, and slanted (1980), that fairly extensive, gendy slop- toward the teedi of the major row. Un- ing sand and gravel bars line the mar- doubtedly this fragile tooth is somedmes gins of preferred stream habitats. The lost during removal of arch or cleaning; species will occupy much smaller however, some individuals do lack teeth in streams that are near the confluence of the minor. Pharyngeal arches of A^. knigi- a major stream. A few specimens at vari- rostris were removed from 20 specimens ous times have been collected in the from each of five drainages from the Yel- main channel of the Alabama and low River in Florida to the Amite River in upper Tombigbee rivers. Louisiana. The pharyngeal tooth formula Notropis ammophilus is a gregarious was 0,4 — 4,0 in 35 specimens and 1,4 — 4,1 species and, in its typical habitat, forms in 32 specimens. Thirty specimens had schools of hundreds of individuals. A either 0,4 — 4,1 or 1,4 — 4,0. Of the remain- school of fish will swim slowly upstream ing three specimens, one had the formula for a short distance and then more or 0,4 — 3,0, one 0,4 — 5,0 and the other 0,5 — less drift downstream while moving in 4,0. N. hmgrostris (Biloxi River sample) has and out of depressions such as troughs somewhat higher lateral line scale and pre- on the lee side of snags. Spawning was dorsal diagonal scale row counts and a observed on numerous occasions. It oc- higher vertebral count (Pascagoula River curred in the shallows and in the man- sample) than the other two species (Tables ner of Notropis longirostris as described 2 and 3); however, N. longirostris has the by Hubbs and Walker (1942); however, lowest body circumferential scale count. A in no instance did we attempt to follow more complete analysis of N. longirostris the activities of a particular male as they populations will appear in a future paper, reported. Based on observations spawn- Geographical Variation. — Not- ing extended from May through Au- ropis ammophilus does not vary signifi- gust. candy throughout its range. The Chilatchee Creek, the type locality, Hatchie River and the Yazoo River is relatively rich in species. The follow- specimens vary slightly from the ing 14 species were collected with the Alabama River specimens in scale and holotype: Campostoma oligolepis, Ericymba fin ray counts (Table 2). Based on re- buccata, Notropis bellus, N. venustus, cent observations, the nuptial coloration Pimephales notatus, Noturus nocturnus, of Hatchie and Yazoo system specimens Gambusia affinis, Lepomis cyanellus, L. is identical with that displayed by Tom- macrochirus, L. megalotis, Etheostoma sp., bigbee and Alabama system specimens. E. nigrum, E. rupestre, and E. stigmaeum. Habitat and Biology. — Notropis Twelve additional collections from the ammophilus usually inhabits small to type locality yielded: Carpiodes velifer, large streams offering the ideal habitat Erimyzon oblongus, Minytrema melanops, No. 2 New Notropis from Southeastern U.S. 61 Moxostoma erythrurum, M. poecilurum, Hybopsis winchelli, Notemigonus crysoleu- cas, Notropis stilbius, N. texanus, N. volucel- lus, Pimephales vigilax, Sem otilus atromac- ulatus, Ictalurus punctatus, Noturus gyrin- us, N. leptacanthus, Fundulus olivaceus, Ambloplites ariommus, Lepomis gulosus, L. microlophus, L. punctatus, Micropterus punctulatus, M. salmoides, Pomoxis nigro- maculatus, Etheostoma parvipinne, E. whip- plei, E. zoniferum, Percina maculata and P. nigrofasciata. Thus, we have a total of 43 species from the type locality. An analysis, of the 12,156 fish col- lected in 13 samples (excluding the 1989 samples) from Chilatchee Creek reveals some interesting information regarding abundance and frequency of occcur- rence. Notropis ammophilus numbered 2976 (24.5 percent of sample), the high- est in relative abundance. Next most abundant was Pimephales notatus (2837 specimens, 23.3 percent), another bot- tom dwelling fish. There were two com- mon mid- water species: Notropis venustus (1952 specimens, 16.1 percent) and A'^. bellus (1667 specimens, 13.7 percent). Thus, the two mid-water species (3619 specimens) represented 29.8 percent, and the two bottom species (5818 speci- mens) represented 47.8 percent of all specimens collected. These four species (9437 specimens) represented 77.6 per- cent of the total, and the remaining 39 species represented 22.4 percent. Range. — Notropis ammophilus is wide- ly distributed in the Mobile basin, essen- tially below the Fall Line (Fig. 2). Dis- junct populations occur in the Yellow Creek system, a relatively small watershed of the Tennessee drainage in northern Mississippi, in the headwaters of the Hatchie River system in northern Mississippi and southwestern Tennes- see, and in the Skuna River system of the Yazoo drainage in northern Missis- sippi. Notropis longirostris is absent from nearly the entire range ot N. am- mophilus; however, there are popula- tions of A^. longirostris in Little River and Majors Creek which are eastern tribu- taries to the Alabama River in the lower part of the Mobile basin (Fig. 2). No N. ammophilus specimens have been taken in either of these two tributaries. Perhaps the parapatry is the result of human activities, but more likely it is the result of a natural faunal interchange between the Little River and the Escam- bia River system at some time in the past when the two systems were connected. There are also populations of N. lon- girostris in the upper Etowah River in northern Georgia. These populations are well above the Fall Line and far above the uppermost populations of A'^. ammophilus in the lower Coosa River (Fig. 2). We believe the Etowah River population also to be of natural occurr- ence and presume that lateral headwa- ter tributaries of the Etowah and the Chattahoochee rivers were at one time connected. Acknowledgments Study materials and distributional records o^ Notropis ammophilus have been accumulating over the years, due to a large extent to the efforts of graduate students and colleagues at Tulane Uni- versity, University of New Orleans, Mis- sissippi State University, and the Uni- versity of Alabama, who have contrib- uted in a very significant way to the study of southeastern fishes. They are in alphabetical order: Paul K. Ander- son, James G. Armstrong, Clyde D. Bar- bour, Eugene C. Beckham, Meredith May Blackwell, Lyn Branch, Richard D. Caldwell, John H. Caruso, Robert C. Cashner, Glenn H. Clemmer, Veronica Trau Colbert, John Van Conner, Michael D. Dahlberg, Keith G. Goodfel- low, Gerald E. Gunning, Charles D. 62 Tulane Studies in Zoology and History Vol. 27 Hancock, Hector Harima, David C. Heins, Henry H. Howell, W. Mike How- ell, Julian Humphries, Maurice F. Met- tee, Rudolph J. Miller, Phillip R. Mundy, Patrick E. O'Neil, John Pagels, J. Malcolm Pierson, John S. Ramsey, Sharon E. Reilly, Kenneth Relyea, Dawn E. Remington, Steven O. Rohmann, Patrick Sorensen, Diana W. Stein, Mark Stevens, Jamie E. Thomerson, Bruce A. Thompson, Charles E. Tucker, Benja- min R. Wall. James D. Williams. We thank David Etnier for loan of specimens from the Hatchie and Etowah River systems. We are grateful to Richard L. Mayden who made availa- ble the extensive UAIC data and to Maurice F. Mettee and Patrick E. O'Neil who furnished the extensive GSA data, and to J. Malcolm Pierson for data from the Cahaba River. Frank J. Pezold (Mis- sissippi State University) made MSU data available. We extend our sincere thanks to Glenn H. Clemmer who as- sisted in many of the early collections and is responsible for most of the ac- cumulated material at MSU. He also as- sisted in making vertebral counts. Reeve M. Bailey reviewed a draft of this paper and offered helpful suggestions, also he graciously withdrew from an early con- sidered co-authorship. Moreover, we are indebted to Reeve M. Bailey for UMMZ collection data and for the use of photo- graphs (Figs. 3-4) that were prepared by William L. Brudon and to Patrick E. O'Neil for the use of his color photo- graph (Fig. 1). Literature Cited BosCHUNG, H. T. 1973. A report on the fishes of the upper Tombigbee River, Yellow and In- dian creek systems of Alabama and Mississippi, first Supplemental Report, Continuing En- vironmental Studies. Vol. 7, Appendix C. .\rmy Corps of Engineers, Mobile District, Mobile, Ala., 144 p. + appendix. BosCHUNG, H. T 1984. A study of the fishes of the upper Tombigbee River drainage system south of the Columbus Lock and Dam Tennes- see-Tombigbee Waterway: U.S. Army Corps of Engineers, Mobile District, Mobile, Ala., 194 p. + appendix. BOSCHUNG, H. T. 1987. Physical factors and the distribution and abundance of fishes in the upper Tombigbee River system of Alabama and Mississippi, with emphasis on the Tennes- see-Tombigbee Waterway, pp. 184-192 in: Matthews, W. J. and D. C. Heins (eds.). Com- munity and Evolutionary Ecology of North American Stream Fishes. University of Ok- lahoma Press, Norman, Okla., 310 p. Bc:)SCHUNG, H. T. 1989. Adas of fishes of the upper Tombigbee River drainage, Alabama- Mississippi. Southeastern Fishes Council 19: 104 p. Cook, F. A. 1959. Freshwater Fishes in Missis- sippi. Miss. Game and Fish Comm., Jackson, 239 p. Gilbert, C. R. 1978. Type catalogue of the North American cyprinicl fish genus Notropis. Bull. Florida State Mus. Biol. Sci. 23(1): 104 p. Gilbert, C. R. 1980. Notropis sabinae Jordan and Gilbert, Sabine shiner, p. 304 in: Lee, D. S., et al. Atlas of North American freshwater fishes. N.C. State Mus. of Nat. Hist., 867 p. Gilbert, C. R., and G. H. Burgess. 1980. Not- ropis lojigirostris (Hay), longnose shiner, p. 283 in: Lee, D. S., el al. Atlas of North American freshwater fishes. N.C. State Mus. of Nat. Hist., 867 p. GiRARD, C. 1856. Researches upon the cyprinoid fishes inhabiting the fresh waters of the United States, west of the Mississippi Valley, from specimens in the museum of the Smithsonian Institution. Proc. Acad. Nat. Sci. Phila. 8(5): 165-213. (also printed as a separate. 1856: 54 p.). Hay, O. P. 1881. On a collection of fishes from eastern Mississippi. Proc. U.S. Nat. Mus. (1880)3: 488-515. Heins, D. C, G. E. Gunning, and J. D. Wil- liams. 1980. Reproducdon and population structure of an undescribed species ot Notropis (Pisces: Cyprinidae) from the Mobile Bay drainage, Alabama. Copeia, 1980(4): 822-830. Hubbard, W. D. 1987. Distribution and diversity of fishes in the Noxubee River drainage, Mis- sissippi. Jour. Mississippi Acad. Sci. 32: 19-29. HUBBS, C. L., and B. W. Walker. 1942. Habitat and breeding behavior of the American cyp- rinid fish Notropis longirostris. Copeia, 194Z(2): 101-104. Jenkins, R. E. 1976. A list of undescribed fresh- water fish species of continental United States and Canada, with additions to the 1970 checklist. Copeia, 1976(3): 642-644. Jordan, D. S., B. W. Evermann, and H. W. Clark. 1930. Checklist of the fishes and fishlike vertebrates of North and Middle America north of the northern boundary of Venezuela and Colombia, Rept. U.S. Fish Comm. (1928) (Appendix X): 670 p. No. 2 New Notropis from Southeastern U.S. 63 Jordan, D. S. and C. H. Gilbert. 1877. On the genera of North American freshwater fishes. Proc. Acad. Nat. Sci. Phila. 29: 83-104. Jordan, D. S., and C. H. Gilbert. 1887. List of fishes collected in Arkansas, Indian Territory, and Texas, in September, 1884. with notes and descriptions. Proc. U.S. Nat. Miis. 9: 1-25. Lee, D. S., C. R. Gilbert, C. H. Hocutt, R. E. Jenkins, D. E. McAllister, and J. R. Stauffer, Jr. 1980. Atlas of North American freshwater fishes. N.C. State Mus. of Nat. Hist., Raleigh, x + 867 p. Mayden, R. L. 1989. Phylogenetic studies of North American minnows, with emphasis on the genus Cyprinella (Teleostei: Cyp- riniformes). Univ. Kansas Mus. Nat. Hist., Misc. Publ. 80: 189 p. Mettee, M. F., p. E. O'Neil, J. J. Pierson, and R. D. SUTTKUS. 1989. Fishes of the Black War- rior River system in Alabama. Alabama Geological Survey, Bull. 133: 201 p. Mettee, M. F., P. E. O'Neil, R. D. Suttkus, and J. M. Pierson. 1987. Fishes of the lower Tombigbee River system in Alabama and Mis- sissippi. Alabama Geological Survey, Bull. 107: 186 p. Pierson, J. M., W. M. Howell, R. A. Stiles, M. F. Mettee, P. E. O'Neil, R. D. Suttkus and J. S. Ramsey. 1989. Fishes of the Cahaba River system in Alabama. Alabama Geological Sur- vey, Bull. 134: 183 p. Pierson, J. M., and C. A. Schultz. 1984. A re- port on the fishes of Bull Mountain Creek, with comments on the status of rare species. Proc. Southeastern Fishes Council 4(3): 1-3. Pierson, J. M., C. A. Schultz, and H. T. Bos- CHUNG. 1986. Fishes of the Buttahatchee River system of Alabama and Mississippi. Proc. Southeastern Fishes Council 4(4): 11-13. Rafinesque, C. S. 1820. Ichthyologia Ohiensis, or natural history of the fishes inhabiting the River Ohio and its tributary streams, preceded by a physical description of the Ohio and its branches. W. G. Hunt, Lexington, Ky., 90 p. Ramsey, J. S. 1985. Freshwater Fishes, in: R. H. Mount, ed. Vertebrate Wildlife of Alabama, Auburn University, 44 p. Robins, C. R., R. M. Bailey, C. E. Bond, J. R. Brooker, E. a. Lac:hner, R. N. Lea, and W. B. Scott. 1980. A list of common and sci- entific names of fishes from the United States and Canada: American Fisheries Society Spe- cial Publication 12, 174 p. Smith-Vaniz, W. F. 1968. Freshwater fishes of Alabama. Agri. Exp. Stat., Auburn Univ., vii + 211 p. Swift, C, C. R. Gilbert, S. A. Bortone, G. H. Burgess, and R. W. Yerger. 1986. Chap. 7, pp. 213-266. in: The Zoogeography of North American Freshwater Fishes. C. H. Hocutt, and E. O. Wiley (eds.). John Wiley and Sons, xiii + 866 p. Timmons, T. J. 1982. Initial changes in fish species composition in two new lakes of the Tennessee-Tombigbee Waterway, Alabama- Mississippi. Proc. Southeastern Fishes Council 4(1): 1-4. Wiley, E. O., and R. L. Mayden. 1985. Species and speciation in phylogenetic systematics, with examples from the North American fish fauna. Ann. Missouri Bot. Card. 72: 596-635. r . : illln i 3 2044 f m 093 361 iiHiini 03 ll m